Effects of Glycosaminoglycans on Cell Proliferation of ... - CiteSeerX

5 downloads 0 Views 140KB Size Report
chondroitin sulphate (CS), dermatan sulphate (DS), heparan sulphate (HS) .... lyase II (heparinase II, no EC number) from Flavobacterium heparinium were ... and AC and analysis of ¢-disaccharides by ion-suppression HPLC and reversed ...
ANTICANCER RESEARCH 25: 2851-2856 (2005)

Effects of Glycosaminoglycans on Cell Proliferation of Normal Osteoblasts and Human Osteosarcoma Cells Depend on their Type and Fine Chemical Compositions D. NIKITOVIC1, A. ZAFIROPOULOS1, G.N. TZANAKAKIS1, N.K. KARAMANOS2 and A.M. TSATSAKIS1 1Departments 2Department

of Histology and Toxicology, School of Medicine, University of Crete, 71003 Heraklion; of Chemistry, Laboratory of Biochemistry, University of Patras, 26110 Patras, Greece

Abstract. Osteoblastic cells produce a complex extracellular matrix (ECM) composed of a mixture of proteoglycans (PGs), collagens and non-collagenous proteins. The interaction of proteoglycans with matrix effector macromolecules via either their glycosaminoglycan (GAG) chains or their protein core is critical in regulating a variety of cellular events. Alterations in the structural composition of the GAG/PG component of the ECM may have important consequences on cell proliferation and / or differentiation. Human osteoblasts and two osteosarcoma cell lines, able to produce galactosaminoglycan (GalAGs) and heparan sulphate (HS)-containing proteoglycans, were treated with their main GAG chain types, and the effects on cell growth were examined. Chondroitin sulphate (CS∞) and dermatan sulphate (DS) inhibited cell proliferation of all osteoblastic cell lines at high concentration (100 Ìg/ml). DS showed the stronger

Abbreviations: GlcA, D-glucuronic acid; GlcN, D-glucosamine; ¢dinonSHA, 2-acetamido-2-deoxy-3-O-(4-deoxy-·-L- threo-hex-4enopyranosyluronic acid)-D-glucose; ¢di-nonSCS, 2-acetamido-2deoxy-3-O-(4-deoxy-·-L- threo-hex-4-enopyranosyluronic acid)-Dgalactose; ¢di-mono4S, ¢di-mono6S, ¢di-monoNS, monosulphated ¢-disaccharides at C-4, C-6 and the amino group of hexosamine, respectively; ¢di-mono2S, monosulphated ¢-disaccharides at C-2 of uronic acid and C-4 of hexosamine, respectively; ¢di-di(2,N)S and ¢di-di(6,N)S, disulphated ¢-disaccharides at C-2 of uronic acid and C-4 of hexosamine and at C-2 and C-6, respectively: ¢di-di(2,N)S and ¢di-di(6,N)S, disulphated ¢-disaccharides at C-2 of uronic acid and the amino group of hexosamine and at C-6 and the amino group, respectively; ¢di-tri(2,4,6)S, trisulphated ¢-disaccharides at C-2 of uronic acid and at C-4 and C-6 of hexosamine. Correspondence to: Dr. A.M. Tsatsakis, Departments of Histology and Toxicology, School of Medicine, University of Crete, 71003 Heraklion, Greece. e-mail: [email protected] or to Dr. N. Karamanos, Department of Chemistry, Laboratory of Biochemistry, University of Patras, 26110 Patras, Greece. e-mail: [email protected] Key Words: Osteosarcoma, glycosaminoglycans, cell proliferation.

