Confocal Raman Microscopy Investigation of Self ... - ACS Publications

Subscriber access provided by University Library of Southern Denmark

Article

Confocal Raman Microscopy Investigation of SelfAssembly of Hybrid Phospholipid Bilayers within Individual Porous Silica Chromatographic Particles Jay P. Kitt, David A. Bryce, Shelley D. Minteer, and Joel M. Harris Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.9b01359 • Publication Date (Web): 14 May 2019 Downloaded from http://pubs.acs.org on May 14, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Confocal Raman Microscopy Investigation of Self-Assembly of Hybrid Phospholipid Bilayers within Individual Porous Silica Chromatographic Particles Jay P. Kitt, David A. Bryce, Shelley D. Minteer, and Joel M. Harris* Department of Chemistry, University of Utah, 315 South 1400 East Salt Lake City, UT 84112-0850 USA Abstract Hybrid-supported phospholipid bilayers are a model structure utilized for measurement of molecular interactions that typically occur at cell membranes. These membrane models are prepared by adsorption of a lipid monolayer onto a stable n-alkyl chain layer that is covalently-bound to a support surface. Hybrid bilayers have been adapted to chromatographic retention measurements of lipophilicity through the assembly of a phospholipid monolayer onto n-alkane-modified silica surfaces in reversed-phase chromatographic particles. Recent Raman microscopy studies of these particles have shown that the acyl chains of the phospholipid interact with the C18-alkyl chains immobilized on the silica surface, where both lipid and C18 alkyl chains become ordered due to chain interdigitation. Confocal Raman microscopy has also been used to investigate the association of small molecules with hybrid-lipid bilayers in C18 chromatographic silica particles; the partitioning of model solutes compares favorably to that in lipid vesicle membranes with similar changes in acyl-chain structure (disordering) with solute partitioning. The present study seeks information about how these membrane-mimetic bilayers assemble onto the C18–derivatized silica surfaces of reversed-phase chromatographic silica particles. Confocal Raman microscopy is capable of interrogating the time-dependent internal composition and structure within individual silica particles. The Raman scattering data can be resolved into component Raman spectra and corresponding composition vectors that describe the timedependent changes in intensity of the component spectra. This analysis provides insight into how the structures of both the lipid and C18 alkyl chains of hybrid lipid bilayers evolve during deposition and organization on the internal surfaces of reversed-phase chromatographic silica particles. * Corresponding author: [email protected]

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 19

2 INTRODUCTION Hybrid-supported phospholipid bilayers (hybrid bilayers) are a stable model system used to investigate how small-molecules or proteins interact with phospholipid membranes.1,2 These model lipid bilayers are formed by self-assembly of an upper leaflet of phospholipid onto a lower leaflet of surface-bound n-alkane chains. Hybrid bilayers were first assembled onto n-alkane thiol monolayers on gold surfaces3 and characterized with infrared reflection-absorption spectroscopy,1,4 surface plasmon resonance,5,6 and surface-enhanced Raman spectroscopy7,8 measurements. Gold-substrate-supported hybrid bilayers have been well-characterized,1,9 where infrared spectroscopy and neutron reflectivity revealed the lower leaflet of n-alkane thiol (HSC18) to be an ordered, self-assembled monolayer at the gold surface, to which phospholipid adsorbs to form a well-ordered upper-leaflet with a thickness and lipid headgroup spacing that are in close agreement with vesicle bilayers.9 The application of hybrid lipid bilayers has been extended to chromatographic retention measurements of lipid-membrane affinity of small molecules and peptides.2,10-12 Hybrid bilayers for chromatographic retention measurements have been prepared by adsorption of phospholipids onto the n-alkane-modified silica surfaces in reversed-phase chromatographic particles. These hybrid bilayers are formed within the nanometer-scale pores of chromatographic silica by adsorption of phospholipid from aqueous solution. The surface tension between an aqueous solution and C18–modified silica surface represents a significant barrier to water being able to wet and penetrate the nanometer-scale pores,13,14 where water porosimetry indicates that wetting (Laplace) pressures can be as high as ~20 MPa.14 To overcome this barrier, an organic modifier (typically isopropanol or methanol) is included in the phospholipid dispersion for hybrid-bilayer assembly,2,10-12 which lowers the interfacial tension and allows the solution to wet and penetrate the interior hydrophobic surfaces of the nanometer-scale pores. Following deposition of lipid onto the C18–modified silica surfaces, the mobile phase composition is switched to water or buffer so that the partitioning of analytes into the immobilized lipid and their corresponding chromatographic retention reflect biologically-relevant conditions. The resulting retention measurements of small molecules and peptides on lipid-modified chromatographic silica columns have been in agreement with measurements of the distribution coefficients of the same molecules into membranes of phospholipid vesicles dispersions,2,10,15 suggesting that these ACS Paragon Plus Environment

