Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Jan 1.
Published in final edited form as: Methods. 2021 Feb 18;197:13–19. doi: 10.1016/j.ymeth.2021.02.010

Topographically smooth and stable supported lipid bilayer for high-resolution AFM studies

Siddhartha Banerjee 1, Yuri L Lyubchenko 1,*
PMCID: PMC8371085  NIHMSID: NIHMS1677087  PMID: 33609699

Abstract

The cellular membrane has been identified to play a critical role in various biological processes including the assembly of biological systems. Membranes are complex, primarily two-dimensional assemblies with varied lipid compositions depending on the particular region of the cell. Supported lipid bilayers are considered as appropriate models for physio-chemical studies of membranes including numerous single molecule techniques. Atomic force microscopy (AFM) as a topographic technique is a fully appropriate single molecule technique capable of direct observation of molecular processes on membranes. However, reliable experimental AFM studies require the preparation of the bilayer with a sub-nanometer smooth morphology, which remains stable over long-time observation. Here we present the methodology, which allows one to prepare a smooth, stable, structurally homogeneous lipid bilayer without the presence of any trapped vesicles. We described the application of such lipid bilayers to probe time-dependent early stages of aggregation of monomeric amyloid proteins. Importantly, the proposed methodology can be extended to bilayers with various compositions, by incorporating different lipids for on-membrane aggregation study including cholesterol. Furthermore, this methodology development allowed us to monitor the aggregation of amyloid protein at its physiologically relevant low protein concentration. The flexibility of altering the membrane composition allows to identify the specific role of a particular lipid towards the aggregation kinetics, revealing the plausible mechanism of disease development.

Keywords: supported lipid bilayer, time-lapse AFM, amyloid, protein aggregation

1. Introduction

The cellular membrane plays a critical role in numerous biological events including inter cellular communication, molecular transport and different biochemical processes [1–4]. The major hurdle in working with natural membrane lies in its severe complexity [5–8]. Developing an experimental model membrane system, which is technically easier to work with, is essential to understanding the on-membrane biological events at a molecular level. Such system includes micelles, bicelles, liposomes or vesicles, lipid nano discs and supported lipid bilayers [9–16]. Among them, supported lipid bilayer mimics the cellular membrane most effectively and the presence of a solid support underneath the lipid bilayer allows the use of them in various surface based biophysical techniques [17–19]

Generation of a smooth, homogeneous, stable supported lipid bilayer has remained a challenge, in effectively developing such a model membrane system. The key shortcoming has been the presence of trapped vesicles, which are spherical in morphology and can easily be misled as protein monomers or protein-protein complexes. The trapped vesicles puts a limit in the usage of a supported lipid bilayer as a membrane for studying protein-lipid interaction, starting from the monomeric state of the protein. Another limitation has been the generation of small lipid bilayer patch on the solid support. These types of lipid bilayers are unstable and get dislodged from the support underneath very easily, resulting in ambiguous results and excluding the possibility for probing the on-membrane events for longer time periods. We have recently described a protocol enabling us to prepare stable and topographically smooth bilayers consisting of phospholipids 1-palmitoyl-2-oleoyl-glycero-3-phosphocholine (POPC) and 1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-L-serine (POPS) [17, 20, 21–22].

Here we extended this methodology for the preparation of a smooth, homogeneous, and stable lipid bilayer containing such an important component of membranes as cholesterol. The major advantage of this methodology is that the composition of the lipid bilayer can be changed by incorporating new lipids or by changing the existing lipid content. The results of the role of specific lipid molecules elucidate information about the biological process under study. We have incorporated cholesterol, which has second highest concentration in the cellular membrane after phospholipids and demonstrated a successful procedure for generating a suitable lipid bilayer comprising of phospholipids and cholesterol. We illustrated the major features of this by generating a lipid bilayer which allows us to probe early-stage aggregates which are the major toxic species in the disease development. We have applied time-lapse AFM, which provides a unique opportunity to monitor the on-membrane aggregation in real time, without any labelling of the protein which can significantly alter the aggregation propensity of Aß42. Furthermore, it allows to visualize the same area of the lipid bilayer over a course of a few hours enabling to visualize the in-situ formation of oligomers.

