The heparan sulfate on the surface of all adherent cells modulates the actions of a large number of extracellular ligands. Members of both cell surface heparan sulfate proteoglycan families, the transmembrane syndecans and the glycosylphosphoinositide-linked glypicans, bind these ligands and enhance formation of their receptor-signaling complexes. These heparan sulfate proteoglycans also immobilize and regulate the turnover of ligands that act at the cell surface. The extracellular domains of these proteoglycans can be shed from the cell surface, generating soluble heparan sulfate proteoglycans that can inhibit interactions at the cell surface. Recent analyses of genetic defects in Drosophila melanogaster, mice, and humans confirm most of these activities in vivo and identify additional processes that involve cell surface heparan sulfate proteoglycans. This chapter focuses on the mechanisms underlying these activities and on the cellular functions that they regulate.

Meso-scale (unit cell of an impregnated textile reinforcement) finite element (FE) modelling of textile composites is a powerful tool for homogenisation of mechanical properties, study of stress-strain fields inside the unit cell, determination of damage initiation conditions and sites and simulation of damage development and associated deterioration of the homogenised mechanical properties of the composite. Meso-FE can be considered as a part of the micro-meso-macro multi-level modelling process, with micro-models (fibres in the matrix) providing material properties for homogenised impregnated yarns and fibrous plies, and macro-model (structural analysis) using results of meso-homogenisation. The paper discusses stages of the meso-FE analysis and proposes a succession of steps (''road map'') and the corresponding algorithms for it: (1) Building a model of internal geometry of the reinforcement; (2) Transferring the geometry into a volume description (''solid'' CAD-model); (3) Preparation for meshing: correction of the interpenetration of volumes of yarns in the solid model and providing space for the thin matrix layers between the yarns; (4) Meshing; (5) Assigning local material properties of the impregnated yarns and the matrix; (6) Definition of the minimum possible unit cell using symmetry of the reinforcement and assigning periodic boundary conditions; (7) Homogenisation procedure; (8) Damage initiation criteria; (9) Damage propagation modelling. The ''road map'' is illustrated by examples of meso-FE analysis of woven and braided composites.

No abstract

A comparative analysis was carried out of heparan sulfate (HS) and chondroitin sulfate (CS) chains of the ectodomains of hybrid type transmembrane proteoglycans, syndecan-1 and -4, synthesized simultaneously by normal murine mammary gland epithelial cells. Although the HS chains were structurally indistinguishable, intriguingly the CS chains were structurally and functionally distinct, probably reflecting the differential regulation of sulfotransferases involved in the synthesis of HS and CS. The CS chains of the two syndecans comprised nonsulfated, 4-O-, 6-O-, and 4,6-O-disulfated N-acetylgalactosamine-containing disaccharide units and were significantly different, with a higher degree of sulfation for syndecan-4. Functional analysis using a BIAcore system showed that basic fibroblast growth factor (bFGF) specifically bound only to the HS chains of both syndecans, whereas midkine (MK) and pleiotrophin (PTN) bound not only to the HS but also to the CS chains. Stronger binding of MK and PTN to the CS chains of syndecan-4 than those of syndecan-1 was revealed, supporting the structural and functional differences. Intriguingly, removal of the CS chains decreased the association and dissociation rate constants of MK, PTN, and bFGF for both syndecans, suggesting the simultaneous binding of these growth factors to both types of chains, producing a ternary complex that transfers the growth factors to the corresponding cell surface receptors more efficiently compared with the HS chains alone. The involvement of the core protein was also shown in the binding of MK and PTN to syndecan-1, suggesting the possibility of cooperation with the HS and/or CS chains in the binding of these growth factors and their delivery to the cell surface receptors. Proteoglycans (PG) 1 bear glycosaminoglycans (GAGs) such as heparan sulfate (HS) and chondroitin sulfate (CS). The repeating disaccharide unit in the HS and CS backbones isGlcUA/IdceA-GlcNAc and GlcUA-GalNAc, respectively, onto which are superimposed specific modification patterns, most notably the addition of sulfate groups by a variety of sulfotransferases (1). PGs are distributed ubiquitously in extracellular matrices and at cell surfaces (2). Syndecans are the major cell surface PGs expressed by virtually all epithelial cells. Four kinds of syndecans form a gene family, the transmembrane and cytoplasmic domain being conserved among all of the members (3-7). These syndecans are expressed with different cell-, tissue-, and developmental stage-specific patterns (8, 9), suggesting distinct functions for each family member (10), although some shared activities have been observed, for example, for syndecan-1 and -4 (3, 5, 11, 12).The majority of GAG chains added to the core proteins of syndecans are of the HS type, although syndecan-1 (13) and syndecan-4 (14) are modified by CS chains as well. The HS chains bind collagen types I, III, and V (15), fibronectin (16), tenascin (17), thrombospondin (18) and basic fibroblast growth factor (bFGF) (19,20), and other components of the cellula...