Page 3 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


3 supported bilayers are a reasonable model for phospholipid vesicle bilayers. In the original development of lipid bilayers in reversed-phase chromatographic silica, the structure of the immobilized lipid was uncertain, where the earliest conjecture was that lipid vesicles (liposomes) remained intact as they adsorbed the n-alkane-modified silica surface.16 The long-term stability of the lipid-modified stationary phase and the corresponding reproducibility of solute retention over weeks of use suggested to Tsirkin and Grushka10 that phospholipids were not adsorbed intact liposomes but rather a self-assembled lipid monolayer with lipid acyl chains oriented toward the hydrophobic bonded phase and the polar head-groups in contact with water. To determine the organization of lipid layers deposited onto the interior surfaces of reversedphase chromatographic silica, it is helpful to have an in situ spectroscopic measurement of lipid bilayer structure. Raman spectroscopy has previously been employed in situ to probe the composition and structure of reversed-phase chromatographic interfaces,17-24 and this approach was recently used in a confocal-microscopy experiment to probe the structure of hybrid lipid bilayers formed within individual chromatographic silica particles.25 Confocal Raman microscopy is well suited to this task because it can probe the internal interfacial composition and structure within individual chromatographic particles; the results showed that the acyl chains of the phospholipid indeed interact with the C18 alkyl chains bound to the silica surface, as suggested by Tsirkin and Grushka.10 The hydrocarbon chains were found to be highly ordered, and this ordering arises from chain interdigitation, as indicated by the surface-density of lipid in the hybrid bilayer, the nearly all-trans conformations of both C18 and lipid alkyl chains upon bilayer formation, and the decoupling of C18 methylene C-H vibrations by deuterated (C-D labeled) lipid acyl chains residing between C18 chains that disrupt C-H dipolar coupling.25 Further insight into the hybrid-bilayer structure was found in temperature-dependent Raman spectra, where a single melting transition was observed that is broader and shifted to higher temperature compared to a lipid vesicle. This result is consistent with an interdigitated structure, where C18 and lipid chains melt together and where the immobility of the covalently-bound C18chains decreases the phase-transition cooperativity and increases the melting temperature compared to a phospholipid vesicle membrane. Confocal Raman microscopy has also been used to investigate the partitioning of small molecules into hybrid-lipid bilayers in C18 chromatographic silica particles compared to their ACS Paragon Plus Environment


Page 4 of 19

4 partitioning into lipid vesicle membranes.26 While there are structural differences between hybrid (interdigitated) versus vesicle (non-interdigitated) lipid bilayers, a comparison of the pHdependent partitioning of a model solute (2-(4-isobutylphenyl) propionic acid) showed an equivalent response at low pH for the neutral form of the solute; at high pH, the deprotonated (anionic) form of the solute showed somewhat greater partitioning into the hybrid-bilayer. This difference likely arises from greater headgroup spacing of the hybrid bilayer that arises from interdigitation,25 which increases solution-phase contact with hydrophobic regions of the interface and contributes to hybrid bilayer association of the surface-active anionic solute by its lowering of the interfacial tension. Confocal Raman microscopy also revealed that in response to partitioning of the model solute, hybrid-supported bilayers exhibit very similar changes in alkylchain structure (disordering) as a vesicle bilayer; thus, despite differences in chain conformation, hybrid-bilayers are a reasonable membrane model of both small-molecule partitioning and corresponding structural response of the lipid bilayer. The above confocal-Raman microscopy investigations have provided insights into the structure of hybrid-lipid bilayers assembled on the C18–modified silica surfaces that are relevant to their applications chromatographic and spectroscopic studies of small-molecule lipophilicity. Information about how these membrane-mimetic hybrid bilayers assemble onto the internal surfaces of reversed-phase chromatographic particles, however, is currently not known. In the present work, confocal Raman microscopy is employed to investigate the time-dependent selfassembly of hybrid supported phospholipid bilayers within individual chromatographic particles. Raman spectra are analyzed using a multivariate statistical technique, self-modeling curve resolution (SMCR),27,28 which allows resolution of the time-dependent Raman scattering data into component Raman spectra and corresponding composition vectors that describe the timedependent changes in intensity of the component spectra. This methodology has been previously used to detect temperature-dependent structural changes in phospholipid membranes of individual optically-trapped lipid vesicles29 and in fully-assembled hybrid bilayers in silica particles.25 In the present work, this analysis provides insight into how the structures of both the lipid and C18 chains of hybrid lipid bilayers evolve during their deposition and organization into ordered structures on reversed-phase chromatographic silica surfaces.


Page 5 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


5 EXPERIMENTAL SECTION Reagents and Materials. Methanol and isopropanol (Optima, ≥99%) were acquired from Fischer Scientific (Waltham, MA). Chloroform (HPLC, ≥99.9%) was obtained from SigmaAldrich (St. Louis, MO). Water used in all experiments was filtered using a Barnstead GenPure UV filtration system and had a minimum resistivity of 18 MΩ∙cm. 1,2-dimyristoyl-sn-glycero-3phosphocholine (DMPC, >99%) and 1,2-dimyristoyl-sn-glycero-3-phosphocholine-d54 (DMPCD54, >99%) were obtained from Avanti Polar Lipids (Alabaster, AL). DMPC samples were diluted in chloroform and stored at -20 °C until use. Solutions for forming hybrid-bilayers were prepared by drying aliquots of DMPC or DMPC-D54 in chloroform under a stream of nitrogen followed by drying under vacuum (100-mTorr) for ~2 h. The lipid was then dissolved in isopropanol and diluted with water to a final isopropanol concentration of 15%. The final concentration of DMPC was 5 mg/mL. Flow cells for time-dependent hybrid-bilayer formation experiments were constructed by drilling a pair of 2.5 mm holes approximately 1-cm apart in a 25-mm diameter, 3-mm-thick Pyrex glass top plate (VWR, Radnor, PA). The top plate was adapted for connecting tubing by gluing 2.4-mm o.d. Luer adapters (Value Plastics, Inc., Fort Collins, CO) into the holes using Devcon 5-minute epoxy (ITW Devcon, Danvers, MA). The top plate is attached to a 22-mm x 22-mm No. 1 glass coverslip (Gold Seal, Erie Scientific Co., Portsmouth, NH) using 140 μm thick double stick tape (TapeCase Ltd., Elk Grove Village, IL). The ‘cell’ portion of the flow cell is created by cutting a 2.5-mm-by-11-mm channel between the Luer adapters prior attaching the tape to the coverslip. A diagram of the microscope flow cell has been published previously.24 Tubing for the flow experiments was 1.6-mm-i.d., 2.4-mm-o.d. Viton elastomer (Cole-Parmer, Vernon Hills, IL). Confocal Raman microscopy. The confocal Raman microscope used in this work has been described in detail in earlier work.30 A brief description including minor updates to the components is presented here. The 647 nm beam of a Kr+ laser (Innova 90, Coherent, Inc., Santa Clara, CA) operating at 50 mW is passed through a band-pass filter (LL01-647-12.5, Semrock, Lake Forest, IL), a 4× beam expander (50-25-4X-657, Special Optics Inc., Wharton, NJ), and directed into the back of a Nikon TE 300 (Nikon, Tokyo, Japan) inverted fluorescence microscope. The beam is then reflected off of a dichroic beamsplitter (Di03-R635, Semrock, ACS Paragon Plus Environment


