Distribution of Filipin-Sterol Complexes on Cultured Cells: Cell-Substratum Contact Areas Associated with Acetylcholine Receptor Clusters Muscle

Specialized areas wi th in broad, close, cell-substratum contacts seen with reflection interference contrast microscopy in cultures of Xenopus embryonic muscle cells were studied. These areas usually contained a distinct pattern of l ight and dark spots suggesting that the closeness of apposit ion between the membrane and the substratum was irregular. They coincided with areas containing acetylcholine receptor clusters identif ied by fluorescence labeled c~-bungarotoxin. Freeze-fracture of the cells confirmed these observations. The membrane in these areas was highly convoluted and contained aggregates of large P-face intramembrane particles (probably representing acetylcholine receptors). If cells were fixed and then treated with the sterol-specific antibiotic f i l ipin before fracturing, the pattern of f i l ipin-sterol c om pl e x dis t r ibut ion closely fo l lowed the pat tern of ce l l subs t ra tum contact . Filipin-sterol complexes were in low densi ty in the regions where the m e m b r a n e con t a ined clustered i n t r a m e m b r a n e particles. These m e m b r a n e regions were away from the subs t ra tum (bright white areas in reflection interference contrast; depressions of the P-face in freeze-fracture). Filipin-sterol complexes were also in reduced density where the membrane was very close to the substratum (dark areas in reflection interference contrast; bulges of the P-face in freezefracture). These areas were not associated with clustered acetylcholine receptors (aggregated particles). This result suggests that f i l ip in treatment causes litt le or no artefact in either acetylcholine receptor distr ibut ion or membrane topography of fixed cells and that the distr ibut ion of f i l ipin-sterol complexes may closely parallel the microheterogeneity of membranes that exist in l iving cells. In recent years, there has been a substantial increase in the application of cytochemical agents that can be used for detecting the distribution of membrane lipids. In particular, the polyene antibiotic, filipin, has seen increasing use for the detection of membrane cholesterol in both artificial and natural systems. This is because filipin forms a specific complex with membrane 3-fl-hydroxy sterols that is easily seen in freezefracture replicas as a distinct round protuberance or pit (3, 8, 15, 27, 36, 40, 42). This allows a simple method for viewing the regional distribution of membrane cholesterol. The use of this methodology in cell systems, however, is dependent upon several major assumptions: that the cytochemical agent is indeed highly specific in its interaction with membrane lipid, that it penetrates evenly into tissues, that it is not prevented from interacting or forming a complex with the lipid when THE JOURNAL OF CELL BIOLOGY VOLUME 96 FEBRUARY 1983 363-372 © The Rockefel ler Univers i ty Press 0021-9525/83/02/0363/10 $1.00 present, and that there is little or no lateral movement of the lipid after fixation or after the formation of the complex with the cytochemical agent. If these agents are to be useful in describing the microheterogeneity of membrane lipid as it exists in the living cell, these assumptions must prove correct. In some cases, the question of specificity has been extensively studied. This is particularly true of filipin (24, 26, 31), contributing to its growing application. Studies on specificity, however, have been primarily restricted to artificial systems because of the ease of interpreting the results. In natural cell systems recent progress has been made using filipin; for instance, Elias et al. (15) have shown that variable penetration cannot explain the regional variation in the distribution of filipin-sterol complexes in sperm membranes, because barriers to filipin penetration can be eliminated in their prep363 on O cber 0, 2017 jcb.rress.org D ow nladed fom aration. In addition, Friend and Bearer (17) have shown that the number of filipin-sterol complexes formed in Drosophila larval cells is proportional to the amount of sterol incorporated during culturing of the ceils. However, several important problems remain. The influence o f proteins or other lipids on the interaction of cholesterol with filipin is unknown; indeed, the absence of fdipin-sterol complexes from particle aggregates or possible cell adhesion sites suggests a possible influence (3, 8, 15, 28, 36). Aldehyde fixation before or during filipin treatment is critical for the prevention o f a filipin-induced redistribution of both intramembrane particles and filipin-sterol complexes (36), suggesting that the movement o f proteins or the complex may not be a problem as long as fixation proceeds or accompanies treatment. Glutaraldehyde, however, does not prevent movement of lipid within the membrane bilayer (19, 23, 34), even though it can interact with phospholipids containing primary amines (11, 18, 21, 25, 30, 38, 47). Glutaraldehyde apparently does not interact with cholesterol (47) and obviously does not prevent the interaction of cholesterol with filipin. It is not known how far cholesterol will migrate to interact with t iepin or how much freedom of movement the complex has in fixed ceils. Therefore, lipid movement before or after formation of a lipid-specific complex, as well as the inhibition of complex formation by proteins, lipids, or regional restraints on membrane deformation, may be potential problems associated with using cytochemical agents such as filipin for specifically mapping the regional