2. Supported lipid bilayer for probing amyloid aggregation

2.1. Preparative steps of supported lipid bilayer

In general, the total procedure for the preparation of a supported lipid bilayer is a multistep process (Fig. 1). First, the lipids which generally remain in a solvent like chloroform or chloroform/methanol are mixed in a glass vial according to the desired composition. Then, the solvents are evaporated by a gentle stream of Argon gas. As a result, a dried lipid film is produced at the bottom of the glass vial. To ensure the complete removal of the solvent, the vial is kept in a vacuum chamber overnight. Then, the lipid film is hydrated by the desired buffer. Occasional vortexing is performed to loosen the lipid film from the glass surface and then the vial is placed into a sonicator to produce the vesicles. Sonication is carried out for 45 minutes to 1 hour, until the solution becomes nearly clear without any presence of visible lipid film residues.

Figure 1. Schematic representation of the steps for preparation of supported lipid bilayer.

Figure 1.

Open in a new tab

(Step I) Vesicles which are prepared by sonication is deposited onto the mica surface. (Step II) The sample is incubated on a hot plate at 60°C for ~1 hr. (Step III) At the end of the incubation period the sample is brought back to the room temperature. (Step IV) The sample is gently rinsed with buffer. (Step V) The prepared lipid bilayer sample is always kept wet by the imaging buffer before the AFM scan.

After the preparation of the vesicles, they are deposited onto a freshly cleaved mica surface (Step I, Fig. 1). 100 μl volume is optimal for generating a liquid droplet onto a 0.5 cm × 0.5 cm mica piece, because a larger volume could overflow the mica substrate and potentially contaminate the lipid bilayer. Then, the sample is placed inside a petri dish and incubated on a hot plate at 60°C for 1 hr (Step II, Fig. 1). Additional buffer can be added to the sample to prevent it from drying out. After the incubation, the sample is allowed to cool to room temperature (Step III, Fig. 1) and then it is rinsed with buffer to remove excess vesicles. The rinsing step is carried out in a very gentle way, since the force exerted by the jet could potentially destroy the lipid bilayer (Step IV, Fig. 1). The imaging buffer is added on top of the substrate to keep the sample wet throughout the experiment (Step V, Fig. 1), as the drying effect could also lead to the rupture of the lipid bilayer. This step completes the lipid bilayer preparation. The sample is scanned by AFM immediately after the preparation.

2.2. AFM imaging of the supported lipid bilayer:

The supported lipid bilayers are a soft, fluidic system. Individual lipid molecules are not covalently tethered to the mica underneath, rather they are physisorbed on to the substrate. This makes the sample highly mobile. To successfully image the lipid bilayer, soft triangular cantilevers are used. We have chosen MSNL-E (Bruker) having the nominal spring constant of ~0.1 N/m and the resonance frequency of 7–9 kHz in buffer. The AFM imaging has been carried out in tapping mode with MFP-3D instrument (Asylum Research, Santa Barbara, CA) at room temperature. The typical scan speed was kept at 1 to 2 Hz. All the imaging experiments has been performed in buffer medium.

To characterize the formation of a lipid bilayer, a large area of the surface is scanned initially. This allows to locate the small defects or edge of the lipid bilayer surface, confirming the bilayer is formed (Fig. 2a). Subsequently, a smaller scan area is imaged to check the smoothness and homogeneity of the surface (Fig. 2b). Special attention is given to identify whether the prepared lipid bilayer contains any trapped vesicles or not. Trapped vesicles are not favored as they appear as globular features which could easily be misled as protein monomer or protein aggregates. The procedure mentioned in the section 2.1 generates smooth, homogeneous lipid bilayer, devoid of any trapped vesicles as shown in Fig. 2b. This type of lipid bilayer is ideal to study on-membrane protein aggregation starting from its monomeric state. Further characterization of the prepared lipid bilayer is performed by measuring the height of the layer from the mica surface (Fig. 2c). This measurement is achieved by taking a cross-section profile through any defect regions of the bilayer (as shown in Fig. 2a with a white line). A typical height value of 4.5 nm confirms the formation of proper lipid bilayer.

