Heidi Tveit

A proteoglycan undergoes different modifications en route to the apical and basolateral surfaces of Madin-Darby canine kidney cells

We have grown polarized epithelial Madin-Darby canine kidney II (MDCK II) cells on filters in the presence of [35S]sulfate, [3H]glucosamine, or [35S]cysteine/[35S]methionine to study proteoglycan (PG) synthesis, sorting, and secretion to the apical and basolateral media. Whereas most of the [35S]sulfate label was recovered in basolateral PGs, the [3H]glucosamine label was predominantly incorporated into the glycosaminoglycan chains of apical PGs, indicating that basolateral PGs are more intensely sulfated than their apical counterparts. Expression of the PG serglycin with a green fluorescent protein tag (SG-GFP) in MDCK II cells produced a protein core secreted 85% apically, which was largely modified by chondroitin sulfate chains. Surprisingly, the 15% of secreted SG-GFP molecules recovered basolaterally were more heavily sulfated and displayed a different sulfation pattern than the apical counterpart. More detailed studies of the differential modification of apically and basolaterally secreted SG-GFP indicate that the protein cores have been designated to apical and basolateral transport platforms before pathway-specific, post-translational modifications have been completed.

Serglycin is implicated in the promotion of aggressive phenotype of breast cancer cells.

Serglycin is a proteoglycan expressed by some malignant cells. It promotes metastasis and protects some tumor cells from complement system attack. In the present study, we show for the first time the in situ expression of serglycin by breast cancer cells by immunohistochemistry in patients’ material. Moreover, we demonstrate high expression and constitutive secretion of serglycin in the aggressive MDA-MB-231 breast cancer cell line. Serglycin exhibited a strong cytoplasmic staining in these cells, observable at the cell periphery in a thread of filaments near the cell membrane, but also in filopodia-like structures. Serglycin was purified from conditioned medium of MDA-MB-231 cells, and represented the major proteoglycan secreted by these cells, having a molecular size of ∼250 kDa and carrying chondroitin sulfate side chains, mainly composed of 4-sulfated (∼87%), 6-sulfated (∼10%) and non-sulfated (∼3%) disaccharides. Purified serglycin inhibited early steps of both the classical and the lectin pathways of complement by binding to C1q and mannose-binding lectin. Stable expression of serglycin in less aggressive MCF-7 breast cancer cells induced their proliferation, anchorage-independent growth, migration and invasion. Interestingly, over-expression of serglycin lacking the glycosaminoglycan attachment sites failed to promote these cellular functions, suggesting that glycanation of serglycin is a pre-requisite for its oncogenic properties. Our findings suggest that serglycin promotes a more aggressive cancer cell phenotype and may protect breast cancer cells from complement attack supporting their survival and expansion.

How many ways through the Golgi maze

The secretory route in eukaryotic cells has been regarded as one common pathway from the endoplasmic reticulum (ER) through the Golgi cisternae to the trans Golgi network where recognition, sorting and exit of cargo molecules are thought to occur. Morphologically, the ribosome‐coated ER is observed throughout the cytoplasm, while the Golgi apparatus usually is confined to a perinuclear position in mammalian cells. However, Golgi outposts have been observed in neuronal dendrites and dispersed Golgi elements in skeletal muscle myofibers. In insects, like in Drosophila melanogaster imaginal disc cells and epidermal cells of Tobacco and Arabidopsis leafs, individual Golgi stacks are distributed throughout the cytoplasm. Golgi stacks do not only differ in their intracellular localization but also in the number of stacks from one to several hundreds. Each stack consists of closely aligned, flattened, membrane‐limited cisternae. The number of cisternae in a Golgi stack is also variable, 2–3 in some ciliates, 10 in many plant cell types and up to 30 in certain euglenoids. The yeast Saccharomyces cerevisiae has a Golgi structure of minimal complexity with scattered solitary cisternae. It is assumed that the number of Golgi cisternae reflects the overall complexity of the enzymatic reactions that occur in their lumen, while the number of stacks reflects the load of macromolecules arriving at the cis side. In this review, we will focus on how the available morphological and biochemical data fit with the current view of protein sorting in the secretory pathway, particularly in polarized cells like neuronal and epithelial cells.

