Exploring Molecular-Biomembrane Interactions ... - ACS Publications

May 25, 2018 - Tzong-Hsien Lee, Daniel J. Hirst, Ketav Kulkarni, Mark P. Del Borgo, and Marie-Isabel Aguilar*. Department of Biochemistry and Molecula...
1 downloads 0 Views 18MB Size
Review Cite This: Chem. Rev. 2018, 118, 5392−5487

pubs.acs.org/CR

Exploring Molecular-Biomembrane Interactions with Surface Plasmon Resonance and Dual Polarization Interferometry Technology: Expanding the Spotlight onto Biomembrane Structure Tzong-Hsien Lee, Daniel J. Hirst, Ketav Kulkarni, Mark P. Del Borgo, and Marie-Isabel Aguilar*

Downloaded via UNIV OF CALIFORNIA SANTA BARBARA on June 24, 2018 at 15:51:47 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

Department of Biochemistry and Molecular Biology and Biomedicine Discovery Institute, Monash University, Clayton, VIC 3800, Australia ABSTRACT: The molecular analysis of biomolecular-membrane interactions is central to understanding most cellular systems but has emerged as a complex technical challenge given the complexities of membrane structure and composition across all living cells. We present a review of the application of surface plasmon resonance and dual polarization interferometry-based biosensors to the study of biomembrane-based systems using both planar mono- or bilayers or liposomes. We first describe the optical principals and instrumentation of surface plasmon resonance, including both linear and extraordinary transmission modes and dual polarization interferometry. We then describe the wide range of model membrane systems that have been developed for deposition on the chips surfaces that include planar, polymer cushioned, tethered bilayers, and liposomes. This is followed by a description of the different chemical immobilization or physisorption techniques. The application of this broad range of engineered membrane surfaces to biomolecular-membrane interactions is then overviewed and how the information obtained using these techniques enhance our molecular understanding of membrane-mediated peptide and protein function. We first discuss experiments where SPR alone has been used to characterize membrane binding and describe how these studies yielded novel insight into the molecular events associated with membrane interactions and how they provided a significant impetus to more recent studies that focus on coincident membrane structure changes during binding of peptides and proteins. We then discuss the emerging limitations of not monitoring the effects on membrane structure and how SPR data can be combined with DPI to provide significant new information on how a membrane responds to the binding of peptides and proteins.

CONTENTS 1. Introduction 2. Optical Principles and Instrumentation 2.1. Surface Plasmon Resonance (SPR) 2.1.1. Propagating Surface Plasmon Resonance 2.1.2. Localized SPR on Nanostructured Metal Substrates 2.1.3. Combination of Propagating and Localized SPR on Nanostructured Arrays 2.1.4. SPR Imaging/Microscopy 2.1.5. Extraordinary Optical Transmission (EOT) Through a Metallic Nanohole/ Nanopore Array 2.2. Dual Polarization Interferometry (DPI) 2.2.1. Optical Waveguide Principle: Far-Field Pattern of Interferometry 2.2.2. Dual Polarization Modes 2.2.3. Anisotropy and Birefringence 2.2.4. Optogeometrical Parameters 2.2.5. Assumptions and Limitations of Calculation of Mass 3. Model Membrane Systems 3.1. Supported Lipid Bilayers (SLBs) 3.1.1. Physisorbed Lipid Bilayers

© 2018 American Chemical Society

3.1.2. Covalently Immobilized Lipid Membranes (Chemisorbed/Tethered) 3.1.3. Polymer-Cushioned Planar Bilayers 3.1.4. Tethered (Chemisorbed) SLBs 3.1.5. Lipid Bicelles and Nanodiscs 3.2. Lipid Vesicles on Solid Supports 3.2.1. Surface Modifications for Liposome Adsorption 3.2.2. Direct Coupling of Liposomes to Surfaces 3.2.3. Liposomes Adsorbed to Protein- and Biopolymer-Modified Surfaces 3.2.4. Classical, Non-Selective Chemistry for Lipid Functionalization 3.2.5. Bio-Orthogonal Coupling Chemistry for Lipid Functionalization 3.3. Micro- and Nano-Arrayed Lipid Bilayers 3.4. Naturally-Derived Model Membrane Systems 4. Major Applications of Optical Biosensors in Biomembrane Research 4.1. Characterization of Membrane Structural and Physical Properties

5393 5396 5396 5396 5400 5401 5402

5403 5405 5405 5406 5407 5408 5408 5409 5411 5411

5413 5413 5415 5416 5417 5418 5419 5419 5422 5423 5426 5427 5430 5430

Received: December 11, 2017 Published: May 25, 2018 5392

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews 4.1.1. Membrane Thickness 4.1.2. Bilayer Order 4.1.3. Bilayer Phase Transition 4.1.4. Lateral Structural Characterization of Membranes by SPRi 4.2. Characterization of Molecule-Biomembrane Interactions by SPR (Mass Only) 4.2.1. Peptide-Membrane Interactions 4.2.2. Curve-Fitting Strategies and Kinetic Models to Determine Affinity Constants for Membrane Interactions 4.2.3. Membrane-Lytic Peptides 4.2.4. Membrane Interaction of Peripheral Membrane Proteins in MembraneMediated Signaling 4.2.5. Protein Translocation Through the Membrane 4.3. Characterization of Peptide Binding and Changes in the Bilayer Integrity by Mass and Birefringence 4.3.1. Kinetic Analysis of Experimental Data to Define Binding Mechanisms 4.3.2. Profiling of Lipid Perturbation and Disruption by Membrane Active Peptides 4.4. Membrane Protein Receptors 4.4.1. Membrane-Based SPR and DPI of G Protein-Coupled Receptors 4.5. Binding of Molecules to Membranes on Nanostructured Substrates 5. Conclusions Author Information Corresponding Author ORCID Notes Biographies Acknowledgments References

Review

these membrane proteins are controlled by the highly sophisticated cooperativity of protein structures, topologies, and assemblies with the intrinsic physical properties of the membrane. 2−5 In addition to the membrane-mediated functions of intrinsic membrane proteins, membranes also serve as templates for molecular interactions involving specialized lipid species or particular membrane physical properties which impact on diverse membrane processes associated with immuoregulation, cellular morphogenesis, cytokinesis, assembly of trafficking vesicles to the function of organelles, membrane partition, or resistance of drugs, cell death regulation, and cell invasion by pathogens.6−16 However, many phenomena still remain to be comprehensively characterized in which the exact mechanisms of how different proteins drive specific types of membrane transformation. Protein interactome maps do not consider such specific interactions and thus cannot predict precise outcomes of the interactions of the involved proteins and particular membrane. These can only be inferred from experimental characterization of reconstituted membrane systems where precise biochemical and physical interactions can be elucidated and differentiated to provide mechanistic details of molecule-membrane interactions. The molecular composition of lipids, proteins, and sugars in biomembranes are highly complex and exhibit great diversity in their molecular and supramolecular structures with varying physical properties and biological functions. The relative proportion and distribution of lipid molecules in different membranes are subjected to regulatory networks to maintain membrane homeostasis for their specialized functions.17 Perturbation of the underlying membrane-derived cooperativity and homeostatic machinery can lead to malfunction of the system and deterioration into pathological states and cell death. Lipids display remarkable structural diversity, driven by factors such as variable chain length and headgroup, a multitude of oxidative, reductive, substitutional, and ring-forming biochemical transformations as well as modification with sugar residues and other functional groups of different biosynthetic origin. The major classes of lipids found in biomembranes are summarized in Figure 1. In eukaryotic membranes, the glycerolphospholipids are the predominant lipid components, while sphingosine-based lipids and sterols also constitute a major fraction. In most prokaryotic membranes, phosphatidylethanolamine (PE), phosphotidylglycerol (PG), and cardiolipin (CL) are major glycerolphospholipids [the main eukaryotic phosphatidylcholine (PC) is absent in prokaryotic cells]. In plant cell membranes, the sugarcontaining glycerol-based lipids, such as monogalactosyl and digalactosyl diacylglycerol, are found in high proportion. In addition to the structural diversity of the head groups and backbone of lipid molecules, the molecular diversity of lipids in membranes is much more complex when the different length, degree of saturation, and structures of the acyl chain are considered. There are no reliable estimates of the number of discrete lipid structures in nature, due to the technical challenges of elucidating chemical structures, although this may change with recent advances in lipidomics.18 Given the importance of these molecules in biomembrane function, cellular homeostasis, and pathological disorder, well-organized databases of lipids are essential to link relevant structural information and related features to the physical properties of lipid and membrane and their role in mediating biomolecular interactions.19−21 A comprehensive list of lipid classes and subclasses, as well as the common abbreviations, can be found

5430 5434 5437 5437 5438 5438

5439 5446

5448 5449

5451 5452

5452 5457 5458 5461 5464 5465 5465 5465 5465 5465 5466 5466

1. INTRODUCTION The propensity of lipid molecules to self-associate into closed structures in aqueous environments is the physical basis for membrane formation in the genesis of cellular life. Biological membranes are outstanding examples of molecular assemblies of extreme complexity whose structure and function cannot be determined from the genome alone and which present some grand challenges to science. Due to the immense importance for life processes, research on biological membranes has for a long time been central and very active in biomedical and biophysical fields. Multiple disciplines including biochemistry, pharmacology, molecular biology, biophysics, and engineering have all contributed to our current knowledge of biological membranes and their wide-ranging functions. However, progress in the fundamental understanding of membranes has arguably not matched the pace of research related to proteins and DNA. Biomembranes exist in every form of life as the most versatile self-assembled system with the capability to incorporate proteins that function as transmembrane ion and photon conducting channels, signaling receptors, active/passive transporters, metabolic enzymes, anchorages for the intracellular cytoskeleton and extracellular matrix, cell−cell recognition, intercellular junctions, and energy production.1 Functioning of 5393

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

membrane surface. Lipid curvature stress controls the organization of lipids into various types of hexagonal (HI and HII), cubic (micellar and bicontinuous), and rhombohedral phases which are associated with a specific packing parameter defined by the relative shape of the headgroup and lipid chain according to the simplistic “molecular shape” model.24 Additional important features of biomembranes are the existence of cross-sectional and lateral asymmetries. The arrangement of particular lipid classes on either side of the bilayer structure and the enzymatic processes responsible for maintaining lipid asymmetry are critical in regulating networks of lipid metabolism, intracellular signaling, and molecular transport. The cytosolic leaflet of the mammalian cell membrane is rich in phosphatidylserine and phosphatidylethanolamine, whereas the extracellular monolayer is composed of high levels of phosphatidylcholine and sphingomyelin. There is also a great deal known about how this asymmetry is exploited in a range of different physiological responses. The lateral asymmetry of lipid distribution has underpinned the concept of membrane rafts and speculation about their possible role in trans-membrane signal transduction. Given the observed diversity of membrane lipids, there is obviously scope for specific interactions between different molecular species of lipids in biological membranes. The interaction between sterols and phospholipids to create ordered domains in monolayer and bilayer membranes is of particular interest in this context.25 These asymmetric properties induce variable lateral pressures, line tension, and transverse forces within the membranes which collectively regulate the structure−activity of membrane receptors and transmembrane channels.26 The combination of these different physical features with different lipid conformations are important in modulating the membrane morphologies in various cellular processes ranging from intracellular vacuole transport to cell locomotion and cytokinesis. The different membrane topologies in combination with different physical properties are also important in regulating the activity of transmembrane proteins and in mediating the binding of peptide/proteins to the membrane. Various analytical technologies have been applied to characterize the molecular properties and chemical compositions of biomembranes and to explore the molecular mechanisms of changes in the membrane structural properties associated with molecule-membrane interactions.27 These techniques include NMR, X-ray scattering, neutron scattering, AFM, CryoEM, SEM/TEM, FTIR, DSC, force microscopy/ spectroscopy, secondary ion mass spectrometry/imaging, fluorescence spectroscopy, high-resolution fluorescence microscopic imaging, and microfluidic optical sensors.28,29 Optical techniques that integrate biosensing are among the highly sensitive biophysical techniques used to obtain both qualitative information and quantitative measurement of biomolecular interactions in real-time.30−35 Such interactions can be detected with specific labeling of biomolecules with fluorescence probes or detected directly without labeling. The specific labeling requires robust chemistry to synthesize or attach specific probes to the molecules of interest. The probe can also alter the structural and physicochemical properties of the biomolecules and interfere with the biomolecular interaction. Label-free biosensing techniques relying on the intrinsic optical properties of the biomolecules are therefore preferable in studying biomolecular interactions. The intrinsic properties including autofluorescence, Raman scattering, refractive index (RI), and optical path length (a product of RI and geometric size) have

in the classification system provided by the Lipid MAPS consortium.20 There are more than 40000 lipid structures that have been characterized of the estimated 180000 lipids based on acyl/alkyl chain and glycan permutations for glycerolipids, glycerophospholipids, and sphingolipids. This number is almost certainly a conservative estimate22 and also excludes the complexity arising from isomeric structures that differ only in double-bond position, backbone substitution, and stereochemistry, or the lipids derived from prenols or polyketides.22,23 Despite the wealth of information obtained from various experimental techniques on membrane structure−function, our understanding of the enormous diversity of lipid structures in biological membranes is still limited. One of the enduring sources of scientific questions with regard to the properties of cell membranes is the highly complex assortment of polar lipids and polymorphic molecular organizations with highly dynamic physical properties that make up the bilayer matrix. Why does such diversity exist when most membrane functions, such as the permeability barrier properties and catalytic activity of intrinsic membrane proteins, can be reconstituted using one or a mixture of a few molecularly defined lipids? Likewise, it is also critical to improve our knowledge on the biochemical and biophysical principles that regulate the membrane homeostasis of lipid compositions for preserving each distinct membrane morphology in living cells.17 Notwithstanding the complexity of biomembrane composition, our understanding of the cell membrane has significantly advanced in recent years with the development of various model membrane systems and improved technologies that have provided increasing insight into the organization and structure of various components and domains that form an integral part of the membrane. Efforts are now underway to selectively identify and target membrane domains, particularly membrane proteins, to modulate their structural integrity or function, thereby leading to the desired downstream effects in the cell. Different classes of membrane proteins are also being explored as drug targets for new therapeutic indications. However, many membrane proteins are yet to be fully exploited as drug targets, partly due to the difficulties in their expression in large quantity, isolation, and purification for further biophysical characterization. Membranes even composed of a single lipid component also have complex physical properties in different length and time scales.4 As shown schematically in Figure 1, the self-assembly of lipid molecules into lamellar phases are highly polymorphic in supramolecular organization and different lyotropic crystalline phases. These lyotropic phases exist in dynamic transition between gel and liquid crystalline phases, thickness modulation, electric charges associated with membrane surface potential, variable lipid cross-sectional area induced membrane curvature, and the packing and ordering of lipid molecules in relation to the elastic expansion, bending, and surface density of membranes. Furthermore, lipid molecules are not covalently bound in membranes and mainly stabilized by van der Waals interaction between chains as metastable complex systems whose molecular arrangement changes dynamically. In addition, membrane interdigitation and domain formation increase the lateral heterogeneity in membrane structures. Although the lamellar phase is the most commonly characterized lipid arrangement for membranes, nonlamellar phases can also be induced by the presence of high levels of nonlamellar-prone lipids such as phosphatidylethanolamine that apply lateral pressure on the polar head groups at the 5394

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Figure 1. Chemical diversity of lipid compositions (left) and the complex physical properties and structural polymorphisms (right) of biomembranes. The glycerophospholipids are varied with different structures of the polar groups linked to the phosphate groups, the linkages of the sn-1 acyl chains to the glycerol backbones and the diverse length, degree of unsaturation, and branches of acyl chains. Other major membrane lipid classes include sphingolipids of various polar moieties (choline and disaccharides) with different acyl chain length and saturation, sterols of various ring and side chains and glycerolipids of mono- or diacyl chain composition with various length and saturation at different position of the glycerol backbone. Lipid molecules can self-assemble into lamellar or nonlamellar phases depending on the lipid geometrical shape and packing parameters and the conditions of the surrounding media. Highly polymorphic structures with various physical properties can coexist in the biomembranes and undergo dynamic transitions between different phases regulated by complex molecular mechanisms of lipid−lipid and protein−lipid interactions.

in proximity to the metal surface also provides great potential in probing molecule-membrane interactions. Another sensing technique, dual polarization interferometry (DPI), is based on an integrated planar optical waveguide interferometer that combines evanescent field sensing and optical phase difference measurement methods and provides the advantage of multiparameter measurements in a single binding assay which yield the optogeometrical properties of density and thickness of the adsorbed layer.36 The significance of these features lies in the ability to now analyze the impact of membrane active peptides and proteins on the structure of the bilayers simultaneously with the mass changes associated with the binding event. The microinjection and microfluidic systems of the biosensor allow high throughput screening of the structural and compositional factors that mediate peptide/protein binding to the membrane.52−55 However, the use of biosensors in studying membrane interactions is still relatively low compared to their application in other nonmembrane molecular binding studies.31,32 This is partly due to the lack of robust protocols to create stable, reliable membrane systems on the biosensor chip in situ and the inability to provide information on the overall geometrical properties of the membrane. Nevertheless, several optical biosensors have been developed to study the formation of biomembranes and to determine the geometrical properties, including mass per unit area, thickness, and density and molecular packing of artificial and native cell membranes either

been exploited for label-free biosensing. Specific signal enhancement is required for measuring the autofluorescence and Raman scattering for biomolecular interactions as in surface-enhanced Raman spectroscopy. The RI and optical path length are highly sensitive and can be measured accurately with techniques such as surface plasmon resonance (SPR) and interferometry. Both the plasmonic and interferometry integrated optical biosensing techniques with the use of evanescent field sensing within 100−200 nm distance above the surface are widely used to monitor the real-time binding process for understanding the mechanisms and kinetics of molecular interactions. The historical development and integration of interferometry and plasmonic phenomena into various biosensing technologies for the characterization of molecule-membrane interaction is highlighted schematically in Figure 2. Since the early discoveries of both optical interference and the SPR phenomenon, the introduction of various commercial SPR and interferometry instruments around the 1990’s36−41 together with the fabrication of versatile model membrane systems,42−44 have advanced our knowledge of the kinetic events and the changes in the physicochemical properties associated with the binding of molecules to membranes.31,45−51 SPR has been the most widely used technique to explore the binding mechanisms of biomolecules to the membrane. Localized SPR (LSPR) with a much smaller evanescent field 5395

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Figure 2. Schematic timelines of the development of interferometric and plasmonic analysis and their applications in biomembrane research.

immobilized or physisorbed onto the sensor chip.31 Below we present a detailed overview of the physical principles and application to membrane interactions of SPR and DPI and discuss the significance of the experimental outputs to our increasing understanding of the role of the membrane in a wide range of biological processes. We first discuss experiments where SPR alone has been used to characterize membrane binding and describe how these studies yielded new insight into membrane interactions and how they provided a significant impetus to more recent studies that focus on coincident membrane structure changes during binding of peptides and proteins. In particular, we discuss the limitations of SPR in terms of monitoring the effects on membrane structure and how SPR data can be combined with DPI to provide significant new information on how a membrane responds to binding of peptides and proteins.

There are two principal types of SPs depending on the geometrical shapes and configurations of the metal-dielectric media (Figure 3). Most commonly used SPR sensor techniques are based on the propagating plasmons on which SPPs are generated and propagated along smooth planar metal surfaces. Another type of SPs, the localized surface plasmons (LSPs) are generated at nanostructured metal surfaces such as nanoparticles and/or isolated nanostructures and a nanoaperture in a metal film.61−63 These two types of SPs differ both in their excitation methods and intrinsic physical properties. Different aspects of membrane structures on the biomolecular interactions can be characterized separately with propagating SPRs and LSPRs. 2.1.1. Propagating Surface Plasmon Resonance. For propagating SPRs, since the propagation constant (or wavevector) of SPPs is greater than the wavevector of the incident light in the dielectric medium, the SPPs on a planar dielectric− metal interface cannot be excited directly by the light incident onto the smooth metal surface (Figure 3A). Hence, a prior enhancement of the light wavevector parallel to the interface is required to match the propagation constant of the SPPs for the excitation of SPPs modes and resulting in resonance. The wavevector of the incident light can be enhanced to match the propagation constant of the SPPs through methods of attenuated total reflection (ATR) or diffraction. The enhancement and subsequent matching of the light wavevector to the SPPs propagation constant are achieved by introducing a coupling device (coupler). Among the various commonly used couplers with different reflection methods such as prism coupler for ATR, grating structure for wave diffraction, and dielectric waveguide coupler (optical fibers) for ATR, prism couplers are the most commonly used approach for optical excitation of SPPs. Both Kretschmann and Otto prism coupling configurations have been employed in SPR-based biosensor designs.64,65 In the most commonly used Kretschmann configuration coupling in ATR conditions, a monochromatic light passes through the high refractive index prism and is totally reflected at the base of the prism. For this conventional resonator, SPPs can only be excited and resonated by p-polarized light at a specific incident angle for a given light wavelength. Thus, the field of a surface plasmon is a transverse-magnetic (TM) polarized mode in which the vector of the magnetic field is perpendicular to the

2. OPTICAL PRINCIPLES AND INSTRUMENTATION 2.1. Surface Plasmon Resonance (SPR)

Surface plasmons (SPs), also commonly known as surface plasmon polaritons (SPPs) or surface plasma waves (SPWs) can be considered as electron density waves at the surface of conducting materials (such as metal) propagating along the interface of dielectric and metal media.56−60 The presence of free conduction electrons at the metal surface is essential for the generation of SPPs at the interface between a conducting material (ie, metal) and a dielectric medium with oppositely charged dielectric constants. These conduction electrons cannot freely propagate and become bound to the metal surface. When the metal-dielectric media are excited with external light (electromagnetic) waves under certain conditions, these electrons respond collectively by oscillating in resonance with the light wave. This collective oscillation of free conduction electrons driven in resonance with the external light waves at the interface of metal-dielectric media is referred to as SPR. The SPP oscillations are associated with an electric field propagating along a metal−dielectric interface, and thus, SPs are also viewed as electromagnetic waves bound strongly to the interface. The surface plasmon field intensity at the interface can be made very high, making SPR a powerful tool for characterizing various molecular interactions at the interface. 5396

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Figure 3. Schematics of optical configuration and data outputs of (A) prism-coupled surface plasmon resonance (SPR), (B) localized surface plasmon resonance (LSPR), (C) extraordinary optical transmission (EOT), and (D) dual polarization interferometry (DPI). (A) The optical configuration of Kretschmann prism-coupled surface plasmon excitation and sensing consists of a prism overlaid with a thin metallic film (Au or Ag) film. The resonant excitation of plasmons by incident light occurs at a defined combination of resonant angle and optical wavelength and generate surface electromagnetic waves propagating in a direction parallel to the metal−dielectric interface. The angular dependence of reflectivity referred to as the SPR spectrum shows a very sharp dip at the resonance condition. The total internal reflection of incidence light from the metal thin film generates an evanescent field in the vicinity of the dielectric−metal interface as a RI sensitive sensing volume with the field strength decaying from the surface. Molecular layers formed on the dielectric−metal interface interact with the evanescence field, thereby altering the resonance excitation process. This results in changes in the intensity of reflected light by the metal film and measured as a shift of the reflectance angle. The formation of a lipid bilayer on the chip surface and the subsequent binding of molecules to the lipid bilayers can be characterized in real time and provides a platform for studying the affinity and kinetics of peptide-membrane interactions and ligand−receptor binding. (B) The resonance of incident light wave with collective oscillation of the conduction elections confined in the subwavelength metallic nanoparticles and around nanoholes or nanowell in thin metal film results in localized nonpropagating plasmon excitations (LSPR). The LSPR can be detected by optical spectroscopy as extinction, scattering, and adsorption spectrum. The bandwidth, height, and position of the spectral extinction maximum or spectral dip minimum depend on the geometric parameters of the nanostructure and the dielectric function of the surrounding medium. Local RI changes induced by biomolecular interactions at the surface of the nanostructures can be monitored via the spectral peak shift and provide the kinetic mechanism of the molecular 5397

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Figure 3. continued interaction. (C) The transmission efficiencies of light through 2D periodic arrays of subwavelength apertures in an optically opaque metal film can exceed unity of normalized total aperture area, which is referred to as extraordinary optical transmission (EOT). This EOT is a consequence of surface plasmon excitations when the metallic nanohole array is illuminated at normal incidence and is characterized by a series of maxima and minima in the zero-order transmission spectrum. The spectral characteristics can be finely tuned by varying the pore diameter, shape, and depth and array periodicity for optimal sensing sensitivity and depends on the optical properties of the metal−deielectric interfaces. The formation of molecule thin film on the array of nanoholes can be characterized by the wavelength shift as a function of bulk RI changes. The spectral peak shift can also be monitored in real-time for the kinetics of molecular interaction. (D) The dual polarization interferometer (DPI) measures the phase changes of interference patterns as lightwaves pass through a dual slab waveguide of two high-refractive index structures (sensing waveguide and reference waveguide). Two orthogonal polarizations, TM and TE, pass through the waveguide and are detected separately as a phase shift in the fringe pattern at the far-field, which provide different sensitivity to the bulk and in close proximity of surfaces which provide the thickness and RI values for the adsorption of molecules onto the surface. Thus, the process of vesicle adsorption-to-unilamellar and the subsequent dynamic impact of peptide binding on the lipid molecular organization can be quantitatively analyzed in real time. Such molecular order of membrane is measured by optical birefringence (Δnf), differences of RI for each TM and TE polarization mode. The changes in the packing, alignment, and degree of order of lipid molecules assembled on the surface upon peptide/protein interaction can be monitored simultaneously with real-time changes in mass and birefringence. The correlation of membrane-bound peptide mass with bilayer disordering provides unique transformation profiles in delineating the mechanisms of the impact of bound peptides on membrane structure.

Table 1. Comparison of Different Signal Modulations Used as SPR Sensor Outputsa

a

Compiled from refs 66 and 67.

direction of propagation. The SPPs propagate along a direction within the planar metal−dielectric interface that is characterized by the electromagnetic field distribution and wavevector (propagation constant). As the excited SPPs are confined in the near-field, the maximum intensity of the electromagnetic field is in the vicinity of the metal−dielectric interface and the intensity of the field decays exponentially into both the dielectric and metal media. The field decay perpendicular to the direction of SPPs propagation into the dielectric medium is denoted as the evanescent field and characterized by the penetration depth. This penetration depth is then defined as the perpendicular distance from the interface at which the intensity of the field is decreased by a factor of the natural

logarithm (e). The penetration depth varies with the wavelength of incident light and the permittivity of the medium. For example, the penetration depth into a dielectric medium with RI = 1.32 of a gold interface increases from 100 to 600 nm as the wavelength increases from 600 to 1000 nm.58,59 The penetration depths vary with the interface material: for example, for a light with wavelength 648 nm incident to a gold film, the depth is 351 nm in air, 191 nm in water, and 27 nm into the gold.33 This interface sensitivity of the evanescent field and the penetration depth are exploited to detect the changes in the dielectric properties (a changing refractive index) from molecular interactions occurring in the vicinity of the interface. 5398

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Typically, the metal films with resonant wavelengths from 500 to 800 nm are commonly used in the wavelength modulation.75,76 However, the wavelength shift in an SPR sensor can also be measured in the near IR region using a Fourier transform spectrometer. The angle or wavelength shift in sensor responses expressed in RUs can be related to changes in the refractive index (Δn) and surface concentration of surface-bound molecules (ΔΓ) when the adsorption of biomolecules occurs within a layer of thickness (h) at the sensor surface less than the penetration depth. On the basis of the de Feijter formula77 in which the refractive index change is linearly dependent on the surfacebound molecular mass/area, the surface bound molecules in mass per unit area can therefore be derived from the experimentally determined refractive changes by

The magnitude of SPP propagation is characterized by the propagation constant (Kspp), which is determined by the dielectric constant of both metal (εm) and dielectric (εd) media for a given angular frequency (ω) of incident light as K spp =

ω C

εm·εd 2π = n p sin θ εm + εd λ

where C is the velocity of light in vacuum. For a particular combination of wavelength (λ) and angle of incident light (θ) through a prism with refractive index (np), the wavevector of the p-polarized light is matched to the SPP propagation constant and couples the evanescent wave to the surface plasmon. This coupling of SPPs with the incident light wavevector only occurs within a narrow range of incidence angles or wavelengths and is accompanied by the transfer of optical energy into the SPs and its dissipation into the metal medium. This absorption of incident light energy by SPPs results in a reduction of the intensity of the reflected light and the angular, and wavelength dependency of reflected light results in characteristic angular, wavelength, light intensity, and phase modulations in the SPR spectrum. Alteration in the optical properties of the metal−dielectric interface as a result of analyte adsorption will change the Kspp leading to shifts in the signal modulations. All these angular, wavelength, light intensity, and phase signal modulations can be readily integrated into the most commonly used prism-coupling methods as a sensor output. Comparison of the characteristics of these four signal modulations is summarized in Table 1. 2.1.1.1. Angular Modulation. The angular modulation is commonly integrated into most commercial SPR sensors in which the reflectivity variation is measured as a function of incident angle of a monochromatic light.68 A characteristic drop in the reflectivity occurs at a particular angle which is denoted as the resonance angle or SPR angle. The position of this SPR angle (minimum reflectivity) is highly sensitive to the refractive index changes in the vicinity of the metal−dielectric interface.69 Interaction of molecules with the ligands immobilized on the metallic surface induces changes in the refractive index near the surface, thus giving rise to a shift of the resonance angle. The angular shift response is normally expressed in resonance units (RUs) and can be monitored in real-time throughout the association−dissociation process of molecular interaction. A He−Ne laser (632.8 nm) is commonly used to measure the shift in resonance angles upon adsorption of molecules. However, dual-wavelength approaches have been developed to determine both the thickness and the dielectric constant (RI) of a molecular layer formed on the metal surface.70,71 In addition, the use of near-IR and IR wavelengths for scanning angle shifts have also been reported to yield smaller angle shifts and sharper minimum with increased dynamic range.72,73 The near-IR and IR wavelength can be used to measure thicker films containing molecules that absorb visible light. 2.1.1.2. Wavelength Modulation. In SPR systems with wavelength modulation, the SPs are excited by polychromatic incident light at a fixed angle which provides an alternative SPR detection method.74,75 In this case, the reflectivity variation as a function of wavelength is measured by a spectrometer. A spectral adsorption minimum occurs at a particular wavelength as the resonant wavelength. Similar to the angular modulation, the resonant wavelength is sensitive to the changes in the refractive index of the dielectric medium on the metal surface and shifts upon adsorption of molecules at the interface.

Δn =

⎛ dn ⎞ ΔΓ ⎜ ⎟ ⎝ dc ⎠ h

where dn/dc is the refractive increment index of the molecule. Typically, the refractive index increment for most biomolecules ranges from 0.1 to 0.3 cm3/g. The dn/dc is about 0.182 cm3/g for proteins and nucleic acids and 0.135 cm3/g for lipids within 8% variability.78−80 Typically, the sensor response changes by the adsorption of 10pg/mm2 proteins to the surface is comparable to the bulk refractive index changes of 10−5 RIU. The relationship between the changes in the refractive index at the metal surface and the sensor output is also dependent on the type of coupling devices, modulation methods and incident wavelength, and the spatial distribution of the refractive index change. Highly sensitive SPR sensors have been developed to measure molecules at concentrations as low as 0.1pg/mm2 in which the optical system is able to resolve shifts of resonant wavelength at 0.5 pm or where the resonant angle is less than 10−5 degree.58 2.1.1.3. Intensity Modulation. The transfer of energy from the incident photon to the SPs under the resonant angle or wavelength is also accompanied by the simultaneous intensity and phase changes of reflective light. This phenomenon is characterized by a drastic drop in reflective intensity and a sharp phase jump.81,82 In contrast to the angular and wavelength modulations used in SPR spectroscopy, the intensity and phase modulations are integrated into the SPR imaging (SPRi) systems which couples the sensitivity of SPR with the spatial capabilities of imaging and is compatible with a microarray format. 83−87 Most SPRi systems adopt the Kretschmann coupling configuration incident at a fixed angle and use a charge-coupled device (CCD) camera, detector, or microscope to measure the spatial variation in reflected light. The intensity-interrogated SPRi techniques measure the reflected light intensity changes caused by the refractive index variations on the metal−dielectric interface at a fixed incidence angle and wavelength. The variation in the reflected light intensity is expressed as the percent reflectivity, %R. The measurement of RI changes from the intensity-interrogated SPRi systems generally have low sensitivity (%R/RIU) and resolution due to the noise generated directly from the light source. The noise level can be reduced by referencing to improve the sensitivity and resolution of the intensity SPRi by 1 order of magnitude with the use of a dual wavelength from two sequentially switched laser diodes or by the introduction of elliptically polarized light and polarization contrast from two crossed polarizers.88,89 5399

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

2.1.1.4. Phase Modulation. On the basis of the Frensel formulas, the light reflected from a metal surface is accompanied by the phase changes of light beam (Δϕ). As the SPs can only be excited by the p-polarization, the ppolarized component experiences significant phase change while the s-polarized component remains unchanged and is used as a reference. The phase difference due to interference observed through spatial displacement of the light beam is the most sensitive to the refractive index changes with the lowest detection limit at the interface for SP phenomena.90−92 This highly sensitive phase modulation is due to the fact that the plasmon wave vector that probes the electric field for the phase modulation is maximal at the very minimum of the SPR curve. This electric field enhancement of the maximal phase change is much larger than the angle shift per unit change of interfacial RI. In addition, the phase noise can be orders of magnitude lower compared to the intensity modulation, thereby providing an improved signal-to-noise ratio. Furthermore, measurement of phase changes with referencing also provides better spatial and temporal mapping and a subsequent signal averaging/ filtering for the image treatment.93 Despite the distinctive features of phase changes implemented as an SPR sensor output, more sophisticated optical phase retrieval methods based on interferometry, polarimetry, and optical heterodyning are required to extract phase information which are less commonly integrated into commercial SPR sensors.58,67,93−95 However, the phase mode has become more prominent in twodimensional microarrays of SPRi for multiplex analysis of molecular binding and microchannel/fiber-based SPR sensors. 2.1.2. Localized SPR on Nanostructured Metal Substrates. The propagation of SPR on a planar metal− dielectric interface with a coupling device in the conventional SPR sensor has proven very effective in the monitoring and characterizing of biomolecular interactions with a sensitivity that usually ranges between 10−5 and 10−7 refractive index units (RIU).58 However, the configuration of the propagating SPR sensors has limited potential in multiplexing and miniaturization of plasmonic sensors. Reducing the dimension of the plasmonic surface on which the evanescent wave is propagated and the introduction of plasmonic nanostructures for enhanced sensitivity have significantly increased the types of SPR sensing platforms and the potential applications for SPR. The type of SPPs is dependent on the dimension of the metallic substrate. For SPPs propagation, the dimension of the substrate has to be longer than the incident wavelength. As the dimension of the metallic structure is smaller than the incident wavelength (propagation distance), the plasmons cannot propagate freely along the interface. When light impinges on the plasmonic metallic nanostructures, the free electrons oscillate collectively relative to the lattice of positive ions at a frequency that matches the frequency of incident light. This collective oscillation of electrons is confined to a finite volume at the localized surface of metallic nanostructures and is referred to as localized surface plasmon resonance (LSPR) (Figure 3B). Due to this resonant oscillation of electrons, plasmonics have unique wavelength-dependent extinction spectra, including absorption and Rayleigh scattering light and produce electromagnetic field enhancements at the nanoparticle surface. The light absorption and scattering by nanoparticles at the UV−visible region generate distinctive colors and also lead to extinction or attenuation in intensity for the incident light. The unique phenomena of light absorption and scattering of metallic nanostructures and electric field

enhancement of LSPR can all be integrated for spectroscopic analysis using LSPR biosensing. In addition, the strong electric fields on the surface of nanostructures greatly enhance the Raman scattering and fluorescence signals for the molecules in the vicinity of the surface and provide these nanoparticles as new probes for biosensing and imaging. The factors controlling the light extinction (absorption and scattering) properties of the metallic nanosphere can be described by the Mie theory, which corresponds to a rigorous analytical solution of Maxwell’s equations with spherical boundary conditions. For a light incident through a suspension of spherical nanoparticles, the molar extinction coefficient is correlated to the extinction cross section of an individual particle. The extinction cross section (Cext) of the metallic nanosphere with a radius (r) much smaller than the incident wavelength (λ) can be calculated using the Mie theory as follows Cext =

24π 2r 3Nεm 3/2 εi × λ ln(10) (εr + Xεm)2 + εi 2

where εm is the dielectric constant of the surrounding media, while εr and εi are the real and imaginary parts of the complex function of the metal dielectric constant (ε = εr + iεi), respectively. A shape factor (X) is also considered in the calculation of Cext. The values of X can be varied from 2 for a nanosphere to as large as 20 for asymmetric nanoparticles with a high aspect ratio. This clearly shows that the peak of adsorption and scattering to absorption ratio of the LSPR are influenced by the size, geometric shape, structure, and composition of the metal nanostructures, incident wavelength, and the surrounding dielectric medium. The molar extinction coefficient of the nanoparticles can be as high as 1011 M−1 cm−1 with several orders of magnitude higher than those of organic dyes. However, Mie theory is only applicable to the dielectric constants of both the nanoparticles, and the dielectric medium are homogeneous and can be described by their bulk dielectric function. Different solutions have been developed for other nonspherical shapes. For nonspherical nanostructures, the SPs are unevenly distributed and polarized to the tip/end of the structure characterized by a shape dependence of the LSPR extinction spectra. The extinction spectra for the plasmon resonance of Au nanorods split into peaks with a strongly redshifted longitudinal mode for polarization perpendicular to the long axis and a weakly blue-shifted transverse mode for polarization perpendicular to the long axis. The separation of the two split peaks can also be enhanced by increasing the aspect ratio of the nanorods. Multiple plasmon resonances with a longitudinal plasmon mode and enhanced electric field are confined to the tips of triangular and squared nanostructures. The Mie theory is best described for nanoparticles much smaller than the incident wavelength with diameter less than 30 nm which accounts only for dipolar oscillation. For larger nanospheres beyond the Rayleigh approximation, the dipolar resonance shifts toward longer wavelength with a substantial broadening caused by retardation effects. Increasing the particle size also enhances the scattering together with lowering the absorbance and impacts on the LSPR sensitivity. Furthermore, the above equation does not account for the interactions between nanostructures and is only applicable for nanoparticles that are well-separated in the solid state or present at low numbers in solution. Coupling of the SPPs occurs as the distance between particles decreases which results in profound 5400

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

electromagnetic field enhancement and this distance-dependent coupling provides a plasmonic molecular ruler enabling highresolution monitoring of molecular conformation. The magnitude and direction of the spectral shift varies with the orientation of the nanoparticle pair with respect to the polarization axis of the incident light. The above solution has also been used to describe the extinction properties of nanospheres, concentric core−shell nanospheres, nanospheroids, and nanocylinders with an infinite length. For other nonspherical geometric nanostructures and boundary conditions, other numerical solutions have been applied to calculate the LSPR extinction spectra of complex nanostructures and their aggregates. These numerical methods include discrete dipole approximation (DDA), finite-difference time-domain (FDTD), the boundary element method (BEM), and the multiple multipole method. The LSPR wavelength shift response of these sensors can be understood using a model of the refractive-index response of propagating surface plasmons on a planar noble metal surface as follows

2.1.3. Combination of Propagating and Localized SPR on Nanostructured Arrays. The established plasmonic sensing and detection using either propagating SPR on a thin metallic film or LSPR on various nanostructures have achieved remarkably low limits of detection and sensitivity. The specific schemes can be implemented to fit a specific sensor purpose by appropriate manipulation and selection of optical properties, detection modulations, and structural parameters. Among various approaches to improve the sensing performance such as tuning the optical properties by changing the size, shape, and composition of the metallic substrate, advances in fabrication and synthesis techniques have greatly improved the throughput, shape control, reproducibility, and optical properties for plasmonic nanostructure arrays. The short-range randomly patterned and long-range ordered periodic nanohole/nanopore arrays have been shown to increase biosensing capability over conventional continuous Au films in the visible-NIR wavelength range and used in other applications such as SERS, optoelectronics, and waveguides. The nanohole array structures exhibit unique SPR characteristics which simultaneously support several SP modes with variable properties.99,100 The length scale of the nanohole array is greater than the wavelength of incident light which support propagating SPs in the semicontinuous region between the holes in the Au film. The subwavelength nanoholes are coupled to the excitation of LSPR, and shorter range plasmons (Bragg SPs mode) are excited by incident light that is diffracted into the sample plane. These SP modes can be excited in either the Kretschmann or transmission configuration and the propagating SPs in Kretchmann configuration exhibited greater sensitivity than the Bragg SPs in transmission mode. The nanohole arrays exhibit propagating or localized SPs depending on the excitation angle in the Kretchmann configuration, and coupling between these two modes occurring at a specific angle greatly enhances the SP resonance for the application of Raman spectroscopy. The optimal sensing performance of the nanohole array can be achieved by tuning the periodicity and diameter of nanoholes, angle, and wavelength of incident light and metallic film thickness (hole depth). Microhole arrays with a hole diameter equal to half the periodicity coupling both the propagating and localized SPs have exhibited enhanced sensing performance with high sensitivity (>3000 nm/RIU) and an RI resolution of 10−6 RIU and a 2-fold improvement in response to the formation of mercapto-hexadecanoic acid (MW = 288 g/ mol) compared to conventional SPR.101 Further improvement in surface sensitivity to RI changes of a nanohole array has been demonstrated by exciting the SPs in the NIR.102 These periodic nanostructure array sensors allow the detection of IgG concentration 1 order of magnitude lower than the conventional SPR and for more accurate calculation of molecular layer thickness. As shown by these studies, the plasmonic modes in the nanohole array become more complex as the dimension and complexities of the plasmonic substrates increase. Optimization of the optical properties of nanohole configurations by tuning the hole diameter, shape, periodicity, and metal film composition and thickness is critical to apply these micro/nanohole arrays for SPR-based sensor applications. Although the detailed mechanisms of the optical properties of a nanohole array are less important in sensing applications, further understanding the optical nature of SP field enhancement that arise from the resonance coupling between different SP modes in relation to different geometrical structural parameters is critical to expand the applications of periodic

⎡ ⎛ −2d ⎞⎤ Δλ = m(na − nm)⎢1 − exp⎜ ⎟⎥ ⎢⎣ ⎝ ld ⎠⎥⎦

where Δλmax is the wavelength shift, m is the refractive-index sensitivity, Δn (na − nm) is the change in refractive index induced by an adsorbate, d is the effective adsorbate layer thickness, and ld is the characteristic electromagnetic field decay length. This model assumes a single exponential decay of the electromagnetic field normal to the planar surface, which is accurate for a propagating surface plasmon but is likely to be an oversimplification for the electromagnetic fields associated with noble metal nanoparticles. While this oversimplified model does not quantitatively capture all aspects of the LSPR nanosensor response, it does provide some guidance for sensor optimization. In particular, the above equation highlights the importance of distance dependence as described by the electromagnetic field decay length, ld. The main differences between the SPR and LSPR phenomena are the different comparative RI sensitivities and the characteristic electromagnetic field decay lengths. For propagation SPR sensors, a planar metallic film is used as the RI-based sensor to detect analyte binding at or near a metal surface. This sensor gains its sensitivity from an extremely large refractive index sensitivity (∼2 × 106 nm/RIU)96 and modest decay length (200−300 nm), and the sensitivity is proportional to the square of the electric field that extends from the metal film. In contrast, the LSPR nanosensor has a modest refractive index sensitivity (∼200 nm/RIU) with a very short EM field decay length of less than 50 nm.97 This difference in refractive index sensitivities is around 4 orders of magnitude lower for the LSPR than the SPR sensor. However, similar actual detection sensitivities were obtained for SPR and LSPR due to the short and tunable EM field decay length that can enhance the sensitivity of the LSPR-based nanosensor. LSPR sensing characteristics demonstrate that the decay length is 5−15 nm or 1−3% of the light’s wavelength and can be finely tuned by selecting the size, shape, and composition of nanoparticles. This is in contrast to the large decay length or 15−25% of the resonant wavelength for the propagating SPR on planar metallic interfaces.98 Overall, the sensitivity of LSPR close to the surface has allowed monitoring of vesicle-to-bilayer transition in the deposition of liposomes. 5401

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

for spatially resolved images.89 An RI resolution of 2 × 10−6 RIU averaging over a 6 mm2 laser beam diameter was achieved. Fabrication of sensor chips with multilayer structures supporting long-range SPR (LR-SPR) has also been developed to improve the sensitivity and RI resolution of SPRi for thinfilm characterization and bioaffinity measurements.111 In comparison to conventional SPR, LR-SPR possesses longer surface propagation lengths along the chip surface, higher electric field strengths, and sharper resonance peaks in the angular resonance curve. In this LR-SPR chip design, a thin film of 1180 nm Cytop, an inert optically transparent amorphous fluoropolymer with an RI close to water, is sandwiched between the SF10 prism and the Au film. Although the implementation of multilayer structures in LR-SPR spectroscopy greatly enhanced the sensitivity 8 fold and RI resolution to 3 × 10−7 ∼ 3 × 10−8 RIU, the use of LR-SPR in imaging only resulted in approximately 20% response enhancement for a DNA hybridization assay compared to the conventional SPR imaging. However, the long penetration depth of LR-SPR has allowed the detection of large analytes such as whole viruses and bacteria.112−114 The performance of SPRi has also been improved using polarization contrast,88,108 in which a spatially patterned multilayer structure on the prism coupler is placed between two cross polarizers and waveplates to create low noise background from an uncoated region of the chip surface. The phase contrast between the p- and s-polarization generates bright images for the sensing regions, and the spatial distribution of intensity and phase changes associated with the RI changes are captured by a CCD camera. A high contrast image can be generated using a nonsensing area as reference where the reflected light is blocked by the output polarizer. An RI resolution of 5 × 10−6 in more than 100 sensing channels was reported and further improvement in RI resolution was achieved by compensating for the fluctuation of the dark current and incidence intensity. While most phase sensitive SPRi systems use a Mach−Zehnder interferometer with fixed waveplates,91,93,95 the RI resolution was further improved by incorporating a liquid crystal variable phase retarder as an electro-optic modulator to create a spatially periodic retardation pattern on the metallic sensing regions with accurate phase shift.115,116 SPR-induced interference fringe patterns recorded with a 2D CCD camera were subsequently reconstructed using a five-step phase-shift algorithm or converting the periodic polarization state into a fringe pattern by a linear polarizer. Such phase contrast SPRi has demonstrated a resolution of 2 × 10−7 RIU with a lateral spatial resolution of 100 μm. By monitoring the real-time changes in the phase shift, the formation of self-assembled alkanethiol monolayers can be spatially characterized on a patterned microarray. While the configurations of conventional prism coupler-interrogated SPRi systems have been optimized to improve the sensitivity and RI resolution, the physical constraint of a prism coupler remains an inherent limitation on the numerical aperture (NA) and magnification of an imaging system. The images generally suffered from poor spatial resolution of about ∼3 μm using a 0.1NA prism substrate illuminated at a wavelength of 633 nm.117 In addition, image blur and distortion are also noticeable due to refraction at the output face of the prism which generates geometrical aberrations strongly altering lateral resolution. Furthermore, the relative movement between sample and the CCD camera can cause the acquired images to move and result in image distortion when the incident angle

nanostructure arrays in studying the kinetics of molecular interactions. 2.1.4. SPR Imaging/Microscopy. While the propagating SPR spectroscopic systems commonly integrated with angular and wavelength modulation are highly sensitive to the molecules binding to the surface for detailed kinetic analysis, the spectroscopic measurement generally lacks the spatial information and is relatively low-throughput with few independent sensing channels. In comparison, the geometrical structure of thin films built for the propagating SPR geometry allows the characterization of biomolecular-surface interaction to be readily derived for the spatial dependence of the electromagnetic field. Surface plasmon microscopy and later as SPR imaging (SPRi) was therefore developed to address these limitations by coupling the sensitivity of SPR with the spatial capability of imaging on an array format and enables spatially resolved, surface sensitive, label-free, high-throughput real-time analysis of biomolecular-surface interaction.83,84,103 The most commonly used SPRi systems are based on the high index prism coupler of the Kretschmann configuration. The SPRi systems integrated with intensity,83,84 angular,104,105 wavelength,106,107 phase modulation,90,91 and polarization contrast88,108 have been developed to monitor the distribution of RI changes along the metal film surface. Among these different modulations, the SPRi systems are generally integrated with intensity modulation in which a collimated p-polarized beam of monochromatic light incidents through a prism coupler onto a thin gold film at a fixed incidence angle close to the coupling angle for the excitation of surface plasmons. The detector module measures the intensity of light reflected from the surface to monitor its spectral or angular distribution. The variation in refractive index in the dielectric layer changes the SPR absorption minimum wavelength which in turn increases the intensity of the reflected light. These changes in the reflective light intensity can be collected by array detectors such as 2D CCD/CMOS cameras. The differential intensity of the neighboring nonsensing area is used to generate high-resolution 2D intensity contrast images for the RI distribution across the sensing surface in real-time with hundreds to thousands of active spots. In contrast to conventional SPR spectroscopy, the intensity-based SPRi has 1 order of magnitude lower RI resolution and sensitivity. The general implementation of prism configured SPRi also showed poor image contrast with low spatial resolution and distortion. A number of improvements in the optics, fabrication of microarray sensor chip, and microfluidic design have been developed to enhance the performance of SPRi in the multiplexed analysis of molecular binding to the micrometresized sensing area. The commonly used monochromatic laser beam has been replaced with a collimated LED white light/ narrow-band near-infrared (NIR) excitation.72 The use of NIR produces sharper resonance with larger reflectivity changes upon absorption and generates high contrast images in the NIR wavelength range 800−1152 nm. The utilization of an incoherent white light source and narrow band-interference filter has also been shown to eliminate interference fringes as seen in the conventional SPRi using coherent laser excitation.109,110 This white light excited SPRi with tunable wavelength provides an RI resolution of 3 × 10−5 RIU with an upper limit of lateral resolution less than 50 μm. In an alternative modification, a dual wavelength-based SPRi uses two sequentially switchable laser diodes as a light source and the difference in the reflectivity for each wavelength was acquired 5402

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

explained by the interplay of various mechanisms involving the coupling between the localized SPs excitations associated with individual nanoholes/pores and grating-coupled Bloch wave SPPs (BW-SPPs) in propagation mode.134−139 In addition, non-SPP diffracted evanescent waves and interference between surface waves and the incident light wave also play a significant role in EOT.136,137,139−143 Furthermore, a nonresonant Rayleigh-Wood anomaly arising from light moving parallel to the metal surface also contributes to the transmission feature as a spectral minimum near the spectral maximum.138,144 In addition, the presence of Fano-type interference effects between the transmitted light and the SPPs causes a further red-shift in the actual transmission maximum compared to that predicted by the approximation based only on the wavelengths of BW-SPPs excited by normal illumination on a subwavelength metallic nanohole array.144,145 The enhanced transmission arising from the interplay of these different modes generally exhibits complex multiple maxima and minima in the wavelength-dependent EOT spectra and also depends on the periodicity of the array and the dielectric function of both the metal and the surrounding dielectric medium.125,130,146 The spectral positions of the transmission peaks (λsp) for a square array of a circular nanohole array excited at normal incidence can be approximated by the following equation

is changed, which limits the use of the angle scanning method. In order to correct these drawbacks of the prism coupler in SPRi, various methods such as double-prism geometry,118 a cylindrical prism geometry, and a tilting CCD camera compensating imaging optics have been used to improve the image quality.119 Alternatively, an optical microscope objectivebased SPR system built on the basis of the Kretschmann configuration using a high numerical aperture oil immersion objective and an inverted microscope has been developed that offers a spatial resolution near the optical diffraction limit of the incident light117 and provides high-resolution SPR images for single molecule optical mapping analysis. A major advantage of objective-type SPRi is that the fixed sample and imaging optical paths throughout incident angle scanning permits a pixel-by-pixel tracking of the reflectivity in the SPR images. Each of these pixels then produce a framed SPR image using minimum angle information. Therefore, image distortion resulting from geometrical aberrations and interferences from the laser intensity can be significantly corrected by using a homogeneous referencing surface.120 Furthermore, the objective-type SPRi also corrects the incident angle divergence at different image fields caused by the abnormality of the objective lens. The use of a high numerical aperture with high magnification can provide a diffraction limit resolution of about 300 nm for the imaging optics.117 A numerical aperture of immersion objective larger than the refractive index of the medium is selected for the SPR angles larger than the critical angles. In contrast to the angle scanning mode in a conventional propagating SPR spectroscopy, the objective lens significantly simplifies the system design which converts the rotational motion of incident light at the sample into linear motion of the stage. This objective-based SPRi with millisecond temporal resolution and submicrometer spatial resolution was applied to study the kinetics of binding between lectin and glycoproteins to map the spatial distribution of nicotinic acetylcholine receptors (nAChRs) in the native membrane environment of a single cell.121 Furthermore, the objectiveSPRi has also been used in single DNA molecules and virus detection and optical mapping of cell−substrate interactions and glycoprotein polarization in native cell membranes during chemotaxis.122−124 2.1.5. Extraordinary Optical Transmission (EOT) Through a Metallic Nanohole/Nanopore Array. The periodic arrays of nanohole and nanopore plasmonic metallic film supporting both propagating and localized SPs in the Kretschmann setup can also be operated in a colinear transmission mode. The elimination of a bulky prism coupler has allowed for simpler collinear optical arrangement which is ideal for system miniaturization and imaging. In this zero-order transmission mode where the incident and detected light are collinear, the light transmitted through arrays of periodic subwavelength nanohole or nanopore milled through a thin Au or Ag film is several orders of magnitude more efficient than predicted by classical Bethe aperture theory. This transmission enhancement is referred to as extraordinary optical transmission (EOT), where the transmission is associated with an efficiency greater than 1 (i.e., the flux of photons per unit area emerging from the aperture is larger than the incident flux per unit area125) (Figure 3C). Since its initial discovery, the EOT effect has been extensively studied both theoretically and experimentally for its potential applications in nanoplasmonic sensing and for understanding the underlying optical mechanisms.126−133 The enhanced transmission has been

λsp(i , j) =

P i2 + j2

×

εm(λ) × εd εm(λ) + εd

where P is the periodicity (lattice spacing) of the nanoholes, i and j are the integer scattering or diffraction orders from the array, and εm(λ) and εd are the real parts of the wavelengthdependent relative dielectric constant of the metal and dielectric constant of the surrounding medium, respectively. Although this approximation provides a simple means to assign the various peaks with the corresponding scattering order and resonance modes, the actual peaks are red-shifted as compared to the above approximation without including the SP scattering via nanoholes and ohmic loss.126,145,147 The spectral sensitivity of EOT for periodic nanohole arrays can be related to variations in the refractive indices of the bulk solution and for the adsorption of molecules to the metal surface.148 This relation can be expressed by the following equation, S=

εm(λ) Δλ =λ× = Δn n(n2 + εm(λ))

P i2 + j2

×

⎛ ε (λ) ⎞3 ⎜ 2 m ⎟ ⎝ n + εm(λ) ⎠

where n is the refractive index of the dielectric and λ is the wavelength in vacuum. In practice, integration at zero-order transmission causes strong interaction between various SPP modes and forms multiple peaks in the transmission spectra. The lower-order SPP modes related to the metal−solution interfaces are more sensitive to RI changes and are generally selected for quantitative analysis. As the optical response of the EOT is a complex coherence of different resonance modes and the various spectral peaks associated with different structures and lattice arrangement of nanohole array, it still remains a challenge to optimize the geometrical parameters for a particular spectral spectroscopic condition in maximizing the sensing performance. Nevertheless, as expected from the SPPs resonance and sensitivity relationship, the EOT phenomenon of nanohole arrays provide a unique method for monitoring the 5403

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

with a small redshift of the main peak of about 30 nm observed as the hole size increased from 148 to 286 nm.156 Changing the hole shape influences not only the propagating properties of the hole but also the scattering to EM modes on either side of the holes.157,158 For periodic arrays of nanoholes changed from circular or square to rectangular with decreased area, the normalized transmission is enhanced by nearly 1 order of magnitude.130,146,152 This enhanced transmission for the elongated shape is mediated by a combination of localized mode associated with the transmission properties of single rectangular holes and collective effects related to the periodicity. This effect is also seen as the spectral position of the main (1,0) peak for SPP-Bloch mode shifts monotonically to longer wavelength regions as the aspect ratio of the holes is increased from a value where the periodicity mainly determines the position of the spectral maximum to a value where the localized mode governs the peak position.151 In addition to a monotonic red-shift, the normalized transmission intensity is also enhanced as a function of the increasing aspect ratio of the elongated hole. To enhance the light transmission, the metal film thickness has to be optically opaque within several-orders of the skindepth of the metal (i.e., the penetration depth for the intensity of incidence light being reduced by 1/e in metal).126,130,152,159,160 Typical skin-depths are on the order of 20 nm for noble metals in the visible regions, and film thickness on the order of 200 nm are more suitable at optical wavelengths. The optimal metal film thickness for wellcharacterized optical properties is around 10−50 nm, allowing for light transmission through the film.160−162 The metal film thickness or hole depth affects the EOT phenomena in numerous ways. For thick films approximately equal to or less than a skin depth, the resonance modes on each side of the film are strongly coupled via the hole functioning as a channel. This coupling effect decreases exponentially with increases in film thickness and the modes. In addition, localized resonances within the holes from the reflection of the waveguide mode is also observed in thick films. The film thickness also affects the scattering of the surface wave and therefore the transmission. The collective result from all these effects is that the transmission resonance reduces exponentially with film thickness, which is attributed to the exponential decay of the waveguide mode within the hole. For thicknesses less than 200 nm but greater than the skin depth, the transmission maximum becomes less sensitive to the film thickness due to the strong coupling of resonance modes. Furthermore, there is minimal enhancement in transmission for a film thickness equal to or less than 10 nm, and the enhancement level increases with increasing film thickness. No further transmission enhancement can be determined as the film is thicker than 200 nm. Due to the unique EOT properties of nanohole arrays, a plethora of nanostructures with various geometrical properties have been fabricated and their optical properties and sensing sensitivity characterized. Among various designs, the circular, square, and triangular nanoholes exhibit higher refractive index sensitivity.152 For sensing in the visible and NIR regions, the typical array periodicity ranging from 400 to 1000 nm with a hole diameter of 100−500 nm exhibits a range of sensitivity from 300 to 1500 nm/RIU, which is generally lower than that of the conventional prism-coupled SPR sensors.133,148 In spite of the lower sensitivity to RI changes, the nanohole array operated in an SP-mediated transmission mode still has great potential in sensor applications due to its smaller finite sensing

molecular layer formation and exploring surface binding mechanism of molecules. Extensive studies have focused on the dependence of EOT on numerous parameters defining the nanohole array.130,149−151 The influence of the metal properties on the EOT depends on the selected wavelength and metal material. On the basis of the transmission intensity of the EOT spectral maximum for different metals as a function of the resonant wavelength, Au and Ag sustain higher EOT (with high absolute values for the real part dielectric constant and greater penetration depth). Ag is ideally suited to obtain a high EM field in the visible region, and Au is selected for its stability and high transmission for incidence wavelength above 600 nm. The Ag film is generally coated with a thin layer of dielectric material to prevent it from oxidation. Extraordinary transmission has been observed for periodic arrays in a wide variety of materials and with more complex structures. These more complex structures such as a metal-dielectric double film, multiple layers of Al/AlN3, and metal−insulator−metal have been implemented in EOT and their optical properties still remain to be fully characterized. In addition to the effect of material properties on the EOT, the spectral characteristics of EOT are also affected by numerous parameters associated with the geometrical structure of nanohole arrays.125,130,133,151−154 These geometrical parameters include array periodicity, aperture size, the shape and aspect ratio of the holes, metal film thickness for the hole depth, symmetry of the structure, finite size effect surface quality (roughness), and imperfections in the hole shape. The periodicity affects the SPP mode governing the transmittance and the contribution of the localized mode to enhance or diminish the peak intensity by modifying the extent of resonance coupling to surface waves. Thus, several different effects on the dispersion of the EM field are manifested by changes in the position of the peaks determined by the periodicity of the array. The spectral peak (maximum transmission wavelength) shifts to a longer wavelength region as the periodicity of the array increases. However, the transmission intensity can be diminished by increasing the periodicity. The changes in the periodicity also influence the sensitivity of SPP response to variation in the refractive indices of the dielectric−metal interface. For example, it has been shown that the sensitivity is linearly increased with increases in the periodicity of the nanohole array at a longer wavelength region. The effect of the pore size was seen in the first report of EOT.125 Changes in the hole diameter can significantly affect the transmission intensity and the bandwidth of the spectral peak as shown for various reciprocal-lattice directions incident at different angles.130,146,152 As the hole diameter increases, both the intensity and the peak-width of the transmission spectral maximum increases. No single power-law scaling can be used to describe the increased transmission over a large range of hole size. A linear relationship between the hole diameter and the intensity and width of the spectral maximum only exists within a limited range of diameter for square holes and is mainly observed for hole sizes below 200 nm, while the transmission reaches a maximum and becomes saturated for hole sizes above 200 nm. In addition, no plasmon excitation is seen for holes of less than 50 nm.155 Thus, the nanoholes are typically fabricated between 50 and 200 nm in size for better plasmon coupling.150 In addition, it has also been shown that the position of the spectral maximum is affected by the hole size 5404

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Figure 4. Schematic diagram of the vertical multiple stack slab waveguide sensor chip with optical coupling. The interference fringes generated at the far-field are captured with CCD and Fourier transformed relating intensity to transverse distance (position). Molecules adsorbed onto the upper sensing waveguide cause a shift in the fringe (Δx) due to a change in the RI.

in which a silicon oxynitride sensing waveguide is vertically stacked on top of the reference waveguide and separated by a cladding layer in an alternating low and high refractive index arrangement on a silicon wafer. The reference waveguide is used as a reference for measuring the contrast phase shifts or amplitude changes in the sensing waveguide. The cladding layer between the bottom reference waveguide and top sensing waveguide is sufficient to isolate the reference waveguide from the influence of sample in contact with the sensing waveguide. 2.2.1. Optical Waveguide Principle: Far-Field Pattern of Interferometry. A plane polarized light from a He−Ne laser at 632.8 nm propagates through the optical waveguide device which excites all modes equally. As the light exits the waveguide, the modes emanating from each waveguide diffract into the far-field, where they interfere and generate Young’s interference fringes on a photodiode array shown schematically in Figure 4. The full spatial distribution of the intensity pattern is described by a set of fringes modulated in the x direction and contained within a Gaussian envelope which results from diffraction. On the basis of the theory of wave interference, the period, a (maxima and minima), of the fringe pattern generated in the far-field is determined by the distance, s, from the waveguide end to the CCD camera capturing the interference fringes and the separation distance, d, between the sensing and reference waveguide, according to a = λs/d, where λ is the wavelength (632.8 nm) of the laser. The linear relationship between the fringe position and phase is used to analyze the relative phase shifts (Δφ) in the sensor from the relative position shift of the fringe maxima or fringe spacing (Δx) through Δφ = 2πΔx/a.

area and high throughput multiplexing measurement capability. In addition, the magnitude of the wavelength shift from a smaller number of molecules in a small area have significant effects on the distribution of the plasmonic field around the nanohole, the penetration depth, and the localized field. The changes in the spectral properties and plasmon degeneracy and bandwidth may provide more insight into the location effect of bound molecules on the surface properties. However, the optical responses of the EOT in nanohole arrays arise from complex coupling effects of different resonance modes associated with different geometric parameters, and it remains a challenging task to optimize these parameters to maximize the sensing performance and the full spectral features to fully investigate molecular binding events. Nevertheless, several configurations of the EOT-based nanoplasmonic sensors have been developed to explore molecular interactions associated with biomembranes. The advances in fabrication methods enable various nanostructures and nanoparticles to be constructed into sensor chips which greatly expand the ability to manipulate the geometrical structure of biomembranes. As a result, each spot in the array can simultaneously provide a wealth of SPR information for mechanistic studies of biomolecular-membrane interactions. 2.2. Dual Polarization Interferometry (DPI)

The DPI system is an integrated planar optical waveguide interferometer that measures optical phase and contrast differences to detect changes in the optical properties of a molecular layer163 (Figure 3D). The core configuration of DPI comprises a vertical double optical path Young’s interferometer 5405

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Figure 5. Analytical solution of TM and TE polarizations using Maxwell’s equation to obtain (A) unique values of thickness and RI at each time point for the formation of an isotropic molecular thin film and (B) either RI or thickness value is fixed to obtain the birefringence at any given point during the formation of a single anisotropic molecular thin film such as a lipid bilayer.

tolerance provides advantages for the disposable stack to be inserted and removed from the optical train without the need for precision alignment. The position shift of the interference fringes is highly sensitive to the changes in the sensing surface, and thus monitoring the temporal phase and contrast changes allows the determination of concentration, affinity, and kinetics for quantitative studies of molecular interactions. However, this description of waveguide techniques does not resolve the molecular layer properties formed on a surface (i.e., a molecular layer with a given thickness and RI (density) on the waveguide surface could cause the same phase change as a layer consisting of the same molecules with a lower thickness and higher RI). As a consequence, the data obtained from this optical setup based on single mode evanescent wave methods are limited to the measurement of mass-only related changes. 2.2.2. Dual Polarization Modes. The interaction of biomolecules with a membrane involve changes in the mass accumulated on the membrane lipids and more complex and less well-characterized changes in the membrane structural properties. To resolve the properties of the molecule layer and the binding-mediated membrane structural changes using an optical biosensor, two independent orthogonal polarized fields at a fixed wavelength or two evanescent fields at different optical wavelengths are required for the experimental determination of two optogeometrical parameters [average refractive index (RI) and thickness] of a uniform, isotropic thin film. This is now possible with DPI using two modes of orthogonal polarizations: transverse electric (TE) mode, in which the electric field is parallel to the slab, and the transverse magnetic (TM) mode, in which the electric field is perpendicular to the slab.37,38 Between the light source and waveguide end-face, a ferroelectric liquid crystal (FLC) halfplate polarization rotator is fitted to switch between TE and TM polarizations which excite the waveguide structure. The high tolerance to relative movement of the incident beam is important as the FLC plate produces a small refractive displacement of the beam during the process of activating the polarizer switch, but the fringe image remains stationary during the switching process. This polarization rotator is controlled by

The phase changes of interest, (Δφ), involve changes in the effective refractive index of the mode in the sensing waveguide. The effective refractive index of the reference waveguide is not affected by changes in the upper sensing waveguide since the evanescent field decays rapidly in the region of the cladding layer between the two waveguides. The phase difference is given by Δφ = 2πLΔNs/λ, where L in the interaction optical path length and ΔNs is the effective refractive index change in the upper sensing waveguide mode. The positional shift of the interference fringes (Δφ) can be analyzed by a fast Fourier transformation relating intensity to position. The direct measurement of Δφ therefore allows direct determination of the effective refractive index of the sensing waveguide. In addition to the phase changes of interference fringes, the fringe contrast, the intensity difference between the maxima (peak apex), and the minima (peak valley) of the fringes are used for measuring the difference in the light intensity propagating through the sensing waveguide, which is influenced by any losses from light scattering or adsorption molecules on the sensing surface. Contrast loss provides an alternative method to calculate the mass of light absorbing molecules and for characterizing crystallization processes. Adsorption to and desorption of molecules from the sensing waveguide interferes with the evanescent field of the light and alters the complex propagation constant of the sensing waveguide modes causing phase and contrast changes. Interferometers offer significant advantages over angular measurements such as those made in SPR-based sensors providing an exceedingly stable measurement platform with both high sensitivity 10−7 refractive index unit (RIU), 0.1pg/ mm2, and resolution below 0.1 Å. The stability of the measurement platform is achieved by the provision of a highfidelity reference signal via the reference waveguide. This allows any minor deviations in the local environment or the output radians to be compensated. Furthermore, the optical setup in the interferometer is highly tolerable to macroscopic movements of the input coupling beam on the order of hundreds of micrometers, causing no change in the interference pattern. This level of optical 5406

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Most significantly in the context of biomembrane applications, the method also fails to resolve molecules self-assembled into layers with uniaxial anisotropic properties such as lipid bilayers. The optical properties of molecular layers depend on the underlying structural arrangement. In the case of a lipid bilayer, the two-dimensional alignment of the cylindrical or rod-likeshaped molecules results in an asymmetric polarizability, and the response of lipid molecules to an applied electric field is termed anisotropy. DPI allows this anisotropy to be measured and exploited by studying how the electric field interacts with the membrane in real time. For anisotropic materials, polarized light experiences different refractive indices along the vertical and horizontal optical axis in the plane of the object. Thus, the optically anisotropic properties of the lipid bilayer with a uniaxial optical axis have two unequal principal refractive indices for two different polarization planes. The difference between the perpendicular and parallel refractive index for the lipid bilayers is defined as the birefringence (Δnf) and

a digital signal processor (DSP) typically switching at 50 Hz. The diffraction fringes emanating from the dual slab waveguide are captured by a 1024 × 1024-element imaging device camera. The captured images of interference fringes are sent to a DSP, and the relative change in fringe position (phase) is updated every 20 ms using a spatial Fourier transform algorithm.36,164 These processes are repeated, and output as real-time changes in the TE and TM polarization mode throughout the course of adlayer formation and molecular interaction. The TE and TM polarizations behave differently within the waveguide and hence the evanescent field profiles are slightly different, and therefore, two different phase shifts can be obtained for the molecular layer formed on the surface. In particular, the TE mode is more confined to the immediate waveguide surface than the TM mode which is more extended into the bulk. Therefore, the TE mode is more sensitive to the changes in the refractive index and density of the sensing layer. Those changes in phases and contrasts in each polarization are solved with Maxwell’s equations to the multilayer dielectric continuum electrodynamic model that fit the experimental data to a uniform layer model. This allows inclusion of an arbitrary number of layers in a model, each of which is represented by its own layer matrix. Having solved Maxwell’s equation for each polarization, a different range of thickness and refractive index can be matched to the phases obtained from each TM and TE mode (Figure 5A). When the TM and TE solution ranges of thickness and RI values are superimposed, these two ranges intersect at a unique point where a set of thickness and refractive index corresponds to the actual optogeometrical properties of the molecular layer at a particular time (t). This determination of the geometrical parameters for a molecular layer assumes the layers are isotropic in nature. The initial waveguide thickness and refractive index prior to forming a molecular layer can be obtained from calibrating the waveguide using two solutions with known RIs as the upper layer. Thus, the structural parameters of molecular layers can be obtained throughout the adsorption process. The RI and thickness values can be used to calculate the total mass, surface concentration, and number of molecules for the molecular footprint on the surface. 2.2.3. Anisotropy and Birefringence. Obtaining both the thickness and refractive index (density) provides valuable information on the geometrical structures of the molecular layer. However, there are situations where the method described above fails to resolve the phases and errors or anomalous values of the thickness and RI are obtained. This outcome can be a result of sparse distribution of molecules adsorbed on the surface that leads to an underestimated thickness and overestimated refractive index when the surface coverage is below a threshold (∼17% coverage).164 The analysis thus assumes that the molecules form a continuous, uniform layer on the surface and that the material responds the same in every direction, that is, the surface is isotropic. When this applied layer model does not fit the real density (RI) distribution of the measured molecules, layer thickness can become underestimated.164−166 This can be a result of a number of factors including (a) the molecules are adsorbed on one end of the surface resulting in a gradient distribution to the other end, (b) molecules are preferentially adsorbed to a particular area, forming patches or domains of various size on the surface, or (c) the molecules undergo conformational changes differently in response to different contact region and area on the surface, resulting in a heterogeneous layer density.