Page 6 of 19

6 Lake Forest, IL) to slightly overfill the back aperture of a 100×, 1.4 N.A., microscope objective (Plan APO VC, Nikon, Tokyo, Japan). Scattered light is collected using the same objective and Rayleigh scattered light is filtered by passing the collected light through the same dichroic beamsplitter and a holographic notch filter (647 nm, Kaiser Optical Systems, Ann Arbor, MI). Raman scattered light is then focused onto the 50-µm entrance slit of a Chromex (Bruker) 500is (Bruker, Billerica, MA), dispersed by a 300 lines/mm grating blazed at 750 nm, and focused onto a charge couple device (CCD) camera (iDus DU401A, Andor, Belfast, UK). Confocal microscopy is achieved by defining a ‘pinhole’ aperture using the entrance slit of the monochromator (set to 50 µm) to define the aperture in the horizontal dimension and by binning three rows of pixels (78 µm) on the CCD camera to define the aperture in the vertical dimension. The resulting confocal probe is ~1 fL in volume with a beam-waist diameter of ~600 nm. To measure time-resolved Raman spectra of hybrid bilayer formation within individual C18 chromatographic particles, a solution of particles, pre-wetted in methanol, was injected into a two-port flow cell31 and particles were left a period of time (~1 hr) allowing them to settle onto the coverslip. By working under the inlet of the flow cell,32 horizontal-shear force on the particle was minimized allowing solution to be flowed through the cell without disturbing the particles which have settled on the coverslip. The confocal probe volume was centered within an individual particle by bringing the laser focus to a visible reflected spot at the coverslip-solution interface and translating the stage in the x and y dimensions until the reflected spot was directly beneath an individual particle; the focused-spot was then brought to the center of the particle by translating the objective upward in the z dimension until the edge of the particle came into tight focus. The solution was then swapped to ‘blank’ sample solution (aqueous 15% isopropanol) and the particle was allowed to equilibrate as confirmed by the absence of changes of the Raman scattering intensity of bands corresponding to surface-bound C18 alkyl chains and the isopropanol/water solution filling the pores in subsequent spectra. At this point, the solution was switched to a sample of hybrid-bilayer-formation solution comprised of 15% aqueous isopropanol containing 0.5 mg/mL DMPC. All experiments were carried out at room temperature (~23 ºC) Raman spectra from within the particle were collected at 30 s intervals until the spectra no longer varied in subsequent frames at which point flow was switched to water to rinse the DMPC solution from the cell and confirm the hybrid bilayer was localized to the particle interior. ACS Paragon Plus Environment

Page 7 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


7 During each step following the particle settling step, flow was maintained at a flow rate of 0.05 mL/min using a syringe pump (PHD 2000, Harvard Apparatus, Holliston, MA) to prevent depletion of DMPC from solution surrounding a particle. Following rinsing, a spectrum of the solution surrounding the particle was collected to confirm solution-phase DMPC was out of the cell and DMPC in the solution filling the pores did not contribute to the within-particle spectrum. Spectral data analysis. All spectra were baseline corrected using a custom Matlab (Mathworks, Natick, MA) script which subtracts a 5th-order polynomial function fit to portions of the spectrum that do not contain strong vibrational bands. To prevent baseline noise from contributing uncertainty to the results of spectral data analysis steps, the spectra were truncated to spectral regions that vary with accumulation of lipid.33 Analysis of the time-dependent spectra of hybrid-bilayer formation was carried out using a combination of self-modeling curve resolution and multivariate least-squares analysis as described in detail in the section below. For a more in-depth discussion of these methods, the reader is directed to previously publications of self-modeling curve resolution and multivariate least-squares techniques27,28 and their applications to the analysis of multicomponent Raman spectra.25,29,34 RESULTS AND DISCUSSION Confocal Raman microscopy of hybrid supported lipid bilayer self-assembly. To investigate hybrid-bilayer self-assembly within a reversed-phase chromatographic particle, the time-dependent changes in phospholipid and surface-bound C18-alkyl chains were monitored insitu as a function of time as lipid adsorbs to the C18-functionalized silica surface from solution (Figure 1). This experiment was accomplished by injecting a dispersion of C18 particles in methanol into a microscopy flow cell and allowing particles to settle and adhere to the coverslip surface. The confocal probe volume was translated into the center of an individual particle and the solution was switched to 15% v/v isopropanol/water. Raman scattering was collected as a function of time as the solution was switched to 15% v/v isopropanol/water containing 5 mg/mL DMPC. Once the spectra were no longer changing (indicating equilibrium DMPC surface coverage), the solution was switched to pure water.



