Mechanical properties of the high cholesterol-containing membrane: An AFM study

Abstract

Cholesterol (Chol) content in most cellular membranes does not exceed 50 mol%, only in the eye lens’s fiber cell plasma membrane, its content surpasses 50 mol%. At this high concentration, Chol induces the formation of pure cholesterol bilayer domains (CBDs), which coexist with the surrounding phospholipid-cholesterol domain (PCD). Here, we applied atomic force microscopy to study the mechanical properties of Chol/phosphatidylcholine membranes where the Chol content was increased from 0 to 75 mol%, relevant to eye lens membranes. The surface roughness of the membrane decreases with an increase of Chol content until it reaches 60 mol%, and roughness increases with a further increment in Chol content. We propose that the increased roughness at higher Chol content results from the formation of CBDs. Force spectroscopy on the membrane with Chol content of 50 mol% or lesser exhibited single breakthrough events, whereas two distinct puncture events were observed for membranes with the Chol content greater than 50 mol%. We propose that the first puncture force corresponds to the membranes containing coexisting PCD and CBDs. In contrast, the second puncture force corresponds to the “CBD water pocket” formed due to coexisting CBDs and PCD. Membrane area compressibility modulus (K_A) increases with an increase in Chol content until it reaches 60 mol%, and with further increment in Chol content, CBDs are formed, and K_A starts to decrease. Our results report the increase in membrane roughness and decrease K_A at very high Chol content (> 60 mol %) relevant to the eye lens membrane.

Graphical abstract

1. Introduction

Cellular membranes are composed of complex molecular composition mostly dominated by phospholipids (PLs) and cholesterol (Chol) besides proteins and carbohydrates [1, 2]. The PL composition of membranes of specific cell types or cell organelles differs vastly to adapt to the particular functionality [3, 4]. Such variation in functionality is promoted by the variation in PL types, including headgroup, hydrocarbon chain length, charge, and the degree of unsaturation of hydrocarbon [3, 5]. The plasma membrane’s component, Chol, plays a vital role in modulating eukaryotic cell’s membrane properties like fluidity, membrane thickness permeability, and rigidity [6]. Furthermore, Chol promotes several other-cellular activities like cell signaling [7, 8], condensing phospholipids [9, 10], exocytosis [11], and endocytosis [12], and trafficking membrane proteins [13]. Chol has a short flexible hydrocarbon chain, a rigid hydrophobic central sterol ring composed of four hydrocarbon rings with a small hydrophilic hydroxyl group attached. Since the PLs have a larger hydrophilic headgroup and long hydrocarbon chain, the cholesterol molecule fits between them. Moreover, the interaction between the phospholipids and cholesterol is mainly explained based on two models. The first model proposes that the acyl chain of PL and cholesterol are packed tightly due to the limited sharing space under phospholipid’s head group acting as an ‘umbrella’ and shields the Chol from exposure to the outer aqueous environment [14]. Second, the template model describes the interaction based on the cholesterol’s planar surface, complementing the phospholipid’s flexible acyl chain to produce tight packing and closed hydrophobic contact [15].

The mechanical property of the membrane depends upon the phase state, temperature, lipid headgroup, hydrocarbon chain as well as the Chol content in the membrane. Several methods are employed to monitor membrane’s mechanical property. Area compressibility and Young’s modulus of the 1-palmitoyl-2- oleoyl-sn-glycero-3-phosphatidylglycerol (POPG) vesicles were studied using dynamic light scattering (DLS) [16, 17]. Several lateral compressibility parameters [18] and bending rigidity [19] of the membrane was measured using X-ray. Cryo-electron microscopy was used to measure the membrane tension of egg phosphatidylcholine (PC) [20]. The most distinctive approach to decipher the membrane’s elastic property is by using micropipette aspiration of giant unilamellar vesicles (GUVs) pioneered by the Evans group [21–23]. Using such aspiration approach on giant plasma membrane vesicles (GPMVs) obtained from cholesterol depleted cells, lower bending rigidity was obtained, whereas larger rigidity was obtained on vesicles obtained from cholesterol-enriched cells [24]. Chol’s modulation in the membrane’s mechanical properties also depends on the hydrocarbon chain length and saturation [19, 25, 26] and the lipid headgroup [27]. Recently, AFM has also gained growing attention in deciphering the membrane mechanical property [28–30]. Using AFM, phase-separated domains were observed in membrane with low Chol, but such separation diminished, and only Chol enriched liquid-ordered phase (lo) was achieved in high Chol lipid membrane [31], although this transition in multicomponent mixture progresses via nanoscopic structures [32]. Also, the mechanical stability on the supported lipid bilayer by cholesterol is found to increase as seen by elevated breakthrough forces [33–35].

Since most of the cellular membrane in the human consists of cholesterol less than 50% [4], the membrane property studies are focused on this regime of the cholesterol content [27, 36]. Nevertheless, the eye lens membrane consists of significantly higher cholesterol than any other cellular membranes [37]. Depending upon the age group Chol/PL molar ratio ranges from 0.6 to 1.8 in the cortex and 0.7 to 4.4 in the nucleus of the human eye lens membrane [38–40]. Also, the Chol/PL ratio in the eye lens membrane varies among the species [41] and age [39]. When the membrane is saturated with cholesterol, phospholipid cholesterol domain (PCD) is formed within the membrane, and with a further increase of Chol content, cholesterol bilayer domains (CBDs) are formed, which coexist with surrounding PCD [38, 42, 43]. In the eye lens membrane, cholesterol is much higher than the Chol saturation limit such that it comprises not only PCD but also CBDs within the membranes [44–47]. The spin-labeling electron paramagnetic resonance (EPR) approach confirms the presence of coexisting CBD and PCDs in the total lipid extracts of animals and humans’ eye lens membranes [38, 40, 47–49] and on the model membrane containing a high Chol/PL mixing ratio [42–44, 46, 50]. We have shown that the physical properties of PCD do not change significantly with age [40, 49] as the CBDs serves as a buffer to maintain cholesterol concentration in the coexisting PCD [51].

Stiffness and other mechanical properties of an eye lens are investigated using ex vivo [52] and in vivo [53, 54] experiments in which the human lens tissues of central nuclear regions are found stiffer than the tissues at the boundary regions, and stiffness increases with aging [55, 56]. Similar results were obtained with the lens cells in the nuclear area and the cortex region (nuclear cells ~20 times stiffer than in the cortex region) [57]. These studies are focused on the cells and the tissues of the eye lenses; so, the investigation of eye lens membrane at the molecular level to study its mechanical property is warranted. Atomic force microscopy (AFM) is widely used to investigate the topography and the mechanical properties of the lipid system [28, 35, 58, 59]. Here, we use AFM to analyze the topography and mechanical properties of the 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphatidylcholine (POPC)-cholesterol membrane where we hypothesized the presence of coexisting PCD and CBDs. To avoid artifacts in investigations of membranes at high Chol content, the membrane suspensions were prepared using the rapid solvent exchange (RSE) method [60, 61], which preserves the samples’ compositional homogeneity at high cholesterol-containing lipid mixture. Since PCs are dominant in short life span animals’ eye lenses [62, 63], we use POPC as a model lipid for our studies. We have successfully employed AFM in a model system to access the supported lipid bilayer membrane topography and mechanical properties [64, 65], where we obtained the decline in area compressibility modulus (K_A) with the addition of peptidomimetic E107-3 on POPC with a 10% POPG bilayer.

