Confocal microscopy confirmed that in phosphatidylcholine giant unilamellar vesicles with very high cholesterol content pure cholesterol bilayer domains form

Abstract

The cholesterol (Chol) content in the fiber cell plasma membranes of the eye lens is extremely high, exceeding the solubility threshold in the lenses of old humans. This high Chol content forms pure Chol bilayer domains (CBDs) and Chol crystals in model membranes and membranes formed from the total lipid extracts from human lenses. CBDs have been detected using electron paramagnetic resonance (EPR) spin-labeling approaches. Here, we confirm the presence of CBDs in giant unilamellar vesicles prepared using the electroformation method from Chol/1-palmitoyl-2-oleoylphosphocholine and Chol/distearoylphosphatidylcholine mixtures. Confocal microscopy experiments using phospholipid analog (1,1’-dioctadecyl-3,3,3’,3’-tetramethylindocarbocyanine-5,5’-disulfonic acid) and cholesterol analog fluorescent probes (23-(dipyrrometheneboron difluoride)-24-norcholesterol) were performed, allowing us to make three major conclusions: (1) In all membranes with a Chol/phospholipid (PL) mixing ratio (expressed as a molar ratio) greater than 2, pure CBDs were formed within the bulk PL bilayer saturated with Chol. (2) CBDs were present as the pure Chol bilayer and not as separate patches of Chol monolayers in each leaflet of the PL bilayer. (3) CBDs, presented as single large domains, were always located at the top of giant unilamellar vesicles, independent of the change in sample orientation (right-side-up/upside-down). Results obtained with confocal microscopy and fluorescent Chol and PL analogs, combined with those obtained using EPR and spin-labeled Chol and PL analogs, contribute to the understanding of the organization of lipids in the fiber cell plasma membranes of the human eye lens.

1. Introduction

The fiber cell plasma membranes of the human eye lens are the only biological membranes in the human body where the cholesterol (Chol) content is so high that pure Chol bilayer domains (CBDs) and/or Chol crystals form (1–7). The need for high Chol content in the eye lens is not well understood, and it is not known whether the appearance of CBDs and/or Chol crystals is harmful or beneficial for lens function. However, the disturbance of Chol homeostasis in fiber cells may result in damages associated with cataracts (8–16). Our investigations indicate that the high Chol content and the presence of CBDs play a significant function in maintaining lens transparency (8, 17–19). Because of that, we used EPR spin labeling (1–3, 20–23), molecular dynamics (MD) simulation (24, 25), and presently confocal fluorescence microscopy to obtain maximum information about the structure, dynamics, and other properties of the CBD immersed within the phospholipid (PL) bilayer.

The first reports that high Chol content can form pure Chol bilayers in PL membranes were based on the X-ray diffraction measurements and the observed 34 Å diffraction pattern in membrane preparations (5, 6, 26). The same 34 Å pattern was observed for pure Chol crystals (27). Also, other reported properties were the same as those for the pseudo-bilayers of Chol in Chol crystals, namely the rigidity of the structure and the conclusion that it is formed by anhydrous Chol molecules.

The major inconsistences between X-ray diffraction and EPR spin-labeling data include the rigid versus fluid structure of the observed domains and the anhydrous versus hydrated form of Chol molecules forming pure Chol bilayers. These inconsistences were discussed in a previous paper (8), the final conclusion being that Chol molecules in the CBD possess a high level of mobility, similar to that of Chol molecules in the surrounding PL bilayer. This conclusion was confirmed through an MD simulation of the behavior of Chol molecules in the pure Chol bilayer and in the PL bilayer saturated with Chol (24, 25). Also, both the EPR spin-labeling approach and MD simulation indicated that the -OH group of Chol in the pure Chol bilayer is highly accessible to water molecules, and, thus, this bilayer cannot be formed by anhydrous Chol molecules (20).

