Cytoplasmic Signals Mediate Apical Early Endosomal Targeting of Endotubin in MDCK Cells

Endotubin is an integral membrane protein that targets into apical endosomes in polarized epithelial cells. Although the role of cytoplasmic targeting signals as mediators of basolateral targeting and endocytosis is well established, it has been suggested that apical targeting requires either N-glycosylation of the ectoplasmic domains or partitioning of macromolecules into glycolipid-rich rafts. However, we have previously shown that the cytoplasmic portion of endotubin possesses signals that are necessary for its proper sorting into the apical early endosomes. To further define the targeting signals involved in this apically directed event, as well as to determine if the cytoplasmic domain was sufficient to mediate apical endosomal targeting, we generated a panel of endotubin and Tac-antigen chimeras and expressed them in Madin–Darby canine kidney cells. We show that both the apically targeting wild-type endotubin and a basolaterally targeted cytoplasmic domain mutant do not associate with rafts and are TX-100 soluble. The cytoplasmic tail of endotubin is sufficient for apical endosomal targeting, as chimeras with the endotubin cytoplasmic domain and Tac transmembrane and extracellular domains are efficiently targeted to the apical endosomal compartment. Furthermore, we show that overexpression of these chimeras results in their missorting to the basolateral membrane, indicating that the apical sorting process is a saturable event. These results show that cells contain machinery in both the biosynthetic and endosomal compartments that recognize cytoplasmic apical sorting signals.     

Definition of distinct compartments in polarized Madin-Darby canine kidney (MDCK) cells for membrane-volume sorting, polarized sorting and apical recycling

Previous studies of fibroblasts have demonstrated that recycling of endocytic receptors occurs through a default mechanism of membrane-volume sorting. Epithelial cells require an additional level of polar membrane sorting, but there are conflicting models of polar sorting, some suggesting that it occurs in early endosomes, others suggesting it occurs in a specialized apical recycling endosome (ARE). The relationship between endocytic sorting to the lysosomal, recycling and transcytotic pathways in polarized cells was addressed by characterizing the endocytic itineraries of LDL, transferrin (Tf) and IgA, respectively, in polarized Madin-Darby canine kidney (MDCK) cells. Quantitative analyses of 3-dimensional images of living and fixed polarized cells demonstrate that endocytic sorting occurs sequentially. Initially internalized into lateral sorting endosomes, Tf and IgA are jointly sorted from LDL into apical and medical recycling endosomes, in a manner consistent with default sorting of membrane from volume. While Tf is recycled to the basolateral membrane from recycling endosomes, IgA is sorted to the ARE prior to apical delivery. Quantifications of the efficiency of sorting of IgA from Tf between the recycling endosomes and the ARE match biochemical measurements of transepithelial protein transport, indicating that all polar sorting occurs in this step. Unlike fibroblasts, rab11 is not associated with Tf recycling compartments in either polarized or glass-grown MDCK cells, rather it is associated with the compartments to which IgA is directed after sorting from Tf. These results complicate a suggested homology between the ARE and the fibroblast perinuclear recycling compartment and provide a framework that justifies previous conflicting models of polarized sorting.     

N-Glycans mediate the apical sorting of a GPI-anchored, raft-associated protein in Madin-Darby canine kidney cells.

Glycosyl-phosphatidylinositol (GPI)- anchored proteins are preferentially transported to the apical cell surface of polarized Madin-Darby canine kidney (MDCK) cells. It has been assumed that the GPI anchor itself acts as an apical determinant by its interaction with sphingolipid-cholesterol rafts. We modified the rat growth hormone (rGH), an unglycosylated, unpolarized secreted protein, into a GPI-anchored protein and analyzed its surface delivery in polarized MDCK cells. The addition of a GPI anchor to rGH did not lead to an increase in apical delivery of the protein. However, addition of N-glycans to GPI-anchored rGH resulted in predominant apical delivery, suggesting that N-glycans act as apical sorting signals on GPI-anchored proteins as they do on transmembrane and secretory proteins. In contrast to the GPI-anchored rGH, a transmembrane form of rGH which was not raft-associated accumulated intracellularly. Addition of N-glycans to this chimeric protein prevented intracellular accumulation and led to apical delivery.     

Direct apical sorting of rat liver dipeptidylpeptidase IV expressed in Madin-Darby canine kidney cells.

