Long-Lived Planar Asymmetric Suspended Membranes for Investigating Protein Dynamics and Membrane Fusion

doi:10.21203/rs.3.pex-2536/v1

Method Article

Long-Lived Planar Asymmetric Suspended Membranes for Investigating Protein Dynamics and Membrane Fusion

https://doi.org/10.21203/rs.3.pex-2536/v1

This work is licensed under a CC BY 4.0 License

This protocol has been posted on Protocol Exchange, an open repository of community-contributed protocols sponsored by Nature Portfolio. These protocols are posted directly on the Protocol Exchange by authors and are made freely available to the scientific community for use and comment.

Version 1

posted

You are reading this latest protocol version

Membrane fusion is pivotal in cellular signaling and trafficking, requiring an intricate understanding of the interactions between proteins and lipids. Addressing the limitations of studying complex protein dynamics in cellular environments, a reductionist in vitro membrane model system is essential. Such a method begins with minimal components and incrementally adds complexity, allowing for the isolation and detailed examination of individual molecular species, an effort unattainable with traditional cellular techniques. Suspended lipid membranes accurately mimic cellular architecture by providing a unique environment with aqueous compartments on both sides, exceptional lateral diffusion properties, a solvent-free environment, and excellent stability for time-lapse imaging. Here, we present a detailed protocol for generating pore-spanning based planar suspended lipid membranes from native and reconstituted membranes that are ideal for time-lapse imaging using total internal reflection and confocal microscopy techniques. Our approach enables precise single-molecule protein counting, comprehensive analysis of protein dynamics, and the conduct of single-vesicle fusion assays. This method stands out for its exceptional temporal resolution, essential for accurately monitoring vesicle docking and fusion kinetics. The natural biological composition and lipid asymmetry of these membranes, combined with their excellent diffusion capabilities, greatly enhance the ability of studying the dynamics of membrane proteins. This protocol marks a significant advancement in membrane fusion research and offers deep insights into the nuances of cellular signaling and mechanisms.

Planar Lipid bilayers

Exocytosis

Membrane Asymmetry

Membrane Protein Dynamics

Eukaryotic organisms utilize SNARE protein-mediated fusion to release chemicals, peptides, and hormones to regulate growth, communication, and metabolism without leaking the content or disturbing lipid asymmetry.^1-5Depending on the physiology, precise constraints over vesicle size and rate of exocytosis control the flow of substance and information to maintain metabolic homeostasis. For example, at neuronal synapses, the first synaptic vesicles (SVs) can release neurotransmitters in 0.2-1 msec after action potential stimulation to keep up the pace of neural circuitry.⁶ Synapses are crowded environments with more than 250 identified proteins, which only a small number have been studied.^7,8 Dissecting the fundamental role of proteins and how they are organized in the active zone to facilitate membrane remodeling and ultra-fast neurotransmitter release is complex by in vivo methods. The knowledge we could gain from in vivo is sometimes clouded by the complexity of the cell and often they require fixing or expanding the samples.^9,10 In contrast, cell-free systems (e.g., liposomes, giant unilamellar vesicles, supported lipid bilayers, and nanodiscs) provide a “clean and minimal” system to study the role of individual lipids and proteins and their functional role in exocytosis.

While efficient, the reductionist approach faces challenges, especially in oversimplified models that fail to mimic physiological conditions accurately. An example is the early minimal reconstructions of Ca²⁺-triggered synaptic vesicle fusion, which utilized liposome-liposome fusion. However, this method falls short in replicating the intricate planar-curved architecture characteristic of cellular systems. To address this, advancements were made with supported-lipid membranes, either incorporating a lipid monolayer at the base or using a polyethylene-glycol-based layer to lift the membrane from the glass support. These modifications significantly improved the reconstitution of membrane proteins and reduced non-specific surface binding, thus providing a more accurate representation of cellular environments. Nevertheless, these models have limitations, especially in accommodating long, multi-pass transmembrane proteins with extensive amino acids on either side, due to the insufficient distance between the lipid bilayer and the glass support. Therefore, developing membrane models that closely replicate the native environment without organic solvents and allow for the exploration of biological processes on both sides of the membrane is crucial for dissecting biological mechanisms. The ultimate goal is to create an in vitro platform that offers precise control over each component with nanometric accuracy, high spatial and temporal resolution, lipid asymmetry, and the capacity to handle large, multipass membrane proteins and controlled protein copy numbers. Such a platform would provide unparalleled insights into the roles of proteins and lipids in synaptic vesicle (SV) exocytosis.

Suspended Lipid Membranes (SLIM) overcome the limitations of other model systems, such as liposomes, droplet lipid bilayers (DIBs), black lipid membranes (BLMs), and supported lipid bilayers (SLBs), by providing a controlled environment home for recalcitrant membrane proteins. It is well established that SLIM platforms can accurately reproduce exocytosis physiology under in vitro conditions.^11-16 Building on this, we have developed an advanced version of the SLIM platform, tailored to enable single-molecule imaging and the formation of planar membranes using cell-derived giant plasma membranes. This innovation paves the way for studying membrane fusion under conditions resembling the natural cellular environment. It also facilitates the examination of complex membrane proteins, such as G-protein-coupled receptors (GPCRs), ion channels, and receptors, without needing to fix and permeabilize cell membranes (Figure.1).¹⁷ This breakthrough not only enhances our understanding of membrane proteins but also opens new avenues for developing drugs targeting these crucial proteins.

In this protocol, we provide a straightforward and efficient method for constructing planar suspended lipid membranes by inducing the rupture of synthetic or native giant unilamellar vesicles. The procedure is designed to ensure the integrity and functionality of the membranes, which are crucial for the subsequent examination of vesicle docking and fusion events. Additionally, the protocol includes guidelines for employing fusion analyzer software to process and analyze the resulting data. This comprehensive guide provides detailed instructions for generating synthetic and native GUVs with intact membrane proteins, facilitating their study under a microscope in a planar configuration. It also discusses the benefits and drawbacks of various methodologies. Moreover, the step-by-step approach enables researchers to adeptly prepare lipid membranes with specific membrane proteins of interest, allowing for the investigation of their transient interactions and dynamics in single-molecule fluorescence settings.

