VIP36 was isolated from MDCK cells as a component of glycolipid-enriched detergent-insoluble complexes. The protein is localized to the Golgi apparatus and the cell surface, and belongs to a new family of legume lectin homologues in the animal secretory pathway that might be involved in the trafficking of glycoproteins, glycolipids or both. Here we show that VIP36 is N-glycosylated and expressed in organs abundant in epithelial cells as well as in non-epithelial organs. Our studies demonstrate that the recombinant exoplasmic/luminal domain of VIP36 binds Ca2+ and that the protein decorates internal membrane structures of MDCK cells in vitro that are distinct from the Golgi apparatus. This binding requires Ca2+ and can be specifically inhibited by N-acetyl-D-galactosamine. The recombinant protein was used for affinity chromatography. Glycopeptides obtained from [3H]galactose-labelled cells bind to VIP36 and can be eluted with N-acetyl-D-galactosamine. Our data imply that VIP36 functions as a lectin in post-Golgi trafficking.

In MDCK cells proteins and lipids are sorted into vesicular carriers in the trans-Golgi network (TGN) and delivered to the apical or basolateral cell surface (Matter and Mellman, 1994; Rodriguez-Boulan and Powell, 1992; Simons and Wandinger-Ness, 1990). In order to identify factors involved in this process we have characterized components of immunoisolated exocytic carrier vesicles (Wandinger-Ness et al., 1990). VIP36 (Vesicular-Integral membrane Protein of 36 kDa) is present in apical and basolateral carriers and was isolated as a component of glycolipid-enriched detergent-insoluble complexes containing the apical marker protein influenza virus haemagglutinin (Fiedler et al., 1993, 1994). The exoplasmic domain of VIP36 is homologous to the N-terminal luminal domain of ERGIC-53, a protein localized to the endoplasmic reticulum (ER)-Golgi intermediate region (Schindler et al., 1993; Schweizer et al., 1988). Both proteins show a significant similarity to leguminous plant lectins (Fiedler and Simons, 1994), suggesting that they might be involved in the delivery of saccharidebearing molecules in the secretory pathway (Fiedler and Simons, 1995). Recent studies have demonstrated that ERGIC-53 is identical to a previously defined mannose-binding lectin from monocytes (Arar et al., 1995; Carpentier et al., 1994; Pimpaneau et al., 1991) but the function of ERGIC-53 remains unknown.

Here we analyze the tissue distribution of VIP36 and show that the recombinant exoplasmic domain of VIP36 (VIP36Exo) binds Ca2+ similarly to legume lectins. We further demonstrate that VIP36Exo binding to MDCK cells and the binding of glycopeptides to VIP36Exo can be competed by N-acetyl-D-galactosamine confirming a glycan affinity of VIP36.

Materials, cells and antibodies

Unless otherwise indicated, all chemicals were obtained from the sources previously described (Fiedler et al., 1993, 1994, 1995). Sephadex G-25 and sugars were purchased from Sigma (St Louis, MO) except for sucrose (Merck, Darmstadt, Germany) and D-Gal (Serva, Heidelberg, Germany). [2-3H]mannose and [6-3H]galactose were from Amersham (Braunschweig, Germany). Pronase from Streptomyces griseus was from Calbiochem (San Diego, CA). Catalase was purchased from Pharmacia (Uppsala, Sweden). Affi-Gel 10 was purchased from BioRad (Munich, Germany).

MDCK strain II cells were grown and passaged using media previously described (Wandinger-Ness et al., 1990).

Anti-VIP36-C antibodies were described previously (Fiedler et al., 1994). Anti-VIP36-N polyclonal sera were raised against synthetic peptides covalently coupled to keyhole limpet haemocyanin using the N-terminal residues 45-62 of VIP36 (DITDGNSEHLKREHSLIK) (VIP36-N) according to Kreis (1986). Sera were collected after the sixth injection of antigen. For affinity purification, the N-terminal peptide was linked directly to CNBr-activated Sepharose 4B according to the manufacturer (Pharmacia, Uppsala, Sweden). Serum (1.4 ml) was rotated with the matrix overnight at 4°C. Bound antibody was eluted with 0.2 M glycine (pH 2.8) and the fractions were neutralized with 1 M Tris-HCl (pH 8.0).