0250-7005/2005 $2.00+.40

inhibitory effect, probably due to the presence of flexible IdoA residues that provide a greater variety in conformation to these macromolecules. Heparin strongly inhibited the proliferation rates of both normal osteoblasts and transformed osteoblastic cells at concentrations ≥1 Ìg/ml. The presence of large amounts of IdoA-derived trisulphated disaccharides, responsible for the overall negative charge of heparin, should be considered as a critical factor for the inhibition of cell proliferation. The obtained results suggest that matrix GAGs are factors which affect cell growth of both malignant and normal cells of the osteoblastic lineage in a concentration-dependent manner. This effect is closely related to the fine chemical structure of GAGs, i.e. the presence of L-iduronic acid and the degree of sulphation. The control of cellular proliferation is closely related to physiological processes, such as cell differentiation, tissue morphogenesis, repair and angiogenesis. Growth factors/mitogens and growth inhibitors, whose effects can be modulated by interactions with the extracellular matrix (ECM), regulate cell growth (1). The matrix is constructed from a variety of fibrous proteins, their respective cell surface receptors, free glycosaminoglycans (GAGs) such as hyualuronan and complex GAGs covalently bound into a protein core and termed proteoglycans (PGs) (2). PGs can be substituted with different types of sulphated GAGs, including chondroitin sulphate (CS), dermatan sulphate (DS), heparan sulphate (HS), heparin or keratan sulphate. Due to the presence of various sulphate esterified groups in the GAG chains, PGs are highly negatively-charged macromolecules (3). Via specific interactions of GAGs with growth factors or acting directly as growth factors receptors, PGs may modulate and modify growth factor activities (4, 5). Exogenously added GAGs can exert growth-regulatory effects on cultured normal or transformed cells. Thus, L-iduronic acid (IdoA)-containing glucosaminoglycans (GlcAGs), i.e. heparin and heparan sulphate (HS), inhibit the proliferation of several cell lines in vitro, such as epithelial cells (6), mesangial cells (7), human

2851

ANTICANCER RESEARCH 25: 2851-2856 (2005) lung fibroblasts (8) and mesenhymal cells (9). The IdoAcontaining galactosaminoglycan, dermatan sulphate (DS), inhibits the proliferation of human lung fibroblasts (8), whereas it stimulates U-937 leukemia cells (10). CS, a nonIdoA-containing galactosaminoglycan (GalAG), effectively stimulates both human lung fibroblasts (8) and U-937 leukemia cells (10). Furthermore, a study by Volpi et al. (11) revealed that leukemia cell proliferation depended on the level of CS sulphation. Syrokou et al. (12) showed that matrix GAGs affect the growth of malignant mesothelioma cells with epithelial and fibroblast-like morphology in a manner that depends on their concentration, on the fine structure of the GAG chain, as well as on cell phenotype. It seems, therefore, that the effects of GAGs on cell proliferation are closely related to the structure of the GAG chain, the level of sulphation and on the cell type (6, 8). Human bone cells in culture produce a large versican-like PG, a heparan sulphate syndecan-like PG, as well DScontaining biglycan and decorin (13, 14). Pathogenesis of osteosarcoma implicates qualitative and quantitative changes in the PG component of the ECM (15). A recent study has shown an absence of the biglycan transcript in osteosarcoma samples, which is in accordance with the non-mineralized osteoid produced by the osteosarcoma cells (15). We have previously reported that the MG-63 and Saos 2 osteosarcoma cell lines differ both in their relative biosynthetic patterns and GAG/PG distribution among culture medium and cell membrane (16). In the present study, the in vitro effects of the main types of GAGs, i.e. CS, DS and heparin, and the relationship of their fine sulphation pattern on the proliferation of normal human osteoblasts and two human osteosarcoma cell lines, are reported.