Being a label-free approach and the ability to scan in physiological conditions, AFM helps provide the missing information about high cholesterol-containing eye lens-alike-membrane’s mechanical properties. The mechanical property of high cholesterol membrane performed in this work will help us understand the role of high Chol in lens membrane elasticity. Furthermore, this study can be extended to understand cholesterol role in presbyopia’s fundamental cause, a loss of ability to accommodate images in the lens for most humans after the age of ’40s, as this condition is implicated in the biomechanical and biophysical characteristics of the crystalline eye lens [53, 66–68]. Such understanding provides the fundamental basis for future advances in curing accommodation related to eye lens dysfunction. Apart from that, the high cholesterol-containing membrane studies can aid in elucidating pathological conditions like cataract [69], atherosclerosis [70], and gallstone formation [71].

2. Methods and Material

2.1. Materials

1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) and cholesterol (Chol) were purchased from Avanti Polar Lipids (Alabaster, AL). All other reagents (reagent grade) were purchased from Sigma Aldrich (St. Louis, MO). Buffer A was prepared with 10 mM HEPES, 150 mM NaCl, and 5 mM of CaCl₂ at pH 7.4, whereas buffer B was prepared with 10 mM HEPES and 150 mM NaCl at pH 7.4. Water used for the buffers’ preparation was ultrapure (Resistivity=18.2 MΩ.cm) obtained from Thermo Scientific Easy pure Rodi purification system (Marietta, OH).

2.2. Lipid vesicles preparation

Large multilamellar vesicles (LMVs) were prepared by the standard rapid solvent exchange method [46, 60, 61]. The small unilamellar vesicles (SUVs) were prepared by sonication of LMV suspensions made of POPC and Chol with varying Chol concentration. Chloroform solutions of POPC and Chol were mixed to obtain respective Chol/POPC mixing ratio ranging from 0 to 3. The POPC concentration was maintained at 2.5 mg/mL, and Chol concentration was varied to obtain the desired mixing ratio. 0.5 mL of HEPES buffer A was added to the test tube containing ~75 μL lipid solution in chloroform to prepare LMVs by the rapid solvent exchange method [46, 60, 61]. A probe-tip sonicator (Fisher Scientific, model FB705) was employed to disperse the LMVs. Five to seven 10s sonication cycles followed by 15s cooling in an ice bath was adequate to transfer the milky suspension of LMVs into a slightly hazy transparent solution. A similar preparation method was used previously to prepare POPC SUVs for EPR experiments [72]. The obtained SUV solution was further centrifuged at 5000 rpm using Eppendorf centrifuge 5424R (Eppendorf AG, Hamburg, Germany) for 5 minutes to remove any possible membrane aggregates and metal particles. The obtained supernatant consisting of SUVs was used to prepare supported membranes for AFM experiments.

Rapid solvent exchange method (RSE) for the preparation of the vesicles is significant when preparing the vesicles with high Chol content. Previously, Huang et al. [60] used RSE method and demonstrated that the vesicles prepared with RSE method can accommodate the maximum (full) amount of Chol in the membrane, and has a true equilibrium Chol solubility limit. They used the X-ray diffraction and confirmed that both monohydrate and anhydrous Chol crystals are formed when the sample is prepared with conventional preparation methods (film deposition and lyophilization), giving falsely low estimates of the Chol solubility threshold, whereas when the sample is prepared with RSE method, only monohydrate Chol crystals are formed [60]. The anhydrous Chol crystals are formed in the conventional preparation methods due to the de-mixing of Chol when the lipid mixture passes through the solid-state intermediate [60]. Chol trapped in these anhydrous cholesterol crystals does not participate in the further liposome formation giving falsely low estimates of the Chol solubility threshold [60]. By using differential scanning calorimetry, we have confirmed that only Chol monohydrate crystals are formed when the vesicles are prepared by the RSE method [46, 50].

2.3. Supported lipid membrane (SLM) preparation

Supported lipid membranes were prepared by fusion of small unilamellar vesicles (SUVs) on top of flat mica surface affixed to thin metallic disk. A smooth surface was identified on a freshly cleaved mica disk placed on top of the piezo stack and was covered with a fluid cell affixed to the AFM head assembly. The SUV supernatant was diluted to 0.2-0.5 mg/mL, and around 75 μL of the solution was transferred to this flat region of the mica disk from one of the inlets in the fluid cell via silicone tubing and syringe pump system. For the preparation of membrane patches, SUV supernatant was diluted to 15 μg/mL, The lower concentration of SUVs was used for imaging of membrane patches because higher concentration results in faster rupture of vesicles, forming the complete supported membrane quickly causing difficulty in capturing the initial stages of formation of membrane patches. Previously, Zhengjian et al. used lower concentration of 5 μg/mL to image supported membrane patches for POPC membrane and discussed about the difficulty in capturing initial stage of membrane formation due to faster rupture of vesicles at higher concentration [73]. While incubating, the AFM tip remained ~ 50 μm above the mica surface. Since the phase transition temperature of POPC (−2°C) is well below the room temperature, we incubate the sample at room temperature. During the incubation for about 20 minutes, the SUV’s fused and ruptured on the mica surface forming lipid patches. The coalition of such patches formed a supported lipid membrane (SLM). The unfused SUVs were flushed out through the other fluid cell outlet with ~ 5 mL of buffer B. The supported membranes were prepared within 2 hours of the SUVs preparation, and the AFM measurements were performed immediately after the membranes are formed and excess SUVs flushed out. In the case of bilayer patches, the imaging was performed in about 15 minutes of incubation without flushing out the vesicles.

2.4. AFM image and force spectroscopy

We used Bruker multimode VIII AFM with scanner-E and a nanoscope V controller (Santa Barbara, CA) capable of scanning in fluid for imaging and force spectroscopy. For imaging and force spectroscopy, we use a DNP-S10 cantilever with a nominal spring constant of 0.35 N/m. The tip end radius, measured by scanning the relatively rough titanium substrate and using tip quantification function in nanoscope analysis, was obtained to be ~32 nm. To exclude the discrepancy in the results arising from the tip radius, we use the same AFM tip for all the images and force curve experiments. Imaging and force spectroscopic measurement procedures are described elsewhere [32, 59, 64, 65]. All the images were captured at a rate of ~1 Hz with pixel size 256×256. For the direct comparison of the features of the membranes, we used the same cantilever and maintained the same imaging parameters for all the experiments. The maximum force, i.e., peak force applied on the membrane, was set to 600 pN. After the acquisition, all images were further flattened using nanoscope analysis software. All the measurements were performed at room temperature. The roughness and the height of the membrane measurements are from the images taken from three independent experiments (two different images each from three different bilayers) for each Chol/POPC samples.

To obtain the force curves, we selected a region of 5 μm×5 μm within the image and specified 10×10 points for force ramping. Force curves were obtained with the approaching and retarding speed of tip at 200 nm/s. 2560 data points were taken for each force curve. We measured the deflection sensitivity and spring constant after each sample change to compensate for the laser spot’s drift reflecting from the tip’s back. The average breakthrough forces, rupture depth, membrane thickness, and the compressibility modulus K_A were estimated from three independent experiments with ~100 force curves in each experiment.