Applications and target audience

The protocol described here focuses on recreating in situ a neuronal synapse, containing a readily releasable pool of vesicles on a chip, released by calcium. This innovative approach grants researchers the ability to analyze the function of each synaptic protein in the context of neurotransmitter release and assess the impact of mutations linked to neurodegenerative diseases. Beyond its applications in neuroscience, the protocol is designed to facilitate reconstitution any membrane protein into a free-standing planar lipid bilayer, creating opportunities to explore a broad spectrum of biological processes. These include protein-protein and protein-ligand interactions, membrane remodeling, and the elucidation of signaling pathways.

The versatility of this system is expected to draw attention from researchers across multiple disciplines, notably those in membrane biophysics, immunology, and cell signaling. The protocol is positioned to have an immediate and extensive impact on life sciences research, particularly given the increasing demand for accurate membrane models that enable drug screening. By allowing for the study of membrane proteins within their native environment, this method significantly enhances our understanding of protein dynamics and interactions, thereby catalyzing advancements in fundamental research and therapeutic development.

Comparison with alternative methods and advantages

When comparing various membrane models employed for the study of exocytosis, the planar membrane geometry stands out due to its close resemblance to the cellular plasma membrane, offering distinct advantages over other methods as outlined. The most prevalent among these is the supported lipid bilayer (SLB) system, which is adept at mimicking the planar-curved architecture of the cell surface and is particularly suitable for single-molecule imaging techniques, providing a clear window into the dynamics of exocytosis. The fluidity of this system allows for the natural mobility of proteins, which is essential for their function during vesicle fusion and release processes. However, one limitation of the SLB is the inherent tendency of the membranes to adhere to the substrate, which can artificially constrain the movement of proteins and does not perfectly replicate the three-dimensional environment they would encounter in vivo. This adherence can affect the physiological relevance of the observations, as it may not accurately reflect the proteins' behavior and interactions within the complex milieu of a living cell.

The planar geometry of suspended lipid membranes (SLIM) overcomes these drawbacks by offering a more biologically faithful platform that minimizes artifacts from surface attachment, thereby furnishing a more accurate representation of membrane fusion and protein dynamics as they occur during exocytosis. This system allows for the recreation of any cellular membrane-protein environment, either by incorporating purified proteins or by using cell-derived membranes from GPMVs. It uniquely provides access to both leaflets of the lipid bilayer and their associated proteins, enhancing the physiological relevance of the model.

The SLIM platform’s free-standing membranes ensure high protein mobility, leading to a greater number of vesicle proceed to fusions compared to other methods, thus enabling precise quantitative analyses and experimental reproducibility. These membranes replicate the in vivo geometry, which previous methodologies, including those employing SLBs, could not achieve due to their lack of access to the inner leaflet and the potential for creating artifacts by impeding protein diffusion. The effectiveness of the SLIM approach is further enhanced using laser scanning confocal microscopy with a resonance scanner that sharply focuses on the lipid bilayer while reducing background interference or TIRF microscopy that focuses below ~200 nm from the bilayer, ensuring clear and accurate observations of the vesicle fusion processes.

Limitations

The Suspended Lipid Membrane (SLIM) platform, developed for in-depth studies of cellular processes at the plasma membrane, exhibits notable strengths in membrane visualization and simulating physiological conditions. However, it also presents certain limitations in its current form, especially regarding the accessibility of membrane components and the flexibility for experimental manipulation, especially with native plasma membranes.

1. Orientation and Accessibility of Membrane Leaflets: In the process of bilayer formation from GPMVs, there's an inherent orientation where the inner leaflet faces upwards and the outer leaflet faces downward into the underlying holes. This specific orientation can restrict experimental designs, particularly those necessitating direct interaction with the outer leaflet. Currently, introducing ligands to the outer leaflet post-bilayer formation is challenging. Future versions of the platform are planned to incorporate flow channels beneath the silicon substrate, enabling access to both membrane leaflets.

2.Variability in GPMV Composition: The bulk collection of GPMVs for bilayer formation, while standardizing the process and enhancing robustness, may result in the dilution of certain proteins or lipid domains. This could lead to some bilayers lacking proteins or lipids of interest. A provisional remedy is the utilization of protein tags and affinity sorting to select GPMVsof interest. Future methodologies aim to selectively generate GPMVs from targeted regions of the plasma membrane.

3.Visualization Challenges Due to Silicon Substrate: The employment of silicon in the SLIM setup limits observation to regions where the membrane is suspended. This limitation obscures the bilayer sections that overlay the silicon, creating challenges when the area of interest includes proteins or domains located on the supported part of the bilayer. Modifying the silicon structure, such as altering the dimensions and spacing of the holes, can offer a partial solution. A more comprehensive approach would entail design changes that permit dynamic adjustments of the membrane’s suspension and support within the system.

Despite these limitations, the SLIM platform represents a significant advancement in the study of membrane-associated phenomena. These challenges underscore the importance of ongoing development and refinement. Future enhancements will aim to increase the accessibility of membrane components, guarantee the presence of specific proteins or lipids within the bilayers, and improve the overall visibility across the entire membrane structure. This continual evolution will ensure that the platform remains at the forefront of membrane biology research.

Biological Materials

Cell lines (available from the corresponding author upon request)

i) Rat Insulinoma Cell Line INS-1 832/3 (Millipore Sigma, Cat. No: SCC208)

ii) Mouse Neuro-2a (N2A) cells (ATCC, Cat. No: CCL-131)

Proteins

· Pre-assembled t-SNAREs (SNAP25/Sytaxin1) (Reagent Setup)

· VAMP2 (Reagent Setup)

· Synaptotagmin (SYT) -1 (Reagent Setup)

· Complexin (Reagent Setup)

Reagents

· Ni-NTA agarose (Thermo Fisher, Cat. No. 30230)

· Imidazole (Thermo Fisher, Cat. No. I5513)

· Glycerol (J.T. Baker, Cat. No. 2136-03)

· TEV protease (Sigma-Aldrich, Cat. No. T4455)

· SUMO protease (Life Technologies, Cat. No. 12588-018)

· Bio-Rad protein assay kit (Bio-Rad, Cat. No. 500-0001)

· Laemmli Loading Buffer, 4X (Bio-Rad, Cat. No. 1610747)

· Tris-glycine SDS running buffer, 10× (Bio-Rad, Cat. No. 1610772)

· Mini protean TGX Stain free protein Gels, 10 well, 50ul (Bio-Rad, Cat. No. 4868094)

· Precision Plus Protein™ WesternC™ Blotting Standards (Bio-Rad, Cat. No. 1610385)

