Regulation of the chondroitin/dermatan fine structure ...

Glycobiology vol. 15 no. 12 pp. 1277–1285, 2005 doi:10.1093/glycob/cwj027 Advance Access publication on August 23, 2005

Regulation of the chondroitin/dermatan fine structure by transforming growth factor-␤1 through effects on polymer-modifying enzymes

5 Department of Cell and Molecular Biology, and 6Physiological Sciences, Lund University, BMC B11, S-221 84 Lund, Sweden

Received on March 22, 2005; revised on July 6, 2005; accepted on August 5, 2005

The chondroitin/dermatan sulfate proteoglycans (CS/DSPGs), biglycan, decorin, and versican play several important roles in extracellular matrix influencing matrix organization, cell proliferation, and recruitment. Moreover, they bind and regulate growth factors in the extracellular matrix. We have previously shown that cultured human lung fibroblasts treated with transforming growth factor-␤ (TGF-␤) alone or in combination with epidermal growth factor and platelet-derived growth factor, increase the production of these PGs. In this report, we describe that the structure of their galactosaminoglycan side chains is altered, albeit there is no alteration of polysaccharide length. The findings showed that iduronic acid content is reduced by 50% in decorin and biglycan, whereas 4-O-sulfation is increased 2-fold in versican. To unravel the mechanism behind these changes, the activities of chondroitin C-5 epimerase and of O-sulfotransferases in cellular fractions prepared from fibroblasts were quantitated, and transcript levels of the relevant sulfotransferases were measured by real time polymerase chain reaction (RT–PCR). The C-5 epimerase activity was reduced by 25% in TGF-␤1 treated cells and 50% in fibroblasts treated with the growth factor combination. No change in activity in dermatan 4-O sulfotransferase was observed, and only a minor decrease in dermatan 4-Osulfotransferase-1 (D4ST-1) mRNA was observed. On the other hand, chondroitin 4-O sulfotransferase activity increased 2-fold upon TGF- 1 treatment and 3-fold upon treatment with the growth factor combination. This is in agreement with a 2-fold up-regulation of chondroitin-4-O-sulfotransferase 1 (C4ST-1) mRNA, and no changes in chondroitin-4-O-sul1These authors contributed equally to this work. 2To whom correspondence should be addressed; e-mail:

[email protected] 3Present address: Department of Anatomy and Cell Biology, McGill University, 3640, University Street, Montreal, Québec, Canada H3A 2B2 4Present address: The Burnham Institute, Program for Glycobiology and Carbohydrate Chemistry, 10901 North Torrey Pines Road, La Jolla, CA 92037

fotransferase 2 (C4ST-2) mRNA. Thus, cellular activity and transcript level correlated well with the changes in the structure of the dermatan/chondroitin sulfate chains. Key words: chondroitin/dermatan/glucuronyl c5-epimerase/ sulfotransferase/glycosaminoglycan Introduction Our understanding of the roles of proteoglycans (PGs) in processes at the cell surface and in the extracellular matrix has increased vastly during the last decade (Iozzo, 1998; Bernfield et al., 1999). Fibroblasts synthesize at least three different chondroitin sulfate/dermatan sulfate PGs (CS/ DSPGs), the small leucine-rich repeat PGs, decorin and biglycan (Westergren-Thorsson et al., 2002), and the large hyaluronan-binding versican (Iozzo, 1998). The core proteins of these PGs are substituted with galactosaminoglycans, which are heteropolysaccharides composed of alternating N-acetyl-D-galactosamine (GalNAc) and D-glucuronic acid (GlcA) residues, sulfated at various positions. In these polysaccharides, some of the GlcA residues are epimerized into L-iduronic acids (IdoA) on polymer level by chondroitin C-5 epimerase (Malmstrom, 1981). This occurs in blocks that vary in length from a few disaccharides to almost a whole polysaccharide. A galactosaminoglycan chain containing IdoA residues is referred to as dermatan sulfate (DS), whereas the related non-IdoA containing counterpart is called chondroitin sulfate (CS). In DS chains, the ratio of IdoA to GlcA varies substantially between different tissues and core proteins (Trowbridge and Gallo, 2002). The function of a PG can reside in its core protein, in the glycosaminoglycan (GAG) side chains or both. GAG chains exert their functions mostly through interactions with proteins. These interactions depend largely on the fine structure of the GAG. However, most knowledge about the structure-function relationship of GAGs comes from research on heparin/heparan sulfate (HS) chains, whereas galactosaminoglycans have been studied much less. During the last few years some roles of galactosaminoglycan chains have been resolved, but the ‘‘biologically active motifs’’ have, in most cases, not been determined yet. Moreover, motifs in DS chains containing alternating GlcA- and IdoA-disaccharides generate self-association that is of importance in collagen fibril organization (Fransson et al.,

© The Author 2005. Published by Oxford University Press. All rights reserved. The online version of this article has been published under an open access model. Users are entitled to use, reproduce, disseminate, or display the open access version of this article for non-commercial purposes provided that: the original authorship is properly and fully attributed; the Journal and Oxford University Press are attributed as the original place of publication with the correct citation details given; if an article is subsequently reproduced or disseminated not in its entirety but only in part or as a derivative work this must be clearly indicated. For commercial re-use, please contact [email protected]

1277

Downloaded from http://glycob.oxfordjournals.org/ at Lund University Libraries, Head Office on November 6, 2012

Kerstin Tiedemann1,3,5, Benny Olander1,5, Erik Eklund1,4,5, Lizbet Todorova5, Martin Bengtsson6, Marco Maccarana5, Gunilla Westergren-Thorsson5, and Anders Malmström2,5