Page 8 of 19

8 Examining Raman reporter bands indicative of acyl- and alkyl-chain structure ((C-C stretching (1030-1150 cm-1), C-H bending (1400-1500 cm-1), CH2 twisting (1260-1340 cm-1), and C-H stretching (2800-3100 cm-1)), increases in bands associated with disordered alkyl-chains such as the broad C-C stretching mode at 1086 cm-1, and the broad CH2-twisting mode at 1303 cm-1 are observed, along with headgroup CN-

Figure 1: Raman spectra collected in situ within an individual C18-silica particle as DMPC is adsorbed from solution to the C18-surface. The red region is the equilibrium region where the overlapping portions of the SMCR analysis were carried out.

stretching at 715 cm-1, confirming phospholipid adsorption. About midway through the lipidaccumulation process, indicators of alkyl-chain ordering begin to appear, most notably transconformer C-C stretching modes at 1061 cm-1 and 1126 cm-1, suggesting hybrid-bilayer ordering as the lipid coverage increases. A significant increase in alkyl-chain ordering is observed as the solution is switched from isopropanol/water to pure water. To analyze these spectra, where many peaks overlap and a model for the spectral changes as a function of time is unknown, we employ a multivariate statistical method, self-modeling curve resolution (SMCR), which allows resolution of correlated spectral changes without prior knowledge of a physical model of the process.25,27-29,34 Self-modeling curve resolution of the spectral data was carried out as follows: spectra collected as a function of time were organized into a w by t data matrix D where each row (w) corresponded to the measured Raman scattering intensity at a given wavenumber and each column (t) corresponded to the time at which the spectrum was collected. With the assumption that each spectrum can be represented by a linear combination of component spectra, where each component represents a structural form of the bilayer, the data matrix, D, is represented by D = AC, the product of a w by n component spectra matrix, A, and C, an n by c matrix of component concentrations, where rows correspond to the relative contribution of each component to the overall spectrum at each point in time. Self-modeling curve resolution begins with principal components analysis (PCA),28 where spectral data are decomposed into a matrix of orthonormal eigenvectors, Q, and respective ACS Paragon Plus Environment

Page 9 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


9 scores, λ. The eigenvectors in Q correspond to the correlated changes in the spectral data and the diagonal elements of the scores matrix, λ, indicate the relative magnitude of the contribution of each eigenvector to the spectral data. By an F-test of the scores together with examining the eigenvectors for significant low-frequency content,28 Q can be truncated to a smaller matrix of principal components, 𝐐, which capture the statistically significant correlation in the data while rejecting uncorrelated noise. The spectral data are then projected onto the composition eigenvectors, 𝐐, to obtain a matrix of principal-component spectral eigenvectors, U = D 𝐐. Both sets of abstract eigenvectors, U and 𝐐, can then be rotated from eigenvector space into real space by finding a rotation matrix K that generates matrices the A = U K-1 and C = K 𝐐T, corresponding to the component spectra and time-dependent composition vectors, respectively. The data, D, can then be represented by the product of A and C, where D = A C = U K-1K 𝐐T. SMCR analysis is used here to resolve spectral changes occurring during phospholipid accumulation on the C18 alkyl chains within a chromatographic particle. The analyzed spectra were limited to regions where structurally-informative spectral changes occur. Over the entire accumulation experiment, PCA indicated that there were four significant components. Because identifying a 16-element rotation matrix K for a four-component SMCR analysis is extremely challenging, the spectra were divided into two overlapping time-segments: a three-component early segment corresponding to adsorption of the phospholipid from isopropanol and a twocomponent later segment corresponding to the switch from isopropanol/water/DMPC to pure water. After completing the self-modeling curve resolution for each time segment, the component spectra in the overlapping isopropanol equilibration region were confirmed to be equivalent and were averaged; the resulting spectra of the four components, A (Figure 2A), were used in a matrix least-squares step to determine the time-dependent concentration vectors over the entire experiment, C = [ATA]-1ATD, and these are plotted in Figure 2B. From these results, the four-state (three-step) hybrid-bilayer self-assembly process proceeds as follows: phospholipid initially adsorbed to the C18 surface is disordered. In the plot of all the spectral data (Figure 1), displacement of isopropanol by the accumulation of phospholipid is indicated by a decrease in intensity of the isopropanol C-C stretching mode35 (816 cm-1). In the component spectrum, the appearance of phospholipid is indicated by the appearance of the CN stretching mode from the phospholipid head group (715 cm-1), and an ACS Paragon Plus Environment


Page 10 of 19

10 increase in the broad C-C gauche-conformer stretching mode at 1086 cm-1 and the CH2twisting mode at 1303 cm-1; increased scattering in these modes is consistent with the growth of disordered chains from the adsorption of lipid.36-38 Additional

indicators

that

the

adsorbed

phospholipid acyl chains are disordered are the less well-resolved C-H bending and scissoring modes at 1436 cm-1 and 1455 cm-1, respectively.3841

In the C-H stretching region, the broad band at

2853 cm-1 also increases; this band is associated with the splitting of the CH2 symmetric-stretching mode and a strong Fermi resonance between the terminal

methyl-group

methylene-stretching

and

mode,

the

symmetric-

associated

with

gauche defects. The 2900-cm-1 band, which is due to the sum of the broad CH-symmetric stretching mode (at 2890 cm-1 when alkyl chains contain a large fraction of gauche defects41) and the Fermi resonance due to the asymmetric CH-bending mode and symmetric CH3-stretching mode (2910

