Exogenous MAL reroutes selected hepatic apical proteins into the direct pathway in WIF-B cells

Unlike simple epithelial cells that directly target newly synthesized glycophosphatidylinositol (GPI)-anchored and single transmembrane domain (TMD) proteins from the trans-Golgi network to the apical membrane, hepatocytes use an indirect pathway: proteins are delivered to the basolateral domain and then selectively internalized and transcytosed to the apical plasma membrane. Myelin and lymphocyte protein (MAL) and MAL2 have been identified as regulators of direct and indirect apical delivery, respectively. Hepatocytes lack endogenous MAL consistent with the absence of direct apical targeting. Does MAL expression reroute hepatic apical residents into the direct pathway? We found that MAL expression in WIF-B cells induced the formation of cholesterol and glycosphingolipid-enriched Golgi domains that contained GPI-anchored and single TMD apical proteins; polymeric IgA receptor (pIgA-R), polytopic apical, and basolateral resident distributions were excluded. Basolateral delivery of newly synthesized apical residents was decreased in MAL-expressing cells concomitant with increased apical delivery; pIgA-R and basolateral resident delivery was unchanged. These data suggest that MAL rerouted selected hepatic apical proteins into the direct pathway.     

Rat liver dipeptidylpeptidase IV contains competing apical and basolateral targeting information.

Madin-Darby canine kidney (MDCK) cells deliver endogenous apical and basolateral proteins directly to the appropriate domains. We are investigating the molecular signals on a model plasma membrane hydrolase, dipeptidylpeptidase IV (DPPIV). Most newly synthesized rat liver DPPIV is delivered directly to the apical surface of transfected MDCK cells; however, about 20% is delivered first to the basolateral surface and reaches the apical surface via transcytosis (Casanova, J. E., Mishumi, Y., Ikehara, Y., Hubbard, A. L., and Mostov, K. E. (1991) J. Biol. Chem. 266, 24428-24432). A soluble form of DPPIV (solDPPIV) containing only the lumenal domain of the protein was efficiently transported and secreted by stably transfected MDCK cells. If this domain contains apical sorting information, we would expect 80% of the soluble protein to be secreted apically. Surprisingly, 95% of the secreted solDPPIV was found in the apical medium. The high efficiency of apical secretion suggested that the transmembrane domain and cytoplasmic tail of DPPIV might contain competing basolateral targeting information. To test this hypothesis, we investigated the trafficking of a chimera in which the cytoplasmic tail and transmembrane domains of DPPIV were joined to lysozyme, an exogenous protein which should not contain sorting information. This protein was delivered predominantly to the basolateral surface. Our results suggest that the lumenal domain of DPPIV carries dominant apical sorting information while the transmembrane domain and cytoplasmic tail of the molecule contains competing basolateral sorting information.     

5''nucleotidase is sorted to the apical domain of hepatocytes via an indirect route [published erratum appears in J Cell Biol 1993 Nov;123(3):following 767]

In hepatocytes, all newly synthesized plasma membrane (PM) proteins so far studied arrive first at the basolateral domain; apically destined proteins are subsequently endocytosed and sorted to the apical domain via transcytosis. A mechanism for the sorting of newly synthesized glycophosphatidylinositol (GPI)-linked proteins has been proposed whereby they associate in lipid microdomains in the trans-Golgi network and then arrive at the apical domain directly. Such a mechanism poses a potential exception to the hepatocyte rule. We have used pulse-chase techniques in conjunction with subcellular fractionation to compare the trafficking of 5' nucleotidase (5NT), an endogenous GPI-anchored protein of hepatocytes, with two transmembrane proteins. Using a one-step fractionation technique to separate a highly enriched fraction of Golgi-derived membranes from ER and PM, we find that both 5NT and the polymeric IgA receptor (pIgAR) traverse the ER and Golgi apparatus with high efficiency. Using a method that resolves PM vesicles derived from the apical and basolateral domains, we find that 5NT first appears at the basolateral domain as early as 30 min of chase. However the subsequent redistribution to the apical domain requires > 3.5 h of chase to reach steady state. This rate of transcytosis is much slower than that observed for dipeptidylpeptidase IV, an apical protein anchored via a single transmembrane domain. We propose that the slow rate of transcytosis is related to the fact that GPI-linked proteins are excluded from clathrin-coated pits/vesicles, and instead must be endocytosed via a slower nonclathrin pathway.     