The mouse cell line producing monoclonal antibody 9E10 against the human c-Myc epitope EQKLISEED (Evan et al., 1985) was purchased from The European Collection of Animal Cultures (no. 85102202).

Northern blot analysis

RNA was prepared from adult mouse tissues according to Lütcke et al. (1993). Total RNA (25 μg) was separated on a 1% agarose gel and transferred in 10× SSC (sodium chloride/sodium citrate) to Gene Screen Plus membranes (Dupont Co., Wilmington, DE). Probes (α-32P-labelled) were prepared by random priming (Amersham, Braunschweig, Germany) using a 795 bp fragment (encoding amino acids 49-314 of VIP36) generated by PCR from MDCK VIP36 cDNA (Fiedler et al., 1994) or using a 283 bp PCR fragment from human glyceraldehyde phosphate dehydrogenase cDNA. Unincorporated nucleotides were removed by a push column (NucTrap; Stratagene, La Jolla, CA). Filters were prehybridized in a solution containing 50% formamide, 5× SSC, 5× Denhardt’s solution and 1% SDS for 2 hours at 42°C, and subsequently hybridized for 18 hours in the same solution containing ∼3.5×106 cpm of the probe. Filters were washed in 2× SSC, 0.5% SDS at 50°C. X-ray films were exposed at −70°C with intensifying screens.

Glycosylation analysis

VIP36 was immunoprecipitated from MDCK cells using the affinitypurified VIP36-N antibody. Cells were metabolically labelled for 15 hours using 26 MBq [35S]methionine (Amersham, Braunschweig, Germany)/75 cm2 tissue-culture flask and chased for 2 hours with medium containing 10× methionine. They were lysed in buffer A (1% NP-40, 0.5% sodium deoxycholate, 2 mM EDTA, protease inhibitor cocktail (CLAP: 20 μg/ml each of chymostatin, leupeptin, antipain and pepstatin) in phosphate buffered saline (PBS)) by rotation at 4°C and cleared by centrifugation in an Eppendorf centrifuge. The supernatant was adjusted to 1 mg/ml BSA and 0.5% SDS final concentration. After incubation for 12 hours at 4°C with VIP36-N antibodies, pre-blocked Protein A-Sepharose was added and incubated for 1 hour at 4°C. Samples (corresponding to ∼1/4 of a 75 cm2 flask) were washed 4× with buffer B (buffer A including 1% SDS final concentration, 1 mg/ml BSA) and 1× with PBS. After removal of the PBS, 33 μl of a solution containing 0.6% SDS, 150 mM sodium citrate (pH 5.5), 1.5% β-mercaptoethanol was added to sample 1 and 33 μl of a solution containing 0.3% SDS, 300 mM sodium phosphate (pH 7.0), 1.5% β-mercaptoethanol was added to sample 2. The samples were heated at 97°C for 4 minutes. After cooling, the samples were adjusted to 100 μl with water including CLAP, 0.5% NP-40 and 5 mU endoglycosidase H (Boehringer Mannheim, Germany) for sample 1 or including CLAP, 0.5% NP-40, 10 mM EDTA and 500 mU N-glycosidase F (Boehringer Mannheim, Germany) for sample 2. Control samples were treated as sample 1 or 2, omitting the enzyme. Following incubation at 37°C for 18 hours SDS-sample buffer was added, the samples were heated at 97°C for 5 minutes and resolved by SDS-PAGE as described (Bennett et al., 1988).