Cell cultures. MG-63, Saos 2 and normal osteblastic (OB)(17) cells were grown at 37ÆC in a humidified atmosphere of 5% (v/v) CO2, in DMEM supplemented with 10% FBS, 4 mM L-glutamine, 2 g/l sodium bicarbonate, 100 IU/ml penicillin and 100 Ìg/ml steptomycin. The cultures were performed in 75-mm 2 tissue culture flasks and the culture medium was changed every other day. Confluent cultures, after being washed with phosphatebuffered saline (PBS), were harvested by trypsinazation with 0.23% (w/v) trypsin in PBS containing 0.1% (w/v) Na2 EDTA for 15 min at 37ÆC and 1 vol. of medium was added to terminate enzymic activity. Cells were collected by centrifugation at 200 x g for 5 min, and their number was measured by suspension in Hanks’ balanced solution, using a Coulter particle-counter (Hialeah, FL, USA). Cell viability was determined by trypan blue dye exclusion. Fine chemical structure of GAGs Determination of IdoA content: The IdoA to GlcA molar ratios were determined by the reversed phase HPLC method of Karamanos et al. (18), following stoichiometric reduction of GAGs with NaBH4, hydrolysis with 2M trifluoroacetic acid at 100ÆC for 8 h and perO-benzoylation of the obtained 1,6-anhydro-L-idose and D-glucose. Derivatives were detected and recorded at 230 nm. Disaccharide composition: The sulphation pattern of CSA and DS was determined following treatments with chondroitinases ABC and AC and analysis of ¢-disaccharides by ion-suppression HPLC and reversed polarity high performance capillary electophoresis (HPCE) (19, 20). On the other hand, the disaccharide composition of heparin was estimated following degradation of the chains with an equi-unit mixture of heparin lyases I, II and III (0.05 units per 25 Ìg of uronic acid) and analysis of ¢-disaccharides by ion-pair HPLC and HPCE (21, 22). Other analytical assays. N-Sulphated hexosamine was determined using the low-pH nitrous acid procedure (23) and N-acetylated hexosamine by the method of Tsuji et al. (24). By applying cellulose acetate membrane electrophoresis (25) and gel-permeation HPLC (26, 27), GAG purity was tested.

Materials and Methods Materials. Chondroitin sulphate type A (CSA) (mainly C-4sulphated CS) from bovine trachea, dermatan sulphate (DS) from porcine skin, heparin from porcine intestinal mucosa and heparin lyase II (heparinase II, no EC number) from Flavobacterium heparinium were obtained from Sigma Chemical Co. (St Louis, MO, USA). Standard preparations of CS/DS and HS-derived disaccharides, i.e. 1¢di-nonSCS, ¢di-nonSHS, ¢di-mono4S, ¢di-mono6S, ¢di-mono2S, ¢di-di(2,4)S, ¢di-di(4,6)S, ¢di-diSB, ¢di-diSD, ¢di-diSE, ¢di-tri(2,4,6)S and ¢di-di-triS came from Seikagaku Kogyo Co. (Tokyo, Japan). Chondroitinase ABC from Proteus vulgaris (EC 4.2.2.4), chondroitinase AC II from Arthrobacter aurescens (EC 4.2.2.5), chondroitinase B (no EC number), chondro-4-(EC 3.1.6/9) and –6-sulphatases (EC 3.1.6.10) from Proteus vulgaris and heparin lyases I and III from Flavobacterium heparinium (EC 4.2.2.7 and EC 4.2.2.8, respectively) were also purchased from Seikagaku Kogyo Co. Fetal bovine serum (FBS), D minimal essential medium (DMEM), penicillin, streptomycin and L-glutamine were all obtained from Biochrom KG (Berlin, Germany). [Methyl]-3H thymidine was purchased from Moravek Biochemicals, USA. All other chemicals used were of the best commercially available grade.

2852

Proliferation assay effects of GAGs. Growing cells from nonconfluent cultures were harvested and seeded in 24-well plates (Costar, USA) at a density of 20x103 cells per well in 1 ml of DMEM. The cells were allowed to rest overnight and the medium was replaced with fresh medium supplemented with heparin, CSA or DSB at concentrations of 1, 10 and 100 Ìg/ml. After 30 h of incubation [methyl]-3H thymidine was added up to the final concentration of 0.15 mCi/ml, and the cells were incubated for an additional 18 h. The cells were washed with PBS and fixed with icecold trichloroacetic acid (5% w/v) for 10 min. Subsequently the plates were washed gently with running tap water and air-dried. The DNA was solubilized by adding 0.2 ml of 1% (w/v) SDS in 0.3 M NaOH under continuous shaking for 15 min. The solubilized DNA was transferred to vials containing 2 ml of scintillation cocktail Lumagel Safe (Lumac, The Netherlands) and radioactivity was measured in a liquid scintillation spectrometer (Beckman Instruments, USA). Statistical analysis. The dependence of cell proliferation on the presence of various GAGs was evaluated using the t-test and the one-way completely randomised variance analysis (ANOVA) using the Microcal Origin (version 5.0) software.