Hierarchy of sorting signals in chimeras of intestinal lactase-phlorizin hydrolase and the influenza virus hemagglutinin

Lactase-phlorizin hydrolase (LPH) is an apical protein in intestinal cells. The location of sorting signals in LPH was investigated by preparing a series of mutants that lacked the LPH cytoplasmic domain or had the cytoplasmic domain of LPH replaced by sequences that comprised basolateral targeting signals and overlapping internalization signals of various potency. These signals are mutants of the cytoplasmic domain of the influenza hemagglutinin (HA), which have been shown to be dominant in targeting HA to the basolateral membrane. The LPH-HA chimeras were expressed in Madin-Darby canine kidney (MDCK) and colon carcinoma (Caco-2) cells, and their transport to the cell surface was analyzed. All of the LPH mutants were targeted correctly to the apical membrane. Furthermore, the LPH-HA chimeras were internalized, indicating that the HA tails were available to interact with the cytoplasmic components of clathrin-coated pits. The introduction of a strong basolateral sorting signal into LPH was not sufficient to override the strong apical signals of the LPH external domain or transmembrane domains. These results show that basolateral sorting signals are not always dominant over apical sorting signals in proteins that contain each and suggest that sorting of basolateral from apical proteins occurs within a common compartment where competition for sorting signals can occur.     

Detergent insoluble microdomains are not involved in transcytosis of polymeric Ig receptor in FRT and MDCK cells

In polarized epithelial cells, sorting of proteins and lipids to the apical or basolateral domain of the plasma membrane can occur via direct or indirect (transcytotic) pathways from the trans Golgi network (TGN). The 'rafts' hypothesis postulates that the key event for direct apical sorting of some transmembrane proteins and the majority of GPI-anchored proteins depends on their association with glycosphingolipid and cholesterol enriched microdomains (rafts). However, the mechanism of indirect sorting to the apical membrane is not clear. The polyimmunoglobulin receptor (pIgR) is one of the best studied proteins that follow the transcytotic pathway. It is normally delivered from the TGN to the basolateral surface of polarized Madin–Darby Canine Kidney (MDCK) cells from where it transports dIgA or dIgM to the apical surface. We have studied the intracellular trafficking of pIgR in Fischer rat thyroid cells (FRT), and have investigated the sorting machinery involved in transcytosis of this receptor in both FRT and MDCK cells. We found that, in contrast with MDCK cells, a significant amount (∼30%) of pIgR reaches the apical surface by a direct pathway. Furthermore, in both cell lines it does not associate with Triton X-100-insoluble microdomains, suggesting that at least in these cells 'rafts' are not involved in basolateral to apical transcytosis.     

Transmembrane domain of influenza virus neuraminidase, a type II protein, possesses an apical sorting signal in polarized MDCK cells.

The influenza virus neuraminidase (NA), a type II transmembrane protein, is directly transported to the apical plasma membrane in polarized MDCK cells. By using deletion mutants and chimeric constructs of influenza virus NA with the human transferrin receptor, a type II basolateral transmembrane protein, we investigated the location of the apical sorting signal of influenza virus NA. When these mutant and chimeric proteins were expressed in stably transfected polarized MDCK cells, the transmembrane domain of NA, and not the cytoplasmic tail, provided a determinant for apical targeting in polarized MDCK cells and this transmembrane signal was sufficient for sorting and transport of the ectodomain of a reporter protein (transferrin receptor) directly to the apical plasma membrane of polarized MDCK cells. In addition, by using differential detergent extraction, we demonstrated that influenza virus NA and the chimeras which were transported to the apical plasma membrane also became insoluble in Triton X-100 but soluble in octylglucoside after extraction from MDCK cells during exocytic transport. These data indicate that the transmembrane domain of NA provides the determinant(s) both for apical transport and for association with Triton X-100-insoluble lipids.     