In Madin-Darby canine kidney (MDCK) cells, apical and basolateral membrane proteins are segregated from each other in the trans-Golgi network (TGN) and are transported to the appropriate membrane domain via separate vesicle populations. In hepatocytes, however, all plasma membrane proteins are delivered basolaterally. Apical proteins are then selectively retrieved and reach the apical surface by transcytosis. The sorting of apical proteins in different cell types may be the result of differences in the cellular sorting machinery, or alternatively, due to expression of cell-specific sorting signals on the proteins themselves. To test this directly, we have stably expressed cDNA encoding an apical protein from rat liver, dipeptidylpeptidase IV (DPPIV), in MDCK cells. We found that approximately 90% of the exogenous DPPIV is expressed on the apical cell surface at steady state. Furthermore, we demonstrate that this distribution is primarily due to vectorial transport from the TGN to the apical plasma membrane. The small pool of mis-sorted DPPIV that appears basolaterally is slowly endocytosed (t1/2 approximately 60 min) and is subsequently transcytosed. These data are consistent with the notion that both hepatocytes and MDCK cells are capable of correctly sorting rat liver DPPIV, but that this sorting occurs at different sites in the two cell types.     

RhoB-dependent modulation of postendocytic traffic in polarized Madin-Darby canine kidney cells

The Rho family of GTPases is implicated in the control of endocytic and biosynthetic traffic of many cell types; however, the cellular distribution of RhoB remains controversial and its function is not well understood. Using confocal microscopy, we found that endogenous RhoB and green fluorescent protein-tagged wild-type RhoB were localized to early endosomes, and to a much lesser extent to recycling endosomes, late endosomes or Golgi complex of fixed or live polarized Madin-Darby canine kidney cells. Consistent with RhoB localization to early endosomes, we observed that expression of dominant-negative RhoBN19 or dominant-active RhoBV14 altered postendocytic traffic of ligand-receptor complexes that undergo recycling, degradation or transcytosis. In vitro assays established that RhoB modulated the basolateral-to-apical transcytotic pathway by regulating cargo exit from basolateral early endosomes. Our results indicate that RhoB is localized, in part, to early endosomes where it regulates receptor egress through the early endocytic system.     

Mucin-like domain of enteropeptidase directs apical targeting in Madin-Darby canine kidney cells

Enteropeptidase, a type II transmembrane protein of the enterocyte brush border, is sorted directly to the apical membrane of Madin-Darby canine kidney II cells. Apical targeting appears to be mediated by an N-terminal segment that contains a 27-amino acid residue O-glycosylated mucin-like domain consisting of two short mucin-like repeats, A and B. Targeting signals within these repeats were characterized by using green fluorescent protein (GFP) as a reporter. Constructs with a cleavable signal peptide and both repeats A and B were secreted apically. Similar constructs lacking mucin repeats were secreted randomly. Either repeat A or B was sufficient to direct apical targeting of GFP. O-linked oligosaccharides alone were not sufficient for targeting because fusion to a different O-glycosylated motif did not alter the random secretion of GFP, and several constructs with mutations in either repeat A or B were O-glycosylated and secreted randomly. In addition, repeat B appears to contain an apical targeting signal that functions in the absence of glycosylation. Density gradient centrifugation indicated that, unlike several other apically targeted membrane and soluble proteins, apical sorting of mucin-GFP chimeric proteins does not appear to utilize lipid rafts.     

Newly synthesized and recycling pools of the apical protein gp135 do not occupy the same compartments

Polarized epithelial cells sort newly synthesized and recycling plasma membrane proteins into distinct trafficking pathways directed to either the apical or basolateral membrane domains. While the trans‐Golgi network is a well‐established site of protein sorting, increasing evidence indicates a key role for endosomes in the initial trafficking of newly synthesized proteins. Both basolateral and apical proteins have been shown to traverse endosomes en route to the plasma membrane. In particular, apical proteins traffic through either subapical early or recycling endosomes. Here we use the SNAP tag system to analyze the trafficking of the apical protein gp135, also known as podocalyxin. We show that newly synthesized gp135 traverses the apical recycling endosome, but not the apical early endosomes (AEEs). In contrast, post‐endocytic gp135 is delivered to the AEE before recycling back to the apical membrane. The pathways pursued by the newly synthesized and recycling gp135 populations do not detectably intersect, demonstrating that the biosynthetic and post‐endocytic pools of this protein are subjected to distinct sorting processes.      