2.5. Statistics

The Studen’s t-test was used to determine the statistical significance between the root mean squared surface roughness (Rq) of membranes at different Chol/POPC mixing ratios. For Rq, we used the Rq values obtained from three independent experiments. The statistical significance was determined by comparing the Rq values among each other for different Chol/POPC membranes. A value of p ≤ 0.05 was considered statistically significant.

3. Results and Discussion

3.1. Topographical images and surface roughness

We prepared the mica supported planar lipid membranes with a representative Chol/POPC mixing ratio between 0 to 3 within the AFM fluid cell. The AFM tip was stationed well above the mica surface to minimize interference during lipid membrane formation. After the removal of excess SUVs, the tip was engaged to acquire the membrane’s topographical images. The representative images of the POPC lipid membrane with different cholesterol content are shown in Fig. 1. We confirmed the uniform membranes with non-detectable height difference by scanning them at various locations. The height profile of the lines across the image is shown in Fig 1.

Figure 1:

Representative height images of the cholesterol-containing phospholipid membrane. Image with Chol/POPC mixing ratio of 0.0, 1.0, 1.5, 2.5, and 3.0 are shown from left to right. The height profiles of the lines on the respective images are shown under the image. The roughness of the membrane decreases initially until the Chol/POPC mixing ratio is ≤ 1.5. However, above this limit, the roughness increases and is similar to the POPC membrane without Chol. Scale bar = 1 μm for all images.

The effect in membrane topography by increasing cholesterol content in the POPC membrane is shown in Fig. 1A to 1E. The root mean squared surface roughness (Rq), the vertical descriptor of surface architecture, for membranes at different Chol/POPC mixing ratios is shown in Table 1. Table 1 shows the average roughness measurement (Rq) with the errors. For a direct comparison of the roughness parameter, we used the same cantilever and imaging parameters for all the images taken. As seen from Table 1, the Rq of the membrane decreases from 134±8 pm to 99±13 pm when the Chol/POPC mixing ratio increases from 0 to 1.5. Statistically significant differences were seen with p ≤ 0.05 when the Rq for Chol/POPC mixing ratio of 0 is compared with Rq of Chol/POPC mixing ratio 1 and 1.5; however, no significant difference with p ≤ 0.05 was observed between Rq of Chol/POPC mixing ratio of 1 and 1.5. However, above Chol/POPC mixing of 1.5, the estimated Rq increases to 138±20 pm, which is similar to Rq of the cholesterol-free membrane (134±8 pm). Statistically significant differences were seen with p ≤ 0.05 when the Rq at Chol/POPC mixing ratio of 1.5 is compared with Rq at Chol/POPC mixing ratio of 2.5 and 3; however, no significant difference with p ≤ 0.05 was observed between Rq of Chol/POPC mixing ratio of 2.5 and 3.

Table 1:

The root mean squared surface roughness (Rq) of the membranes at different Chol/POPC mixing ratios. The values are shown as mean ± SD.

Chol/POPC mixing ratio	0	1	1.5	2.5	3
Rq (pm)	134±8	101±4	99±13	136±11	138±20

One of the main conclusions which can be made from topographical images presented in Fig. 1 is the absence of a clear indication of the existence of CBD domains. Membranes are uniform and relatively smooth at all investigated Chol contents. Because existence of CBDs at Chol content greater than Chol/POPC 1/1 was clearly shown by EPR spin labeling [42, 50] and confocal microscopy methods [74], it suggests that AFM in the present set-up cannot discriminate these domains. The thickness of CBD (pure Chol bilayer) is 34 Å [75–78], and the thickness of the surrounding PCD (POPC bilayer saturated with Chol at Chol/POPC molar ratio of 1/1) is ~ 45 Å [79]; these domains should be discriminated by AFM at Chol content greater than 1/1. Coexistence of these two domains is suggested by double punching observed for membranes at high Chol content (see Fig. 2). So, the logical explanation is that CBDs in these SLMs are small, smaller than the surface of the discriminating AFM tip (~3000 nm²). We discussed the size of CBDs in our previous paper [38] with the final conclusion that the size should be small. However, if some forces exist which induce coalescence of small CBDs, the one large domain can be observed. We showed with fluorescent probes and confocal microscopy [74] that this can happen in membranes of GUVs, and the force which induce coalescence of small CBDs is gravitation force. Thus, we can conclude that CBDs formed in supported POPC membranes are small, with a surface smaller than 3000 nm².

Figure 2:

Force–separation curves on the supported lipid membrane exhibiting single and double puncture events. Single puncture events are obtained in bilayer with low Chol/POPC mixing ratio, whereas bilayers containing Chol/POPC mixing ratio > 1.0 exhibit two puncture events. The puncture event related to lower breakthrough force corresponds to coexisting PCD and CBD and the larger puncture force corresponds to “CBD water pocket”. Force curves shown here are captured in a single supported lipid bilayer for all Chol content.

Comparison of our previous results with those presented here should be done with immense care because different samples were used in these experimental studies. Earlier, we used liposomes (large multilamellar vesicles (LMVs)) prepared by using the rapid solvent exchange method [38, 46, 50]. Our EPR spin-labeling approach shows the formation of coexisting PCDs and CBDs at high Chol content for liposome samples [38, 42, 43, 46, 50]. In the presented research, we use supported lipid bilayers. Previously, Ziblat et al. [80] used grazing incidence X-ray diffraction in polymer supported lipid bilayer made of Chol and PLs and confirmed the presence of crystalline Chol bilayer. Their findings of Chol as single bilayers agrees with our suggested CBD formation in supported Chol/POPC membrane in the study investigated here.

Results with the SLM roughness agree with our previous results and comments above. EPR spin-labeling methods enabled us to obtain the molecular-level information on the organization and dynamics of lipid molecules in the PCD and how these properties change as a function of the membrane depth [38, 41, 47, 49]. Those studies showed that in Chol saturated POPC bilayers, both oxygen transport parameter and hydrophobicity profiles across the bilayer had rectangular shape [41]. In liquid-ordered POPC-Chol bilayers, the rectangular shape of the oxygen transport parameter profile, with an abrupt increase close to the positions of C9 in acyl chains, is observed only for bilayers saturated with Chol (i.e., containing 50 mol% Chol) [41]. This contrasts liquid-ordered PC-Chol bilayers of the lowest Chol content (~30 mol%), for which the profile is bell-shaped and similar to that for bilayers without Chol [38, 44, 81]. Therefore, it was concluded that saturation with Chol decreases vertical fluctuations of membrane lipids (phospholipids and cholesterol). In effect, the phospholipids were vertically aligned, and all Chol rings were immersed to the same membrane depth, which is close to the positions of C9 in PC acyl chains. As a result, the membrane surface was smoother compared to membranes without or with lower Chol content. These experimental results were additionally confirmed by MD simulations carried out also for Chol/POPC membranes [79]. At greater contents of Chol, CBDs are formed within these smooth PCD. These domains cannot be discriminated by the AFM; however, their presence increases the roughness of the membrane surface, which is seen in Fig. 1.