Expression of VIP36Exo and Ca2+ binding

A cDNA encoding a recombinant soluble form of VIP36 (VIP36Exo, amino acids 45-322) including a human c-Myc epitope (EQKLISEED) followed by 6 histidines residues at the C terminus was generated by PCR using MDCK VIP36 cDNA as a template (Fiedler et al., 1994). PCR was carried out according to Chavrier et al. (1992) with the primers 5’-CGCCCATATGGATATAACT-GACGGTAACAGTGAAC-3’ and 5’-GCGGGGATCCTTAG TGGTGGTGGTGGTGGTGCAGGTCCTCCTCGGAGATCAGCT-TCTGCTCCCGCCACCCGGTCAGGGGCCCACTT-3’. Nucleotide sequences were verified using the T7 sequencing kit (Pharmacia, Uppsala, Sweden). The cDNA was subcloned into the pRAT-5 vector (Peränen, J., Rikkonen, M., Hyvönen, M. and Kääriäinen, L., unpublished) which utilizes a T7lac promoter and allows protein expression in an Escherichia coli BL21 (DE3) strain (Novagen, Madison, WI). Log-phase growing cultures were cooled to 15°C and expression was induced with 0.5 mM isopropyl-thiogalactoside. The growth medium was replenished with 0.2 mM CaCl2 and 0.2 mM MnCl2. The cultures were incubated for 22 hours at 15°C. The cells were passed twice through a French press (10,000 psi) at 4°C in buffer C (50 mM Tris-HCl, pH 8.0, 2 mM PMSF, 0.2 mM CaCl2, 0.2 mM MnCl2) including 0.1% Tween-20. Following centrifugation in a SW40 rotor at 30,000 rpm for 1 hour the supernatant was adjusted to 5 mM imidazole (pH 8.0) and incubated for 1 hour at 4°C with Ni-NTA (N-(5-amino-1-carboxypentyl)iminodiacetic acid) agarose (Qiagen, Hilden, Germany). The resin was washed with buffer C, including 5 mM imidazole, 0.1% Tween-20 and 0.5 M NaCl, followed by the same buffer omitting Tween-20. The last wash was carried out with buffer C including 4 mM ATP, 1 mM MgCl2, 10 mM imidazole, 0.5 M NaCl. VIP36Exo was eluted with buffer C, including 200 mM imidazole, 0.5 M NaCl, and extensively dialyzed at 4°C against 50 mM Hepes, pH 7.2, 150 mM NaCl, 1 mM MgCl2, 0.2 mM CaCl2, 0.2 mM MnCl2.

Ca2+ binding to VIP36Exo was determined by equilibrium dialysis using a micro-dialyzer (Bachofer, Reutlingen, Germany) and 45Ca (Amersham, Braunschweig, Germany) as a tracer. Prior to Ca2+ binding, VIP36Exo was extensively dialyzed against buffer D (20 mM Tris-HCl, pH 7.4, 100 mM KCl). Ca2+ binding was carried out for 4 hours at room temperature in buffer D, including 2 mM MgCl2 and 0.4 mM MnCl2. The VIP36Exo concentration was determined by amino acid analysis and was 20 μM in the binding reaction.

Immunofluorescence assay

Subconfluent MDCK cells, grown on glass coverslips, were washed with PBS and fixed in 4 % paraformaldehyde in PBS for 20 minutes. Free aldehyde groups were quenched with 50 mM NH4Cl in PBS for 20 minutes and then the cells were permeabilized for 4 minutes at room temperature in 0.1% Triton X-100 in PBS. After a rinse in Dulbecco’s-PBS (DPBS) the cells were incubated for 1 hour in 0.2% gelatin in DPBS. A solution containing 80 nM VIP36Exo and ∼40 nM of purified ascites of 9E10 anti-c-Myc antibodies, with or without saccharides, in 0.2% gelatin in DPBS was incubated for 30 minutes at room temperature and then added to the coverslips and incubated for 1 hour at 37°C. After 4 short washes in DPBS the cells were fixed in 4% paraformaldehyde in PBS for 10 minutes, quenched with 50 mM NH4Cl in PBS for 10 minutes and incubated for 20 minutes in 0.2% gelatin in PBS. Binding in the absence of Ca2+ was carried out in PBS instead of DPBS in all solutions. In addition, 6 mM EDTA was added to the VIP36Exo/9E10 mixture. Binding was visualized with preadsorbed anti-mouse rhodamine-conjugated antibodies (Dianova, Hamburg, Germany). Immunofluorescence localization of endogenous VIP36 with VIP36-C antibodies was carried out as described (Fiedler et al., 1994). The coverslips were viewed and photographed with an Axiophot photomicroscope (Carl Zeiss, Oberkochen, Germany) using identical exposure times.