Induction of caveolae in the apical plasma membrane of Madin-Darby canine kidney cells

In this paper, we have analyzed the behavior of antibody cross-linked raft-associated proteins on the surface of MDCK cells. We observed that cross-linking of membrane proteins gave different results depending on whether cross-linking occurred on the apical or basolateral plasma membrane. Whereas antibody cross-linking induced the formation of large clusters on the basolateral membrane, resembling those observed on the surface of fibroblasts (Harder, T., P. Scheiffele, P. Verkade, and K. Simons. 1998. J. Cell Biol. 929-942), only small ( approximately 100 nm) clusters formed on the apical plasma membrane. Cross-linked apical raft proteins e.g., GPI-anchored placental alkaline phosphatase (PLAP), influenza hemagglutinin, and gp114 coclustered and were internalized slowly ( approximately 10% after 60 min). Endocytosis occurred through surface invaginations that corresponded in size to caveolae and were labeled with caveolin-1 antibodies. Upon cholesterol depletion the internalization of PLAP was completely inhibited. In contrast, when a non-raft protein, the mutant LDL receptor LDLR-CT22, was cross-linked, it was excluded from the clusters of raft proteins and was rapidly internalized via clathrin-coated pits.Since caveolae are normally present on the basolateral membrane but lacking from the apical side, our data demonstrate that antibody cross-linking induced the formation of caveolae, which slowly internalized cross-linked clusters of raft-associated proteins.     

Rab10 regulates membrane transport through early endosomes of polarized Madin-Darby canine kidney cells

Rab10, a protein originally isolated from Madin-Darby Canine Kidney (MDCK) epithelial cells, belongs to a family of Rab proteins that includes Rab8 and Rab13. Although both Rab8 and Rab13 have been found to mediate polarized membrane transport, the function of Rab10 in mammalian cells has not yet been established. We have used quantitative confocal microscopy of polarized MDCK cells expressing GFP chimeras of wild-type and mutant forms of Rab10 to analyze the function of Rab10 in polarized cells. These studies demonstrate that Rab10 is specifically associated with the common endosomes of MDCK cells, accessible to endocytic probes internalized from either the apical or basolateral plasma membrane domains. Expression of mutant Rab10 defective for either GTP hydrolysis or GTP binding increased recycling from early compartments on the basolateral endocytic pathway without affecting recycling from later compartments or the apical recycling pathway. These results suggest that Rab10 mediates transport from basolateral sorting endosomes to common endosomes.     

rab4 regulates transport to the apical plasma membrane in Madin-Darby canine kidney cells.

The small GTPase rab4 is associated with early endosomes and regulates membrane recycling in fibroblasts. rab4 is present in epithelial cells; however, neither its localization nor function has been established in this cell type. We transfected Madin-Darby canine kidney cells with rab4, the GTPase-deficient mutant rab4Q67L, and the dominant negative mutant rab4S22N that poorly binds guanine nucleotides. Confocal immunofluorescence microscopy showed that rab4 was concentrated on internal structures at the lateral side of the cell around the nucleus. Quantitative immunoelectron microscopy revealed that the majority of rab4 was localized in the upper third of the cytoplasm. In cell surface binding experiments with (125)I-transferrin, we found a redistribution of transferrin receptor from the basolateral to the apical plasma membrane in cells expressing rab4 and rab4Q67L. After accumulation of transferrin at 16 degrees C in basolateral early endosomes, rab4 and rab4Q67L increased the amount of apically targeted transferrin receptor. A qualitatively similar effect was obtained in control cells treated with brefeldin A. The effects of brefeldin A and rab4 on apical targeting of transferrin receptor were not additive, suggesting that brefeldin A and rab4 may act in the same transport pathway from common endosomes.     