3.2. Membrane force measurement

To characterize the POPC-Chol membrane’s mechanical properties, we performed AFM force spectroscopy experiments on the SLM prepared from different Chol/POPC mixing ratios. Fig. 2 shows the collection of force curves obtained from one set of the SLM with Chol/POPC mixing ratios from 0 to 3. During force spectroscopy, there is no interaction when the tip is far from the membrane surface. As the tip approaches the membrane, some non-steric interaction like hydration and van der Waals slightly deviates the force from zero [14]. After the cantilever tip touches the membrane’s surface, the loading force increases, and it deforms the membrane elastically. The tip depresses the membrane with an incremental load until it fails to withstand the applied force. This maximum force that the membrane can stand intact or after which membrane punctures is called the breakthrough force (FB) and is defined by the discontinuity or sudden jump in the approach section of the force-distance curve (see Fig 2). Breakthrough force is considered a direct measurement of the lipid molecule’s lateral interaction [82].

After the tip punctures, the tip passes through the membrane’s hydrocarbon region, crossing the water layer lying underneath the membrane before reaching the mica substrate. This water layer existing between the substrate (mica) and the membrane extends roughly 1 to 2 nm [33, 83, 84] and provides lubrication and maintain mobility for the lipid molecules [84]. For the membranes consisting of the Chol/POPC mixing ratio 0 and 1, we obtained a single puncture event (see Fig. 2). We found the average breakthrough force for the membrane containing 0 mol% Chol and 50 mol% Chol: 5.26±0.28 nN and 9.25±0.57 nN, respectively. The higher puncture force in 50 mol% Chol (i.e., for the membrane with only PCD) represents a more ordered and smoother membrane with the decrease in lipid’s vertical fluctuations (see also Sect. 3.1). For a similar experiment with different ionic concentrations, a smaller breakthrough force was obtained [33]. The bilayer membrane system’s breakthrough force alters significantly depending upon the ionic concentration of the imaging buffer [85]. In addition, the breakthrough force depends on lipid packing, acyl chain length, and degree of unsaturation [58, 86]. Also, breakthrough force is affected by experimental parameters such as cantilever, tip radius, and approaching speed [58, 87, 88].

We obtained two puncture events for the membrane with Chol/POPC mixing ratio > 1.0, as shown in the second row of Fig. 2. The first puncture force (FB₁) in this membrane corresponds to the coexisting PCD and CBDs, which is smaller than the membrane’s puncture force containing only PCD. In contrast, we speculate that the second puncture force (FB₂) corresponds to the “CBD water pocket” formed due to the coexisting CBDs and PCD on the water layer above the mica surface. Since the thickness of CBD is 34 Å [75–78], and the thickness of the surrounding PCD is ~ 45 Å [79], we speculate that this difference in the thickness between CBD and PCD give rise to the formation of “CBD water pocket”. For the membrane with coexisting CBDs and PCD, the CBDs are organized across the transmembrane, exhibiting lower height than the PCD (~0.6 nm on both surfaces). The distal region’s wells are filled with buffer which will be the part of bulk buffer solution above the membrane. However, in the proximal region, such a gap introduces additional buffer thickness, which we term as “CBD water pocket”, in addition to the buffer layer already present with the usual PCD membrane. “CBD water pocket” acts as an additional resistance, apart from the usual leaflet headgroup membrane tension, and hence acts as a barrier for AFM tip to penetrate in a single step. After sufficient AFM tip thrust on the region, it penetrates the layer, and hence the second puncture event evolves.

Previously, we have used EPR spin-labeling method and measured the polarity (water accessibility) for Chol/POPC membrane where CBD coexists with surrounding PCD at higher Chol content (Chol/POPC >1) [42]. Our EPR measurement shows that the membrane surface of CBD is significantly less hydrophobic (i.e., more polar) in comparison to the surrounding PCD, whereas the center of CBD is as hydrophobic as the center of PCD [42]. Significantly lower hydrophobicity (higher polarity) on the CBD surface and higher hydrophobicity (less polarity) on the CBD center gives rise to a second puncture event. Thus, the polarity (water accessibility) in the “CBD water pocket” is significantly greater than the surrounding PCD region that results in a second puncture force, which is about 2 to 2.5 times larger than the first puncture (see Figs 2 and 4B). The probability distribution of the breakthrough forces in different Chol/POPC mixing ratio is shown in Fig 3. The solid black line represents the Gaussian fit for the distribution. The probability distribution of the breakthrough force becomes broader when the Chol/POPC mixing ratio increases from 0 to 1; however, no distinct trend is observed in the distribution with Chol/POPC mixing ratio > 1 (green bar in Fig 3). For the Chol/POPC mixing ratio > 1, the green bar represents the distribution of the breakthrough force for coexisting CBDs and PCD, whereas the gray bar represents the breakthrough force distribution of “CBD water pocket”. When the Chol content increase from 1.5 to 2.5, the distribution becomes narrower, whereas the distribution becomes broader again when the Chol content increases from 2.5 to 3 (green bar in Fig. 3). The “CBD water pocket” distribution is similar but shifted slightly toward the higher puncture force (gray bar in Fig. 3 and Table 1). The rupture depth arising due to “CBD water pocket” increase with Chol content, which will also be discussed below in Sect 3.3. Based on these observation’s, we speculate that the size or amount of CBD that coexist with PCD in SLM increase with an increase in Chol content; however, the distribution of the size of CBDs are not uniform that is reflected in the probability distribution function (green and gray bar in Fig. 3 after Chol/POPC mixing ratio > 1). Nevertheless, CBDs’ size is smaller than the size of AFM tip, as discussed in Sect. 3.1.

Figure 4:

(A) Representative single force curves of the membrane with Chol/POPC mixing ratio from 0 to 3 (B) The average breakthrough forces of the membrane with Chol/POPC mixing ratio from 0 to 3. The circles indicate the breakthrough forces for POPC without Chol (Chol/POPC mixing ratio of 0) along with PCD (Chol/POPC mixing ratio of 1) and PCD coexisting with CBDs (Chol/POPC mixing ratio of 1.5, 2.5, and 3). In contrast, the squares indicate the breakthrough force for the “CBD water pocket”.

Figure 3:

Distribution of breakthrough forces in the bilayers of the lipid composition as indicated. The bars are the experimental data (events) and the solid lines are the gaussian fit for corresponding distribution. The green bar represents the distribution of the breakthrough force of membranes with a Chol/POPC mixing ratio of 0 to 3. For Chol/POPC ratio > 1.0, first distribution of the breakthrough force corresponds to coexisting PCD and CBD (green bar) and the second distribution of breakthrough force correspond to “CBD water pocket” (grey bar).