We have isolated and sequenced cDNA clones that encode the core protein of PG-M-like proteoglycan produced by cultured mouse aortic endothelial cells (Morita, H., Takeuchi, T., Suzuki, S., Maeda, K., Yamada, K., Eguchi, G., and Kimata, K. (1990) Biochem. J. 265, 61-68). A homology search of the cDNA sequence has suggested that the core protein is a mouse equivalent of chick PG-M(V1), one of the alternatively spliced forms of the PG-M core protein, which may correspond to human versican. Northern blot analysis revealed three mRNA species of 10, 9, and 8 kilobases (kb) in size. The analysis of PG-M mRNA species in embryonic limb buds and adult brain revealed the presence of other mRNA species with different sizes; the one with the largest size (12 kb) was found in embryonic limb buds, and the ones with smaller sizes of 7.5 and 6.5 kb were in adult brain. Sequencing of cDNA clones for the smaller forms in the adult brain showed that they were different from PG-M(V1) in encoding the second chondroitin sulfate attachment domain (CS alpha) alone. Occurrence of the PCR products striding over the junction of the first and second chondroitin sulfate attachment domains suggested that a mRNA of 12 kb in size corresponded to a transcript without the alternative splicing (PG-M(V0)). It is likely, therefore, that multiforms of the PG-M core protein may be generated by alternative usage of either or both of the two different chondroitin sulfate attachment domains (alpha and beta) and that molecular forms of PG-M may vary from tissue to tissue by such an alternative splicing.

No abstract

We showed previously that the alternative splicing of chondroitin sulfate attachment domains (CS alpha and CS beta) yielded multiforms of the PG-M core protein in mouse. A transcript encoding a new short form of the core protein PG-M(V3) was found in various mouse tissues using polymerase chain reaction. DNA sequences of the polymerase chain reaction products suggested that PG-M(V3) had no chondroitin sulfate attachment domain. PG-M(V3) was also detected in various human tissues. The presence of a transcript for PG-M(V3) was further supported by Northern blot analysis. Southern blot analysis confirmed that multiforms of the PG-M core protein, including PG-M(V3), were derived from a single genomic locus by an alternative splicing mechanism. Because PG-M(V3) has no chondroitin sulfate attachment region, which is the most distinctive portion of a proteoglycan molecule, this form may have a unique function.

We previously showed not only the presence of multiple RNA transcripts of different sizes encoding the core protein of mouse PG-M, but also their tissue-dependent expression. Major causes for the multiple forms were found to be due to alternative usage of the two different chondroitin sulfate attachment domains (alpha and beta). In this study, genomic DNA analysis has revealed that these domains are encoded by two large exons, exon VII (2880 base pairs) and exon VIII (5229 base pairs). The splice sites of these two exons were consistent with the occurrence of alternative splicing without frameshift. Furthermore, the mouse PG-M gene was shown to have four distinct polyadenylation signals and three candidates for the transcription initiation site as well. These genomic structural variations may contribute to the multiplicity of PG-M transcripts. Northern hybridization analysis showed that at least three different transcripts were generated by different usage of the distinct polyadenylation signals.