Δn f = ne − no

where ne, the extraordinary RI, is the RI perpendicular to the bilayer, and no, the ordinary RI, is the RI parallel to the lipid bilayer.167,168 As an electromagnetic wave passes through the lipid membrane in a direction parallel to the normal, the index of refraction along perpendicular directions to the propagating wave are related to the projections of the acyl chains. As a result, the orthogonal polarizations propagating through the lipid bilayer experience an effective phase shift between the two components of the wave that is termed the retardance (σ). Measurement of the birefringence can be further related to the effective retardance by σ = 2πtΔnf/λ, a parameter that is a function of acyl chain tilt, thickness of the membrane (t), and differences in indices of refractions along the radial and length directions of the lipid molecules. Thus, the optical birefringence of lipid bilayers can be directly related to molecular properties such as polarizability, shape, and orientation. The anisotropy of the bilayer increases as the bilayer becomes more ordered, and the degree of birefringence is a quantitative measurement of this order or lack thereof. Birefringence therefore provides a direct measurement of the quality of the bilayer itself and also changes in the bilayer structure as it undergoes phase transitions or interactions with other molecules such as proteins, polymers, or ions. To resolve the anisotropic layers with DPI, the method requires complementary data to fix either the RI or thickness in order to get a unique solution for a pair of thickness and birefringence values or a pair of RI and birefringence values (Figure 5B). The two different RI, nTM and nTE, associated with two polarization modes TM and TE can be directly measured by DPI where n TM =

ne 2 sin 2 θ + no 2 cos2 θ

n TE = no

and θ is the angle of the lipid molecules with respect to the normal. For lipids fully aligned perpendicular to the surface normal, the angle is 0°. The birefringence values represent an average measurement of the degree of alignment and density of lipid molecules and are highly sensitive to small changes in these parameters. The effective birefringence (nTM − nTE) is determined by fixing the RI of the deposited unilamellar lipid bilayer to 1.47 indicative of a uniform layer complete coverage or by fixing the thickness of bilayer with known values (e.g., 4.6 nm for DMPC bilayer at 20 °C).169 The birefringence and mass 5407

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

A summary of the main assumptions are as follows. (1) The refractive index is linearly dependent on the molecular concentration in solution. (2) The molecules bind homogeneously on the surface without deformation and without strong intermolecular interactions that lead to aggregation. (3) The molecules form a uniform isotropic layer with even density throughout the optical sensing depth. (4) The molecules are highly pure and of single composition. (5) Low surface roughness. Limitations that need to be considered are as follows. (1) Sample size is beyond the evanescent field as in the case of lipid microvesicles and exosomes carrying specific biomarkers. (2) Heterogeneous size distribution, such as molecule aggregates, exosomes, and biopolymers. (3) Complex compositions of different molecular species with uncertainties of dn/dc values such as exosomes are composed of a mixture of lipid, proteins, and RNAs, for which an approximation of dn/dc value is required. (4) The dn/dc values are varied with buffer composition and wavelength. (5) Deformation or unfolding of molecules throughout the process of binding and formation of the layer. (6) Molecules bind to different depth of the preadsorbed layer, such as polymer brush border and tethered liposomes. (7) Molecule layers have substantial birefringence. (8) Validation of dn/dc values and multilayer models for complex film structures. (9) Surface roughness as experienced in the nanoparticle immobilized substrate of LSPR hinders the use of these methods for calculating mass from the layer thickness and RI. Thus, calibrations with solutions of known RI are required for mass determination. In the context of estimating the mass of molecules bound to the surface of an optical sensor in relation to the changes in refractive index, de Feijter extended McCrackin’s formula to determine the mass per unit area for a film containing molecules with a single refractive index increment (dn/dc). The common use of the de Feijter formula in mass calculations assumes the formation of an isotropic uniform layer on the surface on which the molecules distribute homogeneously. The formation of discontinuous layers with isolated islands and/or multilayers with varied density will influence the mass derived for SPR more than for DPI. This is because in DPI, the average film thickness and refractive index are determined compared to the resonant angle shift detected from a small area in SPR. The mass calculation based on the de Feijter formula also depends on accurate dn/dc values (i.e., a linear dependence of RI changes with the molecule concentration).78 Providing the molecules are pure and of sufficient quantity, dn/dc can be readily measured for any specific combination of analytes and buffer using a benchtop refractometer. Prediction of dn/dc values based on the amino acid composition in a bioinformatics approach to a set of predicted proteins from the genome of human and other species has also been developed for proteins which are difficult to purify and obtain in large quantity.170 The de Feijter formula can also be extended for use with films of complex molecular composition with different dn/dc given that the relative fraction of each component is well-defined and characterized. However, the relative compositions of different molecules in natural membranes directly from cells are difficult to fully characterize, and the fraction of different molecules in a membrane is estimated from the buoyancy measurement. This has been shown for estimating the dn/dc values of exosomes composed of proteins, lipids, and RNA.171 With the SPR signal, it has been shown that the use of dn/dc to calculate the mass of exosomes binding to a surface and supported lipids bilayers

of the molecules can be determined simultaneously by DPI, and a typical bilayer has a birefringence of 0.010−0.025 refractive index units. DPI can also measure refractive index increments with high sensitivity and can therefore characterize very subtle changes in the molecular organization of a membrane, revealing mechanisms of molecule-membrane interaction that are difficult if not impossible to see by other means. 2.2.4. Optogeometrical Parameters. The de Feijter formula77 is used to calculate the mass of adsorbed lipid and peptide layers from the DPI measurements. The mass per unit area of the lipid bilayer (ml) and the mass per unit area of the peptides/proteins (mp) binding to the chip are calculated using ml = df (n iso − nbuffer)/(dn/dc)l mp = df (n iso − nbuffer)/(dn/dc)p

where df is the thickness of the bilayer, niso is the average isotropic RI of the attached bilayer, nbuffer is the RI of the bulk buffer, and (dn/dc)l is the specific RI increment of the attached bilayer, whereas (dn/dc)p is the specific RI increment of the peptide/proteins. The de Feijter formula assumes that (dn/dc) remains constant throughout the experiment. For the analysis, the (dn/dc) values of 0.135 and 0.182 mL/g are used for lipids and peptides/proteins, respectively. Overall, DPI allows a single lipid bilayer system to be characterized in terms of a range of dynamic structural parameters that include thickness, mass per unit area, surface area per lipid molecule, and molecular orientational order (birefringence or anisotropy). Most significantly, the ability to accurately measure the orientational order of lipid molecules is a distinctive feature of DPI that allows detection and tracking of changes in the structure of the lipid bilayer. The flow-through system also allows kinetic analysis of the peptide/protein binding and membrane disruption processes in real time. Since the birefringence quantifies the degree of alignment and uniaxial packing of the lipid molecules on the planar surface, changes in birefringence as a function of peptide binding to the membrane provides new insight into the mechanism of binding, and the rate of concentration-dependent changes in lipid packing, phase transition for domain formation, and membrane destabilization. 2.2.5. Assumptions and Limitations of Calculation of Mass. Interpretation of the SPR response is generally achieved by multiplication of the resonant angle shift expressed as response units (RU) with a proportionality constant unique for the adsorbed species and calibrated independently by, for example, radiolabeling. The most commonly used relationship is 1RU corresponds to 1 pg/mm2 for most biomolecules. However, this is only valid for molecules that bind homogeneously into a uniform layer throughout the SPR sensing depth and do not adsorb light at the incident wavelength. Alternatively, the optical thickness of the adlayer in a multilayer model can also be used as a measure of the adsorption and for mass calculation, in particular, when an appropriate calibration factor to determine from the SPR angle shift is not known, such as studying new biointerface interactions. The layer thickness can be solved from the complete reflectivity versus incident angle spectrum using Maxwell’s equations. However, an assumed RI for the adlayer is used for this method. If the assumptions are reasonable, a good approximation of the adsorbed mass can be obtained using the de Feijter formula. 5408

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

(SLBs) during the vesicle-to-bilayer transformation has resulted in an underestimation of the vesicle mass by as much as 30%.171,172 This is due to the different dn/dc values for liposomes and supported lipid bilayers (SLBs), since it is difficult to determine directly. However, this is valid for the mass of final SLBs. Furthermore, the dn/dc value will vary depending on lipid composition and deformation of adsorbed liposomes and ordering of lipids in the SLB. In addition, the size of lipid vesicles and exosomes in a similar order of the evanescent field is also a substantial fraction of errors. Thus, for these particular cases of natural membranes and exosomes, the commonly used de Feijter formula to estimate mass uptake based on the RI changes can lead to significant qualitative and quantitative errors. In cases of uncertainty in determining the dn/dc values, the adsorbed mass can also be determined from the RI and thickness of an adsorbed layer with the approximation of Lorentz−Lorenz relation for the dependence of layer density on the refractive index, as demonstrated for the quantitation of prothrombin binding to phosphatidylserine bilayers by ellipsometry.173 The mass per unit area of molecule adlayers formed on the surface can be estimated from the layer thickness (df) and density (ρ) according to m = df × ρ =

f (n) =

0.3df × f (n) A M

nb2 − 1

( )

−υ

nb2

nav 2 =

n p2 + 2ns 2 3

where the np and ns are the refractive index of the p-polarization and s-polarization, respectively, which are obtained directly from fitting the resonance spectra. Furthermore, complex modeling has also been developed to account for an inhomogeneous layer with uneven density distribution in mass determination with DPI.166

3. MODEL MEMBRANE SYSTEMS Native biomembranes are highly complex in their chemical composition, physical phase, and structural properties. While the well-established DNA recombinant technologies have provided efficient ways to manipulate gene and protein functions in vivo, lipids are not genetically encoded and these approaches cannot be applied to the manipulation of the structure and composition of lipids and physical properties of membranes. To specifically alter the lipid composition, the genes encoding enzymes along the whole biosynthetic pathways for particular lipid structures need to be targeted. These genetic mutations can cause other indirect changes and may not target the primary membrane site. In addition, altering the lipid composition via gene manipulation may induce lethal changes in membrane stability and permeability before the exact molecular role of specific lipids in membrane properties is revealed. Nevertheless, combining in vivo genetic approaches and in vitro artificial biomembrane systems has proven useful in defining lipid function to unravel the molecular basis for the phenotypic changes of a membrane in biological and pathological states. Due to the heterogeneous molecular structures and relative amounts and distribution of lipids in natural biomembranes, comprehensive information on the roles of individual lipid structure in physical properties and functions of membranes and on the structure−activity of peptides/ proteins are commonly obtained through the use of artificial model membrane systems containing lipids, proteins, carbohydrates, and other organic and inorganic species of interest. The bottom-up approaches are widely used for the development of artificial model membrane systems which involve the reconstitution of synthetic or naturally isolated lipids where the concentration, composition, mechanical, and geometrical properties can be well-controlled for specific applications and characterizing techniques. While the lipids from natural cells closely resemble the native membrane, pure synthetic lipids provide significant advantages in understanding the relationships between chemical structure, polymorphic behavior, and dynamic properties of the lipid membrane and the influence of environmental factors, such as temperature, pH, ionic strength, hydration level, and organic solvent on the membrane properties. A broad range of model membrane systems have been developed on various supports of different materials adapted to optical sensing and spectroscopic applications. These model membranes supported on solid substrates have been valuable tools for characterizing the structural properties of membranes in relation to compositional changes and environmental stimuli. These systems are also used to simulate and monitor moleculemembrane interactions for various cell membrane-mediated processes with high reproducibility and high sensitivity. The design and development of model membranes strongly depends on the material nature, physical properties, architectures, and geometrical dimensions of the substrates. Even with

× ( n − n b)

+2

n + nb 2

(n + 2) ·(nb2 + 2)

From the formula above, it follows that the molecular weight (M), the molar refractivity (A), and the partial specific volume (ν) of the adsorbed or stacked molecular species has to be known in order to obtain mass (m) from the thickness (df) and refractive index (n), where n is the refractive index of the adlayer and nb is the bulk refractive index. This method was also validated with known mass of lipid bilayers and using radioisotope labeled proteins using either one or two component analysis. Both the de Feijter formula and Cuypers formula were used to calculate the mass of lipase adsorbed onto a C18-modified surface measured by DPI, and the mass values calculated from both formulas were also compared to those obtained from SPR based on 1RU is equivalent to 1 pg/mm2.174 The use of the Cuypers method led to a higher value of adsorbed mass than that calculated by the de Feijter formula for every protein concentration. However, the calculated mass using either method in DPI were in good agreement with the adsorption isotherm from SPR measurements. In a DPI measurement, a minimum of 17% coverage is required to give a valid measurement of the thickness, since if the molecules adsorbed too sparsely, each of the polarizations propagating through the sensing waveguide senses insufficient molecules for mass determination. The Lorentz−Lorenz relation was further modified for the influence of anisotropy on the interpretation of dn/dc and mass calculation of anisotropic adlayers.168 The mass can be calculated from the layer thickness (df) and average refractive index (nav) as follows: ⎛ n 2 − 1⎞ ⎛M⎞ ⎟ m = 0.1⎜ ⎟ × df × ⎜ av 2 ⎝A⎠ ⎝ nav + 2 ⎠ 5409

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

physisorbed/chemisorbed onto planar substrates, and (3) lipid bilayers formed on the surface of nanostructured substrates such as corrugated, well/hole, and micro/nanoparticles architectures. Model membranes fabricated on substrates with different architectures are complementary to each other in their morphologies and structural organization since the assembly of common amphipathic lipids and other related compounds can be finely tuned to these structural configurations. Thus, integration of both the planar and spherical configuration with numerous surface-sensitive technologies provides a comprehensive analysis of membrane curvature and related structural properties associated with molecule-membrane interactions. Although the commonly used spherical and planar configurations of model membranes have greatly facilitated the characterization and understanding of physicochemical properties of membranes modulated by the substrate and in studying molecule-membrane interactions, biomembranes can also form complex structures with specific roles in modulating the membrane phase separation and protein interactions. Rough sheets, smooth sheets, smooth tubular, multilayer stacks, and folded, bent, and cubic shapes are also characteristic structures found in the membranes of smooth/rough ER, Golgi apparatus, and other intracellular organelles and during cell locomotion, division, and vesicle transport.13 With advances in micro/nanofabrication technologies, membranes formed on the substrates with tubular, well/hole, nanocube, nanodisc, and nonspherical nanostructure configurations impose significant geometrical constraints on the physical nature of the membrane organizations. The formation and physical properties of these nonplanar structured lipid bilayers on the solid substrates are not only determined by their compositions but also dependent on the geometrical structures of the underlying substrate. Therefore, formation of bilayers on supports with different configurations by engineering the substrate provides a unique approach to manipulate the structure and interfacial properties in a defined manner and plays a central role in advancing our understanding of the regulation of protein activity through remodelling of the geometrical structure and physical phases of a membrane. The formation, functionalization, and application of artificial biomembrane models in various formats are described in the following sections. Understanding the quantitative-structure relationships of lipid molecules is highly challenging due to the molecular complexity and technical complications associated with the native membrane. Several advances in model membrane systems have been progressed from simple format with low sensitivity at low throughput to fast, high-throughput, multiplexed array formats with three-dimensional higher-order structure. The integration of various model membranes with flow-over or flow-through channel structures is also important for kinetic studies of molecule-membrane interactions. Since direct contact of bilayers with the surface can restrict the binding-embedding of proteins in the membrane, decoupling of the membrane from the substrate while maintaining it close to the surface is also critical for biosensing technology. Model membranes composed of one or few structurally defined lipids are commonly used to probe the particular physical properties for membrane function. However, it still remains challenging to characterize the physical properties of model membranes composed of complex lipid species.

the simple model system composed of one or few lipid components, there is little understanding of how the physical properties of the simple model membranes measured in vitro relate to their function in vivo. In addition, not all physical properties of lipid bilayers have been fully characterized in many in vitro experiments. The observed changes in membrane properties as a consequence of molecular interaction are generally linked to the structural effect of particular interacting molecules such as peptides or proteins rather than the result of changes in the overall membrane properties. Stability and dynamic fluid properties are two contradictory features in a membrane and need to be well-balanced in the fabrication of SLB systems. Indeed, a membrane bilayer is a self-assembled colloidal system that lacks long-term stability suitable for storage, and significant adhesion force is required to stabilize the macroscopic structure onto solid substrates. Strong and stable interaction of lipid bilayers with solid support surfaces can also enhance the sensitivity of biophysical analytical methods, particularly for the surface sensitive SPR and DPI. While it is crucial to maintain dynamic properties of lipids for optimal performance of the functional units in the model membrane systems, the dynamic fluidity of lipids in the membrane can be severely compromised by the restricted lipid motion and bilayer defect from strong interaction. Membranes are also highly cooperative systems with very effective amplification mechanisms in response to external stimuli. This high degree of molecular cooperativity is of particular importance in characterizing the changes in molecular order and structures induced by phase separation and molecule interaction. Therefore, finding an acceptable compromise between the two essential but contradictory features demands a good understanding of the properties and modifications of the solid substrates, the structural properties of model membrane, and detailed knowledge of the relevant biochemical and physical properties of the SLB systems. Major applications of model membranes involve the following. (A) Biophysical characterization: physical properties (fluidity, phase changes, molecular order in relation to the chemical composition and environmental factors of pH, salt, temperature, and mechanical stress), and formation mechanisms. (B) Functional studies: ligand-membrane receptor interaction, mechanistic binding of biomolecules (peptides, proteins, and drugs), channel regulation (ion, mechanical, and voltaged-gated), membrane fusion, fission, and membrane repair. (C) Detection and screening: small molecules, drugs, nanoparticles, and toxic compounds, etc. (D) Biomaterial engineering: surface guided cell growth, antifouling. (E) Energy transduction: biofuel production, energy conversion (photosynthesis thylakoid membranes, respiratory enzymes, electron transport chain, ATP synthase, and biocapacitor). Depending on the types of application, the platforms of model membranes and the associated analytical methods can vary quite dramatically. The model membrane systems mimicking the complex nature of biomembrane range from simple models of a single phospholipid species to very complex platforms fabricated with various lipids, proteins, and carbohydrates and other biomolecule-conjugated lipids. In general, the generic configurations of widely used model membranes can be classified into (1) spherical vesicles suspended in solution or coated on the porous/nonporous spherical particles and attached onto the solid substrate, (2) planar lipid bilayers free-suspended over apertures with diameters ranging from micrometer to nanometer and 5410

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Figure 6. Versatile lipid bilayers and monolayers that can be supported on planar substrates via physisorption of a bilayer onto (A) an unmodified substrate or (B and C) modified substrates, (D) physisorption of a monolayer onto a modified substrate, (E and F) direct attachment of a monolayer or bilayer directly onto solid substrates through reactive groups coupled either at the headgroup or at the ω-end of the acyl chain, (G) polymercushioned planar bilayer, (H, I, and J) covalent tethering of a lipid monolayer or bilayer onto a modified hydrophilic surface, and (K) adsorption of multilamellar bilayer structures.

Figure 6. In the first type of solid-supported membrane, lipid bilayers can be noncovalently adsorbed (physisorbed) onto hydrophilic surfaces of unmodified substrates (glass, mica, and silica) (Figure 6A) and substrates modified with hydrophilic polymers and self-assembled monolayers (SAMs) terminated with −COOH and − OH (Figure 6, panels B and C). Lipid monolayers have also been adsorbed onto a hydrophobic SAM of either short or long alkyl chains covalently attached to the substrate to form a hybrid lipid bilayer (HLBs) (Figure 6D). The third type of membrane system is the covalently immobilized/anchored lipid monolayer or bilayer on chemically modified hydrophilic surfaces (Figure 6, panels H, I, and J). The control of the physical properties of SLBs in a defined manner remains a challenging task. Various methods have been developed for preparing lipid bilayers on solid supports. These methods typically involve Langmuir−Blodgett (LB)/ Lang-

3.1. Supported Lipid Bilayers (SLBs)

3.1.1. Physisorbed Lipid Bilayers. The formation of planar biomembranes on either unmodified or chemically modified substrates of different material properties are most compatible with the platform configurations commonly developed in sensing systems. The substrates used to construct integrated interferometers for biosensing such as silicon oxynitride, silicon oxide, silicon nitride, and to a lesser extent polymers, allow the direct adsorption of a membrane, while modification of metallic plasmonic sensing substrates with dielectric materials or monolayers of organic molecules assist membrane adsorption. The planar supported lipid membranes are broadly classified into several different categories according to the nature of the substrate surfaces and types of membranesubstrate interactions which are illustrated schematically in 5411

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

muir−Schaefer (LS) deposition,175−177 microcontact printing,178 spin-coating,179 lipid dip-pen nanolithography,180,181 air bubble collapse deposition,182 evaporation-induced assembly,183 lipid/detergent mixed micelles, 184,185 vesicle fusion,186,187 solvent-assisted lipid bilayer formation,188,189 and the droplet interface bilayer method.190,191 Since the early development of molecular monolayers or multilayers supported on solid substrates, the so-called “Langmuir-Blodgett (LB) films” have been widely used to characterize the physicochemical properties of various molecules on the surfaces of solid materials,192 to fabricate new materials with novel surface properties,193,194 and to fabricate model membrane systems,176,177 as well as for the application of membranes in electrochemical biosensors195−198 and structural characterization of membranes with various spectroscopic techniques.176,199−201 The preparation of lipid bilayers with the Langmuir−Blodgett technique involves moving a clean plate in a vertical or horizontal orientation (the solid substrate) through an aqueous solution−air interface resulting in the transfer of the lipid monolayer onto the planar solid support during immersion or withdrawal, and the lipid films can be removed with organic solvents or aqueous detergent solutions. The orientation of lipid molecules in LB films is dependent on the surface properties of the solid substrate. For hydrophobic substrates, the tails of lipids are preferentially adsorbed onto the surface of a solid substrate, and the monolayer is transferred during dipping. In contrast, the head groups of lipid molecules are adsorbed onto polar substrate surfaces and monolayer transfer preferentially occurs during the withdrawal process. The organization, surface density, and number of lipid layers can be controlled by repeated withdrawal and dipping of a hydrophilic substrate through the monolayer and leads to the formation of an alternative head−tail-tail−head multilayer LB film. Such tail-to-tail hydrophobic interactions and head-tohead electrostatic interactions greatly increase the stability of an LB film. Thus, different lipid compositions in each leaflet of a multibilayer can be constructed by alternatively changing the lipid composition during the preparation of the Langmuir monolayer. However, difficulties have been reported in the deposition of the second monolayer in which the first monolayer is retransferred back onto the surface of the Langmuir trough when the covered substrates are immersed in the subphase.202,203 The effectiveness of monolayer adsorption and assembly onto a substrate relies on the preparation of a stable, well-compressed and homogeneous monolayer on the Langmuir trough. Additionally, a constant, steady, and slow (on the order of millimeters per second) speed is required to achieve a stable monolayer transfer to the solid substrate, and above a critical transfer velocity no monolayer can be formed. Controlled evaporation of organic solvent from phospholipids in organic solution has been used as an alternative method to LB films for the formation of ultrathin films, which retain the regular bilayer structure of biomembranes. For a multibilayer membrane, each of the stacked bilayers on the solid surface is separated by a water layer about 1−2.5 nm in thickness. These oriented lipid multibilayers have been frequently used in polarized transmission and attenuated total reflection (ATR)IR studies to provide information on the structure and organization of pure lipid bilayers in response to the binding of molecules.204 The formation of single bilayers on hydrophilic supports can be achieved by incubation of the lipid vesicle dispersions on

substrates42,175 or by dialysis of a detergent solution of the lipid in the presence of the substrate materials.205 Several experimental parameters must be tested and optimized for a specific lipid composition (even with a single lipid species) to obtain a defect-free bilayer on the solid support. These parameters are generally associated with the surface properties of substrates, liposome properties, and the bulk solution properties and conditions. The success and ease of forming a bilayer structure from liposome adsorption onto solid supports is highly dependent on fine-tuning a combination of factors, including material properties, structure (including roughness) and cleanliness of the substrate, composition, size, charge, and concentration of the liposomes, composition of the surrounding medium, temperature, the protein content, and the geometry of and flow dynamics in the fusion cell. Fully fluid supported bilayers by vesicle fusion are known to form on highly hydrophilic mica, silica solid surfaces (i.e., glass, quartz, thermally grown silicon dioxide, and sputtered silicon dioxide), and oxidized poly(dimethylsiloxane) (PDMS) which displays a hydrophilic SiOx surface. However, vesicles adsorbed intact onto a SiO2 surface when the amount of liposomes was below the critical surface coverage.186 In contrast, exposure of lipid vesicles to other hydrophilic solid surfaces resulted in lipid structures other than a fully fluid bilayer in which an intact vesicular layer was formed on oxidized gold, oxidized platinum, or titanium oxide.206−210 In addition to the planar solid substrate, a membrane bilayer can also form on spherical substrates such as polystyrenedivinylbenzene (PSDB), glass, and silica spherical beads.211−213 The dynamic properties of the lipid molecules were retained in a single bilayer of either a single species or a binary mixture of phospholipids adsorbed onto silica particles (pore size 4000 Å, diameter 10 ± 1.5 μm).214 Transition endotherms measured by differential scanning microcalorimetry, revealed that the lipid bilayer systems that consisted of two lipid species on the spherical support exhibited a rather broad and complex feature consistent with partial demixing of two lipids in the gel phase. In contrast, the endotherm for adsorbed bilayers with a single lipid species showed specific heat changes consistent with the presence of a single phospholipid species. In addition, the transition temperatures for the single and binary lipid bilayers were also different. These unique membrane-like thermodynamic properties were then exploited to provide a selective purification system of the peripheral membrane protein, myristoylated alanine-rich C kinase substrate (MARCKS)related protein.215 This bilayer-adsorbed spherical solid support thus provides charge-selective protein separation under biocompatible conditions solely by changes in the column temperature without changes in ionic strength, which allowed the native conformation of the protein to be maintained during the purification. In addition to the adsorbed bilayer, the perfusion of phospholipids dissolved in organic solvent (85% isopropanol) into reversed phase chromatographic supports resulted in a hybrid membrane bilayer in which a monolayer of lipids is noncovalently adsorbed onto the hydrophobic alkyl chains.216 The amount of adsorbed phospholipids was found to be similar to that of lipid vesicles, thereby confirming the formation of a bilayerlike structure. This system was stable in solvents containing less than 35% acetonitrile and was used to study peptide-lipid binding properties For lipid monolayer covered substrates, the predeposited lipid monolayer provides a very stable surface when subjected to harsh conditions. The liposomes that provide the outer lipid 5412

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

chromatographic supports have also been covalently immobilized with ω-end modified phospholipids to form an immobilized lipid monolayer with the head groups exposed to the aqueous environment. The reaction of carboxyl groups attached at the ω-end of phospholipids with an aminosilane on the silica surface resulted in the formation of an immobilized lipid monolayer on the surface of chromatographic supports.221 In addition to the use of carboxyl groups as a functionalized moiety, the synthesis of a new class of immobilizable glycerophospholipids containing amino groups at the ωterminus of each of the sn-1 and sn-2 acyl chains has also been reported.222,223 In addition to the immobilizable phosphatidylcholine, other lipid analogues, such as phosphatidylglycerol, phosphatidylethanolamine, phosphatidylserine, and phosphatidic acids with different charges, size, and hydration properties have also been synthesized.32,224,225 The methods used for the preparation of the immobilized glycerophospholipid monolayer generally consist of three sequential steps: (i) the activation step in which the silica is pretreated and subsequently reacted with an organosilane with defined functionality, whereby it is covalently bound to the silica, (ii) the coupling step in which the activated silica is subjected to reaction with the biomimetic ligands under gentle conditions to couple the ligands, and (iii) the end-blocking step in which the residual activating groups from the coupling steps are removed by means of appropriate reactions. Thus, immobilization of lipid analogues, either as a single species or a mixture in different proportions, onto the activated spherical chromatographic supports therefore provide a series of model biomembranes which can closely mimic the lipid compositions found in naturally occurring membranes. The covalently immobilized lipid monolayers on a silica substrate exhibited long-term stability under aqueous conditions and also in the presence of organic solvent.221−223 In addition, the types of salts, ionic strength, pH, and organic solvent composition in the mobile phase can be optimized to provide baseline separation of drugs and membrane protein mixtures.226 As the residual reactive groups on the surfaces are blocked with small chemical compounds, the leaching of lipids from the column is greatly diminished under slightly acidic conditions and when stored in detergent solution for long periods of time. However, perfusion of an immobilized biomembrane column with 0.1% trifluoroacetic acid (TFA)/H2O, 0.1% TFA/acetonitrile, 35 mM citric acid, 35 mM NH4H2PO4, and 35 mM NH4Cl resulted in approximately 2% of the immobilized lipid molecules leaching from the column. Although the lipid density on the silica particles was similar to that found in natural biomembranes, an important disadvantage associated with the immobilized lipid monolayer is the lack of lipid dynamics, including lateral diffusion, flip-flop, and axial displacement, due to the covalent linkage of the lipid molecules to the silica surface. However, this lack of lipid dynamics is compensated by the increased stability derived from the covalent immobilization of lipids onto solid surfaces. Thus, the immobilized phospholipid monolayers on solid supports have unique applications compared to the immobilized liposomes. In particular, organic solvents and detergents can be included in the buffer for screening drug binding, examining peptide/protein−lipid interactions, and separating membrane-associated proteins. 3.1.3. Polymer-Cushioned Planar Bilayers. The adsorption of lipid bilayers directly onto a substrate provides very stable membrane systems. The ultrathin water layer of 1−2.5

layer must make contact with the alkyl chains of the deposited monolayer. The underlying solid support is important for the process, with platinum (Pt) generally resulting in better transfers than Si/SiO2.217 Liposome fusion onto bare and lipid-coated Pt results in membranes with similar characteristics.218 Liposomes prepared from unsaturated phosphatidylcholines such as POPC or DOPC are most frequently used for forming supported planar bilayer structures, as they are in the liquid crystalline state at room temperature, facilitating rapid spreading and formation of a fluid membrane on different types of support. By comparison, the unsaturated phosphatidylcholines are inferior to DPPC for obtaining membranes with the required properties. A mismatch due to the differences in area occupied per molecule in the densely packed monolayer and the much greater area occupied by each unsaturated phosphatidylcholine in the liposomes might be a reason for an inferior hydrophobic interaction, with consequences for a successful transfer. Divalent cations have also been found to promote attachment of mixed phosphatidylcholine/phosphatidylglycerol vesicles to form supported vesicle layers and bilayer formation from pure phosphatidylcholine vesicles. Adding a high concentration of calcium ions can promote lipid transfer from the liposomes to plain hydrophilic supports and simplify the adsorption pattern. Other “intrinsic” fusogenic agents such as protein and a high DPPE/cholesterol ratio is often also used to promote the transfer of lipids and bilayer formation to the solid surfaces. The combination of LB films and vesicle fusion provides an extra dimension for the preparation of a fluid bilayer on a hydrophilic support.211 In addition to the planar bilayer systems, a single phospholipid bilayer adsorbed onto spherical particles, such as microglass beads (0.3−10 mm in diameter) and polysaccharide nanoparticles (30−75 nm in diameter), have also been developed, and the dynamic properties of the bilayer also characterized by 31P-, 2H NMR, and fluorescence spectroscopy.211,219 The single phospholipid bilayers on hydrophilic supports can be fully immersed in water or physiological buffers. The adsorption or fusion approaches are advantageous for their ease of vesicle preparation and the incorporation of transmembrane proteins from proteoliposomes. The thin film of a bilayer membrane adsorbed on optically transparent substrates such as glass, mica, and quartz is ideally suited for detailed structural studies of membrane structure, phase changes, and specific membrane-associated proteins and peptides via neutron/X-ray diffraction, electron microscopy, circular dichroism, and FT-IR spectroscopy. 3.1.2. Covalently Immobilized Lipid Membranes (Chemisorbed/Tethered). Membrane lipids can be covalently attached directly onto solid substrates through reactive groups coupled either at the headgroup or at the ω-end of the acyl chain (Figure 6, panels E and F). Chemisorption of phosphatidylcholine consisting of thiol groups or heterocyclic disulfide groups at the terminus of fatty acyl chains form a stable phospholipid monolayer on gold with the choline head groups extended outward from the surface.220 As studied by a combination of ellipsometry, contact angle measurements, heterogeneous electron-transfer properties, and X-ray photoelectron spectroscopy, a monolayer consisting of thiol-coupled phosphatidylcholine molecules showed very similar surface properties to biomembranes. However, the liquid-crystalline properties of the monolayer have not been demonstrated in such a lipid monolayer system. Fused silica particles such as 5413

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Figure 7. Schematic of various approaches to prepare polymer-supported lipid mono/bilayers. Adapted with permission from ref 250. Copyright 1999 Elsevier.

incorporated molecules were also demonstrated by the remarkable long-range lateral mobility of lipopeptides embedded into a polymer-supported bilayer.234 The ability of the polymer to self-heal any local defects in the membrane over macro-sized substrates (cm2) ultimately yields defect-free lipid bilayers with increased electrical resistance across the membrane for ion channel characterization.240 To create composite polymer−lipid films, the polymer can be either physisorbed or covalently immobilized onto the supports followed by the formation and stabilization of the membrane on the polymer layer. Optimization of the chemical composition, thickness, density, and swelling properties of the polymer in response to solution conditions is important to prevent the decoupling and dewetting of lipid bilayers and to avoid the formation of polymer blisters/barriers.245,246 To avoid the formation of polymer blisters in a thermodynamically and mechanically stable polymer−lipid composite film on a solid substrate (exposed to air or water), the wetting conditions have to be carefully controlled and the spreading pressure needs to be maintained positively for both the bilayer and the polymer on the solid substrates. The weakly attractive or repulsive interactive forces between membranes and substrate are particularly important to avoid dewetting of the soft layered films following bilayer deposition247,248 and to minimize the interaction forces at the polymer−lipid interface which can restrict the lateral mobility and elastic modulus of the membrane. Thus, in addition to the hydration level and stability of the polymer films, high flexibility and exclusion of highly charged polyelectrolytes are also major concerns in the design of polymer−lipid composite films. In order to assemble lipid membranes on a soft polymer support with a tunable elastic modulus, versatile approaches

nm acts as a lubricant between the model membrane and solid supports and preserves the long-range lateral mobility of lipids and thermodynamic properties found in a natural membrane. However, the lipid head groups in close vicinity to the substrate surfaces impose some fundamental drawbacks. The interactions between the phospholipid head groups and the hydrophilic substrate results in slowing lateral diffusion and breakdown of the two-dimensional fluidic nature of the membrane.227,228 The membrane-substrate interaction can also induce asymmetric lateral mobility in the inner and outer leaflets.228 The small membrane-substrate space also limited the incorporation of transmembrane proteins and reduced the activity and mobility of incorporated proteins due to the conformational effect and strong interaction with the substrate.229−231 Although the adsorbed bilayer and hybrid bilayer systems have been widely used in spectroscopic applications for structural characterization, these drawbacks have limited such systems for several optical and electrochemical biosensor applications. These limitations can be overcome by increasing the thickness of the aqueous lubricant layer by introducing a soft hydrophilic polymer network or polyelectrolyte film (10−100 nm) between the membrane bilayer and the solid substrate to maintain both the dynamic properties of the membrane and the structural integrity of the functional protein−membrane assembly (Figure 6G).232−241 The polymer film, which acts as a spacer between solid substrate and the lipid bilayer, minimizes negative substrate effects, such as defect formation, decreased lateral mobility, and limited self-annealing of the membrane. Furthermore, the polymer interface allows the incorporation of transmembrane proteins in their native form with the surface domains protruding into the hydrophilic polymer network.230,240,242−244 The dynamic properties of lipids and the 5414

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

to phenol derivatives, which can be polymerized at the anode electrode surface.257 This enables selective and directed functionalization of individual electrodes of multielectrode arrays. Bilayer deposition can occur by three methods: (1) monolayer transfer from the air−water interface with Langmuir−Blodgett techniques which enables the deposition of asymmetric bilayers; (2) vesicle fusion in which lipid vesicles are deposited onto the substrate from vesicle suspensions. By reconstitution of ∼10% of the charged lipids, the vesicles open and form adherent bilayer patches that fuse into continuous bilayers after annealing at elevated temperatures (∼50 °C); and (3) single bilayer spreading which is achieved simply by depositing a lipid reservoir from an organic solution onto the solid. Following the addition of water, a single bilayer is spontaneously pulled over the surface by adhesion forces (if it is hydrophilic and sufficiently attractive). The bilayer is continuous and selfhealing because local pores heal rapidly owing to the strong spreading pressure provided by the lipid reservoir.258 The polymer supported bilayer membrane in combination with microelectrophoresis provides a tool to study the migration of molecules within a bilayer in an applied electric field and for the local enrichment of charged molecules including membrane proteins in two-dimensional systems. This local enrichment of receptor proteins using microelectrophoresis not only allows the separation of the membrane proteins in their native form but also leads to signal amplification for the detection of specific binding reactions.232,259 Overall, the polymer-cushion separating the membrane from the solid substrate without direct linkage significantly reduces the frictional coupling between the lipid molecules and the incorporated proteins in the membranes on the substrate, thus preventing the loss of activity through denaturation. However, the surface passivation of thick polymer layers can greatly reduce the detection sensitivity in studies exploring the kinetics of molecule interaction with model membranes on solid supports with surface-base sensing methods. 3.1.4. Tethered (Chemisorbed) SLBs. In addition to the polymer-cushioned membrane systems, direct coupling of lipid molecules modified with long spacer groups to the solid substrate has also been developed for the formation of tethered bilayer systems260 (Figure 6H). The spacer introduced at the lipid headgroup links the membrane to the substrate in a mechanically and chemically stable way by establishing covalent bonds between some of the lipid molecules in the proximal monolayer of the membrane and the tethering units and to specific reactive groups on the modified substrate surface.47,224,238,261−263 Chemical stabilization of the whole complex architecture results in long-term stability, and this tethering system leads to the required structural spatial and functional decoupling of membrane and substrate. As a result, a sufficient submembraneous space can serve both as an ionic reservoir as well as provide adequate space for the unperturbed incorporation of bulky proteins. The lipids covalently attached to the solid surfaces through modified head groups form a monolayer with hydrocarbon chains exposed to the solvent. A thiolipid comprising a hydrophilic oligoethoxy spacer between the lipid headgroup and the reactive sulfur group has been used to form lipid monolayers on gold surfaces via chemisorption.253 Bilayers can be subsequently formed by the LB transfer, solvent spreading, or by the vesicle fusion approach. Cross-linking the thiolipids onto the titanium oxide surfaces of optical waveguides provides

have been developed as schematically shown in Figure 7, to fabricate the defect-free lipid bilayer-polymer composite systems by (1) chemically grafting a film of highly watersoluble natural polymer, such as dextran, hyaluronic acid, cellulose, and poly(acrylamide), to the solid surface239,247,249 and subsequently depositing a lipid bilayer via liposome adsorption, which is stabilized by attractive electrostatic interaction; (2) the displacement of predeposited lipids by adding poly(ethylenimine) (PEI) polymers to a solid-supported bilayer, thus forming a polymer-supported bilayer to which the attractive polymer−substrate interaction appears to be stronger than the polymer-bilayer interaction;250,251 (3) depositing soft hydrophilic or hydrophobic multilayers of rodlike molecules with alkyl side chains and the subsequent LB transfer of lipid monolayers or bilayers;252 and (4) covalent attachment of lipopolymers which contain lipid head groups modified with polymers to the substrate surfaces followed by vesicle fusion. The underlying polymer separates the tethered bilayer from the solid surface,43,253 and the thickness and density of the polymer layer coupled to the substrates can be controlled by spin coating179,254,255 or through layer-by-layer deposition.239,256 The whole membrane system can be stabilized by a balance of different superimposed molecular interactions of mostly electrostatic, van der Waals, and hydrophobic origins. The stability of the entire system depends on attractive potentials at both the polymer−substrate and the polymer− bilayer interfaces and their strength relative to each other. A weaker attraction at the polymer−substrate interface than the lipid-polymer was found for the case of dextran and poly(acrylamide)-based system. On the other hand, the attractive polymer−substrate interaction is stronger than the polymer-bilayer interaction in the case of the PEI-based systems. An alternative approach to stabilization of a polymer-supported lipid bilayer is based on controlled covalent tethering between the polymer cushion and the solid substrate and between the lipid bilayer and the polymer cushion. The tethering density can be controlled by adjusting the density of cross-linker molecules at the substrate through different selfassembly conditions and the density of actual tethering points through different reaction times. There are versatile techniques for covalently tethering polymers to solids. The covalent tethering can be mediated by forming monolayers of alkyl silanes on Si/SiO2 or indium− tin−oxide surfaces or alkylthiols and alkylmercaptans on Au and GaAs surfaces. The end of the acyl chain of a monolayer can be covalently modified with different functional groups including (1) epoxy groups, which are covalently bound to carboxyl groups; (2) amines, which bind covalently to carboxyl groups of chains (activated via coupling of succinimide or imidazole esters); and (3) photo-cross-linking groups (benzophenone silane-functionalized glass), which can be photochemically linked to any polymer segment. The density of the active anchoring sites of the grafted monolayer can be subsequently adjusted by controlled partial deactivation of the epoxy- and photocross linkers through hydrolysis. Because this yields polar groups, the solid surface is simultaneously rendered highly hydrophilic, resulting in a significant reduction in nonspecific binding of biological macromolecules. The deposition of ultrathin polymer films by electrochemical polymerization is another procedure used for the formation of soft functionalized polymer films (or polymer cushions for membranes). Synthetic polymer chains (e.g., polyethylene glycol) or oligopeptides (e.g., epitopes of antigens) are coupled 5415

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

molecules have a high degree of lateral mobility and flip-flop motion, which provides a favorable environment for protein conformational dynamic necessary for biological activity. In addition, the orientation and the surface concentration of the membrane proteins can be specifically controlled on the surface which is crucial for the analysis of protein function and applications in biosensing. The generation of a protein-tethered bilayer lipid membrane (ptBLM) involves the chemical modification of the support, specific affinity adsorption of recombinant proteins, and reconstitution into the lipid environment by detergent substitution. The specific functional groups can be either attached at the N- or C-terminus of target proteins. Alternatively, functional groups can be added with specific enzymes that recognize particular sequences introduced at a specific region of the protein. Protein-tethered bilayer lipid membranes have also been developed on surfaces modified with biotinylated bovine serum albumin or NTA.265−267 For example, a biotinylated neurokinin-1 receptor (a G proteincoupled receptor) was immobilized in a uniform orientation on streptavidin-coated quartz sensor surfaces.267 In another example, the His-tagged membrane protein cytochrome c oxidase was first attached to the surface in its detergent solubilized form. In the second step, the detergent molecules were substituted by lipid molecules, thus forming a lipid bilayer that is tethered to the support by the protein itself. The coupling His-tag provides sufficient intramolecular flexibility to allow for activity in the reconstituted protein. Above all, it renders the method universally applicable to all His-tagged membrane proteins. Both in situ affinity coupling of proteins and in situ dialysis were combined to form a membrane protein-tethered lipid bilayer.266 In addition to the formation of unilamellar SLBs on the dielectric substrates, complex multilamellar lipid bilayer structures can also be constructed and characterized through the changes in refractive index directly related to the polarizability of molecules at optical wavelengths (Figure 6K). The multilamellar membrane assemblies are found with various structural and functional roles in myelin sheath of neuron, stacked membrane cisternae of endoplasmonic reticulum, lung surfactants, and stacked thylakoid membranes of photosynthetic chloroplasts. The formation of multilayered SLBs generally involved the LB transfer or rupture of GUVs onto the preformed SLBs through the electrostatic interaction with oppositely charged lipids.268,269 To increase the space between each lipid lamellar layer, the multibilayers can also be stacked by DNA hybridization,270,271 biotin−streptavidin coupling,272 polyelectrolytes or lipo-copolymers,273,274 silica and graphene oxide templates,275,276 deposition of silica vapor to spin-coated lipid films275 and dip-pin nanotechnologies.277−279 The multimembrane structures allows the amplification of function and energy production through membrane compartmentalization in series, and the fabrication of multimembrane structures on substrates present great potentials for mimicking and harnessing energy of various energy production systems. Integration of transmembrane proteins into the multilayer membranes can be applied to the 3D crystallization,280,281 structural characterization,269 and biosensing.282,283 3.1.5. Lipid Bicelles and Nanodiscs. The self-assembly of lipids and detergents or short chain lipids into nanosized bicelles, a water-soluble, binary disc-shaped assembly varying in shape, size, and composition, provides a roughly circular membrane system particle in aqueous solution.284 Bicelles

bilayer membranes of similar design.47 This strategy provides the basis for the more sophisticated design of protein− membrane systems for biosensor development. Others have synthesized lipid disulfide anchors on polymer backbones or Nterminal sulfhydryl or lipoic acid modified peptide-tethers derived from an α-laminin subunit to mediate the organization of lipid bilayer formation on solid supports.238,260 In a different preparation method, the reaction between vesicles containing DODA-Suc-NHS molecules as anchors and a cysteaminemodified gold surface resulted in the self-assembly of a phospholipid bilayer covalently anchored onto the gold surface.261 Cholesterol modified with oligoethoxythiol has also been tethered onto gold surfaces which provides an anchor for the fusion of lipid vesicles to form a supported bilayer.263 Effects of the tethered cholesterol density on the kinetics of membrane formation on gold surfaces were also studied using SPR techniques. The resulting tethered bilayers provide long-term stability, and the spacer between the surfaces and the headgroups results in favorable elastic properties of the bilayers. Additionally, the water molecules are retained in the hydrophilic spacer region, which serves as a medium to maintain the hydration and ionic properties of the membrane surface. The elastic properties of the spacer region also allow the incorporation of transmembrane proteins in their native functional form. Peptide spacers coupled to the lipid head groups have also been used to form tethered lipid monolayers on gold surfaces. This peptide-based lipid monolayer was used for the bilayer deposition and incorporation of the membrane protein F0F1-ATPase from chloroplasts and E. coli.262 The controlled construction of protein−membrane systems has also been reported through the development of tethered membranes. The combination of thiolipid and modified gramicidin tethered onto a gold substrate forms the basis of a biomembrane-based biosensor224 (Figure 6I). Hydrophilic spacer groups were coupled to a fraction of the tethered lipids which forms a reservoir and improves the conductivity of the membrane. A tethered membrane spanning lipid was also incorporated to increase the stability of the membrane. Such an immobilized gramicidin-membrane system has also provided a useful system for the investigation of molecular antigen− antibody recognition processes and peptide-membrane interactions. In an analogous approach, a rhodopsin-transducin complex has been coupled to patterned membranes on a gold surface264 (Figure 6J). The self-assembled carboxyl-exposing thiols were first tethered onto the gold surfaces and subsequently removed within the designed patterning using UV lithography methods. The resulting spaces were filled with thiolipids consisting of a long spacer group coupled to the headgroup. A micropatterned array of fluid bilayers and incorporation of the rhodopsin receptor was then assembled onto this patterned tethered membrane. This type of patterned supported membrane system offers a novel approach to screen membrane receptor agonists and antagonists and is discussed in more detail below. An alternative approach to construct a tethered bilayer membrane, particularly for membrane proteins, is based on the initial immobilization of proteins onto the surface followed by bilayer deposition. Several protein immobilization strategies can be used based on high affinity interactions between oligohistidine sequences and nitrilotriacetic acid (NTA), glutathione-S-transferase and glutathione, antibody and antigen, streptavidin and biotin, or complementary oligonucleotides. In these membrane protein-tethered bilayer systems, the lipid 5416

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

of appropriate length, and various MSPs constructs have been recombinantly engineered to control the nanodisc size to accommodate the membrane protein complexes in their native states. Longer MSPs are generally used to assemble the nanodiscs up to 17 nm diameter. With the use of circularized MSPs in which a C-terminal consensus sequence (LPGTG) recognized by sortase A and a single glycine at N-terminus of the protein are covalently linked,308 stable nanodiscs with multiple geometric shapes and defined sizes up to 80 nm have been used to greatly enhance the 2D-NMR signal intensity and spectral resolution for both voltage-dependent anion channel protein (VDAC-1) and the neurotensin receptor-1 (NTR1). Nanodiscs of increasing size have also been used for the incorporation of larger molecular complexes, such as the enzymes involved in cellular energy metabolism, oligomers of integral membrane receptors, ion channels, and transporter proteins.298 As recently reviewed, the successful incorporation of a range of fully functional membrane proteins for structural characterization, such as monomeric or trimeric bacteriorhodopsin, monomeric and dimeric bovine rhodopsin, mammalian cytochromes P450, and their complexes with cytochrome P450 reductase, ATP synthase, cytochrome c oxidase, and various transmembrane receptors has been achieved using selfassembly from a mixture of cholate-solubilized MSP, lipids, and target protein.298,301,309−311 Modification of the MSP sequence for affinity tag (Histidine, FLAG, etc.) incorporation allows both site-specific labeling and immobilization of nanodiscs to the substrates modified with complementary tags and immobilization onto an SPR sensor chip has provided a platform to study the interaction of a tissue factor (TF)embedded nanodisc with human blood coagulation factor VII and X using SPR.312 The combination of nanodisc and nanoarray technology also has significant potential for the display of libraries of membrane proteins on sensing substrates where the proteins are directly captured from the membrane environment of the cellular membrane. However, there is no compartmentalization of the two sides of the bilayer in nanodiscs and are therefore, limited to the study of ion and molecular transport and communication across membranes.

provide a model membrane system with a fully hydrated, planar phospholipid bilayer environment for studying peptide/ protein−membrane interactions and drug partition.285,286 Bicelles are generally prepared by mixing long chain (14−18 carbons) and short chain phospholipids (6−8 carbons) or detergents in chloroform. After evaporating the chloroform, the bicelles are formed via hydration in the buffer to reach different lipid concentrations. By electron microscopy, the size of bicelles range from 10 to 50 nm in diameter with a bilayer thickness of about 4−6 nm.287,288 The combination of a mixture of the longchain lipid (DMPC/DPPC) and the short-chain lipid (DHPC/ DCPC) in aqueous solution results in the formation of mixed micelles that possess a bilayer structure. The size and fluidic properties of the bicelles can be adjusted by the ratio of the long-to-short chain lipids and incorporation of negatively charged lipids.289,290 Membrane proteins can also be embedded in both isotropic and aligned bicelles. Both isotropic and magnetically aligned bicelles have mostly been employed as templates for structural and topological characterization of small membrane-inserting and membrane-associated peptides in solution NMR studies and for structure determination of larger membrane proteins with combined solution/solid state NMR analysis.286,288,291−293 Adsorption of bicelles composed of both long and short chain lipids onto polystyrene and SiO2 substrates also provides an alternative approach to promote the formation of SLBs.294,295 The ability of discrete discoidal bicelles to fuse into a fluid bilayer tethered to aldehydefunctionalized polystyrene particles or planar unmodified SiO2 was demonstrated by fluorescence recovery after photobleaching. The optimal formation of SLBs promoted by the adsorption of bicelles to the substrates are mainly determined by the concentration of lipids and the ratio between the short and long chain lipids. The disc-shaped lipid bilayer has also been stabilized by biotinylated PEG,296,297 such as PEGstabilized nanosized bicelles composed of DSPE-PEG2000-biotin immobilized via streptavidin to biotinylated substrate with varying degrees of carboxymethylation and thickness of the dextran matrix. The stably bound bicelles thus provide an alternative model membrane in SPR-based analysis of membrane interaction of membrane proteins.296 The structural characteristics of discoidal high density lipoproteins (HDLs) composed of cholesterol esters, lipids, and proteins stabilized by the amphipathic apolipoprotein A-I has led to the development of nanodiscs for biochemical and biophysical studies of biomolecule-membrane interaction and for quantitative structure−function analysis of membrane proteins.298−302 Nanodiscs are nanosized planar lipid bilayers stabilized by two membrane-scaffold proteins (MSPs) that combine and form a belt around the lipid disc.298,300,303 These nanodiscs have been used extensively for the incorporation of membrane proteins in their native-like membrane environment yet are soluble in aqueous solution with long-term stability.299,304−307 The formation of empty nanodiscs and protein-integrated nanodiscs generally involve the incubation of detergent solubilized phospholipids, target membrane proteins, and MSPs in a stoichiometric-dependent manner followed by gradual removal of detergent via adsorption on hydrophobic biobeads or by dialysis. Optimal ratios of these constituents and the conditions and mode of protein incorporation into the lipid bilayer are crucial for the formation of stable nanodisc assembly. The solution composition, cofactors, metal ions and other cosolvents all control the efficient assembly of nanodiscs. The size of the nanodiscs can be varied using MSPs

3.2. Lipid Vesicles on Solid Supports

Liposomes (lipid vesicles) are versatile model membrane systems commonly used to study the phase behavior of lateral lipid organization,313−316 molecular recognition of ligands to membrane receptors and channels,46,317,318 cell adhesion and motility,319−321 membrane fusion-fission during cell division and autophagy,322−324 membrane trafficking for vacuole/ exosomes secretion,325−327 and pore formation induced by bacterial toxins/antimicrobial peptides.328,329 The aqueous lumen confined within the liposomes resembles the native intracellular environment, while the channel proteins and receptors can be reconstituted into bilayers to monitor responses generated inside the liposome as a result of molecular transport across the membrane, transmembrane receptormediated signaling, and biochemical reactions for functional studies of biorecognition processes.330 The simple, wellestablished methods for preparing and characterizing liposomes are well-integrated into a broad range of biophysical techniques for studying molecule-membrane interactions.331,332 Molecular components either extracted from native cells or chemically synthesized can be reconstituted using any number of methods and the variation in liposome size and number of lamellar layers 5417

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

can be consistently controlled with extrusion and sonication.333 Depending on the method of preparation, liposomes are generally classified based on the number of bilayers as multilamellar or unilamellar and according to their size as small unilamellar vesicles (SUV; 20−100 nm diameter), large unilamellar vesicles (LUV; 80−400 nm diameter), large multilamellar vesicles (LMV or MLV; 100−400 nm diameter), and giant unilamellar vesicles (GUV; 1−100 μm).334−336 Among this broad range of liposome size, liposomes with diameter less than 100 nm are most compatible with evanescent field sensing.27,337 The liposomes are metastable structures with a short shelf life of a few days. However, the surface of liposomes can be modified (decorated) with hydrogel, biopolymer, or disaccharide to increase their stability for long-term storage.338−340 On the basis of the physical properties of bare and chemically modified metallic and dielectric (silica based) supports, the model membranes can be integrated into the biosensing system as intact lipid vesicles via physisorption,186,341−347 chemical immobilization,348−350 or bioconjugation351−356 to the substrates. The physisorption of liposomes onto an unmodified substrate is mainly governed by electrostatic interactions and hydrogen bonds. The advantages of noncovalent, weak adsorption of liposomes to a substrate include the simple experimental setup as neither liposome functionalization nor surface modification is required, and fully reversible adsorption provides easy surface regeneration. For unmodified plasmonic substrates such as Au and Ag, the structural configuration of model membrane systems can be varied from vesicles to structured bilayers depending on their lattice structure, morphology, and roughness.206 The deposition of SUVs onto a thin film of Au evaporated or sputtered onto silicon or glass with a chromium undercoat generally results in inhomogeneous patches of intact liposomes that do not rupture,186,207,342,344 which ultimately leads to irreproducible binding data. The resistance to vesicle rupture is mainly due to the hydrophobic nature and the surface roughness of the evaporated Au film, which are characterized by the granulated structures with diameters of 15−20 nm and roughness of 1 nm.206 In contrast, lipid tubular structures have been reported for the unidirectional fusion of DOPC:DPPC (1:1) and DPOC:DPPC:Chol (1:1:1) liposomes deposited onto atomically flat, single-crystal Au(111) flame annealed to the dielectric support.357 The tubular structures follow the triangular shape of discontinuous Au terraces with a thickness of 9 nm. Planar lipid bilayers can be obtained by depositing the liposomes onto a flat pristine templated strip of Au of a few square micrometers358 and provides useful substrates for imaging of phase separation and distribution of molecules into specific membrane domains. However, since membrane-based SPR systems require homogeneous membrane layers, Au surfaces are generally modified with various surface chemistries to uniformly stabilize the liposomes on SPR substrates. Silicon-based substrates are mainly used in DPI and are also employed as modified plasmonic platforms. The deposition of liposomes onto SiO2 or SiOxNy generally forms bilayers through complex mechanisms of vesicle adsorption, rupture, fusion, and spreading.207,359 Although some liposomes of particular compositions (e.g., high cholesterol content) under some solution conditions (e.g., absence of divalent cations) fail to form a bilayer on silica based-substrates, these physisorbed liposomes formed an inhomogeneous layer with low stability and produce very inconsistent readout from the sensors. Thus,

when intact liposomes are desired, it is important to chemically modify the silica-based substrates to keep the liposomes intact for sensing purposes. 3.2.1. Surface Modifications for Liposome Adsorption. Surface modification of Au and silicon-based substrates with organic monolayers has been extensively studied. However, several considerations in modifying the surfaces are critical in selecting a particular surface chemistry for optical biosensing applications. The ideal surface chemistry increases the stability of membrane, provides consistent surface coverage, specific attachment of liposomes, reduces the nonspecific binding of molecules to the substrate without hampering the sensitivity, controls the coupling density, allows easy regeneration without surface fouling, and retains lipid fluidity. A common strategy employed to modify both Au and Si surfaces to generate suitable physical and chemical properties for liposome adsorption is via the formation of self-assembled monolayer (SAM) of alkanethiols and alkyoxysilanes.360,361 Alkanethiols and dialkyl sulfides both form a thiolate species on the gold surface when the thiol is formed by oxidation. Thiols adsorb randomly onto a Au surface at an initial stage followed by hydrocarbon chain-reorientation to form a relatively uniform monolayer of alkyl chains tilted at an angle of 26−28° from the normal.362 The condensation between the siloxanes of the organosilane and hydroxyl moieties present on a silica-based substrate surface results in self-assembly of a uniform organosilane monolayer.363 Among the vast variety of commercially available alkanethiols and alkyoxisilanes, short alkyl chains have been mainly used for RI-based sensors due to the high evanescent field decay with distance from the surface, while longer alkyl chain lengths results in a more ordered SAM with higher packing density which is important to guide lipid adsorption and packing in the membranes.363 The hydrophobicity, charge, and functionality of the SAM can be tailored by employing different functional groups such as −CH3, −COOH, −NH2, −SH, −OH, −OSO3−, −OPO32−, epoxy, and other bioaffinity moieties at the end of the alkyl chain.360,361,364 The wide range of available functionalities provides the opportunity to optimize liposome or bilayer formation through the selection of either single or mixed SAMs. The formation of liposome layers are strongly influenced by the mode of coordination of ions between the liposomes and the SAM functional groups. The deposition of neutral eggPC liposomes to the −OSO3− and −OPO32−-modified Au surface results in adsorption of intact liposomes in the absence of divalent cation (Mg2+), while bilayers formed when equilibrated with 5 mM Mg2+.344 Similar intact liposome layers can also be obtained on thiolated SAMs with terminal −OH groups. Thus, the vesiclesurface interaction can be precisely controlled for intact liposome or SLB formation on the surface by tuning the composition of the deposition solution. The rupture of adsorbed SUVs and the formation of stable lipid bilayers composed of neutral lipids also depends strongly on the critical surface density of charged groups such as −COO− and NH3+. It has been shown that the deposition of neutral eggPC SUVs onto mixed SAMs on Au formed with different ratios of H2N(CH2)11-SH/HO-(CH2)11-SH resulted in the adsorption of intact vesicles at the surface with 80% −NH3+ or −COO− resulted in the formation of fluidic SLBs. Thus, the vesicle-bilayer trans5418

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

formation at a critical surface charge density of ∼75% and lipid mobility on the SAMs can be controlled by the composition of mixed SAMs, the solution counterions mediating the interaction between the liposomes, and the charged groups on the SAMs. 3.2.2. Direct Coupling of Liposomes to Surfaces. Liposomes can also be directly coupled to substrates that are modified with various functional groups. Covalent coupling provides stable and strong attachment of liposomes to the sensing substrates. A broad range of coupling agents can be used to selectively immobilize liposomes with different functionality to sensing substrates and to pattern discrete liposome arrays, and one of the earliest methods developed is based on amine-coupling chemistry. 1-Ethyl-3-(3(dimethylamino)propyl)-carbodiimide (EDC), a zero-length cross-linking agent, is commonly used to activate −COOH modified substrates for immobilizing primary amines of PEcontaining liposomes or peptide/protein-incorporated liposomes. Since the reactive ester can be rapidly hydrolyzed in aqueous solutions resulting in low coupling efficiency, Nhydroxysuccinimide (NHS) or N-hydroxysulfosuccinimide (sulfo-NHS) are used to enhance the stability of the active ester and therefore the coupling efficiency. On the other hand, the aminated surface can be carboxylated with succinic anhydride and then coupled to a primary amine present at the surface of liposomes or proteoliposomes using EDC/NHS coupling. The carboxylic acid-containing liposomes can be prepared by including carboxylic acid-lipids such as POPE-NG [1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoethanolamine-N(glytaryl)]. Alternatively, the carboxylic acid functionality can be introduced into the PE-containing liposomes by reacting PElipids with a wide range of anhydrides in the presence of triethylamine.366 The carboxylic acid-modified liposomes can be activated with EDC followed by reaction with primary amine groups on the modified sensing substrates. In addition, this coupling method also allows the biofunctionalization of liposome surfaces with various biomolecules such as peptide and protein ligands for specific bioconjugation of liposomes to the substrates.366−369 The main advantages of using −COOHmodified liposomes is that prior ligand modification is not required and protein denaturation is minimized. However, random attachment of proteins with multiple amines may reduce the affinity of bioconjugation to the substrates. In addition, liposomes/proteoliposomes can also be coupled to the functionalized substrates using various homobifunctional and heterobifunctional cross-linkers.370,371 The homobifunctional cross-linkers containing NHS-ester, isothiocyanate, or glutaraldehyde on either or both ends of the linker are reactive to amine-moieties. For the heterobifunctional linker, the other reactive group in the linker such as an NHS-ester, maleimide, or carbodiimide reacts with target affinity groups at primary amine, thiol, or carboxyl groups, respectively. Using a heterobifunctional cross-linker for amine-to-sulfihydryl conjugation such as m-maleimidobenzoyl-N-hydroxysuccinimide ester (MBS), PE-liposomes can also be immobilized to the −SH-modified substrates. The reversible properties of the disulfide bond also allows the thiolate surface to be regenerated by treatment with reducing agents such as dithiothreitol (DTT). However, this approach has not been widely adopted, and more studies have yet to be demonstrated for the optical sensor. The −SH functionalized surface can also be used to immobilize liposomes containing maleimide-modified lipids or

the use of click chemistry between the thiol surface and alkenemodified lipids. 3.2.3. Liposomes Adsorbed to Protein- and Biopolymer-Modified Surfaces. Liposomes can also be adsorbed onto a substrate that has been passivated with proteins or either natural or synthetic polymers.372,373 Common passivation schemes include covering the substrates with a layer of protein such as bovine serum albumin (BSA) or fibrinogen, native or synthetic hydrogels, functionalized SAMs, surface-tethered polymer brushes, physisorbed copolymer brushes, and amphiphilic lipid bilayers. The main purpose of the underlying passivated layer is 3-fold: (1) to reduce the nonspecific binding of molecules to the substrate, (2) to provide a lubricant layer for the reconstituted membrane proteins, and (3) to maintain the lipid diffusivity and lipid bilayer integrity of membranes. The formation of either a bilayer or an intact liposome layer via depositing liposomes onto a BSA-passivated substrate depends on the conformational state of the BSA on the substrate.374 The concentration of BSA used for surface passivation is critical in controlling the surface roughness for the successful formation of SLBs composed of neutral lipids. Glass or silica substrates are passivated with a low concentration of BSA solution for low surface roughness, while the bilayers will not form on substrates passivated with high concentrations of BSA above 0.2 mM which cause a high surface roughness. In contrast, neutral liposomes adsorbed to substrates coated with either heatdenatured or β-mercaptoethanol (β-ME) reduced BSA deposited as a tightly packed intact liposome layer without rupture. In the absence of divalent cation such as Ca2+or Mg2+, the adsorption of anionic liposomes also formed intact liposome layers on substrates passivated with native, heatdenatured or β-ME-reduced BSA. In the presence of Ca2+, the anionic liposomes remained intact on the heat-denatured or βME-denatured BSA films possibly due to the loss of strong Ca2+ chelation sites and the reduction of surface charge below a critical charge density on the surface preventing bilayer formation. Liposomes can also be stabilized on substrates modified with tethering groups which intercalate into the bilayer of liposomes without rupturing. The tethering methods generally involve the covalent attachment of polymers containing hydrophobic acyl chains to the substrate by transferring the preorganized lipocopolymer at the air−water interface to preactivated substrates or immersing the preactivated substrates into a solution of lipocopolymer.244 Alternatively, homopolymers or heteropolymers can be deposited with a layer-by-layer approach followed by covalent attachment of an acyl chain to the polymer layer.375,376 The acyl chain can be immobilized in solution or by transfer of the preorganized lipid monolayer containing reactive anchor groups at the air−water interface to the polymer on the substrates. Films of the alkyl chain-modified biopolymer can also be prepared by evaporating an aqueous solution of soluble lipocopolymer directly onto the substrate. In addition to the transfer approaches, passivation of substrates with hydrophilic polymers such as molecular spacers can be achieved by either “grafting from” or “grafting to” strategies. The “grafting from” technique is commonly used to form a polymer layer of large area, in which the surface is modified with initiators followed by immobilization of the anchoring groups. Both thin and thick films of brush polymer spacer can then be grown from the anchoring groups by various in situ surface-initiated polymerization (SIP) reactions,377,378 including radical chain polymerization,379 living cationic polymer5419

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Figure 8. Functionalization of liposomes, proteoliposomes, and lipid bilayers either physisorbed or chemisorbed onto substrates passivated with various molecule layers. Various passivation chemistries are used to modify solid substrates with polymers such as PEG, PLL-g-PEG, grafted polymers, proteins, SLBs, or self-assembled monolayers with different terminal functionalities. The membrane can be further specifically attached to the passivated surface with different bioconjugation chemistries with and without linkers. These versatile approaches provide stable, reversible, and selective binding of membranes to a surface for multiplexing analysis of membrane binding.

ization,380 and controlled living polymerization such as atomictransfer-radical polymerization.381,382 The monomers can polymerize into linear or branched structures. This “grafting from” approach produces polymer brushes with high density which cannot be further functionalized nor used for patterning of sensors with a low density of functional groups. In contrast, the “grafting to” approach in which the polymer is chemically immobilized to the anchor prior to depositing the lipid bilayers or other functional groups allows the direct patterning of lipid bilayers and functional groups on the sensing substrates. However, depending on the physical extension of the polymer chains and mutual exclusion interactions between the polymers, if the density of the polymer graft is significantly below the threshold required for the extended brush structure, nonspecific binding of molecules to the substrate will occur. This drawback of low graft density can be solved by adsorbing copolymers such as PLL-g-PEG for which the polymer grafting density can be controlled at high density. Thus, the layer thickness, surface roughness, density of anchorage points, and the conformational freedom of the polymer chains in response to environmental properties can be tuned with the chemical composition of the polymer and the immobilization methods, including the crosslinking density to achieve stable immobilized liposomes. With advances in micro/nanopatterning technologies, the hydrophobic alkyl modified polymers can also be patterned onto substrates with spatial precision. To maintain the integrity of the tethered liposomes and the functional activity of transmembrane proteins, several features of the lipocopolymers are considered to be essential for sensing applications. These features include the degree of polymer expansion that is tunable by pH, salt, ionic strength, and temperature that provides an

aqueous environment for lipid head groups and extramembrane domain of proteins. It is also critical to maintain the continuity and homogeneity of the tethered liposomes. The lateral mobility of membrane components will not be significantly reduced by the hydrophobic tether and polymer. Several lipocopolymers used for the formation of liposome layers involve the hydrophilic, nontoxic alkyl chain (hexadecanoic acid)-modified poly(ethyloxazoline-stat-ethylenimine) (PEOxPEI),244 positively charged quaternary ammonium compound (QAC) with a dodecyl chain covalently attached to PEGgrafted poly(L-lysine) (PLL-g-PEG),383 dodecyl chain modified chitosan (C12-chitosan),384 and hydrophobic acyl chain modified carboxymethyl dextran.385 PLL-g-PEG has been commonly used to passivate the substrate for the formation of a supported lipid bilayer. The size and density of the PEG moiety can be controlled to optimize the distance between the liposome lipid bilayer and the substrate in which a hydrophilic space of >10 nm has been estimated for a molecular weight of 2−3.4 kDa per PEG chain at an optimal grafting density.386−388 The effect of the QAC density on the morphology and lipid fluidity of liposome layers has been characterized on a substrate modified with a mixture of 10−100% PLL-g-PEG-QAC in nonfunctionalized PLL-g-PEG.383 Vesicle deformation without rupture was then demonstrated for the adsorption of SUVs and LUVs of different compositions onto the low density PLL-gPEG-QAC modified substrate, and the extent of liposome deformation increased at higher QAC density. Thus, the extent of the vesicle deformation can be tuned by changing the QAC density. The transformation of the deformed vesicles into a bilayer on the PLL-g-PEG-QAC was also induced by adding low molecular weight PEG to cause liposome rupture. The lipid 5420

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Figure 9. Classical and nonclassical ligation chemistries for the functionalization of lipid bilayers. Liposomes or SLBs are commonly postfunctionalized with ligands ranging from small molecules, including imaging probes, peptides to macromolecules such as proteins, antibodies, DNA, and oligo/polysaccharides for biosensing of specific ligand and cell-matrix adhesion studies. These ligation chemistries allow the attachment of ligands without altering the integrity of the lipid bilayers. Classical ligation chemistries including amine-carboxylic acid conjugation, disulfide bridge formation, hydrozone bond formation, and thiol-maleimide addition have been applied to immobilize different moieties to the lipids of preformed liposomes and SLBs. Versatile bioorthogonal chemistries such as native chemical ligation, CuAAC, Cu-free cycloaddition, Staudinger ligation, and IDEEA are now commonly applied to immobilize the functional moieties to the membrane in a more specific, selective manner.

the lipid monolayers on hydrophobic SAMs is about 0.2 μm2/ s.383 The formation of stable vesicle layers has also been shown on a hydrophobic acyl chain modified natural biopolymer in which chitosan was modified with a low density of C12 alkyl groups at 2.5%.384,389 This C12 modified chitosan formed a smooth layer

molecules in this PLL-g-PEG-QAC supported bilayer exhibited a high diffusion coefficient of about 10 μm2/s, which is 5 fold higher than the lipid diffusion at 2 μm2/s in a planar bilayer formed on the UV/ozone cleaned glass via liposome-bilayer transformation, while the lateral diffusion of lipid molecules in 5421

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

with increased water contact angle. The DPPC LUVs tethered to the C12-grafted chitosan through hydrophobic insertion remained intact and densely packed and was able to withstand extensive washing and repetitive glass probe shearing. In contrast, no adsorption of liposomes was observed on the unmodified chitosan film. Liposome immobilization based on the density of hydrophobic tethers leading to the formation of stable vesicle layers also depends on the surface coverage of liposome. The critical role of surface coverage of liposomes on alkyl chain-grafted polymer supports has also been demonstrated by experimental and simulation models.390 The adhesion of SOPC SUVs to alkyl chain terminated PEGs on silica and on poly-L-lysine can be enhanced by a high density of alkyl chains (AC). While the overall vesicle deposition leads to the formation of a bilayer, the rupture of liposomes only occurred at a high surface coverage above 3.45 ng/mm2. This coverage is also higher than the surface coverage threshold of about 2.2 ng/mm2 for vesicle-to-bilayer transformation on unmodified silica. Liposomes are less prone to rupture and merge on the polymer supports, and this higher stability may result from the flexible nature of polymers compared to the hard interface of such silica, which prevents excessive vesicle deformation leading to reduced rupture rate. The integration of microsized-supported liposome and proteoliposome systems described above into optical biosensing systems allows measurement of molecule-membrane and ligand-membrane receptor interactions. However, the physisorption of liposomes onto polymer cushions or tethered through hydrophobic anchor approaches generally lack the selectivity in which liposomes with different compositions can be adsorbed to the same locations on the substrate, leading to variable distribution and surface coverage. Thus, without specific labeling of the liposome, this limits the use of this approach for surface patterning of liposomes or for quantitative analysis of liposome-liposome interactions in the study of, for example, vesicle fusion. With advances in nanofabrication technologies, specific immobilization of liposomes composed of various lipid/protein components forming discrete functional membrane arrays can significantly increase the throughput and multiplexing of ligand binding analysis. To achieve this, various specific surface chemistries and biofunctionalization of either SAMs or polymer-modified substrates have been extensively used to attach the liposome in a specific manner, as shown schematically in Figure 8. The functionalized liposomes can be directly immobilized onto the modified substrate or attached via bioaffinity methods. Functionalization of both the liposome/proteoliposomes and the substrates are required in specific bioaffinity methods, and various conjugation chemistries have been developed to attach a broad range of specific functional groups to surfaces of both liposomes and substrates.360,364,391−396 Some commonly used paired bioaffinity groups are biotin−avidin/streptavidin, metal chelating groups (NTA-Ni-His6), Ab-Ag, DNA−DNA, DNA/RNA aptamer-DNA/proteins/sugars, cell-adhesion motif (eg RGD, IKVAV peptides)-integrin, carbohydrate-lectin, protein A-IgFC, and coiled-coil peptides. These bioaffinity conjugation methods allow highly specific immobilization of liposomes to passivated substrates with optimal densities of the accompanying groups. Furthermore, the intervesicle distance can be controlled on a patterned array for a target spatial resolution of signal response from an individual vesicle or vesicle pairs. Various approaches have been used to introduce the bioaffinity groups to the liposome surfaces.392,395