VIP36Exo affinity chromatography of glycopeptides

Glycopeptides were isolated from MDCK cells. The cells were labelled for 36 hours with 74 MBq [2-3H]mannose or [6-3H]galactose in 5 ml MEM per 10 cm dish. The cells were washed twice with PBS and scraped from the dish in PBS. Following precipitation by the addition of nine volumes acetone for 30 minutes on ice, the samples were pelleted in the Eppendorf centrifuge. The pellets were incubated for 16 hours at 60°C in 1 ml of 0.1 M Tris-HCl, pH 8.0, 1 mM CaCl2, 0.02% NaN3 and 10 mg/ml predigested pronase. The reactions were stopped by heating the samples at 100°C for 10 minutes and the precipitated material was removed by centrifugation for 5 minutes in the Eppendorf centrifuge. The supernatants were dried down and the glycopeptides were desalted on a Sephadex G-25 column (1 cm × 60 cm) equilibrated in 7% propanol. The glycopeptide-containing fractions were dried and dissolved in 20 mM Tris-HCl, pH 7.0, 150 mM NaCl.

Affinity columns were prepared by coupling Ni-NTA purified VIP36Exo or catalase to Affi-Gel according to the manufacturer’s procedure. A 25 mg sample of protein was mixed end-over-end at 4°C for 4 hours with 1 ml Affi-Gel 10 in 3.5 ml (total volume) of 25 mM Hepes, pH 7.2, 0.2 mM CaCl2, 0.2 mM MnCl2, 150 mM NaCl, 0.005% NP-40 (buffer E). The beads were then washed with cold buffer E. The coupling efficiency was ∼86% and ∼56% for VIP36Exo and catalase, respectively. Unreacted groups were blocked by adding 0.1 M ethanolamine in buffer E. After mixing for an additional 2 hours the beads were washed extensively with buffer E and poured into 1.5 cm × 10 cm columns. The columns were stored at 4°C in buffer E with 0.02% NaN3.

For affinity chromatography the columns were equilibrated in HBS (25 mM Hepes, pH 7.0, 150 mM NaCl), 2 mM CaCl2, 2 mM MnCl2, 10% glycerol (buffer F) at 4°C. [2-3H]mannose- or [6-3H]galactoselabelled glycopeptides (∼300,000 cpm) in 2.5 ml buffer F (total volume) was applied to the column and mixed end-over-end at 4°C for 10 hours. The columns were washed with 50 column volumes of buffer F at room temperature. The flow rate was 1 ml/minute and 1 ml fractions were collected. Bound material was eluted with 20 column volumes 0.2 M GalNAc in buffer F at room temperature followed by elution with 20 column volumes HBS, 10% glycerol, 10 mM EDTA. The fractions were analyzed by scintillation counting.

VIP36 is expressed in epithelial and non-epithelial cells

Since VIP36 was originally isolated from MDCK cells (Fiedler et al., 1994) it was interesting to analyze its expression in both epithelial and non-epithelial organs by northern blot analysis (Fig. 1a). The probe reacted with a transcript of 1600 nucleotides, which corresponds to the size of the isolated cDNA and was present in mouse kidney, liver, intestine, lung, spleen, heart and NIH 3T3 fibroblasts. This demonstrates that VIP36 is expressed in organs abundant in epithelial cells as well as in non-epithelial tissues. A very low level of this transcript was present in brain. In addition, the probe reacted with transcripts of ∼3000 and ∼4000 nucleotides that are unknown but may correspond to related family members.