Figure 2: A. Component Raman spectra corresponding to C18-silica prior to lipid adsorption (black), the initial adsorption of DMPC where alkyl chains contain a large population of gauche conformers (red) , the second adsorption step where the alkyl-chains adopt mostly trans conformation (blue) and the final component corresponding to the within–particle hybrid bilayer in water (green). B. Relative concentration vectors showing the proportion of each component in time as assembly of the DMPC hybrid layer occurs.

cm-1),42,43 indicate freedom in the alkyl chains that allows interaction between terminal methyl groups and nearby methylene stretching modes, further evidence of disordered alkyl chains. Interestingly, the spectrum corresponding to the surface-bound disordered C18-chains does not entirely disappear during the initial adsorption (Figure 2B), nor does the disordered component corresponding to adsorption of gauche conformers grow to a relative concentration of unity. Instead, a second step occurs where continued adsorption of phospholipid results in alkylchain ordering. Displacement of isopropanol continues as a decrease in its relative contribution until it disappears at ~750 s, at which time, the growth of the CN-stretching mode rolls over, indicating no further lipid accumulation, and where the spectra no longer change in time and ACS Paragon Plus Environment

Page 11 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


11 adsorption equilibrium is achieved. In this region, the disordered component has decreased by about half relative to its maximum at ~500 seconds, replaced by a rapidly rising ordered component spectrum whose alkyl-chain conformations show trans-conformer C-C stretching modes at 1061 cm-1 and 1126 cm-1, a shift to lower frequency (1298 cm-1) of the CH2-twisting mode, and overlap of the CH-bending modes.36,37,44 Additionally, in the CH-stretching region, the CH-symmetric stretching mode at 2847 cm-1 and the CH2-antisymmetric-stretching mode at 2883 cm-1, which are both sensitive to alkyl-chain packing due to inter-chain vibrational coupling, indicate a gel-phase bilayer with a majority of ordered trans-conformers.41,44 At 1260 s, after within-particle spectra of the hybrid bilayer in isopropanol/water/DMPC solution were stable, the exterior solution was switched to pure water. Further alkyl-chain ordering is observed accompanying displacement of isopropanol. This is not surprising, as disordering effects of alcohols on membranes are well-documented.45 It has been shown that association of ethanol with the glycerol moiety in the phospholipid headgroup and interaction with acyl-chain methylene groups near the headgroup likely leads to disordering;46 a similar interaction likely occurs with isopropanol at the hybrid-bilayer membrane. Thus, removal of isopropanol results in further and significant chain ordering of the hybrid bilayer. The rate of lipid accumulation to form a hybrid-bilayer within these porous particles is considerably

(several-orders-of-magnitude)

slower than diffusional-limited transport of phospholipid

from

the

relatively

high

concentration (0.7-mM) solution to equilibrate with the C18-modified particle surface. This suggests there is a significant kinetic barrier for transferring lipid from solution to the surface, which was also observed for pyrene partitioning

Figure 3. Particle-size dependence of DMPC hybridbilayer formation. A. CN-stretching intensity shows the total lipid coverage versus time (inset: accumulation time-constant versus particle radius). B. Time-evolution of hybrid-bilayer structure (spectral component variation versus time).



Page 12 of 19

12 kinetics into C18-chromatographic particles.32 To investigate the nature of the transfer kinetics, hybrid-bilayer formation was measured for different-sized particles, and the results are plotted in Figure 3. The linear dependence of lipid-accumulation time with particle radius (Figure 3A) indicates that the barrier for transferring lipid to the C18 surface that occurs near the outer surface of the particle. As size of the particle increases, the volume of the particle requiring lipid grows with the radius cubed while the surface area to capture lipid increases with the radius squared. Thus, the time-for-accumulation grows in proportion to the particle radius, as observed for hybrid-bilayer formation and for small-molecule partitioning into C18 particles.32 The non-zero xaxis intercept indicates that lipid diffuses a small distance (~1.3-µm) into the pores of the particle before being captured by the C18 surface. While particle-size effects account for variation in lipid-accumulation times, the multi-step kinetics of hybrid-bilayer structure evolution are consistently observed even as the rate of accumulation changes with particle size (Figure 3B). Raman spectroscopy of hybrid bilayer self-assembly with deuterated phospholipids. In previous work, hybrid bilayers formed in C18 reversed-phase chromatographic silica were demonstrated to have an interdigitated structure, and during bilayer melting, both the upper leaflet of phospholipid and lower leaflet of C18

chains

disorder

simultaneously.25

During hybrid bilayer self-assembly, we observe ordering of alkyl chains; however, it is unclear whether this ordering involves only the upper leaflet of phospholipid, the lower leaflet of C18, or both leaflets. To investigate the nature of hybrid bilayer formation in more detail, the timedependence of bilayer self-assembly were monitored with a phospholipid having

Figure 4: Raman spectra collected in situ within an individual C18-silica particle as deuterated DMPC is adsorbed from solution to the C18-surface. Red section is the equilibrium region where the overlapping portions of the SMCR analysis were carried out.

perdeuterated acyl-chains, DMPC-D54 (Figure 4). This experiment allows Raman scattering from the upper leaflet to be distinguished from Raman scattering of the lower leaflet C18 chains. The spectra in Figure 4 were subjected to self-modeling curve resolution analysis as outlined above, where the principal components analysis again identified four significant components. ACS Paragon Plus Environment