· Trans-Blot Turbo RTA Mini 0.2 µm Nitrocellulose Transfer Kit (Bio-Rad, Cat. No.1704270)

· Precision Plus Protein Dual Color Standards (Bio-Rad, Cat. No. 1610374)

· Bio-Safe Coomassie stain (Bio-Rad, Cat. No. 161-0786)

· ImageJ free software (US National Institutes of Health)

· MatLab (Mathworks)

· Milli-Q water

· VAMP2 mouse antibody (ProteinTech, 67822-1-1g)

· Synaptophysin mouse antibody (ProteinTech, 67864-1-1g)

· 35 mm glass bottom dishes (MatTek, Cat. No. P35G-1.5-14-C)

· Norland Optical Adhesive 61 (NOA 61; Thorlabs, Cat. No. NOA61)

· Sticky-Slide 8 Well high (Ibidi, Cat. No. 80808)

· sticky-Slide 18 Well (Ibidi, Cat. No. 81818)

Lipids

Caution

Please be advised that all lipid stocks are preserved in chloroform, which poses a chemical risk. Avoid inhalation of gases, fumes, vapors, or sprays. It is imperative to don appropriate protective gear when handling these materials. Ensure that all work involving chloroform is conducted within a fume hood.

Caution

Lipids are purchased in chloroform solution from suppliers like Avanti Polar Lipids and should be kept at −20 °C for optimal preservation. Special attention is required for the storage of lipid stocks, especially for hygroscopic varieties such as 1,2-dioleoyl-sn-glycero-3-phospho-L-serine (DOPS) and L-α-phosphatidylinositol-4,5-bisphosphate (PI (4,5) P2). These compounds are susceptible to precipitation if the containers are not sealed properly.

· 1,2-Dioleoyl-sn-glycero-3-phospho-l-serine (sodium salt) (DOPS; Avanti Polar Lipids, Cat. No. 840035)

· 1,2-Dioleoyl-sn-glycero-3-phosphocholine (DOPC; Avanti Polar Lipids, Cat. No. 850375)

· 1,2-Dioleoyl-sn-glycero-3-phosphoethanolamine-N-(7-nitro-2-1,3-benzoxadiazol-4-yl) (ammonium salt) (NBD-PE; Avanti Polar Lipids, Cat. No. 810145)

· L-α-Phosphatidylinositol-4,5-bisphosphate (brain, porcine, ammonium salt) (PI(4,5)P2 or PIP2), Brain extract, CHCl3/CH3OH/H2O (20:9:1) stock, Avanti Polar Lipids, Cat. No. 840046X or synthetic PI(4,5)P2, powder, (Avanti Polar Lipids, Cat. No. 850155P).

· ATTO 647N DOPE (Attotec, Cat. No. AD 647N-161).

· ATTO 488 DOPE (Attotec, Cat. No. AD 488-161).

SUVs

· Bond Breaker TCEP Solution (Thermo Scientific, Cat. No. 77720)

· Potassium Chloride (Sigma-Aldrich, Cat. No. P9541)

· HEPES (Fisher- Bioreagent, Cat. No. BP 310)

· Dithiothreitol (DTT) (RPI, Cat. No. D11000)

· Triton-X-100 (Sigma, CAS 9036-19-5)

· N-octyl-β-D-glucopyranoside (OG; Chem Impex Intl, Cat. No. 00234)

· Bio-Beads^TM SM2 Adsorbent Media (Bio-RAD Cat. No. 152-3920, 20-50 Mesh)

· OptiPrep (STEMCELL Technologies, Cat. No: 07820).

· Chloroform (American Bioanalytical, Cat. No. AB00350)

· EDTA (Sigma-Aldrich, Cat. No. EDS)

· n-Dodecyl-β-maltoside (Thermo Scientific, Cat. No. 89903)

· Sulforhodamine B (Invitrogen, cat. no. S1307)

· Mini dialysis casette 7000MWCO (Thermo Scientific, Cat. No. 69592)

GPMVs

· N-Ethylmaleimide (NEM) (Calbiochem, Cat. No. 34115)

· Sucrose (Fisher Cat. No. S-500 -210554)

· MemGlow 488 (Cytoskeleton, Cat. No. MG01-02)

· Calcium Chloride (Sigma-Aldrich, Cat. No. C5080)

Cell Culture

· Inverted contrast microscope for routine examination of cell cultures

· Water bath at 37 °C to warm cell culture media

· Incubator at 37 °C with 5% CO2

· Greiner centrifuge tube, 50 ml (Sigma-Aldrich, Cat. No. T2318)

· Cell culture dishes 100 mm (Corning Cat. No. 430167)

· Filtered sterile pipette tips (Fisher Scientific Cat. No.02-707-404)

· Pipet-Aid (Drummond, Cat. No. 4-000-101)

· Serological Pipettes (Corning, 10, 25 and 50 mL)

· Scissors (FisherSci)

Proteoliposomes:

¨ Autoclave (LBR Scientific LCC, TUTTNAUER)

¨ Shaker Incubator at 37 °C for E. coli growth (Eppendorf Innova-44R)

¨ Magnetic stirrer (VWR)

¨ High-speed centrifuge (Eppendorf 5427R)

¨ Low Speed Centrifuge for 15/50 ml tubes (Eppendorf, 5702-R)

¨ Ultracentrifuge (Optima L-90k ultracentrifuge, SW60 Ti rotor; SW55 rotor; JS5.0 rotor Beckman)

¨ Table Spin centrifuge (Argos Flexifuge)

¨ Magnetic Stirrer (VWR)

¨ pH meter (Fisher Scientific, AB15)

¨ Refrigerator (−80 °C) (Haier, HYCD-290)

¨ Refrigerator (4 and −20 °C) (Haier, DE-25W262)

¨ Microcentrifuge tube 1.7 ml (Posi-Click Denville, 1149K01)

¨ Microcentrifuge tube 0.6 ml (Posi-Click Denville, C-2176)

¨ Serological pipette 5 ml (FALCON, 357543)

¨ Serological pipette 10 ml (FALCON, 356551)

¨ Serological pipette 25 ml (Basix^TM, 20210904-071-B)

¨ 10 ml petri dish (VWR)

¨ Heating Block (VWR Scientific Products)

¨ Cell Spreader (VWR)

¨ Water Bath (Fisher Scientific)

¨ JS 5.0 liner (Beckman Coulter, Cat. No 368735-40)