Nikitovic et al: Exogenous Glycosaminoglycans Affect Osteoblastic Cell Proliferation

Table I. Fine chemical composition and properties of GAGs estimated after specific enzymic degradation and HPLC / HPCE analyses. Components/properties Uronic acida IdoA / GlcA Hexosaminesb N-acetylated N-sulphonylated Disaccharide compositionc ¢di-nonS ¢di-mono4S ¢di-monoNS ¢di-mono6S ¢di-mono2S ¢di-di(2,N)S ¢di-di(2,4)S ¢di-di(2,6)S ¢di-di(6,N)S ¢di-di(4,6)S ¢di-tri(2,6,N)S

CSA

DS

Heparin

0/100

71/29

67/33

100 ND

100 NDd

13.0 87.0

5 68 17 ≤1 7 2 -

2 71 1 24 2 -

≤0.5 ≤0.5 6 2 5 6 12 68

Results are expressed as a percentage of: atotal uronic acid, btotal hexosamine and ctotal ¢-disaccharides recovered by HPLC and HPCE analyses. dNot detected.

the proliferation of cells cultured in the absence of GAGs with that in the presence of various GAGs at different concentrations. The results were expressed as a percentage of [3H]-thymidine incorporation in relation to controls (mean±SD, n=15). Viability tests showed that viability exceeded 92%. CSA had no statistically significant effect on Saos 2 cell proliferation at any of the concentrations tested (Figure 1A). The proliferation of MG-63 and OB cells was similarly not affected by CSA when up to 10 Ìg/ml was used. This GAG, however, inhibited the growth of these cells at the very high CSA concentration (100 Ìg/ml) (Figure 1A). DS showed an inhibitory effect on Saos 2 and OB cell proliferation at the high concentration (100 Ìg/ml), while at lower concentrations no significant effect was observed. MG-63 cells, however, were strongly inhibited (up to 80%) even at 10 Ìg/ml DS (Figure 1B). Heparin caused the strongest suppression of cell proliferation as compared to the GalAGs tested (Figure 1C). Even at 1 Ìg/ml, the inhibition ranged from 50% for OB up to 80% for MG-63 cells. When heparin was used at 10 Ìg/ml and higher concentrations, the inhibition of proliferation for the three cell lines was almost complete (~90%).

Discussion Results Purity and chemical composition of GAGs. Upon applying the cellulose acetate membrane method (25), all tested GAGs migrated as homogenously-charged populations, without contaminants. Furthemore, gel-permeation HPLC (26, 27) revealed that all the tested GAGs eluted as homogenously-charged populations (data not shown). The chemical compositions of the tested GAGs are summarized in Table I. In DS, IdoA accounts for 71% of the total uronic acid, whereas in CSA only glucuronic acid (GlcA) was detected. Galactosamine (GalN) was found to be exclusively N-acetylated in both types of GalAGs, i.e. DS and CSA, used. HPLC analysis after treatment with chondroitinases showed that the main dissacharide unit in both DS and CSA was ¢di-mono4S (71% and 68%, respectively). Cleavage of heparin with heparin lyases I, II and III in combination and dissacharide analysis by HPLC and HPCE (21, 22) revealed that the trisulphated disaccharide ¢ditri(2,6 N)S was the major dissacharide unit. Furthermore, the majority of uronic acid (67%) is IdoA and most of glucosamine (GlcN) (87%) is N-sulphonylated. Effects of GAGs on proliferation of normal and transformed osteoblasts. The effect of GAGs on the proliferation of normal osteoblasts and MG-63 and Saos 2 osteosarcoma cell lines cultured in DMEM was evaluated by comparing