Intracellular transport of influenza virus hemagglutinin to the apical surface of Madin-Darby canine kidney cells

The intracellular pathway followed by the influenza virus hemagglutinin (HA) to the apical surface of Madin-Darby canine kidney cells was studied by radioimmunoassay, immunofluorescence, and immunoelectron microscopy. To synchronize the migration, we used a temperature- sensitive mutant of influenza WSN, ts61, which, at the nonpermissive temperature, 39.5 degrees C, exhibits a defect in the HA that prevents its exit from the endoplasmic reticulum. Upon transfer to permissive temperature, 32 degrees C, the HA appeared in the Golgi apparatus after 10 min, and on the apical surface after 30-40 min. In the presence of cycloheximide, the expression was not inhibited, indicating that the ts defect is reversible; a wave of HA migrated to the cell surface, where it accumulated with a half time of 60 min. After passage through the Golgi apparatus the HA was detected in a population of smooth vesicles, about twice the size of coated vesicles, located in the apical half of the cytoplasm. These HA-containing vesicles did not react with anti- clathrin antibodies. Monensin (10 microM) delayed the surface appearance of HA by 2 h, but not the transport to the Golgi apparatus. Incubation at 20 degrees C retarded the migration to the Golgi apparatus by approximately 30 min and blocked the surface appearance by acting at a late stage in the intracellular pathway, presumably at the level of the post-Golgi vesicles. The initial appearance of HA on the apical surface was in the center; no preference was observed for the tight-junctional regions.     

Intracellular sorting and basolateral appearance of the G protein of vesicular stomatitis virus in Madin-Darby canine kidney cells

The polarity of the surface distribution of viral glycoproteins during virus infection has been studied in the Madin-Darby canine kidney epithelial cell line on nitrocellulose filters. Using a surface radioimmunoassay on Madin-Darby canine kidney strain I cells that had been infected with vesicular stomatitis virus or with avian influenza fowl plague virus, we found that the surface G protein was 97% basolateral, whereas the fowl plague virus hemagglutinin was 88% apical. Newly synthesized, pulse-labeled vesicular stomatitis virus appeared first on the basolateral plasma membrane as measured by an immunoprecipitation assay in which the anti-G protein antibody was applied to the monolayer either from the apical or the basolateral side. Labeled G protein could be accumulated inside the cell at a late stage of transport by decreasing the temperature to 20 degrees C during the chase. Reversal to 37 degrees C led to its rapid and synchronous transport to the basolateral surface at an initial rate 61-fold greater than that of transport to the apical side. These results demonstrate that the newly synthesized G protein is transported directly to the basolateral membrane and does not pass over the apical membrane en route. Since a previous study of the surface appearance of influenza virus hemagglutinins showed that the newly synthesized hemagglutinins were inserted directly from an intracellular site into the apical membrane (Matlin, K., and K. Simons, 1984, J. Cell Biol., 99:2131-2139), we conclude that the divergence of the transport pathway for the apical and basolateral viral glycoproteins has to occur intracellularly, i.e., before reaching the cell surface.     

Inhibition by brefeldin A of protein secretion from the apical cell surface of Madin-Darby canine kidney cells

The effect of brefeldin A (BFA) on total and polarized protein secretion was examined in MDCK cells. Increasing concentrations of BFA have increasingly inhibitory effects on total protein secretion. The total protein secretion was essentially unaffected by BFA at 0.5 microgram/ml. When the BFA concentration was increased to 10 and 30 micrograms/ml, the total protein secretion was reduced to about 70 and 25%, respectively, of the control level. Consistent with this effect on total protein secretion, the Golgi structure as revealed by C6-NBD-ceramide (a fluorescent ceramide analog) staining was essentially unaltered by 0.5 microgram/ml BFA, while 10 and 30 micrograms/ml BFA significantly dispersed the Golgi apparatus. When the polarity of protein secretion was examined, it was found that the ratio of proteins secreted from the apical to those from the basolateral surface was reduced from 1.5-2.0 to 0.4-0.7 by all three BFA concentrations. Furthermore, several proteins which are preferentially released from the apical surface were found to be released without apparent surface polarity, while several other proteins which were preferentially released from the basolateral surface were unaffected. This study suggests that BFA, at 0.5 microgram/ml, can selectively inhibit protein secretion from the apical surface without affecting total protein secretion. The inhibition of apical secretion results in enhanced protein secretion from the basolateral surface.     

MAL regulates clathrin-mediated endocytosis at the apical surface of Madin-Darby canine kidney cells

MAL is an integral protein component of the machinery for apical transport in epithelial Madin-Darby canine kidney (MDCK) cells. To maintain its distribution, MAL cycles continuously between the plasma membrane and the Golgi complex. The clathrin-mediated route for apical internalization is known to differ from that at the basolateral surface. Herein, we report that MAL depends on the clathrin pathway for apical internalization. Apically internalized polymeric Ig receptor (pIgR), which uses clathrin for endocytosis, colocalized with internalized MAL in the same apical vesicles. Time-lapse confocal microscopic analysis revealed cotransport of pIgR and MAL in the same endocytic structures. Immunoelectron microscopic analysis evidenced colabeling of MAL with apically labeled pIgR in pits and clathrin-coated vesicles. Apical internalization of pIgR was abrogated in cells with reduced levels of MAL, whereas this did not occur either with its basolateral entry or the apical internalization of glycosylphosphatidylinositol-anchored proteins, which does not involve clathrin. Therefore, MAL is critical for efficient clathrin-mediated endocytosis at the apical surface in MDCK cells.     

