Subscriber access provided by CORNELL UNIVERSITY LIBRARY
Article
A microfluidic diffusion platform for characterizing the size of lipid vesicles and the thermodynamics of protein-lipid interactions Hongze Gang, Celine Galvagnion, Georg Meisl, Thomas Müller, Alexander K. Buell, Aviad Levin, Christopher M. Dobson, Bo-Zhong Mu, and Tuomas P.J. Knowles Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b04820 • Publication Date (Web): 09 Jan 2018 Downloaded from http://pubs.acs.org on January 10, 2018
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 free 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 accessible to all readers and 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.
Analytical Chemistry 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 14 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
A microfluidic diffusion platform for characterizing the size of lipid vesicles and the thermodynamics of protein-lipid interactions Hongze Gang‡,†,§, Céline Galvagnion‡,†,◊, Georg Meisl†, Thomas Müller†,¶, Alexander K. Buell†,|, Aviad Levin†, Christopher M. Dobson†,*, Bozhong Mu§,* and Tuomas P.J. Knowles†,#,* †
Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge, CB2 1EW, UK §
State Key Laboratory of Bioreactor Engineering and School of Chemistry and Molecular Engineering, East China University of Science and Technology, 130 Meilong Road, Shanghai, 200237 P.R. China
¶
Fluidic Analytics Ltd, Unit 5 Chesterton Mill, French’s Road, Cambridge, CB4 3NP, UK #
Cavendish Laboratory, Department of Physics, University of Cambridge, J J Thomson Avenue, Cambridge, CB3 1HE, UK
*
Prof. Christopher M. Dobson, E-mail:
[email protected], Phone Number: +44 (0)1223 763070; Prof. Bozhong Mu, E-mail:
[email protected], Phone Number: +86 (0)21 64252063; Prof. Tuomas P.J. Knowles, E-mail:
[email protected], Phone Number: +44 (0)1223 336344
ABSTRACT: Elucidation of the fundamental interactions of proteins with biological membranes under native conditions is crucial for understanding the molecular basis of their biological function and malfunction. Notably, the large surface to volume ratio of living cells provides a molecular landscape for significant interactions of cellular components with membranes modulating their function. However, such interactions can be challenging to probe using conventional biophysical methods due to the heterogeneity of the species and processes involved. Here, we use direct measurements of micron scale molecular diffusivity to detect and quantify the interactions of α-synuclein, associated with the aetiology of Parkinson’s disease, with negatively charged lipid vesicles. We further demonstrate that this microfluidic approach enables the characterisation of vesicles of different binary mixtures size distribution, which is not readily accessible using conventional light scattering techniques. Finally, the size distributions of the two α-synuclein conformations - free α-synuclein and membrane-bound α-synuclein - were resolved under varying lipid:protein ratios, thus allowing the determination of the dissociation constant and the binding stoichiometry associated with this protein-lipid system. The microfluidic diffusional sizing platform allows these measurements to be performed on a time scale of minutes using microlitre volumes, thus establishing the basis for an approach for the study of molecular interactions of heterogenous systems under native conditions.
