Skip to main content
NCBI home page
Chemical & Biomedical Imaging logoLink to Chemical & Biomedical Imaging
. 2025 Apr 9;3(9):605–614. doi: 10.1021/cbmi.5c00011

Quantifying the Shape and Stiffness of Single Extracellular Vesicles in Aqueous Solution via Membrane Diffusivity Measurements

Yihan Wang †, Huihui Gao †, Chu Han †, Liu Liu †, Jingwen Deng †, Hangwei Fan †, Zirui Zhou §, Mengyao Zhang §, Xiaohui Zhang ⊥, Feiyang Cheng †, Xiang Zhan ‡, Hao Ge ∥, Yan-Ling Liu †, Xinwei Zhang †, Wei-Hua Huang †, Wei Yan ⊥, Jing Zhang §, Wei Zhang #, Limin Xiang †,*
PMCID: PMC12458017  PMID: 41000200

Abstract

Quantifying the shape and stiffness of extracellular vesicles (EVs) is essential for understanding their biophysical properties and roles in intercellular communication. However, achieving single-particle resolution under physiological conditions remains a significant challenge. Here, we introduce an approach that integrates single-molecule diffusivity mapping (SMdM) with diffusion models for spherical and discoidal shapes to quantify the geometric and mechanical properties of individual liposomes and EVs in aqueous solution. Our findings identify charged lipids and cholesterol as critical factors that enhance liposome stiffness, driving their shapes closer to spheres. Applying this method to EVs reveals that those derived from tumor cells exhibit lower stiffness compared to EVs from normal cells, consistent with the biomechanical characteristics of their parent cells. This rapid, high-throughput strategy for characterizing the shape and stiffness of single EVs in aqueous solution offers promising applications in cancer biomarker discovery and the development of EV-based therapeutics.

Keywords: Diffusion, Membranes, Vesicles, Stiffness, Super-resolution Imaging

graphic file with name im5c00011_0007.jpg

graphic file with name im5c00011_0006.jpg

Introduction

Amphiphilic phospholipid molecules self-assemble into diverse shapes, including supported bilayers, liposomes, and discoidal micelles, through electrostatic and hydrophobic interactions. − This unique property underlies the formation of complex cellular membrane structures in biological systems, such as extracellular vesicles (EVs), which closely resemble liposomes in morphology. EVs are membrane-bound particles with diameters ranging from 30 nm to several microns, secreted by cells into the extracellular environment, where they act as critical mediators of intercellular communication. − Electron microscopy studies conducted in vacuum conditions have revealed a variety of EV shapes, including concave discs, spherical structures, and dumbbell-like forms. , However, under physiological conditions, EVs mediate communication within aqueous environments. The stiffness of their lipid bilayers is a key determinant of their shape, particularly when interacting with surfaces such as cellular plasma membranes. , Accurately quantifying the stiffness of EVs under physiological conditions is therefore a highly demanded but technically challenging task.

The stiffness of EVs has been reported to influence their uptake efficiency by recipient cells, thereby affecting cell–cell communication. , Recent studies have shown that EVs secreted by tumor cells are generally softer than those from normal cells, , suggesting that EVs with lower stiffness are more readily taken up by other cells. However, the subdiffraction size and heterogeneity of EVs pose significant challenges for single-particle analysis of their stiffness. Nanoindentation experiments using atomic force microscopy (AFM) have been successfully employed to study the shapes and mechanical properties of single EVs derived from various cell lines. , Nonetheless, these measurements suffer from low throughput, cannot be performed in aqueous solution, and may alter the elastic properties of the vesicle membrane due to contact with the probe. Optical methods such as Förster resonance energy transfer (FRET) offer a noninvasive approach to measuring EV curvature, but these techniques rely on ensemble measurements, which lack the resolution to capture size or diameter information on single EVs. Simultaneously quantifying the size and shape of single EVs in aqueous solution represents an ideal strategy for accurately assessing their stiffness.

Single-molecule diffusivity mapping (SMdM) is a super-resolution imaging technique designed to explore spatial heterogeneities in diffusivity and has been widely employed to study diffusion both within cells and on cellular membranes. − This technique involves tracking millions of single-molecule displacements within a specific region, allowing for the precise calculation of diffusion rates and achieving spatial resolution in diffusivity measurements. However, when applied to EVswhose subdiffraction sizes present significant challengesthe measured diffusion rates on their membranes are often underestimated. This phenomenon, known as the confined space effect in diffusivity measurements, , arises because SMdM tracks displacements in the 2D x-y plane, capturing only projections of the actual displacements along the curved membrane surface of EVs. As a result, the measured diffusion rates are systematically lower than their true values. Crucially, the magnitude of this confined space effect, and the degree of underestimation, depends on the geometrical shapes of the EVs. We propose that by comparing the measured diffusion rates with their true values, the geometric shape of EVs can be accurately inferred.

Here we established diffusion models for spherical and discoidal shapes, integrating these models with SMdM measurements to infer the geometric shapes of individual EVs adhered to glass substrates in aqueous solution, thereby quantifying their stiffness. We investigated the impact of lipid composition and found that increasing the content of cholesterol and charged lipids significantly modify the shapes and stiffness of liposomes. Finally, we examined the shape distributions of EVs derived from various cell types, revealing notable heterogeneities between those from tumor cells and normal cells.

Results and Discussion

We began by employing the single-molecule diffusivity mapping (SMdM) technique to measure the diffusion of the membrane probe BDP-TMR-alkyne on supported lipid bilayers (SLBs) composed of 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC). BDP-TMR-alkyne, as a membrane probe, exhibits an approximately 10-fold increase in fluorescence brightness in hydrophobic environments compared to aqueous ones, making it highly suitable for PAINT-type super-resolution imaging. Stroboscopic illumination was utilized to achieve an effective exposure time of 0.1 ms, with a frame-to-frame separation time of 1 ms, enabling the tracking of individual membrane probes across consecutive frames (Figure a). By constructing a displacement histogram for all probes and fitting it with a 2D Brownian motion model, we determined the diffusion rate of the membrane probes on SLBs (denoted as D true) to be 4.4 ± 0.3 μm2/s (Figure b).