variation in cell membrane lipid. To address some of the problems associated with cytochemical agents such as •ipin, it is important to establish that a correlation does exist between specializations of the membrane observed on living cells and membrane areas defmed by the extent of their reaction with filipin. One way to do this is to observe specializations on a single cell while it is living and then see if the same area is detectable following reaction with filipin and freeze-fracturing. This should allow precise identification of membrane areas that are not effected by filipin and, by assessing how closely the areas correlate, determination o f whether migration o f the complex or lipid is occurring. This may, in turn, allow insight into the problem of whether the lack of filipin-sterol complexes in a region results directly from a lack of sterol or indirectly through influences on the ability o f filipin to form a complex. We chose to do a correlation as described above using a combination of techniques. A close association between specialized areas o f contact between the cell and the glass substrate and acetylcholine receptor clusters has been reported in cultures o f rat myotubes (7). Both parameters may be observed in living ceils by using reflection interference contrast (RIC) and fluorescence microscopy. There is evidence suggesting that both ceil adhesion sites and acetylcholine receptor dusters influence the distribution of filipin-sterol complexes (8, 37). By freeze-fracturing cells that were first observed with fluorescence and RIC microscopy in the living state, we have been able to make precise observations on the influence of these cell surface specializations on the distribution o f filipin-sterol complexes. Our results suggest that the distribution of filipin-sterol complexes is highly influenced by both the presence of acetylcholine receptor clusters and specialized very close contacts with the substratum. Filipin-sterol complexes usually do not form within these areas and do not appear able to migrate into them by lateral movement from surrounding regions. This supports the interpretation that the distribution of filipin-sterol complexes reflects the microheterogeneity of membranes in living 364 THE JOURNAlOE Cl:l.L BIOLOGY • VOLUME 96, 1983 cells. Whether this variation results from actual heterogeneity of membrane sterol, however, remains to be seen. MATERIALS AND METHODS Tissue Culture: Muscle cultures were prepared from stage 17-18 (29) Xenopus laevis embryos as described previously (8, 32). In order to view living cultures with a standard light microscope and then prepare the same cultures for freeze-fracture, we used a simple chamber. A single 4-ram circle was scored in the center of a square coverglass (No. 0; 22 mmx 22 ram). The coverglass was cleaned and placed into a plastic culture dish containing culture medium, with the scored surface facing up. Ceils were plated within the 4-ram circle at low density. To view the cells under the microscope, we removed the coverglass from the culture dish and placed it face down onto a standard glass side containing 12 drops of culture medium. Two thin spacers made from strips of No. I coverglass were glued to the slide surface to provide adequate separation between the glass surfaces. Culture medium was carefully absorbed from the top surface of the coverglass to allow layering of immersion oil. The edges of the chamber were not sealed. Small drops of culture medium were added from time to time to compensate for evaporation from the thin fluid layer. Fluorescence Microscopy: Cultures were labeled with monotetramethylrhodamine-abungarotoxin (R-aBGT) (35) for 30 min at room temperature. The concentration of R-aBGT used for staining was determined by titrating with increasing amounts until a maximum brightness of specific staining was achieved. Cultures were washed with eight rapid changes (~5 ml each) of culture medium before they were viewed. Fluorescence microscopy was performed using a standard Zeiss microscope equipped with epillumination and a 150 W halogen light source. We used a Zeiss planapo 63/1.4 phase contrast oil immersion objective. Phase and fluorescent images were photographed on Kodak Tri X film and were developed in Diafine. Reflection Interference Contrast Microscopy: Reflection interference contrast (interference reflection) microscopy was performed using basically the same system as for fluorescence microscopy (1). The Zeiss Antiflex-neofluar 63/1.25 oil immersion objective, designed specifically for RIC microscopy (5), was used. Photography was performed with Kodak Technical Pan film 2415, which was developed in D-19. FreezeFracture: After cells were photographed in the light microscope, the immersion oil was carefully removed from the coverglass surface. The entire chamber was then placed into a large petri dish and the coverglass was floated off the slide by the slow addition of drops of ftxative (0.5% ghitaraldehyde plus 5 mM CaC12 in 0.05 M Na cacodylate buffer, pH 7.4). Control cultures were fixed for 1.5 h. Cultures to be treated with •ipin (gift of T. E. Grady, The Upjohn Co., Kalamazoo, M[) were initially fixed for 30 min. The solution was then exchanged for a fresh fixative solution containing 0.04% fllipin and 1% DMSO (8). Cultures were treated with this solution for 1 h. After washing and equilibration with a 20% glycerol-buffer solution, the 4-ram disk was broken free from the rest of the coverglass. Freeze-fracture procedures with the double replica device by means of the gold-disk-coverglass sandwich technique have been described previously (32, 33, 48).