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
Introduction In nature, the self-assembly of proteins into supra-molecular complexes underlies both biological function and malfunction (1-4). Indeed, the majority of active cellular processes rely on protein-based molecular machinery for their function, while aberrant protein self-assembly processes are implicated in a wide array of pathological conditions, including neurodegenerative disorders (2, 3). In the latter case, the formation of pathogenic amyloid assemblies from peptides and proteins, which are normally found in their soluble state in healthy tissue, has emerged as a crucial upstream event, found to trigger a cascade of pathological processes leading to the onset and progression of disease (3). In addition to protein-protein interactions in solution, other factors may modulate molecular interactions of proteins and their resulting self-assembly propensity, such as the affinity of proteins for biological membranes (5, 6). A striking example of this phenomenon is the role of lipid membranes in triggering the aberrant assembly of α-synuclein, a protein associated with Parkinson’s disease (7), into amyloid fibrils (6, 8-10). α-synuclein is mainly localised in nerve terminals (11) and the monomeric protein has been found to populate two conformations in vivo and in vitro: a cytosolic intrinsically disordered conformation in equilibrium with an α-helical conformation bound to lipid vesicles (12-17). The interaction between the protein and lipid vesicles is believed to play an important role in its suggested biological function, in particular in the context of synaptic plasticity (18), but can also modulate the kinetics of its aberrant aggregation (19). Characterizing the interactions between the protein and membranes is thus of great importance for the determination and understanding of protein self-assembly mechanisms, both in vitro and in vivo. As the small, typically picolitre volumes of cells imply that the surface to volume ratios in living systems can be significantly larger than in bulk in vitro assays, membrane interactions have the potential to play a major role in favoring, modulating or suppressing protein self-assembly processes. However, using conventional biophysical techniques, the study of mixed protein-lipid assemblies has proven to be challenging. This is mainly due to the highly heterogeneous nature of such assemblies which lie in a size range where they can be larger than the complexes commonly studied by solution structural biology techniques such as nuclear magnetic resonance spectroscopy, yet smaller than assemblies which are commonly studied using optical microscopy techniques. Microfluidic techniques have emerged as valuable platforms in the study of biomolecular phenomena, mainly due to their versatility, inherently low sample consumption, as well as high sensitivity and high-throughput analysis (20, 21). Such devices allow analysis of processes on the micron scale (22-25), and can be exploited for the study of molecular diffusivity changes as molecules undergo binding events that change their effective molecular weight and thus their diffusion coefficients. This approach has been recently applied for the study of the interactions between small molecules and proteins (26-29) as well as the determination of the sizes of proteins and protein complexes (30, 31). The performance of such diffusionbased platforms is enabled by the ability to temporally and spatially monitor micronscale mass transport processes and globally analyse the resulting data (30).
ACS Paragon Plus Environment
Page 2 of 14
Page 3 of 14 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
Here, we explore the application of microfluidic diffusional sizing for the study of αsynuclein binding to negatively charged small unilamellar vesicles (SUVs). We first applied the diffusion-based microfluidic platform to measure the size of small and large unilamellar vesicles (SUVs and LUVs, respectively) and to determine the size distribution of heterogeneous mixtures of these vesicles. We found that our platform enables the accurate sizing of lipid vesicles in heterogeneous solutions with components, which differ in size down to only a factor of 2. We then studied the binding of α-synuclein to such lipid vesicles by monitoring the change in the effective hydrodynamic radius (Rh) of the protein with increasing lipid:protein ratios. From these measurements, we were able to determine both the dissociation constant, KD, and stoichiometry, L, defined as the number of lipid molecules bound to one molecule of α-synuclein, characteristic of this system. We further show that the hydrodynamic radii and the dissociation constant obtained from these diffusionbased microfluidic measurements are in agreement with those obtained using traditional bulk techniques such as dynamic light scattering (DLS) and circular dichroism (CD), respectively, but with enhanced resolution for probing heterogeneous systems and reduced sample consumption. Experimental Section Fabrication of microfluidic devices Microfluidic devices were fabricated using polydimethylsiloxane (PDMS) (Sylgard 184 kit, Dow Corning) mixed with carbon powder (Sigma) onto a master which was produced by spinning SU8-3025 (MicroChem Corp. Westborough, MA, USA) onto a silicon wafer and patterning by means of UV lithography (32, 33). The PDMS channels were plasma bonded to a glass slide to obtain microfluidic devices. As shown in Fig 1a, the device has one analyte inlet, one auxiliary fluid inlet, and one outlet. The analyte stream meets the auxiliary stream (buffer of analyte sample was used as an auxiliary fluid) at a nozzle, and the streams flow side by side to the outlet along the channel of a width of 300 µm and a height of 25 µm. Preparation of DOPS vesicles Chloroform solutions of 1,2-dioleoyl-sn-glycero-3-phospho-L-serine (sodium salt) (DOPS) (Avanti Polar) and ATTO-565-1,2-dimyristoyl-sn-glycero-3phosphoethanolamine (ATTO- 565-DMPE) (ATTO-TEC GmbH) were mixed to a final DOPS:ATTO-565-DMPE molar ratio of 200:1. A dry and thin lipid film was then prepared by gently evaporating the chloroform using a nitrogen stream. The film was then placed under vacuum over night to remove all traces of chloroform. Phosphate buffer (20 mM Na2HPO4/NaH2PO4, 0.01% NaN3, pH 6.5) was added to hydrate the lipid film and the solution was stirred at room temperature (above the phase transition temperature of DOPS, -11°C) for 2 hours. The lipid solution was then frozen and thawed 5 times using dry ice and water bath, respectively. The lipid solution was extruded through polycarbonate membranes with the desired pore size (Avanti Polar), and the size distribution of lipid vesicles was then determined using dynamic light scattering (Malvern Instruments, Malvern UK). Three polydisperse vesicle mixtures were prepared by mixing vesicles of different radii (about 20 nm and 40 nm, respectively) at varying molar ratios. Non-labeled vesicles were used for the binding experiments. All stock solutions of the lipid vesicle were stored at 4°C.