Mutations in the Middle of the Transmembrane Domain Reverse the Polarity of Transport of the Influenza Virus Hemagglutinin in MDCK Epithelial Cells

The composition of the plasma membrane domains of epithelial cells is maintained by biosynthetic pathways that can sort both proteins and lipids into transport vesicles destined for either the apical or basolateral surface. In MDCK cells, the influenza virus hemagglutinin is sorted in the trans-Golgi network into detergent-insoluble, glycosphingolipid-enriched membrane domains that are proposed to be necessary for sorting hemagglutinin to the apical cell surface. Site- directed mutagenesis of the hemagglutinin transmembrane domain was used to test this proposal. The region of the transmembrane domain required for apical transport included the residues most conserved among hemagglutinin subtypes. Several mutants were found to enter detergent-insoluble membranes but were not properly sorted. Replacement of transmembrane residues 520 and 521 with alanines converted the 2A520 mutant hemagglutinin into a basolateral protein. Depleting cell cholesterol reduced the ability of wild-type hemagglutinin to partition into detergent-insoluble membranes but had no effect on apical or basolateral sorting. In contrast, cholesterol depletion allowed random transport of the 2A520 mutant. The mutant appeared to lack sorting information but was prevented from reaching the apical surface when detergent-insoluble membranes were present. Apical sorting of hemagglutinin may require binding of either protein or lipids at the middle of the transmembrane domain and this normally occurs in detergent-insoluble membrane domains. Entry into these domains appears necessary, but not sufficient, for apical sorting.     

Aminopeptidase N is directly sorted to the apical domain in MDCK cells

In different epithelial cell types, integral membrane proteins appear to follow different sorting pathways to the apical surface. In hepatocytes, several apical proteins were shown to be transported there indirectly via the basolateral membrane, whereas in MDCK cells a direct sorting pathway from the trans-Golgi-network to the apical membrane has been demonstrated. However, different proteins had been studied in these cells. To compare the sorting of a single protein in both systems, we have expressed aminopeptidase N, which already had been shown to be sorted indirectly in hepatocytes, in transfected MDCK cells. As expected, it was predominantly localized to the apical domain of the plasma membrane. By monitoring the appearance of newly synthesized aminopeptidase N at the apical and basolateral surface, it was found to be directly sorted to the apical domain in MDCK cells, indicating that the sorting pathways are indeed cell type-specific.     

Identification of a novel mono-leucine basolateral sorting motif within the cytoplasmic domain of amphiregulin

Epithelial cells establish apical and basolateral (BL) membranes with distinct protein and lipid compositions. To achieve this spatial asymmetry, the cell utilizes a variety of mechanisms for differential sorting, delivery and retention of cell surface proteins. The EGF receptor (EGFR) and its ligand, amphiregulin (AREG), are transmembrane proteins delivered to the BL membrane in polarized epithelial cells. Herein, we show that the cytoplasmic domain of AREG (ACD) contains dominant BL sorting information; replacement of the cytoplasmic domain of apically targeted nerve growth factor receptor with the ACD redirects the chimera to the BL surface. Using sequential truncations and site-directed mutagenesis of the ACD, we identify a novel BL sorting motif consisting of a single leucine C-terminal to an acidic cluster (EEXXXL). In adaptor protein (AP)-1B-deficient cells, newly synthesized AREG is initially delivered to the BL surface as in AP-1B-expressing cells. However, in these AP-1B-deficient cells, recycling of AREG back to the BL surface is compromised, leading to its appearance at the apical surface. These results show that recycling, but not delivery, of AREG to the BL surface is AP-1B dependent.     

Apical and basolateral endosomes of MDCK cells are interconnected and contain a polarized sorting mechanism

We have evaluated transcytotic routes in MDCK cells for their ability to generate a polarized surface distribution of trafficking proteins by following the intracellular sorting of transferrin receptors (TRs). We find that the selective basolateral expression of TRs is maintained in the face of extensive trafficking between the apical and basolateral surfaces. Biochemical studies of receptors loaded with tracer under conditions approaching steady state indicate that TRs internalized from the two surfaces are extensively colocalized within MDCK cells and that both populations of receptors are selectively delivered to the basolateral surface. Tailless TRs in which the cytoplasmic domain has been deleted display an unpolarized cell surface distribution and recycle in an unpolarized fashion. We show by EM that wild-type receptors internalized from each surface are colocalized within endosomal elements distributed throughout the cytoplasm. By preloading endosomal elements directly accessible from the basolateral surface with transferrin (Tf)-HRP, we show that apically internalized TRs rapidly enter the same compartment. We also show that both transcytosing (apically internalized) and recycling (basolaterally internalized) TRs are delivered to the basolateral border by a distinctive subset of exocytotic, 60-nm-diam vesicles. Together, the biochemical and morphological data show that apical and basolateral endosomes of MDCK cells are interconnected and contain a signal- dependent polarized sorting mechanism. We propose a dynamic model of polarized sorting in MDCK cells in which a single endosome-based, signal-dependent sorting step is sufficient to maintain the polarized phenotype.     