Fig. 4A represents the Chol/POPC membrane’s representative force curves, and Fig. 4B shows the average breakthrough forces. The average breakthrough force (FB₁ and FB₂) of the membrane system is shown in Table 1. Several previous studies have suggested two puncture events in the lipid membrane. Some studies have attributed to double bilayer [64] or the formation of additional bilayer on the tip, either by large tip size [89], coated tip [90], or high ionic concentration in buffer [91]. Similarly, two puncture events on the bilayers are reported in some studies indicating the leaflets’ uncoupling by factors like scanning speed of AFM tip, membrane preparation, and force curve acquisition temperature [58]. We used the faster scanning speed (1 μm/sec) and slower speed (0.2 μm/sec, data presented here) for acquiring data but did not obtain significantly different force curves. Also, the scanning temperature (room temperature ~25°C) is well above the phase transition temperature for POPC to exhibit such a nature. Uncoupling of leaflets, and hence double penetration by AFM tip, below the main phase transition temperature, has been shown even for single-component lipid membranes like POPE [58] and SM [92]. Two-step breakthrough events were also obtained in a binary or tertiary mixture of lipids DPPC and pSM with Cholesterol (≤30%) and Ceramide [92, 93] A recent study by Adhyapak et al. found the uncoupling of DOPC/Chol membrane leaflets even at 20% cholesterol at room temperature [94]. The two puncture events in their experiment are attributed to the higher mechanical stability due to the Chol and decoupling of the membrane leaflets due to asymmetrical cholesterol distribution. The vesicles used for the bilayer preparation for our experiments are prepared from a different approach (RSE method) than the traditional film deposition, which at higher cholesterol-containing membrane introduces artifactual de-mixing of cholesterol during vesicle preparation [60], and disrupts the supported membrane stability [94]. The RSE method of vesicle preparation ensures the compositional homogeneity in high cholesterol-containing membranes [46, 60, 61]. However, it has been shown that the POPC lipid bilayer exhibits a single puncture event for cholesterol level up to 50%, which is the Chol saturation limit for the POPC membrane, using the film deposition technique [33]. Surprisingly, to the best of our knowledge, no AFM study has been reported for high cholesterol-containing POPC membrane (i.e., for Chol/POPC mixing ratio >1). Our results report a single puncture event for Chol/POPC mixing ratio 0 and 1, whereas, for higher Chol content (i.e., above the Chol saturation limit of POPC membrane; Chol/POPC mixing ratio >1) two breakthrough force arise likely due to the formation of coexisting CBDs and PCD and the water pocket introduced by membrane thickness difference as discussed above. Above Chol’s saturation, CBDs are present in both unilamellar and multilamellar vesicles, verified by confocal microscopy [74] and EPR [38, 42, 43, 46, 50]. Ziblat et al [80] investigated polymer supported lipid bilayer using grazing incidence X-ray diffraction where they report the formation of Chol as single bilayers. Our recent data with the EPR spin-labeling method confirmed CBDs’ presence in Chol/POPC vesicles for Chol/POPC mixing ratio greater than 1 [50]. Both in the EPR spin-labeling method in previous studies [50] and AFM study presented here, the LMVs are prepared with the RSE method. In the EPR measurements, to increase the signal-to-noise ratio, LMVs were centrifuged for 15 min at 12000 g, and the loose pellets were filled into a gas-permeable capillary to perform EPR measurements. Also, LMVs for EPR measurements contain 1 mol% of cholesterol analog spin label [50]. However, in the AFM study presented here, LMVs were sonicated to produce SUVs to prepare supported membranes for AFM experiments (see Sect. 2.2) [50]. Our AFM force measurement (this Sect.) and AFM image roughness (Sect. 3.1) results suggest CBD’s existence in SLM, which complements the results obtained with EPR [38, 42, 43, 46, 50], confocal microscopy [74], and grazing incidence X-ray diffraction [80].

For all the Chol/POPC mixing ratios greater than 1, we observed two puncture events. Interestingly, no single force curve with one puncture event was observed in any of the ~100 force curves that we captured for each sample above the Chol/POPC mixing ratio 1. This data suggests that the proposed CBDs co-existing with the surrounding PCD are uniformly distributed in the membrane for all the samples above Chol/POPC mixing ratio greater than 1. If the CBDs were not uniformly distributed, we would have obtained a mixture of single puncture and double puncture events containing force curves; however, we observed two puncture events for all the ~100 force curves. Thus, the first puncture force (FB₁) in this membrane corresponds to the co-existing PCD and CBD where, CBDs are organized across the transmembrane. As discussed above, we hypothesized that the second puncture event (FB₂) occurs in the water layer above the mica surface due to the presence of a “CBD water pocket”, which is formed due to co-existing CBDs and PCD. CBDs’ surface has significantly higher polarity than the surrounding PCD [42], resulting in significantly higher water accessibility in the “CBD water pocket”. Once the AFM tip penetrates the co-existing CBD and PCD region and reaches the “CBD water pocket”, it experiences significantly higher resistance in the vicinity of “CBD water pocket”. The AFM tip experiences this high resistance because the surface of CBD is highly polar, whereas the center is as hydrophobic as the center of PCD, so water molecules are not favorable in moving towards the center of CBD/PCD, causing the significant resistance for the AFM tip as it approaches the mica surface. This results in a very high second rupture force (see Table 3). The second rupture depth data suggest that the AFM tip travels with resistance in the vicinity of “CBD water pocket” before the second puncture event occurs (see Fig. 2 and Table 3).

Table 3:

Breakthrough force (FB₁ and FB₂)), rupture depth (Δx₁ and Δx₂) and membrane area compressibility modulus (K_A) shown in the table. The values are expressed as mean ± SD.

Chol/POPC mixing ratio	Breakthrough force (nN)		Rupture depth (nm)		KA (mN/m)
	FB1	FB2	Δx2	Δx1	KA
0	5.26±0.28		3.65±0.12		272±69
1	9.25±0.57		4.16±0.10		524±104
1.5	4.30±0.60	11.13±0.77	4.27±0.07	0.72±0.11	631±116
2.5	4.37±0.24	11.55±0.95	3.08±0.05	1.59±0.20	492±76
3.0	6.34±1.20	12.74±1.23	3.95±0.54	1.99±0.34	393±138

As we discussed in this section, several hypotheses have been proposed earlier to explain the double puncture events, and there is no consensus on the origin of it. Here, we propose the CBD hypothesis to explain the double puncture events, but we cannot assure it. Other possibilities like membrane asymmetry (asymmetrical distribution of Chol in the upper and lower leaflet), and uncoupling of the upper and lower leaflet cannot be discarded. Also, the combination of two or more possibilities like lipid asymmetry, uncoupling of the upper and lower leaflet, CBD hypothesis proposed in this manuscript may give rise to double puncture events.