1.

1

Open in a new tab

Measurement of diffusion rates on supported lipid bilayers (SLBs) and construction of spherical and discoidal models for confined spaces. (a) Schematic of supported lipid bilayers composed of DOPC, along with the chemical structure of DOPC. The diffusion of membrane probes on SLBs was tracked between every consecutive frames. Scale bar: 2 μm. (b) Displacement histogram for membrane probes on SLBs. Fitting the histogram with a 2D Brownian motion model yielded a diffusion rate of 4.4 μm2/s, which was used as the D true for modeling. (c) Diffusion on the surface of a spherical model. Measured displacements are projections of true displacements onto the x–y plane, resulting in underestimated diffusion rates. (d) Simulated displacement histogram for membrane probes on a sphere with a diameter of 200 nm, yielding a fitted diffusion rate of 2.0 μm2/s. (e) Fitted diffusion rates for spherical models as a function of increasing diameters. (f) Diffusion on the surface of a discoidal model. Measured displacements are calculated as the distance between starting and ending points, without accounting for encounters with boundaries, leading to underestimated diffusion rates. (g) Simulated displacement histogram for membrane probes on a disc with a diameter of 200 nm, yielding a fitted diffusion rate of 2.2 μm2/s. (h) Fitted diffusion rates for discoidal models as a function of increasing diameters.

While the 2D Brownian motion model accurately describes diffusion on SLBs, it fails to account for accurate diffusion measurements within confined 3D spaces, such as liposome membranes, which adopts various shapes. In the case of a spherical surface, displacements are projected onto the x–y plane, leading to an underestimation of the measured displacement (Figure c, where blue indicates true displacement and red represents measured displacement). For a discoidal surface, the primary underestimation arises from displacements encountering the boundary (Figure f, with blue indicating true displacement and green denoting measured displacement). To address these discrepancies, we randomly generated 104 displacements on the model surface by Monte Carlo simulation to investigate the relationship between the measured diffusion rates (D fit) and the diameter of spherical or discoidal models at a given D true (see Methods for details). The localization precision of 12 nm was determined experimentally by repeatedly measuring the position of a single fluorophore and calculating the standard deviation of its localization distribution. This value was then incorporated into the model simulation (Figure S1).

Consistent with our experiments, our simulations generated 104 displacements (d) with a time interval of Δt = 1 ms and D true = 4.4 μm2/s. The resulting displacement histograms for a sphere and a disc, each with a diameter of 200 nm, are shown in Figures d and g, respectively. By fitting these histograms with a 2D Brownian motion model, we obtained the underestimated fitted diffusion rates of D fit_s = 2.0 μm2/s for the sphere and D fit_d = 2.2 μm2/s for the disc. Figures e and h illustrate how D fit values for spheres and discs converge to D true as the diameter increases. Notably, the D fit values for the disc approach D true more rapidly compared to the sphere. These results suggest that the measured D fit can serve as a robust indicator for inferring the geometric shape of liposomes.

To validate our simulations, we employed polystyrene nanoparticles coated with liposomes as a model for spherical shape. , Due to the rigidity of the nanoparticles, their phospholipid-coated structures closely approximate perfect spheres (Figure a). In contrast, discoidal micellesformed when liposomes adsorb onto a glass surface and subsequently fuseserve as an ideal representation of a disc (Figure a). To quantitatively describe shapes that fall between these two extremes, we defined a shape coefficient, α, determined by linearly mixing the displacement histograms of the spherical and discoidal models to best match the experimentally measured displacement histograms (Figure b; see Methods for details).

2.

2

Open in a new tab

Validation of spherical and discoidal models using liposome-coated polystyrene nanoparticles and discoidal micelles. (a) Schematic illustration of lipid-coated nanoparticles and discoidal micelles. (b) Calculation of the shape coefficient (α) by linearly combining histograms of spherical and discoidal models to match experimentally measured histograms, quantifying the proximity to spherical shapes. (c) Single-particle SMdM image of lipid-coated nanoparticles, with each particle color-coded according to its measured diffusion rate. (d) Size-diffusion rate curve of lipid-coated nanoparticles aligns more closely with the spherical model (green), suggesting a spherical shape. (e) Three-dimensional super-resolution imaging validated that particles with a diameter of 200 nm conform to spherical shapes. (f) Analysis of the shape coefficient (α) for individual lipid-coated nanoparticles, overlaid with their diffusion rates. Results beyond the shaded region are considered reliable. (g) Single-particle SMdM image of discoidal micelles, color-coded by measured diffusion rates. (h) Size-diffusion rate curve of discoidal micelles aligns more closely with the discoidal model (red), indicating a discoidal shape. (i) Three-dimensional super-resolution imaging confirmed that discoidal micelles with a diameter of 200 nm conform to discoidal shapes. (j) Analysis of the shape coefficient (α) for individual discoidal micelles, overlaid with their diffusion rates. Scale bar: 600 nm in c and g, 200 nm in e and i.

We first performed SMdM on the liposome-coated polystyrene nanoparticles. Using DBSCAN, a density-based clustering algorithm, we simultaneously extracted diameter and diffusivity information by grouping closely connected single-molecule localizations into clusters, with the full-width at half-maximum (fwhm) of the distribution representing the particle diameter. The accuracy of diameter measurements obtained via DBSCAN analysis in single-molecule localization microscopy (SMLM) images was validated by comparing the results with those from Nanoparticle Tracking Analysis (NTA) (Figure S2). The single-particle SMdM image is shown in Figure c, with individual particles color-coded according to their measured diffusion rates. The size distribution of lipid-coated polystyrene nanoparticles was further validated using NTA (Figure S3), which yielded results consistent with those obtained from DBSCAN. To evaluated the goodness-of-fit between the measured data and various models, the R2 values were used as the statistical criterion. Analysis of individual particle diameters and measured diffusion rates (Figure d) demonstrated a stronger alignment with the spherical model (R2 = 0.833), whereas the discoidal model exhibited a comparatively lower goodness-of-fit (R2 = 0.772), confirming the spherical shape of these rigid nanoparticles.