Characterization of the prion protein 3F4 epitope and its use as a molecular tag

The monoclonal antibody (MAb) 3F4 has for nearly two decades been one of the most commonly used tools in prion research. This MAb has contributed significantly to our understanding of the normal cell biology of the prion protein (PrP(C)), as well as the disease related abnormalities occurring in prion diseases. The 3F4 antibody binds strongly to human and hamster PrP, with a specific requirement of two Met residues at positions 109 and 112 in the human PrP. Other species in which PrP lack one of the Met residues, like cattle and sheep, or both, like rat and mouse, do not react with the 3F4 antibody. These and other observations have led to the commonly accepted notion that the 3F4 epitope consists of the tetra-peptide Met-Lys-His-Met. In this study, we have identified the minimal epitope for 3F4 by studying its binding to synthetic peptides and by analysis of mutated ovine PrP::GFP constructs expressed in cell culture. We have found that the 3F4 epitope consists of a hepta-peptide (Lys-Thr-Asn-Met-Lys-His-Met), which in sheep encompass residues 109-115. We found that Lys 109 is critically important for 3F4 binding, as omission, or substitution of this residue to Ala resulted in no binding. We also demonstrate that the hepta-peptide constituting the minimal 3F4 epitope, can be used as a discrete, moveable high-affinity molecular tag. Thus, the 3F4 antibody can find its use beyond prion research.

A Secretory Golgi Bypass Route to the Apical Surface Domain of Epithelial MDCK Cells

Proteins leave the endoplasmic reticulum (ER) for the plasma membrane via the classical secretory pathway, but routes bypassing the Golgi apparatus have also been observed. Apical and basolateral protein secretion in epithelial Madin‐Darby canine kidney (MDCK) cells display differential sensitivity to Brefeldin A (BFA), where low concentrations retard apical transport, while basolateral transport still proceeds through intact Golgi cisternae. We now describe that BFA‐mediated retardation of glycoprotein and proteoglycan transport through the Golgi apparatus induces surface transport of molecules lacking Golgi modifications, possessing those acquired in the ER. Low concentrations of BFA induces apical Golgi bypass, while higher concentrations were required to induce basolateral Golgi bypass. Addition of the KDEL ER‐retrieval sequence to model protein cores allowed observation of apical Golgi bypass in untreated MDCK cells. Basolateral Golgi bypass was only observed after the addition of BFA or upon cholesterol depletion. Thus, in MDCK cells, an apical Golgi bypass route can transport cargo from pre‐Golgi organelles in untreated cells, while the basolateral bypass route is inducible.

Neutralization of endomembrane compartments in epithelial MDCK cells affects proteoglycan synthesis in the apical secretory pathway.

PGs (proteoglycans) are proteins acquiring long, linear and sulfated GAG (glycosaminoglycan) chains during Golgi passage. In MDCK cells (Madin-Darby canine kidney cells), most of the CS (chondroitin sulfate) PGs are secreted apically, whereas most of the HS (heparan sulfate) PGs are secreted basolaterally. The apical and basolateral secretory routes differ in their GAG synthesis, since a protein core that traverses both routes acquires shorter chains, but more sulfate, in the basolateral pathway than in the apical counterpart [Tveit, Dick, Skibeli and Prydz (2005) J. Biol. Chem. 280, 29596-29603]. Golgi cisternae and the trans-Golgi network have slightly acidic lumens. We therefore investigated how neutralization of endomembrane compartments with the vacuolar H(+)-ATPase inhibitor Baf A(1) (bafilomycin A(1)) affected GAG synthesis and PG sorting in MDCK cells. Baf A(1) induced a slight reduction in basolateral secretion of macromolecules, which was compensated by an apical increase. More dramatic changes occurred to PG synthesis in the apical pathway on neutralization. The difference in apical and basolateral PG sulfation levels observed for control cells was abolished, due to enhanced sulfation of apical CS-GAGs. In addition, a large fraction of apical HS-GAGs was elongated to longer chain lengths. The differential sensitivity of the apical and basolateral secretory pathways to Baf A(1) indicates that the apical pathway is more acidic than the basolateral counterpart in untreated MDCK cells. Neutralization gave an apical GAG output that was more similar to that of the basolateral pathway, suggesting that neutralization made the luminal environments of the two pathways more similar.

Easy HPLC-based separation and quantitation of chondroitin sulphate and hyaluronan disaccharides after chondroitinase ABC treatment.