Basolateral sorting signals differ in their ability to redirect apical proteins to the basolateral cell surface

Polarized sorting of membrane proteins in epithelial cells is mediated by cytoplasmic basolateral signals or by apical signals in the transmembrane or exoplasmic domains. Basolateral signals were generally found to be dominant over apical determinants. We have generated chimeric proteins with the cytoplasmic domain of either the asialoglycoprotein receptor H1 or the transferrin receptor, two basolateral proteins, fused to the transmembrane and exoplasmic segments of aminopeptidase N, an apical protein, and analyzed them in Madin-Darby canine kidney cells. Whereas both cytoplasmic sequences induced endocytosis of the chimeras, only that of the transferrin receptor mediated basolateral expression in steady state. The H1 fusion protein, although still largely sorted to the basolateral side in biosynthetic surface transport, was subsequently resorted to the apical cell surface. We tested whether the difference in sorting between trimeric wild-type H1 and the dimeric aminopeptidase chimera was caused by the number of sorting signals presented in the oligomers. Consistent with this hypothesis, the H1 signal was fully functional in a tetrameric fusion protein with the transmembrane and exoplasmic domains of influenza neuraminidase. The results suggest that basolateral signals per se need not be dominant over apical determinants for steady-state polarity and emphasize an important contribution of the valence of signals in polarized sorting.     

WIF-B cells: an in vitro model for studies of hepatocyte polarity

We have evaluated the utility of the hepatoma-derived hybrid cell line, WIF-B, for in vitro studies of polarized hepatocyte functions. The majority (> 70%) of cells in confluent culture formed closed spaces with adjacent cells. These bile canalicular-like spaces (BC) accumulated fluorescein, a property of bile canaliculi in vivo. By indirect immunofluorescence, six plasma membrane (PM) proteins showed polarized distributions similar to rat hepatocytes in situ. Four apical PM proteins were concentrated in the BC membrane of WIF-B cells. Microtubules radiated from the BC (apical) membrane, and actin and foci of gamma-tubulin were concentrated in this region. The tight junction- associated protein ZO-1 was present in belts marking the boundary between apical and basolateral PM domains. We explored the functional properties of this boundary in living cells using fluorescent membrane lipid analogs and soluble tracers. When cells were incubated at 4 degrees C with a fluorescent analog of sphingomyelin, only the basolateral PM was labeled. In contrast, when both PM domains were labeled by de novo synthesis of fluorescent sphingomyelin from ceramide, fluorescent lipid could only be removed from the basolateral domain. These data demonstrate the presence of a barrier to the lateral diffusion of lipids between the PM domains. However, small soluble FITC- dextrans (4,400 mol wt) were able to diffuse into BC, while larger FITC- dextrans were restricted to various degrees depending on their size and incubation temperature. At 4 degrees C, the surface labeling reagent sNHS-LC-biotin (557 mol wt) had access to the entire PM, but streptavidin (60,000 mol wt), which binds to biotinylated molecules, was restricted to only the basolateral domain. Such differential accessibility of well-characterized probes can be used to mark each membrane domain separately. These results show that WIF-B cells are a suitable model to study membrane trafficking and targeting in hepatocytes in vitro.     

A hepatocyte-specific basolateral membrane protein is targeted to the same domain when expressed in Madin-Darby canine kidney cells

Different mechanisms for polarized sorting of apical and basolateral plasma membrane proteins appear to be operative in different cell types. In hepatocytes, all proteins are first transported to the basolateral surface, where sorting (probably signal-mediated) of apical proteins then takes place. In contrast, in Madin-Darby canine kidney (MDCK) cells, proteins are directly transported from the trans-Golgi network to their appropriate plasma membrane domain. In order to study the differences in the sorting requirements of the two cell types, we have expressed a hepatocyte-specific basolateral membrane protein, the asialoglycoprotein receptor H1, in MDCK cells. H1 was found to be specifically transported to the basolateral domain also in this heterologous system, suggesting that either the same basolateral targeting signal is operative in both cell types or, more likely, that basolateral transport occurs "by default," i.e. without the requirement for a sorting signal.     