Conventionally, biofunctionalized liposomes are prepared by mixing the synthetic lipids with bioaffinity groups, such as biotinylated PE or NTA-modified lipids, with other lipid components at a particular molar ratio using various standard liposome preparation methods. This approach has been widely used due to the simple handling without additional liposome purification, and no further chemical modification to the liposomes is required. However, some synthetic lipids with bioaffinity groups may have limited solubility and stability in solvent or may not be fully compatible with various stages of liposome preparation and/or do not readily form liposomes. The self-assembly of lipids into liposomes is governed by the geometrical packing parameter of the lipid headgroup size, chain length, and the lipid volume. Large bioaffinity groups on the lipid head may thus limit the formation of liposomes and requires optimization of the molar ratio. Alternatively, bioaffinity groups can be introduced by chemical or enzymatic modification to preformed liposomes carrying the native reactive lipids such as PE or PS or synthetic lipidlike molecules with chemically modified reactive groups. The degree of biomolecule conjugation to the liposome surface can be controlled by the fraction of reactive lipids and also depends on the efficiency of coupling chemistry. As shown schematically in Figure 9, various nonselective and orthogonal coupling chemistries have been explored for the functionalization of liposomes for various applications. The ideal coupling chemistries should react rapidly and specifically under mild reaction conditions to minimize destabilization of the bilayer properties and liposome deformation. It is also important that coupling of the bioaffinity groups does not negatively impact on the structure and activity of integrated proteins. 3.2.4. Classical, Non-Selective Chemistry for Lipid Functionalization. 3.2.4.1. Amine- and Carboxylic AcidModified Liposomes. The amino groups of natural PE in liposomes can be functionalized by reaction with dimethyl suberimidate or glutaraldehyde to form an imidoester or imine cross-linking, respectively, for amine−amine cross-linking.397,398 A variety of antibodies have been coupled to a liposome surface with this approach to produce antibodymodified liposomes that are able to bind to the respective antigen-coated matrices with high affinity and specificity.398 The main advantage of this functionalization with a homobifunctional linker is the use of natural lipids present in the membrane without prior derivatization. The free carboxylic groups in 1,2-dipalmitoyl-2-oleoyl-sn-N(glutaryl) (DP-NGPE) can be activated with EDC/NHS for further reaction with amine-containing biomolecules as commonly used for peptides/proteins.366−369 This allows in situ activation of −COOH functional groups without prior modification of ligands to be attached to the liposome surfaces which reduces the risk of ligand denaturation. However, there are two drawbacks of both these approaches: (1) the ligand and liposome can be prone to aggregation and (2) the presence of multiple primary amines in the ligands can result in the random orientation of immobilization and alter the affinity toward the target analytes. 3.2.4.2. Thiol- and Bromoacetyl-Modified Liposomes. Thiol- and bromoacetyl-functionalizations are also commonly used for ligand immobilization to liposome surfaces. The reactive thiol is introduced by conjugating 3-(2-pyridyldithio)propionate (PDP) to PE yielding the dithiopyridine (DTP)phospholipid which forms a disulfide linkage with DTPmodified proteins activated at acidic pH.399,400 The disulfide 5422

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

acid.413 This method has also been used to couple peptides to azide-modified liposomes as shown for the coupling of the Cterminal alkyne-functionlised gH625 peptide to azide-AdOOLys(C(O)CH 2 CH 2 C(O)N-(C 18 H 37 ) 2 )-amide and tetrabranched neurotensin peptides ((NT8−13)4-alkyne) to (C18H37)2-PEG9-N3 incorporated liposomes.419,420 Liposomes doped with 2% azide-modified lipids have also been coupled to alkyne-functionalized ssDNA-lipid in a glass-supported lipid layer.350 Tethering of the vesicles is mediated via hybridization to complementary DNA in the liposomes and stabilized by the formation of triazole linkage. This DNA-templating click reaction allowed the docking and fusion of heterogeneous liposomes carrying different DNA sequences to be studied. A minimum of 2% POPE-N3 was required for the effective coupling of the vesicles which displayed the azide to alkynefunctionalized DNA-lipids.350 Following the development of these azide-modified liposomes, efficient liposome postfunctionalization with biotin allows for further immobilization to a streptavidin-coated microarray and used to generate a multiplexed microarray platform to explore the specific binding of protein kinase Cα to diacylglycerol-doped liposomes.356 Although the CuAAC provides efficient coupling of functional moieties to liposomes in aqueous media at room temperature to form a triazole ring which is stable to temperature and hydrolysis, unsaturated lipid acyl chains are prone to oxidation by copper ions which could cause liposome deformation and degradation of the bilayer. The toxicity, DNA degradation, instability, and interference with protein activity associated with the copper catalyst are also the main drawbacks of this click reaction, and a chemoselective aniline-catalyzed hydrazone coupling has been developed as an alternative coupling method. The strain-promoted azide−alkyne [3 + 2] cycloaddition (SPAAC) referred as “Cu-free click chemistry” eliminates the use of the cytotoxic Cu2+ catalyst by employing cyclooctyne which rapidly reacts with an azido group forming a triazole ring under ambient conditions without altering the function of the immobilized macromolecules.421−425 This method has been used for lipid modification by introducing dibenzocyclooctyne (DBCO) to DSPE-PEG followed by liposome preparation. The DBCO-liposomes then react with azidomannose and azidomodified sialic acids to form the glycoliposomes. Vesicles doped with synthetic lipids modified in this way with bicycle[6.1.0]nonyne (BCN) cyclooctyne groups can be coupled to a variety of azide-containing molecules as demonstrated for the small dye compound, 5′TAMRA-azide, and large azide-modified Ankyrin repeat proteins.425 Postfunctionalization of liposomes using this method has also been demonstrated with the coupling of peptides. For example, the cyclic-RGD peptides modified with a bifunctional polyHis-tag and azide groups have been coupled to liposomes composed of 1% DSPE-PEG 2000 -cyclooctyne and 1% DSPE-PEG 2000 NTA.421 A coupling efficiency of 98% was achieved with this double coupling of affinity NTA-Ni2+-His6 and covalent triazole linkage. The SPAAC was also used to couple a somastostatin receptor targeting peptide (TATE) to the lipid moiety.421 However, the coupling efficiency is generally lower for large proteins, and the reaction did not reach completion. In addition, the coupling reaction can be complicated by the formation of two triazole regioisomers leading to multiple products. Alternatively, the Staudinger ligation of triphenylphosphine (TPP)-carrying liposomes with azide-containing bioaffinity groups to form an amide bond via aza-ylide intermediate

protected derivative PDP-PEG-DSPE can be activated in situ by dithiothreitol (DTT) reduction for further attachment of maleimide-derivatized ligands.401,402 The bromoacetyl-functionalized liposomes allow the conjugation of cysteine-containing peptides/proteins.403,404 The bromoacetyl group was introduced by acylation of PE with 2-[2-[2-[(2-bromoacetyl)amino]ethoxy]ethoxy]-ethoxy acetic acid in the presence of N,N′-dicyclohexylcarbodimide (DCC). As the maleimide reacts with bromoacetyl at pH = 9.0, this allows the selective attachment of different maleimide-modified ligands to liposomes exposing both thiol and bromoacetyl groups at separate acidic and basic pH.404 3.2.4.3. Maleimide-Modified Liposomes. The formation of a thioester bond between maleimide-functionalized liposomes and thiol-derivatized ligands is widely used to functionalize liposomes with targeting ligands by Michael addition.403−410 Liposomes carrying N-(4-(p-maleimidophenyl)butyryl)-PE have been commonly used for direct surface functionalization, while maleimide-PEG-PE is also commonly used for ligand attachment to the distal end of the PEG spacer. The synthetic maleimido-functionalized lipids incorporated during liposome preparation can react with the cysteine under mild conditions. Conjugating the CGGH6 (a hexa-histidine containing peptide) peptide to maleimide-liposomes allowed their association with Ni2+-NTA present on the passivated substrate. The commercially available DSPE-PEG-maleimide phospholipid has been widely used to functionalize the liposomes with thiol-containing affinity groups. The length and polarity of the spacer can influence the reactivity and coupling efficiencies of ligands to the maleimido groups in which a longer polar spacer results in higher coupling efficiency.407 3.2.4.4. p-Nitrophenylcarbonyl-Modified Liposomes. This method has been developed for attaching ligands to the PEG terminus of liposomes without prior derivatization.411,412 The p-nitrophenylcarbonyl-PEG (pNP-PEG) modified liposomes were obtained in a single step by reacting DOPE with bis-pNPPEG which forms stable carbamate bonds with ligands containing primary amines. However, this method is limited by dimerization and lack of site-specific immobilization. In addition, these methods are nonchemoselective, which can lead to heterogeneous conjugations at different or multiple sites and the harsh reaction conditions cause the unravelling, rupture, and fusion of immobilized liposomes. 3.2.5. Bio-Orthogonal Coupling Chemistry for Lipid Functionalization. Site-specific conjugation of bioaffinity groups to liposome surfaces can be achieved via various bioorthogonal coupling chemistries which are chemoselective and are characterized by a fast reaction rate, mild reaction conditions in aqueous medium and high coupling efficiency. As the functional coupling groups are not present in natural bioaffinity groups such as peptides/proteins, the resulting products are homogeneous and free of side products. One of the selective coupling methods to postfunctionalize liposomes is based on the Cu(I)-catalyzed azide/alkyne Huisgen 1,3dipolar cycloaddition (CuAAC).350,413−418 This click chemistry using CuAAC leads to the formation of a triazole ring as shown by coupling of an azido-modified α-D-mannose ligand to a 5− 10 mol % alkyne-modified ether lipid (N-[2-(2-(2-(2-(2,3bis(hexadecyloxy)propoxy)ethoxy)ethoxy)-ethoxy)ethyl]hex-5ynamide) and incorporated into a liposome with 80% coupling efficiency.417 The mannose was readily accessible to binding of concanavalin A. The alkyne group can also be introduced to the preformed liposomes by derivatizing the PE with proliolic 5423

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

bioaffinity groups. These enzymatic methods promote the covalent coupling of specific functional groups under mild conditions particularly for postfunctionalization of liposomes with large proteins in a specific and orientation-defined manner. For example, sortase A is a transpeptidease from Staphylococcus aureus that recognizes the −LPXTG-motif at the C-terminus of a protein and cleaves the sequence between threonine (T) and glycine (G) residues forming a thioester bond with threonine.438−442 The cleavage is then followed by a transpeptidation where the free carboxyl group of threonine is transferred to the substrate containing a glycine with a free amino group (e.g., H2N-GGG-) forming a new peptide bond. This enzymatic approach has been demonstrated by attaching an enhanced green fluorescent protein (eGFP) engineered with a C-terminal LPATG-H6 peptide motif to the liposome composed of 2% DSPE-GG-NH2 and DSPE-PEG2000-GGNH2 by incubating with sortase A at 37 °C.439 The conjugation of eGFP is more efficient with the glycine-lipids with PEG2000 spacer. These enzymatic modifications can be readily expanded to conjugate other peptides and protein bioaffinity groups such, as single chain antibodies (scFv), to the surface of liposomes doped with various sortase substrate modified-lipids.443−446 In addition to the chemical and enzymatic modification of liposomes, hydrophobic insertion of ligands derivatized with natural and synthetic anchoring groups allows the direct incorporation of functional moieties into bilayers. This approach has been well-characterized in vivo in which the anchoring groups such as fatty acids were transferred by enzymes to the proteins for membrane insertion.447,448 Several in vivo acyl modifications of proteins involve prenylation, myristoylation, and palmitoylation. In prenylation, an isoprenoid, such as either a farnesyl or a geranylgeranyl group, is added to a cysteine of a tetrapeptide motif (CAAX) at the Cterminus of proteins.448,449 This in vivo prenylation of proteins can be catalyzed by farnesyltransferase (FTase) using farnesyl pyrophosphate as the prenyl donor or geranylgeranyl transferase type I and II using geranylgeranyl pyrophosphate.450−452 The in vitro addition of a farnesyl group to proteins has been established by incubating the CAAX motif containing substrate with isolated FTase. The farnesylated proteins can also be produced in an E. coli expression system double-transformed with CAAX-protein substrate and FTase.452 The farnesyl pyrophosphate in the E. coli provide natural substrate for the production of recombinant farnesylated proteins. The myristoylation reaction involves the addition of myristic acid to the glycine of an N-terminal MGxxxS/Txx motif in substrate proteins catalyzed by N-myristoyltransferase (NMT).453−455 Palmitoylation of proteins is catalyzed by protein acyltransferase which has been used in vitro for the preparation of palmitoylated peptide amphiphiles self-assembled into monolayers.456 These enzyme-catalyzed additions of hydrophobic anchor groups have been particularly successful in functionalizing a membrane surface with large proteins which can be difficult to couple by chemical methods. In particular, some lipid and protein components can interfere with the chemical reaction. Furthermore, different protein species can be incorporated to the membrane surface, and the ligand density can be well-controlled using this anchoring approach. Other anchoring groups such as cholesterol and bilayerinserting peptides have also been employed to anchor proteins, peptides, and oligonucleotides (DNA, RNA, and aptamer) to membrane surfaces. Glycosylphosphatidyinositol (GPI), a natural lipid anchor for many extracellular membrane-

rearrangement has been employed for site-specific functionalization of liposomes.348,349,426−430 This Staudinger ligation method also allows the immobilization of TPP-liposomes to azide-PEG-modified substrates. For example, terminal TPPderivatized lipids have been prepared by amidation of cholesterol-PEG2000-NH2 and DSPE-PEG2000-NH2 with 3diphenylphosphino-4-methoxycarbonylbenzoic acid NHS active ester.426,428 The density of azido-reactive groups on the liposome surface can be fine-tuned by incorporating TPPderivatized lipids into liposomes at various ratios. The reactivity of these TPP groups on the surface-immobilized liposome toward azide-modified lactose431 were employed for the fabrication of glycomembrane arrays for lectin-binding assays.426 An 80% functionalization of the TPP groups was obtained based on the phenol-sulfuric acid test for carbohydrate quantification. The absence of carboxyfluorescein leakage confirmed the integrity of the liposome under the ligation conditions. In addition, the aggregation of liposomes was prevented by the coupling of azido-lactose. The Staudinger ligation has also been demonstrated for the biotinylation of TPP-PEG-liposomes with azide-modified biotin.349 The biotinylated liposomes were subsequently immobilized onto a streptavidin-micropatterned glass. This method offers great promise for fabricating the patterned array of glycomembrane mimetics through further coupling of azide-containing carbohydrates (e.g., azidoethyllactoside) to the TPP-PEG liposome array of azide-modified or streptavidin-coated substrates. Immobilisation of liposomes carrying a TPP-anchor and ganglioside (GM1 and GM3) onto an azide-modified glass as a glycoliposome array allows the differential analysis of specific binding of lectin and toxin to various ganglioside structures.348 This strategy thus meets a number of important criteria in the fabrication of functional liposome arrays in that the stability of immobilized liposomes is maintained via the PEG spacers on both the substrate and the liposome surface, and a diverse array of azido-compounds can be separately functionalized onto the TPP-liposomes in discrete spots of substrates. Another biorthogonal chemoselective reaction between the trans-cyclooctene (TCO) and tetrazine has also been introduced to postfunctionalize liposomes with peptides.432−435 This reaction is based on the inverse electron demand Diels− Alder cycloaddition (IEDDA) of the dienes of the tetrazine with the dienophile TCO upon retro-[4 + 2] cycloaddition and has found broad application for both in vitro and in vivo celllabeling with specific probes, particularly for imaging and radiotherapy.433 To introduce the reactive TCO groups onto the liposome surface, DSPE-PEG2000-NH2 was reacted with TCO-NHS to form DSPE-PEG2000-TCO, which was mixed with other lipid components at a 2.5% molar ratio for liposome preparation. The reactive TCO-tagged liposomes were then functionalized with a pH (low) insertion peptide (pHLIP) via a maleimide−thiol reaction through the maleimide-tetrazine heterobifunctional linker conjugated to the cysteine-terminated pHLIP.436 The linear pHLIP peptide undergoes random coil− helix transition upon lowering the pH in the surrounding medium which facilitates its insertion into the membranes.437 Overall, a broad selection of chemistries for postfunctionalization can be readily exploited to attach liposomes to a substrate in a specific and location-addressable manner. In addition to the chemical postfunctionalizations of liposomes described above, enzyme-mediated ligations have widespread use in the design of liposomes with multiple 5424

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Figure 10. Versatile lipid bilayer and liposomes supported on different nanostructured substrates. (A) Exposed metallic nanodisks coated with a thin silica layer, (B) embedded metallic nanodisks in optical transparent epoxy coated with a thin silica layer of ∼10 nm, (C) SLBs formed within the micro/nanohole following the shape of the structured substrates; (D) SLBs spread across the nanohole to form suspended membrane nanoarrays, (E) formation of intact lipid vesicles or SLB-coated nanoparticles of various diameters in the nanohole, and (F) the membrane curvature controlled on the structurally corrugated substrates. Reproduced from ref 491. Copyright 2006 American Chemical Society. Reproduced from ref 492. Copyright 2008 American Chemical Society. (G and H) Formation of curved membranes over the nanoparticles adsorbed on the planar substrates and the effect of particle diameter on the topographic structures of the bilayer as continuous or discontinuous bilayers on the nanostructured substrates. Reproduced with permission from ref 493. Copyright 2008 American Chemical Society. Reproduced with permission from ref 494. Copyright 2009 American Chemical Society.

modified with double lipid anchors are more stable in the bilayer.465,468−470 Depending on the sequence properties of the oligonucleotide, the single-strand oligonucleotide (e.g., ssDNA) can interact with a bilayer surface and lower the hybridization efficiency to the affinity group.471,472 The bilayers modified with complementary short ssDNA-lipids and long ssDNA-lipids not only increase the stability of the DNA-lipids in the bilayer but also prevent membrane-DNA association and increase their hybridization efficiency to the target affinity groups.350,471 Native and synthetic peptides capable of inserting into the bilayers are also used as anchoring groups for functionalization. Gramicidin A (gA), a natural ion channel-forming peptide secreted by Bacillus brevis, facilitates a transmembrane flux of monovalent cations upon reversible head-to-head dimerization in a lipid bilayer and has been used as an anchor to function as an ion channel in an artificial membrane. Biotin and hapten linked to gA also provide the affinity groups for binding of streptavidin and hapten-specific Fab antibodies224 to the membrane surface. The modification of gA with various charged functional groups such as converting a tert-butyloxycarbonyl (BOC)-protected glycine group to glycolic acid, hydroxyl, sulfate, and amino groups at the C-terminus allowed its reaction with different analytes in solution473 to create a versatile membrane-based biosensor.

associated proteins, can be coupled to ligands for subsequent incorporation into a membrane via the GPI tail.457,458 The GPIanchored proteins are commonly prepared by in vivo expression systems. However, the structural heterogeneity of the GPI anchors makes them difficult to isolate and purify in a pure form for further applications. Various methods have therefore been developed for the chemical and chemoenzymatic synthesis of GPIs and GPI-anchored peptides/ proteins and glycoproteins for membrane functionalization and other biomedical applications.459−461 The proteins of interest constructed with GPI-modification sequences can then be incorporated into the liposomes and SLBs via insertion of the GPI tail into the bilayer.462,463 Although the GPI-anchored proteins can be purified from the cell lysate, the difficulties in purification and aggregation of GPI anchored proteins have limited the widespread use of this approach in functionalizing model membranes. Liposomes have also been commonly functionalized with synthetic lipidated oligonucleotides with cholesterol, stearyl, and distearyl lipid anchors.464−467 Nucleic acid-lipids provide unique barcode systems for the formation of heterogeneous arrays of SLBs or liposomes composed of specific lipid and protein components at a defined ratio.353 Single lipid acyl or cholesterol-derivatized oligonucleotides are less stable and can dissociate from the bilayer surface, while oligonucleotides 5425

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

physical properties that could be used for addressing, activation, or detection. Coating metallic nanoparticles with SLBs allows quantitative LSPR measurement of protein−membrane interactions495 as demonstrated by the cooperative multivalent binding between a protein and a GM1-containing membrane.496 The deposition of SLBs-coated nanoparticles of liposomes onto patterned nanoholes497,498 with different periodicities also provides a high throughput, multiplex system for the analysis of membrane binding characterized by measuring the wavelength shift based on the LSPR or EOT. Understanding the mechanisms of the formation of lipid bilayer membranes on solid surfaces is important in further development of patterned membrane arrays and membrane protein microarrays. Microarrays consisting of immobilized molecules as spatially-indexed “probes” are used to screen the concentration and binding affinity of target molecules. In principle, if the target concentrations are known, the affinity of the target for different probe microspots can be estimated simultaneously. Conversely, given the known affinities of different target molecules for each probe microspot, the observed binding response may be used to simultaneously estimate the concentrations of multiple analytes in the sample. The attractiveness of microarray technology lies in the ability to obtain highly multiplexed information using small amounts of sample. Microarrays containing components of the cell membrane then provide an attractive platform for efficiently studying fundamental aspects of molecular recognition at the cell surface. These studies are of immediate relevance to diverse applications ranging from drug discovery to the detection of pathogens. The ability to pattern arrays of lipid bilayers has become almost essential for performing multiplexed, high information-content assays. To form a patterned biomembrane microarray, it is essential to control the formation and composition of the supported membrane in a well-defined geometry. The geometric arrangement of barriers to lipid diffusion constrains the membrane and determines its topology. Different lipid compositions and proteins can be deposited or flowed onto individual microspots in the array, thus producing a mosaic pattern of fluid membrane each with a different composition. The charged membrane components can then be rearranged by introducing an electric field across the membrane array. This provides further information about the clustering state of molecules in the membrane and quantitative analysis of transient interactions between membrane components. A variety of techniques have been developed for the patterning of lipid bilayers on a variety of substrates to yield homogeneous arrays consisting of the same lipid composition or as heterogeneous arrays containing different biomolecules.54,239,499 The homogeneous arrays provide a platform useful for screening the binding of different ligands to a membrane at various concentrations while the heterogeneous arrays can be applied to explore the role of different lipid compositions on the binding of ligands in a high-throughput manner. Several multistep, indirect methods have been developed to fabricate physical, chemical, and/or electrostatic barriers prepatterned on the substrate surface to restrict lateral diffusion and mixing of lipid molecules between confined areas.44,54,499 Many of these methods circumvent the requirements for a hydrated environment by employing prepatterned surfaces. Typically, patterns of barrier materials are deposited onto substrate surfaces using controlled deposition techniques such as photolithography,500,501 e-beam lithography, and

Functionalization of a liposome can also be achieved by the noncovalent host−guest interaction approach. This postfunctionalization of a liposome or a lipid bilayer is mediated by incorporating organic macrocycle scaffolds into the membrane. Organic macrocycles such as cyclodextrin, calixarenes, pillarenes, and cucrubiturils can be readily functionalized, and their host cavities can be exploited as receptors for sensing and drug delivery applications.474 Among these organic macrocycles, cyclodextrins (CDs) are a class of cyclic oligosaccharides with 6−8 glucose units forming a toruslike shape. CDs have a hydrophilic exterior and a hydrophobic cavity that form a complex with hydrophobic guest molecules. A broad range of guest compounds such as adamantane, azobenzene, ferrocene, or tert-butylbenzene derivatives can be encapsulated into the hydrophobic cavity.475−477 Synthetic amphiphilic β-CDs have been incorporated into the liposomes478,479 and serve as artificial receptors for postfunctionalization with DNA, carbohydrates, and peptides derivatized with adamantine.479−482 The orthogonal interactions of amphiphilic cyclodextrin host-hydrophobic guest systems and proteins have then been employed to form multilayered liposomes on gold substrate.355 The CD-liposomes were decorated with admantane-derivatized mannose or biotin which allows further interaction with lectin (ConA) and streptavidin, respectively. This host−guest-assisted vesicle immobilization offers a facile approach to microarray fabrication of multilayer glycoliposomes as versatile mimics of artificial tissues. 3.3. Micro- and Nano-Arrayed Lipid Bilayers

The formation of lamellar lipid bilayers on planar supports have significantly advanced our understanding of the physicochemical properties of membranes and their role in modulating molecular interactions and function in a membrane environment. However, native membranes also exhibit diverse nanoscale structures and topologies, and characterizing membranes of different nanostructures has been greatly facilitated by advances in nanofabrication technology which configure the lipid vesicles or lamellar lipid bilayers on a substrate of different nanostructured scaffolds.373,483−487 These model membranes formed on various nanofabricated substrates as exemplified in Figure 10 enable the role of different underlying membrane morphologies that modulate the physical properties and lipid organization to be characterized and allow the molecular basis of membrane structural organization on molecule interactions. Integration of micro- and nanostructures as sensing platforms with nanoplasmonic and zero-mode waveguides then enhance both the sensitivity and lateral resolution of dynamic changes in membrane structures. Nanoparticles covered with supported lipid bilayers which combine the intrinsic properties of metal, oxide, and metal oxide particles with that of lipid bilayers has significant potential in both basic and applied science.488 By uncoupling the inorganic surface from the surrounding aqueous phase, nanoSLBs provide a natural environment for biomolecules, eliminating or reducing possible problems of nonspecific adsorption or protein denaturation, known to be critical in the development of biomaterials. SLB-coated nanoparticles provide numerous possibilities for functionalization for the development of nanovectors with specific molecular targets as well as for carrying and delivering molecules to a selected site. In addition, strategies developed for silica nanoparticles are directly applicable to core−shell silica nanoparticles, such as lipid-modified quantum dots,489,490 extending the range of 5426

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

microcontact printing502−504 which modifies the solid substrate. Lipid vesicles readily adsorb and fuse into a continuous supported membrane over the prepatterned surface, where they do not form a fluid membrane on the barriers. Barrier materials have included metals (Au, Al, Cr, and Ti) and metal oxides (Al2O3 and TiO2),505 photoresists,506 proteins,507 polymers,243 and even photopolymerizable lipids,508 but simple mechanical scratches44 have also proven useful. In addition, soft lithography of poly(dimethylsiloxane) (PDMS) stamps have been applied to direct patterned deposition/stamping of a preformed membrane onto a substrate. Alternatively, polymer lift-off509 or blotting away material510,511 from the lipid membranecoated surface have also proven successful in generating a wide range of membrane compositions on very high spatial densities (>106 corrals/cm2). Alternative methods involved using a spatially directed illumination of preformed supported lipid bilayers by short-wavelength ultraviolet radiation.512,513 Patterns of hydrophilic voids were formed within a fluid membrane as well as isolated membrane corrals over large substrate areas. These voids can be refilled with bilayers of different lipid and protein compositions via vesicle fusion, thereby providing a synthetic means for probing 2D reaction-diffusion processes. By formulating the lipid and protein compositions, membrane protein arrays can be fabricated forming functional microdomains in well-defined patterns. The photochemically patterned high-density membrane array combined a region facilitating transmembrane proton transfer with a surrounding reporter region containing a pH sensitive fluorescence probe such as FITC. By integrating optical transduction with ionchannel transport through the bilayer-embedded gramicidin, spatial separation of the reporter and the transporter membrane regions complemented the measurement of electrical transduction across the membrane. These types of patterned arrays of stable, uniform, and laterally mobile SLBs allow the parallel and stochastic analysis of membrane transport. Furthermore, the substrate can be partitioned into regions of different hydrophilicities by using electron microscopy grids to laterally control the extent of plasma oxidation.514 Addition of vesicles of different compositions results in supported fluid lipid monolayers, adsorbs intact lipid vesicles, or supports fluid bilayers separated by regions devoid of membrane. In all cases, a fluid membrane is partitioned into compartments, which are separated by barriers restricting the lateral diffusion of the lipids. The methods requiring substrate prepatterning depend on the prior deposition of exogenous materials on the substrate surface to form single, permanent patterns.54 Moreover, the mechanisms by which barrier materials compartmentalize membrane patterns remain poorly understood. For instance, while aluminum oxide resisted vesicle spreading, deposition of lipids, albeit immobile, was observed for chrome, gold, and indium−tin oxide surfaces. On the other hand, methods based on PDMS stamps require optimization of the contact time and associated contact pressure for different lipid compositions. Arraying membrane proteins has also been achieved by printing mixtures of the protein and associated lipids using appropriate surface chemistry for the immobilization of lipids. For example, cell membrane preparations containing G proteincoupled receptors (GPCRs) from a cell line overexpressing the receptor was used directly for fabrication of GPCR microarrays. Multiple arrays of GPCR-containing cell membranes were printed on a γ-aminopropylsilane (GAPS)-derivatized surface with rapid kinetics while maintaining the lateral fluidity and

significant mechanical stability after multiple withdraws through an air−water interface without lipid desorption.515 The advent of DNA microarray technology and the development of several printing technologies has also greatly facilitated the fabrication of membrane microarrays. Quill- and solid pin printing have been used for fabricating GPCR-membrane microarrays. Quillpin printing method offers advantages for its high-quality and low sample usage of ∼0.5 nL or less per data point from a single insertion of the pin into the GPCR sample to yield several hundred microspots. The functional activity of the GPCRs such as the human neurotensin receptor, adrenergic receptor, and dopamine receptor was further demonstrated by the specific binding of their corresponding fluorescence-labeled cognate ligands.515,516 The GPCR microarray can therefore contribute to lead identification and validation stages of drug discovery. In addition, the GPCR-G protein complex is largely preserved in the microspot which may provide potential in parallel studies of ligand binding and the downstream activation of G proteins. 3.4. Naturally-Derived Model Membrane Systems

Studying biological membranes directly from the cells of living organisms is the major objective to fully understand the physical and chemical properties of lipid and protein components in controlling the structure−function relationship and the molecular interaction of biomembranes. However, exploring these physicochemical relationships within the native membrane in vivo is challenging. This approach presents several difficulties involving the inherent lack of control of cellular process associated with highly dynamic molecular exchange and turnover of molecular components in native membranes. In addition, the inner leaflet of the cellular membrane and intracellular organelle membranes are inaccessible to molecules in solution, and special approaches to transfer the molecules into the cell can alter the membrane properties. The complex phase and structural polymorphisms can also complicate the extraction of information for molecular interactions related to particular membrane structure and organization. Therefore, in order to study the membrane in a simple format, either top-down or bottom-up approaches are used to construct model membranes for biophysical and biochemical analysis of native membrane mediated cellular processes. Complex native membranes are deconstructed into simpler systems in top-down approaches. This can be achieved by mechanical disruption (homogenization) of cells or tissues followed by density gradient differential centrifugation to isolate the plasma membrane or organelles or by lipid extraction using organic solvent and detergent followed by reconstitution of the extracted lipids into membranes.517 The isolated or reconstituted membranes can then be stabilized in close proximity to the surface of sensing substrates via physisorption or chemisorption approaches. However, fundamental drawbacks of these approaches involve the contamination of membrane components from different organelles, loss of bilayer asymmetry and lateral structural domain organizations, adverse effects from organic solvent and detergent residues on the membrane properties, and conformational and functional changes in the membrane proteins. In order to apply the surface sensitive SPR and DPI technologies to characterize the membrane properties and its related biochemical processes, the native membrane can be coupled to a planar substrate in an orientation-selective 5427

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Figure 11. Formation of natural membranes on solid supports. (A) Erythrocyte membranes (labeled with red) in an inside-out orientation are selectively spread on the cellulose regions micropatterned on a glass support. Different wetting conditions for membrane patterning can be created by microcontact printing of BSA (green) on the cellulose cushion. Reproduced with permission from ref 517. Copyright 2008 American Institute of Physics. Reproduced from ref 519. Copyright 2004 American Chemical Society. (B) Natural membranes with expressed membrane proteins of interest are suspended across nanoaperture arrays of 500 nm holes with 5 μm periodicity in a 500 nm thick SiN film. The orientation of membrane proteins in the cell membrane sheets relative to the microstructured substrates is controlled by membrane transferring protocols. The dimension and configuration of the nanoaperture arrays provide access for analytes from the pore side as shown by the labeling of A488-labeled wheat germ agglutinin and GR-Cy3 to extracellular membranes of HEK293 expressing neurokinin-1 receptor and homopentameric 5HT3R receptor. Reproduced from ref 522. Copyright 2006 American Chemical Society. (C) The formation of bacterial outer membrane supported bilayers by the adsorption of outer membrane vesicles and their fusion into a bilayer induced by the coadsorption of PEG-liposomes. The temporal mass changes showed two-regime kinetics at low (left top) OMV coverage and single-regime kinetics at high OMV coverage (left bottom). Reproduced with permission from ref 523. Copyright 2016 Springer Nature.

manner. The preparation of natural membrane fragments on solid supports has been achieved by spreading human erythrocyte ghost and sarcoplasmic reticulum vesicles onto a 5−10 nm thick hydrated cellulose cushion on glass to form a defect-free polymer-supported native membrane with the cytoplasmic domain exposed to bulk solution518,519 (Figure 11A). The formation of a defect-free membrane was facilitated by adjusting the wetting properties of the polymer cushion, whereas a low surface coverage of an inhomogeneous cell membrane patch that adhered strongly to the solid substrate was obtained when a highly charged polylysine polymer cushion was used. Importantly, the orientation and lateral disruption of transmembrane proteins were maintained in their native state. A native cell membrane can also be transferred to

the substrate, for example, by pressing a polylysine-coated glass to the apical parts of HEK-293 cells followed by removing the glass to rip large regions of the plasma membrane.520,521 The dimension of the resulting transferred cell membranes sheets were typically several hundred square micrometers attached onto the substrate. The fluidity of both leaflets and protein compositions were conserved in these membrane fragments, and the receptors of interest expressed on the cell surface retained their orientation in the native membrane with the cytoplasmic domain exposed to the bulk aqueous phase. However, these native membranes were constrained by the substrate which prevented access to both membrane leaflets, and only one side of membrane proteins was accessible to ligand binding. 5428

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

or DMSO.526 However, the GPMVs induced by organic solvent are still attached to the cell. The formation of GPMVs has also been induced by incubating HeLa cells with the cysteinecontaining pro-apoptotic peptide [KLAKLAK]2-conjugated to the cell-penetrating peptide octa-arginine (R8) and its analogues.532,533 Since the heterogeneous size of GPMVs is beyond the sensing fields of SPR and DPI-based technology, size reduction of GPMVs or spreading of the GPMVs onto the substrate as SLBs are required to be compatible with these surface-sensitive methods. The surface of the GPMVs are generally rendered with negatively charged silica acid residues, which makes them poorly adherent to bare silica/glass substrates. The spreading of GPMVs onto the SLBs on a planar glass/silica substrate can therefore be assisted by the coadsorption of pure liposome with a composition similar to the host cells. This approach has been successfully demonstrated by the formation of viruslike SLBs containing viral transmembrane proteins, such as hemagglutinin (HA) and feline aminopeptidase N (fAPN).242,534 In contrast to the inside-out orientation for the cell spreading and detaching cell membrane approaches, this liposome-assisted GPMVs-SLB transition resulted in the exposure of extracellular protein domains to the bulk solution while the intracellular domains interact with the substrate surface. These viral proteins retained their activity in the SLBs and allowed quantitative determination of kinetic rate constants for intermediate steps in the viral fusion process.242,534 However, these viral receptors have very limited diffusivity in this viral-like SLB with a lipid diffusion coefficient of around 0.25−0.35 μm2/s. This limited mobility of membrane proteins can be solved by introducing polymer cushions between the substrate and the bilayer. A stepwise adsorption of GPMVs followed by PEG-containing liposomes to the glass thus resulted in the formation of SLBs.535 The lipid molecules diffused similarly to those found in the above method, while the proteins became significantly mobile with varied diffusion coefficients for different proteins ranging from 0.5 to 0.75 μm2/s as determined by the single particle tracking method. The transmembrane proteins and GPI-anchored membrane proteins also retained their orientation with the extracellular domain exposed to the bulk medium, and the process of maintaining this bilayer orientation via vesicle rupture and spreading was mediated by the so-called parachute mechanism.536,537 The isolation of native membrane vesicles from a bacterial cell can also be used to develop SLBs with similar molecular components to the native bacterial membrane. Outer membrane vesicles (OMV) are spherical vesicles of 20−250 nm in diameter naturally secreted from G(−) bacteria during cell growth as the membrane blebs pinch off the bacterial outer membrane surface.538,539 Similar to the formation of SLBs via the sequential adsorption of GPMVs and PEG-liposomes on glass substrates, the formation of a native bacterial outer membrane (OM) supported bilayer has also been achieved with a stepwise adsorption of OMV followed by the addition of PEG-liposome to induce vesicle-bilayer formation on a glass substrate523 (Figure 11C). The resulting SLBs have their extracellular leaflet exposed to the bulk aqueous phase via a parachute mechanism. This native bacterial OM SLB was then further applied to characterize the antibacterial mechanisms of a cationic peptide, Polymyxin B, involving membrane-disruption kinetics and changes in membrane mechanical properties. The combination of these bacterial OMVs isolated from different growth conditions and bacterial strains into the nanohole array

The activity of membrane proteins can be also impaired by direct physical interaction with the solid substrate. In order to preserve the intrinsic properties of the membrane components and allow both the intracellular and extracellular domains to be fully accessible to molecular interactions, the native membrane can be spread across nanoapertures in planar supports in a defined orientation. This approach has been demonstrated by pressing silicon nitride (SiN) nanopore arrays with 50−600 nm aperture diameters across 100−500 nm thick SiN films with a periodicity of 2 and 5 μm to a layer of cultured HEK-293 cells expressing either ligand-gated ion channel (5HT3 receptor) or GPCR (neurokinin-1 receptor)522 (Figure 11B). Although the orientation of both transmembrane proteins and membrane leaflets are retained in this supported membrane, the low number of receptors within the nanoapertures were not detectable by fluorescent-antibodies and it is also crucial to completely cover all apertures to seal membrane for ion channel function. In addition, a cumulative measurement from multiple pore-spanning membrane areas rather than from a single transporter confined in each nanopore was obtained for the transmembrane response, potentially negating the advantage of creating single nanopores. In addition to coupling whole cells or isolated organelles to a substrate, membrane vesicles derived from native cells provide an alternative native membrane system to study membrane function in its native form. Giant plasma membrane vesicles (GPMVs) or cell membrane blebs, are spherical native membrane protrusions that form from the plasma membrane during normal physiological processes including cell locomotion, cytokinesis, and apoptosis.524,525 GPMVs produced naturally or induced by chemical or physical means are heterogeneous in size range from ∼50 nm up to 10 μm and provide an excellent model membrane system to explore the structural properties and functional activity of membrane domain organization.526−529 GPMVs have advantages over synthetic liposomes in that the orientation and assembly of complex membrane proteins are preserved in the plasma membrane vesicles. Expressing the membrane proteins of interest in the cells and collecting the GPMVs enriched in the expressed membrane proteins can also be easier than the arduous reconstitution approaches with purified membrane proteins in preparing proteoliposomes. More importantly, the GPMVs are free of cytoskeleton and organelles and the organization of lipid molecules into micro/nanodomains observed in native membranes is maintained. Due to these features of conserved membrane properties, GPMVs are therefore of great interest in biophysical studies as intermediate model systems that can be integrated with surface sensitive techniques for the quantitative analysis of physicochemical properties of coexisting membrane phases, protein−membrane domain interactions, membrane protein assembly, viral fusion, and as delivery vehicles for cell-specific targeting. GPMVs can be artificially induced by treating cells with several chemical methods such as dithiothreitol (DTT) and paraformaldehyde (PFA).526,528 However, several artifacts can be induced with this approach such as nonspecific cross-linking of lipids and proteins by the aldehyde, reduction of protein disulfides and thioesters leading to depalmitoylation, and specific coupling of phosphatidylethanolamine to proteins. Thus, non-cross-linking vesiculants, such as N-ethylmaleimide (NEM), have been used to circumvent the cross-linking/reduction artifacts induced by DTT/PFA. 530,531 GPMVs can also be induced by a combination of polar organic solvents, such as ethanol, acetone, 5429

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

and to study the kinetic mechanisms of vesicle-to-bilayer transformation. The most widely used approach to characterizing the structure of a deposited membrane is the determination of the thickness and will be discussed in this section. The effect of physical properties and modifications on the formation of lipid monolayers and bilayers on metallic Au and Ag substrates has been monitored by SPR.206,207,357 As SLBs do not readily form on the Au substrates used in SPR sensing systems, the Au films deposited on dielectric supports or gold particles are commonly modified with SiO2, thiolated SAM with different end groups, and various polymers. To apply SPR sensing to characterize the geometric structure of the bilayer and to monitor the kinetics of adsorption processes, methods have been established for the quantitative measurement of thickness or surface concentration of the adsorbed lipid bilayer.96,238,263,550 While SPR is most commonly used for the quantitative analysis of kinetics, affinity, and active concentration of molecule interaction, fitting the SPR response to a given layer structure is limited to only a few examples involving quantitative mathematical formalism for interpretation of SPR signals from adsorbed films on a wide variety of substrates.96,238,551−553 The angular position of the SPR spectral minimum is a function of the physical layer thickness and the refractive index of the molecular layers formed on the substrates. The thickness of lipid bilayers has been determined by fitting the SPR reflectivity curve to a theoretical model based on the Fresnel’s equation using WINSPALL.552,554,555 This requires knowledge of the refractive index (n) and the film thickness (d) of each dielectric slab of the multilayer system under investigation. However, the exact refractive index of the compound forming a layer on the metal surface is required to calculate the thickness values, and the approximations for refractive index values used in studies of lipid bilayers can vary from 1.45 to 1.50. For example, an RI value of 1.45 was assumed at a wavelength of 750 nm in an SPR analysis of a lipid monolayer deposited on hydrophobic thiol as hybrid lipid bilayers (HLB) and SLBs,556,557 while RI values of 1.50, 1.49, and 1.493 have also been assumed for determination of the thickness of SLBs.552,553,558−561 The resulting thickness values of the bilayer can vary by about 10 Å between RI values of 1.45 and 1.50.551 Notably, the thickness determined by SPR is also influenced by the decay length which can vary with the distance between lipid bilayers and substrate surface altered by chemical modifications. Thus, the calculated thickness is an exponentially weighted average thickness for vesicles, which underestimates the mass of the adsorbed lipid molecules. This curve fitting program was nevertheless used to characterize the equilibrium bilayer structures in terms of the effects of different Au modifications on vesicle adsorption and bilayer formation. Thickness values around 4.0 nm were determined for eggPC bilayers deposited on an Au film coated with different SiO2 thicknesses from 3.3 to 50 nm.555,562 In contrast, the adsorption of eggPC vesicles to a −COOH-terminated SAM on Au formed a layer with a thickness of 7.8 nm, while the thickness reduced to 3.7 nm after removing adsorbed vesicles by a Milli-Q water rinse. The adsorption of eggPC vesicles onto the hydrophobic −CH3-terminated SAM formed a lipid monolayer about 2 nm. On the other hand, vesicle adsorption onto a surface without rupture was evident from a thickness value of 7.8 nm obtained for eggPC vesicles adsorbed on an −NH2-terminated SAM.555 The effect of hydrophilic polymer modifications on the equilibrium thickness of lipid bilayers has also been shown for

may provide a valuable system with broad application in, for example, probing drug responses to various physicochemical properties of bacterial membranes. In addition to the GPMVs, other naturally secreted membrane vesicles derived from endosomal and plasma membrane origin such as exosomes and microvesicles with 50−100 nm in diameter are actively secreted by mammalian cells.540−543 In particular, there has been a strong focus on the use of exosomes that are shed from various cancer cells carrying molecular information related to cancer development and the potential of biomarkers or targets for diagnosis and treatment.543,544 Other potential uses of exosomes involve the development of platforms for screening diverse disease markers and providing simple native membrane systems for manipulating membrane proteins in a native cell orientation and conformation for molecular interaction analysis.545 Arrays of periodic nanoholes patterned in a metal film functionalized with affinity ligands for different exosomal protein markers have also been developed for the detection546 and profiling of exosome surface proteins.545 On the basis of light transmission through periodic plasmonic nanoholes, the spectral shifts and intensity changes were monitored in real time for quantitative analysis of the target exosomal marker proteins.

4. MAJOR APPLICATIONS OF OPTICAL BIOSENSORS IN BIOMEMBRANE RESEARCH 4.1. Characterization of Membrane Structural and Physical Properties

4.1.1. Membrane Thickness. As described in previous sections, artificial membranes modified with specifically tailored function can be deposited onto sensing substrates with different architectures to mimic particular types of natural biomembranes. The successful application of these membrane systems to fundamental studies of structure−function relationships strongly depends on robust and consistent membrane fabrication methods to allow precise control of biomembrane assembly and function and together with their interface with supporting surfaces. Although model membrane systems are designed to simplify the complex nature of membrane composition, physical properties, and the control of protein activity, there is no universal standard for in vitro membrane systems that can be used for all applications. Indeed, the use of several model systems in an orthogonal manner to study individual phenomena can provide significant insight into membrane function. However, the structural properties of supported lipid membranes vary greatly depending on the constituent lipids and the nature and geometric structures of the underlying substrates and the physicochemical properties of any modifications. Overall, therefore, the systematic characterization of membrane properties and the corresponding lipid− substrate interactions are critical in validating all surfacesensitive measurement approaches (refs 31, 128, 197, 351, 372, 373, and 547). In particular, the formation of complete, defectfree bilayers is essential for the structural characterization of the membrane and molecule-membrane interactions without interference from nonspecific binding to the substrates. The formation of planar bilayers free of organic solvent residue via vesicle adsorption-bilayer transformation process on the sensor surfaces is commonly used in microfluidic systems integrated in conventional SPR and DPI systems.167,187,207,548,549 Various analytical approaches have been established to characterize the geometric parameters of the bilayers formed on the substrates 5430

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

estimated between 100% DOPC and 100% EPC SLBs from the sensor surface, with an assumed bilayer thickness of 4.91 nm for all bilayers composed of DOPC and EPC. SLB formation via vesicle adsorption onto polymer layers of zwitterionic sulfobetaine methacrylate (SBMA) grafted from monolayers of dithiodiundecane-11,1-diylbis[2-bromo-a-methylpropanoate] or (3-(2-bromoisobutyryl)-propyl)-dimethylchlorosilane chemisorbed onto a gold or a glass surface at different atom-transfer radical polymerization (ATRP) conditions have also been characterized by SPR.554 A thickness of 4.8 nm was obtained for a DOPC bilayer formed on poly(SBMA) by fitting the kinetic profile of SPR reflectivity for the adsorption of DOPC liposomes to a 16.8 nm poly(SBMA) layer polymerized for 15 min. However, the adsorption of DOPC liposomes to this thin poly(SBMA) layer was greatly reduced in the absence of NaCl. Furthermore, no DOPC membrane formation was observed on a thick poly(SBMA) of 35 nm after polymerizing for 30 min both in the presence or absence of NaCl, despite its more hydrophilic properties shown by a smaller water contact angle. This phenomenon can be attributed to the lower density of the zwitterionic groups in thinner poly(SBMA) layers formed after shorter polymerization times. The structural characteristics of the SLBs on poly(SBMA) can thus be tuned by controlling the ATRP conditions for different polymer thickness and surface charge density. It has also been shown that the thickness and RI of thin molecule layers can be derived from the SPR reflectivity-angular profiles measured at a single wavelength in two different bulk media such as air and water with known RI values or measured at two different wavelengths in the same bulk medium.70,566−569 The thickness and RI of lipid monolayers and multilayers on the substrates have also been determined by employing three wavelength SPR measurements without the use of assumed RI values in a single wavelength analysis as stated above.570 By measuring the angular SPR reflectivity changes in two different media with a sufficiently large difference in RI, a unique solution of thickness and RI for the molecular layers can be obtained from the intersection point of the two continuum solutions of thickness versus RI. A similar but more complex approach based on the linear approximation of RI dependence on the wavelength was used to find a unique solution for thickness and RI from the intersection point of the two continuum solutions.570 The thickness-RI continuum solutions for the lipid layer deposited on SPR sensing substrates can also be obtained by fitting the SPR angular profiles with a multilayer model using WINSPALL program. These approaches have been applied to the measurement of the thickness and RI values of both hydrogenated eggPC (HSPC) and stearic acid (SA) monolayers and multilayers deposited onto an Au substrate via the LB technique at different compression pressures. However, the experimental errors in the intersection points in the thickness-RI continuum results in a relatively high uncertainty in the experimental values leading to an overestimation in the three-wavelength analysis for a single-deposited monolayer of both HSPC and SA.570 Despite this apparent deviation in thickness values, a linear dependence of thickness on layer number was observed for SA monolayer and multilayer using the three-wavelength analysis. An SA thickness of 2.5 nm obtained from averaging five or more deposited SA layers was in good agreement with a thickness of 2.6 nm derived from a two-medium analysis. These values obtained from either twomedium or three-wavelength analysis correspond well to a

supported membranes composed of neutral POPC and cholesterol at various molar percentages in negatively charged stearoyl-oleoyl-phosphatidylserine (SOPS). The deposition of 100% SOPS SUVs to a positively charged poly(diallyldimethylammonium chloride) (PDDA) polymer adsorbed onto a 11-mercaptoundecanoic acid (MUA)-SAM modified Au substrate formed a stable bilayer of 3.2 nm.552 This thickness value for SOPS was less than that characterized by X-ray and neutron reflectivity for bilayers with the same acyl chain length. In contrast, the zwitterionic POPC SUVs formed vesicle layers or multilayer aggregates of 140 nm on the PDDA/MUA surface.552 It has also been shown that the thickness increased from a single bilayer of about 4 nm to that of multibilayers as the molar percentages of POPC increased above 25% in SOPC. In addition, an intermediate stage for liposome adsorption was noticed as a kink in the adsorption profile of reflectivity versus time for lipid mixtures with POPC at or above 50%. The effect of cholesterol on the thickness of SOPC bilayer formed on the PDDA/MUA was also characterized by a curve-fitting approach based on the Fresnel’s equation. The thickness of the SOPS bilayer increased linearly from 3.2 to 4.5 nm with increasing cholesterol composition up to 20%.551 The formation of multilayers or unruptured vesicle layers with thickness values larger than 5.3 nm was also observed for SOPS vesicles mixed with more than 25% cholesterol. A similar kink in the kinetic profile for an intermediate stage of vesicle adsorption was also observed as the cholesterol content increased above 25%. The number and thickness of hydrated multilayered polymer cushions also influence the equilibrium thickness of bilayers. Lipid membranes cushioned by layer-bylayer polyion polymers adsorbed onto the alkylthiol SAMs (PDDA/PSS/PDDA/MUA) on Au substrates characterized by SPR showed increasing bilayer thickness with increasing POPC content in SOPC which showed a similar trend to that observed with the single PDDA polymer layer.563,564 However, the thickness values of bilayers on the layer-by-layer polymer cushion is larger than those on corresponding lipid bilayers formed on a single PDDA polymer. This increased thickness can be attributed to lower electrostatic attraction between the lipids and the polymers with reduced charge density and less interdigitation of lipid molecules in the bilayer on the layer-bylayer polymer cushion. The effect of electrostatic interaction between lipid molecules and substrates on the final SLB properties has also been demonstrated by LSPR analysis in combination with QCM-D.565 Although bilayer thickness was not determined by LSPR, the electrostatic effect on the decoupling distance of SLBs to a silica-coated Au nanodisc with an average height and diameter of ∼20 and ∼100 nm, respectively, was probed in accordance with the LSPR measurement principle where the LSPR-induced electromagnetic field enhancement is a function of decay length of plasmonic evanescence. The kinetic profile of the redshift in Δλmax was tracked for the process of SLB formation and its final SLB properties. The stronger interaction between increasing positive charge on the surface of palmitoyl-oleoyl-glyceroethylphosphocholine (EPC)/DOPC liposomes resulted in a reduction in the hydration mass associated with the SLB for a thinner hydration layer between the SLBs and the silica substrate.565 On the basis of the distribution of EM field intensity as a function of the distance from the sensor surface approximated by an exponential decay model with a characteristic decay length, a variation of 1.79 nm in distance was 5431

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Figure 12. Temporal changes in the SPR angle for the deposition of 5 different lipid vesicles onto SiO2 surface, PEG-SAM, Dex-6 kDa substrates (Left to right) measured at a wavelength of 670 nm. Lipid compositions are colored as follows: EggPC (red), EggPC+PS (black), EggPC+PS+Chol (blue), EggPC + Label (orange), and HepG2-extact (green). SPR is sensitive to the refractive index of the lipids (density and packing), as well as to the effect of the supporting surface structure on the evanescent field. This makes it relatively difficult to differentiate whether a bilayer or vesicle layer is formed. However, measurements with the label clearly show a tethered vesicle layer formed on the PEG-SAM and SLBs on the other two surfaces. The formation of HepG2 SLBs on the Dex-6 kDa surface induced a larger shift in resonance angles due to the presence of cholesterol and other nonlipid components such as proteins with higher RI and mass. Reproduced from ref 571. Copyright 2014 American Chemical Society.

bilayers of 5.6−6.4 nm observed for eggPC/PS/Chol. The obtained RI values between 1.44 and 1.47 at different wavelengths also indicated that these liposomes formed dense layers on the Dex6 substrate. The curve fitting approach using the Fresnel equation for the determination of thickness and RI of the membrane on the sensor surface as either bilayer or vesicles is generally not applicable to SPR measurements from most commonly used commercial SPR systems. The properties of the membrane formed on various commercial sensor chips are only assessed qualitatively by the SPR signals given in response units (RUs) and the mass of the adsorbed molecule layer by the approximation of 1RU for 1pg/mm2 of proteins and lipids. A variety of the commercial sensor chips with different modifications on the Au substrates have been used in the formation of bilayers or vesicle layers for studying moleculemembrane interactions.45 The hydrophobic alkanethiol SAMmodified Au sensor substrate known as the HPA chip is generally used for the formation of a lipid monolayer on alkylSAMs as a hybrid bilayer membrane (HBM) via the fusion of liposomes. Response ranges from 2000 to 2800 RUs have been reported for monolayer formation of various lipid compositions on the HPA chip,573,574 and these hybrid bilayer surfaces have been used to study many aspects of protein/peptide interactions with a membrane. HBMs modified with bioaffinity groups by the fusion of SOPC liposomes consisting of the metal-chelating lipid, Ni-NTA-DODA, produced an increase of 2800 RUs corresponding to a planar monolayer.574 The high surface coverage of stable Ni-NTA tagged planar HBMs was further demonstrated by the specific docking of (His)6-tagged proteins to the membrane surface. However, HBM systems are limited in their application due to the restriction of peptide/ protein insertion into the membrane and the integration of transmembrane proteins in an active form. Model membranes have also been commonly prepared on the commercial L1 chip, a lipophilic compound-modified carboxymethoxy-dextran on Au, for the stable capture of intact

theoretical length of 2.5 nm of a SA molecule. In contrast to the linear dependence of thickness on the layer number, a nonlinear relationship between RI and the layer number was obtained using the three-wavelength analysis for SA layers. Thus, three-wavelength SPR analysis provides an efficient way to determine the real thickness but only provides the apparent RI values, while the two-medium analysis can be used to probe the real RI and apparent thickness of the molecule layers formed on the substrate. Biomolecule-membrane interactions are defined by a combination of geometrical structural parameters and physical properties of bilayers, and the effect of membrane morphology on the sensing substrate (as either SLBs or lipid vesicles) on membrane interactions can be readily explored using membrane thickness as a tool to characterize the membrane prior to interaction studies. As discussed in earlier sections, an understanding of discrete conditions for the formation of bilayers or vesicles on various substrates facilitates the finetuning of the fabrication method for a particular membrane morphology. Differentiation between the SLBs and vesicles on the substrate can be achieved by characterizing the thickness and density of the membrane layers on the substrates. In addition to fitting the SPR angular profile using the Fresnel equations and matrix formalism with an isotropic multilayer model, the thickness and density of the lipid layers formed on the substrates can also be determined by using a dual wavelength SPR approach.571,572 As shown in Figure 12, different membrane morphologies have been selectively controlled by adsorbing liposomes of different compositions onto an Au substrate modified with either low molecular weight acylated dextran (Dex6) chemisorbed on thiolated SAM or hydrophobic alkyl group-modified PEG-thiolated SAMs (PEGSAM).571 On the basis of the thickness values for the lipid layers obtained from the optical measurement of dualwavelength SPR, liposomes composed of eggPC, eggPC/PS (75:25), and eggPC/PS/Chol (70:25:5) all formed bilayers of 4.7−5.1 nm on the SiO2 and Dex6 substrate with thicker 5432

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Table 2. Structural Properties of SLBs Calculated from the TM and TE Phase Changes Obtained by DPI lipids 14:0/14:0 DMPC DMPC/DMPG DMPC/DMPG/Chol DMPE/DMPG DMPC/DMPS DMPC/DMPS/PI(4)P DMPC/DMPS/PI(4,5)P2 16:0/18:1 POPC POPC/POPS POPC/POPS POPC/POPG POPC/POPE/POPS POPC/POPE/POPS/PI POPC/POPE/POPS/PI POPC/POPE/POPS/PI(4,5)P2 POPC/POPE/POPS/PI/CL 18:1/18:1 DOPC DOPG DOPC/DOPE/DOPS DOPC/DOPE/DOPS DOPC/DOPS DOPE/DOPG undefined acyl chain SoyPC/DOPE/Chol LPS on DOPE/DPOG E. coli extract

ratio

thickness (Å)

birefringence

reference

100 80:20 64:16:20 80:20 80:20 76:20:4 76:20:4

45.2 47.6 34.6 43.7 47.9 47.3 5.09

± ± ± ± ± ± ±

0.7 1.6 0.6 5.0 0.7 2.1 1.4

0.0217 0.0235 0.0164 0.0244 0.0218 0.0226 0.0229

0.0006 0.0015 0.0003 0.0028 0.0006 0.0009 0.0005

4.53 4.77 3.47 4.34 4.82 4.73 5.12

± ± ± ± ± ± ±

0.08 0.16 0.07 0.49 0.07 0.22 0.15

596, 597, and 606−608 596, 597, and 606−608 597 597 597 and 601 601 601

100 80:20 75:25 80:20 60:30:10 50:30:10:10 56:10:30:4 56:10:30:4 48:20:10:10:4

48.7 46.9 38.1 45.4 46.9 44.8 46.1 48.6 48.9

± ± ± ± ± ± ± ± ±

1.5 0.6 6.0 0.4 1.3 0.6 0.4 0.5 1.8

0.0185 ± 0.0007 0.0195 ± 0.0001 0.0177 0.0176 0.0176 0.0198 0.0191 0.0198

± ± ± ± ± ±

0.0011 0.0015 0.0005 0.0005 0.0003 0.0001

4.71 4.55 4.07 4.56 4.70 4.47 4.60 4.82 4.88

± ± ± ± ± ± ± ± ±

0.15 0.03 0.20 0.05 0.12 0.07 0.02 0.04 0.33

167 and 607 167 and 609 610 607 600 599 611 611 599

100 100 20:50:30 50:20:30 80:20 70:30 75:25

46.0 46.7 34.7 52.0 46.0 52.9 45.0 51.0

± ± ± ± ± ± ± ±

3.0 2.6 3.9 3.0 2.0 0.1 4.0 2.5

0.0170 0.0149 0.0124 0.0191 0.0172 0.0250 0.0158 0.0168

± ± ± ± ± ± ± ±

0.0010 0.0005 0.0025 0.0007 0.0004 0.0019 0.0002 0.0018

4.60 4.70 3.30 5.20 4.60

± ± ± ± ±

0.30 0.23 0.54 0.30 0.20

4.52 ± 0.04 5.09 ± 0.15

602 612 612 602 602 167 602 600, 613, and 614

60:30:10

46.3

0.0135

58.3 ± 1.3

0.0195 ± 0.0020

4.63 0.7 5.83 ± 0.07

615 613 597

liposomes or bilayer formation.385,575 Other chips including ProteOn GLM, GLC, and LCP chips, with a modified alginate polymer layer bound to the Au surface designed for general amine coupling, have also been used to prepare vesicle layers.576 These ProteOn chips are different in the architecture of the alginate polymer layer, which is an extended polymer matrix in GLM or a compact polymer layer in GLC. Modification of these alginates with alkylamine, such as undodecylamine with EDC/NHS chemistry, is required for vesicle capture. The LCP sensor chip is designed to capture DNA-labeled lipid assemblies through DNA hybridization to a biotinylated DNA tag onto a NeutrAvidin-functionalized planar SAM-Au chip. Validation of the final membrane configurations as a bilayer or as a vesicle is difficult solely from the maximum values and kinetics of RU changes. While these lipophilic alkyl chain modified sensor chips have been developed for the capture of liposomes and as suggested by confocal fluorescence imaging and leakage experiments,385,577−581 the fusion of liposomes to form bilayers on these chips have also been reported to be lipid composition dependent.575,582 The level of immobilization, defined as RUmax, is generally used as an indicator of lipid coverage and conformation. The complete coverage of the lipids on the chip surface is also assessed by the absence of BSA binding. The maximum RU of lipid coating varies greatly from 4000 to 14000 RU for zwitterionic liposomes with diameters of 50−200 nm.45,385,577,579 The deposition of negatively charged liposomes often results in a suboptimal low coverage of less than 3000 RU due to

± ± ± ± ± ± ±

mass (ng/mm2)

electrostatic repulsion.579 However, the coverage of a negatively charged membrane can be improved by the addition of 1−2 mM Ca2+ during the deposition and removal of Ca2+ by a pulsed rinse with 1 mM EDTA. The use of lipophilic alkyl groups to capture a membrane has a few limitations for the studies of some lipophilic drugs and membrane lytic molecules which can bind irreversibly to the lipophilic groups. This irreversible binding can lead to gradual loss in the liposome capture capacity and therefore prevents long-term sensor analysis of the binding of hydrophobic molecules to membranes. Furthermore, the irreversible binding of molecules to the lipophilic groups on the sensor chip also reduce the number of experimental cycles for reproducible membrane formation. For the particular case of a long injection of the pore forming peptide melittin, the chip surface started to deteriorate (in terms of consecutive lipid immobilization and distinct interaction dynamics) after ∼15 experimental cycles.583 Lipid membrane nanodiscs have been developed as a model membrane to incorporate transmembrane proteins in a lipid bilayer of defined nanosize with a specific protein−lipid stoichiometry. The protein-incorporated nanodiscs are monodisperse in solution and both intracellular and extracellular domains are accessible for ligand binding. The immobilization of lipid/proteolipid nanodiscs onto SPR chips have been used to explore complex peptide-membrane and ligand-membrane receptor interactions.584−594 The immobilization level of lipid nanodiscs, the stability of the captured ligands, and the degree of nonspecific analyte binding vary on different types of SPR 5433

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Table 3. Structural Properties of Vesicle Layer Either Physisorbed or Tethered onto the Silicon Oxynitride Surface Calculated from the TM and TE Phase Changes Obtained by DPIa lipids

a

thickness (Å)

RI

mass (ng/mm2)

reference

1.348 1.361 1.367 1.376 1.375 1.356 1.392 1.408 1.398 1.359 1.362 1.354

7.8

70:30 20:50:30

43.3 56.4 33.9 25.6 18.5 28.0 36.6 87.0 45.7 22.0 19.0 39.0

11.5 11.1 7.8 4.4 22.1 67.1 30.5 5.6 3.8 5.4

603 167 603 603 603 602 603 603 603 603 602 602

100 70:30 20:50:30

56 52 107

1.356 1.355 1.349

8.6 7.9 10.4

602 602 602

ratio

SUV DMPC

100

DPPC POPC DOPC

100 100 100

DPPS DSPA DSPG PE/PG/Chol DOPC/DOPS DOPC/DOPE/DOPS LUV DOPC DOPC/DOPS DOPC/DOPE/DOPS

100 100 100

The vesicle layers were formed by depositing either small unilamellar vesicles (SUV) or large unilamellar vesicles (LUV).

properties. These properties are determined by the degree of saturation and branching of the hydrophobic acyl chains and the spatial organization of various lipid components and cooperate in diverse ways to modulate membrane morphology and ultimately molecule-membrane interactions. Characterizing the thickness together with the ordering of the bilayer as a measure of this anisotropy is therefore important not only for fabrication of model membranes with defined geometrical properties for specific applications but also for establishing molecular mechanisms associated with these changes throughout the binding process. As such, the anisotropic properties of a bilayer can be quantitatively determined by resolving both TM and TE polarization modes in situ by DPI,31,167,595−597 and the application of this approach to membrane interactions will be discussed in more detail in sections below. A range of lipid bilayer structural parameters including thickness/density, mass, lipid surface area, and orientational order have been characterized for individual bilayer systems of different lipid compositions formed on the DPI chip. The values of selected lipid bilayer structural parameters for various lipid compositions derived from TM/TE phase changes are listed in Table 2. In particular, the thickness values correlate closely with the reported bilayers characterized by X-ray and neutron reflectivity.598 These structural parameters were resolved for an anisotropic lipid bilayer and reveal a dependence of the thickness and ordering of the membrane on the length and degree of saturation of the lipid acyl chains. As both the shape and molecular polarizability tensor of the lipid molecules also contribute to the bilayer birefringence, the difference in the isotropic polarizability tensor of the polar head groups also impacts the birefringence values. These effects were evident in the higher molecular order observed for SLBs containing PE. In addition to the dependence of the thickness and organization of a membrane on the lipid composition, the presence of divalent cations such as Ca2+ also influences the packing order of lipid molecules in SLBs.167 The birefringence values were lower for both DOPC and POPC in the presence of Ca2+ compared to higher birefringence determined in the absence of Ca2+. In addition, the isotropically weighted RI and

substrates, which significantly affects the sensitivity of subsequent ligand binding analysis. Lipid nanodiscs have been bound to an NTA-modified substrate via (His)6-Ni2+-NTA conjugation at a high level of nearly 12000 RU, which is required for the sensitivity and detection of small molecules.593 However, the high level of capture was accompanied by lower stability and increased nonspecific binding, which requires more rigorous referencing. Lipid nanodiscs have also been captured by the lipophilic groups on an L1 chip and by immobilized antibodies on CM5 chips. The nanodiscs captured on L1 and CM5 achieved a medium loading of around 3500 RU with the greatest stability and absence of nonspecific binding observed for the capture of nanodiscs on the L1 chip. Although the lipid molecules in the nanodisc are assembled as a tightly packed bilayer, the mobility and expansion of lipid molecules are constrained in the nanodisc which can affect their role in regulating membrane protein function. In addition, systematic control of the bilayer structural parameters such as bilayer thickness and dynamic membrane organization have to be considered for optimal activity of membrane proteins. 4.1.2. Bilayer Order. The determination of thickness and RI for a bilayer using a polarized evanescent field at the surface in most optical biosensors assumes that the SLBs form as isotropic layers on the substrate. Indeed, as described in the section above, the mass of a lipid bilayer can be accurately calculated using the refractive index increment (dn/dc) of lipids without the errors associated with overestimation and underestimation of an isotropic bilayer RI. A low error of ∼0.1% in the calculated mass and discrepancy of ∼2% between the calculated effective birefringence and the layer birefringence have been reported.167 However, the alignment of lipids orthogonal to the substrate and the self-assembly of lipids into highly ordered structures results in anisotropic lipid bilayers. The assumption of an isotropic SLB RI thus leads to significant errors for the layer thickness and the estimated mass of the SLB. In view of the vital role of the dynamic molecular organization and geometrical properties in all membrane functions, the molecular mechanism of membrane-associated cellular processes cannot be fully defined without considering the contribution of these dynamic membrane organization 5434

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Figure 13. DPI characterization of POPC SLBs formation via vesicle adsorption on the unmodified silicon oxynitride at 28 °C at 0.1−0.2 mg/mL lipid concentration and 1−2 mM Ca2+. (A) Temporal changes of the TM polarization mode for depositing POPC SUVs of ∼117 nm diameter. (B) Higher rate of liposome adsorption was observed in the rate of TM phase changes against the TM which is correlated with the lipid mass per unit area on the chip. (C) Changes in POPC mass per unit area during the vesicle-bilayer transformation process with a similar final 3.96 ng/mm2 obtained at different conditions. (D) In contrast to the similar mass/unit area of lipids on the chip surfaces, the lipid molecules are organized differently as demonstrated by different birefringence values where more-ordered POPC bilayer was obtained at 0.15 mg/mL and 2 mM Ca2+ (red line), while the POPC organized into a less-ordered layer at both low and high lipid concentration (black and blue line). Reproduced with permission from ref 549. Copyright 2017 Elsevier.

diameters and could be related to either vesicle deformation and/or to a lower surface coverage, which could lead to an underestimation of the thickness.604 Deformation of vesicles tethered via complement DNA has also been probed by SPR and QCM-D.605 The values for different bilayer structural parameters obtained with solvent-free vesicle-bilayer transformation can be compared to other methods such as LB-deposition and solvent spreading. For example, a low RI of 1.376 with a birefringence value of 0.005 was obtained for a POPC bilayer formed via LB deposition,616,617 while a higher thickness of 5.2 nm, RI of 1.489, and birefringence of 0.0556 was obtained for a POPC bilayer by solvent spreading in a thin Teflon orifice.168 Controlling the physical properties of SLBs in a defined manner remains a challenging task. It is therefore important to characterize the geometrical structural properties for SLBs prepared by highly reproducible assembly methods. This level of quality assurance on the membrane properties demands robust methods involving surface treatment, vesicles preparation, in situ vesicle deposition, and SLB formation and surface regeneration. Characterization of the changes in mass, thickness, and birefringence for the condition-dependence of bilayer formation via liposome adsorption provides a very effective method to selectively tailor the membrane properties. The accurate measurement of optogeometrical and molecular-

thickness values are similar for POPC with and without addition of Ca2+. However, the RI and thickness values of DOPC increased significantly in the presence of Ca2+, showing that the divalent cations have more pronounced effects on the adsorption and bilayer formation for negatively charged lipids. The presence of Ca2+ is also critical for the successful formation of SLBs containing negatively charged PS, PG, CL, and PI.597,599−601 The presence of Ca2+ induces structural changes in hydrocarbon chain packing by inducing tighter packing of the polar head groups and stronger interaction of lipids with the underlying substrates. Depending on the binding angle of divalent cations to lipids and interactions with the substrates, the average alignment of lipids could tilt differently relative to the bilayer normal and therefore give rise to different birefringence values. Intact vesicles can also be either physisorbed or tethered onto the DPI chips.167,602,603 The values of the structural parameters for intact vesicles of various size and composition characterized by the TM//TE phase changes are also listed in Table 3. However, the anisotropic analysis is not applicable for vesicle layers. The LUV and SUV of DOPC/DOPE/DOPS (20:50:30) formed vesicle layers with thicknesses comparable to the corresponding vesicle diameter as characterized by DLS. However, the vesicle layer thickness of DOPC-DOPS (70:30) and DOPC LUVs was much lower than the measured vesicle 5435

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Figure 14. Temperature-induced phase transition of adsorbed (A) DMPC and (B) DMPC/DMPG (4:1) lipid bilayers in the absence (blue) and presence antibacterial peptides (HPA3, red, and HPA3P, green) derived from the N-terminus of Helicobacter pylori ribosomal protein L1. The temperature-induced changes in the molecular ordering are monitored by the changes in optical birefringence. Reproduced with permission from ref 596. Copyright 2010 Elsevier B.V.