Page 13 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


13 The resulting component spectra from SMCR analysis are presented in Figure 5A, and the corresponding composition vectors are plotted in Figure 5B. Similar

adsorption

profiles

are

observed for the perdeuterated phospholipid as the hydrogenated lipid. The initial component, corresponding to the disordered C18 chains disappears

slowly

throughout

the

self-

assembly process. Deuterated phospholipid accumulation is indicated by the appearance of the phospholipid CN stretching mode at 715 cm-1 and the displacement of within-pore isopropanol observed in the disappearance of the isopropanol C-C stretch at 816 cm-1. Interestingly, during deuterated hybrid bilayer self-assembly, accumulation of lipid and acyland alkyl-chain ordering begin simultaneously. Two distinct components are clearly indicated by SMCR analysis: a disordered component with a faster adsorption rate, and an ordered component which appears more gradually. From examination of the component spectra, it is clear that the upper and lower leaflets of the hybrid

bilayer

evolve

in

structure

Figure 5: A. Component Raman spectra corresponding to the C18-silica prior to deuterated lipid adsorption (black), the initial adsorption of DMPC-D54 where the alkyl chains contain a large population of gauche conformers (red) , the second adsorption step where the alkyl chains adopt mostly trans conformation (blue) and the final component corresponding to the within–particle hybrid bilayer in water (green). B. Relative concentration vectors showing the proportion of each component in time as adsorption of DMPC-D54 occurs.

simultaneously during assembly. This is indicated in the disordered component (red spectrum in Figure 5A) by the presence of Raman bands which correspond to exclusively gauche conformers in both deuterated lipid and hydrogenated C18 regions of the spectrum including the distinctly low-intensity and broad CD2-twisting mode (916 cm-1), the low intensity CD2 scissoring mode (981 cm-1), and low intensity symmetric and antisymmetric CD stretching modes (2101 cm-1 and 2195 cm-1) as well as the low-intensity trans-conformer C-C stretching modes (1061 cm-1 and 1126 cm-1), and the corresponding broad and lower-frequency CH2-twisting mode (1303 cm-1) ACS Paragon Plus Environment


Page 14 of 19

14 and the CH stretching modes at 2853 cm-1 and 2900 cm-1 discussed in the previous section. In the ordered component (blue spectrum in Figure 5A), we observe the appearance of bands corresponding to chain ordering, again in both deuterated acyl- and hydrogenated alkylchain Raman modes. These include a more intense, less-broad CD2-twisting mode (916 cm-1), the appearance of a sharp CD-scissoring mode (981 cm-1), and sharper, intense symmetric and antisymmetric CD stretching modes (2101 cm-1 and 2195 cm-1) and in the hydrogenated alkylchains, increased intensity of the trans-conformer C-C stretching modes (1061 cm-1 and 1126 cm-1), increased intensity and sharper CH2-twisting mode (1295 cm-1) and splitting of the CH-scissoring mode (1455 cm-1).47 These results clearly indicate that hybrid bilayer selfassembly including lipid adsorption and alkyl-chain ordering processes occur simultaneously, where the structure of both upper lipid acyl chains and lower leaflet C18 chains evolve together, consistent with forming an interdigitated bilayer. Two interesting differences are apparent between deuterated and hydrogenated hybrid bilayer formation. The first is that lipid adsorption is slower and chain ordering grows slowly and continuously during deuterated bilayer self-assembly compared to a threshold transition to an ordered structure in the case of the hydrogenated lipid. A second difference is that in the case of deuterated phospholipid, the disordered lipid component accumulates and then remains at constant intensity during the self-assembly process, whereas this component first increases and then decreases in intensity during hydrogenated bilayer self-assembly. These results are likely due to differences in the tendency toward ordering of the final hybrid bilayer for these two lipids. The deuterated hybrid bilayer exhibits a lower melting transition centered at ~33°C, while the hydrogenated hybrid bilayer melts at a higher temperature, ~37°C.25 Thus, a significant fraction of gauche defects persists during assembly of the less ordered deuterated hybrid bilayer from alcohol solution, as observed in the results. Acyl-chain ordering of the deuterated lipid, especially in the presence of alcohol,45 must represent a significant barrier to lipid accumulation and organization resulting in slower accumulation and lack of a distinct transition to a more ordered phase in the presence of alcohol. Upon changing the solvent from 30% isopropanol to pure water, however, both deuterated and hydrogenated hybrid bilayers become ordered and dominated by trans-conformer alkyl chains25 as the hydrocarbon region organizes to exclude contact with water. ACS Paragon Plus Environment

Page 15 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


15 CONCLUSIONS In this work, confocal Raman microscopy was employed to investigate the timedependent self-assembly of hybrid-supported phospholipid bilayers within individual chromatographic particles. Self-modeling curve resolution (SMCR) of the Raman scattering data shows that self-assembly is a multi-component process where phospholipid is initially adsorbed to the interface with relatively disordered alkyl-chains, followed by organization of the bilayer into an ordered structure where trans-conformers of the chains are dominant. By using deuterated acyl-chains on the phospholipid and monitoring self-assembly with the same experiment, it was possible to resolve Raman scattering of the upper leaflet of the phospholipid from the lower leaflet of the C18 chains to investigate how the structure of each leaflet evolves during the assembly process. Using SMCR to examine these data, it was determined that both the upperand lower-leaflet chains evolve simultaneously during the assembly, where residual gaucheconformers became all-trans and interdigitated when the surrounding solution was changed to pure water. This work sheds light on the evolution of the structure of hybrid phospholipid bilayer structure within C18-functionalized chromatographic silica particles. It also demonstrates the utility of confocal Raman microscopy for investigating self-assembly processes at buried interfaces within porous materials; this methodology could be easily extended to study such processes within metal-organic frameworks, in porous polymers, nanoporous alumina films, or monolithic columns to guide the dynamic modification of separation supports.

ACKNOWLEDGMENTS This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, Division of Energy, Chemical Sciences, Geosciences, & Biosciences, Award Number DE-FG03-93ER14333, and the Army Research Office MURI, W911NF-14-1-0263.