3.3. Membrane thickness from bilayer patches and force curves

The topographical images and height profiles of the lipid bilayer patches for the Chol/POPC membranes with different Chol content are shown in Fig. 5. The height profiles corresponding to the red lines drawn over the images are displayed under each image. The membrane thickness is measured by taking the height difference between the mica substrate and the top of the bilayer, as shown in Fig. 5. We used the rotating box function of nanoscope analysis software at least at five different locations (patch edges) of each image and calculated the average height from three independent experiments. The membrane patches’ measured height for Chol/POPC mixing ratio of 0.0, 1.0, 1.5, 2.5, and 3.0 is shown in the second row of Table 2. The measured height in Table 2 represents the membrane thickness that includes the water layer thickness above the mica surface. It has been reported earlier that the water layer existing between the mica and the membrane extends roughly 1 nm − 2 nm [33, 83, 84]. We have estimated the water layer thickness for bilayer patches at different Chol content, as shown in Table 2. Water layer thickness for the membrane of Chol/POPC mixing ratio 0 is estimated by subtracting membrane height from bilayer patches and the height of POPC membrane without water layer (3.9±0.33 nm) reported in the literature [79], whereas the water layer thickness for all Chol/POPC mixing ratio ≥ 1 is estimated by subtracting membrane height from bilayer patches and the height of PCD (4.57±0.26 nm) reported in the literature [79]. The same height of PCD is used to subtract for all membrane thickness with Chol/POPC mixing ratio ≥ 1 because we expect no further increase in the height of PCD with an increase in Chol content above 50 mol%. Once the PCD is formed, with the further increase of Chol content, CBDs are formed and coexist with the surrounding PCD. The height of CBDs is 3.4 nm [75–78], which is smaller than the height of PCD. Our estimated water layer thickness for Chol/POPC mixing ratio 0 and 1 agree with the water layer thickness variation of 1-2 nm reported in the literature [33, 83, 84]. Surprisingly, we did not find any measurement of water layer thickness for SLM reported in literature for membrane with higher Chol content (i.e., greater than 50 mol%). We observed the water layer thickness increases with Chol content for Chol/POPC mixing ratio > 1 (see Table 2). As discussed in Sect. 3.2, we speculate that this increase in water layer thickness above the Chol/POPC mixing ratio greater than 1 might be due to CBDs’ formation because surface of CBDs has significantly higher polarity than the surrounding PCDs that increases the water accessibility in the CBD region [42]. The proposed hypothesis of an increase in water layer thickness with an increased Chol content is also supported by the increase in the rupture depth for the second puncture event (see Table 2 and 3, and Sect. 3.4), where the second rupture depth corresponds to water layer thickness after the puncture of “CBD water pocket”.

Figure 5:

Topographical images and height profiles of the lipid bilayer patches at different Chol/POPC mixing ratios. Image with Chol/POPC mixing ratios of 0.0, 1.0, 1.5, 2.5, and 3.0 are shown from left to right. The height profiles across the red lines on the respective images are shown under the image. Scale bar = 1 μm for all images.

Table 2:

Membrane thickness measured from bilayer patches (see Fig. 5) and force curves (see Fig. 2) at different Chol/POPC mixing ratios. The values are shown as mean ± SD.

Chol/POPC mixing ratio		0	1	1.5	2.5	3
Membrane height (nm)	Bilayer patches	6.24±0.13	6.64±0.19	6.71±0.23	7.68±0.44	8.81±0.25
	aForce curves	5.78±0.14	5.87±0.15	7.62±0.20	8.35±0.18	10.58±0.35
Water layer thickness (nm)		2.34±0.35	2.07±0.32	2.14±0.35	3.11±0.51	4.24±0.36

^aForce curves are obtained for complete supported bilayer, and the membrane height includes water layer thickness.

Previously, we have used the EPR spin-labeling method and investigated the Chol/PL membranes prepared from total lipid extract of lens cortex and nucleus, obtained from the human donors of different age groups, and found that the CBD size is determined by the Chol content in the membrane[38]. Our previous results showed that with the increase in the donors’ age, the Chol/PL molar ratio increases in cortical and nuclear lens lipid membranes, resulting in an increase in CBDs’ size (see Fig. 8 in [38] that shows the size of CBDs increases with an increase in Chol/PL molar ratio). Based on this previous observation, we speculate that the CBD size increases with an increase in Chol content in the Chol/POPC membrane investigated in this study. Our previous results also show that CBDs’ surface has significantly higher polarity than the surrounding PCDs’ that increases the water accessibility in the CBD region [42]. Thus, with an increase in CBD size, more water molecules accumulate in the CBD regions, increasing the size of the “CBD water pocket”.

We have also estimated the membrane thickness from the force curves (Fig. 2). The membrane thickness from the force curve represents the distance from the top of the bilayer to the mica substrate, including water layer thickness above the mica surface. The membrane thickness is measured from the x-axis in Fig. 2, representing the distance traveled by the AFM tip after touching the top of the bilayer and reaching the bottom (mica substrate) by puncturing the membrane. The third row of Table 2 shows the membrane thickness estimated from the force curves. The thickness was estimated from the three experiments with ~100 force curves for each sample, as representatively displayed in Fig. 2.

Interestingly, thickness measured from the membrane patches and the force curves shows a similar trend (see the second and third-row in Table 2). With the addition of Chol, the membrane thickness increases, and the further increase in thickness is observed when the Chol content is increased from Chol/POPC mixing ratio 1.5 to 3, as observed in the thickness of membrane patches (see Table 2), supporting our hypothesis of increasing the size of CBD and “CBD water pocket” that leads to increasing water layer thickness, and the overall thickness of the membrane. From Table 2, we can see that although the membrane thickness from bilayer patches and the force curve from the complete membrane shows a similar trend, there is a slight variation in the membrane thickness measured from membrane patches vs. force curves. It has been reported earlier that an AFM tip may experience short-range interactions that depend upon the membrane chemistry before actual contact with the membrane [95] that may result in a difference of membrane thickness measured from force curves vs. bilayer patches.

3.4. Rupture depth

The distance that the AFM probe traverses after puncturing the membrane, before reaching another region/or substrate, is the rupture depth. For the membrane consisting of a single puncture event, the AFM tip contacts the mica substrate after rupturing the membrane. We obtain such a single puncture event for membrane with Chol/POPC mixing ratio up to 1.0 and hence single rupture depth (see Fig. 2). The average rupture depth for the Chol/POPC mixing ratio 0 and 1 are 3.65 ± 0.12 nm and 4.16 ±0.10 nm, respectively. However, two rupture depths are defined for the membrane exhibiting two breakthrough events, as shown in Fig. 6A. We represent the first rupture depth originating after the first breakthrough force (Δx₂), whereas the second rupture depth (Δx₁) as shown in Fig. 6A. The average rupture depth of first puncture events that correspond to coexisting CBDs and PCDs are 4.27 ± 0.07 nm, 3.08 ± 0.05 nm, and 3.95 ± 0.54 nm for Chol/POPC mixing ratio 1.5, 2.5, and 3.0, respectively. There is a slight decrease in average rupture depth for these membrane types as the Chol/POPC mixing ratio reaches 2.5 but increases for mixing ratio to 3.0. Interestingly, the average rupture depth increases linearly for the second puncture event region as the Chol/POPC mixing ratio increases (see Fig. 6B). The second puncture event arises likely due to the “CBD water pocket”. The average rupture depth for this region is 0.72 ± 0.11 nm, 1.59 ± 0.20 nm, and 1.99 ± 0.34 nm for mixing ratios 1.5, 2.5, and 3.0, respectively. The second rupture depth correspond to water layer after the puncture of “CBD water pocket. With the increase in Chol content, the size of CBD and “CBD water pocket” increases, resulting in an increase in water layer thickness followed by an overall increase in membrane thickness as discussed in Sect. 3.3. The increase in water layer thickness with an increase in Chol content above Chol/POPC mixing ratio of 1 can be correlated with an increase in the second puncture even’s rupture depth as a function of Chol content shown in Fig. 6B. We speculate that this increase in rupture depth for the second puncture event can be correlated with an increase in size of CBDs with increasing Chol content (i.e., size of “CBD water pocket” increase with an increase in size of CBD). Previously, we have used EPR-spin labeling method and found that CBDs size increases with an increase in Chol content [38]. The average rupture depth of the membrane system is shown in Table 3.

