Selective anchoring in the specific plasma membrane domain: a role in epithelial cell polarity

The (geometric) mean diffusion coefficients, D, for all labels applied to the apical surface were in the range 1.1-1.3 × 10 -9 cm2/s -l. None differed significantly from any other. Similar D's were measured on the free surface of sparse cells (unpolarized for basolateral antigens). Diffusion coefficients for labels on the basal surface of confluent monolayers were ~50% smaller, in the range 0.6-0.9 x 10 -9 cm2ts -t, but the difference was not statistically significant (Table IV) . The mobile fraction, R, of diffusing molecules was estimated from the extent of recovery of fluorescence after photobleaching. The apparent R were between 38 and 65 %. These values were an underestimate of the recovery of fluorescence, since the fluorescence from labeled cells was 1.5-3.5 times the fluorescence from cell blanks (either unlabeled or treated with only the second Fab). The endogenous fluorescence from cell blanks did bleach but did not recover. Hence, we can make a better approximation to the actual fraction of mobile labels by correcting the apparent R for a significant fraction of fluorescence that was immobile. These corrected values, from experiments in which sufficient (6) (7) (8) (9) (10) (11) (12) blanks were measured to give a reliable estimate, were in the range of 64-84% (Table IV) . They are high but do not reach 90-100%. No difference in R was observed between the apical and the basal surface, for all antigens studied. As shown in a previous work (86), R did not correlate with TX-100 extractability. The overall conclusion from measurements where recovery was observed is compatible with large mobile fractions behaving similarly in the apical or basolateral domains. D and R can only be determined if some recovery is found after photobleaching. In our experiments, we found a large percent of measurements in which no recovery of fluorescence at all could be measured (TIF), even though the ceils appeared to be labeled to the same extent and in the same way (uniform faint fluorescence, rather than the coarse speckles indicating aggregation of label) as cells in which recovery of fluorescence was found. Since 100% of the cells have B1 epitopes, lack of recovery cannot be attributed to unspecific fluorescence in antigen negative ceils. TIF did not correlate with time after labeling (not shown) and were similar at 10°C, therefore, TIF cannot be attributed to endocytosed antigen. Experience with FRAP in fibroblasts indicates Sparse MDCK cells were grown for ",,24 h on glass coverslips. Confluent MDCK cells were grown for the same period on nylon filters coated with native collagen gels separating two chambers. For this condition the probes were added from the medium and there was little or no diffusion to the basal side. The laser beam was focused to the apical domain. For the confluent condition, the probes were added from either the apical or basal chamber. TIF is the percent of measurements (cells) that showed <15% recovery (R). D and R are, respectively, the geometric diffusion coefficient and percent of recovery calculated from measurements with R > 15 %. R is shown as average (corrected for nonrecoverable background) :k SD. The 95 % confidence intervals for D were 0.6 × mean and 2.0 x mean. * TX-100-insoluble fractions are values from Table II (Basal BI and B2-m) and separate video analysis determinations for BI and A2 (apical) extractability on semi-thin frozen sections (not shown), n is the number of measurements (cells).