¨ D-tube dialyzer, 6–8 kDa MWCO (Novagen, cat. no. 71509-3)

¨ Centrifuge tubes, 4 ml, 11 × 60 mm (Thinwall, Ultra-Clear; Beckman, Cat. No. 344062)

¨ Amicon Ultra-0.5, 30-kDa membrane (Millipore, Cat. No. UFC503096)

¨ PVDF membrane, 0.2 μm (Macherey-Nagel, cat. no. 740401.50)

¨ FluoroNunc Maxisorp 96-well white plate (Thermo, Cat. No. 437591)

¨ Poly-Prep chromatography columns (Bio-Rad, Cat. No. 731-1550)

¨ Glass tube, 12 × 75 mm (Fisher, Cat. No. 14-961-26)

¨ Vacuum desiccator

¨ Vacuum pump (Alcatel 2005 C1, ultimate vacuum <10−4 mbar)

¨ Balance

¨ Gas-tight glass syringe (Hamilton, Cat. No. 1701RN, 1710 RN, 1750 RN)

¨ SpectraMax M5 multimode microplate reader

¨ Digital vortex mixer (VWR, Cat. No. 14005-824)

¨ Liquid chromatography system (ÄKTA Explorer; GE Healthcare)

¨ Superdex 200 10/300 GL prepacked column

¨ Orbital shaker (Barnstead Lab, multipurpose rotator (2314Q))

¨ Harrick Plasma Basic Plasma Cleaner

¨ Nitrogen tank and gauge

¨ Automatic high sensitivity osmometer (5004 Micro-Osmette; Precision Systems)

Microscopy

Critical: There is a diverse range of hardware and acquisition softwares available, allowing for various effective combinations. We describe our specific setups below, this protocol can be adapted to any microscope system that includes a minimum of two laser lines, two fluorescence channel detectors, and an enclosure capable of regulating temperature.

Confocal Microscope

Inverted confocal laser-scanning microscope (Leica SP8) equipped with a GalvoZ stage. Following parts of the microscope are necessary to perform membrane fusion experiments.

· Objectives (Leica, model no. HC PL APO 40x/1,10 W CORR CS2 with 0.65 mm long working distance is required for imaging the lipid bilayer.)

· Lasers: argon (488 nm) and helium–neon (HeNe; 633 nm)

· Detectors: 3× spectral PMT and 1× transmission PMT for bright-field microscopy, differential interference contrast (DIC) and polarization, respectively

· Microscope enclosure with temperature control

· GalvoZ stage with inserts for slides and 35-mm dishes.

TIRF Microscope

1. A Nikon Ti-E Sapphire fully motorized body (Nikon Instrument, Japan) with a perfect focus system (PFS)

2. EMCCD camera (Andor IXON 887)

3. Lasers (488 nm, 532 nm, and 633 nm)

4. Dual View opto-split suitable for collecting fluorescence from 488 and 633 nm channels.

5. A Plan Apo 60× NA 1.45 objective lens (Nikon Instruments)

6. Acquisition software: NIS-Elements AR 5.30.02 (Nikon)

The optical configuration should be set according to the dyes used.

EQUIPMENT SETUP

Preparation of BSA-coated coverslips and chambers

Incubate MatTek glass bottom dishes in 1 mL of 1 mg/ml BSA solution for 1 h. Wash two times with GPMV buffer. The coating should be done immediately before use.

Plate reader setup for lipid mixing

Configure the plate reader to incubate at 37 °C. Adjust the excitation and emission wavelengths to 460 nm and 538 nm, respectively, with an automatic cutoff set at 530 nm. Set the photomultiplier tube (PMT) to a low sensitivity level. Program the reader to take measurements at 1-minute intervals over a total acquisition period of approximately 120 minutes.

Preparation of Silicon chips for confocal microscopy

Silicon chips are glued on a glass coverslip using optical transparent glue NOA61 to fit in an Ibidi Sticky-Slide 8 Well high. See PROCEDURE for more details.

Preparation of Silicon Surface for TIRF microscopy

Silicon chip of thickness = 4.7 µm are attached to Ibidi Sticky-Slide 18 well in a format. See PROCEDURE for more details.

Confocal microscopy with incubation systems

Refer to Box 1 for a sample of specific image acquisitions and the settings utilized in our system.

Box 1: Setup for Leica SP8 Confocal Microscope

1. Turn on temperature within the microscope enclosure to maintain 37 °C.

2. In ‘configuration’, turn on the argon (488-nm) and HeNe (633-nm) lasers.

3. Activate PMT 1, tick ‘visible’, set laser power to 2% and spectral range to 500–550 nnm, click ‘PMT 1’ and set gain to 800 (change this later depending on signal strength).

4. Activate PMT 2, tick ‘visible’, set laser power to 5% and spectral range to 650–700 nm, click ‘PMT 2’ and set gain to 800 (change this later depending on signal strength).

5. Activate TMD, tick ‘visible’, set gain to 250 (change this later depending on signal strength).

6. In ‘Acquire’ → ‘Acquisition’:

· Set ‘Acquisition mode’ to XYT

· Set ‘Format’ to 512 × 512

· Set ‘Speed’ to 1800 or 8000 Hz (change this depending on the experimental requirement)

· Set ‘Time interval’ to Minimize

· Set ‘No of Frames’ to 5000

TIRF microscopy with incubation systems

Refer to Box 2 for a sample of specific image acquisitions and the settings utilized in our system.

Box 2: Setup for NIKON-TIRF Microscope

1. Turn on the microscope and required laser lines (for example 488, 532 and 647 nm).

2. Select the 60X TIRF oil objective and put a drop of Nikon Immersion Oil Type F (Index = 1.518 at 23°C).

3. Place the SLIM chip with a slide holder.

4. Turn on the light and focus the array of holes on the chip.

Critical: The SLIM chip is easier to focus with the bright light in Epifluorescence mode.

5. Switch off the bright light and turn on the required laser line.

6. Change the setting to TIRF mode.

Critical: The TIRF angle maybe 2-3 degrees different than the corresponding glass slide. Therefore, it may need initial scanning to find the exact TIRF mode.