Page 16 of 19

16 REFERENCES (1)

Plant, A.L. Supported Hybrid Bilayer Membranes as Rugged Cell Membrane Mimics.

Langmuir 1999, 15, 5128-5135. (2)

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. (3)

Plant, A.L. Self-assembled phospholipid/alkanethiol biomimetic bilayers on gold.

Langmuir 1993, 9, 2764-2767. (4)

Meuse, C.W.; Niaura, G.; Lewis, M.L.; Plant, A.L. Assessing the Molecular Structure of

Alkanethiol Monolayers in Hybrid Bilayer Membranes with Vibrational Spectroscopies. Langmuir 1998, 14, 1604-1611. (5)

Plant, A.L.; Brighamburke, M.; Petrella, E.C.; Oshannessy, D.J.

Phospholipid/Alkanethiol Bilayers for Cell-Surface Receptor Studies by Surface Plasmon Resonance. Anal. Biochem. 1995, 226, 342-348. (6)

Hubbard, J.B.; Silin, V.; Plant, A.L. Self assembly driven by hydrophobic interactions at

alkanethiol monolayers: mechanism of formation of hybrid bilayer membranes. Biophys. Chem. 1998, 75, 163-176. (7)

Levin, C.S.; Kundu, J.; Janesko, B.G.; Scuseria, G.E.; Raphael, R.M.; Halas, N.J.

Interactions of Ibuprofen with Hybrid Lipid Bilayers Probed by Complementary SurfaceEnhanced Vibrational Spectroscopies. J. Phys. Chem. B 2008, 112, 14168-14175. (8)

Millo, D.; Bonifacio, A.; Moncelli, M.R.; Sergo, V.; Gooijer, C.; van der Zwan, G.

Characterization of hybrid bilayer membranes on silver electrodes as biocompatible SERS substrates to study membrane–protein interactions. Colloids Surf., B 2010, 81, 212-216. (9)

Meuse, C.W.; Krueger, S.; Majkrzak, C.F.; Dura, J.A.; Fu, J.; Connor, J.T.; Plant, A.L.

Hybrid Bilayer Membranes in Air and Water: Infrared Spectroscopy and Neutron Reflectivity Studies. Biophys. J. 1998, 74, 1388-1398. (10)

Tsirkin, I.; Grushka, E. Characterization of dynamically prepared phospholipid-modified

reversed-phase columns. J. Chromatogr. A 2001, 919, 245-254. (11)

Luo, H.-B.; Zheng, C.; Cheng, Y.-K. Is phospholipid-saturated alkyl column a convenient

replacement for immobilized-artificial-membrane? J. Chromatogr. A 2007, 1176, 100-106.


Page 17 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


17 (12)

Godard, T.; Grushka, E. The use of phospholipid modified column for the determination

of lipophilic properties in high performance liquid chromatography. J. Chromatogr. A 2011, 1218, 1211-1218. (13)

Walter, T.H.; Iraneta, P.; Capparella, M. Mechanism of retention loss when C8 and C18

HPLC columns are used with highly aqueous mobile phases. J. Chromatogr. A 2005, 1075, 177183. (14)

Gasser-Ramirez, J.L.; Harris, J.M. Confocal Raman Microscopy Investigation of the

Wetting of Reversed-Phase Liquid Chromatographic Stationary Phase Particles. Anal. Chem. 2009, 81, 7632-7638. (15)

Ollila, F.; Halling, K.; Vuorela, P.; Vuorela, H.; Slotte, J.P. Characterization of

flavonoid–biomembrane interactions. Arch. Biochem. Biophys. 2002, 399, 103-108. (16)

Zhang, Y.; Zeng, C.-M.; Li, Y.-M.; Hjertén, S.; Lundahl, P. Immobilized liposome

chromatography of drugs on capillary continuous beds for model analysis of drug-membrane interactions. J. Chromatogr. A 1996, 749, 13-18. (17)

Ho, M.; Cai, M.; Pemberton, J.E. Characterization of Octadecylsilane Stationary Phases

on Commercially Available Silica-Based Packing Materials by Raman Spectroscopy. Anal. Chem. 1997, 69, 2613-2616. (18)

Ho, M.; Pemberton, J.E. Alkyl Chain Conformation of Octadecylsilane Stationary Phases

by Raman Spectroscopy. 1. Temperature Dependence. Anal. Chem. 1998, 70, 4915-4920. (19)

Doyle, C.A.; Vickers, T.J.; Mann, C.K.; Dorsey, J.G. Characterization of C18-bonded

liquid chromatographic stationary phases by Raman spectroscopy:: the effect of mobile phase composition. J. Chromatogr. A 2000, 877, 25-39. (20)

Doyle, C.A.; Vickers, T.J.; Mann, C.K.; Dorsey, J.G. Characterization of C18-bonded

liquid chromatographic stationary phases by Raman spectroscopy:: the effect of temperature. J. Chromatogr. A 2000, 877, 41-59. (21)

Pemberton, J.E.; Ho, M.; Orendorff, C.J.; Ducey, M.W. Raman spectroscopy of

octadecylsilane stationary phase conformational order: Effect of solvent. J. Chromatogr. A 2001, 913, 243-252. (22)

Gasser-Ramirez, J.L.; Harris, J.M. Confocal Raman Microscopy of the Interfacial

Regions of Liquid Chromatographic Stationary-Phase Materials. Anal. Chem. 2009, 81, 28692876.