VIP36 is N-glycosylated

VIP36 has one potential N-linked glycosylation site. In order to study its glycosylation we raised antibodies against an N-terminal peptide of VIP36 (residues 45-62). The affinitypurified antibodies were used to immunoprecipitate VIP36 from metabolically labelled MDCK cells (Fig. 2a, lane 1 and 3). The immunoprecipitate was incubated with endoglycosidase H (Endo H; Fig. 2a, lane 2) or N-glycosidase F (PNGase F; Fig. 2a, lane 4). The decrease in apparent molecular mass to ∼34 kDa after PNGase F treatment demonstrates that VIP36 is N-glycosylated. The N-glycan is sensitive to Endo H treatment after short times of chase (data not shown). Previous experiments have shown that in vitro translated VIP36 and native VIP36 have identical isoelectric points and co-migrate in two-dimensional IEF/SDS-PAGE (Fiedler et al., 1994). Moreover, the isoelectric point is not altered in the presence of influenza virus neuraminidase (Wandinger-Ness et al., 1990). Taken together, this suggests that the majority of VIP36 molecules in MDCK cells carry one N-glycan that is non-sialylated and contains terminal GlcNAc or Gal residues.

Ca2+ binding of VIP36

Legume plant lectins require Ca2+ and Mn2+ for sugar binding (Sharon and Lis, 1990). To test the divalent cation binding of VIP36 we generated a cDNA construct of VIP36 for expression in E. coli. The recombinant cDNA encoded the complete exoplasmic/luminal domain of VIP36 lacking the N-terminal signal sequence and included a c-Myc epitope and (His)6 sequence at the C terminus (VIP36Exo). The protein was purified to homogeneity on a Ni-NTA column (Fig. 2b) and behaved as a soluble monomer in a sucrose sedimentation gradient (data not shown). Contrary to ERGIC-53 (Schweizer et al., 1988), recombinant and endogenous VIP36 did not form disulphide-linked oligomers. Furthermore, no stable non-covalent oligomers could be detected by cross-linking or sedimentation analysis (data not shown). Ca2+ binding was analyzed by equilibrium dialysis using 45Ca as a tracer. The binding was saturable (Fig. 3a) with two binding sites per VIP36Exo molecule (Fig. 3b) and a macroscopic KD of ∼45 μM. No binding was detected in a control experiment with a c-Myc-(His)6 peptide alone. Mn2+ was included during the dialysis but was found not to be required for Ca2+ binding. In the presence of 40 μM free Ca2+ as the sole source of divalent cations 1 mol VIP36 bound ∼0.8 mol Ca2+. No significant changes were detected in the presence of extra added 2 mM MgCl2 and 2 mM MgCl2, 0.4 mM MnCl2. The Ca2+ binding of VIP36Exo implies that the bacterially expressed protein is correctly folded and functional.

Glycan binding of VIP36

VIP36Exo was then used to detect potential VIP36 ligands in MDCK cells by a simple immunofluorescence assay. A ∼2-fold molar excess of VIP36Exo was pre-incubated with anti-c-Myc antibodies and added to fixed, Triton X-100-permeabilized MDCK cells. The binding was visualized with rhodamine-conjugated secondary antibodies in the microscope. A strong labelling of punctate structures throughout the cells was observed (Fig. 4a). These structures are mainly localized in the interior of the cells, since only a weak labelling was detected on the surface of non-permeabilized cells (data not shown). A c-Myc(His)6 peptide or the anti-c-Myc antibodies alone showed negligible binding (data not shown). Only weak binding was detected when VIP36Exo was bound first to permeabilized cells and then visualized with anti-c-Myc antibodies. This suggests that the anti-c-Myc antibody pre-incubation increased the binding avidity and that multivalency is necessary for in vitro binding. Therefore, the pre-incubation was used in all further experiments. The interaction was decreased in the absence of divalent cations (Fig. 4b). Note the two different qualities of the staining: only the small punctate labelling throughout the cells was significantly reduced. The staining of larger structures was typically decreased less and may reflect artefactual binding. For comparison the localization of endogenous VIP36 is shown in Fig. 4g.