To further substantiate this result, we performed three-dimensional super-resolution fluorescence imaging, which verified that particles with diameters around 200 nm exhibit a near-spherical shape (Figure e), consistent with the diffusion measurements. The measured diffusion rates and corresponding α values for individual particles with different diameters were plotted in Figure f, showing that particles with diameters exceeding ∼ 200 nm exhibit α values approaching 1. It is worth noting that as the curves in Figure d began to diverge at ∼ 170 nm (indicated by the dashed line), α values are meaningful only for particles larger than this diameter (beyond the shaded region in Figure f). In contrast, the single-particle SMdM images and diffusion rate versus diameter plots for discoidal micelles are shown in Figures g and h, respectively. The R2 statistical analysis revealed a better goodness-of-fit for the discoidal model (R2 = 0.912) compared to the spherical model (R2 = 0.786), demonstrating closer alignment of the experimental data with the discoidal model.

When combined with three-dimensional super-resolution fluorescence imaging (Figure i), the results confirm that their shapes conform to the discoidal model. Consequently, the calculated α values for individual discs were closer to 0 (Figure j), further corroborating their discoidal shape.

We next investigated the geometric shapes of single liposomes in aqueous solution adhered to glass surfaces. Previous studies have shown that liposome stiffness increases with the chain length and saturation level of their constituent lipids. Furthermore, factors such as charge and cholesterol content also play significant roles in influencing stiffness. As softer liposomes tend to flatten more on glass surfaces and resemble discoidal micelles, their geometric shapes were analyzed to probe their stiffness. Liposomes were prepared using the extrusion method (Figure a) and included DOPC (an unsaturated lipid), DOPC mixed with 10% 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC, a saturated lipid, Figure c), DOPC mixed with 10% 1,2-dioleoyl-sn-glycero-3-phospho-(1’-rac-glycerol) (DOPG, a negatively charged lipid, Figure d), and DOPC mixed with 10% cholesterol (Figure e). The prepared liposomes were deposited onto glass coverslips, and single-particle SMdM analysis was subsequently performed (Figure b). SMdM of supported bilayers (SLBs) with the same components were also performed, with the diffusion rates on SLBs being the D true for simulations in spherical and discoidal models (Figure S4).

3.

3

Open in a new tab

Charged lipids and cholesterol are dominant factors influencing the shape and stiffness of liposomes. (a) Preparation of liposomes using the extrusion method. (b) Liposomes composed of unsaturated lipid (DOPC), saturated lipid (DSPC), charged lipid (DOPG), and cholesterol were deposited onto glass substrates for single-particle SMdM analysis. (c)-(e) Chemical structures of DSPC, DOPG, and cholesterol, respectively. (f) Single-particle SMdM image of liposomes composed solely of DOPC. The size-diffusion rate curve lay between the spherical model (green) and discoidal model (red). Analysis of the shape coefficient (α) for individual liposomes, overlaid with their diffusion rates, indicates an intermediate shape between spheres and discs. (g) Single-particle SMdM image, size-diffusion rate curve, and shape coefficient (α) analysis for liposomes composed of DOPC/DSPC, showing shapes similar to those of DOPC-only liposomes. (h)-(i) Single-particle SMdM image, size-diffusion rate curve, and shape coefficient (α) analysis for liposomes composed of DOPC/DOPG (h) and DOPC/cholesterol (i). These liposomes exhibit shapes closer to spheres, indicating higher stiffness compared to DOPC-only liposomes. (j) Validation of the shapes of liposomes with diameters of ∼ 200 nm using three-dimensional super-resolution imaging. Scale bars: 600 nm in f, g, h, and i; 200 nm in j.

We first performed single-particle SMdM analysis on DOPC liposomes (Figure f). As expected, the measured diffusion rate increased with liposome size, consistent with the confined space effect (left panel of Figure f). The size-diffusion rate curve lay between the predicted curves for spherical and discoidal models (green and red curves in the middle panel of Figure f), with a shape coefficient (α) of 0.40 ± 0.19 (right panel of Figure f). This result indicates that DOPC liposomes adopt an ellipsoidal shape, intermediate between spherical and discoidal shapes. This shape is primarily attributed to the unsaturated chains in DOPC lipids and their phase transition temperature being far below room temperature, which results in lower stiffness and makes the liposomes more prone to deformation upon contact with the glass substrate.

For DOPC/DSPC liposomes, composed of saturated neutral lipids, single-particle SMdM analysis revealed results similar to those of DOPC liposomes (Figure g). The size-diffusion rate curve lay between the spherical and discoidal models (middle panel of Figure g), with a shape coefficient (α) of 0.44 ± 0.23 (right panel of Figure g), confirming that these liposomes also adopt an ellipsoidal shape. In contrast, for DOPC/DOPG liposomes, composed of negatively charged unsaturated lipids, the size-diffusion rate curve was closer to the spherical model (middle panel of Figure h), with a shape coefficient (α) of 0.59 ± 0.31 (right panel of Figure h). This suggests that charged liposomes are stiffer and less prone to deformation, resulting in a shape closer to a sphere on the glass substrate, likely due to the effects of electrostatic repulsion.