Figure 6:

(A) Representative force curves of the lipid membrane exhibiting two puncture events with rupture depth in proximal region (Δx₁) and distal region (Δx₂). The representative force curve is obtained by rupturing the membrane consisting of Chol/POPC mixing ratio of 1.5. For all the Chol/POPC mixing ratio exhibiting two breakthrough forces, the rupture depth of distal region corresponds to PCD coexisting with CBDs whereas the rupture depth of proximal region corresponds to water layer after the puncture of “CBD water pocket”. (B). Rupture depth for Chol/POPC mixing ratio from 0 to 3. Here, for the bilayer exhibiting two puncture events (i.e., Chol/POPC mixing ratio > 1.0) the rupture depth for coexisting PCD and CBD is represented by green, whereas the grey bars represent the rupture depth of water layer after the puncture of “CBD water pocket”. The rupture depth in proximal region (Δx₁) increases with increase in Chol content.

Previously, for a single puncture event, rupture depth is also used as a membrane thickness [86, 96]; however, the elastic deformation from the tip during indentation can alter the measured bilayer thickness [93]. Moreover, rupture depth is a slight underestimation of the membrane thickness, while the depth from the point of contact may be an overestimation, as the AFM tip may experience short-range interactions, which also depends upon the membrane chemistry, before actual contact with the membrane [95]. However, it correlates strictly with the membrane thickness [97]. In our experiment, rupture depth broadens on the bilayer containing Chol/POPC bilayer mixing ratio 1.0, then POPC alone membrane. Such increment in bilayer thickness by Chol, below the Chol saturation limit, on the membrane containing unsaturated lipid, was observed [33, 98, 99]. However, above the Chol saturation limit, we obtained no significant trend in the distal region’s rupture depth, whereas the rupture depth exhibits almost a linear increment in the proximal region (Fig. 6B). This increment can be attributed to the increased size of CBD with incremental Chol content in the membrane and resulting substantially bigger “CBD water pocket” as discussed above.

Since the CBDs are not visualized in the topographical images, the size of the CBDs should be significantly smaller than the size of the AFM tip (the area of Chol molecule in the membrane can be ~0.28 nm² [100], and the surface of the AFM tip is ~3000 nm²). When the AFM tip is compressing the membrane, it may be likely that there are many CBDs under the AFM tip that co-exist with surrounding PCDs. Since the PCD is ~0.6 nm higher height than the CBD on both sides of the bilayer, the AFM tip touches the PCD first and touches CBD after the AFM tip travels ~0.6 nm further. In the presence of CBDs which co-exist with the surrounding PCD, AFM tip experience less force to puncture the membrane, as seen by the sharp decrease in the first puncture force for all the sample above Chol/POPC mixing ratio > 1 (see Fig. 4B). For all the Chol/POPC mixing ratios > 1, the first puncture force is the combined contribution of co-existing PCDs and CBDs, whereas, for the Chol/POPC mixing ratio of 1, the puncture force is only the contribution from PCD. Previously we have studied Chol/POPC membrane with EPR spin labeling method and observed an increase in membrane fluidity for Chol/POPC mixing ratio greater than 1 where the increase in the fluidity was the combined effect of co-existing CBDs and PCD (see Fig 7C in [42] where 1/T1 gives the membrane fluidity). In the case of Chol/POPC mixing ratio equal to 1, the contribution to rupture depth is only from PCD. For all the Chol/POPC mixing ratio > 1, the first rupture depth (Δx₂) originating from the first breakthrough force is the combined contribution of co-existing PCDs and CBDs, whereas the second rupture depth (Δx₁) originating from the second breakthrough force correspond to the rupture depth of water layer above the mica surface. The second rupture depth (Δx₁) increases with an increase in Chol content which correlates interestingly with increased water layer thickness as a Chol content increases (see Table 2 and 3). We hypothesize that water layer thickness increases with Chol content for Chol/POPC mixing ratio greater than 1 due to CBDs’ formation as discussed in Sect. 3.3.

3.5. Area Compressibility modulus (K_A)

To access the membrane’s nanomechanical property, we use the elastic deformation region of the force curves. In this regime, the stretching of the lipid molecules within the AFM tip neighborhood dictates the tip force. Assuming spherical tip radius and tension-free membrane, the tip force is quadratically related with the tip-mica separation ‘s’ given by [29, 64],

$$F=\pi K_{A}R{\left(\frac{D-s}{s}\right)}^2$$ (1)

Where K_A is the membrane’s area compressibility modulus, D is the thickness of the membrane, and R is the end radius of the AFM tip. However, in supported lipid membranes, a thin water layer exists between the substrate and the membrane extending roughly 1 nm − 2 nm [33, 83, 84], providing lubrication and maintaining sufficient mobility for the lipid molecules [84]. Thus, in equation 1, D is the sum of bilayer thickness and thickness of the water layer for non-zero water thickness. Using equation (1) and the force spectroscopy data, we determine the mechanical property (K_A) of the supported lipid membrane for all the Chol/POPC mixing ratios. K_A is used to quantify the energy cost for compressing /stretching the membrane under stress. For Chol/POPC mixing ratio >1, two puncture events are observed (see Fig. 4B); the first puncture event likely corresponds to coexisting CBDs and PCD, whereas the second puncture event likely corresponds to “CBD water pocket” as discussed in Sect. 3.2. In two puncture events, K_A is estimated only for the first puncture event because it represents the lipid bilayer with coexisting CBDs and PCD (see Sect. 3.2). For all Chol/POPC mixing ratios, D in equation 1 represents the sum of bilayer thickness and water thickness. The thickness is taken from the AFM tip’s first contact point in the membranes. In Fig. 7A, we show the fit in the representative force curve’s elastic stretching regime for Chol/POPC mixing ratio of 1. The circles are the data points in the approach section of the force curves, and solid lines are the best fit for equation 1. Since equation 1 is valid only in the elastic regime where lipid bilayers deform elastically, we consider the fit within only 80% of the breakthrough force to avoid the fit range deviating off the elastic regime. The average K_A for the membranes with Chol/POPC mixing ratio 0 to 3 is shown in Fig. 7B. The K_A value increases up to Chol/POPC mixing ratio of 1.5, and with a further increase of Chol content, K_A starts to decrease. The increased K_A value with an increase in Chol content suggests the decrease in membrane elasticity till Chol/POPC mixing ratio of 1.5; however, surprisingly, at higher Chol content (Chol/POPC mixing ratio >1.5), K_A starts to decrease, suggesting in an increase in membrane elasticity. Most of the PC lipids, containing variable chain length and saturation, have their K_A in the range 229-265mN/m [23], which is derived by aspiration of GUVs using micropipette. In comparison, K_A for the POPC only membrane (272 ± 69 mN/m) obtained in our current studies agrees with the previously measured value [26]. Our studies show an increase in K_A value with increasing Chol content till Chol/POPC mixing ratio of 1.5, which agrees with the previously reported trend [22, 26], however to the best of our knowledge, K_A values are reported only for Chol/POPC bilayers with mixing ratio less or equal to 1 [26, 101, 102]. Our attempt is the first to estimate the K_A values for high Chol containing membrane (Chol/POPC mixing ratio >1). Surprisingly, our results suggest an increase in membrane elasticity at very high Chol content (Chol/POPC mixing ratio >1.5). The K_A for membranes with Chol/POPC mixing ratio 0 to 3 is shown in Table 3.