excess amount of available lipids. However, an excess amount of liposomes during the adsorption phase can result in a lower ordered bilayer, as determined from the birefringence. This lower molecular order may be due to the excess liposomes affecting the molecular packing process during the edge fusion of ruptured liposomes on the surface. Overall, a low coverage and low adsorption rate of POPC liposomes at 0.1 mg/mL results in a low birefringence, reflecting a more disordered bilayer from insufficient liposome-liposome interaction. Fine tuning the liposome concentration is thus critical to ensure the optimal liposome-liposome and liposome-substrate interaction for a well-packed defect-free bilayer on the surface. The transbilayer asymmetry distributions of lipid components and mobility are also important in regulating intracellular signaling, membrane remodelling, and engineering-functionalized biomembranes. Control of the asymmetric distribution of lipid composition in SLBs is generally achieved by LBdeposition methods. The asymmetric SLBs prepared by direct vesicle adsorption-fusion has also been reported.619−621 Lipid asymmetry can also be affected by the material nature of underlying substrates, liposome preparation methods, and solution conditions for the vesicle adsorption-bilayer transformation, which collectively vary within typical experimental time scales. For example, in the presence of Ca2+, the mobility of PS molecules is restricted in the leaflet in contact with TiO2, while the mobility of PC molecules is not affected.622 In contrast, both PC and PS are mobile in both leaflets of SLBs formed on SiO2. Lipid asymmetry in SLBs has also been introduced by methyl-β-cyclodextrin-mediated lipid exchanges for SLBs containing sphingomyelin and in the presence of reconstituted glycosylphosphatidylinositol-anchored proteins. 623 The temporal changes in bilayer mass and birefringence were then monitored in DPI for the formation process of bilayer asymmetry from lipid flip-flop induced by the temperature-dependent phase transition in SLBs on TiO2.624

ordering parameters by DPI for lipid bilayer in different environmental conditions therefore provides a basis to track the structural changes associated with an adsorption process to be directly linked to kinetic events of bilayer formation and biomolecular interaction. Several experimental parameters are generally considered for optimizing the conditions to obtain a defect-free bilayer of specific lipid composition on a solid support. The main properties that affect the final bilayer formation and structural properties involve (1) the surface properties and topology of substrates, (2) liposome properties, and (3) the bulk solution properties and conditions. A concentration range of 0.1−0.25 mg/mL liposomes is generally used to achieve a complete bilayer.167,597,618 Although this concentration range (which is above the critical rupture concentration) provides a complete bilayer, the structural organization of lipid molecules can differ significantly for complete bilayers with the same lipid composition. The effect of liposome concentration of the same diameter on the molecular organization of SLBs has been characterized by the real-time phase changes in the TM mode (Figure 13A), the rate of TM phase changes (Figure 13B), the total mass per unit area (Figure 13C), and the birefringence (Figure 13D) under the same adsorption conditions. The TM phase changes and dTM/dt during the process of vesiclebilayer transformation at different POPC liposome concentrations show that increasing the liposome concentration results in faster adsorption and bilayer formation. Similar mass per unit area values of 3.96 ng/mm2 were obtained at a POPC liposome concentration of 0.1−0.2 mg/mL (Figure 13C). However, the POPC molecules are more ordered on the substrate as indicated by a higher birefringence value for 0.15 mg/mL liposome than other liposome concentrations tested. Both the duration of complete bilayer formation and the mass per unit area are indistinguishable for different POPC concentrations, which suggests that the bilayer formed does not require an 5436

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

4.1.4. Lateral Structural Characterization of Membranes by SPRi. Lipid molecules are known to localize into specialized micro/nanodomains in membranes exhibiting different packing, lateral mobility, and detergent solubility. These lipid micro/nanodomains function in mediating transmembrane receptor activity, in recruiting proteins and signaling complexes, and in establishing polarity during membrane remodelling necessary for cell locomotion and are of particular importance in various cellular processes. The size and formation of these membrane domains are extremely dynamic. The geometrical structure parameters of lipid bilayers determined by isotropic SPR analysis and anisotropic DPI analysis assumes that different lipid molecules are distributed homogeneously within the bilayer, and so it is not possible to obtain information on the lateral resolution of a bilayer using these techniques. However, imaging SPR has been used to resolve the lateral changes in the dielectric constant and thickness of lipid bilayers formed on sensing substrates.264,629−634 Specifically, the spatial variation in light extinction generated by the binding of proteins to each region of arrayed lipid bilayers can be imaged in real time, providing the kinetics of the molecule-membrane interaction. Micrometer-sized alternating regions of pure fluid eggPC bilayer separated by thiolipid-tethered bilayers have also been analyzed by SPR imaging methods. Alternating SLBs/tethered SLBs were formed on a photolithography-patterned 11mercaptoundecanoic acid layer on Au substrates via a lipid/ detergent micelle dilution technique.264 The angular position of the respective resonance minimum derived from the intensity of CCD captured images were monitored in real time for the spatial formation of a lipid bilayer and the incorporation of rhodopsin into bilayers at discrete regions. Upon the injection of an eggPC/octylglucoside micellular solution, different angle shifts were obtained for the mercaptoundecanoic acid-modified and thiolipid region. A decrease in RI was observed upon initial dilution of detergent with buffer. However, an abrupt increase in resonance angle was observed when the octylglucoside reached its solution CMC and stable membranes were established after extensive rinsing. Mean angle shifts of 0.4° corresponding to a thickness of 4.7 nm were obtained for an eggPC bilayer formed on the mercaptoundecanoic acidmodified region, while angle shifts of 0.22° corresponding to a thickness of 2.6 nm were obtained for the eggPC monolayer on the tethered-thiolipid region. In contrast, an angle shift of 0.46° was obtained for an eggPC bilayer formed on a nonpatterned, pure mercaptoundecanoic acid-modified Au surface. The structural characteristics of SLBs self-assembled via vesicle-bilayer transformation on a patterned mixed array of hexadecanethiol and mercaptoundecyl tetra(ethylene glycol) SAMs on Au have also been spatially resolved by SPR imaging.633 The hexadecanethiol patterns were created on a Au surface by microcontact printing using a PDMS stamp and the bare Au regions were backfilled with ethylene glycol-terminated thiol monomer. The formation of lipid monolayer (SLMs) and SLBs in patterned SAMs systems were analyzed by temporal reflectivity changes obtained discretely at the center of each square of glycol-terminated region and at the intersection of the two lines that makes up the hexadecanethiol SAM. Several distinctive trends were observed from the variation in image intensities across different SAMs-patterned regions. These include a faster rate of reflectivity increases at the glycolterminated SAMs regions than the alkane-terminated SAMs

4.1.3. Bilayer Phase Transition. The bilayer structure can also be probed more directly through the transition between the gel state and the lamellar liquid crystalline state. The dependence of phase transition on lipid composition is sensitive to temperature and is commonly characterized for liposomes by DSC. The phase transition involves changes in not only the fluidity of lipid molecules but also the packing and alignment of lipids within the bilayer structure. Monitoring the changes in birefringence as a function of temperature also provides a labelfree method to characterize the phase behavior of a pure lipid bilayer and peptide-bound lipid bilayers. Moreover, the Tm values determined by the temperature-induced changes in birefringence are directly analogous to those determined by the excess heat capacity using DSC. By using an uncoated DPI waveguide channel as a reference for correcting the thermal effect on the waveguide structure and the bulk RI, sharp increases in birefringence values defined a phase transition temperature (Tm) of 24 °C for a DMPC bilayer of 4.4 nm at 30 °C by cooling the bilayer down to 20 °C.167,596 In contrast, a lower gel−liquid crystalline transition temperature of 21.8 °C for DMPC bilayers was characterized by measuring the birefringence changes by heating the bilayer from 20 to 37 °C, and this difference in the Tm can be attributed to thermal hysteresis. In addition, a distinctive phase transition was observed at 23.7 °C for the binary DMPC/DMPG (4:1) SLBs.596 The presence of the anionic lipid is consistent with the phase behavior examined by DSC625,626 and reflects a higher ordering for DMPC/DMPG, which is also evident from a higher birefringence value. As shown in Figure 14, the changes in birefringence as a function of temperature also provide a sensitive tool to examine the impact of the HPA3 peptides on the phase transition behavior of the adsorbed DMPC and DMPC/DMPG. Thermal birefringence curves therefore demonstrated the effects of lipid molecular organization on the binding characteristics of the peptide while the lipids undergo a change in structure. In contrast to the lipid bilayers adsorbed onto microsized substrates or present as liposomes at high concentration in solution, the phase behavior of bilayers adsorbed onto the surface of nanoparticles cannot be easily determined by calorimetry methods due to the low quantity of surfaceadsorbed lipids. As LSPR peak shifts are sensitive to the changes in the local dielectric environment, the structural change between gel phase and fluid phase would alter that environment and cause a peak wavelength shift. For lipid bilayers with accessible transition temperatures, Au nanorods were coated with a DMPC bilayer stabilized with DOPG and the LSPR peak wavelength was tracked with the temperature ramped slowly at 0.1 °C/min. The LSPR spectral extinction peaks are broad with full width at half-maxima greater than 50 nm. The extinction spectra allow evaluation of the peak resonance wavelength with a precision of 0.001 nm for a typical extinction value and minute integration time.627 Despite the varied baseline drift rate, a sudden jump was detected at 23 °C in the measured drift of LSPR peak wavelength with temperature628 in agreement with the transition temperature of DMPC vesicles determined by DSC. The stability of the DMPC/DOPG SLBs on the Au nanorods were also verified by the phase transition of LSPR wavelength shift with temperature. The absence of bilayer phase transition upon dilution of the lipid concentration in solution was referred to as the structural collapse from a well-ordered bilayer state to a nonbilayer state. 5437

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

an alternative method for preparation of membrane protein biochips.

regions. Furthermore, a higher signal detected in the glycolSAMs was related to a greater amount of adsorbed lipid molecules than those adsorbed onto the alkane-SAMs and was related to the mass difference between the bilayer versus monolayer. However, the changes in the reflectance is not directly proportional to the amount of adsorbed lipids due to the differences in the inherent RI and the local water ordering between the SLM regions and the SLB regions.635 The enhanced signal was initially observed around the border edge of the glycol- and alkane-terminated SAMs. As SLBs do not form spontaneously on glycol-terminated SAMs, this edge effect of enhancing the signal further demonstrated the critical role of the interface in nucleating bilayer formation over the glycol-terminated region. The kinetics of SLB and SLM formation in the patterned SAMs array were further characterized as a function of pattern dimension and vesicle size. The SLM formation is fast for liposome sizes of 50 and 100 nm and slower for liposome sizes of 30 and 400 nm. However, there was no obvious trends in pattern dimension and no minimum surface packing density required for vesicle rupture and SLM formation on the alkane-SAMs. This is in contrast to approximately 70% critical coverage required for vesicle-bilayer transformation on the unmodified SiO2. This lack of surface area dependence of SPR signal coupled with a single exponential kinetics is in agreement with the Lingler model for SLM formation.636 In contrast, the increase in SPR signals fits to a second-order kinetics for the formation of SLBs on the glycol-SAMs. This indicated that two distinct processes with different rates are involved in SLB formation over the glycol regions via vesicle attachment at the edge of the two patterned SAMs. The structural properties of SLBs tethered to an α-laminin peptide (P19) SAMs on Au substrates have been characterized by SPRi.629 The tSLBs were fabricated in a three-step process, and the Au surface was first modified with a layer of P19 peptides via either direct self-assembly or microcontact printing. PE lipids were then coupled to the P19 peptide layer using EDC/NHS chemistry. Among the various unsaturated and branched PE anchors, only DMPE exhibited optimal packing density and surface coverage while formation of an inhomogeneous layer was observed for unsaturated and branched PE anchors. Lastly, the fusion of liposomes composed of either eggPC or E. coli lipid extract formed the bilayers. The SPR angular scan curve showed that the surface coverage of the cushion layer was significantly improved by the microcontact printing technique, thereby facilitating bilayer formation. The formation of PC or E. coli lipid bilayers on top of P19/DMPE tethered monolayers caused shifts in the resonance angles in the SPR angular curves. A larger angular shift of 1.5° was observed for the E. coli bilayer and a shift of 1.0° was obtained for the PC layer. The incremental thickness increase determined for the PC and E. coli lipid layers were 4.18 and 5.84 nm, respectively. The thickness of the E. coli lipid bilayer was consistent with that formed on the unmodified silicon oxynitride as analyzed by DPI. The functional properties of these two different membrane systems were validated by incorporation of cytochrome bo3 ubiquinol oxidase (Cyt-bo3). The insertion of Cyt-bo3 into different membranes was also monitored by surface plasmon fluorescence spectroscopy (SPFS), which enhances the detection limit of SPR. The formation of a homogeneous lipid bilayer as characterized by SPRi demonstrated that membrane formation can be better controlled with microcontact printing techniques and provides

4.2. Characterization of Molecule-Biomembrane Interactions by SPR (Mass Only)

The interactions between peptides and membranes mediate a wide variety of biological events and characterization of the molecular details of these interactions is central to our understanding of cellular processes such as protein trafficking, cellular signaling and ion-channel formation. A wide range of biophysical techniques have been combined with the use of model membrane systems to study peptide-membrane interactions and have provided important information on the relationship between membrane-active peptide structure and their biological function. Prior to the development of commercial SPR and DPI-based biosensors, however, a detailed analysis of the affinity of peptides for different membrane systems was generally not reported, which was largely due to the difficulty in obtaining this information. For example, membrane binding assays generally require a physical separation step, such as dialysis or centrifugation which can be tedious for large numbers of samples.637−639 Alternatively, peptide partitioning can be directly measured by changes in fluorescence quenching or enhancement of tryptophan fluorescence, which relies on the presence of a fluoroprobe. If an intrinsic fluoroprobe is not present, the probe must be incorporated by covalent modification of the peptide and/or the lipid.640,641 To address this issue, SPR and, more recently, DPI have been applied to the study of biomembrane-based systems using both planar mono- or bilayers or liposomes. This section provides an overview of the application of SPR and DPI to the analysis of membrane interactions and how the information obtained using these techniques enhance our molecular understanding of membrane-mediated peptide and protein function. We first discuss experiments where SPR alone has been used to characterize membrane binding and describe how these studies yielded novel insight into the molecular events associated with membrane interactions and how they provided a significant impetus to more recent studies that focus on coincident membrane structure changes during binding of peptides and proteins. We then discuss the emerging limitations of SPR in terms of monitoring the effects on membrane structure and how SPR data can be combined with DPI to provide significant new information on how a membrane responds to binding of peptides and proteins. 4.2.1. Peptide-Membrane Interactions. The common structural features of membrane-active peptides involve the adoption of a stable secondary structure upon binding to the membrane surface. The induction of a cationic amphipathic αhelical structure has been well-characterized and plays an essential role in the membrane perturbing activities of antimicrobial peptides.642−644 In contrast, similar structure− activity relationships are less well-studied for membrane-active peptides and proteins that adopt a β-sheet structure upon membrane association, although β-sheet antimicrobial peptides such as the defensins and protegrins have been wellcharacterized.644 It is generally considered that the binding of membrane-active peptides to lipid membranes occurs via complex cooperative mechanisms involving at least a two-step process. Some examples of structural and orientational changes that occur during the interaction between peptides and membranes are shown schematically in Figure 15. The peptide 5438

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

membrane structure has only been derived from solid-state NMR and fluorescence studies. However, the advent of SPRbased biosensors allowed a number of new specific questions to be addressed for the first time that include the following. (1) Does the peptide bind reversibly or irreversibly to a particular membrane composition? (2) How does the headgroup structure influence the binding process? (3) What role do electrostatic and hydrophobic interactions play in binding to a membrane? (4) How selective is a peptide for a specific lipid composition? (5) How fast does the peptide bind to a membrane? (6) Are there kinetically distinct stages in membrane binding? (7) Can membrane disruption be observed by SPR? Within this framework of specific questions about the mechanism of interaction, the examples of membrane-based SPR published over the last 20 years fall into one of two distinct categories as follows. (A) What is the relative ability of a peptide/protein to bind to a specific lipid mixture in a bilayer context? (a) Anionic versus neutral → relative role of electrostatic and hydrophobic interactions. (b) What is the lipid headgroup specificity (e.g., PE vs PC, SM, LPS, LPS, PS/ Ca2+, and D-Ala-D-Ala)? Or (B) Evidence for multiple binding steps by kinetic analysis of sensorgrams. (a) Langmuir versus two-step mechanism (1, electrostatic and 2, hydrophobic). (b) HPA versus L1 chip → insertion/pore versus carpet mechanism. Within the context of these mechanistic questions, in the following sections, we overview the approaches to the kinetic analysis of biomolecule-membrane interactions followed by discussion of a range of applications which address different molecular aspects of membrane interactions using SPR. 4.2.2. Curve-Fitting Strategies and Kinetic Models to Determine Affinity Constants for Membrane Interactions. The demonstration that membrane-active peptides exhibited reversible membrane binding represented a significant turning point in the study of the mechanism of action of this class of peptides. This observation allowed a quantitative structure-membrane-binding-activity approach to characterizing the role of membrane interactions in the function of a wide range of peptides. The earlier studies using SPR to measure peptide-membrane interactions thus sought to quantitate the interactions via kinetic analysis of the binding data in a similar manner to methods used for other biomolecular interaction analysis.647−650 While equilibrium measurements were widely used for protein−protein interactions, this approach was largely not feasible for membrane interactions. Kinetic models, with regard to peptide-membrane interactions, are an approximate mathematical representation of the process by which a peptide binds to the membrane. A simple example of such a model is the one-state model (sometimes called “Langmuir binding”) in which the peptide binds and dissociates with the membrane at concentration-dependent rates.32 The peptide does not change state at the membrane, and there is only one way in which it may bind. This may be represented by the reaction:

Figure 15. Coherent changes in the structure, orientation, and assembly of peptide and lipid molecules during the binding process.

first binds electrostatically to the membrane localizing itself near the surface and then reorients or relocates on the surface or inserts further into the hydrocarbon region of the lipid membrane by hydrophobic interactions. To understand peptide-membrane interactions in detail, the role of numerous physicochemical parameters, including peptide charge, hydrophobicity, amphipathicity and the degree of secondary structure angle subtended by the polar face, in the interaction between peptides and membranes have been analyzed. A wide variety of biophysical techniques such as circular dichroism, nuclear magnetic resonance, fluorescence spectroscopy, Fourier transform infrared spectroscopy, attenuated total reflection Fourier transform infrared spectroscopy, and immobilized artificial membrane chromatography combined with model membrane systems have all been used to study biomolecular-membrane interactions.223,331,645,646 These techniques have provided, and continue to provide, important information on specific structure−function relationships regarding peptide-membrane interactions. Since the binding reaction can be very fast, it is often difficult to distinguish the two (or more) binding steps and posed a great challenge to scientists interested in determining this data using conventional techniques. Over the last 15−20 years, SPR has become a widely used technique to study affinity and kinetic events of biomembrane-based interactions and a number of more recent developments in SPR technology platforms has expanded the potential of SPR to provide new information on peptide- and protein-membrane interactions. If one considers the range of techniques listed above that have been used to study peptide/protein interactions, it is apparent that they mostly focus on the structure and behavior of the peptide. Information on the specific nature of the

P + L ↔ PL where P represents the peptide free in solution, L represents the unbound lipid membrane, and PL represents the complex formed by the binding of the peptide to the membrane. Kinetic models are modeled using a system of differential equations describing the rate of change of various states in substances; in this case the reaction is defined by the equation: 5439

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

lipid

5440

PE/PC/PI/PS/Erg (4:2:2:1:3)

DMPC

DMPG DMPC/DMPG (2:1) E. coli extract

bombinin H4 aurein

citropin maculatin caerin

bilayer-L1

bilayer-L1

HBM-HPA

DMPE/chol PC

various % of PA/PC

bilayer-L1

bilayer-L1

HBM-HPA

POPG POPC/POPG (3:1) DMPC

DMPG DMPC/DMPG DMPE

bilayer-L1

DMPE/chol POPC

cecropin B1 (CB1) cecropin B3 (CB3) bombinin H2

cecropin B (CB)

mellitin-21Q

bilayer-L1

bilayer-L1

HBM-HPA

DMPC/DMPG DMPC

mellitin

DMPG DMPC/DMPG eggPC eggPC/chol DMPE

bilayer-L1

Ala8, 13, 18-magainin 2

bilayer-L1

DMPG

bilayer-L1

HBM-HPA

chip

main conclusion

curve fita

b

a,b

c

b

a,b

d

b

a,b

b

maculatin removed PC/PG all peptides bond strongly to anionic membrane membrane selectivity using natural lipid extracts correlated better with the MIC values against G(+) bacteria and hemolytic activity than binding on DPMC and DMPG

aurin removed both PC and PC/PG from chip surface

Proposed CB and CB1 pore in membrane from dye-leakage while CB3 acted by nonpore formation two-state curve fitting with higher membrane affinity for H4 than H2 from a higher affinity at the first step of initial binding

CB and CB1 bind to both neutral and anionic membranes with enhanced response at higher PA content, while no membrane binding for CB3. No correlation between helical content and binding response

a slower step 1 binding showed the critical role of electrostatic interaction in membrane binding and subsequent insertion

the deletion of C-terminal basic residues resulted in significant loss of membrane binding to both HBM and bilayer

different peptide-to-lipid ratios obtained for insertion and membrane lysis

higher response for binding to PE-bilayer, and the binding response to PE-bilayer can be lowered by increasing the bulk NaCl concentration

stronger binding to anionic membrane

melittin binds to all membranes

higher response for anionic membrane at low peptide concentration, inducing lysis to both neutral and anionic membrane

no significant difference in binding responses between HBM and bilayer for surface binding in membrane micellization

antimicrobial peptides (helical) a,b higher SPR responses for peptides binding to anionic membranes

membrane -

DMPC

mixture

DMPC/DMPG eggPC eggPC/chol DMPE DMPE/chol DMPC

magainin 1 and 2

peptides

Table 4. Summary of Studies of Peptide/Protein Binding to Membranes Characterized by SPR

662 and 663

661

659 and 660

655

651

657 and 658

655

32, 651, and 652

606

606

651 and 652

ref

Chemical Reviews Review

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

lipid

5441

L/D lipopepties DDA-LKKLLKKLLKKL DDA-LHHLLHHLLHHL DDA-LRRLLRRLLRRL

β-peptide scrambled β-peptide

D/L diasteromers

12 × kalata analogues

kalata B1 D-kalata B1 Gly6Ala-kalata B1

PC/chol PC/PE/PI/Erg (5:4:1:2)

DMPC/DMPG DMPE DMPE/chol

Egg PG Egg PC/chol E. coli PE/PG PC/SM/PE/chol PC/SM/PE/PS/chol LPS

PE/PG eggPC/chol POPC/chol (4:1) as hRBC POPE/POPG/CL (15:4:1) as E. coli OM DMPC DMPC/DMPG DMPC/DMPG/chol DMPE DMPE/DMPG POPC POPC/POPG (x2) POPC/POPS POPC/chol/SM DPPC POPC/chol POPC/SM POPC/POPE POPC/POPE/chol POPC/POPE/chol/SM Egg PC

ranacyclins

tachyplesin-1 and 5 analogues

DMPC POPE:POPG (3:1)

bovine lipid extract DOPE/DOPG (4:1) DOPE/DOPG/CL (65:23:12)

mixture

thaulin

myxinidin myxinidin mutant (WMR)

peptides

Table 4. continued

b

higher membrane affinity of WMR than myxinidin WMR binds stronger to CL-membrane Ka(bilayer)/Ka(HBM) is higher for WMR as pore formation than myxinidin as surface binding new amphibian peptide, binds stronger to anionic membrane

fita a,b

main conclusion

curve

bilayer-L1

bilayer-L1

bilayer-L1

HBM-HPA

bilayer-L1

bilayer-L1

bilayer-L1

selectivity determined by first binding reduced amphipathicity lowers the membrane insertion of L-peptides only L- and D-peptides bind with similar Rmax to LPS, with L-peptides showed cooperative binding both β-peptide had similar binding on all membranes and higher than Magainin

D-peptides bind similarly to HBM and bilayer for surface binding, while 100 stronger to PE/PG

L-peptide bind 10−25× stronger to PC/PG bilayer than HBM for membrane insertion

stong affinity for PE

higher affinity for PE Gly5 important for binding via possible dimerization

higher binding response for E. coli membrane than hRBC membrane degree of binding did not correlate with hemolytic activity

antifungal peptides b effect of Arg/Lys on binding at pH 5.5 versus 7.4 and antibacterial and/or antifungal activity

b

b

b



antimicrobial peptides (β-hairpin, cyclic and β-peptide) b similar binding responses to both neutral and anionic membranes bilayer-L1

bilayer-L1

HBM-HPA bilayer-L1

chip

membrane -

677

676

674 and 675

671−673

668−670

667

666

665

664

ref

Chemical Reviews Review

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

5442

PC/chol

N-terminal heptad repeat of HIV-1 HIV-GCN4-dimer

T cell TMD peptide

Pep-1 (Ac-KETWWETWWTEWSQPKKKRKV-cysteamine)

HIV fusion inhibitors enfuvirtide sifuvirtide

HIV-GCN4-trimer HIV-GCN4-tetramer N70

bilayer-L1

HBM-HPA bilayer-L1

POPC/POPG (4:1) DMPC DMPG

HBM-HPA bilayer-L1

bilayer-L1

bilayer-L1

HBM-HPA bilayer-L1

bilayer-L1

bilayer-L1

chip

main conclusion

curve fita

mode of interaction depends on whether a peptide exists as a stable, preformed helix or adopts secondary structure only upon membrane binding

binding and structure depnds on anionic lipids binding only on L1 bilayer amyloids

b

cell penetrating peptides b initial fast electrostic binding important for translocation activity

highest affinity for SM

HIV all chimeras bound ireversibly coiled-coil oligomerization increases lipid mixing ability by facilitating stronger binding of the fusion peptides to the target membrane

apolipoproteins-derived peptides b stronger to PG, anionic important

antitumor D,L-amphipathic helix b bind more to anionic lipids but similar binding to HBM and bilayer. Cell selectivity due to increase in anionic lipids in cancer cells

b

sheet peptides bound more to PE/PG W removed specificity and position of W influenced binding

model helical and sheet peptides b helical peptides bound more to PC/PG

membrane -

POPC

POPC POPC:DPPC(1:1) POPC:DPPC(1:2) DPPC EPC SM

DMPC DMPG

class A amphipathic model peptide (18D)

KK-I3,6,13 k8,9K6L

PC/SM/PE/chol (4.5/4.5/1/1) PC/SM/PE/PS/chol (4.35/4.35/1/0.3/1) PC/SM/PE/PS/chol (3.5/3.5/1/2/1)

E. coli PE/Egg PG (7:3)

Egg PC/chol (10:1)

POPC POPG POPC/POPG (4:1) POPE/POPG (4:1)

PE/PG (7:3)

lipid mixture

K−I3,10,13k7,8K6L9 KK-I3,10,13 k7,8K6L

I3,10,13k7,8 K6L9

(KIGAKI)3-NH2 W2-(KIGAKI)3-NH2 W8-(KIGAKI)3-NH2, W18-(KIGAKI)3-NH2 Oct-W8-(KIGAKI)3-NH2 11 × model peptides based on helical AMP template with varied hydrophobic face

(KIAGKIA)3-NH2

DDA-LKHLLKHLLKHL DDA-LRHLLRHLLRHL

peptides

Table 4. continued

686

685

684

683

682

681

680

678 and 679

ref

Chemical Reviews Review

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

5443

lipid

DMPC/lipid A (95:5) DMPC LPS POPC−ergol (4:1) POPC−chol (4:1) POPC POPG POPG/cardiolipin POPE/POPG PC

polymyxin B

vancomycin ristocetin A eremomycin

TA3-1

glycopeptide antibiotics (GPA)

daptomycin

PC/N-docosanoyl-lysyl-(N-ε-acetyl)D-alanyl-D-lactate

POPC/POPG (4:1) POPC/chol (2:1) POPC/GM1 (9:1)

PrPC 106−126- M112W

amphotericin B (AmB) (a polyene)

POPC

DPPG DPPG/chol (4:1) DPPC/DPPG (3:1) DPPC/GM1/chol (5:3:2) DMPC/DMPE/DMPS/DMPG (75:20:2.5:2.5) + chol (30, 40, 60, 80%) DPPC + 12 different gangliosides DMPC + 4 different gangliosides DPPC + 12 different gangliosides

HBM-HPA

bilayer-L1

CM5 bilayer-L1

HBM-HPA

bilayer-L1

role of GPA dimerization

correlation between cytloytic activity and binding

higher binding to G(+) mimics

toxicity not correlated to membrane binding antibiotics specificity for lipid A removes lipid from membrane LPS binds to PMB higher binding to ergosterol

strong electrostatic component to binding

prion peptides b PrPC 106−126 binds less than M112W analogue

ref

573 and 699

695 696 and 697 698

694

692 and 693

689 691 689

689 690

688

HBM-HPA HBM-HPA HBM-HPA

Aβ has high affinty for gangliosides Aβ1−42 only binds in the presence of gangliosides Aβ has high affinty for gangliosides

Aβ has high affinty for gangliosides Aβ1−40 has high affinity for PG and gangliosides

good correlation between the toxic effect of Aβ peptides and their membrane binding cholesterol-enhanced binding and toxicity correlated with lipid binding monomeric Aβ adsorbed quickly but reversibly to lipid bilayers with low affinity, while aggregated Aβ adsorbed slowly but irreversibly

687

687

a

fita bound to synthetic lipid mixtures and to an intact plasma membrane from vascular smooth muscle cells

main conclusion

curve

bilayer-L1

HBM-HPA immobilized liposomesSAM

bilayer-L1

POPC

POPG DPPC + 12 different gangliosides DPPC

bilayer-L1

chip

membrane -

DMPC/DMPE/DMPS/DMPG (75:20:2.5:2.5) + chol (30, 40, 60, 80%)

mixture

PrPC 106−126

Aβ1−38 Aβ25−35 Aβ40−1

Aβ1−42

Aβ1−40

peptides

Table 4. continued

Chemical Reviews Review

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

lipid

5444

PC/PS (75:25) PS/PC/PE (4/76/20) PC/chol/0.1% biotin-X-DHPE

PC/chol/3% GlcCer/0.1% biotin-X-DHPE PC/chol PC/chol/3% GlcCer PC/chol/GM1 (7:2:1) PC/chol/BMP (6:2:2) PC/chol/BMP/GM1 (5:2:2:1)

POPC

factor VIII

factor VIII-light chain (LCh)

sphingolipid activator proteins (saposins) (SAP-A, B, C and D)

HIV antibodies

POPC:chol (2:1) POPC/chol/SM (1:1:1) viral model

POPC DPPE/DPPC/DPPS/chol (40/40/15/5)

2F5, 4E10 2F5−chol conjugated D5−chol conjugated

lipidated Ras constructs recoverin

lysosomal proteins

PC

lactoferrin

fita

b

peripheral membrane proteins HBM-HPA lipidated RAS can insert into hybrid bilayer biotin-liposomes- SAassociation and dissociation of membranes was fast and biphasic (fast and slow components) Ca2+myristoyl switch dependency for membrane association Chip

membrane affinity correlated with antiviral anti-HIV1 activity

GM1 is critical for the enhanced membrane binding od SAP-B prebound epitope required for high affiniy antibody binding

bis(monoacylglycero)phosphate enhance β-galactosidase and SAPs to substrate-carrying membranes

bilayer-L1

bilayer-L1

spike of 20% BMP at high SAP-C concentration result in membrane lysis

weak binding of SAP-C to both lipid vesicle and HBM with and without 3% GlcCer

PS important for F VIII membrane binding

bound to LPS membrane binding proteins a LCh is responsible for membrane binding

cytolytic proteins a specificty for SM

only ProTx-I bound to either membrane. Membrane binding is not essential to act as a gating modifier

protein toxins b two-step binding binding is dependent on pH and Ca2+ a analogues that bind membranes have higher ion channel acitivty

main conclusion

curve

biotin- liposomesSA chip HBM-HPA

bilayer-L1

HBM-HPA

HBM-HPA

HBM-HPA

POPC/POPS (4:1) POPC/SM/CHOL (2.7:4:3.3) POPC/C1P/CHOL (2.7:4:3.3) (C1P = ceramide-1-phosphate) POPC

POPC/POPS (4:1)

bilayer-L1

DOPC POPC

chip

membrane -

HBM-thiolated J1 bilayer-L1 bilayer-L1

mixture

SM/DPPC (1:1) SM/chol (1:1) LPS

lysenin

voltage-gated ion channels (VGICs)

2 GMTs (μ-TRTX-Hd1a and ProTx-I) and 2 pore blockers (ShK and KIIIA)

equinatoxin II perforin spider peptide toxin HwTx-IV (Gatin modifier toxin-GMT)

CE BCE teicoplanin

peptides

Table 4. continued

714 715 and 716

712 and 713

708−711

582, 706, and 707

705

704

703

700 701 702

ref

Chemical Reviews Review

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

5445

lipid

DOPC

LPC

biotinylated PIPs nanodisk POPC/POPE/PI(4,5)P2 POPC/POPE/PI(3,4,5)P3 POPC/POPE/PI(4,5)P2 POPC

POPC/POPS/DOG (99-x:x:1) POPC/POPG/DOG (99-x:x:1) POPC/POPS/DOG (69:30:1) + 1 mM Ca2+ POPC/POPG/DOG (69:30:1) + 1 mM Ca2+ POPC/POPS/DOG (69:30:1) + 7 μM Ca2+ POPC/POPS (7:3) POPC/POPG (7:3) POPS/POPC/POPE (2:5:2)

POPC/POPS (7:3)

PE/PC/PI(4,5)P2 (85:5:10) PE/PC/PI(3,4,5)P3 (85:5:10) DOPC DOPC-PI(3,4,5)P3 (97:3) brain lipid extract POPC

DPPE/DPPC/DPPS/chol (40/40/15/5)

mixture

719 720 and 721

Ca2+-myristoyl switch can function with different lipid moieties PLC-δ1-C2 and PLC-δ3-C2 involved Ca2+-induced electrostatic interactions and PS coordination

bilayer-L1

HBM-alkane-thiolated SAM

nanodisk-CM5

729

728

Ca2+ important for membrane binding

pesticides established SPR method for correlating toxicity and in vivo persistance

725 and 726 727

724

bilayer-L1 correlation between the membrane aggregation activity and the relative affinity for the secondary membrane PI(3,4,5)P3 important for membrane binding of Grp1-PH

723

bilayer-L1

bilayer-L1

580

bilayer-L1

in contrast, PLC-δ4-C2 exhibited a Ca2+-independent membrane binding and no selectivity for PS

718

717

ref

high affinity for PI(3,4,5)P3

PH domain regulates membrane binding via PIPs, specifically PI(4,5)P2

722

a

fita binding to Ca2+ causes a large conformational change to extrude the myristoyl group, and thereby interacted with lipid membranes

main conclusion

curve

bilayer-L1

HBM-HPA bilayer-L1

bilayer-L1

liposomes bind to recoverin-immobilized CM5 HBM-HPA

chip

membrane -

a

Curve fitting with Langmuir 1:1 adsorption model. bCurve fitting with two-state binding model. cCurve fitting with nonlinear binding model. dTransfer matrix model with concentration thresholds for peptide insertion and membrane lysis.

17 different pesticides

C-reactive protein (CRP)

kindlin-3 PH domain

annexin-1

C2 domains of PLC-δ1, PLC-δ3 and PLC-δ4

neurocalcin + different acyl groups protein kinase Cα (PKC-α)

phox homology (PX) domains

phospholipase D (PLD) activity

peptides

Table 4. continued

Chemical Reviews Review

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Figure 16. SPR sensorgrams of magainin 2 binding to (A) DMPC and (B) DMPC/DMPG (4:1) and melitin binding to (C) DMPC and (D) DMPC/DMPG (4:1) bilayers formed on an L1 chip. (A) and (B) Reproduced with permission from ref 606. Copyright 2015 Springer Nature. (C) and (D) Reproduced with permission from ref 655. Copyright 2011 John Wiley and Sons.

P* + T ↔ P*T

dM /dt = ka1CPM max − (kaCP + kd)M

The corresponding differential rate equations for this reaction model are

However, it has been found that this model provides an inadequate representation of binding processes,651−653 and more sophisticated models must be used. A number of different reaction models have been developed for application to kinetic analyses of any biomolecular interaction. However, two different curve fitting algorithms are most relevant on the basis of what is known about the possible binding mechanisms of membrane-mediated binding events. In order to distinguish between the possible binding models, the data can be fitted globally by simultaneously fitting the peptide sensorgrams obtained at ten different concentrations. The two-state model651−653 is often used, so-called because the peptide forms one state on initial binding to the membrane and then undergoes a transition to a second state once it is bound. This may be represented by the reaction:

dR1/dt = ka1 × CP × (R max − R1) − R1 × Kd1

dR 2/dt = ka2 × CP × (R max − R 2) − R 2 × Kd 2

Many of the peptide/protein−membrane interaction examples summarized below have utilized the Langmuir 1:1 or the two-state model to determine membrane affinity constants. The kinetic evaluation of sensorgram data using these numerical integration techniques has therefore provided important insight into the definition of intermediate steps along the pathway of binding and insertion into membranes. There have been a wide range of studies using SPR to understand the mechanism of action of a range of membraneactive peptides. The published applications are listed in Table 4 and illustrate the broad range of peptide and proteins that have been studied and for which new insight into the role of membrane binding was obtained for the first time. The sections below focus on the broad class of membrane-lytic peptides and peripheral membrane proteins. More specific subclasses of molecules which have been characterized by both SPR and DPI are discussed in subsequent sections. 4.2.3. Membrane-Lytic Peptides. A wide range of antimicrobial peptides act via binding to and subsequent disruption of cell membranes. A number of different mechanisms have been proposed to describe the mode of action of these peptides. These models vary from general bilayer disruption following the binding of a critical concentration of peptide, the formation of pores, or through the binding with specific lipophilic components of the membrane such as lipopolysaccharides. Since these models all

P + L ↔ PL1 ↔ PL 2

and mathematically by the system of differential equations dM1/dt = ka1CP(M max − M1 − M 2) − kd1M1 − ka2M1 + kd 2M 2

dM 2 /dt = ka2M1 − kd 2M 2

The parallel reactions model has also been applied and assumes that two simple bimolecular interactions occur in parallel with different rate constants, giving a complex bimolecular interaction as follows. (1) Peptide (P) binds to lipids (L).

P + L ↔ P*L (2) The bound peptide (P*) then provides a surface for the binding of additional peptide molecules [e.g., possible pore formation (T)]. 5446

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

which is known to largely act by a detergent-like mechanism. It also exhibited 100-fold higher affinity for negatively charged vesicles over zwitterionic membranes. In contrast, a ratio of affinity constants = 25 was observed for melittin, indicating that it inserts into the hydrophobic core of the bilayer.652 This presumably reflects the ability of the peptides to penetrate the liposomes, whereas membrane insertion is restricted on the HBM-HPA surface. Indeed it was shown that at sufficiently high concentrations, the cytolytic peptide melittin caused the lysis of the immobilized DMPG liposomes as evidenced by a drop in RU during the association phase. This approach of comparative mono- and bilayer interactions was also applied to investigate the mechanism of action and basis for cell selectivity of a series of naturally derived and artificial AMPs, antifungal, antiparasitic, antiviral, and tumor-lytic peptides (refs 652, 661, 666, 674, 675, 677, 680, 681, 683, and 730). The combined results allowed comprehensive conclusions about the relative contribution of hydrophobic and electrostatic interactions to the different stages of membrane binding together with the role of amphipathicity and structure on the mechanism of binding and selective lysis of normal and tumor cells. The effect of charge on peptide-membrane interactions has also been characterized by the SPR analysis of the binding of magainin, melittin, and its C-terminally truncated analogue (21Q) to neutral and negatively charged membranes supported on either the HPA or L1 chips.49,651,654,655,678 The absence of positively charged C-terminal residues in 21Q resulted in a loss of binding specificity with low binding affinities of this peptide for both DMPC and DMPG. Comparison of the results of melittin and 21Q thus indicates that the positive tail of melittin allows it to bind more rapidly and more strongly to the anionic lipids by electrostatic interactions, thereby enhancing the subsequent hydrophobic binding. These results also clearly demonstrate the role of electrostatic interactions in the initial orientation and binding of these peptides to membrane. In particular, these results suggested that electrostatic interactions may correspond to the initial rapid phase since 21Q exhibited a slower association phase than melittin. The electrostatic binding then enhances the subsequent insertion/peptide aggregation step which may correspond to the slower second phase of binding. While it is possible that the interactions between antimicrobial peptides and membranes are more complex than can be described by the two-state reaction and the parallel reactions models, they provide a more appropriate kinetic analysis of the interactions than the 1:1 Langmuir binding model. The binding of melittin and the 21Q analogue to an expanded range of phospholipid bilayers, including DMPC, DMPE, DMPC/DMPG, DMPC/DMPG/cholesterol, and DMPE/DMPG were also investigated using SPR.655 Melittin bound rapidly to all membrane mixtures, whereas 21Q bound much more slowly on the DMPC and DMPC/ DMPG mixtures again reflecting the role of the initial electrostatic interaction. The loss of the cationic residues also significantly decreased the binding of 21Q with DMPC/ DMPG/cholesterol, DMPE, and DMPE/DMPG. The role of electrostatics was further highlighted with NaCl in the buffer which affected the way melittin bound to the different membranes, causing a more uniform, concentration-dependent increase in response. Overall, the results demonstrated that the positively charged residues at the C-terminus of melittin play an essential role in membrane binding, that modulation of peptide charge influences selectivity of binding to different phospholipids, and that manipulation of the cationic regions of

involve the binding of peptides to the membrane, the determination of the relative affinity of the peptides for a particular target membrane is central to the delineation of the mechanism of action of these peptides. SPR therefore has a very important role in furthering our understanding of the molecular basis of action of this class of peptides.32,49,654−656 Among a significant number of studies that have used SPR to study the interactions of a range of antimicrobial peptides (AMP) with membranes of different composition (refs 651, 652, 655, 668, 671−674, 676, 681, and 730), the structurally well-characterized melittin and magainin are two of the most widely (refs 606, 651, 652, 655, 656, and 731−733) studied AMPs in terms of its structure−function relationships with versatile spectroscopic analysis. SPR has also contributed important further insight into the reversibility of membrane binding of these peptides, the role of electrostatic and hydrophobic interactions in membrane binding, and evidence of a dual-step mechanism of binding and insertion. Once the ability of SPR methods to distinguish between weak and strong membrane binding was established, a number of studies further exploited this technique to explore detailed peptide structure− function relationships. This was facilitated by the ease of preparing and depositing a great variety of phospholipids mixtures deposited on the first manufactured HPA chip and was used to create a hybrid bilayer membrane (HBM) and the subsequent development of the L1 chip to give a bilayer. The SPR analysis of the membrane interactions of magainin and melittin was initially explored through binding to an HBM and lipid vesicles formed on the HPA or to the bilayer deposited onto the L1 chip, respectively. The comparative sensorgrams obtained for each peptide with both DMPC and DMPG membranes demonstrated that magainin 1 binds more strongly to negatively charged lipids (Figure 16, panels A and B), and the strong binding was also accompanied by a high helical content for magainin 1 in liposome solution containing anionic lipids as measured by CD. This result agreed with results reported by other groups, where magainin preferentially acts on bacterial cells containing large amounts of anionic phospholipids via electrostatic interactions.731,734 In contrast, melittin exhibited strong binding in relative terms to both zwitterionic and anionic lipids (Figure 16, panels C and D), which correlated with the fact that melittin binds and lyses both bacterial and eukaryotic cells. It was also demonstrated that melittin and magainin had higher affinities for both DMPC and DMPG membranes formed on the L1 surface compared to those formed on the HPA surface.49,651,652 Kinetic analysis of the data also yielded affinity constants that were very similar to values obtained by other techniques,735,736 thereby validating the use of SPR to provide quantitative information on peptidemembrane interactions. The binding mechanism of magainin and melittin have also been characterized on zwitterionic PC/cholesterol (10/1, w/w) and negatively charged PE/PG (7/3, w/w) lipid mixtures by SPR. Binding affinities were determined by fitting the sensor response curves to a two-state model which demonstrated a strong correlation between the membrane affinity values and the membrane disrupting properties. Furthermore, the relative ratio of affinities of the peptides on the HBM formed on HPA versus the liposomes on L1 chip was again utilized to derive information on the mechanism of action.652 The affinity constants ratio for magainin between bilayers and hybrid monolayers for zwitterionic membranes was approximately = 1, consistent with the preferential surface interaction of magainin 5447

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

signaling molecules such as the myristoylated alanine-richkinase substrate (MARCKS),744 protein kinase C,745 neuromodulin,746 secretory phospholipase A2,747 and the HIV gag protein748 involve recruitment of the protein into receptorinduced signaling complexes at the cytoplasmic surface of the cell membrane. This often involves protein−protein interactions as well as hydrophobic and electrostatic interactions between the positively charged face of a protein domain and the negatively charged membrane surface. Furthermore, discrete cholesterol-containing microdomains in the plasma membrane or so-called lipid rafts have also been shown to play a central role in the function of Ras proteins in signal transduction.749 Indeed, the regulatory consequences of the reversible membrane interactions of proteins classified as “amphitropic”750 highlights the critical importance of membrane interactions in signal transduction. Apart from recruitment, subtle differences in the relative affinity of peptides and proteins for the phospholipids, as well as the orientation and degree of insertion of the peptide into the lipid bilayer, also contributes to the biological function of these proteins. The orientation of a peptide at the membrane surface and/or the degree of insertion into the membrane interior could also play a role in the recruitment and assembly of signaling complexes. The activation of signaling pathways via the binding of ligand to receptor often culminates in the recruitment of signal proteins to the cell membrane, and SPR has been used to study the membrane-binding features of several amphitropic, signaling proteins. Membrane-mediated cell signaling often involves proteins which the critical cysteine residues have undergone posttranslational modification with lipid. However, the preparation of lipid-modified recombinant proteins in high yield presents a significant technical challenge. Bader et al. have reported a method for the covalent coupling of lipidated peptides to recombinant proteins.714 By modifying Ras constructs with different lipid molecules, the SPR binding responses showed that the lipid-modified Ras proteins can insert into artificial membranes prepared on an HPA chip. The activation of signaling pathways via the binding of ligand to receptor often culminates in the recruitment of signal proteins to the cell membrane, and SPR analysis provides more insight into the membrane-binding features of several amphitropic, signaling proteins. Central to many signaling cascades is the Ca2+dependent binding that induces a conformational switch and represents a key regulatory mechanism in membrane recruitment and subsequent signaling cascades. Membrane SPR systems have been widely used to understand these Ca2+mediated membrane binding processes for proteins such as recoverin,715,716 neurocalcin,719 and the C1 and C2 domains in annexin I.725 Lange et al. used SPR to study the mechanism of the Ca2+myristoyl switch of recoverin on biotinylated liposomes which were captured onto streptavidin-modified chips.715 Recoverin is an amphitropic, N-myristoylated Ca2+-binding protein that serves as a calcium sensor in visual transduction. The Ca2+dependent membrane association of recoverin is also dependent on the myristoyl modification (or Ca2+-myristoyl switch). The binding of recoverin to either artificial liposomes or liposomes derived from membranes of the rod outer segment was dependent on both Ca2+ and the myristoyl group. The association kinetics for recoverin was fast and biphasic, while dissociation was faster at lower concentrations Ca2+. In a complementary approach, another group studied the binding of

antimicrobial peptides can be used to modulate membrane selectivity. Kinetic analysis of membrane binding data using either the 1:1 or two-state binding model has clearly demonstrated that many peptides and proteins bind via at least two steps, first electrostatic interactions followed by reorientation and/or insertion into the bilayer. SPR technology thus allowed a more refined description of the mechanism of binding than was otherwise possible. In addition to the studies on AMPs described above, the two-state model has been used to delineate the binding pathway of the protein toxins equinatoxin II and perforin.700,737,738 Many toxins attach to the host membrane followed by insertion and pore formation. The mechanism of pore formation is likely to differ among the large number of different toxins and is generally poorly understood but involves a number of steps incorporating the binding of the water-soluble monomer to the membrane followed by oligomerization of three to four monomers on the membrane surface. In addition, two-state kinetic analysis of the SPR response curves has also been performed for (1) understanding the T cell membrane binding mechanisms and cell entry of the transmembrane sequence of the T-cell antigen receptor α subunit,686 (2) demonstrating the toxicity of prion protein PrPSc fragment 106−126 [PrP(106−126)] is not a consequence of peptide-membrane interaction and pore formation,692,693 and (3) amphipathic α-helical regions of apo A-I for high affinity to lipids.49,682,739 On the basis of these results, a two-step binding process was proposed where the peptide binds and localizes itself on the membrane surface by electrostatic interactions and then relocates itself on the surface via hydrophobic interactions again demonstrating the interplay between electrostatic and hydrophobic interactions in membrane-binding. Other specific questions relate to the influence of secondary structure and the impact of single tryptophan residues on membrane binding and insertion,678,740 illustrating the role of tryptophan in anchoring peptides at the interfacial region of the bilayer revealing significant differences in binding characteristics and kinetics even between peptides with little differences in their structure.676,679,741 SPR-membrane systems have also allowed the role of specific lipid components in the membrane selectivity of peptides and proteins to be investigated. This approach has thus identified the important role of PE in the membrane activity of the cyclotides,668−673 sphingomyelin in the binding of lysenin,704 the role of LPS in the membrane binding of lactoferrin,705 polymyxin B,694,695 and glycopeptide antibiotics,573,699,742 D-Ala-D-Ala of bacterial cell wall in the binding of vancomycin and analogues,743 and Ca2+-dependent binding of damptomycin to G(+) membrane mimics, POPG and POPG/CL (1:1).697,698 Most significantly, SPR has been used to understand the importance of PS/Ca2+ in the activation of factor VIII582,706,707 and the recruitment of proteins involved in intracellular signaling as discussed below. All-in-all, these studies provided several examples of the application of SPR to mechanistic studies of cytolytic peptides to provide estimates of the membrane binding affinity and the identification of at least two steps in the membrane interaction which together with other biophysical techniques are likely to represent membrane binding and insertion. 4.2.4. Membrane Interaction of Peripheral Membrane Proteins in Membrane-Mediated Signaling. In the context of signal transduction, the activation of many intracellular 5448

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Ca2+-containing liposomes to immobilized recoverin.716 Large changes in SPR signals were observed with a range of metals, and they concluded that a large conformational change occurs upon metal binding followed by membrane binding. The membrane binding properties of monoacylated neurocalcins modified with different acyl chains were studied by SPR and cosedimentation.719 Neurocalcin, a member of the family of neuronal calcium sensors that belongs to the superfamily of EF-hand Ca2+-binding proteins, is myristoylated on its Nterminus and can associate with biological membranes in a Ca2+- and myristoyl-dependent manner. The SPR results indicated that neurocalcin was able to associate with membranes comprising brain-derived lipids irrespective of whether it was modified with lauric, myristic, or palmitic acid, indicating that the Ca2+-myristoyl switch can function with different lipid moieties. The membrane-aggregating activity of annexin I has also been examined using SPR.725 Annexins are a family of proteins that reversibly bind membrane-containing anionic phospholipids in a Ca2+-dependent manner, and these studies were aimed at delineating the mechanism of annexin I-mediated membrane aggregation. The binding of annexin I to immobilized liposomes was monitored and then followed by a secondary binding of another solution of liposomes to monitor liposome aggregation. The results revealed a direct correlation between the membrane aggregation activity and the relative affinity for the secondary membrane and support a model whereby annexin I first binds to specific phospholipids in the membrane and then interacts directly with the secondary membrane components via hydrophobic interactions. These results can be compared to a recent study which used a quartz crystal microbalance (QCM) to study the interaction of annexin A1 with solid-supported bilayers immobilized on gold electrodes.726 They investigated the effect of cholesterol on the affinity and rate of membrane interactions with different phospholipids and further illustrate the molecular insight that can be obtained with membrane-based biosensors. Phosphatidylinositol phosphates (PIPs) in the cytoplasmic leaflet of a membrane bilayer are one of the key regulators in cellular signaling. Membrane SPR has been used to investigate the specific roles of different PIPs in the function of phospholipase D and Phox homology (PX) domains. A hybrid bilayer membrane was used to study the regulation of phospholipase D (PLD) activity in relation to its role in the release of phosphatidic acid and subsequent signaling pathways.717 The isoform PLD1 was found to bind specifically to monolayers containing phosphatidylinositol-(4,5)-bisphosphate [PI(4,5)P2] but interacted weakly with surfaces containing phosphatidylserine or phosphatidylinositol-(3,4,5)-triphosphate [PI(3,4,5)P3]. When correlated with the activity of the parent enzyme and mutants proteins with changes in the putative pleckstrin homology (PH) domain, it was concluded that the functional PH domain regulates PLD by mediating its interaction with polyphosphoinositide-containing membranes, which may also be associated with a conformational change and regulation of the catalytic activity. The binding of a series of Phox homology (PX) domain analogues to liposomes containing DOPC with and without PI(3,4,5)P3 were measured by SPR, and the results demonstrated a high affinity for PI(3,4,5)P3, suggesting that PI(3,4,5)P3 may drive recruitment of proteins to the cytoplasmic or vacuolar membranes, where it is present in high amounts.718

The interaction of several peripheral membrane proteins with model membranes have also been characterized by SPR with the L1 chip.580,721−725 These studies particularly focused on the role of electrostatic and hydrophobic interactions between the protein and the membrane through analysis of a series of protein mutants. The role of cationic, aliphatic, and aromatic residues in the membrane binding of five phospholipases A2 was characterized using liposomes composed of either 1,2-di-Ohexadecy-sn-glycero-3-phosphocholine (DHPC) or 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoglycerol (POPG).580,751 On the basis of the binding results, a mechanistic model for the interfacial binding of peripheral proteins was proposed in which the protein binds initially to the membrane through electrostatic interactions, followed by penetration into the membrane via hydrophobic interactions. This approach was also exploited in the study of the role of the C1 and C2 membrane-targeting domains of protein kinase C-α (PKC-α).580,720−724 In particular, these studies investigated the role of Ca2+ in the membrane binding of these domains. On the basis of the differential binding of specifically mutated analogues, they proposed that the initial binding of PKC-α is driven by electrostatic interactions via the C2 domain, which is mediated by a bound Ca2+ ion and specific residues in the C1a domain. This is then followed by disruption of the C1a tethering by a molecule of phosphatidylserine (PS) which in turn leads to membrane penetration and diacylglycerol binding of the C1a domain and PKC activation. In order to further characterize the mechanism by which the C2 domains mediate the membrane targeting of phosphatidylinositol-specific phospholipases C-δ isoforms (PLC), the membrane-binding of the C2 domains of PLC-δ1, PLC-δ3, and PLC-δ4 was studied by SPR using the L1 chip and correlated with subcellular localization with time-lapse confocal microscopy.724 It was found that the membrane binding of PLC-δ1-C2 and PLC-δ3-C2 involved Ca2+-induced electrostatic interactions and PS coordination which in turn controlled the Ca2+-dependent subcellular targeting to the plasma membrane. In contrast, PLC-δ1-C2 and PLC-δ3-C2 involved Ca2+-induced electrostatic interactions and PS coordination, which coincided with a prelocalization to the membrane prior to Ca2+ import and nonselective Ca2+mediated targeting to a range of cellular membranes. Together, these studies demonstrated that C2 domains of PLC-δ isoforms as a Ca2+-dependent membrane-targeting domains and play a role in subcellular localization of these proteins. In a somewhat related system, the role of the membrane in the induction of a bioactive structure was demonstrated for two human monoclonal antibodies 2F5 and 4E10 used in vaccine development against HIV-1. SPR results demonstrated that high-affinity recognition of gp41-derived epitope sequences by 2F5 and 4E10 occurs in a membrane context.712,713 Overall, these studies demonstrate the power of SPR in elucidating the molecular details of cellular signaling processes. 4.2.5. Protein Translocation Through the Membrane. Translocation or transport of proteins and peptides across cell membrane into the cytosol is a complex process involving multistep mechanisms regulated by multicomponent translocation machineries. Many protein toxins from plants and microbial pathogens are able to translocate through the cell membrane and become active, exerting toxic effects by interacting with their counterpart molecules in the cytosol. A comprehensive analysis combining the structural characteristics of proteins and the physicochemical properties of the membrane have begun to unravel the underlying molecular 5449

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Figure 17. Assay for protein translocation across the tethered SLBs characterized by SPR. (A) Design of the SPR sample cell consists of immobilized CaM on the Au substrate and the polymer-tethered SLBs creating a compartment between the SLB and Au substrate. (B) Approach used to characterize CyaA binding to and translocation across the tSLBs. (C) Sensorgram corresponding to CaM immobilization, tSLBs formation via vesicle adsorption, CyaA binding and the translocation of CyaA across the tSLB and binding to the immobilized CaM after applying voltage (−80 mV). Reproduced with permission from ref 752. Copyright 2013 National Academy of Sciences.

Figure 18. Molecular mechanism of membrane disruption consists of a series of intermediate states, each representing an ensemble of closely related bilayer structures. For the examples of antimicrobial peptides and apoptotic peptides induced bilayer disordering, three general membrane-inserted states are defined corresponding to (1) surface parallel-bound state (A−C) , (2) partially inserted state (D−F), and (3) significantly inserted state with membrane lysis or expansion (G−I). Reproduced with permission from ref 31. Copyright 2015 Elsevier.

attached, which created a continuous fluid membrane separating two distinct compartments (Figure 17). The changes in the SPR resonance angles were monitored in real time for the grafting of CaM and the tethered membrane and subsequent binding of CyaA to the membrane before and after applying a voltage. Different angle shifts were observed for the binding of CyaA to the membrane on the cis side and the subsequent translocation of CyaA across the tethered membrane for binding to the immobilized CaM on the trans side. The binding of CyaA to a 54 Å thick membrane of further increased the thickness to around 68 Å. The translocation of CyaA across the tethered membrane and subsequent interaction with immobilized CaM occurred only upon the application of a negative potential across the bilayer. The incubation of antibodies targeted to the catalytic domain at the

mechanisms of various processes involved in toxin translocation across the cell membrane. Membrane translocation assays are well-established for cell or liposome-based assays using specific labels. However, it still remains challenging to characterize the kinetic processes and related changes in the protein conformation and membrane organization throughout the translocation processes under label-free conditions. In this context, a label-free SPR technique was established to explore the voltage- and calcium-dependent translocation of adenylate cyclase (CyaA), a virulence toxin produced by Bordetella pertussis causing whooping cough, across a model membrane tethered on the sensor chip surface.752 The model membrane platform consisted of multilayer assembly of a DSPE-PEG3400SPA-decorated PC membrane tethered on a cysteaminemodified Au substrate with calmodulin (CaM) covalently 5450

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Figure 19. Analysis of changes in bilayer order (as measured by the birefringence, Δnf) as a function of membrane-bound peptide mass (mp) allows the impact of peptide binding in the membrane structure to be determined for various sequential steps corresponding to surface binding, partial-full insertion, membrane expansion, and membrane lysis, whereby decreases in Δnf correspond to disordering in membrane structure. (A) Binding of human blood coagulation factor XI (hFIX) to POPC/POPS (80:20) SLB showed no significant changes in birefringence with a large amount of protein bound to the membrane,606 (B) magainin2 (Mag2) induces small disorder in DMPC bilayer above a threshold concentration and such disorder is fully reversible after the peptide dissociation,606 (C) the accumulative binding of transmembrane domain of proapoptotic Bax protein to the POPC showed little increases in POPC order.599 (D and E) The binding of citropin1.1 and maculatin1.1 to a DMPC bilayer showed no changes in bilayer order at low peptide concentration, while significant disorder in the bilayer was observed as peptide concentration increased.608,754 In contrast, (F) the insertion of HPA peptide into DMPC/DMPG (80:20) induced significant irreversible bilayer disorder with small amounts of membrane-bound peptide mass.596 (G) The lysis of DMPC membrane by aurein1.2 showed multiple sequential changes in bilayer disorder where the birefringence returned to the starting values with significant loss in overall mass,597 while (H) binding of magainin 2 analogue (Ala8,13,18-mag) to DMPC showed significant mass loss with irreversible bilayer disorder as a result of bilayer expansion.606 (I) Binding of jumper ant toxin, Δ-myrtoxinMp1a to bacterial model membrane (DOPE/DOPG = 3:1) resulted in significant irreversible bilayer disorder with increasing membrane-bound peptide mass.600 (F) and (G) Reproduced with permission from refs 596 and 597. Copyright 2010 Elsevier. (D) and (E) Reproduced with permission from ref 608. Copyright 2014 Elsevier. (C) Reproduced from ref 599. Copyright 2014 American Chemical Society. (A) and (B) Reproduced with permission from ref 606. Copyright 2014 Springer Nature.

C-terminal end of CyaA had no influence on membrane binding, but this antibody inhibited the translocation of CyaA across the membrane. This novel system demonstrated the clear advantage of this approach over SLBs and dropletinterface bilayers753 for monitoring the transfer of proteins across a model membrane and detecting the subsequent interaction with intracellular molecules.

an anisotropic system with a unique optical anisotropic property. The degree of molecular order, S, of the uniaxial lipid bilayer defined by the ratio of the principal polarizabilities of the bilayer to the molecular polarizabilities168 is proportional to the birefringence values (Δnf). Thus, it is the birefringence values that represent an averaged measurement of lipid molecular orientation order and lipid acyl chain packing order. High Δnf values are obtained for a fully aligned lipid bilayer, whereas low Δnf indicates a random and disordered lipid bilayer. As listed in Table 2, the birefringence values obtained for planar lipid bilayers of various compositions range from 0.015 to 0.025 refractive index units. These values provide a baseline level for investigating the kinetic changes in membrane structure upon proteins/peptides, polymers, and ion membrane interactions.

4.3. Characterization of Peptide Binding and Changes in the Bilayer Integrity by Mass and Birefringence

As described in earlier sections, the temporal changes in the optical thickness and mass can be used to monitor membrane formation and quality assurance of the lipid bilayer structure. However, the mass or thickness alone does not reveal the dynamic changes in molecular ordering of a lipid bilayer. The ordered orientation of lipid molecules in a membrane creates 5451

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

dM 2 /dt = ka2M1 − kd 2M 2 − ka3M 2 + kd3M3

The ability of DPI to measure both mass and birefringence with high sensitivity allows the characterization of very subtle dynamic changes in orientation and packing order of lipid molecules, revealing mechanisms of membrane-mediated interaction that are difficult if not impossible to see by other means. For the initial analysis, plots of mass versus time (Figure 3D) provide information on binding kinetics analogous to other optical biosensors such as SPR. For further insight, dependence of birefringence on time (Figure 3D) offers unique information on the simultaneous changes in membrane structure associated with binding events. More significantly, the combination of these plots to give the birefringence versus mass plots (as illustrated in Figures 18 and 19) reveals an enormous amount of new information. Both the initial binding of an analyte (such as a peptide), and any subsequent processes, may affect both the mass and structural ordering of the membrane, and the relationship between the two can be visualized using the plot of birefringence versus mass. 4.3.1. Kinetic Analysis of Experimental Data to Define Binding Mechanisms. Studies of membrane-active peptides often focus on the overall process, either in terms of the actual efficacy of the peptide such as cell lysis and cargo transport or the final end-point of mechanisms such as pore formation or micellization. However, in order to fully understand the mechanism of action of the peptides, it is important to characterize the kinetics of sequential events involved in the entire dynamic process. For example, a peptide may bind the membrane, reorient with different topology, insert further into the membrane, destabilize the membrane, and then interact with other peptides to form pores or other structural assemblies. Much research is limited to either the initial peptide binding step with respect to the surface charges or the formation of the final structure of the lipid-peptide complex, neglecting the intermediate steps. However, these intermediate steps may well be critical for the peptide to achieve its function, and understanding these steps may lead to new opportunities for the design of new therapeutics. The question that must be asked is not merely what overall mechanism(s) are involved but what the intermediate steps are, and what factors may affect the sequence of transitions that lead to the ultimate activity of the peptide. Determining the intermediate states requires a thorough characterization of peptide and membrane structural changes and the kinetics associated with different intermediate steps. This information is now emerging with optical waveguide technologies through kinetic analysis of real-time binding data using specific kinetic models and will enable the design of peptides to modulate their activity resulting in more powerful and flexible use for mechanistic understanding of biomolecular membrane interactions. Further elaborations upon the simple 1:1 Langmuir and twostate models stated in previous section are possible: the threestate model653 is the logical extension where the bound peptide, already in its second bound state, may undergo a transition to a third state. Parallel reaction models involve simultaneous reactions. In a simple case, this may involve two one-state reactions; in more complex cases, some or all reactions may have multiple states. Also, in any of above-described models, there may be co-operativity in the reactions (i.e., interactions between peptides on the surface may speed up the transition from one state to another). The three-state model is represented by the equations

dM3/d t = ka3M 2 − kd3M3

Furthermore, multiple-state models used for kinetic analysis that link mass and birefringence outputs together using different proportionalities for each state have been expanded to include deviations from the ideal model such as bilayer expansion, birefringence lag, and mass thresholds, and these changes can result in significant improvements to the fit in some cases. The models fitted to the data can be used to propose hypotheses for the overall mechanism of binding by identifying multiple states and characteristics associated with those states, including changes in membrane structure and integrity. Overall, these studies have provided new insights into the activity of a range of membrane-active peptides. 4.3.2. Profiling of Lipid Perturbation and Disruption by Membrane Active Peptides. The graphical analysis of the correlation between binding and structural changes has been used to evaluate the accumulated impact of a number of peptides to effect progressive changes in membrane structure. The birefringence analysis provides unique real-time information on the impact of peptides on the changes in organization occurring within a lipid bilayer and assists in describing the behavior of membrane-active peptides when interacting with membranes of different lipid compositions. A number of transitions have been described which can be used to evaluate discrete binding behavior and mechanisms for different peptides (Figure 17) (refs 31, 597, 600, 606−608, 611, 653, and 754−757). Increases in mass are characteristics of molecules associated with the membrane (at least initially), and such an increase in mass may also be accompanied by changes in birefringence. For a decrease in birefringence (and hence an increase in membrane disorder), changes in the slope of the birefringence-mass graph provides a measure of sequential events occurring throughout the binding process, from a horizontal to shallow slope representing a surface binding that has only a small effect on birefringence (Figure 18A), through to a near vertical decrease for a substantial interruption of the membrane structure (Figure 18, panels F and I), with other cases that fall between these two extremes (Figure 18). Alternatively, binding may also cause an increase in birefringence, suggesting that a stabilizing or ordering effect is occurring. Membrane changes subsequent to binding may include recovery in birefringence during dissociation from the membrane or further decreases in order with mass loss from the surface. This last possibility may indicate normal dissociation combined with a slow decrease in birefringence from peptide already bound, or it may indicate the loss of mass from the surface, either by membrane thinning and expanding or the removal of material from the surface. With the study of appropriate model systems, it may become possible to characterize peptide-membrane interactions directly by observing the dynamic relationships between mass and structural organization. This approach has greatly facilitated the differentiation of different modes of peptide-membrane interactions. Knowledge of the membrane structure changes significantly informs our understanding of biological processes categorized based on their activity in membrane environment with specific examples including antimicrobial peptides, cell-penetrating peptides, amyloid peptides/proteins and apoptotic peptides, and G protein-coupled receptors described below.

dM1/dt = ka1CP(M max − M1 − M 2) − kd1M1 − ka2M1 + kd 2M 2 5452

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Figure 20. Changes in the membrane order as a function of peptide mass. (A) The changes in the molecular order of supported unilamellar DMPC bilayer induced by the alanine-substituted magainin 2 (Ala8,13,18-Mag2) showed multistage process of membrane destabilization with the increases of membrane-bound peptide mass. The peptides were injected consecutively onto the DMPC bilayer from 1.25 to 40 μM with a 2-fold increment in concentration. The association and dissociation of Ala8,13,18-Mag2 are denoted by the solid arrow and dotted line, respectively. The extent of bilayer disordering from binding leading to membrane expansion correlated well with the sequential changes of spreading of lipid molecules on the solid support bilayer imaged in real-time by AFM. The peptide induced progressive structural changes in membrane involves multiple states a, b, and c of peptide-lipid complexes. Distinctive changes in membrane structures occur at two different peptide/lipid (P/L) ratio labeled as mp*, which corresponds to the initial membrane disordering, and mp**, which corresponds to the onset of membrane disruption with mass loss. (B) Following the initial disordering of lipid molecules occuring at mp* (state b), more peptides continue binding to the membrane and reach a mp* at different molar P/L ratios for various AMP action on either DMPC or DMPC/DMPG. (C) The process of these changes for AMP binding to the membrane is presented schematically for the membrane disruption and expansion that occurs and results in the loss of total mass (state c). (A) and (B) Adapted with permission from ref 31. Copyright 2015 Elsevier. (A) and (B) Adapted with permission from ref 606. Copyright 2015 Springer Nature.

peptidases to LPS,758 novicidin to DOPC/DOPG,759 cationic KYE28, KYE21, and NLF20 derived from human heparin cofactor II binding to LPS and its lipid A moiety,614 and citropin 1.1, maculatin 1.1, and caerin 1.1 binding to DMPC.608 However, this linear decrease in birefringence with increased peptide mass seen in one membrane can become more complex with changes in either peptide sequence or changes in lipid composition. The binding of HPA3 and maculatin 1.1 to the gel-DMPC bilayer displayed similar linear behavior with bilayer structural disordering.596,607,754,760 However, the helical kink in maculatin 1.1 induces a biphasic change with initial increases in membrane-bound peptide mass without significant bilayer disorder followed by an abrupt nonlinear disordering with further increases in mass upon binding to the fluid DMPC bilayer (Figure 19E). In contrast, the linear helical HPA3 showed a linear decrease in bilayer order with the gel-DMPC/ DMPG bilayer (Figure 19F). These studies clearly demonstrated that the role of proline in bilayer perturbation can be delineated by the changes in birefringence, which is not possible by mass-only measurements. These distinctively different features in binding to DMPC/DMPG, despite similar binding behavior with DMPC between HPA3 and maculatin 1.1, also demonstrate the importance in differentiating the membrane structural changes from the peptide binding, since in the absence of the birefringence data, the main conclusion would be that both peptides have similar binding mechanisms. The birefringence-mass profiles can also provide significant new insight into current models of AMP action.31,606,608,656,756,757 While many studies have focused on defining the critical concentration for membrane destruction,761

4.3.2.1. Antimicrobial Peptides. The interaction of AMPs with a membrane can result in structural and topological changes in peptides and perturbation of membrane integrity. The activities of AMPs in selectively destabilizing membrane structure and function are determined by complex factors. Although substantial structural information is available for the AMPs, the dynamic changes in membrane structure in response to the peptide binding is far less studied but requires quantitative characterization in order to delineate the mechanisms of action of AMPs. When considering various models, multiple modes of action may be present simultaneously, and the mode of action may be dependent on variable factors such as the concentration of peptides in solution and membrane-bound, the lipid membrane composition and structure, and the kinetics of binding and structural changes of all interacting partners. DPI analysis of the structural characteristics of an adsorbed lipid bilayer showed that most AMP binding results in an essentially linear decrease in the bilayer birefringence with increasing peptide mass bound to membrane. As shown in Figure 19 (panels D and F) for citropin 1.1 binding to DMPC and HPA binding to DMPC/DMPG (80:20), this decrease in birefringence immediately after peptide binding is a characteristic pattern for peptides incorporated into the membrane without apparent threshold on the membrane surface.597,608 These patterns have been observed for aurein 1.2 binding to E. coli lipid extract and DMPE/DMPG,597 HPA3, and HPA3P binding to DMPC and DMPC/DMPG,597 peptides derived from C-terminal region of human thrombin binding to lipopolysaccharide (LPS),613 C-terminal peptides from S1 5453