The receptor recycling pathway contains two distinct populations of early endosomes with different sorting functions

Receptor recycling involves two endosome populations, peripheral early endosomes and perinuclear recycling endosomes. In polarized epithelial cells, either or both populations must be able to sort apical from basolateral proteins, returning each to its appropriate plasma membrane domain. However, neither the roles of early versus recycling endosomes in polarity nor their relationship to each other has been quantitatively evaluated. Using a combined morphological, biochemical, and kinetic approach, we found these two endosome populations to represent physically and functionally distinct compartments. Early and recycling endosomes were resolved on Optiprep gradients and shown to be differentially associated with rab4, rab11, and transferrin receptor; rab4 was enriched on early endosomes and at least partially depleted from recycling endosomes, with the opposite being true for rab11 and transferrin receptor. The two populations were also pharmacologically distinct, with AlF4 selectively blocking export of transferrin receptor from recycling endosomes to the basolateral plasma membrane. We applied these observations to a detailed kinetic analysis of transferrin and dimeric IgA recycling and transcytosis. The data from these experiments permitted the construction of a testable, mathematical model which enabled a dissection of the roles of early and recycling endosomes in polarized receptor transport. Contrary to expectations, the majority (>65%) of recycling to the basolateral surface is likely to occur from early endosomes, but with relatively little sorting of apical from basolateral proteins. Instead, more complete segregation of basolateral receptors from receptors intended for transcytosis occurred upon delivery to recycling endosomes.     

Protein core-dependent glycosaminoglycan modification and glycosaminoglycan-dependent polarized sorting in epithelial Madin-Darby canine kidney cells

The proteoglycan serglycin (SG) fused to green fluorescent protein (GFP) is secreted predominantly from the apical surface of polarized epithelial Madin-Darby canine kidney (MDCK) cell monolayers, but the minor fraction secreted basolaterally carries more intensely sulfated glycosaminoglycan (GAG) chains (Tveit H, Dick G, Skibeli V, Prydz K. 2005. A proteoglycan undergoes different modifications en route to the apical and basolateral surfaces of Madin-Darby canine kidney cells. J Biol Chem 280: 29596-29603). To investigate whether the domain with GAG attachment sites in SG (i) is sufficient to drive apical protein sorting and (ii) independently generates the sulfation differences observed in the apical and basolateral pathways, the GAG domain of SG was fused into the junction of rat growth hormone (rGH) and GFP and expressed in MDCK cells, either with or without two N-glycosylation sites in the rGH part. Both variants acquired chondroitin sulfate GAG chains and were secreted predominantly to the apical medium, to the same extent as rGH-GFP with two N-glycosylation sites only, and different from the nonsorted variant lacking glycosylation sites. Transfer of the GAG attachment domain from SG to the new rGH context abolished the differences in sulfation intensity and positions observed for SG in the apical and basolateral secretory routes. Thus, these differences are coded by elements outside the GAG attachment domain.     

Polarized apical sorting of guanylyl cyclase C is specified by a cytosolic signal

Receptor guanylyl cyclases respond to ligand stimulation by increasing intracellular cGMP, thereby initiating a variety of cell-signaling pathways. Furthermore, these proteins are differentially localized at the apical and basolateral membranes of epithelial cells. We have identified a region of 11 amino acids in the cytosolic COOH terminus of guanylyl cyclase C (GCC) required for normal apical localization in Madin-Darby canine kidney (MDCK) cells. These amino acids share no significant sequence homology with previously identified cytosolic apical sorting determinants. However, these amino acids are highly conserved and are sufficient to confer apical polarity to the interleukin-2 receptor alpha-chain (Tac). Additionally, we find two molecular weight species of GCC in lysates prepared from MDCK cells over-expressing GCC but observe only the fully mature species on the cell surface. Using pulse-chase analysis in polarized MDCK cells, we followed the generation of this mature species over time finding it to be detectable only at the apical cell surface. These data support the hypothesis that selective apical sorting can be determined using short, cytosolic amino acid motifs and argue for the existence of apical sorting machinery comparable with the machinery identified for basolateral protein traffic.     