K. Tiedemann et al.

1278

was seen after TGF-β1 treatment (Westergren-Thorsson et al., 1992). We hypothesized that the changes in CS/DS co-polymeric structure upon cytokine treatment were due to altered expression of the GAG-modifying enzymes, and therefore performed measurements of activity and mRNA expression on epimerase and O-sulfotransferases, correlating their level and activity with the detailed structures of DS recovered from purified PGs. Results Effects of cytokines on PG production in HFL-1 cells Untreated human fetal lung fibroblasts (HFL-1) cells secrete the CS/DS-carrying PGs versican, biglycan, and decorin, which can be individually separated by a combination of size filtration and hydrophobic chromatography, as described in Materials and methods. Versican represents approximately 35%, biglycan 30%, and decorin 5% of the total 35S-labeled PGs, and the remaining portion represents HSPG and freeGAG-chains (data not shown). Versican, biglycan, and decorin production was increased by TGF-β1 4-, 6.5-, and 1.9-fold, respectively (data not shown). Cytokine combination treatment further increased the production of each of the three PGs 1.5–1.8 fold (data not shown). Content of these PGs before and after cytokine stimulation agree with previous finding reported by our group (Tiedemann et al., 1997). Characterization of galactosaminoglycan side chains Using size-exclusion chromatography, no differences in the length of the DS chains released from decorin and biglycan obtained from untreated control cells, and cells treated with the different cytokines could be observed (Figure 1A and B). As described earlier, the uronic acid composition of the DS chains was determined after digestion with chondroitin AC-1 lyase (cleavage only at glucuronic sites) (Malmstrom et al., 1975). DS chains from decorin and biglycan had a similar composition throughout the experiments (Figure 2). DS chains from control cells had around 70% of the uronic acid residues in the IdoA configuration. These IdoA residues were situated in blocks, as shown by the finding that most of the enzymatically cleaved material was longer than 12 monosaccharides (Figure 2A and B). The content of IdoA in decorin and biglycan DS decreased to 53% upon TGF-β1 treatment, and to 36% upon combination treatment (Figure 2C–F). The increasing amounts of GlcA residues after treatment also appeared in blocks, because most of them were recovered in the disaccharide pool after chondroitin AC-1 lyase digestion (Figure 2C–F). The IdoA content of the CS/DS-chains from versican was 25% and remained unaffected by cytokine treatment (data not shown). The sulfation pattern of the polysaccharide chains in [3H]-glucosamine and [35S]-labeled versican and biglycan/ decorin was studied by separation of the disaccharides obtained after chondroitin ABC lyase digestion. Disaccharides from versican obtained from control fibroblasts were 51% 4-O-sulfated, 28% 6-O-sulfated, and the remaining portion nonsulfated (Figure 3 and Table I). Compared to control, treatment with TGF-β1 resulted in an increased proportion (70%) of 4-O-sulfated disaccharides and in a


1982; Hedbom and Heinegård, 1993). DS also regulates coagulation and affects wound healing/inflammation (Trowbridge and Gallo, 2002). During coagulation, DS, to which the IdoA residues endow conformational flexibility, controls the activity of both thrombin and protein C (Fernandez et al., 1999). This process is achieved via activation of the serpin heparin cofactor-II and requires the presence of at least a hexasaccharide motif with the specific structure (IdoA-2S-GalNAc-4S)3 (Maimone and Tollefsen, 1990). In wound healing/inflammation, DS strongly promotes fibroblast growth factor-2 activity and fibroblast proliferation (Penc et al., 1998). Over-sulfated DS has been shown to promote neurite outgrowth (Hikino et al., 2003), and other motifs with dense sulfation inhibit fibroblast proliferation (Westergren-Thorsson et al., 1991, 1993a). In addition, DS also regulates endothelial ICAM-1 expression (Penc et al., 1999). Biglycan and decorin interact with several cytokines such as transforming growth factor-β (TGF-β) and tumor necrosis factor-α both via the protein cores and the DS side chains (Hildebrand et al., 1994; Tufvesson and Westergren-Thorsson, 2002). Versican, which belongs to the lectican family, binds with its core protein to hyaluronan and microfibrils in the extracellular matrix (Iozzo, 1998; Isogai et al., 2002). CS chains released from versican bind to selectins and platelet factor 4, thereby influencing neutrophil rolling and activity (Petersen et al., 1999; Kawashima et al., 2000, 2002). A complex biosynthetic machinery is required to create the variable heteropolysaccharide pattern of CS/DS (Silbert and Sugumaran, 1995). At least seven enzymes are required for formation and sorting of the tetrasaccharide link region that bridges the polysaccharide to the protein core. Five enzymes all with different properties and organization generate the subsequent polymerization of the polysaccharide backbone (Sato et al., 2003; Yada et al., 2003a,b). The final process of polymer modification, leading to the final highly complex structure, requires at least nine enzymes. The enzyme responsible for the formation of IdoA residues, specific for DS, is the chondroitin C-5 epimerase, which is yet to be cloned. This epimerase requires a non-sulfated chondroitin as a substrate (Malmstrom and Aberg, 1982; Malmstrom, 1984; Hannesson et al., 1996). We have shown that in tissues containing no DS, no epimerase activity is detectable (Tiedemann et al., 2001). To generate a DS chain containing more than 15% IdoA, an efficient subsequent 4-O-sulfation is required (Malmstrom and Aberg, 1982; Eklund et al., 2000; Tiedemann et al., 2001). The main enzymes involved in this process are dermatan 4-O-sulfotransferase1 (D4ST-1) (Kang et al., 2001) and chondroitin 4-O-sulfotransferase-2 (C4ST-2) (Mikami et al., 2003), which both have a preference for IdoA-containing structures. As outlined above, the functions of PGs not only depend on their amount in tissues but also on the structure of the GAG side chains. It has previously been shown that cells treated with TGF-β secrete an increased amount of structurally different PGs. Several reports demonstrate an increase in chain length with respect to the chondroitin sulfate chains of these PGs (Bassols and Massague, 1988; Rapraeger, 1989; Little et al., 2002). However, in a study with human embryonic skin fibroblasts, no change in chain size was noted, whereas a decrease in the amount of L-IdoA