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

birefringence decreased from 0.026 to −0.06. This clearly showed the disordering effect and lipid disintegration in response to the antibacterial polymer. On the basis of the composite model, the surface coverage of antibacterial polymer in the final structure was also calculated with less than 40% of the area covered with lipids at the end of the measurement. Selectivity toward specific target cells is particularly crucial to develop new AMPs and other therapeutic agents acting on the membrane. Peptides binding to model membranes with defined lipid composition and structural properties has demonstrated the complex interplay between charge, hydrophobicity, amphipathicity, bilayer structural organization, presence of sterol, and sphingomyelin, etc. in mediating binding selectivity. In addition, modification of peptide sequence, length, and endtagging with a stretch of either charged or hydrophobic residues can also enhance the selection toward pathogenic bacteria without toxicity toward host cells. What can be concluded is that these peptides have significantly different binding properties related to differences in the effect on membrane structure, and that the bioactivity of these peptides is likely to be mediated by significant changes in membrane structure. This has also been shown for a significant change in the bilayer order upon binding of maculatin 1.1, and this change occurred in a co-operative manner at higher concentrations of maculatin 1.1 with increasing bilayer thickness and order.756 Thus, an optimum bilayer thickness and lipid order may be required for effective membrane perturbation by maculatin 1.1, and increasing the bilayer thickness and order may counteract the activity of maculatin 1.1 and play a role in resistance to AMPs. These structural effects which have not been previously demonstrated may involve the selectivity for specific membrane properties. More direct correlations will be possible when birefringence measurements of bacterial membrane extracts are performed. The selectivity and activities of AMPs on the membrane have also been studied for defensin NaD1 isolated from ornamental tobacco Nicotiana alata on various membranes with DPI and liposome permeabilization assays.611 Defensin is a potent antifungal molecule that also targets tumor cells forming large membrane blebs which cause the cells to burst. Different from the helical structures of most characterized AMPs, plant defensins have a conserved cysteine stabilized α−β motif (CSαβ) composed of one α-helix and three β-strands stabilized by four conserved disulfide bonds. Binding of NaD1 to bilayers containing PI(4,5)P2 is fast and strong in contrast to the relatively slow and weak binding to the control membranes without PI(4,5)P2. No dissociation of NaD1 from the PI(4,5)P2 bilayers indicated that the binding is irreversible and associated with large disordering, as shown by decreases in birefringence. This selective binding to PI(4,5)P2 correlates with the ability of NaD1 to permeabilize only the membranes containing PI(4,5)P2. This finding revealed a novel membrane destabilization mechanism involving oligomerization of the NaD1-lipid complexes and cause a massive loss of membrane order, leading to membrane leakage distinct from the nonspecific action of helical AMPs. As PI(4,5)P2 is not found on the outer leaflet of plasma membranes, multiple mechanisms may exist for the translocation of NaD1 into the cell and to form complexes with the PI(4,5)P2 on the inner leaflet. Nevertheless, the role of PI(4,5)P2 on selective binding and membrane disordering provide a novel interaction model for membrane permeabilization.

characterizing the steps associated with different extents of bilayer perturbation at a distinctive P:L threshold provides a much clearer understanding of the cooperative molecular mechanisms of AMP action on the membrane. The sequential steps of binding, insertion, and bilayer disruption have also been examined in real time with DPI at continuous increments of peptide:lipid (P:L) ratio for various AMPs. Aurein 1.2 and magainin 2 are naturally found in the dorsal secretion of an Australia tree frog and an African frog, respectively.762 Despite the difference in sequence and in particular the length, with 13 and 23 amino acids for aurein 1.2 and magainin 2, respectively, both peptides adopt an amphipathic helix in an anionic membrane environment, while no structure was found for magainin 2 in a neutral membrane environment. The carpet mechanism has been proposed as the mechanism by which both peptides lyse membranes based on similar critical peptide:lipid (P:L) ratios as measured by conventional means. However, DPI can provide information on membrane structure changes immediately prior to, during, and after membrane lysis, and significant differences are apparent for these two peptides which are described as acting by the same mechanism. First, the P:L ratio determined at the point of disruption for an anionic membrane was markedly lower for aurein 1.2 than for magainin 2 (Figure 19, panels B and G, and Figure 20B). Moreover, an alanine-substituted analogue, Ala8,13,18-mag2 with enhanced bactericidal activity also showed anionic membrane disruption at a P:L ratio similar to magainin 2 (Figure 20B). However, the Ala8,13,18-mag2 also disrupted the neutral DMPC membrane (Figures 19H and 20A) at nearly double the P:L ratio. From the P:L ratio determined at the point of membrane disruption, a “carpet mechanism” would be simply attributed to the mechanisms of action of these frog peptides. However, the drop in birefringence was reversible for aurein 1.2 after its disruption of both neutral DMPC and anionic DMPC/DMPG membranes.597,606,608 In contrast, a partially reversible birefringence drop with permanent disordering was obtained for Ala8,13,18-mag2 after membrane disruption. These partially reversible birefringence changes with membrane disruption correlated with the peptide insertion and bilayer expansion as evident from AFM studies (Figure 20A), and both fluorescence dye release and AFM results showed that aurein 1.1, magainin 2, and Ala8,13,18-mag2 disrupt the membrane differently.597,606,608 Overall, these experiments clearly demonstrate that the definition of AMP action in terms of a specific model can be too simplistic. In particular, the reversible packing disorder−order behavior observed for different bilayers is quite distinctive for various AMPs, behavior which is not apparent with other spectroscopic techniques. The analysis of membrane disordering has therefore extended our understanding of how the membrane undergoes a reversible structural change upon exposure to antimicrobial peptides and how the bilayer can either recover or, at a critical peptide concentration, begin to disintegrate (Figure 20C). Studies with the use of another interferometry-based waveguide sensor, optical waveguide light-mode spectroscopy (OWLS), also showed in situ changes of optical birefringence upon the binding of antibacterial polymers to the SLBs.763 The calculation of layer thickness and birefringence values was based on an anisotropic adlayer model while the correlation of layer thickness, and birefringence was assessed comparatively with a composite/exchange model. The addition of antibacterial polymer to a bilayer resulted in initial increases in thickness from 5.2 to 8 nm and decreased to 5.8 nm, while the 5454

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Δ-Myrtoxin-Mp1a, a 49-residue heterodimeric peptide from the venom of the jack jumper bull ant (Myrmecia pilosula) is comprised of a 26-residue A chain and a 23-residue B chain connected by two disulfide bonds in an antiparallel arrangement. The membrane binding characteristics and the effects on membrane structure of Δ-Myrtoxin-Mp1a and various analogues were studied using DPI on geometrically well-defined POPC, POPC/POPE/POPS (60:30:10 molar ratio, mammalian plasma membrane mimic), and DOPE/DOPG (75:25: molar ratio, bacterial membrane mimic) SLBs.600 Quite distinct binding patterns were observed for the heterodimeric Mp1a compared to the individual monomeric Mp3a and 4a peptides. The real time changes in mass for each peptide binding to SLBs showed a rapid increase in mass during the association phase reflecting strong binding. However, the overall amount of each peptide was similar on the mammalian membrane mimics (POPC and POPC/PE/PS) but much higher on the bacterial membrane mimic (DOPE/PG), indicating a higher binding affinity on these membranes. Furthermore, the disordering effect of each peptide was also greater on DOPE/PG than either POPC or POPC/PE/PS and was essentially irreversible (Figure 19I). In contrast, the parallel heterodimer Mp2a showed negligible binding or bilayer-disruptive effects on POPC, thus indicating a selectivity for negatively charged bacterial membranes. Despite the strong membrane binding of Mp4a to DOPE/PG, Mp4a was less effective in bilayerdisrupting as demonstrated by the smaller decreases in birefringence compared to Mp1a, 2a, and 3a relative to the bilayer-bound peptide mass. These membrane binding and disruption characteristics of Mp1a are consistent with the induction of concentration-dependent Ca2+ influx in SH-SY5Y and HEK293 cells and the broad spectrum antibacterial activities with highest potency against G(−) Acinetobacter baumannii. In addition, the monomeric Mp3a and 3b were less active against G(−) strains compared to Mp1a, which is also consistent with the less binding and disordering effect of monomeric peptides on the membrane. These findings demonstrated that the novel analysis of binding and membrane disordering allows the differentiation of the peptide action on membrane in delineating the mechanisms of membrane selectivity. 4.3.2.2. Toxins. The selective binding of the lysinesubstituted μO-conotoxin MfVIA analogues to the negatively charged POPS in modulating the activities of analgesic Nachannel Nav1.8 was also characterized by the changes in massbirefringence measured by DPI and SPR.609 After several rounds of alanine-scanning mutagenesis for the identification of important residues involved in the inhibition of voltage-gated Na channel Nav1.8, a double mutant, (E5K, E8K)MfVIA, was identified with greater positive surface charge. This (E5K, E8K)MfVIA analogue exhibited greater affinity for POPC/ POPS than POPC but very little membrane binding to POPC/ POPS. The binding of (E5K, E8K)MfVIA to POPC/POPS was also found to induce large and irreversible bilayer disordering compared to little and reversible disordering of POPC, even when significant amounts of toxin were bound. The binding and membrane disordering and together with results from differential fluorescence quenching assays lead to the conclusion that (E5K, E8K)MfVIA inserts into the membrane with greater depth than NfVIA. The higher affiinity of (E5K, E8K)MfVIA for membranes was also correlated with enhanced potency on inhibiting Nav1.8 and an apparent decrease in polarization-induced dissociation compared with MfVIA. This

membrane-binding analysis demonstrated that membrane interaction also plays a critical role in the inhibition of Nav1.8 by μO-conotoxin MfVIA driven not only by peptide−ion channel interaction but also via a putative binding site located in the membrane voltage-sensing domain. Building on these membrane interactions, a more potent mutant (E5K, E8K)MfVIA may guide the design of other novel selective inhibitors of NAv1.8 with better therapeutic efficacy. 4.3.2.3. Cell Penetrating Peptides. The transport of peptides and proteins into the interior of the cell remains one of the most challenging obstacles to the development of drug delivery systems due to the size and hydrophilic nature of the biomolecules involved.764,765 The use of cell-penetrating peptides (CPPs) is one of the most promising approaches for overcoming this difficulty in delivering various biomolecules into cells.766,767 While a variety of CPP sequences such as Tat peptide, penetratin, polyarginines, and transportan768 have been determined, Pep-1 is a CPP of particular interest, not only due to outstanding delivery rates but also due to the fact that the mechanism is exclusively physically driven.685 The membrane affinity of Pep-1 was verified to be high in all conditions tested with SPR, but the highest binding was observed on negatively charged phospholipids. Pep-1-membrane interaction was a fast process described by a multistep model initiated by peptide adsorption, primarily governed by electrostatic attractions, and followed by peptide insertion in the hydrophobic membrane core. In the context of a cell-based process, the translocation of Pep-1 is a physical mechanism promoted by peptide primary amphipathicity and asymmetric properties of the membrane. This explains the high efficiency rates of pep-1 when compared with other CPPs. For tracking the progress of peptide binding in real time, both DPI and PWR have been used to study mass and structural changes in the membrane during penetratin binding.755,769 DPI was also used to simultaneously measure changes in mass per unit area and birefringence during the binding and dissociation of penetratin (RQIKIWFQNRRMKWKK), and a biotinylated truncated penetratin, R8K-biotin (RRMKWKKK(Biotin)-NH2), with somewhat reduced efficiency to cross the membrane.765 The mechanism of membrane binding of penetratin and R8K-biotin were further characterized with DPI using multiple-state kinetic analysis.653,755 Strong preference for binding negatively charged POPC/POPG membrane was observed for both peptides, confirming similar results obtained from SPR analysis. However, clearly different mechanisms were obtained for these two peptides binding to the membrane based on the observed birefringence. Penetratin exhibited a distinctive binding pattern, best represented by a three-state kinetic model; on the fluid-state membrane POPC/ POPG compared with that observed on gel-state DMPC/ DMPG, which was adequately represented by a two-state model. The three states observed for POPC/POPG can be characterized as (I) a labile surface-bound state, (II) a securely bound intermediate, and (III) membrane disruption following translocation of the peptide across the bilayer, with a transient state likely to facilitate the transition from (II) to (III) although this cannot be observed or modeled. However, R8K-biotin did not bind well to DMPC/DMPG and showed a more transitory and superficial binding to POPC/POPG. Comparing the two modeling results for POPC/POPG suggests an important role for a securely bound intermediate prior to penetratin insertion and translocation. 5455

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

4.3.2.4. Amyloid Peptides and Proteins. The association of misfolded protein and peptide aggregates with neuronal cell membranes may play a key role in their neurotoxicity in several degenerative disorders, but the underlying mechanisms of amyloid growth and toxicity are still not fully understood. βAmyloid peptides (Aβ) are known to bind to negative membranes, and this binding promotes the formation of toxic β-sheet conformation for their cellular toxicity.770 The quantitative evaluation of affinity constants and specificity of β-amyloid-membrane interactions still remains technically challenging. The development SPR methods for quantitative analysis of a range of peptide and protein interactions have been extended to explore the cytotoxic mechanism of amyloid peptides and proteins action on cell membranes.690,771−774 The SPR binding of Aβ to membranes formed on the L1 surface was compared to a magnetic bead assay that monitored the membrane binding over very long time periods.688 Monomeric Aβ adsorbed quickly and reversibly to POPC and POPG SLBs, while aggregated Aβ adsorbed slowly and irreversibly to both SLBs again reflecting the significant changes in the physical properties of aggregated Aβ in affecting membrane binding. Aβ peptides bound both to synthetic lipid mixtures and to an intact plasma membrane preparation isolated from vascular smooth muscle cells (SMCs) and showed a good correlation between membrane binding and cellular toxicity.687 Increased amount of peptide binding to lipids was also shown for “aged” peptides with increased proportion of oligomeric Aβ species associated with enhanced toxicity. The toxicity of various Aβ analogues also correlated with their lipid binding. Significantly, lipid binding was totally dependent upon the concentration of cholesterol in the lipid mixture. Reduction of cholesterol in vascular SMCs not only reduced the binding of Aβ to purified plasma membrane preparations but also reduced Aβ toxicity in culture. These results strongly supported the view that Aβ toxicity is a direct consequence of binding to lipids in the membrane and that reduction of membrane cholesterol using cholesterol-lowering drugs may be of therapeutic benefit because it reduces Aβ-membrane binding. Glycolipids such as monosialoganglioside, GM1, has been suggested to be actively involved in the development of Alzheimer’s disease due to its ability to seed the aggregation of Aβ.775,776 The binding specificities of amyloid-related peptides to different glycosphingolipids has also been studied using SPR in an effort to gain insight into the role of membrane interactions in the pathology of AD.689,777 On the basis of the relative binding of a series of peptide analogues of the parent peptides Aβ1−40 and Aβ1−42 with liposomes containing different gangliosides, it was demonstrated that 2,3-N-acetylneuraminic acid bound to the internal galactose residue on a neutral oligosaccharide core and is required for Aβ1−40 binding. Furthermore, an additional sialic acid residue linked to the other neutral oligosaccharide core structure increased the binding affinity, thereby providing quantitative information that could be used to rationalize amyloid peptide function. The interaction of a series of neuropeptides including Aβ1−42 with membranes containing different gangliosides was also studied to gain more insight into the role of these glycolipid receptors in various neuronal systems.778 The affinity constants for all peptides ranged from 10−3 ∼ 10−7 M, and the sialic acid moiety on the ganglioside was found to play an important role in the interactions. Higher affinities were observed for polysialo- than monosialiogangliosides. Overall, these studies further highlight the ability of SPR measurements to increase our understanding

of the function of various components in the membrane in Aβinteractions. Formation of amyloid aggregates has more recently been studied by DPI which showed that aggregation is mediated by various types of surfaces, and the kinetics of aggregation is significantly enhanced upon adsorption to a hydrophobic surface.779 Measurements of mass, thickness, and density allowed the very early stages of Aβ40 and Aβ42 aggregation to be characterized by DPI. The initiation of aggregation was accompanied by a decrease in density without significant changes in thickness while during the elongation and higherorder aggregate an abrupt increase in thickness was obtained.779 The role of raft domains in model neuronal membranes on the behavior of raft-dependent Aβ aggregation has been characterized by the spectral shift of s- and p-polarization obtained with PWR spectroscopy.780 Aβ40 peptides bound to all model membranes composed of DOPC, sphingomyelin (SM), SM/Chol, and DOPC/SM/Chol while only aggregates bound in the presence of SM. With the presence of cholesterol in the SM bilayers, a 5-fold stronger binding than to DOPC and SM-only bilayers was accompanied by a mass loss attributed to the removal of lipid molecules from the bilayer and transfer to the Gibbs border, which occurs upon peptide insertion into the bilayer and aggregation leading to lipid displacement. The ability of Aβ to preferentially bind and insert into the densely packed thicker SM microdomain over the less-ordered thinner DOPC domain was characterized as a biphasic binding process on the DOPC/SM/Chol ternary lipid bilayers. A transition from an initial positive spectral shift to a negative spectral shift with increasing membrane-bound peptide mass allowed the process of binding, insertion, and bilayer expansion to be observed for Aβ aggregation on membranes. In addition to characterizing the mass changes, the changes in the optical properties (RI and thickness) of lipid bilayers during the process of Aβ binding and aggregation were quantitatively characterized by spectral simulation of the experimental spectral parameters, which provided further understanding of the effect of amyloid aggregation on membrane perturbation. No change in the bilayer properties was observed during the initial Aβ binding to the bilayer surface of DOPC and SM, while a large decrease in the RI was observed for the cholesterol-containing bilayers, which facilitated the insertion and aggregation of Aβ involving bilayer reorganization. PWR analysis combined with liposome dye-release and cryo-TEM of the binding of various Aβ1−42 variants to DOPG bilayers showed the effect of oligomeric Aβ1−42 on membrane perturbation and cell toxicity.781 The changes in the mass density and structural ordering upon the formation of nontoxic (WT) and toxic HET-s prion peptides-membrane complex has also been characterized by PWR.622,782 The role of lipids in mediating the binding and aggregation of HET-s were differentiated by resolving the angle shift, the rate of binding, and the extent of the membrane reorganization. The peptide accumulation promoted by electrostatic interactions was demonstrated by larger spectral shifts for peptides binding to the negatively charged DOPG SLBs than to neutral DOPC. Such enhanced binding to DOPG was also associated with an increase in membrane thickness of 3 nm. In addition, lateral membrane reorganization of different microdomains exhibiting different optical properties can be measured by the splitting in the resonance spectra. The interaction of toxic protein with DOPG formed three domains with different thicknesses: 60% of the membrane area showed 5456

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

caspases. The function of the Bcl-2 proteins in apoptosis has been linked to their direct interaction with the mitochondrial membrane leading to MOMP.783−787 In recent years, the role of lipids in the Bcl-2 regulated apoptotic pathway has been recognized and is central to the “embedded together” model.788−790 This draws together aspects of both direct and indirect activation but also takes into account the importance of mitochondrial membranes in facilitating key interactions such as those involved in the activation of Bax by the BH3-only protein tBid, and the inhibition of Bax oligomerization by the antiapoptotic family members. The precise role of the membrane in these interactions, however, is still unclear. More specifically, an understanding of the functional role of Bcl-2 protein secondary structure variation caused by protein− membrane interactions and its effect on relative membrane affinity and disorder will be central to understanding the events that lead to MOMP. Recent studies in this area seek to further our understanding of the mechanism by which the Bcl-2 proteins interact with the mitochondrial and plasma membranes, and in turn how these membranes are restructured upon protein contact. Membrane perturbation study based on the birefringence analysis has been focused particularly on the hydrophobic C-terminal transmembrane domain (TMD) of Bcl-2 proteins interacting with membrane mimics including plasma membrane, mitochondrial outer membrane, and mitochondrial outer membrane with cardiolipin.599 Bcl-2 related peptides bound selectively to mitochondrial mimics over plasma membrane mimics and antiapoptotic peptides exerted a different effect on the bilayer to pro-apoptotic peptides. Most significantly, membrane binding behavior of transmembrane domain-derived synthetic peptides indicate an activity-based classification of these peptides in which there is a clear correlation between the membrane disruptive properties and the biological activity.599 The TMD peptides preferentially bound to mitochondrial membranes that in turn induce conformational changes in the peptides. The TMD peptides did not perturb plasma membranes and were therefore nontoxic to cells; however, when the peptides were artificially introduced into cells they targeted mitochondria inducing MOMP and caused cell death. Importantly, at sublethal doses, the TMD peptides exerted a mitochondrial priming activity that enhances the cell deathinducing properties of the chemotherapeutic drug cisplatin. These results clearly showed that even the examination of the isolated transmembrane domain can provide significant insight into the role of the mitochondrial Bcl-2 protein recruitment. In particular, the ability to measure comparative changes in membrane structure of model plasma membrane and mitochondrial membrane by DPI has provided significant new insight into the mechanism of apoptoptic pore formation by Bcl proteins. These studies also lay the foundation for understanding selectivity of Bcl proteins for the mitochondrial membrane relative to the plasma membrane and the role of this selectivity in apoptosis and development of cancer therapeutics.

no change in thickness, while 25% showed an increase of 6 nm and 15% showed an increase thickness of 10 nm upon binding of the toxic prion aggregate. However, no significant difference was observed for the binding kinetics of either the toxic or nontoxic prion aggregate. Unlike ligand−receptor interactions, the affinity constants for amyloid-membrane interactions cannot be accurately determined for the complex coherent changes of peptide binding, self-assembly, and membrane reorganization. The changes in the optogeometrical properties for the binding and subsequent amyloid formation on the membrane have also been characterized for prion protein PrP610 and αsynuclein using DPI, which provide insight into early events in the aggregation.602 On the basis of the DPI phase shift data, the extent of aggregation was significantly enhanced by the electrostatic interaction between the positively charged PrP peptide and negatively charged POPS lipid bilayer, resulting in a thicker and more densely packed protein layer compared to those on POPC bilayers. In addition, further growth of the protein aggregate layer, as indicated by increases in both TM and TE phases, was observed upon incubation of PrP on POPC/POPS bilayer. The role of lipid composition on α-synuclein aggregation and the impact of aggregation on bilayer properties were also characterized by the changes in the optical properties in SLBs and liposomes tethered on DPI chip surfaces.602 Strong binding was mediated by the electrostatic interaction with the DOPScontaining bilayers. α-Synuclein binding was further enhanced by increasing the DOPE content in the membranes. By following the variations in mass and birefringence upon αsynuclein binding, the protein−lipid stoichiometry for these changes in the optical properties of bilayers was correlated with the mode of interaction. Two processes for α-synuclein binding to different SLBs but not DOPC suggested the initial binding caused membrane expansion as a result of insertion followed by a second stage of surface binding without a change in bilayer order. The bilayer disorder induced by α-synuclein showed a low degree of reversibility after rinsing. By comparing the optical properties of SLBs and tethered liposomes, a membrane remodelling mechanism was proposed for the α-synucleinmembrane interaction, in contrast to the transmembrane insertion adopted by AMPs. Specifically, α-synuclein binding to membranes involved insertion at the headgroup region leading to lateral expansion of lipids. The partial insertion was then facilitated and enhanced by defects in lipid packing, and the expansion of lipid resulted in the membrane remodelling and thinning. Overall, the amyloid layer may be anisotropic and its formation onto an existing anisotropic layer may modulate the overall birefringence. However, the formation of the anisotropic peptide assembly may also induce disordering in the bilayer. In addition, amyloid formation is a kinetic process and the molecular mechanism of amyloid deposition may follow a wide range of pathways, which may include rodlike assembly formation in addition to sheets, spheroids, etc. Therefore, new techniques are required to resolve the birefringence changes associated with these systems. 4.3.2.5. Apoptotic Peptides. Mitochondrial outer membrane permeabilization (MOMP) is considered the initial step in intrinsic apoptosis with the formation of the apoptotic pore, which has been related to the release of pro-apoptotic mitochondrial resident proteins, including cytochrome c (cyt c) and smac/diablo to the cytosol and subsequent activation of

4.4. Membrane Protein Receptors

In addition to the mechanistic analysis of molecule-membrane interactions and molecule-induced membrane destabilization, the function of integral membrane proteins involved in a widerange of important biological process including the signal transduction, enzymatic activation, and molecular transport are also modulated by the physicochemical properties of the membrane. A complete understanding of the role of membrane 5457

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Figure 21. Schematic of the preparation of a functional GPCR on the sensor chip surface by capture and reconstitution methods. (A) (1 and 2) capture of detergent-solubilized GPCRs to antibodies immobilized on the sensor chip surface followed by (3 and 4) the reconstitution of GPCRs into the lipid bilayer by injecting a lipid/detergent micelles. (5) Structural integrity of receptors were tested with conformation-dependent mAbs, and (6) functionality of the receptors was assessed by the binding of natural chemokine. (B) (1) Sensorgrams for the capture and reconstitution of CXCR4 and CCR5 to the 1D4 mAbs-immobilized L1 chip. (2) Binding responses obtained for conformation-dependent mAbs 12G5, 44716.111, and 444717.111 to CXCR4, while no binding is observed for anti-CCR5 and 1 Da to the CXCR4. (3) Sensorgrams for SDF-1α binding to CXCR4. Reproduced with permission from ref 799. Copyright 2003 Elsevier.

properties on the structure−function relationship of integral proteins upon interaction with ligands will greatly advance the knowledge on the pathophysiology associated with the functional alteration of membrane proteins and provide a basis for the development of therapeutic leads. Recent developments in chemical, biochemical, and biophysical techniques have advanced the structural and mechanistic understanding of complex membrane protein−ligand interactions. However, improvements in membrane protein expression and purification and methods of membrane protein reconstitution are still required to increase the throughput of ligand-membrane protein binding assays. 4.4.1. Membrane-Based SPR and DPI of G ProteinCoupled Receptors. 4.4.1.1. Functional Reconstitution of Membrane-Bound Receptors for Ligand Screening. G protein-coupled receptors (GPCR) play an important role in signaling cascades, where they participate in the transduction of extracellular signals to intracellular heterotrimeric guanyl nucleotide binding proteins (G proteins). GPCRs are a large superfamily with a structure comprising an extracellular amino terminus, seven transmembrane-spanning α-helices connected by alternating extracellular and intracellular loops, and a cytoplasmic carboxyl terminal region.791 The arrangement of the seven transmembrane-spanning helices and the extracellular domains provides a specific binding site for ligands, which induces a conformational change in the receptor that exposes intracellular regions, which recruit and activate G proteins. As the structures alone do not directly provide insights into the modulation of receptor signaling by allosteric ligands such as small molecules, lipids, and ions, receptors bind specific ligands that initiate signaling cascades and the ability to study these

interactions with purified receptor in a membrane environment is essential for unlocking the molecular details of these signaling events. Prior to the release of the commercially available chips, the preparation of high-quality, lipid-membrane surfaces suitable for SPR-based biophysical studies of peptide and protein−membrane interactions was a significant technical challenge. With the improvement in methods of membrane protein expression, solubilization, purification, and reconstitution into lipid bilayers, commercial biosensors are now being integrated with various receptor-containing membrane preparations45,48,50,792−794 for kinetic analysis of ligand−receptor interaction. The application of label-free optical methods in monitoring a receptor-membrane interaction was pioneered by examining the light-activated receptor rhodopsin.51,264,265,795 SLBs incorporating bovine rhodopsin were formed on a silver film via adsorption of rhodopsin-reconstituted eggPC liposomes,51 which was supported by a Teflon spacer over a silver film. It was found that transducin bound to the membraneincorporated rhodopsin before and after photolysis, which could be monitored by following the SPR spectral changes. This SPR analysis also explored conformational changes and GTP binding upon the light activation by estimating the average thickness of the protein/lipid film. After photolysis, the average thickness of the protein/lipid film increased, which could be related to the formation of metarhodopsin II. In addition, a further change could be seen in the SPR spectra upon addition of GTP resulting from the subsequent GTP/ GDP exchange. The activity of rhodopsin reconstituted into the membrane has also been investigated on a micropatterned membrane array 5458

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

functional activity of receptors maintained throughout the assembly process (Figure 21B3). Interestingly, both receptors can be captured in a functional state even without the introduction of lipids. Curve fitting using 1:1 binding model yielded a dissociation constant (KD) of 160 ± 3 nM for the SDF-1α binding to the CXCR4 receptor in the lipid bilayer on the L1 chip and 156 ± 2 nM for the CXCR4 captured on the L1 chip without lipids, which were also close to a value of ∼200 nM obtained from a cell-based assay. Although the receptors retained their structural integrity and functionality, only 14− 18% of the captured CXCR4 on various surfaces were active and accessible to SDF-1α binding. In addition to the capture and reconstitution of GPCRs onto the sensor surface, CCR5proteoliposomes enriched from immunoprecipitation of cell membrane vesicles could also be captured onto a 1D4 antibody-modified surface with sequential layer-by-layer modifications of streptavidin on the chip, followed by loading biotinylated antibody and final capturing of the receptorspecific 1D4 antibody.800 The capture and reconstitution methods were used to screen a series of membrane receptor solubilization conditions to identify the optimal detergent/lipid/buffer combination for improved activity and stability of CXCR4 and CCR5 on the sensor surfaces. The combination of detergents CHS/DDM/ CHAPS with lipids DOPC/DOPS (7:3) was found to exhibit the best initial receptor activity for both CXCR4 and CCR5 tested with conformational mAbs, 2D7, and 12G5.801 This solubilization condition obtained from the SPR analysis was then used to further develop the affinity chromatographic purification of CCR5 for CD4-dependent binding of HIV gp120 and subsequent screening against cocrystallization conditions.801,802 Further screening of 96 detergent conditions for solubilization of CCR5 with both a serial processing SPR system and an SPR imaging array platform showed that detergents containing maltosides with C9 to C13 alkyl chains were superior in maintaining the structural integrity of CCR5 as determined by the binding response of the conformational mAb 2D7.803 The binding response obtained in SPR analysis is proportional to the molecular weight of analytes. Thus, detection of ligands with a molecular weight less than 500 Da is considerably more challenging than detection of larger proteins or peptides interacting with captured membrane proteins. Furthermore, establishing a stable baseline for the enhancement of detecting small molecules is also critical for membrane protein studies. The dissociation of detergent and lipid loosely associated with the proteins from the chip surface may also change the baseline response upon changes in composition and flow of the buffer. Further development of the methods based on the optimized solubilization, capture, and reconstitution conditions have greatly improved the SPR screening of smallmolecule interactions with GPCRs. As initially demonstrated by the inhibition of gp120 binding to CCR5 by a small-molecule inhibitor, TAK-779, 19 small-molecule inhibitors of average MW 55 Da were screened against CCR5 showing that the binding affinities (KD) obtained from the SPR assay are in good agreement with inhibition constants (Ki) obtained from the whole-cell-based assay.804 Over the years, these approaches have greatly facilitated the improvement in the detection limits of ligand−receptor interaction from large antibodies to peptides and small molecules of less than 500 Da. The increase in the throughput of ligand−receptor interactions analyzed with SPR technologies

by a spatially and time-resolved SPR and SPRi measurements.264,265 The supported lipid membrane initially developed by Vogel et al.796 was prepared on a glass plate on which a thin gold film was first evaporated followed by the self-assembly of a layer of hydrophobic alkanethiol molecules that form covalent sulfur−gold bonds. These lipid bilayer systems possessed the flexibility, space, and buffer compartments required to keep incorporated membrane-spanning proteins in a functionally active form. Changes in the SPR signals were also monitored for the association of G-proteins to rhodopsin upon lightinduced photoisomerization of receptor-bound 11 cis to alltrans-retinal activation and dissociation of G-proteins upon GTP binding. These studies successfully established label-free optical methods for the study of ligand-membrane protein interactions with high sensitivity. The methods for membrane protein incorporation into the membrane were later modified by using a rapid-flow mediated on-surface immobilization and reconstitution on an L1 chip. The purified detergent-solubilized rhodopsin was then immobilized onto the surface via aminecoupling, and the subsequent removal of detergent by buffer allowed the formation of rhodopsin-incorporated membrane.795 The biological activity of the rhodopsin was maintained upon immobilization, and these studies demonstrated potentially very useful methods for the preparation of receptor-based biosensors. SPR assays have also been developed for the validation of chemokine receptors CCR5 and CXCR4 as drug targets. For example, the binding of the HIV envelope protein to the chemokine receptors CCR5 and CXCR4 which were embedded into retrovirus particles has been demonstrated.797 The virus particles containing the receptors were immobilized through amine coupling of the receptors to a C1 or F1 chip, and the functionality of the receptors was demonstrated via binding of specific antibodies and inhibition of binding by peptide inhibitors. This study demonstrated that a wide range of membrane proteins could be immobilized onto a biosensor surface using this viral carrier approach. In a separate study, SPR analysis was established to study the significance of Cterminal receptor phosphorylation for high-affinity arrestin binding to the receptor.798 By coupling the peptides derived from the C-terminus and cytoplasmic loop of CCR5 to CM5 and Sa5-streptoavidin chips, respectively, SPR analysis showed that β-arrestin 1 binds to phosphorylated and nonphosphorylated C-terminal CCR5 peptides with similar affinities, suggesting an additional binding site in receptors for the recruitment of β-arrestin upon chemokine activation. Furthermore, from the SPR binding assay, a conserved Asp-ArgTyr motif within the second intracellular loop of CCR5 was identified as an additional binding site for β-arrestin involved in G-protein activation. Methods for the capture and reconstitution of GPCRs in a lipid bilayer onto a sensor surface from crude detergentsolubilized membrane proteins have also been established for the SPR analysis of chemokine and antibodies binding to CCR5 and CXCR4,799 as shown schematically in Figure 21A. Both receptors were captured and assembled into bilayers at a high density required for the enhanced detection sensitivity of ligand binding. The ability of conformation-dependent mAbs to bind to the captured receptor demonstrated that the structural integrity of CCR5 and CXCR4 was maintained after the capture and reconstitution process (Figure 21B2). In addition, the binding of a natural 7.8 kDa chemokine, stromal cellderived factor 1a (SDF-1a) to CXCR4 demonstrated the 5459

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

firmation of the functionality of the inserted protein, by monitoring ligand-induced conformational changes that reflect the activity of the reconstituted receptors. In one example, it was found that binding of agonist to the human δ-opioid receptor caused an increase in the thickness and molecular packing density of the membrane, while antagonist binding did not show this effect.823,824 The results suggested differences in the degree of transmembrane helical reorientation upon ligand binding and provided direct evidence for the differential effects of agonists and antagonists upon membrane structure. These studies clearly provide important information on the structural changes which occur in both the membrane and the receptor during the early stages of a signaling cascade. The role of lipid composition and membrane properties on the function of GPCRs have been demonstrated for rhodopsin where phosphatidylethanolamine (PE) was reported to promote the metharhodopsin MI-MII transition and even transducing binding.814,825,826 In addition, the membrane domains rich in cholesterol and sphingomyelin (lipid rafts) have been shown to modulate the activity of CCR5 StaR and A2AR in G-protein recruitment and enhancing ligand binding affinities in a new interaction mode beyond pure allosteric modulation.822,827 4.4.1.3. Intracellular Membrane-Tethering of Receptors. The seven transmembrane-spanning helices have been wellcharacterized as the signature motif of GPCRs. An additional highly conserved feature of the prototypic seven-transmembrane spanning arrangement of GPCRs is a helix (termed Helix 8) positioned parallel to the cytoplasmic side of the lipid bilayer, which comprises the first 15−20 amino acids of the intracellular C-terminus. This region of the GPCR is a focal point for many protein−protein interactions that are crucial for receptor coupling to G proteins and other signaling/regulatory molecules.828,829 The angiotensin type 1 receptor (AT1R) is a 359 amino acid GPCR that mediates the important cardiovascular and homeostatic actions of the peptide hormone, angiotensin II (AngII).828,830 It couples primarily to the heterotrimeric G protein, Gq/11, to activate phospholipase Cα (PLCα), which hydrolyses phosphatidylinositol-(4,5)bisphosphate [PI(4,5)P2] to generate the soluble second messengers, diacylglycerol (DAG) and inositol (1,4,5) trisphosphate (IP3).831 These messengers in turn activate protein kinase C and raise intracellular calcium, respectively, thereby promoting cellular responses that are the basis of AngII actions (principally, vasoconstriction, aldosterone release, thirst, and salt appetite).830 Both SPR and DPI have been used to examine the role of anionic lipids and phosphatidylinositides in the membrane tethering of Helix 8 in the AT1A receptor. We used synthetic peptides and model membranes to show that Helix 8 binds with high affinity to phospholipid bilayers via both electrostatic and hydrophobic interactions.601,829,832,833 In particular, the binding characteristics of an AT peptide derived from the proximal region of the AT1R C-terminus (residues 305−325) and its LGJ analogue [substitutions of four basic (Lys) residues putatively involved in molecular recognition] to DMPC and DMPG model membranes on an L1 chip have been analyzed by SPR for the effect of different membrane surface charge.832 The resulting binding responses showed that the AT peptide bound with high affinity to the negatively charged DMPG but poorly to the zwitterionic DMPC bilayer. In contrast, the LGJ analogue displayed poor association with both lipids, indicating that both basic residues in the peptide and the anionic lipid component is crucial to the Helix 8membrane interaction. CD analysis of the peptides in DMPG

has also underpinned the development of robust assays for screening a library of small molecules and fragments.793,805,806 Various methods developed for postfunctionalization of liposomes and immobilization to the solid substrates have also been applied to the immobilization of functional receptors in membranes onto the sensor surface which have been reviewed in the context of GPCRs.396,792 These methods can be applied to any membrane protein and can be categorized as producing either randomly or uniformly oriented protein on the sensor surface. Random orientation may be useful when receptor-containing vesicles are immobilized via a tag on the protein since it guarantees that there are always receptor molecules at the surface with their N-terminal- or C-terminal side available for interaction with analytes. In contrast, a uniform orientation of receptor molecules allows optimum access to the target binding site, either extracellular ligand binding or intracellular signaling molecules eg G protein interactions. By capturing the membrane proteins using different bioaffinity approaches, ligand binding and fragment screening with SPR have also been established for other membrane proteins, including the neurotensin receptor-1 (NTR-1),794,807,808 the human olfactory receptor 17−4 (hOR17−4),809 the neuropeptide Y4 receptor N-terminal domain,810 the adenosine-A2A receptor,811,812 and the β1adrenergic receptor (β1AR).813 The incorporation of GPCRs in a membrane environment on the SPR sensor surface has also allowed the study and screening of allosteric GPCR ligands. For the discovery of allosteric compounds, CCR5 was captured and reconstituted on the sensor surface and the binding of two hundred compounds with an average MW of 362 Da was then screened against both the active and inactive form of the receptor.805 Five real binders were identified for the active form of CCR5, while these binders also showed binding response to inactive CCR5 at higher concentrations. Allosteric modulators that bind to distinct allosteric sites in the receptors were also assayed by screening with pyrazinyl sulphonamide compounds of allosteric CCR4 antagonists which are also weak CCR5 binders in the screening library. These compounds bound the inactive CCR5 with 2-fold higher Rmax than with the active form.805 4.4.1.2. Effect of Membrane Properties on the Activity of GPCRs. Although it has been well-studied that the structures of membrane proteins and characterization of ligand binding mechanisms are crucial for the understanding of their activity and provide further insights into the molecular action of ligands and proteins, the influence of the lipid environment on the function of membrane proteins is less well-established for membranes with more complex molecular compositions and physical properties.814−816 A variant of SPR referred to as coupled plasmon-waveguide or plasmon waveguide resonance (PWR) spectroscopy51,817−821 has been developed to monitor the GPCR reconstitution and ligand binding in a single assay. When applied to membrane-bound receptors for example, this technique allows changes in the structure of the receptor, both parallel and perpendicular to the lipid membrane, to be examined in response to the binding of ligand. Additionally, PWR simultaneously measures mass and anisotropy changes in an oriented system such as lipid bilayers and therefore allows confirmation of the proper orientation of GPCRs in the membrane and membrane anisotropic structures.168 Detailed kinetic analysis with respective rate constants have also provided mechanistic information about the reconstitution process.822 Moreover, PWR allows for an immediate con5460

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

analysis of protein-membrane interactions. In contrast to a planar surface, 3D nanostructures can be engineered to better mimic membrane morphology in a cellular environment. The formation of SLBs or tethered liposomes and the molecular interaction with a membrane on nanostructures can be probed by LSPR, which is sensitive to changes in the RI within approximately 20 nm field length of the metal surface. A model membrane platform has been described by covering individual Au nanorods randomly immobilized on a glass substrate with fluid membrane doped with biotin moieties. The formation the SLBs on the Au nanorods of 56 nm in length and 26 nm in width via vesicle spreading in microcapillaries was probed by a transmission dark-field microscope coupled to an imaging spectrometer.493,839 Up to 20 individual separated nanorods can be simultaneously analyzed from captured darkfield images to generate the scattering spectra. An average red shift of 3.6 nm in the scattering spectra maximum was observed for spreading the membrane to form a corrugated SLB on the surfaces of nanorod-glass substrates. The red shift in the scattering spectral maximum can vary between 3.5 and 5.4 nm depending on the lipid composition and incubation time. The binding of streptavidin to the biotin-functionalized SLBs caused an additional red-shift of an average of 2.9 nm. Due to the short evanescent field on the plasmonic nanoparticles, the use of highly sensitive LSPR to probe the molecule distance from the metal surface was further demonstrated by a smaller spectral shift with narrower dynamic range for the binding of streptavidin to biotin-membrane with a C6 spacer compared that without a spacer.493 This demonstrated the wellcharacterized decay in plasmon sensitivity with distance from the metallic nanoparticle surface. As molecular binding to a membrane on each particle can be analyzed separately from LSPR scattering spectra, the role of lateral compositional and topological heterogeneity of a cellular membrane in mediating the protein binding can potentially be tracked by employing specific probes or proteins on membrane surfaces. An LSPRbased sensor consisting of an alkylthiol SAM-modified Agnanocube randomly deposited onto a glass substrate was developed for characterizing protein-membrane interactions with less influence from solutes in the bulk solution.840 The vesicle-fusion over the alkylthiol SAM-modified Ag-nanocubes formed a hybrid lipid bilayer, while the rupture of vesicles formed lipid bilayers over the bare glass substrates.840 The membrane formation and subsequent protein binding was monitored by the extinction spectral shift of the peak maximum detected by a general spectrophotometer. At the quadripolar LSPR wavelength, the EM field exhibits localized hot spots of amplified intensity with the field being strongest along the edges and corners of the Ag nanocubes. This effect was seen as a large peak at the λmax that dominated the transmission spectrum. The formation of hybrid lipid bilayers (HLBs) over the alkanethiol-modified Ag region caused a 2.4 nm red shift in the quadripolar peak, and the full coverage of bilayers on the substrates was characterized by the absence of significant spectral shift for the addition of BSA to the SLBs. In contrast, a red-shift of 1.26 nm was seen for the addition of neutravidin to the biotinylated HBMs/SLBs. This membrane-nanocube hybrid structure was further applied to the kinetic analysis of the binding of His-tagged yellow fluorescence proteins to the NTA-Ni-modified membrane.840 Metal-nanodiscs deposited on glass with hole-mask colloidal lithography provided a nanoarray for the formation of SLBs and the subsequent binding of CTBx to GM1 characterized by the LSPR shift.841 The SLB formation

and DMPC liposome solutions revealed that AT adopted a helical structure and was partially inserted into the membrane, while LGJ adopted a mixture of sheet and helix structure. These findings suggest that in intact AT1A receptors, the proximal carboxyl-terminus associates with the cytoplasmic face of the cell membrane via a high affinity, anionic phospholipid-specific tethering that serves to increase the amphipathic helicity of this region.832 Anionic phospholipids, including PIPs, interact electrostatically with polybasic proteins,7,834−838 and the Helix 8 region of GPCRs, with its multiple basic residues, is a likely candidate for this interaction. Several studies indicate that anionic phospholipids, including PS and PIPs, may cluster to form microdomains that are significant for receptor function. Birefringence analysis by DPI was therefore performed to characterize the state of the bilayer upon Helix-8 binding to various membranes, and the results were analyzed with multiple-state kinetic modeling techniques to elucidate the intermediate states of peptide binding.601,653 Results demonstrated that Helix 8 discriminated between PIPs and different charges on lipid membranes. Peptide binding to PI(4)P-containing bilayers caused a dramatic change in the birefringence (a measure of membrane order) of the bilayer. Kinetic modeling showed that PI(4)P is held above the bilayer until the mass of bound peptide reaches a threshold, after which the peptides insert further into the bilayer. This suggests that Helix 8 can respond to the presence of PI(4)P by withdrawing from the bilayer, resulting in a functional conformational change in the receptor. Significantly, the observations of the striking difference between the mechanism of binding to DMPC/DMPS and to DMPC/ DMPS/PI(4)P was only possible due to the ability of DPI to measure changes in the membrane structure. In the absence of the birefringence data, mass-only measurements (e.g., with SPR) only reveal increased binding to PI(4)P and the effect of PIPs on membrane structure changes that control Helix 8 orientation were invisible. On the basis of the results of these models, it was concluded that the two-state model, even without bilayer expansion, was sufficient to represent the simpler DMPC and DMPC/DMPS bilayers. However, for membranes containing PIPs, the mass threshold improved the model fit substantially when used with the three-state model with bilayer expansion, and the effect was especially marked for DMPC/DMPS/PI(4)P but also to a lesser extent for DMPC/ DMPS/PI(4,5)P2. Overall, it was concluded that the interaction between Helix 8 and PI(4)P, whereby the PI(4)P molecules pull the AT1RHelix 8 peptide outward from the bilayer, prevents insertion and bilayer disruption until the amount of AT1R-Helix 8 peptide on the surface reaches a critical density, at which point it reinserts. This hypothesis was supported by kinetic modeling of the mass and birefringence changes, using a modified version of our original kinetic modeling technique for the DPI. This biophysical study of the binding mechanism therefore revealed the influence of lipids on AT1R-Helix 8 membrane binding and insertion and provided new insight into the process by which bilayer signals are converted to changes in Helix 8 and hence possible conformational changes in the protein. 4.5. Binding of Molecules to Membranes on Nanostructured Substrates

Membrane-coated plasmonic metallic nanoparticles either attached onto solid substrates or suspended in solution have also been developed as a sensing platform for high-throughput 5461

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

is facilitated by coating the Au nanodiscs with a thin silica film. As the LSPR is limited with short evanescence field above the metal surface, controlling the thickness of the silica film is important in keeping the sensing sensitivity of LSPR. This can be finely tuned by the sol concentration and the sol−gel methods such as dip- or spin-coating producing thicknesses from micrometer to less than 10 nm.842,843 The fusion of 1% GM1-eggPC SUVs to form SLBs on the silica thin films cause an LSPR shift of 8.4 nm for the 0.25% sol-coated Au nanodiscs. This LSPR peak shift is slightly higher than 7.7 nm obtained for silica-sputtered SiO2 due to the thicker sputter silica film of 10 nm.841 The interaction of 200 nM pentameric CTB to 1% GM1-doped eggPC bilayers on 0.25% sol and sputtered silicacoated Au nanodiscs caused an LSPR peak shift of 2.2 and 1.3 nm, respectively. These results demonstrated that the effects of thickness and coverage of silica film on the Au nanodisc on the properties of the SLBs and can affect the characteristics of protein−membrane interactions. In contrast to the LSPR analysis of SLBs formation on nanoparticles immobilized substrates, a solution-phase nanoplasmonic sensor consisting of SLBs-coated plasmonic nanoparticles suspended in solution has been developed for protein−membrane binding using a standard transmission UV−vis spectrophotometer. The SLBs were coated on the surface of SiO2-coated Ag nanocubes upon mixing the nanocubes in liposome solution.495,496 The application of this membrane-coated Ag-nanocube to measuring protein-membrane interactions was demonstrated by the different extents of LSPR spectral shift measured at varying surface density of streptavidin binding to biotin-lipids.495 Furthermore, a correlation of the LSPR shift and protein mass density of 0.191 ng/mm2 nm was established for the binding of CTX-b to the membrane-associated receptor GM1.495,496 The ability of the membrane-coated nanocube sensor in the kinetic analysis of protein-membrane interactions was demonstrated by a higher affinity binding of Ste5, a mitogen-activated protein kinase (MAPK) scaffold protein, to a PI(4,5)P2-bilayer than to a control PI(4,5)P2-free bilayer. These membrane-coated nanocubes have been further developed into a high-throughput lipid bilayer array in microwell format.495 The shifts of the quadrupole LSPR scattering peak were measured to determine the binding affinities, maximum binding capacity, and binding cooperativity of CTX-b to various gangliosides, including GM1, GM2, and fucosyl-GM1.496 Metallic nanohole arrays provide unique options for integrating plasmonic sensing with tethered liposomes or SLBs and freely suspended lipid membrane across the nanohole.149,631,844−847 The development of model membrane systems on nanoplasmonic substrates requires systemic characterization of the dependence of membrane morphology on the geometric and material properties of the substrates. With few examples of liposome-based sensing platform, the predominant SLBs-over nanohole systems are integrated with LSPR or EOT sensing modes for characterizing the bilayer properties and the molecule-membrane and membrane proteins binding processes.149,498,631,844−846,848−850 The metal films perforated with nanoholes exhibit LSPR modes confined to the nanoholes. The randomly arranged nanoholes do not exhibit the long-range higher-order diffractive coupling effects149,846 that have been observed in highly ordered, periodic nanohole arrays.631,844,848 Both formation of SLBs patches and DNA-tethered SUVs confined in nanometric plasmon active apertures in Au or Ag films of 15−55 nm on

transparent SiO2 substrates have been characterized by LSPR and further applied to investigate kinetic mechanisms of protein−membrane interaction. For the formation of the SLBS within the nanoholes in milled Au film on glass, the Au surface was passivated with biotin-BSA to block liposome adsorption and nonspecific molecule binding.149 SUVs were then directed into the hole and adsorbed onto the exposed SiO2 at the bottom of the nanohole. Due to the restricted small area of exposed SiO2, the critical surface area coverage for vesicle−vesicle contact required for vesicle-bilayer transformation is not attainable in the nanohole with only one or few liposomes inside. Vesicle rupture and fusion in the nanohole array has been induced with the addition of Ca2+ to promote the bilayer formation. By monitoring the temporal changes in the extinction measured at 725 nm of the LSPR peak, the saturated changes for NeutrAvidin binding to both biotinBSA on Au and biotin-SLBs on SiO2 was more than a factor of 3 higher than NeutrAvidin binding only to biotinBSA on Au with biotin-free SLBs on SiO2. The membrane-specific binding was further demonstrated by the binding of CTX to GM1-modified SLBs and DNA hybridization to the membrane modified with complementary single strand DNA. The fitting of the kinetics to a single exponential model showed that there were sufficient number of molecules and holes to generate signals proportional to the number of bound molecules. Vesicles have also been introduced into the nanohole via various tethering groups. In one example, lipid vesicles of 100 nm functionalized with cholesterol-tagged DNA were tethered to a complementary ssDNA-modified SLBs passivated on the exposed SiO2 surface at the bottom of nanoholes.845 A red shift of 0.3 nm in the LSPR peak was observed for tethering the complementary ssDNA modified liposomes while no shift in LSPR peak was observed for liposomes modified with noncomplementary ssDNA. Thus, the nanohole arrays can be engineered with biomembranes of different morphology as either in planar bilayers or vesicles of uniform size. Liposomes attached to the tip of Au nanomenhirs within a silicon nitride matrix were integrated with a dual mode plasmonic sensing array.851 Due to the highest sensitivity localized at the edges and corners of asymmetric Au nanomenhirs, multiple plasmon modes with enhanced EM field localized at different regions enable the differentiation of liposomes attached to the apex from the streptavidin binding to the base. By monitoring the peak shift at different wavelengths, the selective localized liposome binding was apparent as a large red-shift at the tip while a small blue-shift was measured at the base. While red-shifts ranging from 0.5 to 4 nm were observed for vesicle adsorption on various metallic nanostructured substrates, the magnitudes of the shifts do not provide the morphological differentiation between SLBs and vesicle on the nanoarray. This differentiation of SLBs from vesicles formed on the plasmonic nanoarray has however been extensively characterized by the kinetic analysis of LSPR extinction maximum wavelength shift and in combination of QCM for liposome adsorption and SLBs formation.128,209,846,852 The kinetic steps of SLB formation via liposome adsorption on the nanoplasmonic substrates vary with the material properties of the substrate, lipid composition, lyotropic crystalline phases, liposome size and concentration, and other environmental conditions, including temperature, osmotic pressure, ionic strength and type, and concentration of divalent cations.209,565,853−859 5462

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

addition of PEG8000. The main advantage of these tethering SLB systems was demonstrated by the multicycle of bilayer formation and surface regeneration with 0.5−1% Triton X100 for consistent bilayer properties, which is critical for multiple, reproducible analysis on a single substrate. The multiplexed protein-membrane binding was further demonstrated by forming different tSLBs, consisting of GM1 or PI(4)P on separated parallel channels. The cholera toxin (CTX) and PIPMultiGrip protein were bound selectively to the GM1 and PI(4)P tSLBs, respectively, which a detection limit of 260 nM was obtained for CTX binding to the GM1. On-demand formation of a SLB microarray has been demonstrated with the use of trehalose-vitrified liposomes on a thin SiO2 (10 nm)coated Au. The lipid vesicles suspended in trehalose solution were spotted and dried on the patterned SiO2/Au array. SPR differential images obtained by subtracting images collected with p-polarization from those imaged with s-polarization were taken before, during, and after hydration in the microfluidic channels. Significant shape deformation was observed from the SPR image during the vitrification of trehalose and removal of water from the environment, and this shape deformation is critical in determining the minimum distance of adjacent spots for eliminating cross reaction, which allows for the highthroughput screening of multiple SLBs with varying components. Upon rehydration, high contrast images between each row with sharp boundaries are indicative of bilayer formation in orderly rows without migration into adjacent spots. The equilibrium binding of CTX to the GM1-SLBs exhibited similar response signals for both bilayers formed with and without desiccation-rehydration processes over a range of CTX concentration. Such similar responses suggested that the GM1 receptor in the preserved vesicles retain biological activity and respective ligand affinities throughout the vitrification, rehydration, and fusion processes. In contrast, a low affinity of CTX toward GM2 and GM3 was evident from the signals that remained at baseline values. The formation of SLBs composed of eggPC with varying amount of the ganglioside GM1 via vesicle fusion onto a SiO2coated-Ag nanohole array has been characterized by multiple channel SPR imaging spectroscopy.631 In contrast to the prismcoupled SPR imaging systems in which the real-time intensitybased 2D images only recorded the reflective intensity at a fixed wavelength or angle, the EOT generated through the metallic nanohole array provides full angular or spectral changes of highresolution SPR imaging in real time. In this nanohole SPRi system, the sensing region of the Ag nanohole array was partitioned with a PDMS microfluidic chip into 50 parallel microfluidic channels which isolate SLBs of different components and receptors. In addition, SPR imaging analysis with parallel microfluidic channels also prevents the imaging artifacts or errors generated from cross-contamination, position-dependent depletion of analyte, and position-dependent binding rates due to mass transport. The full spectra extracted simultaneously from 50 channels were highly reproducible with multiple transmission maxima at different wavelengths characteristic of the dielectric-metal interfacial properties. A 2 nm red-shift in the transmission spectra was observed from the initial (1,0) maximum at wavelength of 723.9 nm for a stable formation of SLBs on the nanohole array. The binding kinetics of cholera toxin b subunit to GM1 receptors were measured in a single analysis. The maximum spectral shifts for injecting 50 nM CTX-b increased proportionally with the amount of GM1 receptor up 2 mol %

Membranes formed on nanohole arrays made in freestanding metallic films are advantageous over the SLBs that are in direct contact with the substrates. The free-standing lipid bilayer over nanoporous metallic films or nanoblack lipid membrane (nanoBLMs) in a periodic ordered array retains bending stability and plasticity upon mechanical stress, and both bilayer leaflets are exposed to solution and accessible to molecules.487,860 The electrochemical properties of the membrane and ion channel have been explored in this nanopore-spanning bilayer.861−863 In addition to measuring the changes of electric current across the bilayer, the kinetics of protein binding to the membrane and transmembrane receptors can be characterized by the plasmon-mediated EOT effect.844,848 The LSPR effect for randomly distributed nanopores with short-range order generates broad extinction spectra which are weaker than the EOT effect through arrays of periodic nanopores with a long-range order.130 Periodic arrays of pore-spanning lipid membranes fabricated over nanopores of Au and Si3N4 layers with a pore diameter of 200 nm using focused ion beams milling was characterized by the transmission spectra for kinetic sensing of protein−membrane interactions.129,848 To facilitate the SLBs formation via vesiclebilayer transformation, a thin SiO2 layer of 20 nm was deposited onto the surface of the Au nanopore array by atomic layer deposition. The formation of the SLBs can be monitored either by acquiring the high-resolution CCD images or by recording the transmission spectra with a fiber optic spectrometer.631 Multiple transmission maxima and minima were observed in the transmission spectra in EOT mode. The positions of these spectral maxima and minima vary with the periodicity and pore diameter, and the RI changes induced by the SLBs formation either over the nanopore region or on the SiO2−Au region cause a 1−2 nm shift of the transmission peaks. Similar lipid diffusivities of about 1.9 μm2/s were obtained for SLBs over nanopores and supported on flat SiO2− Au, suggesting that the continuity and fluidity of the membrane were not affected by the presence of nanopores. Importantly, the lipid molecules diffuse onto the periodic nanopore arrays in a similar way to the lipids of SLBs on the silica surface. The kinetic binding of streptavidin to the biotinylated SLBs on the nanopore array showed either an increasing transmission intensity for a periodicity of 390 nm or a decreasing transmission intensity for a periodicity of 440 nm due to the slope of the transmission resonance peaks at the measuring wavelength.864−866 The application of nanopore arrays to functional assays of transmembrane proteins was demonstrated by the detection of antibodies binding to α-hemolysin incorporated into the membrane over the nanopores. Due to the difficulty in maintaining constant hydration for lipid bilayers during the arraying, handling, and assay procedures, analyzing SLBs micro/nanoarrays with varying bilayer composition remains a challenging task. The tSLBs formed on a biotin-BSA passivated Au surface on a microfluidic cross-patterned array were examined by SPRi,634 and the change in reflectivity for SLB formation and subsequent binding was monitored at a fixed imaging angle. SPR images were analyzed by differential subtraction of a prebinding image from a postbinding image. In addition, images acquired from spolarization were used as a reference for background verification to minimize surface artifacts. Liposomes mimicking mammalian mitochondria compositions mixed with biotinylated DPPE were bound to biotin-BSA layer via NeutrAvidin. The fusion of the bound vesicles into SLBs was induced by the 5463

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Table 5. Comparison of the Key Features of SPR, LSPR, EOT, and DPI as Biomembrane Platforms SPR modulations

sensing field limit of detection

resolution refractive index sensitivity throughput/multiplexing miniaturization spatial imaging enhanced spectroscopy sensitive to temperature variation temperature control for membrane formation and binding surface modification for biomembranes extent of response for SLB formation substrate geometry kinetic models for membrane binding and changes unique features as a biomembrane platform

angular intensity wavelength phase 100−300 nm 10 nM (IgG) 1−5 pg/mm2 1 × 10−7 ∼ 5 × 10−6 RIU ∼2 × 106 nm/RIU low-medium/yes no yes Raman

LSPR

EOT

adsorption scattering Brewster angle

transmission surface SPs localized SPs Brewster angle 30−150 nm 5 nM IgG (SPR) 100 nM IgG (transmission)

10−50 nm 1 nM (IgG)

DPI interferometric TM and TE phases 100−200 nm 0.1 pg/mm2 (mass) 0.01 nm (thickness) 0.0004 g/cm3 (density) 1 × 10−7 RIU

1 × 10−3 RIU

1 × 10−3 ∼ 10−6 RIU

Yes

∼200 nm/RIU high/yes yes yes Raman fluorescence no

50−5000 nm/RIU high/yes yes yes Raman fluorescence no

yes

yes

no

no

yes

dielectric layer (e.g., SiO2) SAM, polymer ∼5000 RU

dielectric layer (e.g., SiO2)

dielectric layer (e.g., SiO2)

not required, (except for tethered liposomes)

SAM, polymer ∼3 nm

SAM, polymer ∼3 nm

∼14 radians (TM)

planar Langmuir

planar/curvature N.A.

planar/hole N.A.

2 state model widely used for membrane binding analysis

measure deformation of liposomes on the surface

suspended SLBs for incorporating transmembrane receptors complex spectral maxima and minima for binding analysis

SLB-coated nanoparticles for binding in solution reduced sensitivity from surface modification of thicker layer

1 × 105 radians/RIU low/no no no N.A.

∼11 radians (TE) planar 2 and 3-state model with/without bilayer expansion model accurate bilayer thickness measurement

lipid molecule packing order versus mass for complex membrane binding mechanisms

significant new insight into the subtleties of membrane binding in terms of relative binding affinity between closely related compounds and new information on the kinetics of membrane binding. These important biophysical data were not previously accessible and has now transformed the field of biomembrane biology and biophysics. Significantly, as in every field, answers to long-term important questions open the door to an ever expanding landscape of novel concepts and questions. Consideration of the topics covered in this review from the design of optical biosensors, to the design of model membrane systems and their deposition onto a myriad of chip surfaces either through physisorption or chemical immobilization, clearly highlights the enormous technological and chemical toolkit at our disposal to study membrane-mediated processes. While this capability has been exploited to some extent in the design of model membrane systems, one could argue that the analytical potential of these systems has not been realized to nearly the same extent in applying these tools to the study of peptide/ protein−membrane interactions. SPR techniques are widely used to measure the binding affinity of an enormous range of biomolecular interactions. The majority of commercially available instruments are limited to mass-only measurements which are then transformed into equilibrium-based derivation of Kds. If the experiments are

concentration in SLBs. The wavelength shift in the transmission spectra showed fast association and slow dissociation for CTX-b binding to GM1. Although the dissociation constant varied from sub-nM to 2 nM (determined based on a 1:1 kinetic interaction model), the obtained values are comparable to those obtained by prism-coupled SPR analysis and fluorescence methods which demonstrated strong binding of CTX-b to GM-1.385,867−869 Overall, a wide range of biomembrane interaction studies have been reported using optical biosensors and a comparative summary of the key features and advantages of each of the four optical biosensor platforms is provided below (Table 5).

5. CONCLUSIONS Characterization of biomolecular-membrane interactions is required to elucidate complex cellular processes. Following the introduction of commercially available SPR instruments with membrane chips, the confirmation that membrane binding was reversible and could be treated in equilibrium terms opened up an entirely new era in the study of membrane interactions. Suddenly, it was possible to investigate the mechanism of action of a large group of peptides and proteins whose biological function was known to be linked to the interaction with a cell or organellar membrane. In particular, the ability to quantitatively measure binding has provided 5464

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

have enormous technical expertise in plasmonic systems and membrane biophysicists with the theoretical and experimental insight into membrane structure and function would allow new technologies that address the next level of mechanistic questions to be developed and allow us to answer questions about membrane structure and its role in biological mechanisms. In summary, the foundation has now been established for future development, which can underpin multiparameter characterization of the molecular interaction associated with membranes for a broad range of biomolecules and their binding mechanisms associated with changes in the membrane structures. These advances will lead to unprecedented mechanistic understanding of membrane-mediated systems and facilitate the development of more sophisticated biosensor platforms and ultimately more effective approaches to the treatment of diseases associated with membrane interactions.

appropriately performed, reliable kinetic analysis can also be performed to further delineate the binding mechanism between two interacting partners. However, even in nonmembranebased systems, it is widely acknowledged that biomolecular interactions are accompanied by conformational changes in one or both partners, and these changes are not captured in the information obtained from mass-only biosensors. Indeed, it still remains a major challenge to fully characterize these structural changes without a combination of spectroscopic techniques probing structures both in the solid and solution phase. The same issues of structural flexibility during interactions also exist in membrane interactions. Indeed, the structural flexibility of a lipid bilayer is arguably the most important feature which allows it to function as a multifaceted gate-keeper of a cell or organelle. However, our familiarity with peptide and protein structure has lead us to make conclusions about the function of membrane-active peptides and proteins largely based on their structure with little or no consideration of the three-dimensional structure of the membrane. Nevertheless, researchers have been able to ask questions about the reversibility, specificity, and kinetics of membrane binding in real time rather than simply asking whether a molecule binds or not. The role of different membrane composition and hence one-dimensional structure can be investigated, but the results are normally interpreted in terms of the effect on the peptide/ protein. Nevertheless, the answers to these questions provides the basis for the next stage to address more complex questions about binding over time and the events after binding− specifically, what are the changes in membrane structure during and after binding, and what are the orientational changes that the peptide/protein and the membrane undergo following the initial binding event? More recently developed technologies such as DPI have allowed researchers to address these more complex questions. It is clearly evident that membrane interactions involve multiple kinetically distinct stages which can be detected by DPI and other technologies. The first important feature of these techniques which allows these observations is the ability to fully characterize the properties of the membrane on the chip surface. This is generally not possible with single wavelength and/or single light beam systems, but the accurate measurement of the thickness and/or density of the deposited membrane is an essential prerequisite to meaningful measurements of membrane structure changes during binding. The ability to also characterize membrane order in terms of birefringence in a real-time format has further expanded our capacity to pose novel questions around membrane structure changes that accompany peptide and protein binding. However, consideration of the physical basis and the operating principals of SPR instruments suggests that substantially more information relevant to membrane systems could be derived. Since technology is the driver of new questions and answers, and one could ask why is there somewhat of a disconnect between the extremely detailed engineering of the instruments plus the surface technology underpinning the bilayer deposition and the relative paucity of molecular details that can be extracted from the biomolecular interaction experiment? One possible reason is that researchers with the capacity to develop more complex technologies have less access to biological questions; as a result, first and second generation instruments are designed to address simple binding questions (i.e., to detect if a molecule binds). Thus, increased collaboration between instrument engineers and physicists who

AUTHOR INFORMATION Corresponding Author

*Address: Department of Biochemistry and Molecular Biology and Biomedicine Discovery Institute, Monash University, Wellington Rd, Clayton, VIC 3800, Australia. E-mail: Mibel. [email protected]. ORCID

Marie-Isabel Aguilar: 0000-0002-0234-4064 Notes

The authors declare no competing financial interest. Biographies Dr. Tzong-Hsien (John) Lee is an Analytical Biochemist at Monash University who is now focusing on developing membrane biosensors and exploring membrane proteomics. He completed his Ph.D. in Biochemistry at Monash University, developing the immobilized phospholipid silica particles for studying the interaction of membrane active peptides and purifying the membrane proteins. He has over 40 publications and currently is a senior research fellow in Prof. Mibel Aguilar’s group developing optical biosensors for membrane interaction studies. Dr. Daniel Hirst earned a B.Sc. (Hons) in 2010 and a Ph.D. in Biochemistry from Monash University in 2016. He worked under Prof. Aguilar as a graduate student, studying the application of dual polarization interferometry and developing kinetic models to characterise peptide-membrane interactions. He is currently a postdoctoral fellow in the laboratory of Prof. Aguilar. Dr. Ketav Kulkarni is a medicinal and peptide chemist whose recent focus is to develop synthetic strategies toward the development of novel self-assembling β-peptide-based materials. He completed his Ph.D. in chemistry at Monash University pursuing multistep synthesis of heterocyclic natural products and developed lab-on-a-chip-based devices with applications in scale-out chemical syntheses and effective sample processing for proteomics. He also has significant postdoctoral experience in synthesis of complex macrocyclic peptides as cell permeable therapeutic target. He is currently a research fellow in Prof Aguilar’s group. Dr. Mark Del Borgo received his Ph.D. in Pharmacology in 2005 from the University of Melbourne and immediately became a research fellow in the Department of Chemistry at the University of Queensland. He then took up a teaching position within the Department of Medicinal Chemistry at the University of Kansas 5465

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

(16) Vigant, F.; Santos, N. C.; Lee, B. Broad-Spectrum Antivirals against Viral Fusion. Nat. Rev. Microbiol. 2015, 13, 426−437. (17) Holthuis, J. C.; Menon, A. K. Lipid Landscapes and Pipelines in Membrane Homeostasis. Nature 2014, 510, 48−57. (18) Ryan, E.; Reid, G. E. Chemical Derivatization and Ultrahigh Resolution and Accurate Mass Spectrometry Strategies for ″Shotgun″ Lipidome Analysis. Acc. Chem. Res. 2016, 49, 1596−1604. (19) Subramaniam, S.; Fahy, E.; Gupta, S.; Sud, M.; Byrnes, R. W.; Cotter, D.; Dinasarapu, A. R.; Maurya, M. R. Bioinformatics and Systems Biology of the Lipidome. Chem. Rev. 2011, 111, 6452−6490. (20) Fahy, E.; Subramaniam, S.; Murphy, R. C.; Nishijima, M.; Raetz, C. R.; Shimizu, T.; Spener, F.; van Meer, G.; Wakelam, M. J.; Dennis, E. A. Update of the Lipid Maps Comprehensive Classification System for Lipids. J. Lipid Res. 2009, 50 (Suppl), S9−14. (21) Fahy, E.; Cotter, D.; Sud, M.; Subramaniam, S. Lipid Classification, Structures and Tools. Biochim. Biophys. Acta, Mol. Cell Biol. Lipids 2011, 1811, 637−647. (22) Yetukuri, L.; Ekroos, K.; Vidal-Puig, A.; Oresic, M. Informatics and Computational Strategies for the Study of Lipids. Mol. BioSyst. 2008, 4, 121−127. (23) Mitchell, T. W.; Pham, H.; Thomas, M. C.; Blanksby, S. J. Identification of Double Bond Position in Lipids: From Gc to Ozid. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci. 2009, 877, 2722−2735. (24) Israelachvili, J. N.; Marcelja, S.; Horn, R. G. Physical Principles of Membrane Organization. Q. Rev. Biophys. 1980, 13, 121−200. (25) Simons, K.; Ikonen, E. Functional Rafts in Cell Membranes. Nature 1997, 387, 569−572. (26) Janmey, P. A.; Kinnunen, P. K. Biophysical Properties of Lipids and Dynamic Membranes. Trends Cell Biol. 2006, 16, 538−546. (27) Christensen, S. M.; Stamou, D. G. Sensing-Applications of Surface-Based Single Vesicle Arrays. Sensors 2010, 10, 11352−11368. (28) Gozen, I.; Jesorka, A. Instrumental Methods to Characterize Molecular Phospholipid Films on Solid Supports. Anal. Chem. 2012, 84, 822−838. (29) Stone, M. B.; Shelby, S. A.; Veatch, S. L. Super-Resolution Microscopy: Shedding Light on the Cellular Plasma Membrane. Chem. Rev. 2017, 117, 7457−7477. (30) Ciminelli, C.; Campanella, C. M.; Dell’Olio, F.; Campanella, C. E.; Armenise, M. N. Label-Free Optical Resonant Sensors for Biochemical Applications. Prog. Quantum Electron. 2013, 37, 51−107. (31) Lee, T. H.; Hirst, D. J.; Aguilar, M. I. New Insights into the Molecular Mechanisms of Biomembrane Structural Changes and Interactions by Optical Biosensor Technology. Biochim. Biophys. Acta, Biomembr. 2015, 1848, 1868−1885. (32) Lee, T. H.; Mozsolits, H.; Aguilar, M. I. Measurement of the Affinity of Melittin for Zwitterionic and Anionic Membranes Using Immobilized Lipid Biosensors. J. Pept. Res. 2001, 58, 464−476. (33) Abbas, A.; Linman, M. J.; Cheng, Q. New Trends in Instrumental Design for Surface Plasmon Resonance-Based Biosensors. Biosens. Bioelectron. 2011, 26, 1815−1824. (34) Konradi, R.; Textor, M.; Reimhult, E. Using Complementary Acoustic and Optical Techniques for Quantitative Monitoring of Biomolecular Adsorption at Interfaces. Biosensors 2012, 2, 341−376. (35) Fernandez Gavela, A.; Grajales Garcia, D.; Ramirez, J. C.; Lechuga, L. M. Last Advances in Silicon-Based Optical Biosensors. Sensors 2016, 16, 285. (36) Cross, G. H.; Reeves, A. A.; Brand, S.; Popplewell, J. F.; Peel, L. L.; Swann, M. J.; Freeman, N. J. A New Quantitative Optical Biosensor for Protein Characterization. Biosens. Bioelectron. 2003, 19, 383−390. (37) Cross, G. H.; Ren, Y.; Freeman, N. J. Young’s Fringes from Vertically Integrated Slab Waveguides: Applications to Humidity Sensing. J. Appl. Phys. 1999, 86, 6483−6488. (38) Cross, G. H.; Ren, Y.; Swann, M. J. Refractometric Discrimination of Void-Space Filling and Swelling During Vapour Sorption in Polymer Films. Analyst 2000, 125, 2173−2175. (39) Liedberg, B.; Nylander, C.; Lundstrom, I. Biosensing with Surface Plasmon Resonance–How It All Started. Biosens. Bioelectron. 1995, 10, i−ix.

before returning to Australia. Since 2009, Dr. Del Borgo has been a research fellow in the Department of Biochemistry & Molecular Biology. He is now working on the generation of peptide-based nanomaterials with application in drug delivery and tissue engineering. Professor Mibel Aguilar is a Bioanalytical and Biophysical Chemist at Monash University whose research focuses on biomembrane nanotechnology, biomaterials, and peptidomimetic drug design. She completed her Ph.D. in Chemistry at the University of Melbourne studying the metabolism and toxicity of paracetamol, and then her move to Monash University coincided with an increasing focus on bioanalytical and biophysical methods for studying peptide and protein structure and function. Her group has published over 220 papers and book chapters and focuses on peptide-based drug and biomaterial design and biomembrane nanotechnology, developing novel compounds that allow us to exploit the potential of peptides as drugs.