The sulphation patterns of glycosaminoglycan (GAG) chains are decisive for the biological activity of their proteoglycan (PG) templates for sugar chain polymerization and sulphation. The amounts and positions of sulphate groups are often determined by HPLC analysis of disaccharides resulting from enzymatic degradation of the GAG chains. While heparan sulphate (HS) and heparin are specifically degraded by heparitinases, chondroitinases not only degrade chondroitin sulphate (CS) and dermatan sulphate (DS), but also the protein-free and unsulphated GAG hyaluronan (HA). Thus, disaccharide preparations derived by chondroitinase degradation may be contaminated by HA disaccharides. The latter will often comigrate in HPLC chromatograms with unsulphated disaccharides derived from CS. We have investigated how variation of pH, amount of enzyme, and incubation time affects disaccharide formation from CS and HA GAG chains. This allowed us to establish conditions where chondroitinase degrades CS completely for quantification of all the resulting disaccharides, with negligible degradation of HA, allowing subsequent HA analysis. In addition, we present simple methodology for disaccharide analysis of small amounts of CS attached to a hybrid PG carrying mostly HS after immune isolation. Both methods are applicable to small amounts of GAGs synthesized by polarized epithelial cells cultured on permeable supports.

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.

Ephrin-B3 binds to a sulfated cell-surface receptor.

The ephrins are a family of proteins known to bind the Eph (erythropoietin-producing hepatocellular) receptor tyrosine kinase family. In the present paper, we provide data showing that ephrin-B3 binds a sulfated cell-surface protein on HEK-293T (human embryonic kidney-293 cells expressing the large T-antigen of simian virus 40) and HeLa cells, a binding that is nearly completely blocked by treatment of these cell lines with chlorate or heparinase, or by addition of the heavily sulfated glycosaminoglycan heparin. This indicates that heparan sulfate on these cells is essential for cell-surface binding of ephrin-B3. Heparin did not affect ephrin-B3 binding to EphB receptors expressed on transfected HEK-293T cells, indicating further that ephrin-B3 binds an alternative receptor which is a heparan sulfate proteoglycan. Site-directed mutagenesis analysis revealed that Arg178 and Lys179 are important for heparin binding of ephrin-B3 and also for ephrin-B3 binding to cells. These amino acids, when introduced in the non-heparin-binding ephrin-B1, conferred the heparin-binding property. Functional studies reveal that ephrin-B3 binding to cells induces cellular signalling and influences cell rounding and cell spreading. In conclusion, our data provide evidence for an unknown ephrin-B3-binding cell-surface proteoglycan involved in cellular signalling.

N-Glycan synthesis in the apical and basolateral secretory pathway of epithelial MDCK cells and the influence of a glycosaminoglycan domain

Different classes of glycans are implicated as mediators of apical protein sorting in the secretory pathway of epithelial cells, but recent research indicates that sorting to the apical and basolateral surfaces may occur before completion of glycan synthesis. We have previously shown that a proteoglycan (PG) core protein can obtain different glycosaminoglycan (GAG) structures in the apical and basolateral secretory routes (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) of epithelial Madin-Darby canine kidney (MDCK) cells. We have now also determined the detailed N-glycan structures acquired by a single glycoprotein species in the same apical and basolateral secretory pathways. For this purpose, rat growth hormone (rGH) with two N-glycan sites (rGH-2N) inserted into the rGH portion (NAS and NFT) was fused to green fluorescent protein (GFP) and expressed in MDCK cells. Immunoisolated rGH variants were analyzed for site occupancy and N-glycan structure by mass spectrometry. The extent of NAS and NFT site occupancy was different, but comparable for rGH-2N secreted apically and basolaterally. Microheterogeneity existed for the glycans attached to each N-glycan site, but no major differences were observed in the apical and basolateral pathways. Transfer of the GAG modification domain from the PG serglycin to the fusion site of rGH-2N and GFP allowed polymerization of GAG chains onto the novel protein variant and influenced the microheterogeneity of the N-glycans toward more acidic glycans, but did not alter the relative site occupancy. In conclusion, no major differences were observed for N-glycan structures obtained by the expressed model proteins in the apical and basolateral secretory pathways of epithelial MDCK cells, but insertion of a GAG attachment domain shifted the N-glycans to more acidic structures.