An array of saccharides was included in the assay to test the involvement of glycans in the binding reaction. The binding could not be inhibited with 200 mM to 400 mM Glc, sucrose, raffinose, inositol, GlcN, Gal, D-Fuc, L-Fuc, Man, lactose, ManN, GlcNAc, Sia (all D stereoisomers except when indicated) but was strongly reduced in the presence of 400 mM GalNAc (Fig. 4c), 40 mM GalNAc (Fig. 4d) and 4 mM GalNAc (Fig. 4e). Again, only the small punctate labelling but not the labelling of the larger structures was significantly reduced. The presence of 0.4 mM GalNAc only slightly decreased the binding (Fig. 4f). The GalNAc-sensitive labelling pattern was distinct from the localization of the majority of endogenous VIP36 (Fig. 4g).

To further analyze the glycan affinity of VIP36 we prepared glycopeptides from [2-3H]mannose- or [6-3H]galactoselabelled MDCK cells and used VIP36Exo for affinity chromatography. [2-3H]mannose and [6-3H]galactose are incorporated into N-linked glycans and N- and O-linked glycans, respectively. Bound material was eluted with 0.2 M GalNAc in the presence of divalent cations followed by elution with EDTA (Fig. 5). Only glycopeptides isolated from [6-3H]galactose-labelled cells bound to the VIP36Exo column and could be eluted with GalNAc (∼0.1% of total sample applied). [2-3H]mannose-labelled glycopeptides did not bind and could not be eluted with either GalNAc or EDTA. A catalase control column showed only negligible binding.

These results further confirm a glycan affinity of VIP36.

In this study we show that VIP36 is a Ca2+-binding glycoprotein. We have produced and purified a recombinant soluble form of the luminal/exoplasmic domain of VIP36 and determined that there are two Ca2+ binding sites per molecule. The metal binding is in accordance with the homology of VIP36 to legume plant lectins (Fiedler et al., 1994). Leguminous lectins, however, bind one Ca2+ and one additional Mn2+ (Sharon and Lis, 1990) and Mn2+ is required for Ca2+ binding. We could not observe any Mn2+-dependence of Ca2+ binding for VIP36. Recombinant VIP36 was used to decorate putative endogenous ligands of VIP36 in a simple imunofluorescence assay. The cell binding was Ca2+-dependent and could be specifically competed with GalNAc. This suggests that, like deglycosylated legume lectins (Lis and Sharon, 1986), the carbohydrate moieties of VIP36 are not required for biological activity. The GalNAc concentrations required for the inhibition of binding were in the mM range, which is in agreement with the typically low affinities of lectins to monosaccharides (Sharon and Lis, 1990). The GalNAcsensitive labelling pattern was distinct from the localization of the majority of endogenous VIP36, confirming that the recombinant protein does not simply interact with its endogenous counterpart present in the Golgi apparatus. The binding of VIP36 to these structures was enhanced by antibody clustering of recombinant VIP36, which indicates that the binding is of low affinity. The punctate labelling probably represents post-Golgi vesicular carriers, either biosynthetic, endocytic or both (Fiedler et al., 1994). The putative unidentified glycan ligands of VIP36 are thus segregated from the bulk of the endogenous molecule, which is consistent with VIP36 being involved in a sorting process.

Recombinant VIP36 was further used for affinity chromatography of glycopeptides isolated from [2-3H]mannose- or [6-3H]galactose-labelled cells. Only [6-3H]galactose-labelled glycopeptides bound to VIP36 and could be eluted with GalNAc, suggesting that VIP36 binds to O-linked glycans. However, it is clear that this simple procedure only detects interactions of relatively high affinity. Therefore, at present the possibility cannot be excluded that [2-3H]mannose-labelled N-glycans may also bind to VIP36 but are only retarded on the affinity column. It should also be noted that although the binding of recombinant VIP36 to cells can be selectively inhibited with GalNAc it is possible to observe an inhibition of lectin binding with mono-saccharides that are unrelated to the specific, more complex in vivo ligand (Merkle and Cummings, 1987).