Biosynthesis of dermatan/chondroitin sulfate

3000

Decorin

A

3000

2500

A

Biglycan 3000

Control

Oligo

2000

B

Oligo

Control

12 10 8

6

2000 12 10 8

6

4

2

1000

4

2

55

65

75

65

75

1000

2000 0 15 2000

25

35

C

45

55

65

75

15 2000

TGF-β1

25

35

45

D

TGF-β1

35

1500

0

S (dpm)

Control TGF TGF, EGF, PDGF

1000

1000

Vt

500

1000

0

0 15

0

1200

0

20

40

60

25

35

E

45

55

65

75

TGF-β1 EGF PDGF-BB

800

35

Control TGF TGF, EGF, PDGF

1500 1000

55

TGF-β1 EGF PDGF-BB

0 25

35

45

Fraction

2000

45

400

15

2500

35

800

0

B

25

F

1200

400

3000

15

55

65

75

15

25

35

45

55

65

75

Fraction

Fig. 2. Chondroitin AC-I lyase cleavage pattern of galactosaminoglycans released from decorin and biglycan. Decorin and biglycan GAG chains, obtained as described in the Materials and methods from control cells, TGF-β1-treated cells and TGF-β1/EGF/PDGF-BB-treated cells were digested with chondroitin AC-I lyase and the resulting oligosaccharides separated on a Biogel P6 column. Standard elution positions of di-, tetra-, hexa-, octa-, deca- and dodecasaccharides are indicated with their respective sugar numbers. “Oligo” represents oligosaccharides larger than dodecasaccharides.

500 0 0

20

40

60

Fraction Fig. 1. Length of galactosaminoglycan chains from decorin and biglycan. The small DSPGs decorin and biglycan were obtained as described in the Materials and methods. The core proteins were removed using alkali elimination and the side chains of decorin (A) and biglycan (B) from control cells, TGF-β1-treated cells and TGF-β1/EGF/PDGF-BB-treated cells were separated using size filtration on a Superose-6 column.

decreased proportion of 6-O- and nonsulfated disaccharides (19% and 11%, respectively; Figure 3 and Table I). The combination of cytokines enhanced these changes further. On the other hand, almost all disaccharides from decorin/ biglycan were 4-O-sulfated (85%), and this proportion did not significantly change upon treatment (Figure 3 and Table I). Analysis of the disaccharides from versican as well as from decorin/biglycan separated on an econosphere column, showed that no mono 2-O-sulfated disaccharides were present (data not shown). Disulfated disaccharides were not analyzed in our study, representing a minor proportion (around 2%) of the PGs produced by cultured human fibroblasts (Coster et al., 1991). Enzyme activities The activity of chondroitin-glucuronate C-5 epimerase, the enzyme that catalyzes the epimerization of GlcA to IdoA, was determined using [5–3H] GlcA-labeled chondroitin as

substrate. Epimerase from control cells catalyzed the release of 750 dpm/h/mg protein (Figure 4). Cells treated with TGF-β1 had a 25% reduction in epimerase activity, and cells treated with the cytokine combination had a 57% reduction. epidermal growth factor (EGF) and plateletderived growth factor-BB (PDGF-BB) alone did not influence epimerase activity (data not shown). Chemically desulfated DS and CS were used as substrates to study the galactosaminoglycan O-sulfotransferase capacity of microsomes. No differences in the sulfotransferase activity toward dermatan in microsomes from untreated, TGF-β1 and TGF-β1/EGF/PDGF-BB treated cells were seen. Nor did treatment with EGF and PDGFBB alone affect this activity (data not shown). However, using chondroitin as acceptor, a 2-fold increase in sulfotransferase activity in TGF-β1 treated cells and a 2.2-fold increase in TGF-β1/EGF/PDGF-BB treated cells, compared to controls, was seen (Figure 5). EGF or PDGF-BB alone did not affect this activity significantly (data not shown). To characterize what type of sulfation occurred, the substrate chondroitin, after incubation, was subjected to chondroitin AC-I/ABC lyase digestions followed by disaccharide separation on a Lichrosorb-NH2 column. Approximately 80% of the 35S-labeled disaccharides resulting from incubation with control cell microsomes were 6– O-sulfated. Chondroitin 6-O-sulfotransferase activity increased by 60% upon TGF-β1 treatment but did not increase further with the cytokine combination (Figure 6). Chondroitin-4-O-sulfotransferase activity increased 2-fold with TGF-β1 and 3.5-fold with the growth factor combination (Figure 6). 1279


S (dpm)

V0

K. Tiedemann et al.

Versican

Biglycan/Decorin

2000

2000

Control

1500

0-S

1000

1000

500

500

0

6-S

10

20

30

40

50

60

2000

10

20

30

40

2000

TGF-β1

1500

0

70

50

60

70

TGF-β1

1500

1000

1000

500

500

35

10

20

30

40

1500

50

60

70

0

10

20

30

40

400

200

50

60

Control

70

TGF- β1

1500

TGF-β1 EGF 1000 PDGF-BB

1000

TGF-β1 EGF PDGF-BB

500

500

0

0 0

10

20

30

40

50

60

0

70

10

20

Fraction

30

40

50

60

70

Fraction

Fig. 3. Analysis of non- and mono-sulfated disaccharides from chondroitin ABC lyase digested galactosaminoglycans from versican and biglycan/decorin. 3H-glucosamine and 35S-labeled versican and biglycan/ decorin side chains from control cells, TGF-β1-treated cells and TGF-β1/ EGF/PDGF-BB-treated cells were digested overnight with chondroitin ABC lyase and the resulting disaccharides were separated on a LichrosorbNH2 column. A standard solution containing the unsaturated disaccharides 2-acetamido-2-deoxy-3-O-(β-D-gluco-4-enepyranosylhexuronic acid)-Dgalactose (ΔDi-0S), 2-acetamido-2-deoxy-3-O-(β-D-gluco-4-enepyranosylhexuronic acid)-4-O-sulfo-D-galactose (ΔDi-4S) and 2-acetamido-2deoxy-3-O-(β-D-gluco-4-enepyranosylhexuronic acid)-6-O-sulfo-D-galactose (ΔDi-6S) were used to calibrate the column (indicated with arrows). The filled circles represent 3H-glucosamine label and the open squares represent 35S- label.