Figure 2. Characterization of supported lipid bilayer.

Figure 2.

Open in a new tab

(a) Large scan of the lipid bilayer on the mica substrate. The presence of few defects reveals the formation of the lipid bilayer on to mica surface. (b) zoomed image of the lipid bilayer. The bilayer is smooth, homogeneous, devoid of any trapped vesicles. (c) the cross-section profile drawn over a defect region as shown in panel ‘a’. The height value is measured as 4.5 nm which is typical for lipid bilayer.

2.3. The preparation of the lipid bilayer with different composition

This methodology for the preparation of lipid bilayer provides the unique opportunity to alter the composition of the membrane and to follow its effect in protein aggregation. We have worked with lipid molecules including POPC, POPS and cholesterol. Lipid bilayer composed of either single lipid or a combination of them have been prepared. Although the overall process as mentioned above (section 2.1) remains same, but few alterations are necessary as the lipid composition changes. For example, the incorporation of cholesterol does not allow the formation of a bilayer on the mica surface. The incubation of vesicles onto mica surface in PBS buffer generates very small lipid patches (Fig. 3a), where the majority of the mica substrate are not covered by the lipid bilayer. Such small lipid patches are not stable and cannot be utilized for longer period time-lapse imaging, as they are prone to be disrupted by repeated scanning of the AFM tip. The cross-section profile on such lipid patches exceeds 5.0 nm value indicating the incomplete flattening or spreading of the vesicles on mica surface (Fig. 3b). Incorporation of CaCl2 in the buffer containing 20 mM HEPES, 150 mM NaCl, pH 7.4 forms significantly larger patches (Fig. 3c) and the 4.5 nm height value of the layer indicates the formation of proper lipid bilayer (Fig. 3d). Such lipid bilayers also remain free from trapped vesicles, which is reflected in very low value of root mean square surface roughness. The roughness values were found to be 75 pm [23]. The percentage of lipid bilayer patch area in the presence and absence of CaCl2 has been plotted in two situations (Fig. 4), where the area of the lipid bilayer area has been calculated from a total area of 15 μm × 15 μm. Dramatic improvement in the bilayer area has been observed in presence of CaCl2 (Fig. 4). Cholesterol is generally known to make the lipid bilayer mechanically stiffer by restricting the movement of adjacent lipid molecules. We have measured the Young’s modulus of the lipid bilayer and compared the stiffness values in presence and absence of cholesterol by applying PeakForce Quantitative Nanomechanical Property Mapping mode (QNM) in NanoScope MultiMode 8 system (Bruker, Santa Barbara, California). This mode provides simultaneous data capturing for the topograph and the modulus map of the surface [23, 24]. Lipid bilayers with and without cholesterol are prepared and the same tip (MSNL from Bruker, Santa Barbara, California) is used for the measurements of the two samples. The spring constant of the cantilever used is determined by the thermal tune method with the help of the scanning software in-built in NanoScope MultiMode 8 system. Samples have been scanned in QNM mode keeping both the topograph and modulus channels on. After recording the data, analysis have been performed in Nanoscope Analysis software, where the modulus values are extracted from each pixel of the scanned surface and plotted into a histogram (Fig. 5). The Young’s modulus of the lipid bilayer without cholesterol has been measured to be 17.4 ± 3.0 MPa, whereas the presence of cholesterol raised the value up to 46.0 ± 14.5 MPa. It clearly demonstrated the higher mechanical stiffness of the surface in presence of cholesterol.

Figure 3. Effect of CaCl2 in preparation of lipid bilayer containing cholesterol.

Figure 3.

Open in a new tab

(a) Small lipid patches are observed when a buffer without CaCl2 is used. (b) Cross-section profile for one of those small patches. The height value is 5.4 nm which is higher compared to the typical lipid bilayer height. (c) Large lipid patch on mica surface with the buffer containing CaCl2. (d) Cross-section profile shows the height value of 4.5 nm.