Our most significant finding was that the physical properties of the lipid bilayer of the eye lens membranes are independent of the drastic changes in PL composition, but only when these membranes are saturated with Chol (1, 19). Such a state is only possible when the membrane possesses a Chol reservoir in the form of pure CBDs, ensuring the lipid bilayer—which acts as a buffer, releasing and incorporating Chol molecules—is saturated with cholesterol (18, 19). This explains why the drastic changes in the PL composition that occur with age (28–32), together with the lack of protein turnover (33, 34), do not impair the homeostasis of the fiber cell membranes and lens. Additionally, the saturating Chol content and the presence of CBDs increase the barrier to oxygen permeation into the lens interior (35–39) as well as the membrane hydrophobicity (2, 3, 23, 40–42), which decreases permeation of polar molecules across the membrane. All these properties are significant in maintaining lens transparency (see (17, 19, 20) for more explanations).

The investigations of CBDs were mainly performed on model membranes made of PL, which are the major constituents of the plasma membrane of the eye lens fiber cells (19, 20, 22, 40, 43, 44), and on lens lipid membranes made of the total lipid extracts from the cortical and nuclear portions of eye lenses (1–3, 21, 23). The conclusion of these works stated that the formation of CBDs precedes the formation of Chol crystals, and the appearance of each depends on the type of PL forming the bilayer (20). When the Chol content in the PL bilayer increases above the upper limit that can be accommodated within the bilayer (i.e., the Chol saturation limit), excess Chol forms CBDs supported by the surrounding bilayer (20), forming a structured (45) (or dispersed (46)) liquid ordered phase. The next limit is the Chol solubility threshold, which indicates the total Chol content in the PL bilayer that can be dissolved in and supported by the bilayer (in the form of CBDs). Above this limit, a new phase is formed, namely Chol crystals (20, 27). Recently, using an EPR spin-labeling approach, we were able to detect CBDs in intact fiber cell plasma membranes from porcine and human eye lenses (7). The Chol content for both the cortical and nuclear fiber cells of lenses obtained from human donors was always high enough to induce the formation of CBDs in intact membranes. CBDs in intact membranes exhibit properties like those observed for CBDs in model membranes.

The EPR spin-labeling approach cannot confirm that the CBD is actually formed as a pure transmembrane Chol bilayer and not by independently located patches of Chol monolayers in each leaflet of the PL-Chol bilayer. Also, reliable information about the size of CBDs in model and intact membranes is missing. Confocal microscopy has been widely used to study the effects of cholesterol on domain formation in model membranes (47–51). This technique, used with PL analog (1,1’-dioctadecyl-3,3,3’,3’-tetramethylindocarbocyanine-5,5’-disulfonic acid) and Chol analog fluorescent probes (23-(dipyrrometheneboron difluoride)-24-norcholesterol), should help provide this missing information. The major advantage of this approach is that the PL and Chol analog fluorescence probes can be present simultaneously in the investigated membranes. However, because of the difference in their excitation and fluorescence wavelengths, the locations of the Chol and PL analog fluorescent probes can be monitored separately in the same samples (membranes). This is advantageous because confocal microscopy can be used to visualize domains where both Chol and PL analog fluorescence probes are incorporated, as well as domains where only one of these probes is incorporated.

2. Materials and Methods

2.1. Materials

The one-palmitoyl-2-oleoylphosphatidylcholine (POPC), distearoylphosphatidylcholine (DSPC), Chol, and Bodipy CHOL fluorescence probes were obtained from Avanti Polar Lipids, Inc. (Alabaster, AL). The PL analog DiIC₁₈(3) fluorescence probe was purchased from Invitrogen, Thermo Fisher Scientific, (Waltham, MA). Chemical structures of these compounds and their approximate locations in the lipid bilayer are presented in Fig. 1. Other chemicals of at least reagent grade were purchased from Sigma-Aldrich (St. Louis, MO).

Fig. 1.