The human osteosarcoma cell lines studied (MG-63 and Saos 2) produce HA, secreted and cell-associated GalAGs, i.e. CS and/or DS, as well as secreted and cell-associated HS in amounts varying with cell phenotype (16). Human bone cells in culture produce a large versican-like PG containing CS chains, a heparan sulphate syndecan-like PG, as well as the GalAGs-containing proteoglycans biglycan and decorin (13, 14). GAGs are highly heterogenous macromolecules, their heterogeneity being derived from the dissacharide unit that constitutes the primary structure. Differences are found in the types of uronic acid and hexosamine, in the number and position of the sulphate residues, in the presence of N-acetyl and/or N-sulphate groups and in the relative molecular mass. The variations in primary structure produce molecules with various chemical and biological properties. It is known that GAGs participate in the regulation of several cellular events, including cell proliferation, adhesion and migration. In this study, we demonstrated a strong inhibition in proliferation rates of both normal osteoblasts and transformed osteoblastic cells at heparin concentrations ≥1 Ìg/ml. Heparin has been reported to exert a significant antiproliferative effect on several cell types (6, 11, 12). These studies revealed that the high contents of IdoA and N-sulphonylated GlcN are essential components in heparin, correlated to its ability to inhibit cell proliferation. Furthermore, heparin has been

2853

ANTICANCER RESEARCH 25: 2851-2856 (2005)

Figure 1. Effects of CSA (A), DS (B) and heparin (C) on the cell proliferation, determined as [3H]-thymidine incorporation, of the osteosarcoma MG63 and Saos 2 cells and normal OB osteoblasts. Results are presented as percentage of [3H]-thymidine incorporation with respect to control (100%). The results are the average of five experiments in triplicate (p≤0.005).

2854

Nikitovic et al: Exogenous Glycosaminoglycans Affect Osteoblastic Cell Proliferation

reported to have both inhibitory and promoting effects on the proliferation of osteoblastic cells (28-30). These conflicting descriptions appear mainly to be related to different experimental conditions. The discrepancies could be explained by the variations in the commercially available GAG preparations. Therefore, we employed a detailed biochemical characterization prior to measurement of cell growth. The inhibitory effect of heparin, demonstrated for all tested osteoblastic cell lines in this study, correlates with the presence of high amounts of IdoA-derived trisulphated disaccharides, responsible for the overall negative charge of this GlcAG. Thus the presence of high amounts of IdoAderived trisulphated disaccharides as compared to GlcAderived disaccharides appears to be a critical factor in the inhibition of cell proliferation (6, 12, 31). Indeed, a recent study correlated specific fine structure and the sulphation pattern of heparin with its capacity to modulate bone morphogenetic protein dependent osteoblastic differentiation and proliferation in mesenchymal cells (9). Heparin, which is not synthesized by osteoblasts but only by connective-tissue type mast cells, has wide clinical use in the prevention and treatment of thromboembolic disease, and is associated with an increased risk of osteoporosis (32). Due to a shared biosynthetic pathway, heparin is structurally related to heparan sulphate chains (33, 34). Thus, heparin interferes with the ligand-binding activities of heparan sulphate chains by competing for binding sites on heparinbinding proteins, interfering with the normal growth factor responses of osteoblastic cells (35). CSA showed no inhibitory effect on osteoblastic cell proliferation for a concentration up to 10 Ìg/ml. The observed inhibitory effects of CSA at the very high concentrations, most pronounced in the case of MG-63 cells (30% at 100 Ìg/ml), may be due to its indirect effects and/or cytotoxicity and were not further examined. DS exerts a stronger inhibitory effect at 10 Ìg/ml only for MG-63 cells. The effect of DS is probably due to the presence of flexible IdoA residues that provide a greater variety in conformation to these macromolecules, which may be critical for MG-63 osteosarcoma cells. Relating the biological activities of macromolecules to their type and sulphation pattern, the major disaccharide unit [-IdoAGalNAc-(4-O-SO3)-] of the DS chain (71%) correlates with increased inhibitory activity. Interestingly, the iduronate contents of the DS-containing osteoblast PGs, biglycan and decorin, increase linearly with donor age (14) and may be correlated to their decreased proliferative capacity. An important role for GAG chains is the formation of complexes with growth factors, whereby they regulate growth factor activities by protecting them from proteolytic cleavage or by presenting them to their receptors (36). Osteoblasts produce an ECM rich in GAG chains. Changes in the composition of this ECM are, therefore, able to effect the fine regulation of proliferation or even influence the cell phenotype.

