Extracellular matrix proteoglycans and cell-substratum adhesion of human endothelial cells: the effect of methyl beta-D-xylopyranoside

Chondroitin sulphate composition and structure in alternatively spliced CD44 fusion proteins

Previous studies have indicated that CD44 isoforms, spliced with variant exons, are heterogeneously glycanated with chondroitin sulphate and heparan sulphate chains. Because such alternative splicing may regulate divergent biological effects of the specific isoforms, we analysed the consequences of this process on the composition and structure of the chondroitin-sulphate chains. Recombinant chimaeras were engineered with and without exons V3-10 or V3,8-10 and expressed as Ig fusion proteins in COS cells. In addition, the chondroitin sulphates of wild-type isoforms were contrasted with those of isoforms mutated with serine-to-alanine codon substitutions at a putative Ser-Gly-Ser-Gly glycosaminoglycan acceptor site within exon V3. The chondroitin sulphates contained both 4- and 6-sulphated galactosamine residues, although there was a high content of non-sulphated galactosamine-containing repeat units. Splicing of exons V4-7, which contain no Ser-Gly consensus motifs, resulted in increased glycanation with chondroitin-sulphate chains, as well as increased sulphation levels of the polymers. Comparison of wild-type and acceptor-site mutant isoforms showed that chondroitin-sulphate content declined by more than 60-80% in the mutant, indicating that assembly of chondroitin-sulphate chains occurs there, and a general decrease in the sulphation level of the remaining chains was observed. Undersulphation of the recombinant chondroitin sulphates was shown by parallel analyses with native human keratinocyte CD44 molecules and is most probably an artifact of transient expression in COS cells. Our data indicate that combinatorial exon splicing exerts complex and distal effects on glycanation patterns and structure, which presumably modulate those functions that may be mediated though the chondroitin-sulphate moieties, such as motility and matrix invasion.

Extracellular matrix heparan sulfate modulates endothelial cell susceptibility to Staphylococcus aureus

The ability of extracellular matrix heparan sulfate to alter the susceptibility of human endothelial cells to S. aureus was investigated. Endothelial cells grown on extracellular matrix synthesized by S. aureus-infected endothelial cells were more susceptible to subsequent staphylococcal infection than endothelial cells grown on the extracellular matrix synthesized by untreated endothelial cells. Endothelial cells were more susceptible to S. aureus infection when 1) grown on heparitinase-treated extracellular matrix that removed heparan sulfate chains, 2) grown on extracellular matrix produced by chlorate-treated endothelial cells that reduced sulfation in the matrix heparan sulfate proteoglycans, 3) grown on heparan sulfate purified from extracellular matrix elaborated by infected endothelial cells, and 4) endothelial cells were chlorate-treated and therefore expressed desulfated cellular heparan sulfate proteoglycans. Extracellular matrix produced by S. aureus-infected endothelial cells contained heparan sulfate proteoglycans with reduced sulfation. The altered extracellular matrix with reduced sulfated heparan sulfate proteoglycans signalled the uninfected endothelial cells to produce under sulfated cellular heparan sulfate proteoglycans that increased S. aureus adherence to the endothelial cells.

Divergent regulation of proteoglycan and glycosaminoglycan free chain expression in human keratinocytes and melanocytes

Keratinocytes and melanocytes, which together form units of structure and function within human epidermis, are known to differ in expression of autocrine growth factors, particularly those with heparin binding affinity. Because such cytokines could be regulated by the endogenous heparinlike glycosaminoglycan, heparan sulfate, proteoglycan synthesis was compared between human keratinocytes and melanocytes cultured from a common donor. Following steady-state isotopic labeling under conditions of active growth (low density cultures) and growth inhibition (high density cultures), the sulfated polymers were isolated from conditioned media and cell extracts. We found that keratinocytes produced substantially more sulfated glycosaminoglycans than did the melanocytes. There was no evidence for hyaluronic acid synthesis by the melanocytes. The majority of [35S]-sulfate labeling was in the heparan sulfates of the keratinocytes and in the chondroitin sulfates of the melanocytes. During the transition from active growth to growth inhibition, there was increased heparan sulfate proteoglycan and free chain synthesis by keratinocytes but not by melanocytes, and chondroitin sulfate proteoglycan production declined in both cell lineages. The differences may reflect divergent evolution as each cell type came to exploit those complex polysaccharides in different ways to regulate molecular pathways of growth and differentiation. The coupling of growth inhibition with augmented synthesis of heparan sulfates observed for the keratinocytes suggests a regulatory role in growth factor signaling in that cell type.