These findings suggest that charge plays a more significant role in determining liposome shape than the saturation level. To further validate this conclusion, we analyzed the shapes of liposomes composed of DOPC mixed with 10% of either a negatively charged saturated lipid (1,2-distearoyl-sn-glycero-3-phospho-(1́-rac-glycerol), DSPG) or a positively charged lipid (octadecylamine, ODA). The results demonstrated that the shapes of these charged liposomes consistently aligned more closely with spherical models (Figure S5), highlighting the critical role of charge as a key determinant of liposome shape and stiffness.

Cholesterol is known to promote the organization of lipid bilayers into more ordered and tightly packed phases. , Recent studies have shown that cholesterol-enriched exosomes are more efficiently taken up by tumor cells, underscoring its critical role in regulating cell–cell communication. The single-particle SMdM analysis of DOPC/cholesterol liposomes is presented in Figure i. Consistent with findings for charged liposomes, the addition of cholesterol led to more spherical liposome shapes, with a shape coefficient (α) of 0.64 ± 0.25 (right panel of Figure i). This effect is likely attributed to cholesterol enhancing lipid packing density, thereby increasing the stiffness of the lipid membrane. These observations for all the liposomes with different components were further validated through three-dimensional super-resolution fluorescence imaging, which confirmed the shapes of liposomes at a diameter of ∼ 200 nm (Figure j). The statistical distributions of α values for all liposome samples are summarized in Figure S6a. The observed discrepancies in α values within each sample likely arise from experimental errors and the inherent heterogeneity among individual liposomes. Despite these variations, statistical analysis successfully captured the differences in α values between different samples.

We employed the same approach to investigate the diffusion rates and geometric shapes of membrane vesicles from live cells. First, we analyzed the diffusion of BDP-TMR-alkyne molecules on the membranes of normal breast epithelial cells (MCF-10A, Figures a and b). The diffusion rate on the plasma membrane (white box in Figure a) was used as the D true. Dehydration-induced membrane vesicles (DIMVs) were prepared following established protocols, and their single-particle SMdM image is shown in Figure b.

4.

4

Open in a new tab

DIMVs derived from tumor cells exhibit lower stiffness compared to those from normal cells. (a) SMdM image showing diffusion on the plasma membrane of an MCF-10A cell (normal cell). Diffusion rates on plasma membrane (white box) was taken as the D true. (b) Single-particle SMdM image of dehydration-induced membrane vesicles (DIMVs) derived from MCF-10A cells. (c) Size-diffusion rate curve of individual MCF-10A-derived DIMVs, aligning more closely with the spherical model (green). (d) SMdM image showing diffusion on the plasma membrane (white box) of an MDA-MB-231 cell (tumor cell).(e)-(f) Single-particle SMdM image and size-diffusion rate curve of individual MDA-MB-231-derived DIMVs. The curve falls between the spherical model (green) and discoidal model (red), indicating an intermediate shape between sphere and disc. (g)-(h) Shape coefficients (α) of DIMVs derived from MCF-10A (g) and MDA-MB-231 (h) cells, overlaid with their diffusion rates. (i) Statistical analysis of the shape coefficients (α) for the DIMVs from MCF-10A and MDA-MB-231, suggesting that DIMVs from tumor cells exhibit lower stiffness and are more prone to deformation. Scale bars: 2 μm in a and d, 600 nm in b and e.

The experimental results revealed that the size-diffusion rate curve for MCF-10A-derived DIMVs aligned more closely with the spherical model (Figure c), with a shape coefficient (α) of 0.75 ± 0.25 (Figure g), indicating that DIMVs from MCF-10A are relatively stiffer. Similar results were obtained from experiments with another type of normal cell line (COS-7, Figure S7).

In comparison, single-particle SMdM analysis of DIMVs derived from the highly invasive breast cancer cell line MDA-MB-231 (Figure d) was presented in Figures e, f, and h. A comparison of the shape coefficients (α) for all DIMVs with diameters larger than 150 nm from these two cell lines was shown in Figure i. The results demonstrated that DIMVs derived from tumor cells with a shape coefficient (α) of 0.46 ± 0.21 exhibited a size-diffusion rate curve closer to the discoidal model compared to those from normal cells. This suggests that DIMVs from tumor cells are more easily deformed upon contact with the glass substrate, indicating lower stiffness. Previous studies have reported that EVs exhibit biomechanical properties similar to those of their parent cells, with vesicles secreted by highly invasive breast cancer cells displaying lower Young’s modulus compared to those from normal cells. , These findings align with our observations. Based on our results for liposomes with different components, the differences in DIMVs stiffness between tumor and normal cells may be attributed to variations in the content of charged lipids and/or cholesterol.

Additionally, we investigated the shape of naturally secreted extracellular vesicles (EVs). EVs from MCF-10A and MDA-MB-231 cells were isolated via ultracentrifugation and characterized according to MISEV 2023 guidelines (Figure a, d, and Figure S8). Single-particle SMdM analysis of naturally secreted EVs from MCF-10A and MDA-MB-231 cells was presented in Figures b and e, with their TEM images in Figure a and d, respectively. Most EVs exhibited significantly lower diffusion rates than those predicted by the spherical model when using plasma membrane diffusion rates as D true (Figure c and f). This discrepancy may stem from the predominance of protein-rich exosomes within the EV population, as exosomes typically contain abundant transmembrane proteins such as CD9, CD63, and TSG101. In contrast, DIMVs mainly consist of microvesicles that share similar lipid and protein compositions with the plasma membrane, thereby aligning more closely with our model predictions.

5.

5

Open in a new tab

Naturally secreted extracellular vesicles from MCF-10A and MDA-MB-231 cells exhibit significant heterogeneities in diffusion rates. (a) TEM image of MCF-10A-derived extracellular vesicles (EVs). (b) Single-particle SMdM image of naturally secreted extracellular vesicles from MCF-10A cells. (c) Size-diffusion rate curve of individual MCF-10A-derived EVs, with most diffusion rates falling below the predicted values of the spherical model (green line). (d) TEM image of MDA-MB-231-derived vesicles. (e) Single-particle SMdM image of naturally secreted extracellular vesicles from MDA-MB-231 cells. (f) Size-diffusion rate curve of individual MDA-MB-231-derived EVs, with most diffusion rates falling below the predicted values of the spherical model (green line). Scale bars: 600 nm in (b) and (e).