An important finding of this study was that both normal and transformed cells of the osteoblastic lineage showed a similar proliferation response to exogenously added GAGs. This indicates that GAGs may exert their influence on cells of the osteoblastic lineage through common pathways, where, however, the presence of IdoA-residues and sulphation type seem to play critical roles.

References 1 Galagher JT: The interaction and regulation of basic and acidic fibroblast growth factors by heparan sulfate. Trends Glycosci Glycotechnol 10: 137-144, 1998. 2 Kjellen L and Lindahl U: Proteoglycans: structures and interactions. Annu Rev Biochem 60: 853-856, 1991. 3 Prydz K and Dalen KT: Synthesis and sorting of proteoglycans. J Cell Science 113: 193-205, 2000. 4 Wight TN, Kinsell MG and Qwarnstroem E: The role of proteoglycans in cell adhesion, migration and proliferation. Curr Opin Cell Biol 4: 793-801, 1992. 5 Tzanakakis GN, Karamanos NK and Hjerpe A: Effects on glycosaminoglycan synthesis in cultured malignant mesothelioma cells of transforming, epidermal and fibroblast growth factors and their combinations with platelet-derived growth factor. Exp Cell Res 220: 130-137, 1995. 6 Wright TC, Castellot JJ, Diamond JR and Karnovsky MJ: Structural determinants of heparin’s growth inhibitory activity. Interdependence of oligosaccharide size and charge. J Biol Chem 264: 1534-1542, 1989a. 7 Castellot JJ, Wonk K, Herman B, Hoover RL, Albertini DF and Wright TC: Binding and internalization of heparin by vascular smooth muscle cells. J Cell Physiol 124: 21-28, 1985. 8 Westergren-Thorsson G, Persson S, Onnervik PO, Malmstrom A and Fransson LA: L-iduronate-rich glycosaminoglycans inhibit growth of normal fibroblasts independently of serum or added growth factors. Exp Cell Res 206: 93-99, 1993. 9 Takada T, Katagiri T, Ifuku M, Morimura N, Kobayashi M, Hasegawa K, Ogamo A and Kamijo R: Sulfated polysaccharides enhance the biological activities of bone morphogenetic proteins. J Biol Chem 278: 43229-43235, 2003. 10 Volpi N, Bolognani L, Conte A et al: Effects of glycosaminoglycans on U-937 leukaemia cell proliferation and differentiation. Exp Cell Res 215: 119-130, 1994. 11 Volpi N, Bolognani L, Conte A and Petrini M: Effects of chondroitin sulphates with different structures on leukemia cells: U-937 cell proliferation and differentiation. Leukemia Res 17: 789-798, 1993. 12 Syrokou A., Tzanakakis G N, Tsegenidis T, Hjerpe A and Karamanos N: Effects of glycosaminoglycans on proliferation of epithelial and fibroblast human malignant mesothelioma cells: a structure-function relationship. Cell Prolif 32: 85-99, 1999. 13 Beresford JN, Fedarko NS, Fisher LW, Midura RJ, Yanagashita M, Termine JD and Robey PG: Analysis of the proteoglycans synthesized by human bone cells in vitro. J Biol Chem 262: 17164-17172, 1987. 14 Grzesik WJ, Frazier CR, Shapiro JR, Sponseller PD, Gehron Robey P and Fedarko NS: Age-related changes in human bone proteoglycan structure. J Biol Chem 277: 43638-43647, 2002.