Chemical structures of phospholipid (DiIC₁₈(3)) and cholesterol (Bodipy CHOL) analogue fluorescent probes together with structures of saturated (DSPC) and unsaturated (POPC) phospholipids and cholesterol (Chol). Approximate locations of these molecules across the lipid bilayer membrane are illustrated.

2.2. Electroformation

GUVs of 10–100 μm diameter were made starting from Chol/POPC and Chol/DSPC mixtures in chloroform with molar ratios from 1/2–5/1 and were labeled simultaneously with both PL (DiIC₁₈(3)) and Chol (Bodipy CHOL) analog fluorescence probes (with DiIC₁₈(3)/PL and Bodipy CHOL/Chol molar ratios of 1/100). They were prepared using the electroformation method (52, 53). Appropriate aliquots (up to 30 μL) were spread at room temperature as a thin film on a 25 mm × 37.5 mm glass slide (a 25 mm × 75 mm slide was cut in half) (CG-90INS115 Delta Technologies, Loveland, CO) coated with indium-tin-oxide. The final lipid mass was 0.25 mg. A coated slide was placed under reduced pressure for at least 30 min to remove the remaining solvent. After the chloroform evaporated, the films were visually inspected to ensure that they were translucent and produced the smooth color spectrum characteristic of thin-film interference. Lipid films that showed evidence of lipid demixing and did not produce a smooth color spectrum were re-spread using chloroform at a high temperature (~50°C). Two slides (lipid coated and uncoated) were coupled face-to-face (the two conducting sides faced inward) with a 0.5 mm Teflon spacer to form a capacitor. The capacitor was filled with 18 mΩ-cm water, and the ends were sealed with vacuum grease. Vesicles were grown for 1.5 h at 60°C under an AC voltage of 1 V at 10 Hz and then kept at the same temperature for 1 h before measurements were taken. Most of the lipid film was incorporated into vesicles, and only a small amount remained as a film on the slides. During our preparation (modified film deposition method), the lipid mixture passed through the solid-state intermediate at which solid-state demixing of cholesterol can occur. It decreased the Chol/PL molar ratios in membranes as compared with that in the chloroform mixture (as the mixing ratio). In all cases, we found the lipid compositions to be the same as they were when prepared in the organic solvent before GUV growth (as a mixing ratio). We have not confirmed that the resulting GUVs have identical lipid compositions. To detach the vesicles from the substrate, we collected the vesicle solution in a syringe with a 22-gauge needle. Before diluting samples for confocal microscopy, the electroformation efficiency for each sample was confirmed using epifluorescence microscopy (Fig. 3). Slides were mounted on an inverted Nikon Eclipse TE2000U microscope outfitted with a motorized, computer-controlled stage. Both green (Ex = 490/20 nm, Emm = 528/38 nm) and red (Ex = 555/28 nm, Emm = 617/73 nm) fluorescence channels were imaged using a 20× objective.

Fig. 3.

The epifluorescence micrograph of GUVs made of Chol/DSPC mixtures with a mixing ratio of 4/1 labeled by both DiIC₁₈(3) and Bodipy CHOL fluorescent probes and grown in a chamber made of two indium-tin-oxide-coated slides. The image was collected using a green fluorescence channel and shows the initial suspension of vesicles generated by the electroformation method.

2.3. Preparation of samples for confocal microscopy

To view the sample, a vesicle suspension that had been diluted 100 to 200 times was placed between two coverslips coupled with a 0.36 mm thick spacer made of double-sided tape (to prevent evaporation).