VIP36 is ubiquitously expressed in epithelial and nonepithelial organs. How does this finding fit with its postulated role in apical transport? If the apical sorting machinery were unique to epithelial cells this would speak against a role of VIP36 in this process. However, it is possible that nonpolarized cells also have two routes for the transport from the TGN to the cell surface that are equivalent to the apical and basolateral routes in epithelial cells (Fiedler and Simons, 1995; Matter and Mellman, 1994). The difference would be the polarized delivery and vesicle docking to two different surface domains in epithelial cells, while the sorting and vesicle budding machinery would be similar in both cell types.

The VIP36 homologue ERGIC-53 (Fiedler and Simons, 1994; Schindler et al., 1993) has been identified as a mannosespecific lectin (Arar et al., 1995; Carpentier et al., 1994; Pimpaneau et al., 1991) but endogenous ligands have not been characterized. It therefore remains to be shown whether VIP36 and ERGIC-53 interact with glycoproteins, glycolipids or glycosylphosphatidyl inositol-anchored molecules in vivo. More work will be necessary to elucidate whether ERGIC-53 is involved in the transport of glycoproteins from the ER to the Golgi apparatus or in the quality control of protein folding by recycling unfolded glycosylated proteins back to the ER; for further discussion see Fiedler and Simons (1995). Similarly, a confirmation of the involvement of VIP36 in glycoprotein and/or glycolipid sorting in epithelial cells will have to await the identification of its natural glycan ligand.

We thank Johan Peränen, Bernard Hoflack and Michael Veit for critical comments on the manuscript; Ten Feizi for hospitality and advice; Chun-Ting Yuen and Robert Childs for reagents and help during K.F.’s visit to the Glycosciences Laboratory; all members of the Simons’ lab for stimulating discussions; Anne Lütcke for the mouse mRNAs; Roland Kellner and Tony Houthaeve for amino acid analysis; and Hilkka Virta for technical assistance. This work was supported by the Boehringer Ingelheim Fonds (to K.F.) and SFB352 of the Deutsche Forschungsgemeinschaft.