Table I. Sulfation of CS/DS chains from versican and biglycan/decorin ΔDi-0S (%) Versican

3H

ΔDi-4S (%) 3H

ΔDi-6S (%) 35S

3H

35S

Control

21

51

66

28

34

TGF-β1

11

70

78

19

22

TGF-β1/EGF/PDGF-BB

10

77

86

13

14

Biglycan/decorin Control

14

85

99

1

1

TGF-β1

14

82

94

4

6

4

95

99

1

1

TGF-β1/EGF/PDGF-BB

Data were calculated from Figure 3 and presented as percent of total disaccharides.

mRNA expression To determine which of the 4-O-sulfotransferases were affected by the cytokine treatment, the mRNA for 4-O-sulfotransferases were estimated using real time polymerase 1280

600

0

0 0

800

TGF- β1 EGF PDGF-BB

Fig. 4. The effect of cytokines on chondroitin C-5 epimerase activity. Particulate cellular fractions sedimenting between 10,000 and 105,000 × g from untreated, TGF-β1-treated and TGF-β1/EGF/PDGF-BB treated cells were incubated with radiolabeled, defructosylated K4 polysaccharide for 24 h at 37°C. The reaction mixtures were distilled and the radioactivity in the resulting distillates was determined. The values represent the mean of three separate experiments ± SE. Student’s t-test gave a value for p = 0.08 for TGF-β-treated cells compared to control cells and p = 0.01 for growth factor combination compared to control cells.

300% 250% 200% 150% 100% 50% 0% Control

TGF-β1

TGF-β1, EGF, PDGF-BB

Fig. 5. Sulfotransferase activity in microsomes incubated with chondroitin as substrate. Particulate cellular fractions from untreated cells, TGF-β1-treated cells, and cells treated with TGF-β1, EGF and PDGF-BB was incubated with [35S]-PAPS and chondroitin as acceptor substrate. The reaction products were quantified as described in Methods. The values represent the means of three separate experiments ± S.E. Student’s t-test gave a value for p = 0.01 for TGF-β-treated cells compared to control cells and p = 0.05 for growth factor combination compared to control cells.

chain reaction (RT–PCR). TGF-β1 increased the mRNA of chondroitin-4-O-sulfotransferase 1 (C4ST1) 2-fold and the cytokine combination by 3.3-fold (Figure 7). The mRNA of C4ST2 did not significantly change, whereas mRNA for dermatan-4-O-sulfotransferase (D4ST1) decreased 60% after TGF-β1 stimulation and


0

Change in specific activity

3

0-S

0

0

S, H (dpm)

Control

1500

6-S

1000

4-S

Released 3H (dpm/h/mg protein)

4-S


Discussion

4

Di-6S

Relative change in sulfation

3.5

Di-4S

3 2.5 2 1.5 1

0

Control

TGF- β1

TGF-β1/EGF/PDGF-BB

Fig. 6. Characterization and quantification of sulfation of chondroitin incubated as a substrate with microsomes. Chondroitin was incubated with [35S]-PAPS and microsomal proteins, as in Figure 5. After incubation the reaction mixtures were spotted on Whatman paper and washed as described in the Materials and methods. Subsequently the reaction products were eluted with water, desalted, digested with chondroitin AC-I/ABC lyase, and the resultant unsaturated disaccharides separated on a Lichrosorb-NH2 column. The amount of 6-O-sulfated (grey bars) and 4-O-disaccharides (white bars) were quantified and expressed as percent of the amount of 6-O-sulfated and 4-O-disaccharides obtained from control cells. The values represent the mean of three separate experiments ± S.E. Student’s t-test gave a value for p = 0.003 for the increase in 4–0-sulfation and 6–0-sulfation after treatment with TGF-β, EGF, and PDGF-BB compared to the control cells.

4

Control TGF-β1 TGF-β1, EGF, PDGF-BB

Relative expression

3.5 3 2.5 2 1.5 1 0.5 0 C4ST-1

C4ST-2

D4ST-1

Fig. 7. mRNA expression of 4-O-sulfotransferases. Fibroblasts were treated with TGF-β1 or the combination of TGF-β1, EGF, and PDGF-BB. Total RNA was prepared and reversely transcribed as outlined in the Materials and methods. The cDNA formed was used for RT–PCR. C4ST1, C4ST2, and D4ST1 were quantified and normalized against 18s RNA. White bars represent control cells, grey bars TGF-β1-treated cells and black bars the combination-treated cells. The results are expressed as relative increase compared to control ± S.E. Student’s t-test gave a value for p = 0.03 for the increase in C4ST-1 mRNA after TGF-β1 treatment and p = 0.01 after combination treatment with TGF-β1, EGF, and PDGF-BB. The decrease of D4ST-1 mRNA gave a p = 0.001 after TGF-β1 treatment.

40% after combination treatment (Figure 7). No transcript for C4ST3 was detected either in the control or after treatment.