The role of secretory and endocytic pathways in the maintenance of cell polarity

Epithelial cells line virtually every organ cavity in the body and are important for vectorial transport through epithelial monolayers such as nutrient uptake or waste product excretion. Central to these tasks is the establishment of epithelial cell polarity. During organ development, epithelial cells set up two biochemically distinct plasma membrane domains, the apical and the basolateral domain. Targeting of correct constituents to each of these regions is essential for maintaining epithelial cell polarity. Newly synthesized transmembrane proteins destined for the basolateral or apical membrane domain are sorted into separate transport carriers either at the TGN (trans-Golgi network) or in perinuclear REs (recycling endosomes). After initial delivery, transmembrane proteins, such as nutrient receptors, frequently undergo multiple rounds of endocytosis followed by re-sorting in REs. Recent work in epithelial cells highlights the REs as a potent sorting station with different subdomains representing individual targeting zones that facilitate the correct surface delivery of transmembrane proteins.     

The proamphiregulin cytoplasmic domain is required for basolateral sorting, but is not essential for constitutive or stimulus-induced processing in polarized Madin-Darby canine kidney cells

In this study, the role of the amphiregulin precursor (pro-AR) cytoplasmic domain in the basolateral sorting and cell-surface processing of pro-AR in polarized epithelial cells has been investigated using Madin-Darby canine kidney cells stably expressing various human pro-AR forms. Our results demonstrate that newly synthesized wild-type pro-AR (50 kDa) is delivered directly to the basolateral membrane domain with >95% efficiency, where it is sequentially cleaved within the ectodomain to release several soluble amphiregulin (AR) forms. Analyses of a pro-AR cytoplasmic domain truncation mutant (ARTL27) and two pro-AR secretory mutants (ARsec184 and ARsec190) indicated that the pro-AR cytoplasmic domain is not required for efficient delivery to the plasma membrane, but does contain essential basolateral sorting information. We show that the pro-AR cytoplasmic domain truncation mutant (ARTL27) is not sorted in polarized Madin-Darby canine kidney cells, with approximately 65% of the newly synthesized protein delivered to the apical cell surface. Under base-line conditions, ARTL27 was preferentially cleaved from the basolateral surface with 4-fold greater efficiency compared with cleavage from the apical membrane domain. However, ARTL27 ectodomain cleavage could be stimulated equivalently from either membrane domain by a variety of different stimuli. The metalloprotease inhibitor BB-94 could inhibit both base-line and stimulus-induced ectodomain cleavage of wild-type pro-AR and ARTL27. These results indicate that the pro-AR cytoplasmic domain is required for basolateral sorting, but is not essential for ectodomain processing. Preferential constitutive cleavage of ARTL27 from the basolateral cell surface also suggests that the metalloprotease activity involved in base-line and stimulus-induced ARTL27 ectodomain cleavage may be regulated differently in the apical and basolateral membrane domains of polarized epithelial cells.     

Targeting of transmembrane protein shrew-1 to adherens junctions is controlled by cytoplasmic sorting motifs

We recently identified transmembrane protein shrew-1 and showed that it is able to target to adherens junctions in polarized epithelial cells. This suggested shrew-1 possesses specific basolateral sorting motifs, which we analyzed by mutational analysis. Systematic mutation of amino acids in putative sorting signals in the cytoplasmic domain of shrew-1 revealed three tyrosines and a dileucine motif necessary for basolateral sorting. Substitution of these amino acids leads to apical localization of shrew-1. By applying tannic acid to either the apical or basolateral part of polarized epithelial cells, thereby blocking vesicle fusion with the plasma membrane, we obtained evidence that the apically localized mutants were primarily targeted to the basolateral membrane and were then redistributed to the apical domain. Further support for a postendocytic sorting mechanism of shrew-1 was obtained by demonstrating that mu1B, a subunit of the epithelial cell-specific adaptor complex AP-1B, interacts with shrew-1. In conclusion, our data provide evidence for a scenario where shrew-1 is primarily delivered to the basolateral membrane by a so far unknown mechanism. Once there, adaptor protein complex AP-1B is involved in retaining shrew-1 at the basolateral membrane by postendocytic sorting mechanisms.     