By integrating diffusion models for spherical and discoidal shapes with single-particle SMdM measurements, we quantified the shape and stiffness of individual liposomes with various components upon contact with a substrate in an aqueous solution. Softer vesicles are more prone to deformation upon glass contact, and our measurements of the shape coefficient α can serve as indirect indicators of membrane softness. Our findings revealed that the presence of charged lipids and cholesterol are key factors that increase liposome stiffness, resulting in a shape closer to a sphere upon contact with a glass substrate. We further applied this approach to DIMVs derived from tumor and normal cells, observing that DIMVs from tumor cells were more easily deformed upon contact with the glass substrate, indicative of lower stiffness. These observations were corroborated by three-dimensional super-resolution imaging. While three-dimensional super-resolution imaging provides accurate shape coefficients (α) for liposomes with diameters around 200 nm, liposomes with varying compositions consistently show a trend of decreasing α values as their size increases (Figure S6). This suggests that the limitations in axial resolution (for smaller diameters) and imaging depth (for larger diameters) in three-dimensional super-resolution imaging make it challenging to precisely quantify the shapes of individual liposomes across a broad size range. , Nevertheless, three-dimensional images remain a valuable reference for estimating the shapes of liposomes with diameters of approximately 200 nm.

A key prerequisite for using this approach to evaluate the shape and stiffness of liposomes and extracellular vesicles (EVs) was the availability of a true diffusion rate without confinement (D true). For liposomes, D true was obtained from diffusion measurements on supported lipid bilayers (SLBs), and for DIMVs, it was derived from diffusion on the plasma membrane. Interestingly, while charged lipids increase the stiffness of liposomes, they do not affect diffusion on SLBs (Figure S4). In contrast, cholesterol content significantly impedes diffusion on SLBs, consistent with previous findings (Figure S4). Since the diffusion rate on the plasma membranes of tumor cells is faster than that on the plasma membranes of normal cells (Figure a, d and Figure S7), we hypothesize that the differences in DIMVs stiffness primarily stem from variations in cholesterol content. Additionally, differences in membrane protein composition on DIMVs may also contribute to the observed changes in stiffness. , Notably, while spatial averaging of diffusion rates across the vesicle surface inherently overlooks regional heterogeneity, such variations are more likely attributed to differences in confined space effects across regions rather than intrinsic variations in lipid and protein composition (Figure S9).

In the diffusion models for spherical and discoidal shapes, the size-diffusion rate curves began to diverge at approximately 150 nm, limiting the accurate quantification of shapes to liposomes and EVs with diameters smaller than 150 nm. Employing membrane curvature-sensing probes in combination with super-resolution imaging techniques offers a promising alternative for quantifying the shape and stiffness of liposomes and EVs with diameters smaller than 150 nm. ,

Conclusions

In summary, leveraging single-particle SMdM analysis and theoretical modeling of diffusion in confined spaces on spherical and discoidal shapes, we developed an approach to quantify the shapes and stiffness of single liposomes and EVs under physiological conditions. Our findings identified charged lipids and cholesterol as key factors that increase liposome stiffness and revealed that DIMVs derived from tumor cells exhibit lower stiffness compared to those from normal cells. This approach enables the rapid characterization of the shapes and stiffness of hundreds to thousands of individual vesicles within minutes. Future studies evaluating EVs from various tumor cell lines to elucidate their stiffness differences and underlying molecular mechanisms represent a promising direction for further application.

Supplementary Material

im5c00011_si_001.pdf (959.3KB, pdf)

Acknowledgments

We acknowledge financial supports from National Key R&D Program of China (2022YFA1305400), National Natural Science Foundation of China (22274122, 22104113), Fundamental Research Funds for the Central Universities interdisciplinary (2042023kf1012), and Innovative Talents Foundation from Renmin Hospital of Wuhan University (JCRCFZ-2022-010).

Glossary

Abbreviations

EVs

extracellular vesicles

SMdM

single-molecule diffusivity mapping

SLBs

supported lipid bilayers

DIMVs

dehydration-induced extracellular vesicles.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/cbmi.5c00011.

  • Additional experimental details, materials, and methods, including SMdM images of diffusion on supported lipid bilayers (SLBs), plasma membrane and extracellular vesicles and the statistics of their shape parameters (PDF)

Yihan Wang: Conceptualization, Methodology, Software, Investigation, Formal Analysis, Writing-Original Draft; Huihui Gao: Investigation, Data Curation, Software; Chu Han: Formal Analysis, Resources; Liu Liu, Jingwen Deng, Hangwei Fan, Zirui Zhou, Mengyao Zhang, Xiaohui Zhang and Feiyang Cheng: Validation; Xiang Zhan, Hao Ge, Yan-Ling Liu, Xinwei Zhang, Wei-Hua Huang, Wei Yan, Jing Zhang and Wei Zhang: Supervision; Limin Xiang: Conceptualization, Funding Acquisition, Resources, Supervision, Writing-Review and Editing

The authors declare no competing financial interest.