ACKNOWLEDGMENTS The financial support of the National Health & Medical Research Council (Grants 1084648 and 1084462) is gratefully acknowledged. REFERENCES (1) Cournia, Z.; Allen, T. W.; Andricioaei, I.; Antonny, B.; Baum, D.; Brannigan, G.; Buchete, N. V.; Deckman, J. T.; Delemotte, L.; Del Val, C. Membrane Protein Structure, Function, and Dynamics: A Perspective from Experiments and Theory. J. Membr. Biol. 2015, 248, 611−640. (2) Whitty, A. Cooperativity and Biological Complexity. Nat. Chem. Biol. 2008, 4, 435−439. (3) von Heijne, G. Membrane-Protein Topology. Nat. Rev. Mol. Cell Biol. 2006, 7, 909−918. (4) Bloom, M.; Evans, E.; Mouritsen, O. G. Physical Properties of the Fluid Lipid-Bilayer Component of Cell Membranes: A Perspective. Q. Rev. Biophys. 1991, 24, 293−397. (5) Hianik, T. Structure and Physical Properties of Biomembranes and Model Membranes. Acta Phys. Slovaca 2006, 56, 687−805. (6) McMahon, H. T.; Gallop, J. L. Membrane Curvature and Mechanisms of Dynamic Cell Membrane Remodelling. Nature 2005, 438, 590−596. (7) Phillips, R.; Ursell, T.; Wiggins, P.; Sens, P. Emerging Roles for Lipids in Shaping Membrane-Protein Function. Nature 2009, 459, 379−385. (8) Vigh, L.; Escriba, P. V.; Sonnleitner, A.; Sonnleitner, M.; Piotto, S.; Maresca, B.; Horvath, I.; Harwood, J. L. The Significance of Lipid Composition for Membrane Activity: New Concepts and Ways of Assessing Function. Prog. Lipid Res. 2005, 44, 303−344. (9) Paluch, E.; Heisenberg, C. P. Biology and Physics of Cell Shape Changes in Development. Curr. Biol. 2009, 19, R790−799. (10) Shnyrova, A. V.; Frolov, V. A.; Zimmerberg, J. Domain-Driven Morphogenesis of Cellular Membranes. Curr. Biol. 2009, 19, R772− 780. (11) Peetla, C.; Vijayaraghavalu, S.; Labhasetwar, V. Biophysics of Cell Membrane Lipids in Cancer Drug Resistance: Implications for Drug Transport and Drug Delivery with Nanoparticles. Adv. Drug Delivery Rev. 2013, 65, 1686−1698. (12) Mukherjee, S.; Maxfield, F. R. Role of Membrane Organization and Membrane Domains in Endocytic Lipid Trafficking. Traffic 2000, 1, 203−211. (13) Jarsch, I. K.; Daste, F.; Gallop, J. L. Membrane Curvature in Cell Biology: An Integration of Molecular Mechanisms. J. Cell Biol. 2016, 214, 375−387. (14) Wu, W.; Shi, X.; Xu, C. Regulation of T Cell Signalling by Membrane Lipids. Nat. Rev. Immunol. 2016, 16, 690−701. (15) Agmon, E.; Stockwell, B. R. Lipid Homeostasis and Regulated Cell Death. Curr. Opin. Chem. Biol. 2017, 39, 83−89. 5466

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

(40) Lofas, S.; Johnsson, B. A Novel Hydrogel Matrix on Gold Surfaces in Surface Plasmon Resonance Sensors for Fast and Efficient Covalent Immobilization of Ligands. J. Chem. Soc., Chem. Commun. 1990, 1526−1528. (41) Löfås, S.; Malmqvist, M.; Rönnberg, I.; Stenberg, E.; Liedberg, B.; Lundström, I. Bioanalysis with Surface Plasmon Resonance. Sens. Actuators, B 1991, 5, 79−84. (42) Brian, A. A.; McConnell, H. M. Allogeneic Stimulation of Cytotoxic T Cells by Supported Planar Membranes. Proc. Natl. Acad. Sci. U. S. A. 1984, 81, 6159−6163. (43) Sackmann, E. Supported Membranes: Scientific and Practical Applications. Science 1996, 271, 43−48. (44) Groves, J. T.; Ulman, N.; Boxer, S. G. Micropatterning Fluid Lipid Bilayers on Solid Supports. Science 1997, 275, 651−653. (45) Besenicar, M.; Macek, P.; Lakey, J. H.; Anderluh, G. Surface Plasmon Resonance in Protein-Membrane Interactions. Chem. Phys. Lipids 2006, 141, 169−178. (46) Cooper, M. A. Advances in Membrane Receptor Screening and Analysis. J. Mol. Recognit. 2004, 17, 286−315. (47) Heyse, S.; Vogel, H.; Sanger, M.; Sigrist, H. Covalent Attachment of Functionalized Lipid Bilayers to Planar Waveguides for Measuring Protein Binding to Biomimetic Membranes. Protein Sci. 1995, 4, 2532−2544. (48) Maynard, J. A.; Lindquist, N. C.; Sutherland, J. N.; Lesuffleur, A.; Warrington, A. E.; Rodriguez, M.; Oh, S. H. Surface Plasmon Resonance for High-Throughput Ligand Screening of MembraneBound Proteins. Biotechnol. J. 2009, 4, 1542−1558. (49) Mozsolits, H.; Aguilar, M. I. Surface Plasmon Resonance Spectroscopy: An Emerging Tool for the Study of Peptide-Membrane Interactions. Biopolymers 2002, 66, 3−18. (50) Patching, S. G. Surface Plasmon Resonance Spectroscopy for Characterization of Membrane Protein-Ligand Interactions and Its Potential for Drug Discovery. Biochim. Biophys. Acta, Biomembr. 2014, 1838, 43−55. (51) Salamon, Z.; Wang, Y.; Soulages, J. L.; Brown, M. F.; Tollin, G. Surface Plasmon Resonance Spectroscopy Studies of Membrane Proteins: Transducin Binding and Activation by Rhodopsin Monitored in Thin Membrane Films. Biophys. J. 1996, 71, 283−294. (52) Sackmann, E. K.; Fulton, A. L.; Beebe, D. J. The Present and Future Role of Microfluidics in Biomedical Research. Nature 2014, 507, 181−189. (53) Ainla, A.; Gozen, I.; Hakonen, B.; Jesorka, A. Lab on a Biomembrane: Rapid Prototyping and Manipulation of 2d Fluidic Lipid Bilayers Circuits. Sci. Rep. 2013, 3, 2743. (54) Groves, J. T.; Boxer, S. G. Micropattern Formation in Supported Lipid Membranes. Acc. Chem. Res. 2002, 35, 149−157. (55) Marsh, D.; Bartucci, R.; Sportelli, L. Lipid Membranes with Grafted Polymers: Physicochemical Aspects. Biochim. Biophys. Acta, Biomembr. 2003, 1615, 33−59. (56) Cardona, M. Fresnel Reflection and Surface Plasmons. Am. J. Phys. 1971, 39, 1277−1277. (57) Swalen, J. D.; Gordon, J. G., II; Philpott, M. R.; Brillante, A.; Pockrand, I.; Santo, R. Plasmon Surface Polariton Dispersion by Direct Optical Observation. Am. J. Phys. 1980, 48, 669−672. (58) Homola, J. Surface Plasmon Resonance Sensors for Detection of Chemical and Biological Species. Chem. Rev. 2008, 108, 462−493. (59) Homola, J. In Surface Plasmon Resonance Based Sensors; Homola, J., Ed.; Springer: Berlin, 2006. (60) Raether, H. In Surface Plasmons on Smooth and Rough Surfaces and on Gratings; Springer: Berlin, 1988. (61) Mayer, K. M.; Hafner, J. H. Localized Surface Plasmon Resonance Sensors. Chem. Rev. 2011, 111, 3828−3857. (62) Stewart, M. E.; Anderton, C. R.; Thompson, L. B.; Maria, J.; Gray, S. K.; Rogers, J. A.; Nuzzo, R. G. Nanostructured Plasmonic Sensors. Chem. Rev. 2008, 108, 494−521. (63) Willets, K. A.; Van Duyne, R. P. Localized Surface Plasmon Resonance Spectroscopy and Sensing. Annu. Rev. Phys. Chem. 2007, 58, 267−297.

(64) Kretschmann, E.; Raether, H. Notizen: Radiative Decay of Non Radiative Surface Plasmons Excited by Light. Z. Naturforsch., A: Phys. Sci. 1968, 23, 2135−2136. (65) Otto, A. Excitation of Nonradiative Surface Plasma Waves in Silver by the Method of Frustrated Total Reflection. Z. Phys. A: Hadrons Nucl. 1968, 216, 398−410. (66) Shalabney, A.; Abdulhalim, I. Sensitivity-Enhancement Methods for Surface Plasmon Sensors. Laser Photon. Rev. 2011, 5, 571−606. (67) Huang, Y. H.; Ho, H. P.; Kong, S. K.; Kabashin, A. V. PhaseSensitive Surface Plasmon Resonance Biosensors: Methodology, Instrumentation and Applications. Ann. Phys. 2012, 524, 637−662. (68) Matsubara, K.; Kawata, S.; Minami, S. Optical Chemical Sensor Based on Surface Plasmon Measurement. Appl. Opt. 1988, 27, 1160− 1163. (69) Akimoto, T.; Sasaki, S.; Ikebukuro, K.; Karube, I. Effect of Incident Angle of Light on Sensitivity and Detection Limit for Layers of Antibody with Surface Plasmon Resonance Spectroscopy. Biosens. Bioelectron. 2000, 15, 355−362. (70) Peterlinz, K. A.; Georgiadis, R. Two-Color Approach for Determination of Thickness and Dielectric Constant of Thin Films Using Surface Plasmon Resonance Spectroscopy. Opt. Commun. 1996, 130, 260−266. (71) Peterlinz, K. A.; Georgiadis, R. M.; Herne, T. M.; Tarlov, M. J. Observation of Hybridization and Dehybridization of Thiol-Tethered DNA Using Two-Color Surface Plasmon Resonance Spectroscopy. J. Am. Chem. Soc. 1997, 119, 3401−3402. (72) Nelson, B. P.; Frutos, A. G.; Brockman, J. M.; Corn, R. M. NearInfrared Surface Plasmon Resonance Measurements of Ultrathin Films. 1. Angle Shift and Spr Imaging Experiments. Anal. Chem. 1999, 71, 3928−3934. (73) Brockman, J. M.; Nelson, B. P.; Corn, R. M. Surface Plasmon Resonance Imaging Measurements of Ultrathin Organic Films. Annu. Rev. Phys. Chem. 2000, 51, 41−63. (74) Zhang, L. M.; Uttamchandani, D. Optical Chemical Sensing Employing Surface Plasmon Resonance. Electron. Lett. 1988, 24, 1469−1470. (75) Liu, X.; Song, D.; Zhang, Q.; Tian, Y.; Ding, L.; Zhang, H. Wavelength-Modulation Surface Plasmon Resonance Sensor. TrAC, Trends Anal. Chem. 2005, 24, 887−893. (76) Sun, Y.; Liu, X.; Song, D.; Tian, Y.; Bi, S.; Zhang, H. Sensitivity Enhancement of Wavelength Modulation Surface Plasmon Resonance Biosensor by Improving the Baseline Solution. Anal. Chim. Acta 2006, 569, 21−26. (77) De Feijter, J. A.; Benjamins, J.; Veer, F. A. Ellipsometry as a Tool to Study the Adsorption Behavior of Synthetic and Biopolymers at the Air−Water Interface. Biopolymers 1978, 17, 1759−1772. (78) Voros, J. The Density and Refractive Index of Adsorbing Protein Layers. Biophys. J. 2004, 87, 553−561. (79) Erbe, A.; Sigel, R. Tilt Angle of Lipid Acyl Chains in Unilamellar Vesicles Determined by Ellipsometric Light Scattering. Eur. Phys. J. E: Soft Matter Biol. Phys. 2007, 22, 303−309. (80) Zangle, T. A.; Teitell, M. A. Live-Cell Mass Profiling: An Emerging Approach in Quantitative Biophysics. Nat. Methods 2014, 11, 1221−1228. (81) Nylander, C.; Liedberg, B.; Lind, T. K. Gas Detection by Means of Surface Plasmon Resonance. Sens. Actuators 1982, 3, 79−88. (82) Sepulveda, B.; Calle, A.; Lechuga, L. M.; Armelles, G. Highly Sensitive Detection of Biomolecules with the Magneto-Optic SurfacePlasmon-Resonance Sensor. Opt. Lett. 2006, 31, 1085−1087. (83) Rothenhausler, B.; Knoll, W. Surface−Plasmon Microscopy. Nature 1988, 332, 615−617. (84) Yeatman, E.; Ash, E. A. Surface Plasmon Microscopy. Electron. Lett. 1987, 23, 1091−1092. (85) Puiu, M.; Bala, C. SPR and SPR Imaging: Recent Trends in Developing Nanodevices for Detection and Real-Time Monitoring of Biomolecular Events. Sensors 2016, 16, 87010.3390/s16060870. (86) Spoto, G.; Minunni, M. Surface Plasmon Resonance Imaging: What Next? J. Phys. Chem. Lett. 2012, 3, 2682−2691. 5467

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

(87) Wong, C. L.; Olivo, M. Surface Plasmon Resonance Imaging Sensors: A Review. Plasmonics 2014, 9, 809−824. (88) Piliarik, M.; Vaisocherova, H.; Homola, J. A New Surface Plasmon Resonance Sensor for High-Throughput Screening Applications. Biosens. Bioelectron. 2005, 20, 2104−2110. (89) Zybin, A.; Grunwald, C.; Mirsky, V. M.; Kuhlmann, J.; Wolfbeis, O. S.; Niemax, K. Double-Wavelength Technique for Surface Plasmon Resonance Measurements: Basic Concept and Applications for Single Sensors and Two-Dimensional Sensor Arrays. Anal. Chem. 2005, 77, 2393−2399. (90) Nelson, S. G.; Johnston, K. S.; Yee, S. S. High Sensitivity Surface Plasmon Resonace Sensor Based on Phase Detection. Sens. Actuators, B 1996, 35, 187−191. (91) Kabashin, A. V.; Nikitin, P. I. Surface Plasmon Resonance Interferometer for Bio- and Chemical-Sensors. Opt. Commun. 1998, 150, 5−8. (92) Nikitin, P. I.; Grigorenko, A. N.; Beloglazov, A. A.; Valeiko, M. V.; Savchuk, A. I.; Savchuk, O. A.; Steiner, G.; Kuhne, C.; Huebner, A.; Salzer, R. Surface Plasmon Resonance Interferometry for Micro-Array Biosensing. Sens. Actuators, A 2000, 85, 189−193. (93) Kabashin, A. V.; Patskovsky, S.; Grigorenko, A. N. Phase and Amplitude Sensitivities in Surface Plasmon Resonance Bio and Chemical Sensing. Opt. Express 2009, 17, 21191−21204. (94) Huang, Y. H.; Ho, H. P.; Wu, S. Y.; Kong, S. K.; Wong, W. W.; Shum, P. Phase Sensitive Spr Sensor for Wide Dynamic Range Detection. Opt. Lett. 2011, 36, 4092−4094. (95) Kashif, M.; Bakar, A. A.; Arsad, N.; Shaari, S. Development of Phase Detection Schemes Based on Surface Plasmon Resonance Using Interferometry. Sensors 2014, 14, 15914−15938. (96) Jung, L. S.; Campbell, C. T.; Chinowsky, T. M.; Mar, M. N.; Yee, S. S. Quantitative Interpretation of the Response of Surface Plasmon Resonance Sensors to Adsorbed Films. Langmuir 1998, 14, 5636−5648. (97) Malinsky, M. D.; Kelly, K. L.; Schatz, G. C.; Van Duyne, R. P. Chain Length Dependence and Sensing Capabilities of the Localized Surface Plasmon Resonance of Silver Nanoparticles Chemically Modified with Alkanethiol Self-Assembled Monolayers. J. Am. Chem. Soc. 2001, 123, 1471−1482. (98) Haes, A. J.; Zou, S.; Schatz, G. C.; Van Duyne, R. P. A Nanoscale Optical Biosensor: The Long Range Distance Dependence of the Localized Surface Plasmon Resonance of Noble Metal Nanoparticles. J. Phys. Chem. B 2004, 108, 109−116. (99) Kelf, T. A.; Sugawara, Y.; Cole, R. M.; Baumberg, J. J.; Abdelsalam, M. E.; Cintra, S.; Mahajan, S.; Russell, A. E.; Bartlett, P. N. Localized and Delocalized Plasmons in Metallic Nanovoids. Phys. Rev. B: Condens. Matter Mater. Phys. 2006, 74, 245415. (100) Live, L. S.; Dhawan, A.; Gibson, K. F.; Poirier-Richard, H. P.; Graham, D.; Canva, M.; Vo-Dinh, T.; Masson, J. F. Angle-Dependent Resonance of Localized and Propagating Surface Plasmons in Microhole Arrays for Enhanced Biosensing. Anal. Bioanal. Chem. 2012, 404, 2859−2868. (101) Live, L. S.; Bolduc, O. R.; Masson, J. F. Propagating Surface Plasmon Resonance on Microhole Arrays. Anal. Chem. 2010, 82, 3780−3787. (102) Kegel, L. L.; Boyne, D.; Booksh, K. S. Sensing with PrismBased near-Infrared Surface Plasmon Resonance Spectroscopy on Nanohole Array Platforms. Anal. Chem. 2014, 86, 3355−3364. (103) Campbell, C. T.; Kim, G. Spr Microscopy and Its Applications to High-Throughput Analyses of Biomolecular Binding Events and Their Kinetics. Biomaterials 2007, 28, 2380−2392. (104) Ruemmele, J. A.; Golden, M. S.; Gao, Y.; Cornelius, E. M.; Anderson, M. E.; Postelnicu, L.; Georgiadis, R. M. Quantitative Surface Plasmon Resonance Imaging: A Simple Approach to Automated Angle Scanning. Anal. Chem. 2008, 80, 4752−4756. (105) Zhou, C.; Jin, W.; Zhang, Y.; Yang, M.; Xiang, L.; Wu, Z.; Jin, Q.; Mu, Y. An Angle-Scanning Surface Plasmon Resonance Imaging Device for Detection of Mismatched Bases in Caspase-3 DNA. Anal. Methods 2013, 5, 2369−2373.

(106) Wong, C. L.; Ho, H. P.; Wong, P. L.; Wu, S. Y.; Guo, F. In 2003 IEEE Conference on Electron Devices and Solid-State Circuits, Edssc 2003; Institute of Electrical and Electronics Engineers Inc., 2003. (107) Wong, C. L.; Chen, G. C.; Ng, B. K.; Agarwal, S.; Lin, Z.; Chen, P.; Ho, H. P. Multiplex Spectral Surface Plasmon Resonance Imaging (Spri) Sensor Based on the Polarization Control Scheme. Opt. Express 2011, 19, 18965−18978. (108) Piliarik, M.; Homola, J. Self-Referencing Spr Imaging for Most Demanding High-Throughput Screening Applications. Sens. Actuators, B 2008, 134, 353−355. (109) Jordan, C. E.; Frutos, A. G.; Thiel, A. J.; Corn, R. M. Surface Plasmon Resonance Imaging Measurements of DNA Hybridization Adsorption and Streptavidin/DNA Multilayer Formation at Chemically Modified Gold Surfaces. Anal. Chem. 1997, 69, 4939−4947. (110) Jordan, C. E.; Corn, R. M. Surface Plasmon Resonance Imaging Measurements of Electrostatic Biopolymer Adsorption onto Chemically Modified Gold Surfaces. Anal. Chem. 1997, 69, 1449− 1456. (111) Wark, A. W.; Lee, H. J.; Corn, R. M. Long-Range Surface Plasmon Resonance Imaging for Bioaffinity Sensors. Anal. Chem. 2005, 77, 3904−3907. (112) Vala, M.; Etheridge, S.; Roach, J. A.; Homola, J. Long-Range Surface Plasmons for Sensitive Detection of Bacterial Analytes. Sens. Actuators, B 2009, 139, 59−63. (113) Wang, Y.; Knoll, W.; Dostalek, J. Bacterial Pathogen Surface Plasmon Resonance Biosensor Advanced by Long Range Surface Plasmons and Magnetic Nanoparticle Assays. Anal. Chem. 2012, 84, 8345−8350. (114) Wong, W. R.; Krupin, O.; Sekaran, S. D.; Mahamd Adikan, F. R.; Berini, P. Serological Diagnosis of Dengue Infection in Blood Plasma Using Long-Range Surface Plasmon Waveguides. Anal. Chem. 2014, 86, 1735−1743. (115) Chen, S. J.; Su, Y. D.; Hsiu, F. M.; Tsou, C. Y.; Chen, Y. K. Surface Plasmon Resonance Phase-Shift Interferometry: Real-Time DNA Microarray Hybridization Analysis. J. Biomed. Opt. 2005, 10, 034005. (116) Su, Y. D.; Chen, S. J.; Yeh, T. L. Common-Path Phase-Shift Interferometry Surface Plasmon Resonance Imaging System. Opt. Lett. 2005, 30, 1488−1490. (117) Huang, B.; Yu, F.; Zare, R. N. Surface Plasmon Resonance Imaging Using a High Numerical Aperture Microscope Objective. Anal. Chem. 2007, 79, 2979−2983. (118) Barchiesi, D.; Otto, A. Excitations of Surface Plasmon Polaritons by Attenuated Total Reflection, Revisited. Rivista del Nuovo Cimento 2013, 36, 173−209. (119) Shumaker-Parry, J. S.; Campbell, C. T. Quantitative Methods for Spatially Resolved Adsorption/Desorption Measurements in Real Time by Surface Plasmon Resonance Microscopy. Anal. Chem. 2004, 76, 907−917. (120) Laplatine, L.; Leroy, L.; Calemczuk, R.; Baganizi, D.; Marche, P. N.; Roupioz, Y.; Livache, T. Spatial Resolution in Prism-Based Surface Plasmon Resonance Microscopy. Opt. Express 2014, 22, 22771−22785. (121) Wang, W.; Yang, Y.; Wang, S.; Nagaraj, V. J.; Liu, Q.; Wu, J.; Tao, N. Label-Free Measuring and Mapping of Binding Kinetics of Membrane Proteins in Single Living Cells. Nat. Chem. 2012, 4, 846− 853. (122) Yu, H.; Shan, X.; Wang, S.; Chen, H.; Tao, N. Plasmonic Imaging and Detection of Single DNA Molecules. ACS Nano 2014, 8, 3427−3433. (123) Wang, S.; Shan, X.; Patel, U.; Huang, X.; Lu, J.; Li, J.; Tao, N. Label-Free Imaging, Detection, and Mass Measurement of Single Viruses by Surface Plasmon Resonance. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 16028−16032. (124) Wang, W.; Wang, S.; Liu, Q.; Wu, J.; Tao, N. Mapping SingleCell-Substrate Interactions by Surface Plasmon Resonance Microscopy. Langmuir 2012, 28, 13373−13379. 5468

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Arrays of Subwavelength Holes. Annu. Rev. Phys. Chem. 2008, 59, 179−202. (147) West, P. R.; Ishii, S.; Naik, G. V.; Emani, N. K.; Shalaev, V. M.; Boltasseva, A. Searching for Better Plasmonic Materials. Laser Photon. Rev. 2010, 4, 795−808. (148) Pang, L.; Hwang, G. M.; Slutsky, B.; Fainman, Y. Spectral Sensitivity of Two-Dimensional Nanohole Array Surface Plasmon Polariton Resonance Sensor. Appl. Phys. Lett. 2007, 91, 123112. (149) Dahlin, A.; Zach, M.; Rindzevicius, T.; Kall, M.; Sutherland, D. S.; Hook, F. Localized Surface Plasmon Resonance Sensing of LipidMembrane-Mediated Biorecognition Events. J. Am. Chem. Soc. 2005, 127, 5043−5048. (150) Dahlin, A. B. Sensing Applications Based on Plasmonic Nanopores: The Hole Story. Analyst 2015, 140, 4748−4759. (151) Monteiro, J. P.; Carneiro, L. B.; Rahman, M. M.; Brolo, A. G.; Santos, M. J. L.; Ferreira, J.; Girotto, E. M. Effect of Periodicity on the Performance of Surface Plasmon Resonance Sensors Based on Subwavelength Nanohole Arrays. Sens. Actuators, B 2013, 178, 366− 370. (152) Gordon, R.; Brolo, A. G.; Sinton, D.; Kavanagh, K. L. Resonant Optical Transmission through Hole-Arrays in Metal Films: Physics and Applications. Laser Photonics Rev. 2010, 4, 311−335. (153) Lesuffleur, A.; Im, H.; Lindquist, N. C.; Oh, S. H. Periodic Nanohole Arrays with Shape-Enhanced Plasmon Resonance as RealTime Biosensors. Appl. Phys. Lett. 2007, 90, 243110. (154) Lesuffleur, A.; Im, H.; Lindquist, N. C.; Lim, K. S.; Oh, S. H. Plasmonic Nanohole Arrays for Real-Time Multiplex Biosensing. Proc. SPIE 2008, 7035, 703504. (155) Xiong, K.; Emilsson, G.; Dahlin, A. B. Biosensing Using Plasmonic Nanohole Arrays with Small, Homogenous and Tunable Aperture Diameters. Analyst 2016, 141, 3803−3810. (156) van der Molen, K. L.; Segerink, F. B.; van Hulst, N. F.; Kuipers, L. Influence of Hole Size on the Extraordinary Transmission through Subwavelength Hole Arrays. Appl. Phys. Lett. 2004, 85, 4316−4318. (157) Klein Koerkamp, K. J.; Enoch, S.; Segerink, F. B.; van Hulst, N. F.; Kuipers, L. Strong Influence of Hole Shape on Extraordinary Transmission through Periodic Arrays of Subwavelength Holes. Phys. Rev. Lett. 2004, 92, 183901. (158) Ekşioğlu, Y.; Cetin, A. E.; Petrácě k, J. Optical Response of Plasmonic Nanohole Arrays: Comparison of Square and Hexagonal Lattices. Plasmonics 2016, 11, 851−856. (159) Rockstuhl, C.; Lederer, F.; Zentgraf, T.; Giessen, H. Enhanced Transmission of Periodic, Quasiperiodic, and Random Nanoaperture Arrays. Appl. Phys. Lett. 2007, 91, 151109. (160) Sannomiya, T.; Scholder, O.; Jefimovs, K.; Hafner, C.; Dahlin, A. B. Investigation of Plasmon Resonances in Metal Films with Nanohole Arrays for Biosensing Applications. Small 2011, 7, 1653− 1663. (161) Braun, J.; Gompf, B.; Kobiela, G.; Dressel, M. How Holes Can Obscure the View: Suppressed Transmission through an Ultrathin Metal Film by a Subwavelength Hole Array. Phys. Rev. Lett. 2009, 103, 203901. (162) Dahlin, A. B.; Wittenberg, N. J.; Höök, F.; Oh, S. H. Promises and Challenges of Nanoplasmonic Devices for Refractometric Biosensing. Nanophotonics 2013, 2, 83−101. (163) Kozma, P.; Kehl, F.; Ehrentreich-Forster, E.; Stamm, C.; Bier, F. F. Integrated Planar Optical Waveguide Interferometer Biosensors: A Comparative Review. Biosens. Bioelectron. 2014, 58, 287−307. (164) Cross, G. H.; Freeman, N. J.; Swann, M. J. In Handbook of Biosensors and Biochips; Marks, R. S., Lowe, C. R., Cullen, D. C., Weetall, H. H., Karube, I., Eds.; John Wiley & Sons, Ltd, 2008. (165) Cross, G. H. Fundamental Limit to the Use of Effective Medium Theories in Optics. Opt. Lett. 2013, 38, 3057−3060. (166) Coffey, P. D.; Swann, M. J.; Waigh, T. A.; Mu, Q.; Lu, J. R. The Structure and Mass of Heterogeneous Thin Films Measured with Dual Polarization Interferometry and Ellipsometry. RSC Adv. 2013, 3, 3316−3324. (167) Mashaghi, A.; Swann, M.; Popplewell, J.; Textor, M.; Reimhult, E. Optical Anisotropy of Supported Lipid Structures Probed by

(125) Ebbesen, T. W.; Lezec, H. J.; Ghaemi, H. F.; Thio, T.; Wolff, P. A. Extraordinary Optical Transmission through Sub-Wavelength Hole Arrays. Nature 1998, 391, 667−669. (126) Genet, C.; Ebbesen, T. W. Light in Tiny Holes. Nature 2007, 445, 39−46. (127) Gordon, R.; Sinton, D.; Kavanagh, K. L.; Brolo, A. G. A New Generation of Sensors Based on Extraordinary Optical Transmission. Acc. Chem. Res. 2008, 41, 1049−1057. (128) Jackman, J. A.; Rahim Ferhan, A.; Cho, N. J. Nanoplasmonic Sensors for Biointerfacial Science. Chem. Soc. Rev. 2017, 46, 3615− 3660. (129) Lesuffleur, A.; Lim, K. S.; Lindquist, N. C.; Im, H.; Warrington, A. E.; Rodriguez, M.; Oh, S. H. Plasmonic Nanohole Arrays for LabelFree Kinetic Biosensing in a Lipid Membrane Environment. Conf. Proc. IEEE Eng. Med. Biol. Soc. 2009, 2009, 1481−1484. (130) Garcia-Vidal, F. J.; Martin-Moreno, L.; Ebbesen, T. W.; Kuipers, L. Light Passing through Subwavelength Apertures. Rev. Mod. Phys. 2010, 82, 729−787. (131) Rodrigo, S.; de León -Pér ez, F.; Martin-Moreno, L. Extraordinary Optical Transmission: Fundamentals and Applications. Proc. IEEE 2016, 104, 2288−2306. (132) Couture, M.; Live, L. S.; Dhawan, A.; Masson, J. F. Eot or Kretschmann Configuration? Comparative Study of the Plasmonic Modes in Gold Nanohole Arrays. Analyst 2012, 137, 4162−4170. (133) Lindquist, N. C.; Nagpal, P.; McPeak, K. M.; Norris, D. J.; Oh, S. H. Engineering Metallic Nanostructures for Plasmonics and Nanophotonics. Rep. Prog. Phys. 2012, 75, 036501. (134) Gao, H.; Henzie, J.; Odom, T. W. Direct Evidence for Surface Plasmon-Mediated Enhanced Light Transmission through Metallic Nanohole Arrays. Nano Lett. 2006, 6, 2104−2108. (135) Huang, C. P.; Wang, Q. J.; Zhu, Y. Y. Dual Effect of Surface Plasmons in Light Transmission through Perforated Metal Films. Phys. Rev. B: Condens. Matter Mater. Phys. 2007, 75, 245421. (136) Cao, Q.; Lalanne, P. Negative Role of Surface Plasmons in the Transmission of Metallic Gratings with Very Narrow Slits. Phys. Rev. Lett. 2002, 88, 057403. (137) Lezec, H.; Thio, T. Diffracted Evanescent Wave Model for Enhanced and Suppressed Optical Transmission through Subwavelength Hole Arrays. Opt. Express 2004, 12, 3629−3651. (138) Chang, S. H.; Gray, S.; Schatz, G. Surface Plasmon Generation and Light Transmission by Isolated Nanoholes and Arrays of Nanoholes in Thin Metal Films. Opt. Express 2005, 13, 3150−3165. (139) Pacifici, D.; Lezec, H. J.; Atwater, H. A.; Weiner, J. Quantitative Determination of Optical Transmission through Subwavelength Slit Arrays in Ag Films: Role of Surface Wave Interference and Local Coupling between Adjacent Slits. Phys. Rev. B: Condens. Matter Mater. Phys. 2008, 77, 115411. (140) Gay, G.; Alloschery, O.; Viaris de Lesegno, B.; O’Dwyer, C.; Weiner, J.; Lezec, H. J. The Optical Response of Nanostructured Surfaces and the Composite Diffracted Evanescent Wave Model. Nat. Phys. 2006, 2, 262. (141) Gay, G.; Alloschery, O.; Weiner, J.; Lezec, H. J.; O’Dwyer, C.; Sukharev, M.; Seideman, T. The Response of Nanostructured Surfaces in the near Field. Nat. Phys. 2006, 2, 792−792. (142) Weiner, J. The Physics of Light Transmission through Subwavelength Apertures and Aperture Arrays. Rep. Prog. Phys. 2009, 72, 064401. (143) Weiner, J. The Electromagnetics of Light Transmission through Subwavelength Slits in Metallic Films. Opt. Express 2011, 19, 16139−16153. (144) Sarrazin, M.; Vigneron, J. P.; Vigoureux, J. M. Role of Wood Anomalies in Optical Properties of Thin Metallic Films with a Bidimensional Array of Subwavelength Holes. Phys. Rev. B: Condens. Matter Mater. Phys. 2003, 67, 085415. (145) Genet, C.; van Exter, M. P.; Woerdman, J. P. Fano-Type Interpretation of Red Shifts and Red Tails in Hole Array Transmission Spectra. Opt. Commun. 2003, 225, 331−336. (146) Coe, J. V.; Heer, J. M.; Teeters-Kennedy, S.; Tian, H.; Rodriguez, K. R. Extraordinary Transmission of Metal Films with 5469

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

(188) Hohner, A. O.; David, M. P.; Radler, J. O. Controlled SolventExchange Deposition of Phospholipid Membranes onto Solid Surfaces. Biointerphases 2010, 5, 1−8. (189) Tabaei, S. R.; Choi, J. H.; Haw Zan, G.; Zhdanov, V. P.; Cho, N. J. Solvent-Assisted Lipid Bilayer Formation on Silicon Dioxide and Gold. Langmuir 2014, 30, 10363−10373. (190) Bayley, H.; Cronin, B.; Heron, A.; Holden, M. A.; Hwang, W. L.; Syeda, R.; Thompson, J.; Wallace, M. Droplet Interface Bilayers. Mol. BioSyst. 2008, 4, 1191−1208. (191) Funakoshi, K.; Suzuki, H.; Takeuchi, S. Lipid Bilayer Formation by Contacting Monolayers in a Microfluidic Device for Membrane Protein Analysis. Anal. Chem. 2006, 78, 8169−8174. (192) Orrit, M. From Langmuir−Blodgett Films to Single Molecules. Colloids Surf., B 2009, 74, 396−400. (193) Fendler, J. H. Membrane Mimetic Chemistry; Wiley: New York, 1982. (194) Ulman, A. An Introduction to Ultrathin Organic Films from Langmuir-Blodgett to Self-Assembly; Academic Press: Boston, 1991. (195) Grieshaber, D.; MacKenzie, R.; Voros, J.; Reimhult, E. Electrochemical Biosensors - Sensor Principles and Architectures. Sensors 2008, 8, 1400−1458. (196) Madrid, E.; Horswell, S. L. Effect of Electric Field on Structure and Dynamics of Bilayers Formed from Anionic Phospholipids. Electrochim. Acta 2014, 146, 850−860. (197) Castellana, E. T.; Cremer, P. S. Solid Supported Lipid Bilayers: From Biophysical Studies to Sensor Design. Surf. Sci. Rep. 2006, 61, 429−444. (198) Siontorou, C.; Nikoleli, G.-P.; Nikolelis, D.; Karapetis, S. Artificial Lipid Membranes: Past, Present, and Future. Membranes 2017, 7, 38. (199) Fragneto, G.; Charitat, T.; Daillant, J. Floating Lipid Bilayers: Models for Physics and Biology. Eur. Biophys. J. 2012, 41, 863−874. (200) Basu, J. K.; Sanyal, M. K. Ordering and Growth of Langmuir− Blodgett Films: X-Ray Scattering Studies. Phys. Rep. 2002, 363, 1−84. (201) Lipkowski, J. In Advances in Planar Lipid Bilayers and Liposomes; Iglič, A., Kulkarni, C. V., Eds., 2014; Vol. 20. (202) Cui, D. F.; Howarth, V. A.; Petty, M. C.; Ancelin, H.; Yarwood, J. The Deposition and Characterization of Phosphatidic Acid Langmuir-Blodgett Films. Thin Solid Films 1990, 192, 391−396. (203) Lotta, T. I.; Laakkonen, L. J.; Virtanen, J. A.; Kinnunen, P. K. Characterization of Langmuir-Blodgett Films of 1,2-Dipalmitoyl-SnGlycero-3-Phosphatidylcholine and 1-Palmitoyl-2-[10-(Pyren-1-Yl)Decanoyl]-Sn-Glycero-3-Phosphat Idylcholine by Ftir-Atr. Chem. Phys. Lipids 1988, 46, 1−12. (204) Fringeli, U. P.; Gunthard, H. H. Infrared Membrane Spectroscopy. Mol. Biol., Biochem. Biophys. 1981, 31, 270−332. (205) Singh, S.; Keller, D. J. Atomic Force Microscopy of Supported Planar Membrane Bilayers. Biophys. J. 1991, 60, 1401−1410. (206) Marquês, J. T.; de Almeida, R. F. M.; Viana, A. S. Lipid Bilayers Supported on Bare and Modified Gold − Formation, Characterization and Relevance of Lipid Rafts. Electrochim. Acta 2014, 126, 139−150. (207) Reimhult, E.; Zach, M.; Hook, F.; Kasemo, B. A Multitechnique Study of Liposome Adsorption on Au and Lipid Bilayer Formation on Sio2. Langmuir 2006, 22, 3313−3319. (208) Reviakine, I.; Rossetti, F. F.; Morozov, A. N.; Textor, M. Investigating the Properties of Supported Vesicular Layers on Titanium Dioxide by Quartz Crystal Microbalance with Dissipation Measurements. J. Chem. Phys. 2005, 122, 204711. (209) Zan, G. H.; Jackman, J. A.; Kim, S.-O.; Cho, N.-J. Controlling Lipid Membrane Architecture for Tunable Nanoplasmonic Biosensing. Small 2014, 10, 4828−4832. (210) Noda, M.; Shimanouchi, T.; Okuyama, M.; Kuboi, R. A BioThermochemical Microbolometer with Immobilized Intact Liposome on Sensor Solid Surface. Sens. Actuators, B 2008, 135, 40−45. (211) Bayerl, T. M.; Bloom, M. Physical Properties of Single Phospholipid Bilayers Adsorbed to Micro Glass Beads. A New Vesicular Model System Studied by 2h-Nuclear Magnetic Resonance. Biophys. J. 1990, 58, 357−362.

Waveguide Spectroscopy and Its Application to Study of Supported Lipid Bilayer Formation Kinetics. Anal. Chem. 2008, 80, 3666−3676. (168) Salamon, Z.; Tollin, G. Optical Anisotropy in Lipid Bilayer Membranes: Coupled Plasmon-Waveguide Resonance Measurements of Molecular Orientation, Polarizability, and Shape. Biophys. J. 2001, 80, 1557−1567. (169) Johnson, S. J.; Bayerl, T. M.; McDermott, D. C.; Adam, G. W.; Rennie, A. R.; Thomas, R. K.; Sackmann, E. Structure of an Adsorbed Dimyristoylphosphatidylcholine Bilayer Measured with Specular Reflection of Neutrons. Biophys. J. 1991, 59, 289−294. (170) Zhao, H.; Brown, P. H.; Schuck, P. On the Distribution of Protein Refractive Index Increments. Biophys. J. 2011, 100, 2309− 2317. (171) Rupert, D. L.; Lasser, C.; Eldh, M.; Block, S.; Zhdanov, V. P.; Lotvall, J. O.; Bally, M.; Hook, F. Determination of Exosome Concentration in Solution Using Surface Plasmon Resonance Spectroscopy. Anal. Chem. 2014, 86, 5929−5936. (172) Reimhult, E.; Larsson, C.; Kasemo, B.; Hook, F. Simultaneous Surface Plasmon Resonance and Quartz Crystal Microbalance with Dissipation Monitoring Measurements of Biomolecular Adsorption Events Involving Structural Transformations and Variations in Coupled Water. Anal. Chem. 2004, 76, 7211−7220. (173) Cuypers, P. A.; Corsel, J. W.; Janssen, M. P.; Kop, J. M.; Hermens, W. T.; Hemker, H. C. The Adsorption of Prothrombin to Phosphatidylserine Multilayers Quantitated by Ellipsometry. J. Biol. Chem. 1983, 258, 2426−2431. (174) Sonesson, A. W.; Callisen, T. H.; Brismar, H.; Elofsson, U. M. A Comparison between Dual Polarization Interferometry (Dpi) and Surface Plasmon Resonance (Spr) for Protein Adsorption Studies. Colloids Surf., B 2007, 54, 236−240. (175) Tamm, L. K.; McConnell, H. M. Supported Phospholipid Bilayers. Biophys. J. 1985, 47, 105−113. (176) Brand, I. In Advances in Planar Lipid Bilayers and Liposomes; Iglič, A., Kulkarni, C. V., Eds.; Academic Press: Oxford, U.K., 2013; Vol. 18, Chapter 2. (177) Girard-Egrot, A. P.; Blum, L. J. In Nanobiotechnology of Biomimetic Membranes; Martin, D. K., Ed.; Springer: Boston, MA, 2007. (178) Jung, S. Y.; Holden, M. A.; Cremer, P. S.; Collier, C. P. TwoComponent Membrane Lithography Via Lipid Backfilling. ChemPhysChem 2005, 6, 423−426. (179) Mennicke, U.; Salditt, T. Preparation of Solid-Supported Lipid Bilayers by Spin-Coating. Langmuir 2002, 18, 8172−8177. (180) Lenhert, S.; Mirkin, C. A.; Fuchs, H. In Situ Lipid Dip-Pen Nanolithography under Water. Scanning 2010, 32, 15−23. (181) Lenhert, S.; Sun, P.; Wang, Y.; Fuchs, H.; Mirkin, C. A. Massively Parallel Dip-Pen Nanolithography of Heterogeneous Supported Phospholipid Multilayer Patterns. Small 2007, 3, 71−75. (182) Mager, M. D.; Melosh, N. A. Lipid Bilayer Deposition and Patterning Via Air Bubble Collapse. Langmuir 2007, 23, 9369−9377. (183) Brinker, C. J.; Lu, Y. F.; Sellinger, A.; Fan, H. Y. EvaporationInduced Self-Assembly: Nanostructures Made Easy. Adv. Mater. 1999, 11, 579−585. (184) Tiberg, F.; Harwigsson, I.; Malmsten, M. Formation of Model Lipid Bilayers at the Silica-Water Interface by Co-Adsorption with Non-Ionic Dodecyl Maltoside Surfactant. Eur. Biophys. J. 2000, 29, 196−203. (185) Vacklin, H. P.; Tiberg, F.; Thomas, R. K. Formation of Supported Phospholipid Bilayers Via Co-Adsorption with Beta-DDodecyl Maltoside. Biochim. Biophys. Acta, Biomembr. 2005, 1668, 17− 24. (186) Reimhult, E.; Höök, F.; Kasemo, B. Intact Vesicle Adsorption and Supported Biomembrane Formation from Vesicles in Solution: Influence of Surface Chemistry, Vesicle Size, Temperature, and Osmotic Pressure. Langmuir 2003, 19, 1681−1691. (187) Richter, R. P.; Berat, R.; Brisson, A. R. Formation of SolidSupported Lipid Bilayers: An Integrated View. Langmuir 2006, 22, 3497−3505. 5470

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

(212) Loidl-Stahlhofen, A.; Eckert, A.; Hartmann, T.; Schottner, M. Solid-Supported Lipid Membranes as a Tool for Determination of Membrane Affinity: High-Throughput Screening of a Physicochemical Parameter. J. Pharm. Sci. 2001, 90, 599−606. (213) Retzinger, G. S.; Meredith, S. C.; Lau, S. H.; Kaiser, E. T.; Kezdy, F. J. A Method for Probing the Affinity of Peptides for Amphiphilic Surfaces. Anal. Biochem. 1985, 150, 131−140. (214) Loidl-Stahlhofen, A.; Kaufmann, S.; Braunschweig, T.; Bayerl, T. M. The Thermodynamic Control of Protein Binding to Lipid Bilayers for Protein Chromatography. Nat. Biotechnol. 1996, 14, 999− 1002. (215) Schmitz, A. A.; Schleiff, E.; Rohring, C.; Loidl-Stahlhofen, A.; Vergeres, G. Interactions of Myristoylated Alanine-Rich C Kinase Substrate (Marcks)-Related Protein with a Novel Solid-Supported Lipid Membrane System (Transil). Anal. Biochem. 1999, 268, 343− 353. (216) Krause, E.; Dathe, M.; Wieprecht, T.; Bienert, M. Noncovalent Immobilized Artificial Membrane Chromatography, an Improved Method for Describing Peptide-Lipid Bilayer Interactions. J. Chromatogr. A 1999, 849, 125−133. (217) Puu, G.; Gustafson, I.; Artursson, E.; Ohlsson, P. A. Retained Activities of Some Membrane Proteins in Stable Lipid Bilayers on a Solid Support. Biosens. Bioelectron. 1995, 10, 463−476. (218) Puu, G.; Gustafson, I. Planar Lipid Bilayers on Solid Supports from Liposomes–Factors of Importance for Kinetics and Stability. Biochim. Biophys. Acta, Biomembr. 1997, 1327, 149−161. (219) Major, M.; Prieur, E.; Tocanne, J. F.; Betbeder, D.; Sautereau, A. M. Characterization and Phase Behaviour of Phospholipid Bilayers Adsorbed on Spherical Polysaccharidic Nanoparticles. Biochim. Biophys. Acta, Biomembr. 1997, 1327, 32−40. (220) Fabianowski, W.; Coyle, L. C.; Weber, B. A.; Granata, R. D.; Castner, D. G.; Sadownik, A.; Regen, S. L. Spontaneous Assembly of Phosphatidylcholine Monolayers Via Chemisorption onto Gold. Langmuir 1989, 5, 35−41. (221) Pidgeon, C.; Venkataram, U. V. Immobilized Artificial Membrane Chromatography: Supports Composed of Membrane Lipids. Anal. Biochem. 1989, 176, 36−47. (222) Lee, T. H.; Rivett, D.; Werkmeister, J.; Hewish, D.; Aguilar, M. I. Interaction of Amphipathic Peptides with an Immobilised Model Membrane. Lett. Pept. Sci. 1999, 6, 371−380. (223) Mozsolits, H.; Lee, T. H.; Wirth, H. J.; Perlmutter, P.; Aguilar, M. I. The Interaction of Bioactive Peptides with an Immobilized Phosphatidylcholine Monolayer. Biophys. J. 1999, 77, 1428−1444. (224) Cornell, B. A.; Braach-Maksvytis, V. L.; King, L. G.; Osman, P. D.; Raguse, B.; Wieczorek, L.; Pace, R. J. A Biosensor That Uses IonChannel Switches. Nature 1997, 387, 580−583. (225) Qiu, X.; Ong, S.; Bernal, C.; Rhee, D.; Pidgeon, C. A General Synthetic Route for Preparing Ether Phospholipids Suitable for Immobilization: A Phosphotriester Approach. J. Org. Chem. 1994, 59, 537−543. (226) Pidgeon, C.; Cai, S. J.; Bernal, C. Mobile Phase Effects on Membrane Protein Elution During Immobilized Artificial Membrane Chromatography. J. Chromatogr. A 1996, 721, 213−230. (227) Richter, R. P.; Maury, N.; Brisson, A. R. On the Effect of the Solid Support on the Interleaflet Distribution of Lipids in Supported Lipid Bilayers. Langmuir 2005, 21, 299−304. (228) Zhang, L.; Granick, S. Lipid Diffusion Compared in Outer and Inner Leaflets of Planar Supported Bilayers. J. Chem. Phys. 2005, 123, 211104. (229) Goennenwein, S.; Tanaka, M.; Hu, B.; Moroder, L.; Sackmann, E. Functional Incorporation of Integrins into Solid Supported Membranes on Ultrathin Films of Cellulose: Impact on Adhesion. Biophys. J. 2003, 85, 646−655. (230) Renner, L.; Pompe, T.; Lemaitre, R.; Drechsel, D.; Werner, C. Controlled Enhancement of Transmembrane Enzyme Activity in Polymer Cushioned Supported Bilayer Membranes. Soft Matter 2010, 6, 5382−5389. (231) Roder, F.; Waichman, S.; Paterok, D.; Schubert, R.; Richter, C.; Liedberg, B.; Piehler, J. Reconstitution of Membrane Proteins into

Polymer-Supported Membranes for Probing Diffusion and Interactions by Single Molecule Techniques. Anal. Chem. 2011, 83, 6792− 6799. (232) Dietrich, C.; Tampe, R. Charge Determination of Membrane Molecules in Polymer-Supported Lipid Layers. Biochim. Biophys. Acta, Biomembr. 1995, 1238, 183−191. (233) Elender, G.; Kuhner, M.; Sackmann, E. Functionalisation of Si/ Sio2 and Glass Surfaces with Ultrathin Dextran Films and Deposition of Lipid Bilayers. Biosens. Bioelectron. 1996, 11, 565−577. (234) Kuhner, M.; Tampe, R.; Sackmann, E. Lipid Mono- and Bilayer Supported on Polymer Films: Composite Polymer-Lipid Films on Solid Substrates. Biophys. J. 1994, 67, 217−226. (235) Majewski, J.; Kuhl, T. L.; Wong, J. Y.; Smith, G. S. X-Ray and Neutron Surface Scattering for Studying Lipid/Polymer Assemblies at the Air-Liquid and Solid-Liquid Interfaces. Rev. Mol. Biotechnol. 2000, 74, 207−231. (236) Munro, J. C.; Frank, C. W. In Situ Formation and Characterization of Poly(Ethylene Glycol)-Supported Lipid Bilayers on Gold Surfaces. Langmuir 2004, 20, 10567−10575. (237) Naumann, C.; Prucker, O.; Lehmann, T.; Ruhe, J.; Knoll, W.; Frank, C. W. The Polymer-Supported Phospholipid Bilayer: Tethering as a New Approach to Substrate-Membrane Stabilization. Biomacromolecules 2002, 3, 27−35. (238) Spinke, J.; Yang, J.; Wolf, H.; Liley, M.; Ringsdorf, H.; Knoll, W. Polymer-Supported Bilayer on a Solid Substrate. Biophys. J. 1992, 63, 1667−1671. (239) Tanaka, M.; Sackmann, E. Polymer-Supported Membranes as Models of the Cell Surface. Nature 2005, 437, 656−663. (240) Wagner, M. L.; Tamm, L. K. Tethered Polymer-Supported Planar Lipid Bilayers for Reconstitution of Integral Membrane Proteins: Silane-Polyethyleneglycol-Lipid as a Cushion and Covalent Linker. Biophys. J. 2000, 79, 1400−1414. (241) Wlodek, M.; Kolasinska-Sojka, M.; Szuwarzynski, M.; Zapotoczny, S.; Warszynski, P. Formation and Characterization of (1-Palmitoyl-2-Oleoyl-Sn-Glycero-3-Phosphocholine)/(1-Palmitoyl-2Oleoyl-Sn-Glycero-3-Phospho-L-Serine) Supported Lipid Bilayers on Polyelectrolyte Multilayer Films. Thin Solid Films 2015, 592, 1−7. (242) Costello, D. A.; Millet, J. K.; Hsia, C. Y.; Whittaker, G. R.; Daniel, S. Single Particle Assay of Coronavirus Membrane Fusion with Proteinaceous Receptor-Embedded Supported Bilayers. Biomaterials 2013, 34, 7895−7904. (243) Sackmann, E.; Tanaka, M. Supported Membranes on Soft Polymer Cushions: Fabrication, Characterization and Applications. Trends Biotechnol. 2000, 18, 58−64. (244) Knoll, W.; Frank, C. W.; Heibel, C.; Naumann, R.; Offenhausser, A.; Ruhe, J.; Schmidt, E. K.; Shen, W. W.; Sinner, A. Functional Tethered Lipid Bilayers. Rev. Mol. Biotechnol. 2000, 74, 137−158. (245) Lin, Y.-H.; Minner, D.; Herring, V.; Naumann, C. Physisorbed Polymer-Tethered Lipid Bilayer with Lipopolymer Gradient. Materials 2012, 5, 2243. (246) Siegel, A. P.; Murcia, M. J.; Johnson, M.; Reif, M.; Jordan, R.; Ruhe, J.; Naumann, C. A. Compartmentalizing a Lipid Bilayer by Tuning Lateral Stress in a Physisorbed Polymer-Tethered Membrane. Soft Matter 2010, 6, 2723−2732. (247) Elender, G.; Sackmann, E. Wetting and Dewetting of Si/Sio $_2$-Wafers by Free and Lipidmonolayer Covered Aqueous Solutions under Controlled Humidity. J. Phys. II 1994, 4, 455−479. (248) Kühner, M.; Sackmann, E. Ultrathin Hydrated Dextran Films Grafted on Glass: Preparation and Characterization of Structural, Viscous, and Elastic Properties by Quantitative Microinterferometry. Langmuir 1996, 12, 4866−4876. (249) McCabe, I. P.; Forstner, M. B. Polymer Supported Lipid Bilayers. Open J. Biophys. 2013, 03, 59−69. (250) Wong, J. Y.; Majewski, J.; Seitz, M.; Park, C. K.; Israelachvili, J. N.; Smith, G. S. Polymer-Cushioned Bilayers. I. A Structural Study of Various Preparation Methods Using Neutron Reflectometry. Biophys. J. 1999, 77, 1445−1457. 5471

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

(271) Tabaei, S. R.; Jonsson, P.; Branden, M.; Hook, F. SelfAssembly Formation of Multiple DNA-Tethered Lipid Bilayers. J. Struct. Biol. 2009, 168, 200−206. (272) Murray, D. H.; Tamm, L. K.; Kiessling, V. Supported Double Membranes. J. Struct. Biol. 2009, 168, 183−189. (273) Han, X.; Achalkumar, A. S.; Cheetham, M. R.; Connell, S. D.; Johnson, B. R.; Bushby, R. J.; Evans, S. D. A Self-Assembly Route for Double Bilayer Lipid Membrane Formation. ChemPhysChem 2010, 11, 569−574. (274) Heath, G. R.; Li, M.; Polignano, I. L.; Richens, J. L.; Catucci, G.; O’Shea, P.; Sadeghi, S. J.; Gilardi, G.; Butt, J. N.; Jeuken, L. J. C. Layer-by-Layer Assembly of Supported Lipid Bilayer Poly-L-Lysine Multilayers. Biomacromolecules 2016, 17, 324−335. (275) Gupta, G.; Iyer, S.; Leasure, K.; Virdone, N.; Dattelbaum, A. M.; Atanassov, P. B.; López, G. P. Stable and Fluid Multilayer Phospholipid−Silica Thin Films: Mimicking Active Multi-Lamellar Biological Assemblies. ACS Nano 2013, 7, 5300−5307. (276) Frost, R.; Jonsson, G. E.; Chakarov, D.; Svedhem, S.; Kasemo, B. Graphene Oxide and Lipid Membranes: Interactions and Nanocomposite Structures. Nano Lett. 2012, 12, 3356−3362. (277) Lenhert, S.; Brinkmann, F.; Laue, T.; Walheim, S.; Vannahme, C.; Klinkhammer, S.; Xu, M.; Sekula, S.; Mappes, T.; Schimmel, T.; et al. Lipid Multilayer Gratings. Nat. Nanotechnol. 2010, 5, 275−279. (278) Salaita, K.; Wang, Y.; Mirkin, C. A. Applications of Dip-Pen Nanolithography. Nat. Nanotechnol. 2007, 2, 145−155. (279) Nafday, O. A.; Lenhert, S. High-Throughput Optical Quality Control of Lipid Multilayers Fabricated by Dip-Pen Nanolithography. Nanotechnology 2011, 22, 225301. (280) Tayebi, L.; Ma, Y.; Vashaee, D.; Chen, G.; Sinha, S. K.; Parikh, A. N. Long-Range Interlayer Alignment of Intralayer Domains in Stacked Lipid Bilayers. Nat. Mater. 2012, 11, 1074−1080. (281) Caffrey, M. Crystallizing Membrane Proteins for Structure Determination: Use of Lipidic Mesophases. Annu. Rev. Biophys. 2009, 38, 29−51. (282) Bechinger, B. Lipid Multilayers: Domains Stack Up. Nat. Mater. 2012, 11, 1005−1006. (283) Lowry, T. W.; Hariri, H.; Prommapan, P.; Kusi-Appiah, A.; Vafai, N.; Bienkiewicz, E. A.; Van Winkle, D. H.; Stagg, S. M.; Lenhert, S. Quantification of Protein-Induced Membrane Remodeling Kinetics in Vitro with Lipid Multilayer Gratings. Small 2016, 12, 506−515. (284) Prosser, R. S.; Evanics, F.; Kitevski, J. L.; Al-Abdul-Wahid, M. S. Current Applications of Bicelles in Nmr Studies of MembraneAssociated Amphiphiles and Proteins. Biochemistry 2006, 45, 8453− 8465. (285) Holland, L. A.; Leigh, A. M. Bilayered Phospholipid Micelles and Capillary Electrophoresis: A New Additive for Electrokinetic Chromatography. Electrophoresis 2003, 24, 2935−2939. (286) Durr, U. H.; Gildenberg, M.; Ramamoorthy, A. The Magic of Bicelles Lights up Membrane Protein Structure. Chem. Rev. 2012, 112, 6054−6074. (287) Sanders, C. R., 2nd; Schwonek, J. P. Characterization of Magnetically Orientable Bilayers in Mixtures of Dihexanoylphosphatidylcholine and Dimyristoylphosphatidylcholine by Solid-State Nmr. Biochemistry 1992, 31, 8898−8905. (288) Tjandra, N.; Bax, A. Direct Measurement of Distances and Angles in Biomolecules by Nmr in a Dilute Liquid Crystalline Medium. Science 1997, 278, 1111−1114. (289) MacEachern, L.; Sylvester, A.; Flynn, A.; Rahmani, A.; Morrow, M. R. Dependence of Bicellar System Phase Behavior and Dynamics on Anionic Lipid Concentration. Langmuir 2013, 29, 3688−3699. (290) Luchette, P. A.; Vetman, T. N.; Prosser, R. S.; Hancock, R. E.; Nieh, M. P.; Glinka, C. J.; Krueger, S.; Katsaras, J. Morphology of FastTumbling Bicelles: A Small Angle Neutron Scattering and Nmr Study. Biochim. Biophys. Acta, Biomembr. 2001, 1513, 83−94. (291) Diller, A.; Loudet, C.; Aussenac, F.; Raffard, G.; Fournier, S.; Laguerre, M.; Grelard, A.; Opella, S. J.; Marassi, F. M.; Dufourc, E. J. Bicelles: A Natural ’Molecular Goniometer’ for Structural, Dynamical and Topological Studies of Molecules in Membranes. Biochimie 2009, 91, 744−751.

(251) Wong, J. Y.; Park, C. K.; Seitz, M.; Israelachvili, J. PolymerCushioned Bilayers. Ii. An Investigation of Interaction Forces and Fusion Using the Surface Forces Apparatus. Biophys. J. 1999, 77, 1458−1468. (252) Wiegand, G.; Jaworek, T.; Wegner, G.; Sackmann, E. Studies of Structure and Local Wetting Properties on Heterogeneous, Micropatterned Solid Surfaces by Microinterferometry. J. Colloid Interface Sci. 1997, 196, 299−312. (253) Lang, H.; Duschl, C.; Vogel, H. A New Class of Thiolipids for the Attachment of Lipid Bilayers on Gold Surfaces. Langmuir 1994, 10, 197−210. (254) Hertrich, S.; Stetter, F.; Ruhm, A.; Hugel, T.; Nickel, B. Highly Hydrated Deformable Polyethylene Glycol-Tethered Lipid Bilayers. Langmuir 2014, 30, 9442−9447. (255) Shen, W. W.; Boxer, S. G.; Knoll, W.; Frank, C. W. PolymerSupported Lipid Bilayers on Benzophenone-Modified Substrates. Biomacromolecules 2001, 2, 70−79. (256) Cassier, T.; Sinner, A.; Offenhäuser, A.; Möhwald, H. Homogeneity, Electrical Resistivity and Lateral Diffusion of Lipid Bilayers Coupled to Polyelectrolyte Multilayers. Colloids Surf., B 1999, 15, 215−225. (257) Heiduschka, P.; Göpel, W.; Beck, W.; Kraas, W.; Kienle, S.; Jung, G. Microstructured Peptide-Functionalised Surfaces by Electrochemical Polymerisation. Chem. - Eur. J. 1996, 2, 667−672. (258) Raedler, J.; Strey, H.; Sackmann, E. Phenomenology and Kinetics of Lipid Bilayer Spreading on Hydrophilic Surfaces. Langmuir 1995, 11, 4539−4548. (259) Stelzle, M.; Miehlich, R.; Sackmann, E. Two-Dimensional Microelectrophoresis in Supported Lipid Bilayers. Biophys. J. 1992, 63, 1346−1354. (260) Rebaud, S.; Maniti, O.; Girard-Egrot, A. P. Tethered Bilayer Lipid Membranes (Tblms): Interest and Applications for Biological Membrane Investigations. Biochimie 2014, 107, 135−142. (261) Brink, G.; Schmitt, L.; Tampe, R.; Sackmann, E. Self Assembly of Covalently Anchored Phospholipid Supported Membranes by Use of Doda-Suc-Nhs-Lipids. Biochim. Biophys. Acta, Biomembr. 1994, 1196, 227−230. (262) Naumann, R.; Jonczyk, A.; Kopp, R.; van Esch, J.; Ringsdorf, H.; Knoll, W.; Gräber, P. Incorporation of Membrane Proteins in Solid-Supported Lipid Layers. Angew. Chem., Int. Ed. Engl. 1995, 34, 2056−2058. (263) Williams, L. M.; Evans, S. D.; Flynn, T. M.; Marsh, A.; Knowles, P. F.; Bushby, R. J.; Boden, N. Kinetics of the Unrolling of Small Unilamellar Phospholipid Vesicles onto Self-Assembled Monolayers. Langmuir 1997, 13, 751−757. (264) Heyse, S.; Ernst, O. P.; Dienes, Z.; Hofmann, K. P.; Vogel, H. Incorporation of Rhodopsin in Laterally Structured Supported Membranes: Observation of Transducin Activation with Spatially and Time-Resolved Surface Plasmon Resonance. Biochemistry 1998, 37, 507−522. (265) Bieri, C.; Ernst, O. P.; Heyse, S.; Hofmann, K. P.; Vogel, H. Micropatterned Immobilization of a G Protein-Coupled Receptor and Direct Detection of G Protein Activation. Nat. Biotechnol. 1999, 17, 1105−1108. (266) Giess, F.; Friedrich, M. G.; Heberle, J.; Naumann, R. L.; Knoll, W. The Protein-Tethered Lipid Bilayer: A Novel Mimic of the Biological Membrane. Biophys. J. 2004, 87, 3213−3220. (267) Martinez, K. L.; Meyer, B. H.; Hovius, R.; Lundstrom, K.; Vogel, H. Ligand Binding to G Protein-Coupled Receptors in Tethered Cell Membranes. Langmuir 2003, 19, 10925−10929. (268) Kaizuka, Y.; Groves, J. T. Structure and Dynamics of Supported Intermembrane Junctions. Biophys. J. 2004, 86, 905−912. (269) Haas, H.; Steitz, R.; Fasano, A.; Liuzzi, G. M.; Polverini, E.; Cavatorta, P.; Riccio, P. Laminar Order within Langmuir-Blodgett Multilayers from Phospholipid and Myelin Basic Protein: A Neutron Reflectivity Study. Langmuir 2007, 23, 8491−8496. (270) Chung, M.; Lowe, R. D.; Chan, Y. H.; Ganesan, P. V.; Boxer, S. G. DNA-Tethered Membranes Formed by Giant Vesicle Rupture. J. Struct. Biol. 2009, 168, 190−199. 5472

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

(292) Park, S. H.; Prytulla, S.; De Angelis, A. A.; Brown, J. M.; Kiefer, H.; Opella, S. J. High-Resolution Nmr Spectroscopy of a Gpcr in Aligned Bicelles. J. Am. Chem. Soc. 2006, 128, 7402−7403. (293) Poget, S. F.; Cahill, S. M.; Girvin, M. E. Isotropic Bicelles Stabilize the Functional Form of a Small Multidrug-Resistance Pump for Nmr Structural Studies. J. Am. Chem. Soc. 2007, 129, 2432−2433. (294) Kolahdouzan, K.; Jackman, J. A.; Yoon, B. K.; Kim, M. C.; Johal, M. S.; Cho, N.-J. Optimizing the Formation of Supported Lipid Bilayers from Bicellar Mixtures. Langmuir 2017, 33, 5052−5064. (295) Saleem, Q.; Zhang, Z.; Petretic, A.; Gradinaru, C. C.; Macdonald, P. M. Single Lipid Bilayer Deposition on Polymer Surfaces Using Bicelles. Biomacromolecules 2015, 16, 1032−1039. (296) Lundquist, A.; Hansen, S. B.; Nordstrom, H.; Danielson, U. H.; Edwards, K. Biotinylated Lipid Bilayer Disks as Model Membranes for Biosensor Analyses. Anal. Biochem. 2010, 405, 153−159. (297) Johansson, E.; Engvall, C.; Arfvidsson, M.; Lundahl, P.; Edwards, K. Development and Initial Evaluation of Peg-Stabilized Bilayer Disks as Novel Model Membranes. Biophys. Chem. 2005, 113, 183−192. (298) Denisov, I. G.; Sligar, S. G. Nanodiscs in Membrane Biochemistry and Biophysics. Chem. Rev. 2017, 117, 4669−4713. (299) Bayburt, T. H.; Carlson, J. W.; Sligar, S. G. Reconstitution and Imaging of a Membrane Protein in a Nanometer-Size Phospholipid Bilayer. J. Struct. Biol. 1998, 123, 37−44. (300) Bayburt, T. H.; Grinkova, Y. V.; Sligar, S. G. Self-Assembly of Discoidal Phospholipid Bilayer Nanoparticles with Membrane Scaffold Proteins. Nano Lett. 2002, 2, 853−856. (301) Denisov, I. G.; Sligar, S. G. Nanodiscs for Structural and Functional Studies of Membrane Proteins. Nat. Struct. Mol. Biol. 2016, 23, 481−486. (302) Rouck, J. E.; Krapf, J. E.; Roy, J.; Huff, H. C.; Das, A. Recent Advances in Nanodisc Technology for Membrane Protein Studies (2012−2017). FEBS Lett. 2017, 591, 2057−2088. (303) Schuler, M. A.; Denisov, I. G.; Sligar, S. G. Nanodiscs as a New Tool to Examine Lipid-Protein Interactions. Methods Mol. Biol. 2013, 974, 415−433. (304) Bayburt, T. H.; Carlson, J. W.; Sligar, S. G. Single Molecule Height Measurements on a Membrane Protein in Nanometer-Scale Phospholipid Bilayer Disks. Langmuir 2000, 16, 5993−5997. (305) Bayburt, T. H.; Sligar, S. G. Single-Molecule Height Measurements on Microsomal Cytochrome P450 in NanometerScale Phospholipid Bilayer Disks. Proc. Natl. Acad. Sci. U. S. A. 2002, 99, 6725−6730. (306) Bayburt, T. H.; Sligar, S. G. Self-Assembly of Single Integral Membrane Proteins into Soluble Nanoscale Phospholipid Bilayers. Protein Sci. 2003, 12, 2476−2481. (307) Civjan, N. R.; Bayburt, T. H.; Schuler, M. A.; Sligar, S. G. Direct Solubilization of Heterologously Expressed Membrane Proteins by Incorporation into Nanoscale Lipid Bilayers. Biotechniques 2003, 35, 556−560. (308) Nasr, M. L.; Baptista, D.; Strauss, M.; Sun, Z. J.; Grigoriu, S.; Huser, S.; Pluckthun, A.; Hagn, F.; Walz, T.; Hogle, J. M.; et al. Covalently Circularized Nanodiscs for Studying Membrane Proteins and Viral Entry. Nat. Methods 2017, 14, 49−52. (309) Nath, A.; Atkins, W. M.; Sligar, S. G. Applications of Phospholipid Bilayer Nanodiscs in the Study of Membranes and Membrane Proteins. Biochemistry 2007, 46, 2059−2069. (310) Bayburt, T. H.; Sligar, S. G. Membrane Protein Assembly into Nanodiscs. FEBS Lett. 2010, 584, 1721−1727. (311) Denisov, I. G.; Sligar, S. G. Cytochromes P450 in Nanodiscs. Biochim. Biophys. Acta, Proteins Proteomics 2011, 1814, 223−229. (312) Shaw, A. W.; Pureza, V. S.; Sligar, S. G.; Morrissey, J. H. The Local Phospholipid Environment Modulates the Activation of Blood Clotting. J. Biol. Chem. 2007, 282, 6556−6563. (313) Wesolowska, O.; Michalak, K.; Maniewska, J.; Hendrich, A. B. Giant Unilamellar Vesicles - a Perfect Tool to Visualize Phase Separation and Lipid Rafts in Model Systems. Acta Biochim. Polym. 2009, 56, 33−39.