The effect of the beta-D-xyloside naroparcil on circulating plasma glycosaminoglycans. An explanation for its known antithrombotic activity in the rabbit

Beta-D-Xylosides are known to initiate or prime free glycosaminoglycan (GAG) chain synthesis in cell and tissue culture. As such, the effect of the venous antithrombotic beta-D-xyloside, naroparcil, was investigated on the plasma GAG profile in the rabbit after oral administration. Using dose-response experiments, we showed that antithrombin activity via antithrombin III and heparin cofactor II was increased in parallel with GAG plasma levels compared to control. A more detailed qualitative examination of plasma GAGs by cellulose acetate electrophoresis and ion-exchange chromatography, following oral administration of naroparcil at 400 mg/kg, revealed the presence of higher density charged molecules compared to control. The extracted GAGs were found to activate inhibition of thrombin by heparin cofactor II and contained approximately 25% of a dermatan sulfate-like compound (undetectable in control), which could be responsible for the antithrombotic effect. Using radiolabeled naroparcil, we found radiolabeled GAG fractions and the fact that naroparcil was a substrate for galactosyltransferase I, the second enzyme responsible for GAG chain polymerization, suggested that the compound could initiate in vivo the biosynthesis of antithrombotic free GAG chains. This is, to our knowledge, the first description of the in vivo effect of a beta-D-xyloside on GAG biosynthesis; furthermore, this is correlated with an antithrombotic action.

The effects of beta-D-xyloside on the synthesis of proteoglycans by skeletal muscle: lack of effect on decorin and differential polymerization of core protein-bound and xyloside-linked chondroitin sulfate

Developing skeletal muscle cells, as both myoblasts and myotubules, synthesize a distinctive large chondroitin sulfate proteoglycan. To probe the role of this proteoglycan in myogenesis, chick embryonic muscle cells in culture were treated with beta-D-xyloside, a compound which interferes with proteoglycan synthesis by acting as an artificial acceptor for glycosaminoglycan synthesis and thereby competing with the proteoglycan core protein. Analysis of the proteoglycans indicates that with increasing concentrations of beta-D-xyloside, synthesis of the chondroitin sulfate proteoglycan is inhibited, with concomitant massive synthesis of xyloside-linked chondroitin sulfate glycosaminoglycans. Xyloside does not appear to inhibit glycosaminoglycan attachment onto the small heparan sulfate and dermatan sulfate proteoglycans which are synthesized in the muscle cultures, even though, because of the mechanism of action of beta-xyloside, these proteoglycans should be affected. At submaximal concentrations of beta-xyloside, there is synthesis of both large chondroitin sulfate proteoglycans and xyloside-linked chondroitin sulfate. The xyloside-linked chondroitin sulfate chains have the same sulfation pattern as the core protein-bound skeletal muscle chondroitin sulfate (90% 6-sulfated isomer), but are much smaller (24,000 vs. 65,000 in molecular weight). The discrepancy in size but identify of sulfation indicates that, although sulfation takes place normally on either the core protein or the xyloside acceptor, termination of glycosylation occurs earlier for xyloside-initiated chondroitin sulfate. In spite of these dramatic effects on chondroitin sulfate proteoglycan synthesis, beta-xyloside elicits no observable effects on in vitro myogenesis. This suggests that the function served by the large chondroitin sulfate proteoglycan is not required in the more simplified environment of the muscle cultures.

Proteoglycan and glycosaminoglycan free chain expression in keratinocytes, endothelium, and mesenchymal cells

Cultured fibroblasts, bovine aortic endothelial cells, and human keratinocytes synthesize both proteoglycans and glycosaminoglycan free chains, the proportions varying between cell types. The major metabolic labeling is in proteoglycans, except for keratinocytes with approximately 60% of product as free chains. The proteoglycans range from approximately 50- greater than 1000 kDa, and the glycosaminoglycan side chains derived by alkaline elimination are approximately 30- greater than 100 kDa. The glycosaminoglycan free chains, in contrast, are smaller, from approximately 7-40 kDa in mass. The proteoglycans are both medium and cell layer constituents, whereas the glycosaminoglycan free chains are essentially confined to cells. The cellular proteoglycans and a portion of the free chains are accessible to in situ digestion by Flavobacterial glycosaminoglycan lyases, presumably reflecting localization to the cell surface. Collectively, the data show the free chains to be a common feature of all cells studied and to be partly expressed on cell surfaces. We hypothesize that the processing that creates these free chains occurs on cell surfaces, in which location they could serve ligand receptor functions.