1281


0.5

Individual growth factors and their combinations not only change the expression of PGs produced and secreted by fibroblasts, but they also modify the detailed structure of the polysaccharide chain, which invariably will affect biological properties. The mechanisms for these conspicuous changes are not yet fully understood. Here, we present that the activity and the mRNA of three of the polymer-modifying enzymes, chondroitin C-5 epimerase, D4ST1, and C4ST1 are regulated by TGF-β1 alone or in combination with EGF plus PDGF-BB. These variations are correlated with the changes of the CS/DS chain structures carried by three isolated PGs, that is, versican, biglycan, and decorin. The multifunctional growth factor TGF-β1 is a potent enhancer of PG synthesis in fibroblasts (Kähäri et al., 1991; Tiedemann et al., 1997). Fetal primary lung fibroblasts, stimulated by TGF-β1 produced approximately 7, 4, and 2-fold more biglycan, versican, and decorin, respectively, measured as increased 35S incorporation. This increase in PG production is not due to an increase in chain length, a result which differs from other reports (Bassols and Massague, 1988; Little et al., 2002). Following chain extension, epimerization of GlcA to IdoA can occur to an extremely variable extent, ranging from few percent of IdoA to almost 100% of epimerized residues (Tiedemann et al., 1997; Bao et al., 2004). The proportion of the two epimers depends on the protein core, the cell type, and tissue localization (Tiedemann et al., 2001). In our analysis, TGF-β treatment reduced the IdoA content in biglycan and decorin DS chains by 50% (from 70 to 73% to 36–35%), whereas it did not have any effect on versican (25%). These data would indicate that the protein cores differentially influence the extent of epimerization, as shown previously (Seidler et al., 2002). These structural changes correlated well with a marked decrease in microsomal epimerase activity, which fell 25% after TGF-β1 alone and 57% with the combination treatment. To address the mechanism behind these changes, cloning of the enzyme is a prerequisite. During CS/DS biosynthesis, sulfation of GalNAc in 4and 6-position is carried out, as well as in 2-position of iduronic acid. The 2-O-sulfation has not been detected in any of the mono-sulfated disaccharides obtained from labeled PGs from fibroblasts. 4-O-Sulfation on GalNAc residues adjacent to IdoA residues is thought to be carried out mainly by D4ST1 and C4ST2 (Evers et al., 2001; Mikami et al., 2003) and is shown to be a process linked in the biosynthesis with the epimerization (Malmstrom and Fransson, 1975). As biglycan and decorin are extensively 4-Osulfated already in the control cell, a 50% reduction in IdoA content must be accompanied at least by a similar reduction in −IdoA-GalNAc-4S-IdoA– structures. D4ST1 mRNA was reduced 50% after cytokine treatment, but no changes in enzyme activity, using desulfated DS as acceptor was observed. On the other hand, biglycan and decorin remained extensively 4-O-sulfated after cytokine treatment, which results in an increased proportion of −GlcA-GalNAc-4S-GlcA– structures. Also, the 4-O-sulfation of versican increases upon treatment (from 51 to 77%). C4ST1, C4ST2, and C4ST3 are the enzymes responsible for sulfation in GlcA-rich structures (Mikami et al., 2003). C4ST1

K. Tiedemann et al.

1282

related to inflammation. The expression, structure and possibly also their function have been shown to change in atherosclerotic plaques. The amount of epimerization has also been implicated in cell proliferation, albeit with some contradictory results. We have previously shown that a high degree of epimerization and sulfation negatively affects fibroblast proliferation. A decrease in the ratio of IdoUA/ GlcUA mediated by TGF-β could therefore create an environment supporting fibroblast growth, a mechanism important in fibrosis for example. This is however in contradiction to other reports regarding Fibroblast Growth Factor (FGF) signaling showing that epimerization is required. However, in several reports binding of FGF-2 to iduronosyl-2S areas are shown, but no resulting stimulation of activity could be demonstrated (Bao et al., 2004; Taylor et al., 2005). In conclusion, we suggest that cytokines are involved not only in the up-regulation of the expression of PGs, but also in the regulation of the structure of the GAG side chains. This is achieved by controlling relevant enzymes in the biosynthetic pathway of the GAGs. Further studies are required to precisely define the molecular mechanisms behind these effects and to outline the biological consequences

Materials and methods Materials HFL-1 were obtained from ATCC, Rockville, MD. Cell culture medium was from Gibco, Paisley, UK. Na235SO4 (1310 Ci/mmol) was purchased from ICN (Asse-Relegem, Belgium) and 3H-glucosamine from Amersham Radiochemical Center (Amersham, UK). The ion-exchange resin DEAE-52 was purchased from Whatman, Maidstone, UK, Sephacryl-500 and octyl-Sepharose CL-4B resins were from Amersham Biosciences, Uppsala, Sweden, Biogel P6 was from BioRad (Sundbyberg, Sweden). Econosphere-NH2 and Lichrosorb-NH2 columns were from Alltec and Merck, Darmstadt, Germany, respectively. Chondroitin ABC (EC 4.2.2.4) and AC-I lyases (EC 4.2.2.5) and unsaturated disaccharides from chondroitin ABC lyase-digested CS and DS were products of Seikagaku, Tokyo, Japan. TGF-β1 was purchased from British Biotechnology (Abingdon, UK), and recombinant EGF and PDGF-BB were from Novakemi AB, Enskede, Sweden. Cell culture conditions, cytokine treatment, and metabolic labeling of PGs HFL-1 fibroblasts, passages 15–25, were grown in 75 cm2 flasks in Eagle’s minimal essential medium (MEM) supplemented with 10% donor calf serum. At near confluence, the serum concentration was lowered to 1% and the cells were allowed to adapt for 2 h. The cytokines TGF-β1 (10 ng/mL), EGF (50 ng/mL) and PDGF-BB (10 ng/mL) (WestergrenThorsson et al., 1992; Tiedemann et al., 1997), diluted in sulfate-deprived MEM with 0.1% serum, were then added in various combinations. After another 2 h, 35SO4 was added to a final concentration of 50 μCi/mL. In some experiments, 25 μCi/mL of 3H-glucosamine was also added. The radiolabeling was performed for 24 h, and the medium was thereafter