Page 18 of 19

18 (23)

Gasser-Ramirez, J.L.; Harris, J.M. Quantitative Confocal Raman Microscopy Study of

Ion-Interaction Retention at Reversed-Phase Chromatographic Interfaces. Anal. Chem. 2010, 82, 5743-5750. (24)

Kitt, J.P.; Harris, J.M. Confocal Raman Microscopy for In-Situ Detection of Solid-Phase

Extraction of Pyrene into Single C18-Silica Particles. Anal. Chem. 2014. (25)

Kitt, J.P.; Harris, J.M. Confocal Raman Microscopy of Hybrid-Supported Phospholipid

Bilayers within Individual C18-Functionalized Chromatographic Particles. Langmuir 2016, 32, 9033-9044. (26)

Kitt, J.P.; Bryce, D.A.; Minteer, S.D.; Harris, J.M. Confocal Raman Microscopy for In-

situ Measurement of Phospholipid-Water Partitioning into Model Phospholipid Bilayers within Individual Chromatographic Particles. Anal. Chem. 2018, 90, 7048-7055. (27)

Lawton, W.H.; Sylvestre, E.A. Self Modeling Curve Resolution. Technometrics 1971, 13,

617-633. (28)

Malinowski, E.R. Factor Analysis in Chemistry (2nd edition); John Wiley and Sons: New

York, 1991. (29)

Fox, C.B.; Uibel, R.H.; Harris, J.M. Detecting Phase Transitions in Phosphatidylcholine

Vesicles by Raman Microscopy and Self-Modeling Curve Resolution. J. Phys. Chem. B 2007, 111, 11428-11436. (30)

Bridges, T.E.; Houlne, M.P.; Harris, J.M. Spatially Resolved Analysis of Small Particles

by Confocal Raman Microscopy: Depth Profiling and Optical Trapping. Anal. Chem. 2003, 76, 576-584. (31)

Kitt, J.P.; Harris, J.M. Confocal Raman Microscopy for in Situ Measurement of Octanol–

Water Partitioning within the Pores of Individual C18-Functionalized Chromatographic Particles. Anal. Chem. 2015, 87, 5340-5347. (32)

Bryce, D.A.; Kitt, J.P.; Harris, J.M. Confocal Raman Microscopy Investigation of

Molecular Transport into Individual Chromatographic Silica Particles. Anal. Chem. 2017, 89, 2755-2763. (33)

Kitt, J.P.; Bryce, D.A.; Harris, J.M. Calorimetry-Derived Composition Vectors to

Resolve Component Raman Spectra in Phospholipid Phase Transitions. Appl. Spectrosc. 2016, 70, 1165-1175. (34)

Malinowski, E.R.; Cox, R.A.; Haldna, U.L. Factor analysis for isolation of the Raman

spectra of aqueous sulfuric acid components. Anal. Chem. 1984, 56, 778-781. ACS Paragon Plus Environment

Page 19 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


19 (35)

Green, J. Thermodynamic properties of organic oxygen compounds. Part 12.—

Vibrational assignment and calculated thermodynamic properties 0–1000° K of isopropyl alcohol. Trans. Faraday Soc. 1963, 59, 1559-1563. (36)

Lippert, J.; Peticolas, W. Raman active vibrations in long-chain fatty acids and

phospholipid sonicates. Biochim. Biophys. Acta, Biomembr. 1972, 282, 8-17. (37)

Grell, E. Membrane Spectroscopy; Springer-Verlag, 1981.

(38)

Orendorff, C.J.; Ducey, M.W.; Pemberton, J.E. Quantitative correlation of Raman

spectral indicators in determining conformational order in alkyl chains. J. Phys. Chem. A 2002, 106, 6991-6998. (39)

Yellin, N.; Levin, I.W. Hydrocarbon chain trans-gauche isomerization in phospholipid

bilayer gel assemblies. Biochemistry 1977, 16, 642-647. (40)

Wallach, D.F.; Verma, S.P.; Fookson, J. Application of laser Raman and infrared

spectroscopy to the analysis of membrane structure. Biochim. Biophys. Acta, Rev. Biomembr. 1979, 559, 153-208. (41)

Wong, P.T.T. Raman Spectroscopy of Thermotropic and High-Pressure Phases of

Aqueous Phospholipid Dispersions. Annu. Rev. Biophys. Bioeng. 1984, 13, 1-24. (42)

Hill, I.R.; Levin, I.W. Vibrational spectra and carbon–hydrogen stretching mode

assignments for a series of n‐alkyl carboxylic acids. J. Chem. Phys. 1979, 70, 842-851. (43)

Cirak, J.; Horvath, L. Effect of lateral chain-chain spacing on the Raman spectrum of

dipalmitoyl phosphatidylcholine bilayers. Chem. Phys. Lipids 1988, 49, 197-204. (44)

Orendorff, C.J.; Ducey, M.W.; Pemberton, J.E. Quantitative Correlation of Raman

Spectral Indicators in Determining Conformational Order in Alkyl Chains. J. Phys. Chem. A 2002, 106, 6991-6998. (45)

Ingólfsson, H.I.; Andersen, O.S. Alcohol's effects on lipid bilayer properties. Biophys. J.

2011, 101, 847-855. (46)

Holte, L.L.; Gawrisch, K. Determining Ethanol Distribution in Phospholipid Multilayers

with MAS−NOESY Spectra. Biochemistry 1997, 36, 4669-4674. (47)

Bryce, D.A.; Kitt, J.P.; Harris, J.M. Confocal-Raman Microscopy Characterization of

Supported Phospholipid Bilayers Deposited on the Interior Surfaces of Chromatographic Silica. J. Am. Chem. Soc. 2018, 140, 4071-4078.


Confocal Raman Microscopy Investigation of Self ... - ACS Publications

Recommend Documents