Arar
,
C.
,
Carpentier
,
V.
,
Le Caer
,
J.-P.
,
Monsigny
,
M.
,
Legrand
,
A.
and
Roche
,
A.-C.
(
1995
).
ERGIC-53, a membrane protein of the endoplasmic reticulum-golgi intermediate compartment, is identical to MR60, an intracellular mannose-specific lectin of myelomonocytic cells
.
J. Biol. Chem
.
270
,
3551
3553
.
Bennett
,
M.
,
Wandinger-Ness
,
A.
and
Simons
,
K.
(
1988
).
Release of putative exocytic transport vesicles from perforated MDCK cells
.
EMBO J
.
7
,
4075
4085
.
Carpentier
,
V.
,
Vassard
,
C.
,
Plessis
,
C.
,
Motta
,
G.
,
Monsigny
,
M.
and
Roche
,
A. C.
(
1994
).
Characterization and cellular localization by monoclonal antibodies of the 60 kDa mannose specific lectin of human promyelocytic cells, HL60
.
Glycoconjugate J
.
11
,
333
338
.
Chavrier
,
P.
,
Simons
,
K.
and
Zerial
,
M.
(
1992
).
The high complexity of the rab and rho GTP-binding protein subfamilies revealed by a PCR cloning approach
.
Gene
112
,
261
264
.
Evan
,
G. I.
,
Lewis
,
G. K.
,
Ramsay
,
G.
and
Bishop
,
J. M.
(
1985
).
Isolation of monoclonal antibodies specific for human c-myc proto-oncogene product
.
Mol. Cell. Biol
.
5
,
3610
3616
.
Fiedler
,
K.
,
Kobayashi
,
T.
,
Kurzchalia
,
T. V.
and
Simons
,
K.
(
1993
).
Glycosphingolipid-enriched detergent-insoluble complexes in protein sorting in epithelial cells
.
Biochemistry
32
,
6365
6373
.
Fiedler
,
K.
,
Parton
,
R. G.
,
Kellner
,
R.
,
Etzold
,
T.
and
Simons
,
K.
(
1994
).
VIP36, a novel component of glycolipid rafts and exocytic carrier vesicles in epithelial cells
.
EMBO J
.
13
,
1729
1740
.
Fiedler
,
K.
and
Simons
,
K.
(
1994
).
A putative novel class of animal lectins in the secretory pathway homologous to leguminous lectins
.
Cell
77
,
625
626
.
Fiedler
,
K.
,
Lafont
,
F.
,
Parton
,
R. G.
and
Simons
,
K.
(
1995
).
Annexin XIIIb: a novel epithelial specific annexin is implicated in vesicular traffic to the apical plasma membrane
.
J. Cell Biol
.
128
,
1043
1053
.
Fiedler
,
K.
and
Simons
,
K.
(
1995
).
The role of N-glycans in the secretory pathway
.
Cell
81
,
309
312
.
Kreis
,
T. E.
(
1986
).
Microinjected antibodies against the cytoplasmic domain of vesicular stomatitis virus glycoprotein, block its transport to the cell surface
.
EMBO J
.
5
,
931
941
.
Lis
,
H.
and
Sharon
,
N.
(
1986
).
Lectins as molecules and as tools
.
Annu. Rev. Biochem
.
55
,
35
67
.
Lütcke
,
A.
,
Jansson
,
S.
,
Parton
,
R. G.
,
Chavrier
,
P.
,
Valencia
,
A.
,
Huber
,
L. A.
,
Lehtonen
,
E.
and
Zerial
,
M.
(
1993
).
Rab17, a novel small GTPase, is specific for epithelial cells and is induced during cell polarization
.
J. Cell Biol
.
121
,
553
564
.
Matter
,
K.
and
Mellman
,
I.
(
1994
).
Mechanisms of cell polarity: Sorting and transport in epithelial cells
.
Curr. Opin. Cell Biol
.
6
,
545
554
.
Merkle
,
R. K.
and
Cummings
,
R. D.
(
1987
).
Lectin affinity chromatography of glycopeptides
.
Meth. Enzymol
.
138
,
232
259
.
Pimpaneau
,
V.
,
Midoux
,
P.
,
Monsigny
,
M.
and
Roche
,
A. C.
(
1991
).
Characterization and isolation of an intracellular D-mannose-specific receptor from human promyelocytic HL60 cells
.
Carbohydrate Res
.
213
,
95
108
.
Rodriguez-Boulan
,
E.
and
Powell
,
S. K.
(
1992
).
Polarity of epithelial and neuonal cells
.
Annu. Rev. Cell Biol
.
8
,
395
427
.
Schindler
,
R.
,
Itin
,
C.
,
Zerial
,
M.
,
Lottspeich
,
F.
and
Hauri
,
H.-P.
(
1993
).
ERGIC-53, a membrane protein of the ER-Golgi intermediate compartment, carries an ER retention motif
.
Eur. J. Cell Biol
.
61
,
1
9
.
Schweizer
,
A.
,
Fransen
,
J. A. M.
,
Bachi
,
T.
,
Ginsel
,
L.
and
Hauri
,
H. P.
(
1988
).
Identification, by a monoclonal antibody, of a 53-kD protein associated with a tubular vesicular compartment at the cis-side of the Golgi apparatus
.
J. Cell Biol
.
107
,
1643
1653
.
Sharon
,
N.
and
Lis
,
H.
(
1990
).
Legume lectins - a large family of homologous proteins
.
FASEB J
.
4
,
3198
3208
.
Simons
,
K.
and
Wandinger-Ness
,
A.
(
1990
).
Polarized sorting in epithelia
.
Cell
62
,
207
210
.
Wandinger-Ness
,
A.
,
Bennet
,
M. K.
,
Antony
,
C.
and
Simons
,
K.
(
1990
).
Distinct transport vesicles mediate the delivery of plasma membrane proteins to the apical and basolateral domains of MDCK cells
.
J. Cell Biol
.
111
,
987
1000
.