2855

ANTICANCER RESEARCH 25: 2851-2856 (2005) 15 Benayahu D, Shur I, Marom R, Meller I and Issakov J: Cellular and molecular properties associated with osteosarcoma cells. J Cellular Biochem 84: 108-114, 2001. 16 Nikitovic D, Tsatsakis AM., Karamanos NK and Tzanakakis GN: The effects of genistein on the synthesis and distribution of glycosaminoglycans/proteoglycans by two osteosarcoma cell lines depends on tyrosine kinase and the estrogen receptor density. Anticancer Res 23: 459-464, 2003. 17 Kletsas D, Basdra EK and Papavassiliou AG: Effect of protein kinase inhibitors on the stretch-elicited c-Fos and c-Jun upregulation in human PDL osteoblast-like cells. J Cell Physiol 190: 313-321, 2002. 18 Karamanos NK, Hjerpe A, Tsegenidis T, Engfeldt B and Antonopoulos CA: Determination of iduronic acid and glucuronic acid in glycosaminoglycans after stoichiometric reduction and depolarization using high-performance liquid chromatography and ultraviolet detection. Anal Biochem 172: 410-419, 1988. 19 Karamanos NK, Syrokou A, Vanky P, Nurminen M and Hjerpe A: Determination of twenty-four variously sulfated galactosaminoglycans and hyaluronan-derived disaccharides by high performance liquid chromatography. Anal Biochem 221: 189199, 1994. 20 Karamanos NK, Axelsson S, Vanky P, Tzanakakis GN and Hjerpe A: Determination of hyaluronan and galactosaminoglycan disaccharides by high performance capillary electrophoresis at the attomole level. Applications to analyses of tissue and cell culture proteoglycans. J Chromatogr A 696: 295-305, 1995. 21 Karamanos NK, Vanky P, Tzanakakis GN and Hjerpe A: Highperformance capillary electrophoresis method to characterize heparin and heparan sulphate disaccharides. Electrophoresis 17: 391-399, 1996a. 22 Karamanos NK, Vanky P, Tzanakakis GN, Tsegenidis T and Hjerpe A: Ion-pair high-performance liquid chromatography for detecting disaccharide composition in heparin and heparan sulphate. J Chromatogr A 765: 169-179, 1997. 23 Shively JE and Conrad HE: Formation of anhydro-sugars in the chemical depolymerization of heparin. Biochemistry 15: 39323942, 1976. 24 Tsuji A, Kinoshita T and Hoshino M: Analytical chemical studies on amino sugars. II. Determination of hexosamines using 3-methyl-2-benzothiazolone hydrazone hydrochloride. Chem Pharm Bull 17: 1505-1510, 1969. 25 Hata R and Nagai T: A rapid and micro method for separation of acidic glycosaminoglycans by two-dimensional electrophoresis. Anal Biochem 45: 462-468, 1972.

2856

26 Engfeldt B, Hjerpe A, Mengarelli S, Reinholdt FP and Wikstrom B: Morphological and biochemical analysis of biopsy specimens in disorders of skeletal development. Acta Paediat Scand 45: 353-363, 1982. 27 Arvaniti A, Karamanos NK, Dimitracopoulos G and Anastassiou ED: Isolation and characterization of a novel 20kDa sulfated polysaccharide from the extracellular slime layer of Staphylococus epidermidis. Arch Biochem Biophys 308: 432438, 1994. 28 Kock HJ and Handschin AE: Osteoblast growth inhibition by unfractionated heparin and by low molecular weight heparins: an in vitro investigation. Clin Appl Thromb Hemost 8: 251255, 2002. 29 Matziolis G, Erli HJ, Rau HM, Klever P, Bosserhoff AK and Paar O: Heparin facilitates proliferation of human osteoblasts in vitro. Orthopade 31: 575-581, 2002. 30 Hausser HJ and Brenner RE: Low doses and high doses of heparin have different effects on osteoblast-like Saos-2 cells in vitro. J Cell Biochem 91: 1062-1073, 2004. 31 Petitou M, Lormeau JC, Choay L and Recherche S: Chemical synthesis of glycosaminoglycans: new approaches to antithrombotic drugs. Nature 350: 30-33, 1991. 32 Nelson-Piercy C: Heparin-induced osteoporosis: Scand J Rheumatol Suppl 107: 68-71, 1998. 33 Lindahl U: ‘Heparin’ – from anticoagulant drug into the new biology. Glycoconjugate J 17: 597-605, 2000. 34 Rabenstein DL: Heparin and heparan sulfate: structure and function. Nat Prod Rep 19: 312-331, 2002. 35 Lindahl U: What else can "heparin’ do? Haemostasis 29(suppl 1): 38-47, 1999. 36 Saksela O, Moscatelli D, Sommer A and Rifkin DB: Endothelial cell-derived heparan sulphate binds basic fibroblast growth factor and protects it from proteolytic degradation. J Cell Biol 107: 743-751, 1988.

Received December 2, 2004 Accepted April 14, 2005