(314) Veatch, S. L.; Keller, S. L. Separation of Liquid Phases in Giant Vesicles of Ternary Mixtures of Phospholipids and Cholesterol. Biophys. J. 2003, 85, 3074−3083. (315) Feigenson, G. W. Phase Diagrams and Lipid Domains in Multicomponent Lipid Bilayer Mixtures. Biochim. Biophys. Acta, Biomembr. 2009, 1788, 47−52. (316) Pomorski, T. G.; Nylander, T.; Cardenas, M. Model Cell Membranes: Discerning Lipid and Protein Contributions in Shaping the Cell. Adv. Colloid Interface Sci. 2014, 205, 207−220. (317) Voskuhl, J.; Ravoo, B. J. Molecular Recognition of Bilayer Vesicles. Chem. Soc. Rev. 2009, 38, 495−505. (318) Tomas, S.; Milanesi, L. Mutual Modulation between Membrane-Embedded Receptor Clustering and Ligand Binding in Lipid Membranes. Nat. Chem. 2010, 2, 1077−1083. (319) Murrell, M. P.; Voituriez, R.; Joanny, J.-F.; Nassoy, P.; Sykes, C.; Gardel, M. L. Liposome Adhesion Generates Traction Stress. Nat. Phys. 2014, 10, 163−169. (320) Lemière, J.; Carvalho, K.; Sykes, C. Cell-Sized Liposomes That Mimic Cell Motility and the Cell Cortex. Methods Cell Biol. 2015, 128, 271−285. (321) Sens, P.; Plastino, J. Membrane Tension and Cytoskeleton Organization in Cell Motility. J. Phys.: Condens. Matter 2015, 27, 273103. (322) Diao, J.; Liu, R.; Rong, Y.; Zhao, M.; Zhang, J.; Lai, Y.; Zhou, Q.; Wilz, L. M.; Li, J.; Vivona, S.; et al. Atg14 Promotes Membrane Tethering and Fusion of Autophagosomes to Endolysosomes. Nature 2015, 520, 563−566. (323) Martens, S.; McMahon, H. T. Mechanisms of Membrane Fusion: Disparate Players and Common Principles. Nat. Rev. Mol. Cell Biol. 2008, 9, 543−556. (324) Ma, M.; Bong, D. Controlled Fusion of Synthetic Lipid Membrane Vesicles. Acc. Chem. Res. 2013, 46, 2988−2997. (325) Sato, Y. T.; Umezaki, K.; Sawada, S.; Mukai, S.; Sasaki, Y.; Harada, N.; Shiku, H.; Akiyoshi, K. Engineering Hybrid Exosomes by Membrane Fusion with Liposomes. Sci. Rep. 2016, 6, 21933. (326) Gyorgy, B.; Szabo, T. G.; Pasztoi, M.; Pal, Z.; Misjak, P.; Aradi, B.; Laszlo, V.; Pallinger, E.; Pap, E.; Kittel, A.; et al. Membrane Vesicles, Current State-of-the-Art: Emerging Role of Extracellular Vesicles. Cell. Mol. Life Sci. 2011, 68, 2667−2688. (327) Colombo, M.; Raposo, G.; Thery, C. Biogenesis, Secretion, and Intercellular Interactions of Exosomes and Other Extracellular Vesicles. Annu. Rev. Cell Dev. Biol. 2014, 30, 255−289. (328) Lohner, K.; Sevcsik, E.; Pabst, G. In Advances in Planar Lipid Bilayers and Liposomes; Leitmannova Liu, A., Ed.; Academic Press: Oxford, UK, 2008; Vol. 6, Chapter 5. (329) Patel, H.; Tscheka, C.; Heerklotz, H. Characterizing Vesicle Leakage by Fluorescence Lifetime Measurements. Soft Matter 2009, 5, 2849−2851. (330) Liu, Q.; Boyd, B. J. Liposomes in Biosensors. Analyst 2013, 138, 391−409. (331) Deleu, M.; Crowet, J. M.; Nasir, M. N.; Lins, L. Complementary Biophysical Tools to Investigate Lipid Specificity in the Interaction between Bioactive Molecules and the Plasma Membrane: A Review. Biochim. Biophys. Acta, Biomembr. 2014, 1838, 3171−3190. (332) Saliba, A. E.; Vonkova, I.; Gavin, A. C. The Systematic Analysis of Protein-Lipid Interactions Comes of Age. Nat. Rev. Mol. Cell Biol. 2015, 16, 753−761. (333) Cho, N.-J.; Hwang, L. Y.; Solandt, J. J. R.; Frank, C. W. Comparison of Extruded and Sonicated Vesicles for Planar Bilayer Self-Assembly. Materials 2013, 6, 3294. (334) Jesorka, A.; Orwar, O. Liposomes: Technologies and Analytical Applications. Annu. Rev. Anal. Chem. 2008, 1, 801−832. (335) Akbarzadeh, A.; Rezaei-Sadabady, R.; Davaran, S.; Joo, S. W.; Zarghami, N.; Hanifehpour, Y.; Samiei, M.; Kouhi, M.; Nejati-Koshki, K. Liposome: Classification, Preparation, and Applications. Nanoscale Res. Lett. 2013, 8, 102. 5473

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

(357) Marques, J. T.; de Almeida, R. F. M.; Viana, A. S. Biomimetic Membrane Rafts Stably Supported on Unmodified Gold. Soft Matter 2012, 8, 2007−2016. (358) Wang, X.; Shindel, M. M.; Wang, S. W.; Ragan, R. A Facile Approach for Assembling Lipid Bilayer Membranes on TemplateStripped Gold. Langmuir 2010, 26, 18239−18245. (359) Benes, M.; Billy, D.; Benda, A.; Speijer, H.; Hof, M.; Hermens, W. T. Surface-Dependent Transitions During Self-Assembly of Phospholipid Membranes on Mica, Silica, and Glass. Langmuir 2004, 20, 10129−10137. (360) Love, J. C.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R. G.; Whitesides, G. M. Self-Assembled Monolayers of Thiolates on Metals as a Form of Nanotechnology. Chem. Rev. 2005, 105, 1103−1169. (361) Ulman, A. Formation and Structure of Self-Assembled Monolayers. Chem. Rev. 1996, 96, 1533−1554. (362) Nuzzo, R. G.; Dubois, L. H.; Allara, D. L. Fundamental Studies of Microscopic Wetting on Organic Surfaces. 1. Formation and Structural Characterization of a Self-Consistent Series of Polyfunctional Organic Monolayers. J. Am. Chem. Soc. 1990, 112, 558−569. (363) Haensch, C.; Hoeppener, S.; Schubert, U. S. Chemical Modification of Self-Assembled Silane Based Monolayers by Surface Reactions. Chem. Soc. Rev. 2010, 39, 2323−2334. (364) Hermanson, G. T. In Bioconjugate Techniques, 3rd ed.; Academic Press: Boston, 2013. (365) Cha, T.; Guo, A.; Zhu, X. Y. Formation of Supported Phospholipid Bilayers on Molecular Surfaces: Role of Surface Charge Density and Electrostatic Interaction. Biophys. J. 2006, 90, 1270−1274. (366) Kung, V. T.; Redemann, C. T. Synthesis of Carboxyacyl Derivatives of Phosphatidylethanolamine and Use as an Efficient Method for Conjugation of Protein to Liposomes. Biochim. Biophys. Acta, Biomembr. 1986, 862, 435−439. (367) Maruyama, K.; Takizawa, T.; Yuda, T.; Kennel, S. J.; Huang, L.; Iwatsuru, M. Targetability of Novel Immunoliposomes Modified with Amphipathic Poly(Ethylene Glycol)S Conjugated at Their Distal Terminals to Monoclonal Antibodies. Biochim. Biophys. Acta, Biomembr. 1995, 1234, 74−80. (368) Ogawa, Y.; Kawahara, H.; Yagi, N.; Kodaka, M.; Tomohiro, T.; Okada, T.; Konakahara, T.; Okuno, H. Synthesis of a Novel Lipopeptide with Alpha-Melanocyte-Stimulating Hormone Peptide Ligand and Its Effect on Liposome Stability. Lipids 1999, 34, 387−394. (369) Yagi, N.; Ogawa, Y.; Kodaka, M.; Okada, T.; Tomohiro, T.; Konakahara, T.; Okuno, H. Preparation of Functional Liposomes with Peptide Ligands and Their Binding to Cell Membranes. Lipids 2000, 35, 673−680. (370) Hermanson, G. T. In Bioconjugate Techniques, 3rd ed.; Academic Press: Boston, 2013. (371) Hermanson, G. T. In Bioconjugate Techniques, 3rd ed.; Academic Press: Boston, 2013. (372) Reimhult, E.; Hook, F. Design of Surface Modifications for Nanoscale Sensor Applications. Sensors 2015, 15, 1635−1675. (373) Reimhult, E.; Baumann, M.; Kaufmann, S.; Kumar, K.; Spycher, P. Advances in Nanopatterned and Nanostructured Supported Lipid Membranes and Their Applications. Biotechnol. Genet. Eng. Rev. 2010, 27, 185−216. (374) Silva-Lopez, E. I.; Edens, L. E.; Barden, A. O.; Keller, D. J.; Brozik, J. A. Conditions for Liposome Adsorption and Bilayer Formation on Bsa Passivated Solid Supports. Chem. Phys. Lipids 2014, 183, 91−99. (375) Knoll, W.; Köper, I.; Naumann, R.; Sinner, E.-K. Tethered Bimolecular Lipid Membranesa Novel Model Membrane Platform. Electrochim. Acta 2008, 53, 6680−6689. (376) Knoll, W.; Bender, K.; Förch, R.; Frank, C.; Götz, H.; Heibel, C.; Jenkins, T.; Jonas, U.; Kibrom, A.; Kügler, R. In Polymer Membranes/Biomembranes; Meier, W. P., Knoll, W., Eds.; Springer: Berlin, 2010. (377) Zoppe, J. O.; Ataman, N. C.; Mocny, P.; Wang, J.; Moraes, J.; Klok, H. A. Surface-Initiated Controlled Radical Polymerization: Stateof-the-Art, Opportunities, and Challenges in Surface and Interface Engineering with Polymer Brushes. Chem. Rev. 2017, 117, 1105−1318.

(336) Maherani, B.; Arab-Tehrany, E.; Mozafari, M. R.; Gaiani, C.; Linder, M. Liposomes: A Review of Manufacturing Techniques and Targeting Strategies. Curr. Nanosci. 2011, 7, 436−452. (337) Christensen, A. L.; Lohr, C.; Christensen, S. M.; Stamou, D. Single Vesicle Biochips for Ultra-Miniaturized Nanoscale Fluidics and Single Molecule Bioscience. Lab Chip 2013, 13, 3613−3625. (338) Torchilin, V. P. Recent Advances with Liposomes as Pharmaceutical Carriers. Nat. Rev. Drug Discovery 2005, 4, 145−160. (339) Ingvarsson, P. T.; Yang, M.; Nielsen, H. M.; Rantanen, J.; Foged, C. Stabilization of Liposomes During Drying. Expert Opin. Drug Delivery 2011, 8, 375−388. (340) Immordino, M. L.; Dosio, F.; Cattel, L. Stealth Liposomes: Review of the Basic Science, Rationale, and Clinical Applications, Existing and Potential. Int. J. Nanomedicine 2006, 1, 297−315. (341) Marquês, J. T.; Viana, A. S.; de Almeida, R. F. M. A Biomimetic Platform to Study the Interactions of Bioelectroactive Molecules with Lipid Nanodomains. Langmuir 2014, 30, 12627−12637. (342) Keller, C. A.; Kasemo, B. Surface Specific Kinetics of Lipid Vesicle Adsorption Measured with a Quartz Crystal Microbalance. Biophys. J. 1998, 75, 1397−1402. (343) Pfeiffer, I.; Petronis, S.; Koper, I.; Kasemo, B.; Zach, M. Vesicle Adsorption and Phospholipid Bilayer Formation on Topographically and Chemically Nanostructured Surfaces. J. Phys. Chem. B 2010, 114, 4623−4631. (344) Ekeroth, J.; Konradsson, P.; Höök, F. Bivalent-Ion-Mediated Vesicle Adsorption and Controlled Supported Phospholipid Bilayer Formation on Molecular Phosphate and Sulfate Layers on Gold. Langmuir 2002, 18, 7923−7929. (345) Richter, R.; Mukhopadhyay, A.; Brisson, A. Pathways of Lipid Vesicle Deposition on Solid Surfaces: A Combined Qcm-D and Afm Study. Biophys. J. 2003, 85, 3035−3047. (346) Nollert, P.; Kiefer, H.; Jahnig, F. Lipid Vesicle Adsorption Versus Formation of Planar Bilayers on Solid Surfaces. Biophys. J. 1995, 69, 1447−1455. (347) Cho, N. J.; Frank, C. W.; Kasemo, B.; Hook, F. Quartz Crystal Microbalance with Dissipation Monitoring of Supported Lipid Bilayers on Various Substrates. Nat. Protoc. 2010, 5, 1096−1106. (348) Ma, Y.; Sobkiv, I.; Gruzdys, V.; Zhang, H.; Sun, X. L. Liposomal Glyco-Microarray for Studying Glycolipid-Protein Interactions. Anal. Bioanal. Chem. 2012, 404, 51−58. (349) Ma, Y.; Zhang, H.; Gruzdys, V.; Sun, X. L. Azide-Reactive Liposome for Chemoselective and Biocompatible Liposomal Surface Functionalization and Glyco-Liposomal Microarray Fabrication. Langmuir 2011, 27, 13097−13103. (350) van Lengerich, B.; Rawle, R. J.; Boxer, S. G. Covalent Attachment of Lipid Vesicles to a Fluid-Supported Bilayer Allows Observation of DNA-Mediated Vesicle Interactions. Langmuir 2010, 26, 8666−8672. (351) Bally, M.; Bailey, K.; Sugihara, K.; Grieshaber, D.; Voros, J.; Stadler, B. Liposome and Lipid Bilayer Arrays Towards Biosensing Applications. Small 2010, 6, 2481−2497. (352) Vermette, P.; Griesser, H. J.; Kambouris, P.; Meagher, L. Characterization of Surface-Immobilized Layers of Intact Liposomes. Biomacromolecules 2004, 5, 1496−1502. (353) Stadler, B.; Bally, M.; Grieshaber, D.; Voros, J.; Brisson, A.; Grandin, H. M. Creation of a Functional Heterogeneous Vesicle Array Via DNA Controlled Surface Sorting onto a Spotted Microarray. Biointerphases 2006, 1, 142−145. (354) Klenkar, G.; Brian, B.; Ederth, T.; Stengel, G.; Hook, F.; Piehler, J.; Liedberg, B. Addressable Adsorption of Lipid Vesicles and Subsequent Protein Interaction Studies. Biointerphases 2008, 3, 29. (355) Roling, O.; Wendeln, C.; Kauscher, U.; Seelheim, P.; Galla, H.J.; Ravoo, B. J. Layer-by-Layer Deposition of Vesicles Mediated by Supramolecular Interactions. Langmuir 2013, 29, 10174−10182. (356) Losey, E. A.; Smith, M. D.; Meng, M.; Best, M. D. MicroplateBased Analysis of Protein−Membrane Binding Interactions Via Immobilization of Whole Liposomes Containing a Biotinylated Anchor. Bioconjugate Chem. 2009, 20, 376−383. 5474

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

(398) Torchilin, V. P.; Khaw, B. A.; Smirnov, V. N.; Haber, E. Preservation of Antimyosin Antibody Activity after Covalent Coupling to Liposomes. Biochem. Biophys. Res. Commun. 1979, 89, 1114−1119. (399) Martin, F. J.; Hubbell, W. L.; Papahadjopoulos, D. Immunospecific Targeting of Liposomes to Cells: A Novel and Efficient Method for Covalent Attachment of Fab’ Fragments Via Disulfide Bonds. Biochemistry 1981, 20, 4229−4238. (400) Leserman, L. D.; Barbet, J.; Kourilsky, F.; Weinstein, J. N. Targeting to Cells of Fluorescent Liposomes Covalently Coupled with Monoclonal Antibody or Protein A. Nature 1980, 288, 602−604. (401) Allen, T. M.; Brandeis, E.; Hansen, C. B.; Kao, G. Y.; Zalipsky, S. A New Strategy for Attachment of Antibodies to Sterically Stabilized Liposomes Resulting in Efficient Targeting to Cancer Cells. Biochim. Biophys. Acta, Biomembr. 1995, 1237, 99−108. (402) Mercadal, M.; Domingo, J. C.; Petriz, J.; Garcia, J.; de Madariaga, M. A. A Novel Strategy Affords High-Yield Coupling of Antibody to Extremities of Liposomal Surface-Grafted Peg Chains. Biochim. Biophys. Acta, Biomembr. 1999, 1418, 232−238. (403) Frisch, B.; Boeckler, C.; Schuber, F. Synthesis of Short Polyoxyethylene-Based Heterobifunctional Cross-Linking Reagents. Application to the Coupling of Peptides to Liposomes. Bioconjugate Chem. 1996, 7, 180−186. (404) Schelte, P.; Boeckler, C.; Frisch, B.; Schuber, F. Differential Reactivity of Maleimide and Bromoacetyl Functions with Thiols: Application to the Preparation of Liposomal Diepitope Constructs. Bioconjugate Chem. 2000, 11, 118−123. (405) Beduneau, A.; Saulnier, P.; Hindre, F.; Clavreul, A.; Leroux, J. C.; Benoit, J. P. Design of Targeted Lipid Nanocapsules by Conjugation of Whole Antibodies and Antibody Fab’ Fragments. Biomaterials 2007, 28, 4978−4990. (406) Crosasso, P.; Brusa, P.; Dosio, F.; Arpicco, S.; Pacchioni, D.; Schuberî, F.; Cattel, L. Antitumoral Activity of Liposomes and Immunoliposomes Containing 5-Fluorouridine Prodrugs. J. Pharm. Sci. 1997, 86, 832−839. (407) Fleiner, M.; Benzinger, P.; Fichert, T.; Massing, U. Studies on Protein-Liposome Coupling Using Novel Thiol-Reactive Coupling Lipids: Influence of Spacer Length and Polarity. Bioconjugate Chem. 2001, 12, 470−475. (408) Garnier, B.; Bouter, A.; Gounou, C.; Petry, K. G.; Brisson, A. R. Annexin A5-Functionalized Liposomes for Targeting Phosphatidylserine-Exposing Membranes. Bioconjugate Chem. 2009, 20, 2114−2122. (409) Janssen, A. P.; Schiffelers, R. M.; ten Hagen, T. L.; Koning, G. A.; Schraa, A. J.; Kok, R. J.; Storm, G.; Molema, G. Peptide-Targeted Peg-Liposomes in Anti-Angiogenic Therapy. Int. J. Pharm. 2003, 254, 55−58. (410) Park, J. W.; Kirpotin, D. B.; Hong, K.; Shalaby, R.; Shao, Y.; Nielsen, U. B.; Marks, J. D.; Papahadjopoulos, D.; Benz, C. C. Tumor Targeting Using Anti-Her2 Immunoliposomes. J. Controlled Release 2001, 74, 95−113. (411) Torchilin, V. P.; Levchenko, T. S.; Lukyanov, A. N.; Khaw, B. A.; Klibanov, A. L.; Rammohan, R.; Samokhin, G. P.; Whiteman, K. R. P-Nitrophenylcarbonyl-Peg-Pe-Liposomes: Fast and Simple Attachment of Specific Ligands, Including Monoclonal Antibodies, to Distal Ends of Peg Chains Via P-Nitrophenylcarbonyl Groups. Biochim. Biophys. Acta, Biomembr. 2001, 1511, 397−411. (412) Torchilin, V. P.; Rammohan, R.; Weissig, V.; Levchenko, T. S. Tat Peptide on the Surface of Liposomes Affords Their Efficient Intracellular Delivery Even at Low Temperature and in the Presence of Metabolic Inhibitors. Proc. Natl. Acad. Sci. U. S. A. 2001, 98, 8786− 8791. (413) Cavalli, S.; Tipton, A. R.; Overhand, M.; Kros, A. The Chemical Modification of Liposome Surfaces Via a Copper-Mediated [3 + 2] Azide-Alkyne Cycloaddition Monitored by a Colorimetric Assay. Chem. Commun. 2006, 3193−3195. (414) Frisch, B.; Hassane, F. S.; Schuber, F. Conjugation of Ligands to the Surface of Preformed Liposomes by Click Chemistry. Methods Mol. Biol. 2010, 605, 267−277. (415) Meldal, M.; Torno̷ e, C. W. Cu-Catalyzed Azide-Alkyne Cycloaddition. Chem. Rev. 2008, 108, 2952−3015.

(378) Barbey, R.; Lavanant, L.; Paripovic, D.; Schuwer, N.; Sugnaux, C.; Tugulu, S.; Klok, H. A. Polymer Brushes Via Surface-Initiated Controlled Radical Polymerization: Synthesis, Characterization, Properties, and Applications. Chem. Rev. 2009, 109, 5437−5527. (379) Prucker, O.; Rühe, J. Mechanism of Radical Chain Polymerizations Initiated by Azo Compounds Covalently Bound to the Surface of Spherical Particles. Macromolecules 1998, 31, 602−613. (380) Jordan, R.; West, N.; Ulman, A.; Chou, Y.-M.; Nuyken, O. Nanocomposites by Surface-Initiated Living Cationic Polymerization of 2-Oxazolines on Functionalized Gold Nanoparticles. Macromolecules 2001, 34, 1606−1611. (381) Ejaz, M.; Tsujii, Y.; Fukuda, T. Controlled Grafting of a WellDefined Polymer on a Porous Glass Filter by Surface-Initiated Atom Transfer Radical Polymerization. Polymer 2001, 42, 6811−6815. (382) Matyjaszewski, K.; Xia, J. Atom Transfer Radical Polymerization. Chem. Rev. 2001, 101, 2921−2990. (383) Ye, Q.; Konradi, R.; Textor, M.; Reimhult, E. Liposomes Tethered to Omega-Functional Peg Brushes and Induced Formation of Peg Brush Supported Planar Lipid Bilayers. Langmuir 2009, 25, 13534−13539. (384) Dowling, M. B.; Javvaji, V.; Payne, G. F.; Raghavan, S. R. Vesicle Capture on Patterned Surfaces Coated with Amphiphilic Biopolymers. Soft Matter 2011, 7, 1219−1226. (385) Cooper, M. A.; Hansson, A.; Lofas, S.; Williams, D. H. A Vesicle Capture Sensor Chip for Kinetic Analysis of Interactions with Membrane-Bound Receptors. Anal. Biochem. 2000, 277, 196−205. (386) Heuberger, M.; Drobek, T.; Spencer, N. D. Interaction Forces and Morphology of a Protein-Resistant Poly(Ethylene Glycol) Layer. Biophys. J. 2005, 88, 495−504. (387) Drobek, T.; Spencer, N. D.; Heuberger, M. Compressing Peg Brushes. Macromolecules 2005, 38, 5254−5259. (388) Pasche, S.; Textor, M.; Meagher, L.; Spencer, N. D.; Griesser, H. J. Relationship between Interfacial Forces Measured by ColloidProbe Atomic Force Microscopy and Protein Resistance of Poly(Ethylene Glycol)-Grafted Poly(L-Lysine) Adlayers on Niobia Surfaces. Langmuir 2005, 21, 6508−6520. (389) Zheng, R.; Arora, J.; Boonkaew, B.; Raghavan, S. R.; Kaplan, D. L.; He, J.; Pesika, N. S.; John, V. T. Liposomes Tethered to a Biopolymer Film through the Hydrophobic Effect Create a Highly Effective Lubricating Surface. Soft Matter 2014, 10, 9226−9229. (390) Peel, M. J.; Cross, S. J.; Birkholz, O.; Aladag, A.; Piehler, J.; Peel, S. Rupture of Stochastically Occurring Vesicle Clusters Limits Bilayer Formation on Alkane-Peg-Type Supports: Uncoupling Clustering from Surface Coverage. Langmuir 2015, 31, 8830−8840. (391) Banuls, M. J.; Puchades, R.; Maquieira, A. Chemical Surface Modifications for the Development of Silicon-Based Label-Free Integrated Optical (Io) Biosensors: A Review. Anal. Chim. Acta 2013, 777, 1−16. (392) Jølck, R. I.; Feldborg, L. N.; Andersen, S.; Moghimi, S. M.; Andresen, T. L. In Biofunctionalization of Polymers and Their Applications; Nyanhongo, G. S., Steiner, W., Gübitz, G., Eds.; Springer: Berlin, 2011. (393) Ariga, K.; Nakanishi, T.; Michinobu, T. Immobilization of Biomaterials to Nano-Assembled Films (Self-Assembled Monolayers, Langmuir-Blodgett Films, and Layer-by-Layer Assemblies) and Their Related Functions. J. Nanosci. Nanotechnol. 2006, 6, 2278−2301. (394) Wang, C. H.; Pun, S. H. Substrate-Mediated Nucleic Acid Delivery from Self-Assembled Monolayers. Trends Biotechnol. 2011, 29, 119−126. (395) Marqués-Gallego, P.; de Kroon, A. I. P. M. Ligation Strategies for Targeting Liposomal Nanocarriers. BioMed Res. Int. 2014, 2014, 1. (396) Früh, V.; Ijzerman, A. P.; Siegal, G. How to Catch a Membrane Protein in Action: A Review of Functional Membrane Protein Immobilization Strategies and Their Applications. Chem. Rev. 2011, 111, 640−656. (397) Torchilin, V. P.; Goldmacher, V. S.; Smirnov, V. N. Comparative Studies on Covalent and Noncovalent Immobilization of Protein Molecules on the Surface of Liposomes. Biochem. Biophys. Res. Commun. 1978, 85, 983−990. 5475

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

in Vivo Bioorthogonal Ligation with Trans-Cyclooctenes. J. Am. Chem. Soc. 2012, 134, 2898−2901. (436) Daumar, P.; Wanger-Baumann, C. A.; Pillarsetty, N.; Fabrizio, L.; Carlin, S. D.; Andreev, O. A.; Reshetnyak, Y. K.; Lewis, J. S. Efficient (18)F-Labeling of Large 37-Amino-Acid Phlip Peptide Analogues and Their Biological Evaluation. Bioconjugate Chem. 2012, 23, 1557−1566. (437) Andreev, O. A.; Engelman, D. M.; Reshetnyak, Y. K. PhSensitive Membrane Peptides (Phlips) as a Novel Class of Delivery Agents. Mol. Membr. Biol. 2010, 27, 341−352. (438) Antos, J. M.; Truttmann, M. C.; Ploegh, H. L. Recent Advances in Sortase-Catalyzed Ligation Methodology. Curr. Opin. Struct. Biol. 2016, 38, 111−118. (439) Guo, X.; Wu, Z.; Guo, Z. New Method for Site-Specific Modification of Liposomes with Proteins Using Sortase a-Mediated Transpeptidation. Bioconjugate Chem. 2012, 23, 650−655. (440) Schmohl, L.; Schwarzer, D. Sortase-Mediated Ligations for the Site-Specific Modification of Proteins. Curr. Opin. Chem. Biol. 2014, 22, 122−128. (441) Ilangovan, U.; Ton-That, H.; Iwahara, J.; Schneewind, O.; Clubb, R. T. Structure of Sortase, the Transpeptidase That Anchors Proteins to the Cell Wall of Staphylococcus Aureus. Proc. Natl. Acad. Sci. U. S. A. 2001, 98, 6056−6061. (442) Ton-That, H.; Mazmanian, S. K.; Faull, K. F.; Schneewind, O. Anchoring of Surface Proteins to the Cell Wall of Staphylococcus Aureus. Sortase Catalyzed in Vitro Transpeptidation Reaction Using Lpxtg Peptide and Nh(2)-Gly(3) Substrates. J. Biol. Chem. 2000, 275, 9876−9881. (443) Tabata, A.; Ohkubo, Y.; Anyoji, N.; Hojo, K.; Tomoyasu, T.; Tatematsu, Y.; Ohkura, K.; Nagamune, H. Development of a Sortase aMediated Peptide-Labeled Liposome Applicable to Drug-Delivery Systems. Anticancer Res. 2015, 35, 4411−4417. (444) van Vught, R.; Pieters, R. J.; Breukink, E. Site-Specific Functionalization of Proteins and Their Applications to Therapeutic Antibodies. Comput. Struct. Biotechnol. J. 2014, 9, e201402001. (445) Tabata, A.; Anyoji, N.; Ohkubo, Y.; Tomoyasu, T.; Nagamune, H. Investigation on the Reaction Conditions of Staphylococcus Aureus Sortase a for Creating Surface-Modified Liposomes as a Drug-Delivery System Tool. Anticancer Res. 2014, 34, 4521−4527. (446) Silvius, J. R.; Leventis, R. A Novel “Prebinding” Strategy Dramatically Enhances Sortase-Mediated Coupling of Proteins to Liposomes. Bioconjugate Chem. 2017, 28, 1271−1282. (447) Resh, M. D. Trafficking and Signaling by Fatty-Acylated and Prenylated Proteins. Nat. Chem. Biol. 2006, 2, 584−590. (448) Casey, P. J.; Seabra, M. C. Protein Prenyltransferases. J. Biol. Chem. 1996, 271, 5289−5292. (449) Nguyen, U. T.; Guo, Z.; Delon, C.; Wu, Y.; Deraeve, C.; Franzel, B.; Bon, R. S.; Blankenfeldt, W.; Goody, R. S.; Waldmann, H.; et al. Analysis of the Eukaryotic Prenylome by Isoprenoid Affinity Tagging. Nat. Chem. Biol. 2009, 5, 227−235. (450) Hougland, J. L.; Fierke, C. A. Getting a Handle on Protein Prenylation. Nat. Chem. Biol. 2009, 5, 197−198. (451) Nadolski, M. J.; Linder, M. E. Protein Lipidation. FEBS J. 2007, 274, 5202−5210. (452) Weinrich, D.; Lin, P. C.; Jonkheijm, P.; Nguyen, U. T.; Schroder, H.; Niemeyer, C. M.; Alexandrov, K.; Goody, R.; Waldmann, H. Oriented Immobilization of Farnesylated Proteins by the Thiol-Ene Reaction. Angew. Chem., Int. Ed. 2010, 49, 1252−1257. (453) Boutin, J. A. Myristoylation. Cell. Signalling 1997, 9, 15−35. (454) Yamauchi, S.; Fusada, N.; Hayashi, H.; Utsumi, T.; Uozumi, N.; Endo, Y.; Tozawa, Y. The Consensus Motif for N-Myristoylation of Plant Proteins in a Wheat Germ Cell-Free Translation System. FEBS J. 2010, 277, 3596−3607. (455) Heal, W. P.; Wright, M. H.; Thinon, E.; Tate, E. W. Multifunctional Protein Labeling Via Enzymatic N-Terminal Tagging and Elaboration by Click Chemistry. Nat. Protoc. 2011, 7, 105−117. (456) Pakalns, T.; Haverstick, K. L.; Fields, G. B.; McCarthy, J. B.; Mooradian, D. L.; Tirrell, M. Cellular Recognition of Synthetic

(416) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B. A Stepwise Huisgen Cycloaddition Process: Copper(I)-Catalyzed Regioselective ″Ligation″ of Azides and Terminal Alkynes. Angew. Chem., Int. Ed. 2002, 41, 2596−2599. (417) Said Hassane, F.; Frisch, B.; Schuber, F. Targeted Liposomes: Convenient Coupling of Ligands to Preformed Vesicles Using ″Click Chemistry″. Bioconjugate Chem. 2006, 17, 849−854. (418) Torno̷ e, C. W.; Christensen, C.; Meldal, M. Peptidotriazoles on Solid Phase: [1,2,3]-Triazoles by Regiospecific Copper(I)Catalyzed 1,3-Dipolar Cycloadditions of Terminal Alkynes to Azides. J. Org. Chem. 2002, 67, 3057−3064. (419) Accardo, A.; Ringhieri, P.; Tesauro, D.; Morelli, G. Liposomes Derivatized with Tetrabranched Neurotensin Peptides Via Click Chemistry Reactions. New J. Chem. 2013, 37, 3528−3534. (420) Tarallo, R.; Accardo, A.; Falanga, A.; Guarnieri, D.; Vitiello, G.; Netti, P.; D’Errico, G.; Morelli, G.; Galdiero, S. Clickable Functionalization of Liposomes with the Gh625 Peptide from Herpes Simplex Virus Type I for Intracellular Drug Delivery. Chem. - Eur. J. 2011, 17, 12659−12668. (421) Bak, M.; Jolck, R. I.; Eliasen, R.; Andresen, T. L. Affinity Induced Surface Functionalization of Liposomes Using Cu-Free Click Chemistry. Bioconjugate Chem. 2016, 27, 1673−1680. (422) Jewett, J. C.; Bertozzi, C. R. Cu-Free Click Cycloaddition Reactions in Chemical Biology. Chem. Soc. Rev. 2010, 39, 1272−1279. (423) Koo, H.; Lee, S.; Na, J. H.; Kim, S. H.; Hahn, S. K.; Choi, K.; Kwon, I. C.; Jeong, S. Y.; Kim, K. Bioorthogonal Copper-Free Click Chemistry in Vivo for Tumor-Targeted Delivery of Nanoparticles. Angew. Chem., Int. Ed. 2012, 51, 11836−11840. (424) Mahal, L. K.; Yarema, K. J.; Bertozzi, C. R. Engineering Chemical Reactivity on Cell Surfaces through Oligosaccharide Biosynthesis. Science 1997, 276, 1125−1128. (425) Oude Blenke, E.; Klaasse, G.; Merten, H.; Pluckthun, A.; Mastrobattista, E.; Martin, N. I. Liposome Functionalization with Copper-Free ″Click Chemistry″. J. Controlled Release 2015, 202, 14− 20. (426) Ma, Y.; Zhang, H.; Sun, X. L. Surface-Bound Cytomimetic Assembly Based on Chemoselective and Biocompatible Immobilization and Further Modification of Intact Liposome. Bioconjugate Chem. 2010, 21, 1994−1999. (427) Saxon, E.; Bertozzi, C. R. Cell Surface Engineering by a Modified Staudinger Reaction. Science 2000, 287, 2007−2010. (428) Vabbilisetty, P.; Sun, X. L. Liposome Surface Functionalization Based on Different Anchoring Lipids Via Staudinger Ligation. Org. Biomol. Chem. 2014, 12, 1237−1244. (429) Zhang, H.; Ma, Y.; Sun, X. L. Chemically-Selective Surface Glyco-Functionalization of Liposomes through Staudinger Ligation. Chem. Commun. 2009, 3032−3034. (430) Zhang, H.; Ma, Y.; Sun, X. L. Chemically Selective Liposome Surface Glyco-Functionalization. Methods Mol. Biol. 2011, 751, 269− 280. (431) Sun, X. L.; Grande, D.; Baskaran, S.; Hanson, S. R.; Chaikof, E. L. Glycosaminoglycan Mimetic Biomaterials. 4. Synthesis of Sulfated Lactose-Based Glycopolymers That Exhibit Anticoagulant Activity. Biomacromolecules 2002, 3, 1065−1070. (432) Blackman, M. L.; Royzen, M.; Fox, J. M. Tetrazine Ligation: Fast Bioconjugation Based on Inverse-Electron-Demand Diels-Alder Reactivity. J. Am. Chem. Soc. 2008, 130, 13518−13519. (433) Emmetiere, F.; Irwin, C.; Viola-Villegas, N. T.; Longo, V.; Cheal, S. M.; Zanzonico, P.; Pillarsetty, N.; Weber, W. A.; Lewis, J. S.; Reiner, T. (18)F-Labeled-Bioorthogonal Liposomes for in Vivo Targeting. Bioconjugate Chem. 2013, 24, 1784−1789. (434) Liu, D. S.; Tangpeerachaikul, A.; Selvaraj, R.; Taylor, M. T.; Fox, J. M.; Ting, A. Y. Diels-Alder Cycloaddition for Fluorophore Targeting to Specific Proteins inside Living Cells. J. Am. Chem. Soc. 2012, 134, 792−795. (435) Seitchik, J. L.; Peeler, J. C.; Taylor, M. T.; Blackman, M. L.; Rhoads, T. W.; Cooley, R. B.; Refakis, C.; Fox, J. M.; Mehl, R. A. Genetically Encoded Tetrazine Amino Acid Directs Rapid Site-Specific 5476

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Peptide Amphiphiles in Self-Assembled Monolayer Films. Biomaterials 1999, 20, 2265−2279. (457) Englund, P. T. The Structure and Biosynthesis of Glycosyl Phosphatidylinositol Protein Anchors. Annu. Rev. Biochem. 1993, 62, 121−138. (458) Ferguson, M. A.; Williams, A. F. Cell-Surface Anchoring of Proteins Via Glycosyl-Phosphatidylinositol Structures. Annu. Rev. Biochem. 1988, 57, 285−320. (459) Guo, Z. Synthetic Studies of Glycosylphosphatidylinositol (Gpi) Anchors and Gpi-Anchored Peptides, Glycopeptides, and Proteins. Curr. Org. Synth. 2013, 10, 366−383. (460) Heider, S.; Dangerfield, J. A.; Metzner, C. Biomedical Applications of Glycosylphosphatidylinositol-Anchored Proteins. J. Lipid Res. 2016, 57, 1778−1788. (461) Tsai, Y. H.; Liu, X.; Seeberger, P. H. Chemical Biology of Glycosylphosphatidylinositol Anchors. Angew. Chem., Int. Ed. 2012, 51, 11438−11456. (462) Ronzon, F.; Morandat, S.; Roux, B.; Bortolato, M. Insertion of a Glycosylphosphatidylinositol-Anchored Enzyme into Liposomes. J. Membr. Biol. 2004, 197, 169−177. (463) Milhiet, P.-E.; Giocondi, M.-C.; Baghdadi, O.; Ronzon, F.; Le Grimellec, C.; Roux, B. Afm Detection of Gpi Protein Insertion into Dopc/Dppc Model Membranes. Single Mol. 2002, 3, 135−140. (464) Beales, P. A.; Vanderlick, T. K. Specific Binding of Different Vesicle Populations by the Hybridization of Membrane-Anchored DNA. J. Phys. Chem. A 2007, 111, 12372−12380. (465) Beales, P. A.; Vanderlick, T. K. DNA as Membrane-Bound Ligand-Receptor Pairs: Duplex Stability Is Tuned by Intermembrane Forces. Biophys. J. 2009, 96, 1554−1565. (466) Beales, P. A.; Vanderlick, T. K. Application of Nucleic AcidLipid Conjugates for the Programmable Organisation of Liposomal Modules. Adv. Colloid Interface Sci. 2014, 207, 290−305. (467) van der Meulen, S. A.; Dubacheva, G. V.; Dogterom, M.; Richter, R. P.; Leunissen, M. E. Quartz Crystal Microbalance with Dissipation Monitoring and Spectroscopic Ellipsometry Measurements of the Phospholipid Bilayer Anchoring Stability and Kinetics of Hydrophobically Modified DNA Oligonucleotides. Langmuir 2014, 30, 6525−6533. (468) Stengel, G.; Simonsson, L.; Campbell, R. A.; Höök, F. Determinants for Membrane Fusion Induced by Cholesterol-Modified DNA Zippers. J. Phys. Chem. B 2008, 112, 8264−8274. (469) Beales, P. A.; Geerts, N.; Inampudi, K. K.; Shigematsu, H.; Wilson, C. J.; Vanderlick, T. K. Reversible Assembly of Stacked Membrane Nanodiscs with Reduced Dimensionality and Variable Periodicity. J. Am. Chem. Soc. 2013, 135, 3335−3338. (470) Chan, Y. H.; van Lengerich, B.; Boxer, S. G. Lipid-Anchored DNA Mediates Vesicle Fusion as Observed by Lipid and Content Mixing. Biointerphases 2008, 3, FA17. (471) Pfeiffer, I.; Höök, F. Bivalent Cholesterol-Based Coupling of Oligonucletides to Lipid Membrane Assemblies. J. Am. Chem. Soc. 2004, 126, 10224−10225. (472) Schade, M.; Berti, D.; Huster, D.; Herrmann, A.; Arbuzova, A. Lipophilic Nucleic Acids–a Flexible Construction Kit for Organization and Functionalization of Surfaces. Adv. Colloid Interface Sci. 2014, 208, 235−251. (473) Capone, R.; Blake, S.; Rincon Restrepo, M.; Yang, J.; Mayer, M. Designing Nanosensors Based on Charged Derivatives of Gramicidin A. J. Am. Chem. Soc. 2007, 129, 9737−9745. (474) Delbianco, M.; Bharate, P.; Varela-Aramburu, S.; Seeberger, P. H. Carbohydrates in Supramolecular Chemistry. Chem. Rev. 2016, 116, 1693−1752. (475) Wang, C.; Li, Z.; Cao, D.; Zhao, Y. L.; Gaines, J. W.; Bozdemir, O. A.; Ambrogio, M. W.; Frasconi, M.; Botros, Y. Y.; Zink, J. I.; et al. Stimulated Release of Size-Selected Cargos in Succession from Mesoporous Silica Nanoparticles. Angew. Chem., Int. Ed. 2012, 51, 5460−5465. (476) Casas-Solvas, J. M.; Ortiz-Salmeron, E.; Fernandez, I.; GarciaFuentes, L.; Santoyo-Gonzalez, F.; Vargas-Berenguel, A. FerroceneBeta-Cyclodextrin Conjugates: Synthesis, Supramolecular Behavior,

and Use as Electrochemical Sensors. Chem. - Eur. J. 2009, 15, 8146− 8162. (477) Gonzalez-Campo, A.; Eker, B.; Gardeniers, H. J.; Huskens, J.; Jonkheijm, P. A Supramolecular Approach to Enzyme Immobilization in Micro-Channels. Small 2012, 8, 3531−3537. (478) Ravoo, B. J.; Darcy, R. Cyclodextrin Bilayer Vesicles. Angew. Chem., Int. Ed. 2000, 39, 4324−4326. (479) Kauscher, U.; Stuart, M. C. A.; Drücker, P.; Galla, H.-J.; Ravoo, B. J. Incorporation of Amphiphilic Cyclodextrins into Liposomes as Artificial Receptor Units. Langmuir 2013, 29, 7377−7383. (480) Nalluri, S. K.; Voskuhl, J.; Bultema, J. B.; Boekema, E. J.; Ravoo, B. J. Light-Responsive Capture and Release of DNA in a Ternary Supramolecular Complex. Angew. Chem., Int. Ed. 2011, 50, 9747−9751. (481) Samanta, A.; Stuart, M. C. A.; Ravoo, B. J. Photoresponsive Capture and Release of Lectins in Multilamellar Complexes. J. Am. Chem. Soc. 2012, 134, 19909−19914. (482) Voskuhl, J.; Stuart, M. C.; Ravoo, B. J. Sugar-Decorated Sugar Vesicles: Lectin-Carbohydrate Recognition at the Surface of Cyclodextrin Vesicles. Chem. - Eur. J. 2010, 16, 2790−2796. (483) Collier, C. P. Bilayer Membrane Interactions with Nanofabricated Scaffolds. Chem. Phys. Lipids 2015, 192, 75−86. (484) Groves, J. T. In Biomems and Biomedical Nanotechnology: Vol. Iii Therapeutic Micro/Nanotechnology; Ferrari, M., Desai, T., Bhatia, S., Eds.; Springer: Boston, MA, 2007. (485) Kam, L. C. Capturing the Nanoscale Complexity of Cellular Membranes in Supported Lipid Bilayers. J. Struct. Biol. 2009, 168, 3− 10. (486) Oliver, A. E.; Parikh, A. N. Templating Membrane Assembly, Structure, and Dynamics Using Engineered Interfaces. Biochim. Biophys. Acta, Biomembr. 2010, 1798, 839−850. (487) Reimhult, E.; Kumar, K. Membrane Biosensor Platforms Using Nano- and Microporous Supports. Trends Biotechnol. 2008, 26, 82−89. (488) Troutier, A. L.; Ladaviere, C. An Overview of Lipid Membrane Supported by Colloidal Particles. Adv. Colloid Interface Sci. 2007, 133, 1−21. (489) Medintz, I. L.; Uyeda, H. T.; Goldman, E. R.; Mattoussi, H. Quantum Dot Bioconjugates for Imaging, Labelling and Sensing. Nat. Mater. 2005, 4, 435−446. (490) Voura, E. B.; Jaiswal, J. K.; Mattoussi, H.; Simon, S. M. Tracking Metastatic Tumor Cell Extravasation with Quantum Dot Nanocrystals and Fluorescence Emission-Scanning Microscopy. Nat. Med. 2004, 10, 993−998. (491) Brozell, A. M.; Muha, M. A.; Sanii, B.; Parikh, A. N. A Class of Supported Membranes: Formation of Fluid Phospholipid Bilayers on Photonic Band Gap Colloidal Crystals. J. Am. Chem. Soc. 2006, 128, 62−63. (492) Sanii, B.; Smith, A. M.; Butti, R.; Brozell, A. M.; Parikh, A. N. Bending Membranes on Demand: Fluid Phospholipid Bilayers on Topographically Deformable Substrates. Nano Lett. 2008, 8, 866−871. (493) Baciu, C. L.; Becker, J.; Janshoff, A.; Sonnichsen, C. ProteinMembrane Interaction Probed by Single Plasmonic Nanoparticles. Nano Lett. 2008, 8, 1724−1728. (494) Roiter, Y.; Ornatska, M.; Rammohan, A. R.; Balakrishnan, J.; Heine, D. R.; Minko, S. Interaction of Lipid Membrane with Nanostructured Surfaces. Langmuir 2009, 25, 6287−6299. (495) Wu, H. J.; Henzie, J.; Lin, W. C.; Rhodes, C.; Li, Z.; Sartorel, E.; Thorner, J.; Yang, P.; Groves, J. T. Membrane-Protein Binding Measured with Solution-Phase Plasmonic Nanocube Sensors. Nat. Methods 2012, 9, 1189−1191. (496) Worstell, N. C.; Krishnan, P.; Weatherston, J. D.; Wu, H. J. Binding Cooperativity Matters: A Gm1-Like Ganglioside-Cholera Toxin B Subunit Binding Study Using a Nanocube-Based Lipid Bilayer Array. PLoS One 2016, 11, e0153265. (497) Wittenberg, N. J.; Johnson, T. W.; Oh, S. H. High-Density Arrays of Submicron Spherical Supported Lipid Bilayers. Anal. Chem. 2012, 84, 8207−8213. (498) Wittenberg, N. J.; Im, H.; Johnson, T. W.; Xu, X.; Warrington, A. E.; Rodriguez, M.; Oh, S. H. Facile Assembly of Micro- and 5477

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Nanoarrays for Sensing with Natural Cell Membranes. ACS Nano 2011, 5, 7555−7564. (499) Han, X.; Qi, G.; Xu, X.; Wang, L. Lipid Bilayer Membrane Arrays: Fabrication and Applications. Adv. Biochem. Eng./Biotechnol. 2012, 131, 121−152. (500) Morigaki, K.; Baumgart, T.; Offenhausser, A.; Knoll, W. Patterning Solid-Supported Lipid Bilayer Membranes by Lithographic Polymerization of a Diacetylene Lipid. Angew. Chem., Int. Ed. 2001, 40, 172−174. (501) van Oudenaarden, A.; Boxer, S. G. Brownian Ratchets: Molecular Separations in Lipid Bilayers Supported on Patterned Arrays. Science 1999, 285, 1046−1048. (502) Jenkins, A. T. A.; Bushby, R. J.; Boden, N.; Evans, S. D.; Knowles, P. F.; Liu, Q.; Miles, R. E.; Ogier, S. D. Ion-Selective Lipid Bilayers Tethered to Microcontact Printed Self-Assembled Monolayers Containing Cholesterol Derivatives. Langmuir 1998, 14, 4675−4678. (503) Kung, L. A.; Groves, J. T.; Ulman, N.; Boxer, S. G. Printing Via Photolithography on Micropartitioned Fluid Lipid Membranes. Adv. Mater. 2000, 12, 731−734. (504) Srinivasan, M. P.; Ratto, T. V.; Stroeve, P.; Longo, M. L. Patterned Supported Bilayers on Self-Assembled Monolayers: Confinement of Adjacent Mobile Bilayers. Langmuir 2001, 17, 7951−7954. (505) Groves, J. T.; Ulman, N.; Cremer, P. S.; Boxer, S. G. Substrate−Membrane Interactions: Mechanisms for Imposing Patterns on a Fluid Bilayer Membrane. Langmuir 1998, 14, 3347−3350. (506) Glasmästar, K.; Gold, J.; Andersson, A.-S.; Sutherland, D. S.; Kasemo, B. Silicone Transfer During Microcontact Printing. Langmuir 2003, 19, 5475−5483. (507) Kung, L. A.; Kam, L.; Hovis, J. S.; Boxer, S. G. Patterning Hybrid Surfaces of Proteins and Supported Lipid Bilayers. Langmuir 2000, 16, 6773−6776. (508) Morigaki, K.; Baumgart, T.; Jonas, U.; Offenhäusser, A.; Knoll, W. Photopolymerization of Diacetylene Lipid Bilayers and Its Application to the Construction of Micropatterned Biomimetic Membranes. Langmuir 2002, 18, 4082−4089. (509) Orth, R. N.; Kameoka, J.; Zipfel, W. R.; Ilic, B.; Webb, W. W.; Clark, T. G.; Craighead, H. G. Creating Biological Membranes on the Micron Scale: Forming Patterned Lipid Bilayers Using a Polymer LiftOff Technique. Biophys. J. 2003, 85, 3066−3073. (510) Hovis, J. S.; Boxer, S. G. Patterning Barriers to Lateral Diffusion in Supported Lipid Bilayer Membranes by Blotting and Stamping. Langmuir 2000, 16, 894−897. (511) Hovis, J. S.; Boxer, S. G. Patterning and Composition Arrays of Supported Lipid Bilayers by Microcontact Printing. Langmuir 2001, 17, 3400−3405. (512) Yang, T. H.; Yee, C. K.; Amweg, M. L.; Singh, S.; Kendall, E. L.; Dattelbaum, A. M.; Shreve, A. P.; Brinker, C. J.; Parikh, A. N. Optical Detection of Ion-Channel-Induced Proton Transport in Supported Phospholipid Bilayers. Nano Lett. 2007, 7, 2446−2451. (513) Yee, C. K.; Amweg, M. L.; Parikh, A. N. Direct Photochemical Patterning and Refunctionalization of Supported Phospholipid Bilayers. J. Am. Chem. Soc. 2004, 126, 13962−13972. (514) Lenz, P.; Ajo-Franklin, C. M.; Boxer, S. G. Patterned Supported Lipid Bilayers and Monolayers on Poly(Dimethylsiloxane). Langmuir 2004, 20, 11092−11099. (515) Fang, Y.; Frutos, A. G.; Lahiri, J. Membrane Protein Microarrays. J. Am. Chem. Soc. 2002, 124, 2394−2395. (516) Fang, Y.; Lahiri, J.; Picard, L. G Protein-Coupled Receptor Microarrays for Drug Discovery. Drug Discovery Today 2003, 8, 755− 761. (517) Tanaka, M.; Rossetti, F. F.; Kaufmann, S. Native Supported Membranes: Creation of Two-Dimensional Cell Membranes on Polymer Supports (Review). Biointerphases 2008, 3, FA12. (518) Tanaka, M.; Kaufmann, S.; Nissen, J.; Hochrein, M. Orientation Selective Immobilization of Human Erythrocyte Membranes on Ultrathin Cellulose Films. Phys. Chem. Chem. Phys. 2001, 3, 4091−4095.

(519) Tanaka, M.; Wong, A. P.; Rehfeldt, F.; Tutus, M.; Kaufmann, S. Selective Deposition of Native Cell Membranes on Biocompatible Micropatterns. J. Am. Chem. Soc. 2004, 126, 3257−3260. (520) Perez, J. B.; Martinez, K. L.; Segura, J. M.; Vogel, H. Supported Cell-Membrane Sheets for Functional Fluorescence Imaging of Membrane Proteins. Adv. Funct. Mater. 2006, 16, 306−312. (521) Perez, J. B.; Segura, J. M.; Abankwa, D.; Piguet, J.; Martinez, K. L.; Vogel, H. Monitoring the Diffusion of Single Heterotrimeric G Proteins in Supported Cell-Membrane Sheets Reveals Their Partitioning into Microdomains. J. Mol. Biol. 2006, 363, 918−930. (522) Danelon, C.; Perez, J. B.; Santschi, C.; Brugger, J.; Vogel, H. Cell Membranes Suspended across Nanoaperture Arrays. Langmuir 2006, 22, 22−25. (523) Hsia, C. Y.; Chen, L.; Singh, R. R.; DeLisa, M. P.; Daniel, S. A Molecularly Complete Planar Bacterial Outer Membrane Platform. Sci. Rep. 2016, 6, 32715. (524) Charras, G.; Paluch, E. Blebs Lead the Way: How to Migrate without Lamellipodia. Nat. Rev. Mol. Cell Biol. 2008, 9, 730−736. (525) Charras, G. T. A Short History of Blebbing. J. Microsc. 2008, 231, 466−478. (526) Baumgart, T.; Hammond, A. T.; Sengupta, P.; Hess, S. T.; Holowka, D. A.; Baird, B. A.; Webb, W. W. Large-Scale Fluid/Fluid Phase Separation of Proteins and Lipids in Giant Plasma Membrane Vesicles. Proc. Natl. Acad. Sci. U. S. A. 2007, 104, 3165−3170. (527) Sezgin, E.; Levental, I.; Mayor, S.; Eggeling, C. The Mystery of Membrane Organization: Composition, Regulation and Roles of Lipid Rafts. Nat. Rev. Mol. Cell Biol. 2017, 18, 361−374. (528) Sezgin, E.; Kaiser, H. J.; Baumgart, T.; Schwille, P.; Simons, K.; Levental, I. Elucidating Membrane Structure and Protein Behavior Using Giant Plasma Membrane Vesicles. Nat. Protoc. 2012, 7, 1042− 1051. (529) Levental, K. R.; Levental, I. Giant Plasma Membrane Vesicles: Models for Understanding Membrane Organization. Curr. Top. Membr. 2015, 75, 25−57. (530) Levental, I.; Lingwood, D.; Grzybek, M.; Coskun, U.; Simons, K. Palmitoylation Regulates Raft Affinity for the Majority of Integral Raft Proteins. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 22050−22054. (531) Holowka, D.; Baird, B. Structural Studies on the MembraneBound Immunoglobulin E-Receptor Complex. 1. Characterization of Large Plasma Membrane Vesicles from Rat Basophilic Leukemia Cells and Insertion of Amphipathic Fluorescent Probes. Biochemistry 1983, 22, 3466−3474. (532) Ellerby, H. M.; Arap, W.; Ellerby, L. M.; Kain, R.; Andrusiak, R.; Rio, G. D.; Krajewski, S.; Lombardo, C. R.; Rao, R.; Ruoslahti, E.; et al. Anti-Cancer Activity of Targeted Pro-Apoptotic Peptides. Nat. Med. 1999, 5, 1032−1038. (533) Lauster, D.; Vazquez, O.; Schwarzer, R.; Seitz, O.; Herrmann, A. Potential of Proapoptotic Peptides to Induce the Formation of Giant Plasma Membrane Vesicles with Lipid Domains. ChemBioChem 2015, 16, 1288−1292. (534) Costello, D. A.; Hsia, C. Y.; Millet, J. K.; Porri, T.; Daniel, S. Membrane Fusion-Competent Virus-Like Proteoliposomes and Proteinaceous Supported Bilayers Made Directly from Cell Plasma Membranes. Langmuir 2013, 29, 6409−6419. (535) Richards, M. J.; Hsia, C. Y.; Singh, R. R.; Haider, H.; Kumpf, J.; Kawate, T.; Daniel, S. Membrane Protein Mobility and Orientation Preserved in Supported Bilayers Created Directly from Cell Plasma Membrane Blebs. Langmuir 2016, 32, 2963−2974. (536) Reimhult, E.; Kasemo, B.; Hook, F. Rupture Pathway of Phosphatidylcholine Liposomes on Silicon Dioxide. Int. J. Mol. Sci. 2009, 10, 1683−1696. (537) Fuhrmans, M.; Muller, M. Mechanisms of Vesicle Spreading on Surfaces: Coarse-Grained Simulations. Langmuir 2013, 29, 4335− 4349. (538) Roier, S.; Zingl, F. G.; Cakar, F.; Durakovic, S.; Kohl, P.; Eichmann, T. O.; Klug, L.; Gadermaier, B.; Weinzerl, K.; Prassl, R.; et al. A Novel Mechanism for the Biogenesis of Outer Membrane Vesicles in Gram-Negative Bacteria. Nat. Commun. 2016, 7, 10515. 5478

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

(539) Kulp, A.; Kuehn, M. J. Biological Functions and Biogenesis of Secreted Bacterial Outer Membrane Vesicles. Annu. Rev. Microbiol. 2010, 64, 163−184. (540) Raposo, G.; Stoorvogel, W. Extracellular Vesicles: Exosomes, Microvesicles, and Friends. J. Cell Biol. 2013, 200, 373−383. (541) Tkach, M.; Thery, C. Communication by Extracellular Vesicles: Where We Are and Where We Need to Go. Cell 2016, 164, 1226−1232. (542) Ibrahim, A.; Marban, E. Exosomes: Fundamental Biology and Roles in Cardiovascular Physiology. Annu. Rev. Physiol. 2016, 78, 67− 83. (543) Thery, C.; Zitvogel, L.; Amigorena, S. Exosomes: Composition, Biogenesis and Function. Nat. Rev. Immunol. 2002, 2, 569−579. (544) Vlassov, A. V.; Magdaleno, S.; Setterquist, R.; Conrad, R. Exosomes: Current Knowledge of Their Composition, Biological Functions, and Diagnostic and Therapeutic Potentials. Biochim. Biophys. Acta, Gen. Subj. 2012, 1820, 940−948. (545) Im, H.; Shao, H.; Park, Y.-I.; Peterson, V. M.; Castro, C. M.; Weissleder, R.; Lee, H. J. Label-Free Detection and Molecular Profiling of Exosomes with a Nano-Plasmonic Sensor. Nat. Biotechnol. 2014, 32, 490−495. (546) Thakur, A.; Qiu, G.; Ng, S. P.; Guan, J.; Yue, J.; Lee, Y.; Wu, C. L. Direct Detection of Two Different Tumor-Derived Extracellular Vesicles by Sam-Aunis Lspr Biosensor. Biosens. Bioelectron. 2017, 94, 400−407. (547) Mashaghi, A.; Mashaghi, S.; Reviakine, I.; Heeren, R. M.; Sandoghdar, V.; Bonn, M. Label-Free Characterization of Biomembranes: From Structure to Dynamics. Chem. Soc. Rev. 2014, 43, 887− 900. (548) Hardy, G. J.; Nayak, R.; Zauscher, S. Model Cell Membranes: Techniques to Form Complex Biomimetic Supported Lipid Bilayers Via Vesicle Fusion. Curr. Opin. Colloid Interface Sci. 2013, 18, 448− 458. (549) Lee, T. H.; Aguilar, M. I. In Advances in Biomembranes and Lipid Self-Assembly; Iglič, A., Garcia-Sáez, A., Rappolt, M., Eds.; Academic Press: Oxford, UK, 2017; Vol. 25, Chapter 5. (550) Williams, L. M.; Evans, S. D.; Flynn, T. M.; Marsh, A.; Knowles, P. F.; Bushby, R. J.; Boden, N. Kinetics of Formation of Single Phospholipid Bilayers on Self-Assembled Monolayer Supports, as Monitored by Surface Plasmon Resonance. Supramol. Sci. 1997, 4, 513−517. (551) Zhang, L.; Booth, C. A.; Stroeve, P. Phosphatidylserine/ Cholesterol Bilayers Supported on a Polycation/Alkylthiol Layer Pair. J. Colloid Interface Sci. 2000, 228, 82−89. (552) Zhang, L.; Longo, M. L.; Stroeve, P. Mobile Phospholipid Bilayers Supported on a Polyion/Alkylthiol Layer Pair. Langmuir 2000, 16, 5093−5099. (553) Bunjes, N.; Schmidt, E. K.; Jonczyk, A.; Rippmann, F.; Beyer, D.; Ringsdorf, H.; Gräber, P.; Knoll, W.; Naumann, R. ThiopeptideSupported Lipid Layers on Solid Substrates. Langmuir 1997, 13, 6188−6194. (554) Santonicola, M. G.; Memesa, M.; Meszynska, A.; Ma, Y.; Vancso, G. J. Surface-Grafted Zwitterionic Polymers as Platforms for Functional Supported Phospholipid Membranes. Soft Matter 2012, 8, 1556−1562. (555) Tawa, K.; Morigaki, K. Substrate-Supported Phospholipid Membranes Studied by Surface Plasmon Resonance and Surface Plasmon Fluorescence Spectroscopy. Biophys. J. 2005, 89, 2750−2758. (556) Hubbard, J. B.; Silin, V.; Plant, A. L. Self Assembly Driven by Hydrophobic Interactions at Alkanethiol Monolayers: Mechanisms of Formation of Hybrid Bilayer Membranes. Biophys. Chem. 1998, 75, 163−176. (557) Stelzle, M.; Weissmueller, G.; Sackmann, E. On the Application of Supported Bilayers as Receptive Layers for Biosensors with Electrical Detection. J. Phys. Chem. 1993, 97, 2974−2981. (558) Knoll, W. Optical Characterization of Organic Thin Films and Interfaces with Evanescent Waves. MRS Bull. 1991, 16, 29−39. (559) Salamon, Z.; Huang, D.; Cramer, W. A.; Tollin, G. Coupled Plasmon-Waveguide Resonance Spectroscopy Studies of the Cyto-

chrome B6f/Plastocyanin System in Supported Lipid Bilayer Membranes. Biophys. J. 1998, 75, 1874−1885. (560) Huang, C.; Thompson, T. E. Properties of Lipid Bilayer Membranes Separating Two Aqueous Phases: Determination of Membrane Thickness. J. Mol. Biol. 1965, 13, 183−193. (561) Schouten, S.; Stroeve, P.; Longo, M. L. DNA Adsorption and Cationic Bilayer Deposition on Self-Assembled Monolayers. Langmuir 1999, 15, 8133−8139. (562) Morigaki, K.; Tawa, K. Vesicle Fusion Studied by Surface Plasmon Resonance and Surface Plasmon Fluorescence Spectroscopy. Biophys. J. 2006, 91, 1380−1387. (563) Ma, C.; Srinivasan, M. P.; Waring, A. J.; Lehrer, R. I.; Longo, M. L.; Stroeve, P. Supported Lipid Bilayers Lifted from the Substrate by Layer-by-Layer Polyion Cushions on Self-Assembled Monolayers. Colloids Surf., B 2003, 28, 319−329. (564) Vidu, R.; Zhang, L.; Waring, A. J.; Lehrer, R. I.; Longo, M. L.; Stroeve, P. Phospholipid Bilayers on a Polyion-Alkylthiol Layer Pair: Microprobe Imaging, Electrochemical Properties and Peptide Association. Mater. Sci. Eng., B 2002, 96, 199−208. (565) Ferhan, A. R.; Jackman, J. A.; Cho, N. J. Probing Spatial Proximity of Supported Lipid Bilayers to Silica Surfaces by Localized Surface Plasmon Resonance Sensing. Anal. Chem. 2017, 89, 4301− 4308. (566) Sadowski, J. W.; Korhonen, I. K. L.; Peltonen, J. P. Characterization of Thin Films and Their Structures in Surface Plasmon Resonance Measurements. Opt. Eng. 1995, 34, 2581−2586. (567) Grassi, J. H.; Georgiadis, R. M. Temperature-Dependent Refractive Index Determination from Critical Angle Measurements: Implications for Quantitative Spr Sensing. Anal. Chem. 1999, 71, 4392−4396. (568) Liang, H.; Miranto, H.; Granqvist, N.; Sadowski, J. W.; Viitala, T.; Wang, B.; Yliperttula, M. Surface Plasmon Resonance Instrument as a Refractometer for Liquids and Ultrathin Films. Sens. Actuators, B 2010, 149, 212−220. (569) Zhou, M.; Otomo, A.; Yokoyama, S.; Mashiko, S. Estimation of Organic Molecular Film Structures Using Surface−Plasmon Resonance Spectroscopy. Thin Solid Films 2001, 393, 114−118. (570) Granqvist, N.; Liang, H.; Laurila, T.; Sadowski, J.; Yliperttula, M.; Viitala, T. Characterizing Ultrathin and Thick Organic Layers by Surface Plasmon Resonance Three-Wavelength and Waveguide Mode Analysis. Langmuir 2013, 29, 8561−8571. (571) Granqvist, N.; Yliperttula, M.; Valimaki, S.; Pulkkinen, P.; Tenhu, H.; Viitala, T. Control of the Morphology of Lipid Layers by Substrate Surface Chemistry. Langmuir 2014, 30, 2799−2809. (572) Rupert, D. L.; Shelke, G. V.; Emilsson, G.; Claudio, V.; Block, S.; Lasser, C.; Dahlin, A.; Lotvall, J. O.; Bally, M.; Zhdanov, V. P.; et al. Dual-Wavelength Surface Plasmon Resonance for Determining the Size and Concentration of Sub-Populations of Extracellular Vesicles. Anal. Chem. 2016, 88, 9980−9988. (573) Cooper, M. A.; Try, A. C.; Carroll, J.; Ellar, D. J.; Williams, D. H. Surface Plasmon Resonance Analysis at a Supported Lipid Monolayer. Biochim. Biophys. Acta, Biomembr. 1998, 1373, 101−111. (574) Radler, U.; Mack, J.; Persike, N.; Jung, G.; Tampe, R. Design of Supported Membranes Tethered Via Metal-Affinity Ligand-Receptor Pairs. Biophys. J. 2000, 79, 3144−3152. (575) Erb, E. M.; Chen, X.; Allen, S.; Roberts, C. J.; Tendler, S. J.; Davies, M. C.; Forsen, S. Characterization of the Surfaces Generated by Liposome Binding to the Modified Dextran Matrix of a Surface Plasmon Resonance Sensor Chip. Anal. Biochem. 2000, 280, 29−35. (576) Canovi, M.; Lucchetti, J.; Stravalaci, M.; Re, F.; Moscatelli, D.; Bigini, P.; Salmona, M.; Gobbi, M. Applications of Surface Plasmon Resonance (Spr) for the Characterization of Nanoparticles Developed for Biomedical Purposes. Sensors 2012, 12, 16420−16432. (577) Baird, C. L.; Courtenay, E. S.; Myszka, D. G. Surface Plasmon Resonance Characterization of Drug/Liposome Interactions. Anal. Biochem. 2002, 310, 93−99. (578) Kim, K.; Cho, S.; Park, J. H.; Byun, Y.; Chung, H.; Kwon, I. C.; Jeong, S. Y. Surface Plasmon Resonance Studies of the Direct 5479

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Interaction between a Drug/Intestinal Brush Border Membrane. Pharm. Res. 2004, 21, 1233−1239. (579) Anderluh, G.; Besenicar, M.; Kladnik, A.; Lakey, J. H.; Macek, P. Properties of Nonfused Liposomes Immobilized on an L1 Biacore Chip and Their Permeabilization by a Eukaryotic Pore-Forming Toxin. Anal. Biochem. 2005, 344, 43−52. (580) Stahelin, R. V.; Cho, W. Differential Roles of Ionic, Aliphatic, and Aromatic Residues in Membrane-Protein Interactions: A Surface Plasmon Resonance Study on Phospholipases A2. Biochemistry 2001, 40, 4672−4678. (581) Honing, S.; Ricotta, D.; Krauss, M.; Spate, K.; Spolaore, B.; Motley, A.; Robinson, M.; Robinson, C.; Haucke, V.; Owen, D. J. Phosphatidylinositol-(4,5)-Bisphosphate Regulates Sorting Signal Recognition by the Clathrin-Associated Adaptor Complex Ap2. Mol. Cell 2005, 18, 519−531. (582) Saenko, E.; Sarafanov, A.; Ananyeva, N.; Behre, E.; Shima, M.; Schwinn, H.; Josic, D. Comparison of the Properties of Phospholipid Surfaces Formed on Hpa and L1 Biosensor Chips for the Binding of the Coagulation Factor Viii. J. Chromatogr. A 2001, 921, 49−56. (583) Olaru, A.; Gheorghiu, M.; David, S.; Polonschii, C.; Gheorghiu, E. Quality Assessment of Spr Sensor Chips; Case Study on L1 Chips. Biosens. Bioelectron. 2013, 45, 77−81. (584) Adamson, R. J.; Watts, A. Kinetics of the Early Events of Gpcr Signalling. FEBS Lett. 2014, 588, 4701−4707. (585) Bocquet, N.; Kohler, J.; Hug, M. N.; Kusznir, E. A.; Rufer, A. C.; Dawson, R. J.; Hennig, M.; Ruf, A.; Huber, W.; Huber, S. RealTime Monitoring of Binding Events on a Thermostabilized Human A2a Receptor Embedded in a Lipid Bilayer by Surface Plasmon Resonance. Biochim. Biophys. Acta, Biomembr. 2015, 1848, 1224−1233. (586) Borch, J.; Torta, F.; Sligar, S. G.; Roepstorff, P. Nanodiscs for Immobilization of Lipid Bilayers and Membrane Receptors: Kinetic Analysis of Cholera Toxin Binding to a Glycolipid Receptor. Anal. Chem. 2008, 80, 6245−6252. (587) Gluck, J. M.; Koenig, B. W.; Willbold, D. Nanodiscs Allow the Use of Integral Membrane Proteins as Analytes in Surface Plasmon Resonance Studies. Anal. Biochem. 2011, 408, 46−52. (588) Goluch, E. D.; Shaw, A. W.; Sligar, S. G.; Liu, C. Microfluidic Patterning of Nanodisc Lipid Bilayers and Multiplexed Analysis of Protein Interaction. Lab Chip 2008, 8, 1723−1728. (589) Hornschemeyer, P.; Liss, V.; Heermann, R.; Jung, K.; Hunke, S. Interaction Analysis of a Two-Component System Using Nanodiscs. PLoS One 2016, 11, e0149187. (590) Jang, H.; Abraham, S. J.; Chavan, T. S.; Hitchinson, B.; Khavrutskii, L.; Tarasova, N. I.; Nussinov, R.; Gaponenko, V. Mechanisms of Membrane Binding of Small Gtpase K-Ras4b Farnesylated Hypervariable Region. J. Biol. Chem. 2015, 290, 9465− 9477. (591) Proverbio, D.; Roos, C.; Beyermann, M.; Orban, E.; Dotsch, V.; Bernhard, F. Functional Properties of Cell-Free Expressed Human Endothelin a and Endothelin B Receptors in Artificial Membrane Environments. Biochim. Biophys. Acta, Biomembr. 2013, 1828, 2182− 2192. (592) Tavoosi, N.; Smith, S. A.; Davis-Harrison, R. L.; Morrissey, J. H. Factor Vii and Protein C Are Phosphatidic Acid-Binding Proteins. Biochemistry 2013, 52, 5545−5552. (593) Trahey, M.; Li, M. J.; Kwon, H.; Woodahl, E. L.; McClary, W. D.; Atkins, W. M. Applications of Lipid Nanodiscs for the Study of Membrane Proteins by Surface Plasmon Resonance. In Current Protocols in Protein Science; John Wiley & Sons: Hoboken, NJ, 2015; Vol. 8110.1002/0471140864.ps2913s81. (594) Xu, H.; Hill, J. J.; Michelsen, K.; Yamane, H.; Kurzeja, R. J.; Tam, T.; Isaacs, R. J.; Shen, F.; Tagari, P. Characterization of the Direct Interaction between Kcsa-Kv1.3 and Its Inhibitors. Biochim. Biophys. Acta, Biomembr. 2015, 1848, 1974−1980. (595) Escorihuela, J.; Gonzalez-Martinez, M. A.; Lopez-Paz, J. L.; Puchades, R.; Maquieira, A.; Gimenez-Romero, D. Dual-Polarization Interferometry: A Novel Technique to Light up the Nanomolecular World. Chem. Rev. 2015, 115, 265−294.