Extracellular matrix derived from Trypanosoma cruzi infected endothelial cells directs phenotypic expression

Infection of confluent human umbilical vein endothelial cells by the parasite Trypanosoma cruzi results in the appearance of an altered heparan sulfate proteoglycan (HSPG) isolated from the extracellular matrix of infected endothelial cells (ECMi). HSPG from ECMi differed from HSPG obtained from the extracellular matrix of uninfected endothelial cells (ECMu) by virtue of an 8-10-fold increase in sulfation and a different elution pattern using DEAE Sepharose chromatography. Analysis of the HSPG that binds to acidic fibroblast growth factor (aFGF) revealed that infection increased the proportion of HSPG that binds to aFGF by 35%. Heparitinase and alkaline borohydride treatment of aFGF-binding HSPG and chromatographic resolution on Sepharose CL4B column revealed an infection-associated 10-fold increase in sulfation of the GAG side chain with no significant change in the migration of the core protein. In addition, the aFGF binding HSPG isolated from ECMi demonstrated a markedly attenuated synergistic mitogenic activity with aFGF in a cell proliferation assay. All of the infection associated changes in HSPG could be demonstrated in HSPG obtained from uninfected endothelial cell cultures grown on ECMi. Hence, the ECMi is associated with signals capable of modulating the ECM associated metabolism of uninfected endothelial cells. This facility of ECMi was also shown to extend to patterns of Gs protein synthesis as revealed by Western blot analysis. The observation that the ECM produced by infected endothelial cells can direct the synthetic patterns of uninfected endothelial cells in a manner uniquely observed in infected endothelial cells suggests a plausible pathway by which infection of only a few cells can ultimately result in the coordinate responses of neighboring uninfected cells.

The distinctive pattern of proteoglycan and glycosaminoglycan free chain synthesis by cultured human epidermal keratinocytes

The in vitro synthesis of proteoglycans and glycosaminoglycan free chains was studied in human epidermal keratinocytes. Preconfluent and confluent cultures established on 3T3 feeders were steady state labeled with [35S]-sulfate and [3H]-glucosamine after removal of the 3T3 cells. Products in nonionic detergent extracts of keratinocytes and in the medium were analyzed in the presence of protease inhibitors. Glycosaminoglycans as proteoglycans and as free chains were defined by susceptibility or resistance, respectively, to alkaline borohydride reduction. Products associated with the cells were approximately 30% proteoglycans and approximately 70% glycosaminoglycan free chains, whereas in the medium virtually all was proteoglycan. The heparan and chondroitin sulfate proteoglycans were small compared to those of many other cell types. Their Kav on Sepharose CL-4B was 0.56 (estimated 50 kDa), whereas the free chain Kav was 0.74 (estimated 12 kDa). Relative amounts of the sulfated products varied with confluence and differentiation; heparan and chondroitin sulfates were equally represented within the free chains and proteoglycans of the cells in preconfluent, proliferating cultures, whereas in postconfluent, differentiated cultures the major labeling was in the heparan sulfate products, consistent with our prior reports (J Invest Dermatol 88:215-9, 1987 and 91:492-8, 1988). The cellular localization of the products was probed with glycosaminoglycan degrading enzymes added to isotopically prelabeled cultures. The proteoglycans appeared to be located on the external surface of plasma membranes, whereas the glycosaminoglycan free chains resisted digestion and are either intracellular or membrane associated, but otherwise inaccessible. These data establish the distinctive pattern of low Mr proteoglycans and abundant cell-associated glycosaminoglycan free chains synthesized by keratinocytes.

Extracellular matrix heparan sulfate proteoglycans modulate the mitogenic capacity of acidic fibroblast growth factor