2.4. Confocal microscopy

Confocal microscopy image acquisitions of GUVs were completed on the same day at 22°C on a Carl Zeiss LSM 510 confocal inverted microscope (Jena, Germany), using C-Apochromat 63×/1.2 NA W corr M27 (water immersion objective). To select the best imaging parameters, individual fluorophores were used to check for cross excitation and bleedover, and the best dichroic, filter set, and excitation/emission settings were chosen. DiIC₁₈(3) was excited with a 561 nm laser and Bodipy CHOL was excited with 488 nm. Emission filters BP 500–530 nm and LP 575 nm were used to capture emitted light from Bodipy CHOL and DiIC₁₈(3), respectively. Images were 8 bit with a pixel time of 2.88 ms. A unidirectional line scan was performed to register the fluorescence location in the multitrack mode. We collected green and red signals in an independent mode where each fluorophore was excited by its own individual excitation wavelength, and the resultant emission for the first fluorophore (green) was collected. Any cross-excitation fluorescence was not collected. This was followed by the next fluorophore (red) data, and this sequence was repeated over different Z-levels. This enables each fluorophore to be captured independently and in microseconds apart from each other in order to prevent bleaching artifacts. To gain a three-dimensional distribution of DiIC₁₈(3) and Bodipy CHOL in the GUV, Z-scans were captured at Nyquist using Aim 4.3 software. Each Z-scan image was an average of four images and Z-scans were converted into maximum intensity projections (MIP) using AxioVision 4.8 (Jena, Germany) software. Images presented in this manuscript are MIP images because the two-dimensional raw data were mere rings of red, green, and yellow pixels and do not provide information about the lipid bilayer organization or dynamics. Three-dimensional renderings of the Z-stack images were performed under the same settings, including threshold (lower limit set at 0.05 for both green and red) orientation, projected image brightness, etc. In order to verify membrane localization, GUVs (mixing ratios from 1/2–5/1) were imaged from both sides (successively) and the registered Z-stacks were analyzed. At least 10 vesicles were analyzed for each sample.

3. Results and Discussion

3.1. PL- and Chol-analog fluorescent probes discriminate and characterize the CBD

Fluorescently labeled PL and Chol analogs may be added together if their excitation wavelengths and fluorescence wavelengths differ significantly. Here, we chose DiIC₁₈(3) as a PL analog (with an excitation wavelength of 549 nm and an emission wavelength of 565 nm) and Bodipy CHOL as a Chol analog (with an excitation wavelength of 495 nm and an emission wavelength of 507 nm). Thus, these fluorescent probes can be simultaneously added to the investigated membranes, and their localization can be observed either as images constructed of the two simultaneously acquired fluorescence signals or as images coming from the fluorescence signals of only one of the probes (DiIC₁₈(3) or Bodipy CHOL). This highlights the major difference in experimental approaches resulting from application of two different techniques (EPR and fluorescence).

As shown in Fig. 2, Bodipy CHOL, as a Chol analog, can be located in both purported domains, namely the phospholipid-cholesterol domain (PCD; PL-bilayer saturated with Chol) and the CBD, existing in PC bilayers at high Chol content. However, DiIC₁₈(3), as a PL analog, should be located only in the PCD and should not partition into the CBD. This fact was used as the foundation for the strategy applied in the presented investigations. In these experiments, we looked directly at the physical location of fluorescent probe molecules. In our study, the absence of the fluorescence signal from the PL analog DiIC₁₈(3) clearly indicates that PLs are not present at a detectable level in one domain, while the fluorescence signal from the Chol analog Bodipy CHOL images both domains. Thus, we can state that in our experiments, the CBD is discriminated using the PL analog florescent probe. In the EPR approach, only the Chol analog spin label can discriminate the presence of the CBD from the surrounding PCD (1–3, 21, 54, 55) because we are looking at different properties of the spin label in the discriminated domains (i.e., the probe molecule must be present in both domains).

Fig. 2.

Schematic drawing of the CBD formed at high Chol content as an integral part of the PC bilayer (here with saturated acyl chains). The distribution and approximate locations of DiIC₁₈(3) and Bodipy CHOL fluorescent probes are indicated. The fluorescent moieties of fluorescent probes, PL analog DiIC₁₈(3) and Chol analog Bodipy CHOL, are indicated by black dots.