mRNA increased 2–3 fold upon treatment, whereas C4ST2 remained unchanged, indicating that the former has a predominant role in CS biosynthesis in fibroblasts. No transcript for C4ST3 was detected. This is not surprising, because it has been shown that C4ST3 is expressed mainly in the liver (Kang et al., 2002). In summary, C4ST1 transcript data and 4-O-sulfation enzyme activity show a 2–3fold increase in the sulfation of GalNAc flanked by GlcA. This is in conclusion with a high-throughput induction gene trap study that indicated C4ST as a target for the TGF-β superfamily proteins during embryogenesis (Kluppel et al., 2002). When cellular extracts were incubated with chondroitin as substrate, 6-O-sulfation activity increased 60% upon treatment. This increase was not reflected in the in vivo structures: 6-O-sulfates remained a minor component on versican and even decreased on biglycan/decorin (from 28 to 19/13%), which instead becomes more 4-O-sulfated. A general finding of our results is that the biosynthetic enzymes modifying CS/DS can cope with increased versican, biglycan, and decorin core protein production. In other terms, if we integrate the increased PG production with the measurements of the single modifications, we discover that the total absolute amount of IdoA, 4-O and 6-Osulfates synthesized by the cell increased approximately 2-, 6-, and 3-fold, respectively. It is therefore apparent that the studied growth factors modify the final CS/DS structures by regulating a biosynthetic machinery which is not working at its highest possible speed in any of its members. One could imagine that different biosynthetic complexes are competing for a limited number of growing chains, as it has been hypothesized for HS. The regulation of growth factors could affect the composition of these biosynthetic complexes, their affinity for the substrate, and their localization within the Golgi. TGF-β has also been shown to modify other glycan structures such as decreasing the mRNA of hyaluronan synthase 2 and 3 in keratinocytes (Pasonen-Seppanen et al., 2003) and up-regulating β-1,6-Nglucosaminyltransferase V involved in the modification of N-linked oligosaccharides (Miyoshi et al., 1995). Thus, in this report, we extend the action of TGF-β1 to the CS/DSmodifying epimerase and C4ST1 in lung fibroblasts. The effect of these structural changes is difficult to address because not much is known about the biological function of specific epitopes in CS/DS. It could be hypothesized, though that these changes are important in processes such as inflammation. GAG chains on versican have for example been implicated in the binding of L- and P-selectin thereby influencing the recruitment of neutrophils. In addition, GAG chains can also bind chemokines, and a changed copolymeric structure could negatively or positively affect the availability of these mediators for the target cells. Evidence of this regards the chemokine CCL 11. DS but not chondroitin was able to bind and inhibit its action on eosinophiles (Culley et al., 2003). TGF-β produced by macrophages during an inflammatory process could therefore by reducing the amount of epimerization, as shown in this study, indirectly affect the recruitment and activation of inflammatory cells. This could be important for examples in allergies and asthma. Evidence has also emerged of the role of CS/DS-PGs in atherosclerosis, a process closely


removed, 25 μg CS-6 as carrier was added, and the sample frozen until further analysis. When microsomes or RNA were prepared from cells, 35SO4 addition was omitted and total 26 h incubation with cytokines was performed, as above.

sucrose, pH 6.5. The cells were then homogenized by Potter and microsomes sedimenting between 10,000 × g and 105,000 × g were used as the enzyme source (Malmstrom et al., 1982). Preparation of O-sulfotransferase substrates

Purification of PGs

Determination of the chain length Purified decorin and biglycan from the octyl-Sepharose column were propanol precipitated, and the GAG-chains were released from the protein cores by β-elimination in 0.5 M NaOH/0.1 M NaBH4 at room temperature for 24 h. The samples were neutralized with acetic acid, and the polysaccharides recovered by anionic exchange chromatography on DEAE-cellulose columns as described above. Samples were subsequently run on a Superose-6 column eluted in 4 M guanidine hydrochloride/50 mM acetate, pH 5.8.

DS and CS-6 was prepared as described earlier (Malmstrom, 1984). DS was treated with chondroitin AC-I lyase to remove the GlcA residues. CS-6 and chondroitin AC-I lyase-treated DS were de-sulfated and fractionated on Sephadex G-100. Fractions with Kav values between 0.39 and 0.66, consisting of oligosaccharides with five to eight disaccharides, were selected as substrates. The uronic acid content of the different fractions was determined using the carbazole assay (Bitter and Muir, 1962). Chondroitin-glucuronate 5-epimerase assay Epimerase (EC 5.1.3.19) activity in 10–25 μg of microsomal protein was assayed in a buffer containing 25 mM Hepes, pH 6.5, 10 mM MnCl2, 0.25 % Nonidet NP-40 in a final volume of 100 μL. Radiolabeled, defructosylated K4 polysaccharide (1.4 × 106 dpm/μmol of hexuronic acid, 30,000 dpm/incubation) (Hannesson et al., 1996) was used as substrate. The samples were incubated for 24 h at 37°C and then boiled for 3 min. They were thereafter distilled, and the radioactivity in the distillate was measured. Protein concentrations were estimated according to the method of Lowry (Lowry et al., 1951). Sulfotransferase assay

Purified decorin and biglycan side chains were digested with 10mU/mL chondroitin AC-I lyase in 0.1 M-Tris/acetate buffer, pH 7.3 at room temperature for 4 h. The split products were separated on a column (1 × 100 cm) of Bio-Gel P6 eluted in 0.5 M NH4HCO3. The amount of IdoA was determined as described previously (Malmstrom et al., 1975).

Sulfotransferase activity in 20–25 μg of microsomal enzyme was assayed for 10 min in a buffer containing 0.2 M MES, 10 mM sodium fluoride, 10 mM MnCl2, 1% Triton X-100, and 0.2 mM PAPS/ 35S-PAPS (2.5 μCi), pH 6.5, in a total volume of 50 μl (Eklund et al., 2000). Dermatan or chondroitin oligosaccharides were used as sulfate acceptors at a final concentration of 0.5 mM uronic acid. After completion of the incubation, sodium sulfate was added to a final concentration of 0.15 M, the samples were boiled for 2 min, and carrier DS (100 μg) was added. The entire reaction mixtures were applied to disks (2x2 cm) of Whatman 3MM papers, which were washed in 6.7 M isobutyric acid/0.19 M ammonia (Sugahara et al., 1985). The disks were then dried and subjected to scintillation counting in 10 mL of omnifluor/toluene mixture.