References

  1. Bayburt T. H., Grinkova Y. V., Sligar S. G.. Self-Assembly of Discoidal Phospholipid Bilayer Nanoparticles with Membrane Scaffold Proteins. Nano Lett. 2002;2:853–856. doi: 10.1021/nl025623k. [DOI] [Google Scholar]
  2. Brea R. J., Cole C. M., Devaraj N. K.. In Situ Vesicle Formation by Native Chemical Ligation. Angew. Chem., Int. Ed. 2014;53:14102–14105. doi: 10.1002/anie.201408538. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Reviakine I., Brisson A.. Formation of Supported Phospholipid Bilayers from Unilamellar Vesicles Investigated by Atomic Force Microscopy. Langmuir. 2000;16:1806–1815. doi: 10.1021/la9903043. [DOI] [Google Scholar]
  4. Kalluri R., LeBleu V. S.. The Biology, Function, and Biomedical Applications of Exosomes. Science. 2020;367:eaau6977. doi: 10.1126/science.aau6977. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Raposo G., Stahl P. D.. Extracellular Vesicles: A New Communication Paradigm? Nat. Rev. Mol. Cell Biol. 2019;20:509–510. doi: 10.1038/s41580-019-0158-7. [DOI] [PubMed] [Google Scholar]
  6. Tkach M., Théry C.. Communication by Extracellular Vesicles: Where We Are and Where We Need to Go. Cell. 2016;164:1226–1232. doi: 10.1016/j.cell.2016.01.043. [DOI] [PubMed] [Google Scholar]
  7. Kapoor K. S., Kong S., Sugimoto H., Guo W., Boominathan V., Chen Y.-L., Biswal S. L., Terlier T., McAndrews K. M., Kalluri R.. Single Extracellular Vesicle Imaging and Computational Analysis Identifies Inherent Architectural Heterogeneity. ACS Nano. 2024;18:11717–11731. doi: 10.1021/acsnano.3c12556. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Kotrbová A., Štěpka K., Maška M., Pálenik J. J., Ilkovics L., Klemová D., Kravec M., Hubatka F., Dave Z., Hampl A., Bryja V., Matula P., Pospíchalová V.. TEM ExosomeAnalyzer: A Computer-Assisted Software Tool For Quantitative Evaluation of Extracellular Vesicles in Transmission Electron Microscopy Images. J. Extracell. Vesicles. 2019;8:1560808. doi: 10.1080/20013078.2018.1560808. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Feng Y., Liu M., Li X., Li M., Xing X., Liu L.. Nanomechanical Signatures of Extracellular Vesicles from Hematologic Cancer Patients Unraveled by Atomic Force Microscopy for Liquid Biopsy. Nano Lett. 2023;23:1591–1599. doi: 10.1021/acs.nanolett.3c00093. [DOI] [PubMed] [Google Scholar]
  10. Ge Z., Dai S., Yu H., Zhao J., Yang W., Tan W., Sun J., Gan Q., Liu L., Wang Z.. Nanomechanical Analysis of Living Small Extracellular Vesicles to Identify Gastric Cancer Cell Malignancy Based on a Biomimetic Peritoneum. ACS Nano. 2024;18:6130–6146. doi: 10.1021/acsnano.3c02285. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Sun J., Zhang L., Wang J., Feng Q., Liu D., Yin Q., Xu D., Wei Y., Ding B., Shi X., Jiang X.. Tunable Rigidity of (Polymeric Core)–(Lipid Shell) Nanoparticles for Regulated Cellular Uptake. Adv. Mater. 2015;27:1402–1407. doi: 10.1002/adma.201404788. [DOI] [PubMed] [Google Scholar]
  12. Hui Y., Yi X., Wibowo D., Yang G., Middelberg A. P. J., Gao H., Zhao C.-X.. Nanoparticle Elasticity Regulates Phagocytosis and Cancer Cell Uptake. Sci. Adv. 2020;6:eaaz4316. doi: 10.1126/sciadv.aaz4316. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. LeClaire M., Wohlschlegel J. A., Chang H., Wadehra M., Yu W., Rao J., Elashoff D., Gimzewski J. K., Sharma S.. Nanoscale Extracellular Vesicles Carry the Mechanobiology Signatures of Breast Cancer Cells. ACS Appl. Nano Mater. 2021;4:9876–9885. doi: 10.1021/acsanm.1c02299. [DOI] [Google Scholar]
  14. Whitehead B., Wu L., Hvam M. L., Aslan H., Dong M., Dyrskjøt L., Ostenfeld M. S., Moghimi S. M., Howard K. A.. Tumour Exosomes Display Differential Mechanical and Complement Activation Properties Dependent on Malignant State: Implications in Endothelial Leakiness. J. Extracell. Vesicles. 2015;4:29685. doi: 10.3402/jev.v4.29685. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Ridolfi A., Brucale M., Montis C., Caselli L., Paolini L., Borup A., Boysen A. T., Loria F., van Herwijnen M. J. C., Kleinjan M., Nejsum P., Zarovni N., Wauben M. H. M., Berti D., Bergese P., Valle F.. AFM-Based High-Throughput Nanomechanical Screening of Single Extracellular Vesicles. Anal. Chem. 2020;92:10274–10282. doi: 10.1021/acs.analchem.9b05716. [DOI] [PubMed] [Google Scholar]
  16. Ye S., Li W., Wang H., Zhu L., Wang C., Yang Y.. Quantitative Nanomechanical Analysis of Small Extracellular Vesicles for Tumor Malignancy Indication. Adv. Sci. 2021;8:2100825. doi: 10.1002/advs.202100825. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Büber E., Schröder T., Scheckenbach M., Dass M., Franquelim H. G., Tinnefeld P.. DNA Origami Curvature Sensors for Nanoparticle and Vesicle Size Determination with Single-Molecule FRET Readout. ACS Nano. 2023;17:3088–3097. doi: 10.1021/acsnano.2c11981. [DOI] [PubMed] [Google Scholar]