Figure 4. Comparison of the lipid coverage with and without CaCl2.

Figure 4.

Open in a new tab

The bar plot shows the percentage coverage of the lipid bilayer on mica surface in two situations. The grey bar shows the bilayer coverage when CaCl2 is present, whereas the black bar shows the bilayer coverage without CaCl2. The coverage has been calculated from multiple segments of 15 μm × 15 μm areas.

Figure 5. Comparison of Young’s modulus in different lipid bilayer composition.

Figure 5.

Open in a new tab

(a) The Young’s modulus value of lipid bilayer obtained from the QNM imaging mode in absence of cholesterol and (b) in presence of cholesterol. The modulus value is obtained from each pixel of the scanned surface and then plotted as histogram. The plot is fitted with Gaussian distribution to obtain the peak value. The errors indicate the standard deviation.

3. Application of supported lipid bilayer to in situ aggregation of amyloid proteins

3.1. Time-lapse imaging of on-membrane amyloid aggregation

We have extensively applied the above-mentioned methodology to prepare a smooth, homogeneous, stable lipid bilayer which is devoid of trapped vesicles and used to monitor the aggregation of Aβ42. Fig. 6 shows the time-lapse imaging of Aβ42 aggregation on this model membrane system. The initial frame shows the lipid bilayer which remains clean and just before the addition of physiologically relevant concentration of Aβ42, 10 nM (Fig. 6a). Time-lapse imaging has been continued and aggregates were observed to be formed on the lipid bilayer surface within 2 hr (Fig. 6b). More aggregates were observed after the 5 hr experiment. To minimize, the tip effect on the Aβ42 aggregation and their appearances on the lipid membrane surface, scanning was not carried out continuously. After recording the data, the scan is paused, which leads to the withdraw of the tip electronically from the surface. This allows to probe the aggregation event with minimal perturbations by the AFM tip. It is to be noted here that the lipid bilayers remain stable for relatively long duration of the time-lapse experiment as shown here for 5 hrs. The lipid bilayer patch does not get destroyed by repeated scanning of the same area of the surface.

Figure 6. Time-lapse AFM imaging of the aggregation of 10 nM Aβ42.

Figure 6.

Open in a new tab

(a) The initial frame of the lipid bilayer which is clean. (b) The same area of the bilayer after 2 hr protein addition. The globular features are protein aggregates formed on the lipid bilayer surface. (c) More aggregates are observed after 5 hr of incubation.

3.2. Visualization of in-situ aggregate formation

The ability to scan the same area of the lipid bilayer allows to monitor the in-situ formation of aggregates on the membrane. Fig. 7 shows the series of high-resolution AFM images where the gradual generation of the oligomer with specific morphological feature has been observed. Fig 7a shows the presence of globular features which are Aβ42 aggregates formed on the membrane surface. The subsequent time-lapse images show the accumulation of the aggregates (Fig. 7 b–e). The cross-sectional profile has been generated across the feature to identify the growth of the aggregates along the z direction. The height of the feature increased with time, indicating the progressive on-membrane oligomer formation. The profile does not show a singular height peak (Fig. 7e). It indicates a specific annular shaped feature of the aggregate.

Figure 7. Visualization of in-situ aggregate formation.

Figure 7.

Open in a new tab

(a-f) Series of AFM images of the same area of the lipid bilayer surface showing the gradual formation of a large oligomer. Proteins accumulate on top of the already attached proteins to form larger aggregate. The cross-section profiles under each image represents the height values. Gradual increase in the height value is observed, indicating the elongation of the aggregate in the z-direction. The trough at the center of the profile in frame ‘e’ indicates the annular shape of the aggregate.