7. For simultaneous dual color monitoring, insert the corresponding dual viewer to split the beam.

Reagent Setup

t-SNAREs (SNAP25, Syntaxin1b)

t-SNARE proteins can be expressed using a bicistronic vector, pTW34 as previously detailed.¹⁸ After purification, aliquot and flash freeze the protein in liquid nitrogen in buffer consisting of 400 mM KCl, 25 mM HEPES, 1% (wt/vol) n-Octyl β-D-glucopyranoside (OG), 1 mM DTT, and 10% (v/v) glycerol. Adjust the pH of this solution to 7.4. For long-term storage, maintain this preparation at −80 °C, where it remains stable and active for up to six months.

6×-His-SUMO-VAMP2

Express and purify full-length, wild-type mouse VAMP2 as previously described.¹⁹ After purification, aliquot and flash freeze the protein in liquid nitrogen in buffer consisting of 400 mM KCl, 25 mM HEPES, 1% (wt/vol) n-Octyl β-D-glucopyranoside (OG), 1 mM DTT, and 10% (v/v) glycerol. VAMP2 maintains its activity for up to 6 months when stored at −80 °C.

Synaptotagmin 1-5xHis

SYT1 protein used in this assay is a truncated protein (residues 47-421) consisting of the transmembrane domain and the complete cytosolic domain. Expression and purification have been detailed previously.¹³ After purification, aliquot and flash freeze the protein in liquid nitrogen in buffer consisting of 400 mM KCl, 25 mM HEPES, 1% (wt/vol) n-Octyl β-D-glucopyranoside (OG), 1 mM DTT, and 10% (v/v) glycerol. SYT1 maintains its activity for up to 2 months when stored at −80 °C.

Complexin 1

Complexin 1 can be expressed as previously described.^20,21 After purification, aliquot and flash freeze the protein in liquid nitrogen in buffer consisting of 140 mM KCl, 25 mM HEPES, 1 mM TCEP and 10% (v/v) glycerol. Complexin 1 retains its activity for up to 6 months when stored at −80 °C. Do not re-freeze thawed aliquots.

Synthetic SUVs (V-SUVs and V-Syt1 SUVs)

Make SUVs using DOPC/DOPS/ATTO-647N = 77 / 20 / 3 mol%. For constructing VAMP2-SUVs (vSUVs), maintain a Lipid/Protein ratio of 100. To create VAMP2-Synaptotagmin1 SUVs (SYT1-vSUVs), adjust the Lipid/Protein ratio of VAMP2/SYT1 to 100:250. Refer to the 'Preparation of SUVs' section in the PROCEDURE for detailed instructions.

Note! The protein-to-lipid ratio can be changed depending on the experiment performed. The reconstituted liposomes are assumed to have 55% protein facing outward and 45% facing inward.²²

GPMV buffer (10 mM HEPES, 150 mM KCl, 2 mM CaCl2, 10mM Sucrose, pH7.4):

Mix 18.75 mL of KCl (2M Stock), 0.5 mL of CaCl2 (1M Stock), and 2.5 mL of HEPES (1M Stock - pH 7.4), 5 mL of sucrose (0.5M Stock) in distilled water and bring to 250 mL. This can be stored at room temperature or at 4 °C for 6 months.

Note: Sucrose is mandatory to burst GUVs

GPMV working buffer (2 mM NEM):

Dissolve 2.5 mg NEM in 10 mL of GPMV buffer. The solution should be freshly prepared.

Dialysis Buffer A (5X):

Prepare 2 L of a buffer (pH 7.4) containing 600 mM KCl, 125 mM HEPES, and 5 mM DTT. This buffer should be freshly prepared.

Dialysis Buffer B (1X):

Take 1 L of Buffer A and make up to 5 L with Milli-Q water in a beaker.

t-SUV reconstitution buffer

Add 1% (wt/vol) OG to 10 mL of a dialysis buffer A (5X).

v-SUV reconstitution buffer

Add 1% (wt/vol) OG to 10 mL of a dialysis buffer B (1X).

MemGlow™ 488: Fluorogenic Membrane Probe Solution (20 µM):

For creating a 20 µM stock solution, reconstitute the lyophilized powder in 100 µl of anhydrous DMSO. Then, aliquot this stock into 5 µL aliquots and store them at -80 °C.

Preparation of Giant Unilamellar Vesicles

1. Use option A to prepare giant plasma membrane (cell-derived) vesicles and option B to prepare synthetic giant unilamellar vesicles (GUVs) from small unilamellar vesicles (SUVs) with proteins of interest (Figure 2).

(A) Preparation of Giant Plasma Membrane Vesicles (GPMVs)

▲ CRITICAL STEP: This procedure has been tailored for cells cultured in 100-mm diameter cell culture dishes. The yield of GPMVs is closely linked to factors such as cell quantity, incubation time and temperature. Optimal results were observed at an incubation time of 1 hour and 30 minutes. However, it's important to fine-tune these parameters for each unique cell type. The conditions outlined in this protocol have been successfully applied to various cell lines, including INS1, N2A, HEK293, HeLa, and HUVEC.

i. Culture the cells in a 100 mm dish until the cells reach 80% confluency (1-2 days). Refer to the American Type Culture Collection database at (http://www.atcc.org/) or refer to the guidelines provided by the supplier to determine the suitable conditions and media for use.

ii. Remove the cell culture media and wash the cells three times with 5 mL of PBS and 5 mL of GPMV buffer.

iii. Add 1 mL of GPMV working buffer to the cells and incubate for 1h 30 mins. [Optional] While this protocol uses NEM to maintain protein functionality, the same process can be carried out using 25 mM paraformaldehyde and 2 mM DTT to produce GPMVs.

iv. Transfer the GPMV-rich cellular supernatant into a microcentrifuge tube using cut-pipette tips.

■ PAUSE After isolation, GPMVs can be stored at 4 °C or ice to transfer to the silicon chip.

[optional] For microscopic visualization of GPMVs, they can be labeled by incubating with MemGlow™ 488 Fluorogenic Membrane Probe Solution at a concentration ranging from 2 to 20 nM. Afterward, place the labeled GPMVs onto BSA-coated MatTek glass-bottom dishes for confocal microscopy imaging.

? TROUBLESHOOTING

Troubleshooting advice can be found in Table 1.

(B) Preparation of Giant Unilamellar Vesicles (GUVs) from SUVs

i. Mix DOPC, DOPS, PIP2, and ATTO465 in a molar ratio of 76.5:20:3:0.5 in a chloroform-rinsed glass tube, aiming for a final concentration of 7 mM in a 50 µL volume. Subsequently, dry the mixture under a stream of nitrogen to remove the solvent. Throughout this procedure, shield the glass tubes with aluminum foil to protect the fluorescent lipids.