  18. Xiang L., Chen K., Yan R., Li W., Xu K.. Single-Molecule Displacement Mapping Unveils Nanoscale Heterogeneities in Intracellular Diffusivity. Nat. Methods. 2020;17:524–530. doi: 10.1038/s41592-020-0793-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Choi A. A., Xiang L., Li W., Xu K.. Single-Molecule Displacement Mapping Indicates Unhindered Intracellular Diffusion of Small (≲1 kDa) Solutes. J. Am. Chem. Soc. 2023;145:8510–8516. doi: 10.1021/jacs.3c00597. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Yan R., Chen K., Xu K.. Probing Nanoscale Diffusional Heterogeneities in Cellular Membranes through Multidimensional Single-Molecule and Super-Resolution Microscopy. J. Am. Chem. Soc. 2020;142:18866–18873. doi: 10.1021/jacs.0c08426. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Śmigiel W. M., Mantovanelli L., Linnik D. S., Punter M., Silberberg J., Xiang L., Xu K., Poolman B.. Protein Diffusion in Escherichia Coli Cytoplasm Scales with the Mass of the Complexes and is Location Dependent. Sci. Adv. 2022;8:eabo5387. doi: 10.1126/sciadv.abo5387. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Tran B. M., Punter C. M., Linnik D., Iyer A., Poolman B.. Single-protein Diffusion in the Periplasm of Escherichia coli. J. Mol. Biol. 2024;436:168420. doi: 10.1016/j.jmb.2023.168420. [DOI] [PubMed] [Google Scholar]
  23. Andrecka J., Spillane K. M., Ortega-Arroyo J., Kukura P.. Direct Observation and Control of Supported Lipid Bilayer Formation with Interferometric Scattering Microscopy. ACS Nano. 2013;7:10662–10670. doi: 10.1021/nn403367c. [DOI] [PubMed] [Google Scholar]
  24. Hughes L. D., Rawle R. J., Boxer S. G.. Choose Your Label Wisely: Water-Soluble Fluorophores Often Interact with Lipid Bilayers. PLoS One. 2014;9:e87649. doi: 10.1371/journal.pone.0087649. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Liu L., Pan D., Chen S., Martikainen M.-V., Kårlund A., Ke J., Pulkkinen H., Ruhanen H., Roponen M., Käkelä R., Xu W., Wang J., Lehto V.-P.. Systematic Design of Cell Membrane Coating to Improve Tumor Targeting of Nanoparticles. Nat. Commun. 2022;13:6181. doi: 10.1038/s41467-022-33889-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Liu L., Bai X., Martikainen M.-V., Kårlund A., Roponen M., Xu W., Hu G., Tasciotti E., Lehto V.-P.. Cell Membrane Coating Integrity Affects the Internalization Mechanism of Biomimetic Nanoparticles. Nat. Commun. 2021;12:5726. doi: 10.1038/s41467-021-26052-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Hahsler M., Piekenbrock M., Doran D.. dbscan: Fast Density-Based Clustering with R. J. Stat. Softw. 2019;91:1–30. doi: 10.18637/jss.v091.i01. [DOI] [Google Scholar]
  28. Lim H.-J., Kim G. W., Heo G. H., Jeong U., Kim M. J., Jeong D., Hyun Y., Kim D.. Nanoscale Single-Vesicle Analysis: High-Throughput Approaches Through AI-Enhanced Super-Resolution Image Analysis. Biosens. Bioelectron. 2024;263:116629. doi: 10.1016/j.bios.2024.116629. [DOI] [PubMed] [Google Scholar]
  29. Huang B., Wang W., Bates M., Zhuang X.. Three-Dimensional Super-Resolution Imaging by Stochastic Optical Reconstruction Microscopy. Science. 2008;319:810–813. doi: 10.1126/science.1153529. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Dai Z., Yu M., Yi X., Wu Z., Tian F., Miao Y., Song W., He S., Ahmad E., Guo S., Zhu C., Zhang X., Li Y., Shi X., Wang R., Gan Y.. Chain-Length- and Saturation-Tuned Mechanics of Fluid Nanovesicles Direct Tumor Delivery. ACS Nano. 2019;13:7676–7689. doi: 10.1021/acsnano.9b01181. [DOI] [PubMed] [Google Scholar]
  31. Zhang L., Granick S.. How to Stabilize Phospholipid Liposomes (Using Nanoparticles) Nano Lett. 2006;6:694–698. doi: 10.1021/nl052455y. [DOI] [PubMed] [Google Scholar]
  32. Dumitru A. C., Mohammed D., Maja M., Yang J., Verstraeten S., del Campo A., Mingeot-Leclercq M.-P., Tyteca D., Alsteens D.. Label-Free Imaging of Cholesterol Assemblies Reveals Hidden Nanomechanics of Breast Cancer Cells. Adv. Sci. Lett. 2020;7:2002643. doi: 10.1002/advs.202002643. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Richter R., Mukhopadhyay A., Brisson A.. Pathways of Lipid Vesicle Deposition on Solid Surfaces: A Combined QCM-D and AFM Study. Biophys. J. 2003;85:3035–3047. doi: 10.1016/S0006-3495(03)74722-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Yu M., Song W., Tian F., Dai Z., Zhu Q., Ahmad E., Guo S., Zhu C., Zhong H., Yuan Y., Zhang T., Yi X., Shi X., Gan Y., Gao H.. Temperature- and Rigidity-Mediated Rapid Transport of Lipid Nanovesicles in Hydrogels. Proc. Natl. Acad. Sci. U.S.A. 2019;116:5362–5369. doi: 10.1073/pnas.1818924116. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Liu S., Wen Y., Shan X., Ma X., Yang C., Cheng X., Zhao Y., Li J., Mi S., Huo H., Li W., Jiang Z., Li Y., Lin J., Miao L., Lu X.. Charge-Assisted Stabilization of Lipid Nanoparticles Enables Inhaled mRNA Delivery for Mucosal vaccination. Nat. Commun. 2024;15:9471. doi: 10.1038/s41467-024-53914-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Danylchuk D. I., Jouard P.-H., Klymchenko A. S.. Targeted Solvatochromic Fluorescent Probes for Imaging Lipid Order in Organelles under Oxidative and Mechanical Stress. J. Am. Chem. Soc. 2021;143:912–924. doi: 10.1021/jacs.0c10972. [DOI] [PubMed] [Google Scholar]