The colocalization of both fluorescent probes allowed us to diminish, or even completely cancel, artifacts from the GUV sedimentation. During sedimentation, GUVs can leave the confocal (focusing) area, giving it “empty space.” In our approach, the confocal fluorescence images of the GUV were obtained first from an overlay of two simultaneously acquired fluorescence signals (each of them collected in the individual color channel); the combined images, were then post-processed to obtain separate images for each fluorescent probe component.

3.2. At high Chol content, CBDs form in GUVs made from saturated and unsaturated PCs

Using the confocal microscopy with Chol and PL analog fluorescent probes allowed us to confirm the presence of the pure CBDs in GUVs made from both saturated and unsaturated PCs (DSPC and POPC). The results are presented in Figs. 4 and 5.

Fig. 4.

Confocal fluorescence images of GUVs made of Chol/DSPC mixtures, with mixing ratios ranging from 0.5 to 5 and recorded at 22°C. Each vesicle contains both fluorescent probes, DiIC₁₈(3) (red) and Bodipy CHOL (green). In the upper row, each image is an overlay of two simultaneously acquired fluorescence signals from DiIC₁₈(3) and Bodipy CHOL. These same data are presented in the middle and bottom rows, but the fluorescence signals originate from DiIC₁₈(3) in the middle row and Bodipy CHOL in the bottom row, respectively. We collected green and red signals in an independent mode. Each fluorophore was excited by its own individual excitation wavelength, and the resultant emission for the first fluorophore (green) was collected. This was followed by the next fluorophore (red) data, and this sequence was repeated over different Z-levels. This enables each fluorophore to be captured independently and in microseconds apart from each other. To gain a three-dimensional distribution of DiIC₁₈(3) and Bodipy CHOL in the GUV Z-scan images were converted into MIPs. Three-dimensional renderings of the Z-stack images were performed under the same settings, including threshold (lower limit set at 0.05 for both green and red) orientation, projected image brightness, etc. Arrows show the lack of fluorophore in certain parts of the vesicles.

Fig. 5.

Confocal fluorescence images of GUVs made of Chol/POPC mixtures, with mixing ratios from 1 to 4 and recorded at 22°C. Each vesicle contains both fluorescent probes, DiIC₁₈(3) (red) and Bodipy CHOL (green). In the upper row, each image is an overlay of two simultaneously acquired fluorescence signals from DiIC₁₈(3) and Bodipy CHOL. These same data are presented in the middle and bottom rows, but fluorescence signals originate from DiIC₁₈(3) in the middle row and Bodipy CHOL in the bottom row, respectively. We collected green and red signals in an independent mode. Each fluorophore was excited by its own individual excitation wavelength, and the resultant emission for the first fluorophore (green) was collected. This was followed by the next fluorophore (red) data, and this sequence was repeated over different Z-levels. This enables each fluorophore to be captured independently and in microseconds apart from each other. To gain a three-dimensional distribution of DiIC₁₈(3) and Bodipy CHOL in the GUV Z-scan images were converted into MIPs. Three-dimensional renderings of the Z-stack images were performed under the same settings, including threshold (lower limit set at 0.05 for both green and red) orientation, projected image brightness, etc. Arrows show the lack of fluorophore in certain parts of the vesicles.

3.3. CBDs form as transmembrane pure Chol bilayers in GUVs

We had no strong experimental evidence that the CBD is actually formed as a transmembrane pure Chol bilayer and not as patches of the pure Chol monolayers independently located in each leaflet of the PL-Chol bilayer. However, following the first X-ray diffraction communications about the existence of CBDs in membranes oversaturated with Chol (5, 6, 26), we depicted the CBD as a pure Chol bilayer immersed into the PL-Chol bilayer (1–3, 21–23, 41, 44, 56). This hypothesized CBD structure was never reliably confirmed experimentally. Finally, with the results presented here, we are able to unequivocally demonstrate that the CBD is formed as a transmembrane pure Chol bilayer.