▲ CRITICAL STEP: The amount of lipids can be adjusted depending on the experiment

ii. Rehydrate the dried lipids with a combination of t-liposome reconstitution buffer and t-SNARE proteins, adjusting the total volume to 50 μL to achieve a protein: lipid ratio of 1:800.

▲ CRITICAL STEP: The protein to ratio can be adjusted depending on the experiment

iii. Mix by vortexing for 15 min at room temperature.

iv. Dialyze the mixture against 1 liter of dialysis buffer A overnight at 4 °C.

v. Next day, collect the t-SUVs and store them at 4 °C

■ PAUSE SUVs can be stored at 4 °C for up to 2-3 days

vi. Add sucrose (final concentration of 2 mM) to the t-SUV solution.

vii. A few 2 μL drops of t-SUV-sucrose mixture solution is placed on the MatTek glass bottom dish and dried at room temperature under atmospheric pressure.

viii. Rehydrate the dried lipid film spots with 6 μL of deionized water and dry again under the same conditions. In the next step, the lipid film is rehydrated again with 10 µL of deionized water for 40-50 mins.

? TROUBLESHOOTING

Troubleshooting advice can be found in Table 1.

ix. Gently pipette up-down 1-5 times using a pipette with a cut tip at the edge of the droplet to collect 10 µL from each droplet. The two droplets can be transferred to the prepared silicon surface.

■ PAUSE: Multiple drops can be pooled and stored at 4 °C.

▲ CRITICAL STEP: The quality of GUVs is better if they are recovered before 50 mins

Cleaning the Silicon Chip:

2. Add 400 μL of 10% RBS solution to each chip and gently shake for 5-10 minutes.

3. Submerge the chips in running water for 5 minutes to eliminate any residual RBS solution.

4. Thoroughly wash the chips with MiliQ water five times, shaking gently for 1 minute each time.

5. Apply 400 μL of 200-proof ethanol, shake for 1 minute, and repeat this process three times.

6. Utilize 400 μL of methanol with 1-minute gentle shaking, repeating the process three times.

▲CRITICAL STEP: Ensure meticulous washing to eliminate any trace of RBS detergent, as this can impact subsequent bilayer formation.

7. Discard all solutions and either air-dry the chip or incubate at 37 °C for 15 minutes.

Formation of Planar Bilayer on the Chip.

8. After cleaning and drying the chip, treat the chip for 5 minutes to clean and activate the surface.

9. Add 400 µL of 200 proof ethanol and incubate for 5 minutes with gentle rocking. This step removes all the air trapped within the chip.

10. Replace the ethanol with water, being careful to avoid air bubbles entering the chip. Mix water 1:1 (v/v) with the solvent in the chamber 5 times. ▲CRITICAL STEP: After ethanol wash, the chip should be kept in water and not be left dry as this will introduce air bubbles into the chip and the bilayer will not form smoothly.

11. For the confocal chip, water should be replaced with 25 mM HEPES (pH 7.4), 120 mM KCl, 0.2 mM TCEP and 5 mM MgCl₂ buffer. For the TIRF SLIM chip, water should be replaced with buffer containing 45% OptiPrep, 25 mM HEPES (pH 7.4), 120 mM KCl, 0.2 mM TCEP and 5 mM MgCl₂. This 45% OptiPrep is required to match the refractive index for the TIRF setup.

▲CRITICAL STEP: A 5 mM MgCl₂supplement is crucial to adhere and burst the GUV or GPMV on the silicon Chip.

12. Cut the tip of a 20 µL tip to make the orifice bigger. Gently collect previously prepared GUV or GPMV solution and dispense onto the silicon chip.

13. Incubate the lipids on the chip for 20 minutes to form the bilayer. Place the chip in the JS5.3 holder and centrifuge the chip at 100 x g at RT for 5 minutes.

▲CRITICAL STEP: This gentle spin helps to settle GUVs/GPMVs and spread onto the chip nicely.

14. Gently exchange the buffer on the top side of the plasma membrane to remove any unbound GUVs or GPMVs.

15. Visualize the membranes using a confocal or TIRF microsopcy depending on the chip used. For the expected result, see Figure 3

Preparation of SUVs

16. Mix lipids (see REAGENT SETUP) at the desired ratio in glass tubes and evaporate the solvent using nitrogen gas. Remove trace amounts of organic solvent under a high vacuum for 2 hours.

17. Hydrate the dried lipids with v-SUV reconstitution buffer. Appropriate amounts of protein are added to mimic the physiological copy number of SYT1 (~15-20 copies) and VAMP2 (~60-70) copies. Maintain the detergent concentration ∼2 times above its critical micelle concentration.

18. Add sulforhodamine B solution to achieve a final concentration of 50 mM.

19. Vortex the tubes containing solution for thorough component dispersal.

20. Add 2 times the volume of dialysis buffer B to rapidly dilute the mixture below CMC, reducing detergent concentration below its critical micelle concentration. Rapid vortexing is necessary while adding the dialysis buffer B. Incubate the solution for 20 mins in a mixer.

21. Pass the sample solution through a prewashed and pre-equilibrated CL-4B column. This step is essential for separating detergents and other impurities from the protein-lipid-detergent mixture and for removing any unencapsulated sulforhodamine B dyes. Before introducing the sample, ensure that the CL-4B column is properly prepared with a prewash and equilibration using vesicle buffer. Once the sample is in the column's stationary phase, use vesicle buffer as the eluent.

22. Dialyze the collected vesicle solution in a 7 kDa MWCO dialysis cassette into 4 liters of dialysis buffer B overnight at 4 °C with 4 g of Bio-Beads.

23. The next day, set up a float-up of the samples using OptiPrep, mixing dialyzed samples 1:1 (vol/vol) with 60% (vol/vol), and the layer of 250 µL of 20% (vol/vol) and 50 µL of 0% in dialysis buffer B in 14 mm × 89 mm Ultra-Clear Beckman centrifuge tubes. Centrifuge in a Beckman SW55 Ti rotor at 48,000 rpm for 4 hours at 4–8 °C.

▲CRITICAL STEP: Maintain the various layers, which must be carefully overlaid to prevent any disturbance to the interface between them.

24. Collect 75 μl of liposomes from the 0% to 20% interface, aiming for a clean, concentrated fraction. The collected SUVs should be used on the same day for better results.