(596) Lee, T. H.; Hall, K. N.; Swann, M. J.; Popplewell, J. F.; Unabia, S.; Park, Y.; Hahm, K. S.; Aguilar, M. I. The Membrane Insertion of Helical Antimicrobial Peptides from the N-Terminus of Helicobacter Pylori Ribosomal Protein L1. Biochim. Biophys. Acta, Biomembr. 2010, 1798, 544−557. (597) Lee, T. H.; Heng, C.; Swann, M. J.; Gehman, J. D.; Separovic, F.; Aguilar, M. I. Real-Time Quantitative Analysis of Lipid Disordering by Aurein 1.2 During Membrane Adsorption, Destabilisation and Lysis. Biochim. Biophys. Acta, Biomembr. 2010, 1798, 1977−1986. (598) Nagle, J. F.; Tristram-Nagle, S. Structure of Lipid Bilayers. Biochim. Biophys. Acta, Rev. Biomembr. 2000, 10, 159−195. (599) Andreu-Fernandéz, V.; Genoves, A.; Lee, T. H.; Stellato, M.; Lucantoni, F.; Orzáez, M.; Mingarro, I.; Aguilar, M. I.; Pérez-Payá, E. Peptides Derived from the Transmembrane Domain of Bcl-2 Proteins as Potential Mitochondrial Priming Tools. ACS Chem. Biol. 2014, 9, 1799−1811. (600) Dekan, Z.; Headey, S. J.; Scanlon, M.; Baldo, B. A.; Lee, T. H.; Aguilar, M. I.; Deuis, J. R.; Vetter, I.; Elliott, A. G.; Amado, M. DeltaMyrtoxin-Mp1a Is a Helical Heterodimer from the Venom of the Jack Jumper Ant That Has Antimicrobial, Membrane-Disrupting, and Nociceptive Activities. Angew. Chem., Int. Ed. 2017, 56, 8495−8499. (601) Hirst, D. J.; Lee, T. H.; Pattenden, L. K.; Thomas, W. G.; Aguilar, M. I. Helix 8 of the Angiotensin- Ii Type 1a Receptor Interacts with Phosphatidylinositol Phosphates and Modulates Membrane Insertion. Sci. Rep. 2015, 5, 9972. (602) Ouberai, M. M.; Wang, J.; Swann, M. J.; Galvagnion, C.; Guilliams, T.; Dobson, C. M.; Welland, M. E. Alpha-Synuclein Senses Lipid Packing Defects and Induces Lateral Expansion of Lipids Leading to Membrane Remodeling. J. Biol. Chem. 2013, 288, 20883− 20895. (603) Popplewell, J. F.; Swann, M. J.; Freeman, N. J.; McDonnell, C.; Ford, R. C. Quantifying the Effects of Melittin on Liposomes. Biochim. Biophys. Acta, Biomembr. 2007, 1768, 13−20. (604) Khan, T. R.; Grandin, H. M.; Mashaghi, A.; Textor, M.; Reimhult, E.; Reviakine, I. Lipid Redistribution in PhosphatidylserineContaining Vesicles Adsorbing on Titania. Biointerphases 2008, 3, FA90. (605) Graneli, A.; Edvardsson, M.; Hook, F. DNA-Based Formation of a Supported, Three-Dimensional Lipid Vesicle Matrix Probed by Qcm-D and Spr. ChemPhysChem 2004, 5, 729−733. (606) Hall, K.; Lee, T. H.; Mechler, A. I.; Swann, M. J.; Aguilar, M. I. Real-Time Measurement of Membrane Conformational States Induced by Antimicrobial Peptides: Balance between Recovery and Lysis. Sci. Rep. 2015, 4, 5479. (607) Hirst, D. J.; Lee, T. H.; Swann, M. J.; Unabia, S.; Park, Y.; Hahm, K. S.; Aguilar, M. I. Effect of Acyl Chain Structure and Bilayer Phase State on Binding and Penetration of a Supported Lipid Bilayer by Hpa3. Eur. Biophys. J. 2011, 40, 503−514. (608) Lee, T. H.; Heng, C.; Separovic, F.; Aguilar, M. I. Comparison of Reversible Membrane Destabilisation Induced by Antimicrobial Peptides Derived from Australian Frogs. Biochim. Biophys. Acta, Biomembr. 2014, 1838, 2205−2215. (609) Deuis, J. R.; Dekan, Z.; Inserra, M. C.; Lee, T. H.; Aguilar, M. I.; Craik, D. J.; Lewis, R. J.; Alewood, P. F.; Mobli, M.; Schroeder, C. I.; et al. Development of a Muo-Conotoxin Analogue with Improved Lipid Membrane Interactions and Potency for the Analgesic Sodium Channel Nav1.8. J. Biol. Chem. 2016, 291, 11829−11842. (610) Sanghera, N.; Swann, M. J.; Ronan, G.; Pinheiro, T. J. Insight into Early Events in the Aggregation of the Prion Protein on Lipid Membranes. Biochim. Biophys. Acta, Biomembr. 2009, 1788, 2245− 2251. (611) Payne, J. A.; Bleackley, M. R.; Lee, T. H.; Shafee, T. M.; Poon, I. K.; Hulett, M. D.; Aguilar, M. I.; van der Weerden, N. L.; Anderson, M. A. The Plant Defensin Nad1 Introduces Membrane Disorder through a Specific Interaction with the Lipid, Phosphatidylinositol 4,5 Bisphosphate. Biochim. Biophys. Acta, Biomembr. 2016, 1858, 1099− 1109. 5480

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

(612) Balhara, V.; Schmidt, R.; Gorr, S. U.; Dewolf, C. Membrane Selectivity and Biophysical Studies of the Antimicrobial Peptide Gl13k. Biochim. Biophys. Acta, Biomembr. 2013, 1828, 2193−2203. (613) Singh, S.; Kalle, M.; Papareddy, P.; Schmidtchen, A.; Malmsten, M. Lipopolysaccharide Interactions of C-Terminal Peptides from Human Thrombin. Biomacromolecules 2013, 14, 1482−1492. (614) Singh, S.; Papareddy, P.; Kalle, M.; Schmidtchen, A.; Malmsten, M. Importance of Lipopolysaccharide Aggregate Disruption for the Anti-Endotoxic Effects of Heparin Cofactor Ii Peptides. Biochim. Biophys. Acta, Biomembr. 2013, 1828, 2709−2719. (615) Karst, J. C.; Barker, R.; Devi, U.; Swann, M. J.; Davi, M.; Roser, S. J.; Ladant, D.; Chenal, A. Identification of a Region That Assists Membrane Insertion and Translocation of the Catalytic Domain of Bordetella Pertussis Cyaa Toxin. J. Biol. Chem. 2012, 287, 9200−9212. (616) Horváth, R.; Fricsovszky, G.; Papp, E. Application of the Optical Waveguide Lightmode Spectroscopy to Monitor Lipid Bilayer Phase Transition. Biosens. Bioelectron. 2003, 18, 415−428. (617) Ramsden, J. J. Molecular Orientation in Lipid Bilayers. Philos. Mag. B 1999, 79, 381−386. (618) Anderson, T. H.; Min, Y.; Weirich, K. L.; Zeng, H.; Fygenson, D.; Israelachvili, J. N. Formation of Supported Bilayers on Silica Substrates. Langmuir 2009, 25, 6997−7005. (619) Rossetti, F. F.; Textor, M.; Reviakine, I. Asymmetric Distribution of Phosphatidyl Serine in Supported Phospholipid Bilayers on Titanium Dioxide. Langmuir 2006, 22, 3467−3473. (620) Stanglmaier, S.; Hertrich, S.; Fritz, K.; Moulin, J. F.; HaeseSeiller, M.; Radler, J. O.; Nickel, B. Asymmetric Distribution of Anionic Phospholipids in Supported Lipid Bilayers. Langmuir 2012, 28, 10818−10821. (621) Wacklin, H. P. Composition and Asymmetry in Supported Membranes Formed by Vesicle Fusion. Langmuir 2011, 27, 7698− 7707. (622) Harte, E.; Maalouli, N.; Shalabney, A.; Texier, E.; Berthelot, K.; Lecomte, S.; Alves, I. D. Probing the Kinetics of Lipid Membrane Formation and the Interaction of a Nontoxic and a Toxic Amyloid with Plasmon Waveguide Resonance. Chem. Commun. 2014, 50, 4168−4171. (623) Visco, I.; Chiantia, S.; Schwille, P. Asymmetric Supported Lipid Bilayer Formation Via Methyl-Beta-Cyclodextrin Mediated Lipid Exchange: Influence of Asymmetry on Lipid Dynamics and Phase Behavior. Langmuir 2014, 30, 7475−7484. (624) Jing, Y.; Kunze, A.; Svedhem, S. Phase Transition-Controlled Flip-Flop in Asymmetric Lipid Membranes. J. Phys. Chem. B 2014, 118, 2389−2395. (625) Lewis, R. N.; Zhang, Y. P.; McElhaney, R. N. Calorimetric and Spectroscopic Studies of the Phase Behavior and Organization of Lipid Bilayer Model Membranes Composed of Binary Mixtures of Dimyristoylphosphatidylcholine and Dimyristoylphosphatidylglycerol. Biochim. Biophys. Acta, Biomembr. 2005, 1668, 203−214. (626) Tristram-Nagle, S.; Nagle, J. F. Lipid Bilayers: Thermodynamics, Structure, Fluctuations, and Interactions. Chem. Phys. Lipids 2004, 127, 3−14. (627) Dahlin, A. B.; Chen, S.; Jonsson, M. P.; Gunnarsson, L.; Kall, M.; Hook, F. High-Resolution Microspectroscopy of Plasmonic Nanostructures for Miniaturized Biosensing. Anal. Chem. 2009, 81, 6572−6580. (628) Matthews, J. R.; Payne, C. M.; Hafner, J. H. Analysis of Phospholipid Bilayers on Gold Nanorods by Plasmon Resonance Sensing and Surface-Enhanced Raman Scattering. Langmuir 2015, 31, 9893−9900. (629) Arslan Yildiz, A.; Yildiz, U. H.; Liedberg, B.; Sinner, E. K. Biomimetic Membrane Platform: Fabrication, Characterization and Applications. Colloids Surf., B 2013, 103, 510−516. (630) Hinman, S. S.; Ruiz, C. J.; Drakakaki, G.; Wilkop, T. E.; Cheng, Q. On-Demand Formation of Supported Lipid Membrane Arrays by Trehalose-Assisted Vesicle Delivery for Spr Imaging. ACS Appl. Mater. Interfaces 2015, 7, 17122−17130. (631) Lee, S. H.; Lindquist, N. C.; Wittenberg, N. J.; Jordan, L. R.; Oh, S. H. Real-Time Full-Spectral Imaging and Affinity Measurements

from 50 Microfluidic Channels Using Nanohole Surface Plasmon Resonance. Lab Chip 2012, 12, 3882−3890. (632) Liu, L.; Wang, Q.; Yang, Z.; Wang, W.; Hu, N.; Luo, H.; Liao, Y.; Zheng, X.; Yang, J. Characterization of Lipid Films by an AngleInterrogation Surface Plasmon Resonance Imaging Device. Colloids Surf., B 2015, 128, 287−295. (633) Strulson, M. K.; Maurer, J. A. Mechanistic Insight into Patterned Supported Lipid Bilayer Self-Assembly. Langmuir 2012, 28, 13652−13659. (634) Taylor, J. D.; Linman, M. J.; Wilkop, T.; Cheng, Q. Regenerable Tethered Bilayer Lipid Membrane Arrays for Multiplexed Label-Free Analysis of Lipid-Protein Interactions on Poly(Dimethylsiloxane) Microchips Using Spr Imaging. Anal. Chem. 2009, 81, 1146−1153. (635) Pertsin, A. J.; Grunze, M. Computer Simulation of Water near the Surface of Oligo(Ethylene Glycol)-Terminated Alkanethiol SelfAssembled Monolayers. Langmuir 2000, 16, 8829−8841. (636) Lingler, S.; Rubinstein, I.; Knoll, W.; Offenhäusser, A. Fusion of Small Unilamellar Lipid Vesicles to Alkanethiol and Thiolipid SelfAssembled Monolayers on Gold. Langmuir 1997, 13, 7085−7091. (637) Zhelev, D. V.; Stoicheva, N.; Scherrer, P.; Needham, D. Interaction of Synthetic Ha2 Influenza Fusion Peptide Analog with Model Membranes. Biophys. J. 2001, 81, 285−304. (638) Wimley, W. C.; White, S. H. Quantitation of Electrostatic and Hydrophobic Membrane Interactions by Equilibrium Dialysis and Reverse-Phase Hplc. Anal. Biochem. 1993, 213, 213−217. (639) Jacobs, R. E.; White, S. H. The Nature of the Hydrophobic Binding of Small Peptides at the Bilayer Interface: Implications for the Insertion of Transbilayer Helices. Biochemistry 1989, 28, 3421−3437. (640) Heymann, J. B.; Zakharov, S. D.; Zhang, Y. L.; Cramer, W. A. Characterization of Electrostatic and Nonelectrostatic Components of Protein–Membrane Binding Interactions. Biochemistry 1996, 35, 2717−2725. (641) Maier, O.; Oberle, V.; Hoekstra, D. Fluorescent Lipid Probes: Some Properties and Applications (a Review). Chem. Phys. Lipids 2002, 116, 3−18. (642) Fuertes, G.; Giménez, D.; Esteban-Martin, S.; García-Sáez, A.; Sánchez, O.; Salgado, J. Role of Membrane Lipids for the Activity of Pore Forming Peptides and Proteins. In Advances in Experimental Medicine and Biology; Springer: New York, 2010; Vol. 677, pp 31− 5510.1007/978-1-4419-6327-7_4. (643) Teixeira, V.; Feio, M. J.; Bastos, M. Role of Lipids in the Interaction of Antimicrobial Peptides with Membranes. Prog. Lipid Res. 2012, 51, 149−177. (644) Yeaman, M. R.; Yount, N. Y. Mechanisms of Antimicrobial Peptide Action and Resistance. Pharmacol. Rev. 2003, 55, 27−55. (645) Galdiero, S.; Falanga, A.; Cantisani, M.; Vitiello, M.; Morelli, G.; Galdiero, M. Peptide-Lipid Interactions: Experiments and Applications. Int. J. Mol. Sci. 2013, 14, 18758−18789. (646) Bechinger, B. The Structure, Dynamics and Orientation of Antimicrobial Peptides in Membranes by Multidimensional Solid-State Nmr Spectroscopy. Biochim. Biophys. Acta, Biomembr. 1999, 1462, 157−183. (647) Karlsson, R.; Falt, A. Experimental Design for Kinetic Analysis of Protein-Protein Interactions with Surface Plasmon Resonance Biosensors. J. Immunol. Methods 1997, 200, 121−133. (648) Karlsson, R.; Michaelsson, A.; Mattsson, L. Kinetic Analysis of Monoclonal Antibody-Antigen Interactions with a New Biosensor Based Analytical System. J. Immunol. Methods 1991, 145, 229−240. (649) O’Shannessy, D. J.; Brigham-Burke, M.; Soneson, K. K.; Hensley, P.; Brooks, I. Determination of Rate and Equilibrium Binding Constants for Macromolecular Interactions Using Surface Plasmon Resonance: Use of Nonlinear Least Squares Analysis Methods. Anal. Biochem. 1993, 212, 457−468. (650) Roden, L. D.; Myszka, D. G. Global Analysis of a Macromolecular Interaction Measured on Biacore. Biochem. Biophys. Res. Commun. 1996, 225, 1073−1077. (651) Mozsolits, H.; Wirth, H. J.; Werkmeister, J.; Aguilar, M. I. Analysis of Antimicrobial Peptide Interactions with Hybrid Bilayer 5481

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Membrane Systems Using Surface Plasmon Resonance. Biochim. Biophys. Acta, Biomembr. 2001, 1512, 64−76. (652) Papo, N.; Shai, Y. Exploring Peptide Membrane Interaction Using Surface Plasmon Resonance: Differentiation between Pore Formation Versus Membrane Disruption by Lytic Peptides. Biochemistry 2003, 42, 458−466. (653) Hirst, D. J.; Lee, T. H.; Swann, M. J.; Aguilar, M. I. Combined Mass and Structural Kinetic Analysis of Multistate Antimicrobial Peptide-Membrane Interactions. Anal. Chem. 2013, 85, 9296−9304. (654) Hall, K.; Aguilar, M. I. Surface Plasmon Resonance Spectroscopy for Studying the Membrane Binding of Antimicrobial Peptides. Methods Mol. Biol. 2010, 627, 213−223. (655) Hall, K.; Lee, T. H.; Aguilar, M. I. The Role of Electrostatic Interactions in the Membrane Binding of Melittin. J. Mol. Recognit. 2011, 24, 108−118. (656) Lee, T. H.; Hall, K. N.; Aguilar, M. I. Antimicrobial Peptide Structure and Mechanism of Action: A Focus on the Role of Membrane Structure. Curr. Top. Med. Chem. 2016, 16, 25−39. (657) Gheorghiu, M.; Olaru, A.; Tar, A.; Polonschii, C.; Gheorghiu, E. Sensing Based on Assessment of Non-Monotonous Effect Determined by Target Analyte: Case Study on Pore-Forming Compounds. Biosens. Bioelectron. 2009, 24, 3517−3523. (658) Gheorghiu, M.; David, S.; Olaru, A.; Polonschii, C.; Gheorghiu, E. In Optical Nano- and Microsystems for Bioanalytics; Fritzsche, W., Popp, J., Eds.; Springer: Berlin, 2012. (659) Wang, W.; Smith, D. K.; Moulding, K.; Chen, H. M. The Dependence of Membrane Permeability by the Antibacterial Peptide Cecropin B and Its Analogs, Cb-1 and Cb-3, on Liposomes of Different Composition. J. Biol. Chem. 1998, 273, 27438−27448. (660) Wang, W.; Smith, D. K.; Chen, H. M. The Effect of Ph on the Structure, Binding and Model Membrane Lysis by Cecropin B and Analogs. Biochim. Biophys. Acta, Gen. Subj. 1999, 1473, 418−430. (661) Mangoni, M. L.; Papo, N.; Saugar, J. M.; Barra, D.; Shai, Y.; Simmaco, M.; Rivas, L. Effect of Natural L- to D-Amino Acid Conversion on the Organization, Membrane Binding, and Biological Function of the Antimicrobial Peptides Bombinins H. Biochemistry 2006, 45, 4266−4276. (662) Gehman, J. D.; Luc, F.; Hall, K.; Lee, T. H.; Boland, M. P.; Pukala, T. L.; Bowie, J. H.; Aguilar, M. I.; Separovic, F. Effect of Antimicrobial Peptides from Australian Tree Frogs on Anionic Phospholipid Membranes. Biochemistry 2008, 47, 8557−8565. (663) Chia, C. S. B.; Gong, Y.; Bowie, J. H.; Zuegg, J.; Cooper, M. A. Membrane Binding and Perturbation Studies of the Antimicrobial Peptides Caerin, Citropin, and Maculatin. Biopolymers 2011, 96, 147− 157. (664) Lombardi, L.; Stellato, M. I.; Oliva, R.; Falanga, A.; Galdiero, M.; Petraccone, L.; D’Errico, G.; De Santis, A.; Galdiero, S.; Del Vecchio, P. Antimicrobial Peptides at Work: Interaction of Myxinidin and Its Mutant Wmr with Lipid Bilayers Mimicking the P. Aeruginosa and E. Coli Membranes. Sci. Rep. 2017, 7, 44425. (665) Marani, M. M.; Perez, L. O.; de Araujo, A. R.; Placido, A.; Sousa, C. F.; Quelemes, P. V.; Oliveira, M.; Gomes-Alves, A. G.; Pueta, M.; Gameiro, P.; et al. Thaulin-1: The First Antimicrobial Peptide Isolated from the Skin of a Patagonian Frog Pleurodema Thaul (Anura: Leptodactylidae: Leiuperinae) with Activity against Escherichia Coli. Gene 2017, 605, 70−80. (666) Mangoni, M. L.; Papo, N.; Mignogna, G.; Andreu, D.; Shai, Y.; Barra, D.; Simmaco, M. Ranacyclins, a New Family of Short Cyclic Antimicrobial Peptides: Biological Function, Mode of Action, and Parameters Involved in Target Specificity. Biochemistry 2003, 42, 14023−14035. (667) Edwards, I. A.; Elliott, A. G.; Kavanagh, A. M.; Blaskovich, M. A. T.; Cooper, M. A. Structure-Activity and -Toxicity Relationships of the Antimicrobial Peptide Tachyplesin-1. ACS Infect. Dis. 2017, 3, 917. (668) Hall, K.; Lee, T. H.; Daly, N. L.; Craik, D. J.; Aguilar, M. I. Gly(6) of Kalata B1 Is Critical for the Selective Binding to Phosphatidylethanolamine Membranes. Biochim. Biophys. Acta, Biomembr. 2012, 1818, 2354−2361.

(669) Kamimori, H.; Hall, K.; Craik, D. J.; Aguilar, M. I. Studies on the Membrane Interactions of the Cyclotides Kalata B1 and Kalata B6 on Model Membrane Systems by Surface Plasmon Resonance. Anal. Biochem. 2005, 337, 149−153. (670) Sando, L.; Troeira Henriques, S. T.; Foley, F.; Simonsen, S. M.; Daly, N. L.; Hall, K. N.; Gustafson, K. R.; Aguilar, M. I.; Craik, D. J. A Synthetic Mirror Image of Kalata B1 Reveals That Cyclotide Activity Is Independent of a Protein Receptor. ChemBioChem 2011, 12, 2456− 2462. (671) Henriques, S. T.; Huang, Y. H.; Castanho, M. A.; Bagatolli, L. A.; Sonza, S.; Tachedjian, G.; Daly, N. L.; Craik, D. J. Phosphatidylethanolamine Binding Is a Conserved Feature of Cyclotide-Membrane Interactions. J. Biol. Chem. 2012, 287, 33629− 33643. (672) Henriques, S. T.; Huang, Y. H.; Chaousis, S.; Sani, M. A.; Poth, A. G.; Separovic, F.; Craik, D. J. The Prototypic Cyclotide Kalata B1 Has a Unique Mechanism of Entering Cells. Chem. Biol. 2015, 22, 1087−1097. (673) Henriques, S. T.; Huang, Y. H.; Rosengren, K. J.; Franquelim, H. G.; Carvalho, F. A.; Johnson, A.; Sonza, S.; Tachedjian, G.; Castanho, M. A.; Daly, N. L.; et al. Decoding the Membrane Activity of the Cyclotide Kalata B1: The Importance of Phosphatidylethanolamine Phospholipids and Lipid Organization on Hemolytic and AntiHiv Activities. J. Biol. Chem. 2011, 286, 24231−24241. (674) Papo, N.; Shai, Y. Effect of Drastic Sequence Alteration and DAmino Acid Incorporation on the Membrane Binding Behavior of Lytic Peptides. Biochemistry 2004, 43, 6393−6403. (675) Papo, N.; Shai, Y. A Molecular Mechanism for Lipopolysaccharide Protection of Gram-Negative Bacteria from Antimicrobial Peptides. J. Biol. Chem. 2005, 280, 10378−10387. (676) Hall, K.; Aguilar, M. I. Membrane Interactions of Antimicrobial Beta-Peptides: The Role of Amphipathicity Versus Secondary Structure Induction. Biopolymers 2009, 92, 554−564. (677) Makovitzki, A.; Shai, Y. Ph-Dependent Antifungal Lipopeptides and Their Plausible Mode of Action. Biochemistry 2005, 44, 9775− 9784. (678) Jin, Y.; Mozsolits, H.; Hammer, J.; Zmuda, E.; Zhu, F.; Zhang, Y.; Aguilar, M. I.; Blazyk, J. Influence of Tryptophan on Lipid Binding of Linear Amphipathic Cationic Antimicrobial Peptides. Biochemistry 2003, 42, 9395−9405. (679) Kamimori, H.; Blazyk, J.; Aguilar, M. I. Lipid MembraneBinding Properties of Tryptophan Analogues of Linear Amphipathic Beta-Sheet Cationic Antimicrobial Peptides Using Surface Plasmon Resonance. Biol. Pharm. Bull. 2005, 28, 148−150. (680) Zelezetsky, I.; Pacor, S.; Pag, U.; Papo, N.; Shai, Y.; Sahl, H. G.; Tossi, A. Controlled Alteration of the Shape and Conformational Stability of Alpha-Helical Cell-Lytic Peptides: Effect on Mode of Action and Cell Specificity. Biochem. J. 2005, 390, 177−188. (681) Papo, N.; Shai, Y. New Lytic Peptides Based on the D,LAmphipathic Helix Motif Preferentially Kill Tumor Cells Compared to Normal Cells. Biochemistry 2003, 42, 9346−9354. (682) Mozsolits, H.; Lee, T. H.; Clayton, A. H.; Sawyer, W. H.; Aguilar, M. I. The Membrane-Binding Properties of a Class a Amphipathic Peptide. Eur. Biophys. J. 2004, 33, 98−108. (683) Wexler-Cohen, Y.; Sackett, K.; Shai, Y. The Role of the NTerminal Heptad Repeat of Hiv-1 in the Actual Lipid Mixing Step as Revealed by Its Substitution with Distant Coiled Coils. Biochemistry 2005, 44, 5853−5861. (684) Cao, P.; Dou, G.; Cheng, Y.; Che, J. The Improved Efficacy of Sifuvirtide Compared with Enfuvirtide Might Be Related to Its Selectivity for the Rigid Biomembrane, as Determined through Surface Plasmon Resonance. PLoS One 2017, 12, e0171567. (685) Henriques, S. T.; Castanho, M. A.; Pattenden, L. K.; Aguilar, M. I. Fast Membrane Association Is a Crucial Factor in the Peptide Pep-1 Translocation Mechanism: A Kinetic Study Followed by Surface Plasmon Resonance. Biopolymers 2010, 94, 314−322. (686) Amon, M. A.; Ali, M.; Bender, V.; Hall, K.; Aguilar, M. I.; Aldrich-Wright, J.; Manolios, N. Kinetic and Conformational Proper5482

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

ties of a Novel T-Cell Antigen Receptor Transmembrane Peptide in Model Membranes. J. Pept. Sci. 2008, 14, 714−724. (687) Subasinghe, S.; Unabia, S.; Barrow, C. J.; Mok, S. S.; Aguilar, M. I.; Small, D. H. Cholesterol Is Necessary Both for the Toxic Effect of Abeta Peptides on Vascular Smooth Muscle Cells and for Abeta Binding to Vascular Smooth Muscle Cell Membranes. J. Neurochem. 2003, 84, 471−479. (688) Kremer, J. J.; Murphy, R. M. Kinetics of Adsorption of BetaAmyloid Peptide Abeta(1−40) to Lipid Bilayers. J. Biochem. Biophys. Methods 2003, 57, 159−169. (689) Ariga, T.; Kobayashi, K.; Hasegawa, A.; Kiso, M.; Ishida, H.; Miyatake, T. Characterization of High-Affinity Binding between Gangliosides and Amyloid Beta-Protein. Arch. Biochem. Biophys. 2001, 388, 225−230. (690) Lin, M. S.; Chiu, H. M.; Fan, F. J.; Tsai, H. T.; Wang, S. S.; Chang, Y.; Chen, W. Y. Kinetics and Enthalpy Measurements of Interaction between Beta-Amyloid and Liposomes by Surface Plasmon Resonance and Isothermal Titration Microcalorimetry. Colloids Surf., B 2007, 58, 231−236. (691) Valdes-Gonzalez, T.; Inagawa, J.; Ido, T. Neuropeptides Interact with Glycolipid Receptors: A Surface Plasmon Resonance Study. Peptides 2001, 22, 1099−1106. (692) Henriques, S. T.; Pattenden, L. K.; Aguilar, M. I.; Castanho, M. A. Prp(106−126) Does Not Interact with Membranes under Physiological Conditions. Biophys. J. 2008, 95, 1877−1889. (693) Henriques, S. T.; Pattenden, L. K.; Aguilar, M. I.; Castanho, M. A. The Toxicity of Prion Protein Fragment Prp(106−126) Is Not Mediated by Membrane Permeabilization as Shown by a M112w Substitution. Biochemistry 2009, 48, 4198−4208. (694) Thomas, C. J.; Surolia, N.; Surolia, A. Surface Plasmon Resonance Studies Resolve the Enigmatic Endotoxin Neutralizing Activity of Polymyxin B. J. Biol. Chem. 1999, 274, 29624−29627. (695) Thomas, C. J.; Surolia, A. Kinetics of the Interaction of Endotoxin with Polymyxin B and Its Analogs: A Surface Plasmon Resonance Analysis. FEBS Lett. 1999, 445, 420−424. (696) Oka, M.; Kamimori, H. Lipid Membrane-Binding Properties of Amphotericin B Deoxycholate (Fungizone) Using Surface Plasmon Resonance. Anal. Sci. 2013, 29, 697−702. (697) Onishi, M.; Kamimori, H. High-Throughput and Sensitive Assay for Amphotericin B Interaction with Lipid Membrane on the Model Membrane Systems by Surface Plasmon Resonance. Biol. Pharm. Bull. 2013, 36, 658−663. (698) Kinouchi, H.; Onishi, M.; Kamimori, H. Lipid MembraneBinding Properties of Daptomycin Using Surface Plasmon Resonance. Anal. Sci. 2013, 29, 297−301. (699) Cooper, M. A.; Williams, D. H. Binding of Glycopeptide Antibiotics to a Model of a Vancomycin-Resistant Bacterium. Chem. Biol. 1999, 6, 891−899. (700) Hong, Q.; Gutierrez-Aguirre, I.; Barlic, A.; Malovrh, P.; Kristan, K.; Podlesek, Z.; Macek, P.; Turk, D.; Gonzalez-Manas, J. M.; Lakey, J. H.; et al. Two-Step Membrane Binding by Equinatoxin Ii, a PoreForming Toxin from the Sea Anemone, Involves an Exposed Aromatic Cluster and a Flexible Helix. J. Biol. Chem. 2002, 277, 41916−41924. (701) Praper, T.; Besenicar, M. P.; Istinic, H.; Podlesek, Z.; Metkar, S. S.; Froelich, C. J.; Anderluh, G. Human Perforin Permeabilizing Activity, but Not Binding to Lipid Membranes, Is Affected by Ph. Mol. Immunol. 2010, 47, 2492−2504. (702) Agwa, A. J.; Lawrence, N.; Deplazes, E.; Cheneval, O.; Chen, R. M.; Craik, D. J.; Schroeder, C. I.; Henriques, S. T. Spider Peptide Toxin Hwtx-Iv Engineered to Bind to Lipid Membranes Has an Increased Inhibitory Potency at Human Voltage-Gated Sodium Channel Hnav1.7. Biochim. Biophys. Acta, Biomembr. 2017, 1859, 835−844. (703) Deplazes, E.; Henriques, S. T.; Smith, J. J.; King, G. F.; Craik, D. J.; Mark, A. E.; Schroeder, C. I. Membrane-Binding Properties of Gating Modifier and Pore-Blocking Toxins: Membrane Interaction Is Not a Prerequisite for Modification of Channel Gating. Biochim. Biophys. Acta, Biomembr. 2016, 1858, 872−882.

(704) Yamaji, A.; Sekizawa, Y.; Emoto, K.; Sakuraba, H.; Inoue, K.; Kobayashi, H.; Umeda, M. Lysenin, a Novel Sphingomyelin-Specific Binding Protein. J. Biol. Chem. 1998, 273, 5300−5306. (705) Chapple, D. S.; Mason, D. J.; Joannou, C. L.; Odell, E. W.; Gant, V.; Evans, R. W. Structure-Function Relationship of Antibacterial Synthetic Peptides Homologous to a Helical Surface Region on Human Lactoferrin against Escherichia Coli Serotype O111. Infect. Immun. 1998, 66, 2434−2440. (706) Saenko, E.; Sarafanov, A.; Greco, N.; Shima, M.; Loster, K.; Schwinn, H.; Josic, D. Use of Surface Plasmon Resonance for Studies of Protein-Protein and Protein-Phospholipid Membrane Interactions. Application to the Binding of Factor Viii to Von Willebrand Factor and to Phosphatidylserine-Containing Membranes. J. Chromatogr. A 1999, 852, 59−71. (707) Saenko, E. L.; Scandella, D.; Yakhyaev, A. V.; Greco, N. J. Activation of Factor Viii by Thrombin Increases Its Affinity for Binding to Synthetic Phospholipid Membranes and Activated Platelets. J. Biol. Chem. 1998, 273, 27918−27926. (708) Wilkening, G.; Linke, T.; Sandhoff, K. Lysosomal Degradation on Vesicular Membrane Surfaces. Enhanced Glucosylceramide Degradation by Lysosomal Anionic Lipids and Activators. J. Biol. Chem. 1998, 273, 30271−30278. (709) Wilkening, G.; Linke, T.; Uhlhorn-Dierks, G.; Sandhoff, K. Degradation of Membrane-Bound Ganglioside Gm1. Stimulation by Bis(Monoacylglycero)Phosphate and the Activator Proteins Sap-B and Gm2-Ap. J. Biol. Chem. 2000, 275, 35814−35819. (710) Werth, N.; Schuette, C. G.; Wilkening, G.; Lemm, T.; Sandhoff, K. Degradation of Membrane-Bound Ganglioside Gm2 by Beta -Hexosaminidase A. Stimulation by Gm2 Activator Protein and Lysosomal Lipids. J. Biol. Chem. 2001, 276, 12685−12690. (711) Linke, T.; Wilkening, G.; Lansmann, S.; Moczall, H.; Bartelsen, O.; Weisgerber, J.; Sandhoff, K. Stimulation of Acid Sphingomyelinase Activity by Lysosomal Lipids and Sphingolipid Activator Proteins. Biol. Chem. 2001, 382, 283−290. (712) Veiga, A. S.; Pattenden, L. K.; Fletcher, J. M.; Castanho, M. A.; Aguilar, M. I. Interactions of Hiv-1 Antibodies 2f5 and 4e10 with a Gp41 Epitope Prebound to Host and Viral Membrane Model Systems. ChemBioChem 2009, 10, 1032−1044. (713) Augusto, M. T.; Hollmann, A.; Troise, F.; Veiga, A. S.; Pessi, A.; Santos, N. C. Lipophilicity Is a Key Factor to Increase the Antiviral Activity of Hiv Neutralizing Antibodies. Colloids Surf., B 2017, 152, 311−316. (714) Bader, B.; Kuhn, K.; Owen, D. J.; Waldmann, H.; Wittinghofer, A.; Kuhlmann, J. Bioorganic Synthesis of Lipid-Modified Proteins for the Study of Signal Transduction. Nature 2000, 403, 223−226. (715) Lange, C.; Koch, K. W. Calcium-Dependent Binding of Recoverin to Membranes Monitored by Surface Plasmon Resonance Spectroscopy in Real Time. Biochemistry 1997, 36, 12019−12026. (716) Ozawa, T.; Fukuda, M.; Nara, M.; Nakamura, A.; Komine, Y.; Kohama, K.; Umezawa, Y. How Can Ca2+ Selectively Activate Recoverin in the Presence of Mg2+? Surface Plasmon Resonance and Ft-Ir Spectroscopic Studies. Biochemistry 2000, 39, 14495−14503. (717) Hodgkin, M. N.; Masson, M. R.; Powner, D.; Saqib, K. M.; Ponting, C. P.; Wakelam, M. J. Phospholipase D Regulation and Localisation Is Dependent Upon a Phosphatidylinositol 4,5Biphosphate-Specific Ph Domain. Curr. Biol. 2000, 10, 43−46. (718) Yu, J. W.; Lemmon, M. A. All Phox Homology (Px) Domains from Saccharomyces Cerevisiae Specifically Recognize Phosphatidylinositol 3-Phosphate. J. Biol. Chem. 2001, 276, 44179−44184. (719) Beven, L.; Adenier, H.; Kichenama, R.; Homand, J.; Redeker, V.; Le Caer, J. P.; Ladant, D.; Chopineau, J. Ca2+-Myristoyl Switch and Membrane Binding of Chemically Acylated Neurocalcins. Biochemistry 2001, 40, 8152−8160. (720) Cho, W. Membrane Targeting by C1 and C2 Domains. J. Biol. Chem. 2001, 276, 32407−32410. (721) Cho, W.; Bittova, L.; Stahelin, R. V. Membrane Binding Assays for Peripheral Proteins. Anal. Biochem. 2001, 296, 153−161. (722) Stahelin, R. V.; Cho, W. Roles of Calcium Ions in the Membrane Binding of C2 Domains. Biochem. J. 2001, 359, 679−685. 5483

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

(723) Bittova, L.; Stahelin, R. V.; Cho, W. Roles of Ionic Residues of the C1 Domain in Protein Kinase C-Alpha Activation and the Origin of Phosphatidylserine Specificity. J. Biol. Chem. 2001, 276, 4218−4226. (724) Ananthanarayanan, B.; Das, S.; Rhee, S. G.; Murray, D.; Cho, W. Membrane Targeting of C2 Domains of Phospholipase C-Delta Isoforms. J. Biol. Chem. 2002, 277, 3568−3575. (725) Bitto, E.; Li, M.; Tikhonov, A. M.; Schlossman, M. L.; Cho, W. Mechanism of Annexin I-Mediated Membrane Aggregation. Biochemistry 2000, 39, 13469−13477. (726) Kastl, K.; Ross, M.; Gerke, V.; Steinem, C. Kinetics and Thermodynamics of Annexin A1 Binding to Solid-Supported Membranes: A Qcm Study. Biochemistry 2002, 41, 10087−10094. (727) Ni, T.; Kalli, A. C.; Naughton, F. B.; Yates, L. A.; Naneh, O.; Kozorog, M.; Anderluh, G.; Sansom, M. S.; Gilbert, R. J. Structure and Lipid-Binding Properties of the Kindlin-3 Pleckstrin Homology Domain. Biochem. J. 2017, 474, 539−556. (728) Goda, T.; Miyahara, Y. Calcium-Independent Binding of Human C-Reactive Protein to Lysophosphatidylcholine in Supported Planar Phospholipid Monolayers. Acta Biomater. 2017, 48, 206−214. (729) Moriwaki, H.; Yamada, K.; Nakanishi, H. Evaluation of the Interaction between Pesticides and a Cell Membrane Model by Surface Plasmon Resonance Spectroscopy Analysis. J. Agric. Food Chem. 2017, 65, 5390−5396. (730) Papo, N.; Shahar, M.; Eisenbach, L.; Shai, Y. A Novel Lytic Peptide Composed of Dl-Amino Acids Selectively Kills Cancer Cells in Culture and in Mice. J. Biol. Chem. 2003, 278, 21018−21023. (731) Matsuzaki, K. Why and How Are Peptide-Lipid Interactions Utilized for Self-Defense? Magainins and Tachyplesins as Archetypes. Biochim. Biophys. Acta, Biomembr. 1999, 1462, 1−10. (732) Sani, M. A.; Separovic, F. How Membrane-Active Peptides Get into Lipid Membranes. Acc. Chem. Res. 2016, 49, 1130−1138. (733) Rady, I.; Siddiqui, I. A.; Rady, M.; Mukhtar, H. Melittin, a Major Peptide Component of Bee Venom, and Its Conjugates in Cancer Therapy. Cancer Lett. 2017, 402, 16−31. (734) Ghosh, A. K.; Rukmini, R.; Chattopadhyay, A. Modulation of Tryptophan Environment in Membrane-Bound Melittin by Negatively Charged Phospholipids: Implications in Membrane Organization and Function. Biochemistry 1997, 36, 14291−14305. (735) Beschiaschvili, G.; Seelig, J. Melittin Binding to Mixed Phosphatidylglycerol/Phosphatidylcholine Membranes. Biochemistry 1990, 29, 52−58. (736) Niemz, A.; Tirrell, D. A. Self-Association and MembraneBinding Behavior of Melittins Containing Trifluoroleucine. J. Am. Chem. Soc. 2001, 123, 7407−7413. (737) Anderluh, G.; Lakey, J. Preface. In Proteins: Membrane Binding and Pore Formation; Advances in Experimental Medicine and Biology; Springer: New York, 2010; Vol. 677, pp v−vi. (738) Anderluh, G.; Macek, P.; Lakey, J. H. Peeking into a Secret World of Pore-Forming Toxins: Membrane Binding Processes Studied by Surface Plasmon Resonance. Toxicon 2003, 42, 225−228. (739) Clayton, A. H.; Sawyer, W. H. The Structure and Orientation of Class-a Amphipathic Peptides on a Phospholipid Bilayer Surface. Eur. Biophys. J. 1999, 28, 133−141. (740) Blazyk, J.; Wiegand, R.; Klein, J.; Hammer, J.; Epand, R. M.; Epand, R. F.; Maloy, W. L.; Kari, U. P. A Novel Linear Amphipathic Beta-Sheet Cationic Antimicrobial Peptide with Enhanced Selectivity for Bacterial Lipids. J. Biol. Chem. 2001, 276, 27899−27906. (741) Chen, H. M.; Clayton, A. H.; Wang, W.; Sawyer, W. H. Kinetics of Membrane Lysis by Custom Lytic Peptides and Peptide Orientations in Membrane. Eur. J. Biochem. 2001, 268, 1659−1669. (742) Cooper, M. A.; Williams, D. H.; Cho, Y. R. Surface Plasmon Resonance Analysis of Glycopeptide Antibiotic Activity at a Model Membrane Surface. Chem. Commun. 1997, 1625−1626. (743) Kinouchi, H.; Arimoto, H.; Nishiguchi, K.; Oka, M.; Maki, H.; Kitagawa, H.; Kamimori, H. Binding Properties of Antimicrobial Agents to Lipid Membranes Using Surface Plasmon Resonance. Biol. Pharm. Bull. 2014, 37, 1383−1389. (744) Victor, K.; Jacob, J.; Cafiso, D. S. Interactions Controlling the Membrane Binding of Basic Protein Domains: Phenylalanine and the

Attachment of the Myristoylated Alanine-Rich C-Kinase Substrate Protein to Interfaces. Biochemistry 1999, 38, 12527−12536. (745) Mosior, M.; McLaughlin, S. Peptides That Mimic the Pseudosubstrate Region of Protein Kinase C Bind to Acidic Lipids in Membranes. Biophys. J. 1991, 60, 149−159. (746) Wertz, S. L.; Savino, Y.; Cafiso, D. S. Solution and Membrane Bound Structure of a Peptide Derived from the Protein Kinase C Substrate Domain of Neuromodulin. Biochemistry 1996, 35, 11104− 11112. (747) Berg, O. G.; Yu, B. Z.; Rogers, J.; Jain, M. K. Interfacial Catalysis by Phospholipase A2: Determination of the Interfacial Kinetic Rate Constants. Biochemistry 1991, 30, 7283−7297. (748) Zhou, W.; Parent, L. J.; Wills, J. W.; Resh, M. D. Identification of a Membrane-Binding Domain within the Amino-Terminal Region of Human Immunodeficiency Virus Type 1 Gag Protein Which Interacts with Acidic Phospholipids. J. Virol. 1994, 68, 2556−2569. (749) Prior, I. A.; Harding, A.; Yan, J.; Sluimer, J.; Parton, R. G.; Hancock, J. F. Gtp-Dependent Segregation of H-Ras from Lipid Rafts Is Required for Biological Activity. Nat. Cell Biol. 2001, 3, 368−375. (750) Johnson, J. E.; Cornell, R. B. Amphitropic Proteins: Regulation by Reversible Membrane Interactions (Review). Mol. Membr. Biol. 1999, 16, 217−235. (751) Das, S.; Cho, W. Roles of Catalytic Domain Residues in Interfacial Binding and Activation of Group Iv Cytosolic Phospholipase A2. J. Biol. Chem. 2002, 277, 23838−23846. (752) Veneziano, R.; Rossi, C.; Chenal, A.; Devoisselle, J. M.; Ladant, D.; Chopineau, J. Bordetella Pertussis Adenylate Cyclase Toxin Translocation across a Tethered Lipid Bilayer. Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 20473−20478. (753) Fischer, A.; Holden, M. A.; Pentelute, B. L.; Collier, R. J. Ultrasensitive Detection of Protein Translocated through Toxin Pores in Droplet-Interface Bilayers. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, 16577−16581. (754) Fernandez, D. I.; Lee, T. H.; Sani, M. A.; Aguilar, M. I.; Separovic, F. Proline Facilitates Membrane Insertion of the Antimicrobial Peptide Maculatin 1.1 Via Surface Indentation and Subsequent Lipid Disordering. Biophys. J. 2013, 104, 1495−1507. (755) Hirst, D. J.; Lee, T. H.; Kulkarni, K.; Wilce, J. A.; Aguilar, M. I. The Impact of Cell-Penetrating Peptides on Membrane Bilayer Structure During Binding and Insertion. Biochim. Biophys. Acta, Biomembr. 2016, 1858, 1841−1849. (756) Lee, T. H.; Sani, M. A.; Overall, S.; Separovic, F.; Aguilar, M. I. Effect of Phosphatidylcholine Bilayer Thickness and Molecular Order on the Binding of the Antimicrobial Peptide Maculatin 1.1. Biochim. Biophys. Acta, Biomembr. 2018, 1860, 300. (757) Sani, M. A.; Lee, T. H.; Aguilar, M. I.; Separovic, F. Proline-15 Creates an Amphipathic Wedge in Maculatin 1.1 Peptides That Drives Lipid Membrane Disruption. Biochim. Biophys. Acta, Biomembr. 2015, 1848, 2277−2289. (758) Singh, S.; Kasetty, G.; Schmidtchen, A.; Malmsten, M. Membrane and Lipopolysaccharide Interactions of C-Terminal Peptides from S1 Peptidases. Biochim. Biophys. Acta, Biomembr. 2012, 1818, 2244−2251. (759) Nielsen, S. B.; Otzen, D. E. Impact of the Antimicrobial Peptide Novicidin on Membrane Structure and Integrity. J. Colloid Interface Sci. 2010, 345, 248−256. (760) Fernandez, D. I.; Le Brun, A. P.; Lee, T. H.; Bansal, P.; Aguilar, M. I.; James, M.; Separovic, F. Structural Effects of the Antimicrobial Peptide Maculatin 1.1 on Supported Lipid Bilayers. Eur. Biophys. J. 2013, 42, 47−59. (761) Huang, H. W. Molecular Mechanism of Antimicrobial Peptides: The Origin of Cooperativity. Biochim. Biophys. Acta, Biomembr. 2006, 1758, 1292−1302. (762) Xu, X.; Lai, R. The Chemistry and Biological Activities of Peptides from Amphibian Skin Secretions. Chem. Rev. 2015, 115, 1760−1846. (763) Horvath, R.; Kobzi, B.; Keul, H.; Moeller, M.; Kiss, E. Molecular Interaction of a New Antibacterial Polymer with a 5484

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

Supported Lipid Bilayer Measured by an in Situ Label-Free Optical Technique. Int. J. Mol. Sci. 2013, 14, 9722−9736. (764) Lukanowska, M.; Howl, J.; Jones, S. Bioportides: Bioactive Cell-Penetrating Peptides That Modulate Cellular Dynamics. Biotechnol. J. 2013, 8, 918−930. (765) Fischer, P. M.; Zhelev, N. Z.; Wang, S.; Melville, J. E.; Fahraeus, R.; Lane, D. P. Structure-Activity Relationship of Truncated and Substituted Analogues of the Intracellular Delivery Vector Penetratin. J. Pept. Res. 2000, 55, 163−172. (766) Stewart, K. M.; Horton, K. L.; Kelley, S. O. Cell-Penetrating Peptides as Delivery Vehicles for Biology and Medicine. Org. Biomol. Chem. 2008, 6, 2242−2255. (767) Lindgren, M.; Hallbrink, M.; Prochiantz, A.; Langel, U. CellPenetrating Peptides. Trends Pharmacol. Sci. 2000, 21, 99−103. (768) Fonseca, S. B.; Pereira, M. P.; Kelley, S. O. Recent Advances in the Use of Cell-Penetrating Peptides for Medical and Biological Applications. Adv. Drug Delivery Rev. 2009, 61, 953−964. (769) Salamon, Z.; Lindblom, G.; Tollin, G. Plasmon-Waveguide Resonance and Impedance Spectroscopy Studies of the Interaction between Penetratin and Supported Lipid Bilayer Membranes. Biophys. J. 2003, 84, 1796−1807. (770) Hertel, C.; Terzi, E.; Hauser, N.; Jakob-Rotne, R.; Seelig, J.; Kemp, J. A. Inhibition of the Electrostatic Interaction between BetaAmyloid Peptide and Membranes Prevents Beta-Amyloid-Induced Toxicity. Proc. Natl. Acad. Sci. U. S. A. 1997, 94, 9412−9416. (771) Aguilar, M. I.; Small, D. H. Surface Plasmon Resonance for the Analysis of Beta-Amyloid Interactions and Fibril Formation in Alzheimer’s Disease Research. Neurotoxic. Res. 2005, 7, 17−27. (772) Murphy, R. M. Kinetics of Amyloid Formation and Membrane Interaction with Amyloidogenic Proteins. Biochim. Biophys. Acta, Biomembr. 2007, 1768, 1923−1934. (773) Hou, X.; Small, D. H.; Aguilar, M. I. Surface Plasmon Resonance Spectroscopy: A New Lead in Studying the Membrane Binding of Amyloidogenic Transthyretin. Methods Mol. Biol. 2011, 752, 215−228. (774) Small, D. H.; Maksel, D.; Kerr, M. L.; Ng, J.; Hou, X.; Chu, C.; Mehrani, H.; Unabia, S.; Azari, M. F.; Loiacono, R.; et al. The BetaAmyloid Protein of Alzheimer’s Disease Binds to Membrane Lipids but Does Not Bind to the Alpha7 Nicotinic Acetylcholine Receptor. J. Neurochem. 2007, 101, 1527−1538. (775) Cebecauer, M.; Hof, M.; Amaro, M. Impact of Gm1 on Membrane-Mediated Aggregation/Oligomerization of Beta-Amyloid: Unifying View. Biophys. J. 2017, 113, 1194−1199. (776) Matsuzaki, K. How Do Membranes Initiate Alzheimer’s Disease? Formation of Toxic Amyloid Fibrils by the Amyloid BetaProtein on Ganglioside Clusters. Acc. Chem. Res. 2014, 47, 2397− 2404. (777) Ariga, T.; Yu, R. K. Gm1 Inhibits Amyloid Beta-ProteinInduced Cytokine Release. Neurochem. Res. 1999, 24, 219−226. (778) Valdes-Gonzalez, T.; Inagawa, J.; Ido, T. Neuropeptides Interact with Glycolipid Receptors: A Surface Plasmon Resonance Study. Peptides 2001, 22, 1099−1106. (779) Zhai, J.; Lee, T. H.; Small, D. H.; Aguilar, M. I. Characterization of Early Stage Intermediates in the Nucleation Phase of Abeta Aggregation. Biochemistry 2012, 51, 1070−1078. (780) Devanathan, S.; Salamon, Z.; Lindblom, G.; Grobner, G.; Tollin, G. Effects of Sphingomyelin, Cholesterol and Zinc Ions on the Binding, Insertion and Aggregation of the Amyloid Abeta(1−40) Peptide in Solid-Supported Lipid Bilayers. FEBS J. 2006, 273, 1389− 1402. (781) Henry, S.; Vignaud, H.; Bobo, C.; Decossas, M.; Lambert, O.; Harte, E.; Alves, I. D.; Cullin, C.; Lecomte, S. Interaction of Abeta(1− 42) Amyloids with Lipids Promotes ″Off-Pathway″ Oligomerization and Membrane Damage. Biomacromolecules 2015, 16, 944−950. (782) Ta, H. P.; Berthelot, K.; Coulary-Salin, B.; Castano, S.; Desbat, B.; Bonnafous, P.; Lambert, O.; Alves, I.; Cullin, C.; Lecomte, S. A Yeast Toxic Mutant of Het-S Amyloid Disrupts Membrane Integrity. Biochim. Biophys. Acta, Biomembr. 2012, 1818, 2325−2334.

(783) Chi, X.; Kale, J.; Leber, B.; Andrews, D. W. Regulating Cell Death at, on, and in Membranes. Biochim. Biophys. Acta, Mol. Cell Res. 2014, 1843, 2100−2113. (784) Chipuk, J. E.; Green, D. R. How Do Bcl-2 Proteins Induce Mitochondrial Outer Membrane Permeabilization? Trends Cell Biol. 2008, 18, 157−164. (785) Leber, B.; Lin, J.; Andrews, D. W. Embedded Together: The Life and Death Consequences of Interaction of the Bcl-2 Family with Membranes. Apoptosis 2007, 12, 897−911. (786) Czabotar, P. E.; Lessene, G.; Strasser, A.; Adams, J. M. Control of Apoptosis by the Bcl-2 Protein Family: Implications for Physiology and Therapy. Nat. Rev. Mol. Cell Biol. 2014, 15, 49−63. (787) Peña-Blanco, A.; García-Saéz, A. J. Bax, Bak and Beyond Mitochondrial Performance in Apoptosis. FEBS J. 2018, 285, 416. (788) Billen, L. P.; Kokoski, C. L.; Lovell, J. F.; Leber, B.; Andrews, D. W. Bcl-Xl Inhibits Membrane Permeabilization by Competing with Bax. PLoS Biol. 2008, 6, e147. (789) Lovell, J. F.; Billen, L. P.; Bindner, S.; Shamas-Din, A.; Fradin, C.; Leber, B.; Andrews, D. W. Membrane Binding by Tbid Initiates an Ordered Series of Events Culminating in Membrane Permeabilization by Bax. Cell 2008, 135, 1074−1084. (790) Leber, B.; Lin, J.; Andrews, D. W. Still Embedded Together Binding to Membranes Regulates Bcl-2 Protein Interactions. Oncogene 2010, 29, 5221−5230. (791) Rosenbaum, D. M.; Rasmussen, S. G.; Kobilka, B. K. The Structure and Function of G-Protein-Coupled Receptors. Nature 2009, 459, 356−363. (792) Locatelli-Hoops, S.; Yeliseev, A. A.; Gawrisch, K.; Gorshkova, I. Surface Plasmon Resonance Applied to G Protein-Coupled Receptors. Biomed. Spectrosc. Imaging 2013, 2, 155−181. (793) Aristotelous, T.; Hopkins, A. L.; Navratilova, I. Surface Plasmon Resonance Analysis of Seven-Transmembrane Receptors. Methods Enzymol. 2015, 556, 499−525. (794) Harding, P. J.; Hadingham, T. C.; McDonnell, J. M.; Watts, A. Direct Analysis of a Gpcr-Agonist Interaction by Surface Plasmon Resonance. Eur. Biophys. J. 2006, 35, 709−712. (795) Karlsson, O. P.; Lofas, S. Flow-Mediated on-Surface Reconstitution of G-Protein Coupled Receptors for Applications in Surface Plasmon Resonance Biosensors. Anal. Biochem. 2002, 300, 132−138. (796) Terrettaz, S.; Stora, T.; Duschl, C.; Vogel, H. Protein Binding to Supported Lipid Membranes: Investigation of the Cholera ToxinGanglioside Interaction by Simultaneous Impedance Spectrosopy and Surface Plasmon Spectroscopy. Langmuir 1993, 9, 1361−1369. (797) Hoffman, T. L.; Canziani, G.; Jia, L.; Rucker, J.; Doms, R. W. A Biosensor Assay for Studying Ligand-Membrane Receptor Interactions: Binding of Antibodies and Hiv-1 Env to Chemokine Receptors. Proc. Natl. Acad. Sci. U. S. A. 2000, 97, 11215−11220. (798) Hüttenrauch, F.; Nitzki, A.; Lin, F. T.; Höning, S.; Oppermann, M. Beta-Arrestin Binding to Cc Chemokine Receptor 5 Requires Multiple C-Terminal Receptor Phosphorylation Sites and Involves a Conserved Asp-Arg-Tyr Sequence Motif. J. Biol. Chem. 2002, 277, 30769−30777. (799) Stenlund, P.; Babcock, G. J.; Sodroski, J.; Myszka, D. G. Capture and Reconstitution of G Protein-Coupled Receptors on a Biosensor Surface. Anal. Biochem. 2003, 316, 243−250. (800) Silin, V. I.; Karlik, E. A.; Ridge, K. D.; Vanderah, D. J. Development of Surface-Based Assays for Transmembrane Proteins: Selective Immobilization of Functional Ccr5, a G Protein-Coupled Receptor. Anal. Biochem. 2006, 349, 247−253. (801) Navratilova, I.; Sodroski, J.; Myszka, D. G. Solubilization, Stabilization, and Purification of Chemokine Receptors Using Biosensor Technology. Anal. Biochem. 2005, 339, 271−281. (802) Navratilova, I.; Pancera, M.; Wyatt, R. T.; Myszka, D. G. A Biosensor-Based Approach toward Purification and Crystallization of G Protein-Coupled Receptors. Anal. Biochem. 2006, 353, 278−283. (803) Rich, R. L.; Miles, A. R.; Gale, B. K.; Myszka, D. G. Detergent Screening of a G-Protein-Coupled Receptor Using Serial and Array Biosensor Technologies. Anal. Biochem. 2009, 386, 98−104. 5485

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

(804) Navratilova, I.; Dioszegi, M.; Myszka, D. G. Analyzing Ligand and Small Molecule Binding Activity of Solubilized Gpcrs Using Biosensor Technology. Anal. Biochem. 2006, 355, 132−139. (805) Navratilova, I.; Besnard, J.; Hopkins, A. L. Screening for Gpcr Ligands Using Surface Plasmon Resonance. ACS Med. Chem. Lett. 2011, 2, 549−554. (806) Shepherd, C. A.; Hopkins, A. L.; Navratilova, I. Fragment Screening by Spr and Advanced Application to Gpcrs. Prog. Biophys. Mol. Biol. 2014, 116, 113−123. (807) Huber, S.; Casagrande, F.; Hug, M. N.; Wang, L.; Heine, P.; Kummer, L.; Pluckthun, A.; Hennig, M. Spr-Based Fragment Screening with Neurotensin Receptor 1 Generates Novel Small Molecule Ligands. PLoS One 2017, 12, e0175842. (808) Harding, P. J.; Attrill, H.; Ross, S.; Koeppe, J. R.; Kapanidis, A. N.; Watts, A. Neurotensin Receptor Type 1: Escherichia Coli Expression, Purification, Characterization and Biophysical Study. Biochem. Soc. Trans. 2007, 35, 760−763. (809) Kaiser, L.; Graveland-Bikker, J.; Steuerwald, D.; Vanberghem, M.; Herlihy, K.; Zhang, S. Efficient Cell-Free Production of Olfactory Receptors: Detergent Optimization, Structure, and Ligand Binding Analyses. Proc. Natl. Acad. Sci. U. S. A. 2008, 105, 15726−15731. (810) Zou, C.; Kumaran, S.; Markovic, S.; Walser, R.; Zerbe, O. Studies of the Structure of the N-Terminal Domain from the Y4 Receptor - a G Protein-Coupled Receptor - and Its Interaction with Hormones from the Npy Family. ChemBioChem 2008, 9, 2276−2284. (811) Zhukov, A.; Andrews, S. P.; Errey, J. C.; Robertson, N.; Tehan, B.; Mason, J. S.; Marshall, F. H.; Weir, M.; Congreve, M. Biophysical Mapping of the Adenosine A2a Receptor. J. Med. Chem. 2011, 54, 4312−4323. (812) Congreve, M.; Rich, R. L.; Myszka, D. G.; Figaroa, F.; Siegal, G.; Marshall, F. H. Fragment Screening of Stabilized G-ProteinCoupled Receptors Using Biophysical Methods. Methods Enzymol. 2011, 493, 115−136. (813) Christopher, J. A.; Brown, J.; Dore, A. S.; Errey, J. C.; Koglin, M.; Marshall, F. H.; Myszka, D. G.; Rich, R. L.; Tate, C. G.; Tehan, B. Biophysical Fragment Screening of the Beta1-Adrenergic Receptor: Identification of High Affinity Arylpiperazine Leads Using StructureBased Drug Design. J. Med. Chem. 2013, 56, 3446−3455. (814) Botelho, A. V.; Gibson, N. J.; Thurmond, R. L.; Wang, Y.; Brown, M. F. Conformational Energetics of Rhodopsin Modulated by Nonlamellar-Forming Lipids. Biochemistry 2002, 41, 6354−6368. (815) Brown, M. F. Modulation of Rhodopsin Function by Properties of the Membrane Bilayer. Chem. Phys. Lipids 1994, 73, 159−180. (816) Oates, J.; Watts, A. Uncovering the Intimate Relationship between Lipids, Cholesterol and Gpcr Activation. Curr. Opin. Struct. Biol. 2011, 21, 802−807. (817) Salamon, Z.; Tollin, G.; Alves, I.; Hruby, V. Chapter 6. Plasmon Resonance Methods in Membrane Protein Biology Applications to Gpcr Signaling. Methods Enzymol. 2009, 461, 123− 146. (818) Salamon, Z.; Brown, M. F.; Tollin, G. Plasmon Resonance Spectroscopy: Probing Molecular Interactions within Membranes. Trends Biochem. Sci. 1999, 24, 213−219. (819) Salamon, Z.; Cowell, S.; Varga, E.; Yamamura, H. I.; Hruby, V. J.; Tollin, G. Plasmon Resonance Studies of Agonist/Antagonist Binding to the Human Delta-Opioid Receptor: New Structural Insights into Receptor-Ligand Interactions. Biophys. J. 2000, 79, 2463−2474. (820) Salamon, Z.; Devanathan, S.; Alves, I. D.; Tollin, G. PlasmonWaveguide Resonance Studies of Lateral Segregation of Lipids and Proteins into Microdomains (Rafts) in Solid-Supported Bilayers. J. Biol. Chem. 2005, 280, 11175−11184. (821) Salamon, Z.; Wang, Y.; Tollin, G.; Macleod, H. A. Assembly and Molecular Organization of Self-Assembled Lipid Bilayers on Solid Substrates Monitored by Surface Plasmon Resonance Spectroscopy. Biochim. Biophys. Acta, Biomembr. 1994, 1195, 267−275. (822) Calmet, P.; De Maria, M.; Harte, E.; Lamb, D.; Serrano-Vega, M.; Jazayeri, A.; Tschammer, N.; Alves, I. D. Real Time Monitoring of

Membrane Gpcr Reconstitution by Plasmon Waveguide Resonance: On the Role of Lipids. Sci. Rep. 2016, 6, 36181. (823) Alves, I. D.; Ciano, K. A.; Boguslavski, V.; Varga, E.; Salamon, Z.; Yamamura, H. I.; Hruby, V. J.; Tollin, G. Selectivity, Cooperativity, and Reciprocity in the Interactions between the Delta-Opioid Receptor, Its Ligands, and G-Proteins. J. Biol. Chem. 2004, 279, 44673−44682. (824) Alves, I. D.; Cowell, S. M.; Salamon, Z.; Devanathan, S.; Tollin, G.; Hruby, V. J. Different Structural States of the Proteolipid Membrane Are Produced by Ligand Binding to the Human DeltaOpioid Receptor as Shown by Plasmon-Waveguide Resonance Spectroscopy. Mol. Pharmacol. 2004, 65, 1248−1257. (825) Alves, I. D.; Salgado, G. F.; Salamon, Z.; Brown, M. F.; Tollin, G.; Hruby, V. J. Phosphatidylethanolamine Enhances Rhodopsin Photoactivation and Transducin Binding in a Solid Supported Lipid Bilayer as Determined Using Plasmon-Waveguide Resonance Spectroscopy. Biophys. J. 2005, 88, 198−210. (826) Jastrzebska, B.; Debinski, A.; Filipek, S.; Palczewski, K. Role of Membrane Integrity on G Protein-Coupled Receptors: Rhodopsin Stability and Function. Prog. Lipid Res. 2011, 50, 267−277. (827) Guixa-Gonzalez, R.; Albasanz, J. L.; Rodriguez-Espigares, I.; Pastor, M.; Sanz, F.; Marti-Solano, M.; Manna, M.; Martinez-Seara, H.; Hildebrand, P. W.; Martín, M.; et al. Membrane Cholesterol Access into a G-Protein-Coupled Receptor. Nat. Commun. 2017, 8, 14505. (828) Venkatakrishnan, A. J.; Deupi, X.; Lebon, G.; Tate, C. G.; Schertler, G. F.; Babu, M. M. Molecular Signatures of G-ProteinCoupled Receptors. Nature 2013, 494, 185−194. (829) Huynh, J.; Thomas, W. G.; Aguilar, M. I.; Pattenden, L. K. Role of Helix 8 in G Protein-Coupled Receptors Based on StructureFunction Studies on the Type 1 Angiotensin Receptor. Mol. Cell. Endocrinol. 2009, 302, 118−127. (830) de Gasparo, M.; Catt, K. J.; Inagami, T.; Wright, J. W.; Unger, T. International Union of Pharmacology. XXIII. The Angiotensin II Receptors. Pharmacol. Rev. 2000, 52, 415−472. (831) Hunyady, L.; Catt, K. J. Pleiotropic At1 Receptor Signaling Pathways Mediating Physiological and Pathogenic Actions of Angiotensin Ii. Mol. Endocrinol. 2006, 20, 953−970. (832) Mozsolits, H.; Unabia, S.; Ahmad, A.; Morton, C. J.; Thomas, W. G.; Aguilar, M. I. Electrostatic and Hydrophobic Forces Tether the Proximal Region of the Angiotensin Ii Receptor (At1a) Carboxyl Terminus to Anionic Lipids. Biochemistry 2002, 41, 7830−7840. (833) Kamimori, H.; Unabia, S.; Thomas, W. G.; Aguilar, M. I. Evaluation of the Membrane-Binding Properties of the Proximal Region of the Angiotensin Ii Receptor (At1a) Carboxyl Terminus by Surface Plasmon Resonance. Anal. Sci. 2005, 21, 171−174. (834) Lemmon, M. A. Membrane Recognition by PhospholipidBinding Domains. Nat. Rev. Mol. Cell Biol. 2008, 9, 99−111. (835) Heo, W. D.; Inoue, T.; Park, W. S.; Kim, M. L.; Park, B. O.; Wandless, T. J.; Meyer, T. Pi(3,4,5)P3 and Pi(4,5)P2 Lipids Target Proteins with Polybasic Clusters to the Plasma Membrane. Science 2006, 314, 1458−1461. (836) Li, L.; Shi, X.; Guo, X.; Li, H.; Xu, C. Ionic Protein-Lipid Interaction at the Plasma Membrane: What Can the Charge Do? Trends Biochem. Sci. 2014, 39, 130−140. (837) Di Paolo, G.; De Camilli, P. Phosphoinositides in Cell Regulation and Membrane Dynamics. Nature 2006, 443, 651−657. (838) Golebiewska, U.; Gambhir, A.; Hangyas-Mihalyne, G.; Zaitseva, I.; Radler, J.; McLaughlin, S. Membrane-Bound Basic Peptides Sequester Multivalent (Pip2), but Not Monovalent (Ps), Acidic Lipids. Biophys. J. 2006, 91, 588−599. (839) Sönnichsen, C.; Franzl, T.; Wilk, T.; von Plessen, G.; Feldmann, J.; Wilson, O.; Mulvaney, P. Drastic Reduction of Plasmon Damping in Gold Nanorods. Phys. Rev. Lett. 2002, 88, 077402. (840) Galush, W. J.; Shelby, S. A.; Mulvihill, M. J.; Tao, A.; Yang, P.; Groves, J. T. A Nanocube Plasmonic Sensor for Molecular Binding on Membrane Surfaces. Nano Lett. 2009, 9, 2077−2082. (841) Bruzas, I.; Unser, S.; Yazdi, S.; Ringe, E.; Sagle, L. Ultrasensitive Plasmonic Platform for Label-Free Detection of Membrane-Associated Species. Anal. Chem. 2016, 88, 7968−7974. 5486

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487

Chemical Reviews

Review

(842) Faustini, M.; Louis, B.; Albouy, P. A.; Kuemmel, M.; Grosso, D. Preparation of Sol−Gel Films by Dip-Coating in Extreme Conditions. J. Phys. Chem. C 2010, 114, 7637−7645. (843) Kambhampati, D. K.; Jakob, T. A. M.; Robertson, J. W.; Cai, M.; Pemberton, J. E.; Knoll, W. Novel Silicon Dioxide Sol−Gel Films for Potential Sensor Applications: A Surface Plasmon Resonance Study. Langmuir 2001, 17, 1169−1175. (844) Im, H.; Lee, S. H.; Wittenberg, N. J.; Johnson, T. W.; Lindquist, N. C.; Nagpal, P.; Norris, D. J.; Oh, S. H. TemplateStripped Smooth Ag Nanohole Arrays with Silica Shells for Surface Plasmon Resonance Biosensing. ACS Nano 2011, 5, 6244−6253. (845) Dahlin, A. B.; Jonsson, M. P.; Höök, F. Specific Self-Assembly of Single Lipid Vesicles in Nanoplasmonic Apertures in Gold. Adv. Mater. 2008, 20, 1436−1442. (846) Jonsson, M. P.; Jonsson, P.; Dahlin, A. B.; Hook, F. Supported Lipid Bilayer Formation and Lipid-Membrane-Mediated Biorecognition Reactions Studied with a New Nanoplasmonic Sensor Template. Nano Lett. 2007, 7, 3462−3468. (847) Lee, B. K.; Lee, H. Y.; Kim, P.; Suh, K. Y.; Kawai, T. Nanoarrays of Tethered Lipid Bilayer Rafts on Poly(Vinyl Alcohol) Hydrogels. Lab Chip 2009, 9, 132−139. (848) Im, H.; Wittenberg, N. J.; Lesuffleur, A.; Lindquist, N. C.; Oh, S. H. Membrane Protein Biosensing with Plasmonic Nanopore Arrays and Pore-Spanning Lipid Membranes. Chem. Sci. 2010, 1, 688−696. (849) Wittenberg, N. J.; Im, H.; Xu, X.; Wootla, B.; Watzlawik, J.; Warrington, A. E.; Rodriguez, M.; Oh, S. H. High-Affinity Binding of Remyelinating Natural Autoantibodies to Myelin-Mimicking Lipid Bilayers Revealed by Nanohole Surface Plasmon Resonance. Anal. Chem. 2012, 84, 6031−6039. (850) Xu, X.; Denic, A.; Jordan, L. R.; Wittenberg, N. J.; Warrington, A. E.; Wootla, B.; Papke, L. M.; Zoecklein, L. J.; Yoo, D.; Shaver, J.; et al. A Natural Human Igm That Binds to Gangliosides Is Therapeutic in Murine Models of Amyotrophic Lateral Sclerosis. Dis. Models & Mech. 2015, 8, 831−842. (851) Kumar, K.; Dahlin, A. B.; Sannomiya, T.; Kaufmann, S.; Isa, L.; Reimhult, E. Embedded Plasmonic Nanomenhirs as Location-Specific Biosensors. Nano Lett. 2013, 13, 6122−6129. (852) Jonsson, M. P.; Jonsson, P.; Hook, F. Simultaneous Nanoplasmonic and Quartz Crystal Microbalance Sensing: Analysis of Biomolecular Conformational Changes and Quantification of the Bound Molecular Mass. Anal. Chem. 2008, 80, 7988−7995. (853) Dacic, M.; Jackman, J. A.; Yorulmaz, S.; Zhdanov, V. P.; Kasemo, B.; Cho, N. J. Influence of Divalent Cations on Deformation and Rupture of Adsorbed Lipid Vesicles. Langmuir 2016, 32, 6486− 6495. (854) Ferhan, A. R.; Jackman, J. A.; Cho, N. J. Integration of Quartz Crystal Microbalance-Dissipation and Reflection-Mode Localized Surface Plasmon Resonance Sensors for Biomacromolecular Interaction Analysis. Anal. Chem. 2016, 88, 12524−12531. (855) Ferhan, A. R.; Jackman, J. A.; Cho, N. J. Investigating How Vesicle Size Influences Vesicle Adsorption on Titanium Oxide: A Competition between Steric Packing and Shape Deformation. Phys. Chem. Chem. Phys. 2017, 19, 2131−2139. (856) Jackman, J. A.; Choi, J. H.; Zhdanov, V. P.; Cho, N. J. Influence of Osmotic Pressure on Adhesion of Lipid Vesicles to Solid Supports. Langmuir 2013, 29, 11375−11384. (857) Jackman, J. A.; Yorulmaz Avsar, S.; Ferhan, A. R.; Li, D.; Park, J. H.; Zhdanov, V. P.; Cho, N. J. Quantitative Profiling of Nanoscale Liposome Deformation by a Localized Surface Plasmon Resonance Sensor. Anal. Chem. 2017, 89, 1102−1109. (858) Jackman, J. A.; Zhdanov, V. P.; Cho, N. J. Nanoplasmonic Biosensing for Soft Matter Adsorption: Kinetics of Lipid Vesicle Attachment and Shape Deformation. Langmuir 2014, 30, 9494−9503. (859) Oh, E.; Jackman, J. A.; Yorulmaz, S.; Zhdanov, V. P.; Lee, H.; Cho, N. J. Contribution of Temperature to Deformation of Adsorbed Vesicles Studied by Nanoplasmonic Biosensing. Langmuir 2015, 31, 771−781.

(860) Mey, I.; Stephan, M.; Schmitt, E. K.; Muller, M. M.; Ben Amar, M.; Steinem, C.; Janshoff, A. Local Membrane Mechanics of PoreSpanning Bilayers. J. Am. Chem. Soc. 2009, 131, 7031−7039. (861) Korman, C. E.; Megens, M.; Ajo-Franklin, C. M.; Horsley, D. A. Nanopore-Spanning Lipid Bilayers on Silicon Nitride Membranes That Seal and Selectively Transport Ions. Langmuir 2013, 29, 4421− 4425. (862) Kant, K.; Priest, C.; Shapter, J. G.; Losic, D. The Influence of Nanopore Dimensions on the Electrochemical Properties of Nanopore Arrays Studied by Impedance Spectroscopy. Sensors 2014, 14, 21316− 21328. (863) Osaki, T.; Takeuchi, S. Artificial Cell Membrane Systems for Biosensing Applications. Anal. Chem. 2017, 89, 216−231. (864) Lesuffleur, A.; Im, H.; Lindquist, N. C.; Lim, K. S.; Oh, S. H. Laser-Illuminated Nanohole Arrays for Multiplex Plasmonic Microarray Sensing. Opt. Express 2008, 16, 219−224. (865) Lindquist, N. C.; Lesuffleur, A.; Im, H.; Oh, S. H. Sub-Micron Resolution Surface Plasmon Resonance Imaging Enabled by Nanohole Arrays with Surrounding Bragg Mirrors for Enhanced Sensitivity and Isolation. Lab Chip 2009, 9, 382−387. (866) Im, H.; Lesuffleur, A.; Lindquist, N. C.; Oh, S. H. Plasmonic Nanoholes in a Multichannel Microarray Format for Parallel Kinetic Assays and Differential Sensing. Anal. Chem. 2009, 81, 2854−2859. (867) Kuziemko, G. M.; Stroh, M.; Stevens, R. C. Cholera Toxin Binding Affinity and Specificity for Gangliosides Determined by Surface Plasmon Resonance. Biochemistry 1996, 35, 6375−6384. (868) MacKenzie, C. R.; Hirama, T.; Lee, K. K.; Altman, E.; Young, N. M. Quantitative Analysis of Bacterial Toxin Affinity and Specificity for Glycolipid Receptors by Surface Plasmon Resonance. J. Biol. Chem. 1997, 272, 5533−5538. (869) Shi, J.; Yang, T.; Kataoka, S.; Zhang, Y.; Diaz, A. J.; Cremer, P. S. Gm1 Clustering Inhibits Cholera Toxin Binding in Supported Phospholipid Membranes. J. Am. Chem. Soc. 2007, 129, 5954−5961.

5487

DOI: 10.1021/acs.chemrev.7b00729 Chem. Rev. 2018, 118, 5392−5487