3.3. Visualization of in-situ fibril formation on membrane surface

The stability and large smooth surface area of the supported lipid bilayer allows to monitor the on-membrane aggregation event for a longer period of time. Such an example is shown in fig. 8. 10 nM Aβ42 has been incubated for 24 hr on the lipid bilayer composed of POPC, POPS and cholesterol. Figure 8a shows the larger scan area after 24 hr of incubation. The white arrows indicate the presence of protofibrillar aggregates. These aggregates are formed in-situ on the membrane surface at low protein concentration triggered by membrane-peptide interaction. The protofibrillar aggregates are formed homogeneously all over the surface, since they were observed in different areas of the surface (Fig. 8b–c). Fig. 8d shows a gallery of such protofibrils which are straight and devoid of any twisting or branches. These features are not very long which are generally found in the solution generated fibrils. The length of these protofibrillar aggregates are measured using Femtoscan software (http://www.nanoscopy.net/en/) and plotted as histogram (Fig. 8e). The data has been fitted with gaussian distribution to identify the peak value.

Figure 8. Formation of protofibrils after longer incubation of 10 nM Aβ42 on PC-PS-Chol lipid bilayer in presence of free cholesterol.

Figure 8.

Open in a new tab

(a) AFM topographic images of the lipid bilayer surface after 12 hr of incubation with Aβ42. The white arrows in ‘a’ indicate the presence of protofibrils on the surface. (b-c) Other areas also possess the protofibrils. (d) A gallery of protofibrillar aggregates found on the lipid bilayer surface (area 400 nm × 400 nm). (e) The histogram shows distribution of protofibril length. The length value indicates that these are not typical large mature amyloid fibrils.

4. Conclusion

In this work, we present the methodology for generating a properly supported lipid bilayer which can be used to probe lipid-protein interaction. The main features of these lipid bilayers have been smooth, vesicle free, homogeneous surface which allows to visualize the molecular events on the lipid bilayers unambiguously. Our approach, provides the opportunity to modify the model membrane composition by changing the lipid ratios or even incorporating new lipid molecules to mimic cellular membrane, which is a highly complex system. The necessary changes in the buffer composition, which is required when new lipids are introduced in the bilayer has also been discussed. This methodology is applied to probe the Aβ42 aggregation on lipid bilayer surface and the results brought a paradigm shift in understanding of the mechanism for disease development in which interaction with membranes drives the protein aggregation at physiologically low concentrations. The smoothness of the bilayer allowed us to identify very early aggregates which are generated on surface, at concentration of Aβ42 at low nanomolar range. The stability of the bilayer for longer period time provided the platform to monitor the aggregation beyond the oligomers state. In-situ formation of protofibrillar aggregates are also observed on the lipid bilayer. Overall, this development in the area of generating experimental model membrane system would provide new tools for directly monitoring various lipid-protein interactions.

Acknowledgements

The work was supported by the NIH grants GM096039 and GM118006 to Y.L.L. The authors thank Shaun Fillaux for the manuscript proofreading and useful comments.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Conflict of interest