25. Assess the protein incorporation ratio in Syt1-vSUVs through SDS-PAGE gel electrophoresis and Coomassie blue staining. Load 20–50 μl of the vesicle solution into the gel initially. If the protein bands appear too faint or overloaded, make volume adjustments and rerun the gel. For the expected result, see Figure 4. Ensure that the synaptobrevin-to-synaptotagmin ratio approximates 4.6:1.

26. Characterize the size distribution of SUVs by dynamic light scattering or cryoelectron microscopy.

Experiment

Vesicle Docking and Fusion

27. Verify the membrane fluidity using the confocal FRAP module.

28. Incubate the chamber with 2 µM of Complexin protein.

29. Dilute 5 µL of the Syt1-vSUVs stock in 995 µLof dialysis buffer B. Then add 5 µL of liposomes to the chamber containing 400 µL of buffer.

▲CRITICAL STEP: The dilution has to be adjusted to find a good compromise between the number of vesicles docked per hole. High dilution results in fewer vesicles per hole.

? TROUBLESHOOTING

Troubleshooting advice can be found in Table 1.

30. Observe the same area under continuous illumination using medium laser intensity. Allow vesicles to dock on the membrane.

[optional]: Check that the green and red lasers are aligned correctly by imaging a multicolor fluorescent bead on a MaTeK glassbottom dish. Ensure that the individual laser beams are imaging on the same area of the sample plane.

31. After the vesicles are docked on the membranes, the excess of the vesicles is removed by exchanging them with a new buffer.

32. Start the 'live' acquisition for 5000 frames at 1800 or 8000 Hz (depending on the experimental requirement) to check the bilayer and vesicle dye fluorescence intensity imaging channels and the focus. The same protocol can be applicable for TIRF microscopy.

33. After starting live data acquisition for a small strip or area to monitor the fusion at a high temporal resolution, wait approximately 5 seconds and carefully introduce a drop of Ca²⁺ solution (to reach a final Ca2+ concentration in vesicle buffer) using a pipette.

34. Save movies of fields of view for both bilayer and vesicle channel channels.

Single Molecule imaging

35. Start the 647-nm laser to observe Alexa-647 signals, and/or open the shutter for the 488-nm laser to observe docked vesicle signals.

36. Drop immersion oil with a refractive index (n = 1.518) that matches that of the coverslip on a 60× numerical aperture (NA) 1.49 objective lens.

37. Place the samples on the stage of the TIRF microscope and adjust the lens position. Next, find a lipid bilayer area, and optimize the focus, intensity and incidence angle of the TIRF microscope excitation lasers.

38. Add the vesicles and Alexa-647 fluorophore conjugated Complexin protein.

39. Record live images under TIRFM by using the NIS-Elements software at a rate of100 ms.

Data Analysis

40. Run the ImageJ and/or MATLAB software.

A) Stepwise bleaching-step analysis

41. Load and visualize the image sequences in TIFF or JPEG format.

42. Adjust the threshold to identify the vesicle puncta on the 488 nm channel images. Mark the ROI and apply the same ROI to the 647 nm channel.

43. Using plot Z-stack option to 'Track all frames' to measure the intensity at 488 and 647 nm channel.

44. Save the data in txt format and plot using any graphing software.

B) Vesicle-Docking and Fusion analysis

45. Vesicle docking and fusion analysis can be performed using ImageJ or MATLAB. Our previously published Fusion Analysis software, along with its source code, is accessible on Matlab Central.¹⁶

46. In summary, the dynamic puncta tracking algorithm extracts the information of vesicles in a space-time trajectory (x,y,t) and analyzes lipid diffusion and cargo release fusion in a two dimensional (2-D) dataset. Centroid and Gaussian ﬁt were employed to track the location of the vesicles with high accuracy.

47. Dynamic ROI coordinates of the ﬂuorescent lipid of the vesicle channel were used to track the ﬂuorescence of the content mixing channel at a given time.

Step

Problem

Possible reason

Solution

1A (iv)

Not Enough GPMVs

Not enough cells, vesiculation reagents, incubation time

Try different concentration of reagents, time and number of cells, minimize cell adherence to dish

1B(viii)

Not Enough GUVs

Low concentration of lipids; Not enough drying and rehydrating cycles

Increase the concentration of lipids. Repeated drying and hydration cycles can improve the yield and size of GUVs, likely from lipid reorganization in each cycle.

Not enough vesicle docked

Low concentration of liposomes

Increase the amount of liposomes added to the chamber

Steps 1 GUV preparation: ∼2-3 h

Steps 2–7 cleaning chips: ∼40 mins

Steps 8–15, Formation of lipid bilayer: ∼2-3 h

Steps 16–26, vesicle preparation: 18-24 h

Steps 27-39, Single molecule counting and Fusion experiment: ~2-3 h

Steps 40–86 Data analysis: ∼10 h

FRAP Analysis:Ensuring a uniform and fluid lipid bilayer is essential for meaningful fusion experiments. Figure 5 presents FRAP data from various experiments, demonstrating reproducibility when maintaining consistent parameters. The FRAP data, along with the MATLAB files used for analysis and corresponding instructions, are available in the Supplementary Methods.

Vesicle Docking and Fusion Analysis: Prior to Ca2+ injection, both content and lipid dye fluorescence intensities at each observed spot should remain relatively constant (Figure 6A). Post-Ca2+ injection, statistically significant jumps in these intensities are expected at specific times, depending on experimental conditions and system factors. Fluorescence intensities should exhibit stability before and after each jump, with occasional gradual drifts due to photobleaching. These intensity jumps categorize events into immediate fusion after Ca2+ addition (Figure 6B). Vesicle leakage is distinguishable, characterized by a small rapid increase and subsequent rapid complete decay of content dye fluorescence intensity. Analysis of hundreds to thousands of time traces is typically necessary for statistically significant histograms.

Synchronized calcium-triggered fusion, mediated by SNAREs and SYT1 in the presence of complexin. Particularly, complexin significantly boosted immediate fusion, especially at a lower (100 μM) Ca2+ concentration. Ongoing investigations explore the impact of additional factors on our system, including Munc18 and Munc13.

Single Molecule Counting: Fluorescent subunit counting data were collected by measuring Alexa-647 Complexin signals. The fluorescence signal of each trace was plotted as a function of time using imageJ and bleaching steps were counted (Figure 7). This protocol presented here measure the CPX signals. However, the methodology can be extended to counting other fusion protein as well as any membrane protein of interest.