Figure 7:

(A) Example of a model fit for force separation curves for bilayer area compressibility modulus (K_A) using equation 1. The hollow green circles are the experimental data and the solid purple line is the fit for Chol/POPC mixing ratio of 1. (B) K_A for the membrane with different Chol/POPC mixing ratio. The circles represent the average area compressibility modulus for POPC membrane without Chol (Chol/POPC mixing ratio of 0), PCD membrane (Chol/POPC mixing ratio of 1), and coexisting PCD and CBD region (Chol/POPC mixing ratio > 1). The modulus increases as the PCD is formed on the membrane and peaks at Chol/POPC mixing ratio 1.5. On further addition of Chol in the POPC membrane, the modulus decreases with increased Chol in the membrane.

The increase in membrane elasticity (i.e., decrease in K_A) results from the formation of coexisting PCD and CBDs at higher Chol content. This result is supported by AFM image roughness measurement (Sect. 3.1) and AFM puncture force measurement (Sect 3.2). As discussed in Sect. 3.1, membrane roughness decreases with increase in Chol content till Chol/POPC mixing ratio of 1.5 and above Chol/POPC mixing ratio of 1.5, membrane roughness increase is likely due to the formation of coexisting PCD and CBDs. Similarly, a sharp decrease in puncture force was observed above Chol/POPC mixing ratio >1 due to the formation of coexisting CBDs and PCD (see Fig. 4B). Furthermore, Chol/POPC mixing ratio >1 AFM tip experiences second puncture force in “CBD water pocket” that is formed due to the difference in thickness of coexisting CBDs and PCD as discussed in Sect. 3.2 (see also Fig. 4B). Our data suggest that the CBDs formation at high Chol content increases the membrane elasticity (i.e., decrease K_A). Previously, we have used EPR spin labeling method and investigated the fluidity of Chol/POPC membrane for Chol/POPC mixing ratio 0 to 3 [42]. EPR spin-labeling method results suggest that membrane fluidity decrease with increase in Chol content up to Chol/POPC mixing ratio of 1, however at high Chol content (Chol/POPC mixing ratio >1) CBDs are formed, and membrane fluidity increase suggesting that formation of coexisting CBDs and PCD leads to increase in membrane fluidity (see Fig. 7C in [42]). K_A values obtained with AFM (results presented here) follow a similar trend as the fluidity parameter (1/T₁) obtained with EPR [42]. AFM data showing the increase in membrane elasticity at very high Chol content (see Fig. 7B) complement with EPR data that shows an increase in membrane fluidity at very high Chol content (see Fig. 7C in [42]).

Previously Yip et al. [103], used fluorescence technique and investigated the membrane fluidity by adding exogenous Chol up to 50 % in total brain lipid extracts and observed the non-linear pattern in membrane fluidity. For the same total brain lipid extracts system, Shamitko-Klingensmith et al. added exogenous Chol up to 50% and observed the similar non-linear pattern membrane thickness, and membrane mechanical properties [104]. Interestingly we have also observed the non-linear pattern in K_A and the first rupture depth (Δx₂) when the Chol content was increased from 0 to 75 mol% (see Table 3). The K_A and the first rupture depth (Δx₂) increases when Chol/POPC mixing ratio was increased from 0 to 1.5 and decrease above Chol/POPC mixing ratio of 1.5. We suggest that non-linear pattern is likely due to the formation of CBDs as discussed in this manuscript. Formation of CBDs in the membrane depends upon the phospholipid composition [50]. We found previously that CBDs starts to form in POPE membrane when Chol content is greater than 33 mol% whereas in the case of POPC membrane CBDs starts when the Chol content is greater than 50 mol% [50]. We think that CBDs hypothesis in this new manuscript may have broad application to understanding the non-linear pattern in other systems like brain lipid extracts.

Thus, our result demonstrates the membrane mechanical property variation in a very high Chol containing lipid membrane, which is very distinct from the membrane with low Chol content. Based on the puncturing information of the Chol containing membrane by AFM tip, we propose the membrane architecture prospect as presented in Fig. 8. For the Chol/POPC mixing ratio ≤ 1, the lipid environment is speculated in Fig. 8A, comprising PL bilayer for 0 mol% Chol and PCD for 50 mol% Chol with a single breakthrough event at “b”. However, for the Chol/POPC mixing ratio > 1, the lipid environment is speculated in Fig. 8B, comprising coexisting CBD and PCD.

Figure 8:

Proposed schematics of supported lipid bilayer membrane puncture with AFM tip along with representative force curves for single and double puncture events. A) For the membrane with Chol/POPC mixing ratio 0 (left; PL bilayer) and 1 (right; PCD), single breakthrough event is observed. B) For the membrane with Chol/POPC mixing ratio > 1, double puncture events are observed. The dotted vertical lines in the force curve are the solid substrate. The colored horizontal lines in membrane represent the tentative position of the AFM tip during membrane penetration with corresponding alphabetical position in the force curves. For first figure (A), only single breakthrough event is obtained as the AFM tip ruptures the membrane. For membranes with Chol/POPC >1.0 (B), the first breakthrough event is experienced due to the presence of coexisting PCD and CBD whereas second breakthrough event is experienced due to “CBD water pocket” evolved by thickness mismatch of CBD and PCD. High polarity of CBD surface compared to surrounding PCD results in a second puncture event with higher breakthrough force.

We believe the water pocket introduced by the thickness mismatch of these coexisting CBDs and PCD seeds as an impediment for AFM tip progression. The CBD surface has significantly lower hydrophobicity (higher polarity) in comparison to surrounding PCD [42], whereas the CBD center is as hydrophobic as the center of PCD [42], which results in the second puncture force. Based on the force curves, we define the two different regions: the PCD-CBD coexisting region (“a” to “d”, with an initial breakthrough event at “b”; see Fig. 8B) and the “CBD-water pocket” region plus the water layer (“c” to the substrate, with the second breakthrough event at “d”; see Fig. 8B).

4. Conclusion

This study provides incredibly insightful information regarding the very high cholesterol-containing lipid membrane’s mechanical property using atomic force microscopy. Based on the surface roughness measurement, membrane force measurement, and compressibility modulus data suggests that high Chol content membrane leads to the formation of coexisting CBDs and PCD in SLM where Chol and CBDs play a significant role in modulating membrane elasticity. This study’s significant conclusion is that high Chol content in the membrane increase membrane elasticity, as documented by a decrease in KA at high Chol content (> 60 mol %). Since the human eye lens membrane contains an extremely high Chol content level, this study presented here will help us understand the role of cholesterol in the lens membrane to regulate lens membrane elasticity. Furthermore, this study can be extended to understand high cholesterol’s role in presbyopia’s fundamental cause, i.e., loss of ability to accommodate images in the lens [53, 66–68].

Highlights.

Mechanical properties of the high cholesterol-containing membrane using AFM. Cholesterol and CBDs play a significant role in modulating membrane elasticity. Membrane compressibility modulus decreases at high cholesterol content.

The surface roughness of the membrane increases at high cholesterol content. AFM confirms the formation of coexisting CBDs and PCD in SLM.