The upper rows of confocal fluorescence images presented in Figs. 4 and 5 are an overlay of two simultaneously acquired fluorescence signals from PL and Chol analog fluorescent probes. Fluorescent probes simulate the distribution of their parent molecules—PLs and Chol—in GUVs; thus, the shape of an individual GUV is visible in each image. The middle row presents the same images as the upper row, but only the fluorescence signal from the Chol analog is visible (green color); fluorescence from the PL analog is omitted. Thus, these images show the distribution of Chol molecules within GUVs. These images are practically the same as those in the upper row, which indicates that Chol molecules are distributed (present) uniformly across the entire GUV surface. The lower row presents the opposite situation: The fluorescence signal from the Chol analog is omitted and only the signal from the PL analog (red color) is visible, showing the distribution (localization) of PL molecules across the surface of the GUV. It can be clearly seen that the tops of the GUVs, with a Chol/PL mixing ratio (expressed as a molar ratio) >2.5, have no red color (as compared with the images in the two upper rows). It indicates that the PL molecules are excluded from the top surface areas of the GUVs. This is consistent with the detection limits we set to see the fluorescence. The images in the middle row show that cholesterol molecules are present across the entire GUV. Thus, the top parts of the GUVs are formed by Chol molecules only. GUVs are formed by the lipid bilayer; therefore, our results allow us to conclude that the top parts of the GUVs are formed by a pure transmembrane Chol bilayer.

3.4. CBDs form as large patches of pure Chol bilayers in GUVs

As shown in Figs. 4, 5, and 6, in confocal fluorescence images of GUVs made of saturated and unsaturated PCs (DSPC and POPC), CBDs present as large patches of pure Chol bilayers. For Chol contents near the Chol solubility threshold, the total area occupied by CBDs should be maximal. For PC membranes, the Chol solubility threshold, which is defined as the maximal Chol content in the PL bilayer, above which Chol crystals are formed outside the lipid bilayer, is ~66 mol% (Chol/PC molar ratio of 2/1) (27, 57, 58). Also, for PC membranes, the Chol concentration that saturates the PC bilayer, and above which the pure CBDs begin to form within the PC-Chol bilayer, is ~50 mol% (Chol/PC molar ratio of 1/1) (20). Thus, the ratio of the area occupied by the CBD to that occupied by the PCD at the maximal Chol content in the PC membranes will be equal to the ratio of the surface occupied by one Chol molecule in the CBD to the surface occupied by one PC and one Chol in the PCD. According to data from the literature, it should be ~1/2.5 (25). A very brief evaluation indicates that the area occupied by the CBD, at a saturating Chol content (at a Chol/PC mixing ratio >3), is about 30% of the total area of the GUV. It fits with the data that CBD formation precedes formation of Chol crystals, and CBDs begin to form at a Chol/PC molar ratio of 1/1 (20). The maximal amount of Chol that can be supported by the PC-Chol bilayer in the form of a CBD is 50% Chol added to the PC liposome membranes. We would like to stress again that the Chol content in the GUV membranes is significantly lower than the mixing ratio (due to the Chol admixing in the form of anhydrous Chol crystals during the solid phase state [drying] of sample preparation). Once Chol crystals are formed, Chol molecules in Chol crystals do not participate further in the formation of GUVs. Thus, Chol content in membranes is decreased as compared with the mixing ratio. We see the formation of significantly large CBDs only at a mixing ratio of 2.5, which is significantly greater than 1, the actual Chol content in PC membranes at which CBDs start to form.

Fig. 6.