1 Rothman, J. E. The Principle of Membrane Fusion in the Cell (Nobel Lecture). Angew Chem Int Edit 53, 12676-12694 (2014). https://doi.org/10.1002/anie.201402380

2 Trimbuch, T. & Rosenmund, C. Should I stop or should I go? The role of complexin in neurotransmitter release. Nature Reviews Neuroscience 17, 118-125 (2016). https://doi.org/10.1038/nrn.2015.16

3 Kaeser, P. S. & Regehr, W. G. Molecular mechanisms for synchronous, asynchronous, and spontaneous neurotransmitter release. Annu Rev Physiol 76, 333-363 (2014). https://doi.org/10.1146/annurev-physiol-021113-170338

4 Brunger, A. T., Choi, U. B., Lai, Y., Leitz, J. & Zhou, Q. Molecular Mechanisms of Fast Neurotransmitter Release. Annu Rev Biophys 47, 469-497 (2018). https://doi.org/10.1146/annurev-biophys-070816-034117

5 Rizo, J. Molecular Mechanisms Underlying Neurotransmitter Release. Annu Rev Biophys 51, 377-408 (2022). https://doi.org/10.1146/annurev-biophys-111821-104732

6 Sudhof, T. C. Neurotransmitter release: the last millisecond in the life of a synaptic vesicle. Neuron 80, 675-690 (2013). https://doi.org/10.1016/j.neuron.2013.10.022

7 Boyken, J. et al. Molecular Profiling of Synaptic Vesicle Docking Sites Reveals Novel Proteins but Few Differences between Glutamatergic and GABAergic Synapses. Neuron 78, 285-297 (2013). https://doi.org/10.1016/j.neuron.2013.02.027

8 Weingarten, J. et al. The proteome of the presynaptic active zone from mouse brain. Mol Cell Neurosci 59, 106-118 (2014). https://doi.org/10.1016/j.mcn.2014.02.003

9 Holderith, N., Heredi, J., Kis, V. & Nusser, Z. A High-Resolution Method for Quantitative Molecular Analysis of Functionally Characterized Individual Synapses. Cell Rep 32 (2020). https://doi.org/ARTN 107968

10.1016/j.celrep.2020.107968

10 Gallagher, B. R. & Zhao, Y. X. Expansion microscopy: A powerful nanoscale imaging tool for neuroscientists. Neurobiol Dis 154 (2021). https://doi.org/ARTN 105362

10.1016/j.nbd.2021.105362

11 Bera, M., Ramakrishnan, S., Coleman, J., Krishnakumar, S. S. & Rothman, J. E. Molecular determinants of complexin clamping and activation function. Elife 11 (2022). https://doi.org/10.7554/eLife.71938

12 Ramakrishnan, S., Bera, M., Coleman, J., Rothman, J. E. & Krishnakumar, S. S. Synergistic roles of Synaptotagmin-1 and complexin in calcium-regulated neuronal exocytosis. Elife 9 (2020). https://doi.org/10.7554/eLife.54506

13 Ramakrishnan, S. et al. Synaptotagmin oligomers are necessary and can be sufficient to form a Ca(2+) -sensitive fusion clamp. FEBS Lett 593, 154-162 (2019). https://doi.org/10.1002/1873-3468.13317

14 Coleman, J. et al. PRRT2 Regulates Synaptic Fusion by Directly Modulating SNARE Complex Assembly. Cell Rep 22, 820-831 (2018). https://doi.org/10.1016/j.celrep.2017.12.056

15 Heo, P. et al. Highly Reproducible Physiological Asymmetric Membrane with Freely Diffusing Embedded Proteins in a 3D-Printed Microfluidic Setup. Small 15, e1900725 (2019). https://doi.org/10.1002/smll.201900725

16 Ramakrishnan, S. et al. High-Throughput Monitoring of Single Vesicle Fusion Using Freestanding Membranes and Automated Analysis. Langmuir 34, 5849-5859 (2018). https://doi.org/10.1021/acs.langmuir.8b00116

17 Kalyana Sundaram, R. V. et al. Native Planar Asymmetric Suspended Membrane for Single-Molecule Investigations: Plasma Membrane on a Chip. Small n/a, 2205567 (2022). https://doi.org/https://doi.org/10.1002/smll.202205567

18 Weber, T. et al. SNAREpins: minimal machinery for membrane fusion. Cell 92, 759-772 (1998). https://doi.org/10.1016/s0092-8674(00)81404-x

19 Shen, J., Tareste, D. C., Paumet, F., Rothman, J. E. & Melia, T. J. Selective activation of cognate SNAREpins by Sec1/Munc18 proteins. Cell 128, 183-195 (2007).

20 Giraudo, C. G., Eng, W. S., Melia, T. J. & Rothman, J. E. A clamping mechanism involved in SNARE-dependent exocytosis. Science 313, 676-680 (2006). https://doi.org/10.1126/science.1129450

21 Kummel, D. et al. Complexin cross-links prefusion SNAREs into a zigzag array. Nat Struct Mol Biol 18, 927-933 (2011). https://doi.org/10.1038/nsmb.2101

22 Ji, H. et al. Protein determinants of SNARE-mediated lipid mixing. Biophys J 99, 553-560 (2010). https://doi.org/10.1016/j.bpj.2010.04.060

SI1Chipdesigns.zip
Chip Designs
SI2VesicleFusion.zip
Vesicle Fusions

Download PDF

Version 1

posted

You are reading this latest protocol version

Long-Lived Planar Asymmetric Suspended Membranes for Investigating Protein Dynamics and Membrane Fusion

Status:

Version 1

Abstract

Figures

Introduction

Reagents

Equipment

Procedure

(B) Preparation of Giant Unilamellar Vesicles (GUVs) from SUVs

Troubleshooting

Time Taken

Anticipated Results

References

Supplementary Files

Associated Publications

Status:

Version 1

Privacy Policy

Terms of Service

Long-Lived Planar Asymmetric Suspended Membranes for Investigating Protein Dynamics and Membrane Fusion

Status:

Version 1

Abstract

Figures

Introduction

Reagents

Equipment

Procedure

(B) Preparation of Giant Unilamellar Vesicles (GUVs) from SUVs

Troubleshooting

Time Taken

Anticipated Results

References

Supplementary Files

Associated Publications

Status:

Version 1

Privacy Policy

Terms of Service

Manage Cookie Preferences