Authors have no conflict of interest

References

  • [1].Coskun U, Simons K, Cell membranes: the lipid perspective, Structure 19(11) (2011) 1543–1548. [DOI] [PubMed] [Google Scholar]
  • [2].Nicolson GL, The fluid-mosaic model of membrane structure: still relevant to understanding the structure, function and dynamics of biological membranes after more than 40 years, Biochim Biophys Acta 1838(6) (2014)1451–1466. [DOI] [PubMed] [Google Scholar]
  • [3].Lombard J, Once upon a time the cell membranes: 175 years of cell boundary research, Biology Direct 9(32) (2014) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [4].Owen JS, McIntyre N, Gillett MPT, Lipoproteins, cell membranes and cellular functions, Trends in Biochemical Sciences 9(5) (1984) 238–242. [Google Scholar]
  • [5].Serna JB, Schütz GJ, Eggeling C, Cebecauer M, There is no simple model of the plasma membrane organization, Front Cell Dev Biol. 4(106) (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [6].Goñi FM, The basic structure and dynamics of cell membranes: an update of the singer–nicolson model, Biochimica et Biophysica Acta (BBA) – Biomembranes 1838(6) (2014) 1467–1476. [DOI] [PubMed] [Google Scholar]
  • [7].Yang NJ, Hinner MJ, Getting across the cell membrane: an overview for small molecules, peptides, and proteins, Methods Mol Biol. 1266 (2015) 29–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [8].Ernst R, Ballweg S, Levental I, Cellular mechanisms of physicochemical membrane homeostasis, Curr Opin Cell Biol. 53 (2018) 44–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [9].Dürr UHN, Gildenberg M, Ramamoorthy A, The magic of bicelles lights up membrane protein structure, Chemical Reviews. 112(11) (2012) 6054–6074. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [10].Richter RP, Bérat R, Brisson AR, Formation of solid-supported lipid bilayers: an integrated view, Langmuir 22(8) (2006) 3497–3505. [DOI] [PubMed] [Google Scholar]
  • [11].Castellana ET, Cremer PS, Solid supported lipid bilayers: from biophysical studies to sensor design, Surf Sci Rep 61(10) (2006) 429–444. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [12].McLean MA, Gregory MC, Sligar SG, Nanodiscs: a controlled bilayer surface for the study of membrane proteins, 47 (2018) 107–124. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [13].Schuler MA, Denisov IG, Sligar SG, Nanodiscs as a new tool to examine lipid-protein interactions, Methods Mol Biol. 974 (2013) 415–433. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [14].Glazier R, Salaita K, Supported lipid bilayer platforms to probe cell mechanobiology, Biochim Biophys Acta. 1859(9 Pt A) (2017) 1465–1482. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [15].Lind TK, Understanding the formation of supported lipid bilayers via vesicle fusion—A case that exemplifies the need for the complementary method approach, Biointerphases 11(2) (2016) 020801–12. [DOI] [PubMed] [Google Scholar]
  • [16].Skotland T, Sagini K, Sandvig K, Llorente A, An emerging focus on lipids in extracellular vesicles, Advanced Drug Delivery Reviews 159 (2020) 308–321. [DOI] [PubMed] [Google Scholar]
  • [17].Lv Z, Banerjee S, Zagorski K, Lyubchenko YL, Supported lipid bilayers for atomic force microscopy studies, Methods Mol Biol. 1814 (2018) 129–143. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [18].Joe S, Véronique V, Biomimetic models to investigate membrane biophysics affecting lipid–protein interaction, Frontiers in Bioengineering and Biotechnology, 8 (2020) 270. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [19].Deleu M, Crowet JM, Nasir MN, Lins L, Complementary biophysical tools to investigate lipid specificity in the interaction between bioactive molecules and the plasma membrane: A review, Biochimica et Biophysica Acta (BBA) – Biomembranes 1838(12) (2014) 3171–3190. [DOI] [PubMed] [Google Scholar]
  • [20].Lv Z, Hashemi M, Banerjee S, Zagorski K, Rochet JC, Lyubchenko YL, Assembly of α-synuclein aggregates on phospholipid bilayers, Biochim Biophys Acta Proteins Proteom 1867(9) (2019) 802–812. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [21].Banerjee S, Hashemi M, Zagorski K, L Lyubchenko Y, Interaction of Aβ42 with membranes triggers the self-assembly into oligomers, Int J Mol Sci 21(3) (2020) 1129. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [22].Banerjee S, Hashemi M, Zagorski K, L Lyubchenko Y, Cholesterol in membranes facilitates aggregation of amyloid β protein at physiologically relevant concentrations, ACS Chem Neurosci (2021) DOI: 10.1021/acschemneuro.0c00688 [DOI] [PubMed] [Google Scholar]
  • [23].Dokukin ME, Sokolov I, Quantitative mapping of the elastic modulus of soft materials with HarmoniX and PeakForce QNM AFM modes, Langmuir 28(46) (2012) 16060–71. [DOI] [PubMed] [Google Scholar]
  • [24].Hu J, Chen S, Huang D, Zhang Y, Lu S, Long M, Global mapping of live cell mechanical features using PeakForce QNM AFM, Biophysics Reports 6 (2020) 9–18. [Google Scholar]

RESOURCES