Confocal fluorescence images of GUVs made of Chol/DSPC mixtures, with mixing ratios of 2.5 and 4 and recorded at 22°C. Each vesicle contains both fluorescent probes, DiIC₁₈(3) (red) and Bodipy CHOL (green). In order to confirm that the CBD always forms at the top of the GUV, confocal fluorescence images first were obtained for one sample orientation; then, images were obtained for the same sample, but the orientation was reversed (i.e., turned upside-down). Fluorescence signals originate from Bodipy CHOL for vesicles presented on the left and from DiIC₁₈(3) for vesicles presented on the right. Arrows indicate that for both samples and both orientations, the CBD was always detected at the top of the GUV.

Based on the images presented in Figs. 4, 5, and 6, we can roughly conclude that the process of forming (increasing size) of CBDs ends at the Chol/PC mixing ratio of 3/1. Thus, at this mixing ratio, the actual Chol content in membranes reaches the Chol solubility threshold (which is at a Chol/PC molar ratio of 2/1), and a further increase in the Chol content (Chol/PC mixing ratio) only adds more Chol into the Chol crystals. Why are the observed CBDs so large? We think that a process occurs in GUVs wherein smaller, individual CBDs coalesce. Because of the “gravitational” preference, small CBDs accumulate at the tops of GUVs. It should be energetically preferable to form one large CBD than a few smaller ones. The EPR spin-labeling approach allows the discrimination of CBDs and evaluates the total amount of Chol forming CBDs; however, it tells us little about the size of the individual CBD (1). Some evaluations suggest that CBDs discriminated by EPR spin labeling are rather small (1). It should also be mentioned that EPR measurements were performed on multilamellar liposomes with curvature much greater than that of the GUVs. See Sect. 3.5 for additional explanations.

3.5. CBDs locate at tops of GUVs

According to the results presented in Figs. 4 and 5, CBDs always formed at the top of the GUVs. This behavior was observed for GUVs made of saturated (DSPC) and unsaturated (POPC) PLs. This was an unexpected result. To confirm this observation, we repeated the confocal fluorescent imaging procedure for saturated and unsaturated GUVs after changing the sample orientation (i.e., turning it upside-down) (Fig. 6). Again, the CBDs were detected at the tops of GUVs. We think that GUVs, with their diameter of ~20 μm, can be treated as macroscopic spheres suspended in water. Thus, gravitational forces can affect the orientation of such kinds of “macroscopic” structures floating in buffer.

The presented results indicated that an individual GUV is composed of two completely separate domains, CBD and PCD, which differ significantly in their surface density. The surface density of the CBD can be evaluated as ~336 ng/cm² and the surface density of the PCD as ~438 ng/cm² for saturated POPC-Chol bilayers with a Chol/POPC molar ratio of 1/1 (25). The surface density of the PCD is substantially greater than that of the CBD, meaning that the center of mass of the whole vesicle will shift outside of the vesicle center, to the opposite side of the CBD domain. This should enhance the orientation, with the less dense part of the membrane (i.e., the CBD) at the top of the GUV.

4. Final Conclusions

The confocal microscopy and fluorescent probes used here can clearly show the locations of the Chol and PL molecules, their separation, or their co-localization. All confocal fluorescence images (Figs. 4, 5, and 6) clearly show that when the GUVs are prepared at a Chol/PL mixing ratio greater than 2.5, two domains are present. One domain simultaneously showed both the fluorescence signals from DiIC₁₈(3) (red) and Bodipy CHOL (green). Thus, in this domain, the PL and Chol molecules coexist. In the second domain, the fluorescence signal from DiIC₁₈(3) (red) was not present, but the fluorescence signal from Bodipy CHOL (green) was present. Thus, the DiIC₁₈(3) and other PL molecules are excluded from this domain and Bodipy CHOL and other Chol molecules remain in this domain. In this approach, fluorescent probes located in both membrane leaflets contribute to the recorded fluoresce signals. Thus, all the presented results provide strong evidence that the CBD is formed as a pure transmembrane Chol bilayer, free from any contamination from the PL. This finding complements our studies on the structure of the CBD and the organization of lipids in membranes loaded and overloaded with Chol.