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 4 of 14
Protein solution The wild-type and N122C variant of α-synuclein were purified as previously described (6, 34). The α-synuclein N122C variant was labeled with maleimidemodified Alexa Fluor® 647 dye (Invitrogen, Carlsbad, CA, USA) via the cysteine thiol moiety, as previously described (35). For the diffusion measurements, we used a 20:1 molar ratio of unlabelled to labelled protein. Circular dichroism measurements α-synuclein was incubated in the absence and presence of DOPS at 20°C. For each α-synuclein:DOPS ratio, far UV spectra were obtained by averaging five individual spectra recorded between 250 and 200 nm with a bandwidth of 1 nm, a data pitch of 0.2 nm, a scanning speed of 50 nm.min−1, and a response time of 1 s. Each value of the CD signal intensity reported at 222 nm corresponds to the average of five measurements, each acquired for 10 s. For each protein sample, the CD signal of the buffer used to solubilize the protein was recorded and subtracted from the CD signal of the protein. The fraction of protein bound to the vesicles in the presence of increasing DOPS concentration was calculated using the CD signal measured at 222 nm as described previously (6) (see Eq. 1). In brief, the measured circular dichroism signal (CD ) can be expressed as: CD = x CD + x CD
(1)
where CD and CD are the CD signal of α-synuclein free in solution and bound to the lipid vesicles and x and x are the fractions of α-synuclein free in solution and bound to the vesicles, respectively. Determination of binding constant We considered a non-cooperative binding Langmuir-Hill adsorption model to fit the change in the fraction of protein bound to the vesicles with increasing concentration of lipid, as described previously (6), using the following equation: x =
!"#$
%
(2)
where α − syn and DOPS are the total concentrations of α-synuclein and DOPS in solution, respectively, ./ (in M) is the dissociation constant and 0 is the number of lipid molecules interacting with one molecule of α-synuclein. The experimental CD and diffusional sizing data were fitted to Eq. 2 using AmyloFit (36). Dynamic light scattering measurements The size distributions of α-synuclein, lipid vesicles, mixtures of vesicles, mixture of αsynuclein and vesicles were assessed using DLS (Zetasizer Nano ZSP, Malvern Instruments, Malvern, UK) at 20°C. The size distribution by particle number was used. Microfluidic diffusion measurements Prior to the measurement, the device was fully prefilled with buffer to remove all air inside the channels. Thereafter, a glass syringe (Hamilton) was used to withdraw
ACS Paragon Plus Environment
Page 5 of 14 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
liquid from the outlet to generate steady-state laminar flow in the channel and to inject analyte and auxiliary fluid which are loaded into reservoirs at the device inlets (see Fig 1a). A stream of analyte co-flows with the auxiliary fluid at the nozzle, and diffuses to either side of the interface when travelling downstream. Once the spatiotemporal steady state of the diffusive gradient was reached, epifluorescence images at the 12 measurement positions along the channel (as noted in Fig 1b) were taken using an inverted microscope (Zeiss, A1 Observer) coupled to a CCD camera (Photometrics Evolve 512). Flow rates were controlled by syringe pumps (neMESYS, Cetoni GmbH). All the diffusion experiments were conducted at 20 ± 1°C, as monitored by a USB temperature data logger (TFA Dostmann). The time required to reach steady state depended on the flow rates, which were related to the expected size of the analyte. Fitting of the experimental concentration profiles In brief, predicted mass-distributions for a given particle size 12 , 3 45 , were generated by solving numerically for the experimental geometry the diffusion advection equation describing mass transport as reported previously (37). Concentration profiles, 6789, giving the concentration of analytes integrated over the height of the microfluidic device, were obtained from the 12 epifluorescence images (as shown in Fig 1b) which correspond to 12 diffusion times. These diffusion profiles were fitted to a linear combination of the simulated basis functions by performing the minimization: min