Differential distribution of low-density lipoprotein-receptor-related protein (LRP) and megalin in polarized epithelial cells is determined by their cytoplasmic domains

Megalin and the low-density lipoprotein (LDL) receptor-related protein (LRP) are two large members of the LDL receptor family that bind and endocytose multiple ligands. The molecular and cellular determinants that dictate the sorting behavior of these receptors in polarized epithelial cells are largely unknown. Megalin is found apically distributed, whereas the limited information on LRP indicates its polarity. We show here that in Madin-Darby canine kidney cells, both endogenous LRP and a minireceptor containing the fourth ligand-binding, transmembrane and LRP cytosolic domains were basolaterally sorted. In contrast, minireceptors that either lacked the cytoplasmic domain or had the tyrosine in the NPTY motif mutated to alanine showed a preferential apical distribution. In LLC-PK1 cells, endogenous megalin was found exclusively in the apical membrane. Studies were also done using chimeric proteins harboring the cytosolic tail of megalin, one with the fourth ligand-binding domain of LRP and the other two containing the green fluorescent protein as the ectodomain and transmembrane domains of either megalin or LRP. Findings from these experiments showed that the cytosolic domain of megalin is sufficient for apical sorting, and that the megalin transmembrane domain promotes association with lipid rafts. In conclusion, we show that LRP and megalin both contain sorting information in their cytosolic domains that directs opposite polarity, basolateral for LRP and apical for megalin. Additionally, we show that the NPTY motif in LRP is important for basolateral sorting and the megalin transmembrane domain directs association with lipid rafts .     

Myosin VI is required for sorting of AP-1B-dependent cargo to the basolateral domain in polarized MDCK cells

In polarized epithelial cells, newly synthesized membrane proteins are delivered on specific pathways to either the apical or basolateral domains, depending on the sorting motifs present in these proteins. Because myosin VI has been shown to facilitate secretory traffic in nonpolarized cells, we investigated its role in biosynthetic trafficking pathways in polarized MDCK cells. We observed that a specific splice isoform of myosin VI with no insert in the tail domain is required for the polarized transport of tyrosine motif containing basolateral membrane proteins. Sorting of other basolateral or apical cargo, however, does not involve myosin VI. Site-directed mutagenesis indicates that a functional complex consisting of myosin VI, optineurin, and probably the GTPase Rab8 plays a role in the basolateral delivery of membrane proteins, whose sorting is mediated by the clathrin adaptor protein complex (AP) AP-1B. Our results suggest that myosin VI is a crucial component in the AP-1B-dependent biosynthetic sorting pathway to the basolateral surface in polarized epithelial cells.     

Raft-mediated trafficking of apical resident proteins occurs in both direct and transcytotic pathways in polarized hepatic cells: role of distinct lipid microdomains

In polarized hepatic cells, pathways and molecular principles mediating the flow of resident apical bile canalicular proteins have not yet been resolved. Herein, we have investigated apical trafficking of a glycosylphosphatidylinositol-linked and two single transmembrane domain proteins on the one hand, and two polytopic proteins on the other in polarized HepG2 cells. We demonstrate that the former arrive at the bile canalicular membrane via the indirect transcytotic pathway, whereas the polytopic proteins reach the apical membrane directly, after Golgi exit. Most importantly, cholesterol-based lipid microdomains ("rafts") are operating in either pathway, and protein sorting into such domains occurs in the biosynthetic pathway, largely in the Golgi. Interestingly, rafts involved in the direct pathway are Lubrol WX insoluble but Triton X-100 soluble, whereas rafts in the indirect pathway are both Lubrol WX and Triton X-100 insoluble. Moreover, whereas cholesterol depletion alters raft-detergent insolubility in the indirect pathway without affecting apical sorting, protein missorting occurs in the direct pathway without affecting raft insolubility. The data implicate cholesterol as a traffic direction-determining parameter in the direct apical pathway. Furthermore, raft-cargo likely distinguishing single vs. multispanning membrane anchors, rather than rafts per se (co)determine the sorting pathway.     

Apical sorting by galectin-3-dependent glycoprotein clustering

Epithelial cells are characterized by their polarized organization based on an apical membrane that is separated from the basolateral membrane domain by tight junctions. Maintenance of this morphology is guaranteed by highly specific sorting machinery that separates lipids and proteins into different carrier populations for the apical or basolateral cell surface. Lipid-raft-independent apical carrier vesicles harbour the beta-galactoside-binding lectin galectin-3, which interacts directly with apical cargo in a glycan-dependent manner. These glycoproteins are mistargeted to the basolateral membrane in galectin-3-depleted cells, dedicating a central role to this lectin in raft-independent sorting as apical receptor. Here, we demonstrate that high-molecular-weight clusters are exclusively formed in the presence of galectin-3. Their stability is sensitive to increased carbohydrate concentrations, and cluster formation as well as apical sorting are perturbed in glycosylation-deficient Madin-Darby canine kidney (MDCK) II cells. Together, our data suggest that glycoprotein cross-linking by galectin-3 is required for apical sorting of non-raft-associated cargo.