Confluent cultures of human endothelial cells deposit into extracellular matrix (ECM) distinct heparan sulfate proteoglycans (HSPG) which modulate acidic fibroblast growth factor's (aFGF) ability to stimulate human endothelial cell mitogenic capacity. Extracellular matrix 35S-HSPG were isolated from cultures metabolically labelled with Na235SO4 by DEAE-Sepharose, Sepharose CL-4B, and aFGF-Affi-Gel 15 column chromatography and identified by resistance to chondroitinase ABC and sensitivity to nitrous acid. Fifty to sixty percent of the 35S-HSPG deposited into ECM do not bind aFGF. The bound 35S-HSGP (40-50% of the total counts applied) eluted from the aFGF-Affi-Gel column after the addition of buffer containing 2 M NaCl. aFGF-binding and aFGF-nonbinding 35S-HSPG were individually pooled and further purified by Sepharose CL-4B column chromatography. 35S-HSPG which bind aFGF, designated HSPGP, were 100-fold superior to heparin in augmenting the mitogenic efficacy of aFGF in sparse proliferating cultures. In contrast, however, 35S-HSPG, which did not bind aFGF, designated HSPG1, inhibited aFGF-stimulated proliferation in both sparse and subconfluent endothelial cell cultures. The majority of the biological activity of both aFGF-potentiating HSPGP and aFGF-inhibitory HSPG1 was contained in the glycosaminoglycan chains released by alkaline borohydride treatment of intact HSPGP or HSPG1, respectively. 3H-Core protein derived from HSPGP or HSPG1 contained only minor biological activity. The ability of heparitinase or heparinase (Flavobacterium heparinum) to abolish biological activity differed, depending upon the HSPG tested, also suggested that these are two distinct HSPGs.

Role of proteoglycans in renal development

The role of proteoglycans (PGs) in morphogenesis was investigated. Fetal kidneys were obtained from 13-day-old mouse embryos and maintained for 7 days in culture. The biosynthesis of PGs was perturbed by addition of p-nitrophenyl-beta-D-xylopyranoside in the culture medium. The kidneys were processed for morphological and biochemical studies. The morphological studies included staining of tissues with anti-basement membrane antibodies and ruthenium red. [35S]sulfate was used as the precursor product for biosynthetic and autoradiographic studies. The kidneys treated with xyloside had loose mesenchyme, inhibition of ureteric bud branching, diminution in the population of developing nephron elements, decreased immunofluorescence with anti-proteoglycan antibodies and staining with ruthenium red, and a reduced [35S]sulfate incorporation into poorly organized extracellular matrices. The biochemical studies included characterization of PGs/glycosaminoglycans (GAGs) by Sepharose CL-4B, -6B, and DEAE-Sephacel chromatographies and cellulose acetate electrophoresis. Under the influence of xyloside, the total radioactivities decreased 2 to 4-fold in tissues and increased 18 to 42-fold in media fractions. A reduction in the size of macromolecular form of PGs, i.e., from MW approximately 2.5 X 10(6) to approximately 2.5 X 10(4), was noted. The PGs/GAGs synthesized were mainly made up of heparan sulfate and small amounts of chondroitin sulfate. They eluted at a lower salt concentration as compared to the controls. A similar diminution in the size of media PGs, i.e., from MW approximately 1.8 X 10(5) to approximately 2.8 X 10(4), was observed. Additional studies with [3H]xyloside indicated that the chains initiated on xyloside residues were similar in size and composition to GAG-chains. These findings indicate that a perturbance in the biosynthesis of PGs/GAGs leads to abnormalities in renal organogenesis.

Inhibition of endothelial cell morphogenetic interactions in vitro by alpha- and beta-xylosides

Bovine aortic endothelial cells retain the ability to undergo histiotypic morphogenetic interactions in vitro as evidenced by a) the reversible expression of a sprouting cell phenotype and b) the patterned self-association of these sprouting cells into three-dimensional meshworks and tubule-like structures. These morphogenetic events are inhibited by xylosides in a dose-dependent manner. Two types of beta-xylosides (p-nitrophenyl-beta-D-xylopyranoside and 4-methylumbelliferyl-beta-D-xylopyranoside) and one alpha-xyloside (p-nitrophenyl-alpha-D-xylopyranoside) were tested. beta-xylosides are well characterized acceptors of glycosaminoglycan chains, whereas alpha-xylosides do not function in this capacity and have been extensively used as negative controls when studying the effects of beta-xylosides. Both alpha- and beta-xylosides inhibited endothelial morphogenetic interactions. This inhibition was slowly reversed during the 6- to 7-d period following removal of the xyloside. Inhibition of morphogenetic interactions by xylosides occurred at concentrations (0.5 to 2.0 mM) that had no demonstrable effects on cell proliferation, migration, or adhesion to 2-D plastic or collagen substrata. The xylosides seemed to inhibit cell spreading on a 3-D environment, they also inhibited the incorporation of [3H]-proline and Na2 35SO4 into the extracellular matrix deposited by the cells, suggesting that the inhibition of morphogenesis may be related to the inhibition of matrix deposition. Endothelial morphogenetic interactions were not inhibited by the extracellular matrix or by the conditioned medium produced by cells cultured in the presence of xylosides.