The MAL proteolipid is necessary for normal apical transport and accurate sorting of the influenza virus hemagglutinin in Madin-Darby canine kidney cells

The MAL (MAL/VIP17) proteolipid is a nonglycosylated integral membrane protein expressed in a restricted pattern of cell types, including T lymphocytes, myelin-forming cells, and polarized epithelial cells. Transport of the influenza virus hemagglutinin (HA) to the apical surface of epithelial Madin-Darby canine kidney (MDCK) cells appears to be mediated by a pathway involving glycolipid- and cholesterol- enriched membranes (GEMs). In MDCK cells, MAL has been proposed previously as being an element of the protein machinery for the GEM-dependent apical transport pathway. Using an antisense oligonucleotide-based strategy and a newly generated monoclonal antibody to canine MAL, herein we have approached the effect of MAL depletion on HA transport in MDCK cells. We have found that MAL depletion diminishes the presence of HA in GEMs, reduces the rate of HA transport to the cell surface, inhibits the delivery of HA to the apical surface, and produces partial missorting of HA to the basolateral membrane. These effects were corrected by ectopic expression of MAL in MDCK cells whose endogenous MAL protein was depleted. Our results indicate that MAL is necessary for both normal apical transport and accurate sorting of HA.     

Postendocytotic sorting of the ligand for the polymeric immunoglobulin receptor in Madin-Darby canine kidney cells

The polymeric immunoglobulin receptor (pIg-R) is responsible for the receptor-mediated transcytosis of polymeric immunoglobulins (IgA and IgM) across various epithelia. We have expressed the cDNA for the pIg-R in Madin-Darby canine kidney (MDCK) cells and found that this system mimics that found in vivo (Mostov, K. E., and D. L. Deitcher. 1986. Cell. 46:613-621). We have now investigated the postendocytotic pathway of the ligand for the pIg-R. After a 5-min internalization at the basolateral surface, approximately 45% of internalized ligand recycles to the basolateral medium and 30% is transcytosed to the apical medium. We have also examined why transcytosis of ligand is unidirectional, going only from basolateral to apical, but not from apical to basolateral. Several factors could explain this, such as proteolytic cleavage of the pIg-R at the apical surface, decreased apical endocytosis of ligand, or an intracellular sorting event. In this report, we show that the protease inhibitor, leupeptin, inhibits the cleavage of the pIg-R but does not alter the unidirectionality of transcytosis. In addition, we demonstrate that there is a significant amount of apical endocytosis of ligand (70% of that observed basolaterally). Finally, we demonstrate that apically endocytosed ligand can return only to the apical surface. Thus, once ligand reaches the apical surface, it is "trapped" and cannot return to the basolateral surface. We propose that the unidirectionality of transcytosis is the result of intracellular sorting, and that this results from a signal(s) present on the pIg-R.     

Formation of the apical pole of epithelial (Madin-Darby canine kidney) cells: polarity of an apical protein is independent of tight junctions while segregation of a basolateral marker requires cell-cell interactions

The time course of development of polarity of an apical (184-kD) and a basolateral (63-kD) plasma membrane protein of Madin-Darby canine kidney cells was followed using semiquantitative immunofluorescence on semithin (approximately 0.5-micron) frozen sections and monoclonal antibody probes. The 184-kD protein became highly polarized to the apical pole within the initial 24 h both in normal medium and in 1-5 microM Ca2+, which results in well-spread, dome-shaped cells, lacking tight junctions and other lateral membrane interactions. In contrast, the basolateral 63-kD membrane protein developed full polarity only after incubation in normal Ca2+ concentrations for greater than 72 h, a time much longer than that required for the formation of tight junctions (approximately 18 h) and failed to polarize in 1-5 microM Ca2+. These results demonstrate that intradomain restriction mechanisms independent of tight junctions, such as self-aggregation or specific interactions with the submembrane cytoskeleton, participate in the regionalization of at least some epithelial plasma membrane proteins. The full operation of these mechanisms depends on the presence of normal cell-cell interactions in the case of the basolateral 63-kD antigen but not in the case of the apical 184-kD protein.