Separation of 2-, 4-, and 6-O-mono-sulfated disaccharides

RNA purification

Decorin, biglycan, and versican side chains were digested with 10mU/mL chondroitin AC-I/ABC lyases (EC 4.2.2.6 and 4) overnight and chromatographed on a column (0.46 x 15 cm) of Lichrosorb-NH2 eluted in 0.1 M acetate, pH 5.0 (Hjerpe et al., 1979). Separation of 2-O-sulfated disaccharides was performed on a Econospere-NH2 column eluted in 50 mM NaH2PO4, pH 2.7 (modified version of method described in [Karamanos et al., 1994]). Elution positions were determined using unsaturated disaccharide standards detected with UV-light at 232 nm.

Total RNA was isolated from cells cultivated for 26 h with cytokines . A spin protocol including a deoxyribonuclease treatment according to the manufacturer’s instructions (Qiagen, Hilden, Germany) was used. Quantification and purity was measured spectrophotometrically with a NanoDrop ND-1000 (NanoDrop Technologies, Wilmington, DE) and the quality of the RNA was determined on a 1.5% TAE-agarose gel.

Determination of the relative proportion of IdoA residues in DS chains

Enzyme preparations HFL-1 cells, treated with cytokines for 26 h, were detached by scraping with a rubber policeman in 50 mM Hepes/0.25 M

cDNA synthesis A first strand cDNA synthesis kit for RT–PCR from Roche was used to transcribe comparative amounts of purified total RNA. The reaction was performed in a total volume of 20 μL containing 1 × reaction buffer, 5 mM MgCl2, 1 1283


The PGs in the cell medium were purified as described previously (Tiedemann et al., 1997). In brief, the medium samples were applied to DEAE-cellulose columns pre-equilibrated in a buffer containing 6 M urea/50 mM acetate buffer, pH 5.8, supplemented with various enzyme inhibitors (10 mM EDTA, 10 mM EACA, 5 mM benzamidine and 5 mM NEM). The columns were subsequently washed with 60 volumes of the same buffer and with 6 volumes of 6 M urea buffer with 0.5 M acetate, pH 5.8. The PGs were then eluted with 4 M guanidine hydrochloride/50 mM acetate buffer, pH 5.8, followed by separation into large (versican) and small (biglycan and decorin) components by gel filtration using a Sephacryl-500 HR column eluted in 4 M guanidine hydrochloride/50 mM acetate, pH 5.8. The small PG pools were further separated into decorin and biglycan using an octyl-Sepharose CL-4B. The radioactivity of the fractions was determined by liquid scintillation counting.

K. Tiedemann et al.

mM deoxynucleotide mix, 0.08 units of random primers, 50 units of RNAse inhibitors, 20 units of AMV reverse transcriptase, and sterile water. Reaction was performed according to manufacturer’s protocol. RT–PCR

Primers Primers were selected with a web-based software called Primer3 and ordered from A/S DNA Technology, Denmark. C4ST-1: forward 5′-aaccaccgcttgaaaagcta-3′and reverse 5′ttgatgatcttggtgccgta-3′. C4ST-2: forward 5′-tcatcgtgtactgggacagc-3′and reverse 5′-cggcttagagaaggacgtgt-3′. C4ST-3: forward 5′-ccggcatttggaaacagag-3′ and reverse 5′-gggtcctgatccaggtcata-3′, D4ST-1: forward 5′-gatgtcacattccccgagtt3′ and reverse 5′-catccaatgctcattcatgc-3′. 18S-RNA: forward 5′-cgaacgtctgccctatcaac-3′ and reverse 5′-tgccttccttggatgtggta-3′. Acknowledgments We acknowledge the technical assistance of Lena Åberg and Camilla Dahlqvist. This work was supported by grants from the Swedish Medical Research Council (7479 and 11550), Greta and John Kock, A. Österlund, Anna-Greta Craaford Foundations, the Swedish Rheumatism Association, Gustaf V’s 80year Fund, the Heart-Lung Foundation, Bergvall foundation, Djurskyddsmyndigheten, AB Polysaccharide Research, Mitzutani Foundation, and the Medical Faculty of Lund University. Abbreviations C4ST1 and 2, chondroitin-4-O-sulfotransferase 1 and 2; C6ST1, chondroitin-6-O-sulfotransferase 1; CS, chondroitin sulfate; D4ST1, dermatan-4-O-sulfotransferase 1; DS, dermatan sulfate; EGF, epidermal growth factor; GAGs, glycosaminoglycans; GalNAc, N-acetyl-D-galactosamine, GlcA D-glucuronic acid; HFL-1, human fetal lung fibroblasts; HS, heparan sulfate; IdoA, L-iduronic acid; MEM, Eagle’s minimal essential medium; PAPS, 3′phosphoadenosine-5′-phosphosulfate; PDGF-BB, plateletderived growth factor-BB; PGs, proteoglycans; TGF-β1, transforming growth factor-β1. 1284