  37. Bagheri Y., Ali A. A., Keshri P., Chambers J., Gershenson A., You M.. Imaging Membrane Order and Dynamic Interactions in Living Cells with a DNA Zipper Probe. Angew. Chem., Int. Ed. 2022;61:e202112033. doi: 10.1002/anie.202112033. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Zhuo Y., Luo Z., Zhu Z., Wang J., Li X., Zhang Z., Guo C., Wang B., Nie D., Gan Y., Hu G., Yu M.. Direct Cytosolic Delivery of siRNA Via Cell Membrane Fusion Using Cholesterol-Enriched Exosomes. Nat. Nanotechnol. 2024;19:1858–1868. doi: 10.1038/s41565-024-01785-0. [DOI] [PubMed] [Google Scholar]
  39. Liu J., Shen T., Zhang Y., Wei X., Bao Y., Ai R., Gan S., Wang D., Lai X., Zhao L., Zhou W., Fang X.. Cell Dehydration Enables Massive Production of Engineered Membrane Vesicles with Therapeutic Functions. J. Extracell. Vesicles. 2024;13:e12483. doi: 10.1002/jev2.12483. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Royo F., Gil-Carton D., Gonzalez E., Mleczko J., Palomo L., Perez-Cormenzana M., Mayo R., Alonso C., Falcon-Perez J. M.. Differences in the Metabolite Composition and Mechanical Properties of Extracellular Vesicles Secreted by Hepatic Cellular Models. J. Extracell. Vesicles. 2019;8:1575678. doi: 10.1080/20013078.2019.1575678. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Massey A., Stewart J., Smith C., Parvini C., McCormick M., Do K., Cartagena-Rivera A. X.. Mechanical Properties of Human Tumour Tissues and Their Implications for Cancer Development. Nat. Rev. Phys. 2024;6:269–282. doi: 10.1038/s42254-024-00707-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Welsh J. A., Goberdhan D. C. I., O’Driscoll L., Buzas E. I., Blenkiron C., Bussolati B., Cai H., Di Vizio D., Driedonks T. A. P., Erdbrügger U., Falcon-Perez J. M., Fu Q.-L., Hill A. F., Lenassi M., Lim S. K., Mahoney M. G., Mohanty S., Möller A., Nieuwland R., Ochiya T., Sahoo S., Torrecilhas A. C., Zheng L., Zijlstra A., Abuelreich S., Bagabas R., Bergese P., Bridges E. M., Brucale M., Burger D., Carney R. P., Cocucci E., Crescitelli R., Hanser E., Harris A. L., Haughey N. J., Hendrix A., Ivanov A. R., Jovanovic-Talisman T., Kruh-Garcia N. A., Ku’ulei-Lyn Faustino V., Kyburz D., Lässer C., Lennon K. M., Lötvall J., Maddox A. L., Martens-Uzunova E. S., Mizenko R. R., Newman L. A., Ridolfi A., Rohde E., Rojalin T., Rowland A., Saftics A., Sandau U. S., Saugstad J. A., Shekari F., Swift S., Ter-Ovanesyan D., Tosar J. P., Useckaite Z., Valle F., Varga Z., van der Pol E., van Herwijnen M. J. C., Wauben M. H. M., Wehman A. M., Williams S., Zendrini A., Zimmerman A. J., Consortium M., Théry C., Witwer K. W.. Minimal Information for Studies of Extracellular Vesicles (MISEV2023): From Basic to Advanced Approaches. J. Extracell. Vesicles. 2024;13:e12404. doi: 10.1002/jev2.12404. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Aquino D., Schönle A., Geisler C., Middendorff C. v., Wurm C. A., Okamura Y., Lang T., Hell S. W., Egner A.. Two-Color Nanoscopy of Three-Dimensional Volumes by 4Pi Detection of Stochastically Switched Fluorophores. Nat. Methods. 2011;8:353–359. doi: 10.1038/nmeth.1583. [DOI] [PubMed] [Google Scholar]
  44. Vorselen D., van Dommelen S. M., Sorkin R., Piontek M. C., Schiller J., Döpp S. T., Kooijmans S. A. A., van Oirschot B. A., Versluijs B. A., Bierings M. B., van Wijk R., Schiffelers R. M., Wuite G. J. L., Roos W. H.. The Fluid Membrane Determines Mechanics of Erythrocyte Extracellular Vesicles and is Softened in Hereditary Spherocytosis. Nat. Commun. 2018;9:4960. doi: 10.1038/s41467-018-07445-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Ridolfi A., Conti L., Brucale M., Frigerio R., Cardellini J., Musicò A., Romano M., Zendrini A., Polito L., Bergamaschi G., Gori A., Montis C., Panella S., Barile L., Berti D., Radeghieri A., Bergese P., Cretich M., Valle F.. Particle Profiling of EV-Lipoprotein Mixtures by AFM Nanomechanical Imaging. J. Extracell. Vesicles. 2023;12:e12349. doi: 10.1002/jev2.12349. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Morton L. A., Yang H., Saludes J. P., Fiorini Z., Beninson L., Chapman E. R., Fleshner M., Xue D., Yin H.. MARCKS-ED Peptide as a Curvature and Lipid Sensor. ACS Chem. Biol. 2013;8:218–225. doi: 10.1021/cb300429e. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Snead D., Wragg R. T., Dittman J. S., Eliezer D.. Membrane Curvature Sensing by the C-Terminal Domain of Complexin. Nat. Commun. 2014;5:4955. doi: 10.1038/ncomms5955. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

im5c00011_si_001.pdf (959.3KB, pdf)

Articles from Chemical & Biomedical Imaging are provided here courtesy of American Chemical Society

RESOURCES