Bao, X., Nishimura, S., Mikami, T., Yamada, S., Itoh, N., and Sugahara, K. (2004) Chondroitin sulfate/dermatan sulfate hybrid chains from embryonic pig brain, which contain a higher proportion of L-iduronic acid than those from adult pig brain, exhibit neuritogenic and growth factor binding activities. J. Biol. Chem., 279, 9765–9776. Bassols, A. and Massague, J. (1988) Transforming growth factor beta regulates the expression and structure of extracellular matrix chondroitin/ dermatan proteoglycans. J. Biol. Chem., 263, 3039–3045. Bernfield, M., Gotte, M., Park, P.W., Reizes, O., Fitzgerald, M.L., Lincecum, J., and Zako, M. (1999) Functions of cell surface heparan sulfate proteoglycans. Annu. Rev. Biochem., 68, 729–777. Bitter, T., and Muir, H.M. (1962) A modified uronic acid carbazole reaction. Anal. Biochem., 4, 330–334. Bustin, S.A. (2000) Absolute quantification of mRNA using real-time reverse transcription polymerase chain reaction assays. J. Mol Endocrinol., 25, 169–193. Bustin, S.A. (2002) Quantification of mRNA using real-time reverse transcription PCR (RT-PCR): trends and problems. J. Mol Endocrinol., 29, 23–39. Coster, L., Hernnas, J., and Malmstrom, A. (1991) Biosynthesis of dermatan sulphate proteoglycans. The effect of beta-D-xyloside addition on the polymer-modification process in fibroblast cultures. Biochem. J., 276, 533–539. Culley, F.J., Fadlon, E.J., Kirchem., A., Williams, T.J., Jose, P.J., and Pease, J.E. (2003) Proteoglycans are potent modulators of the biological responses of eosinophils to chemokines. Eur. J. Immunol., 33, 1302– 1310. Eklund, E., Roden, L., Malmstrom, M., and Malmstrom, A. (2000) Dermatan is a better substrate for 4-O-sulfation than chondroitin: implications in the generation of 4-O-sulfated, L-iduronate-rich galactosaminoglycans. Arch. Biochem. Biophys., 383, 171–177. Evers, M.R., Xia, G., Kang, H.G., Schachner, M., and Baenziger, J.U. (2001) Molecular cloning and characterization of a dermatan-specific N-acetylgalactosamine, 4-O-sulfotransferase. J. Biol. Chem., 276, 36344–36353. Fernandez, J.A., Petaja, J., and Griffin, J.H. (1999) Dermatan sulfate and LMW heparin enhance the anticoagulant action of activated protein C. Thromb. Haemost., 82, 1462–1468. Fransson, L.A., Coster, L., Malmstrom, A., and Sheehan, J.K. (1982) Self-association of scleral proteodermatan sulfate. Evidence for interaction via the dermatan sulfate side chains. J. Biol. Chem., 257, 6333– 6338. Hannesson, H.H., Hagner-McWhirter, A., Tiedemann, K., Lindahl, U., and Malmstrom, A. (1996) Biosynthesis of dermatan sulphate. Defructosylated Escherichia coli K4 capsular polysaccharide as a substrate for the D-glucuronyl C-5 epimerase, and an indication of a two-base reaction mechanism. Biochem. J., 313, 589–596. Hedbom, E. and Heinegård, D. (1993) Binding of fibromodulin and decorin to separate sites on fibrillar collagens. J. Biol. Chem., 268, 27307– 27312. Hikino, M., Mikami, T., Faissner, A., Vilela-Silva, A.C., Pavao, M.S., and Sugahara, K. (2003) Oversulfated dermatan sulfate exhibits neurite outgrowth-promoting activity toward embryonic mouse hippocampal neurons: implications of dermatan sulfate in neuritogenesis in the brain. J. Biol. Chem., 278, 43744–43754. Hildebrand, A., Romaris, M., Rasmussen, L.M., Heinegard, D., Twardzik, D.R., Border, W.A., and Ruoslahti, E. (1994) Interaction of the small interstitial proteoglycans biglycan, decorin and fibromodulin with transforming growth factor beta. Biochem. J., 302, 527–534. Hjerpe, A., Antonopoulos, C.A., and Engfeldt, B. (1979) Determination of sulphated disaccharides from chondroitin sulphates by high-performance liquid chromatography. J. Chromatogr., 171, 339–344. Iozzo, R.V. (1998) Matrix proteoglycans: from molecular design to cellular function. Annu. Rev. Biochem., 67, 609–652. Isogai, Z., Aspberg, A., Keene, D.R., Ono, R.N., Reinhardt, D.P., and Sakai, L.Y. (2002) Versican interacts with fibrillin-1 and links extracellular microfibrils to other connective tissue networks. J. Biol. Chem., 277, 4565–4572. Kähäri, V.M., Larjava, H., and Uitto, J. (1991) Differential regulation of extracellular matrix proteoglycan (PG) gene expression. Transforming


Equal amounts of cDNA from the RT-reaction were used as a template in the PCR. Reaction was performed in small glass capillaries in the Lightcycler machine from Roche (Bromma, Sweden). Each reaction contained: 1 × Lightcycler DNA Master SYBR Green (Roche), 0.5 μM of each primer, 3–4 mM MgCl2 and water to a total volume of 20 μL. After amplification, a melting curve analysis was performed, and no peaks for primer/dimer formation was noted. Relative quantification of mRNA expression was made based on differences in cycle number and normalized to the housekeeping gene 18S-RNA (Bustin, 2000, 2002). Analysis of the result was performed with the Lightcycler software. To confirm that the correct product had been amplified, the reaction mix was further analyzed on a 1.5% agarose gel.

References


Penc, S.F., Pomahac, B., Eriksson, E., Detmar, M., and Gallo, R.L. (1999) Dermatan sulfate activates nuclear factor-kappab and induces endothelial and circulating intercellular adhesion molecule-1. J. Clin. Invest, 103, 1329–1335. Penc, S.F., Pomahac, B., Winkler, T., Dorschner, R.A., Eriksson, E., Herndon, M., and Gallo, R.L. (1998) Dermatan sulfate released after injury is a potent promoter of fibroblast growth factor-2 function. J. Biol. Chem., 273, 28116–28121. Petersen, F., Brandt, E., Lindahl, U., and Spillmann, D. (1999) Characterization of a neutrophil cell surface glycosaminoglycan that mediates binding of platelet factor, 4. J. Biol. Chem., 274, 12376–12382. Rapraeger, A. (1989) Transforming growth factor (type beta) promotes the addition of chondroitin sulfate chains to the cell surface proteoglycan (syndecan) of mouse mammary epithelia. J. Cell Biol., 109, 2509–2518. Sato, T., Gotoh, M., Kiyohara, K., Akashima, T., Iwasaki, H., Kameyama, A., Mochizuki, H., Yada, T., Inaba, N., Togayachi, A., et al. (2003) Differential roles of two N-acetylgalactosaminyltransferases, CSGalNAcT-1, and a novel enzyme, CSGalNAcT-2. Initiation and elongation in synthesis of chondroitin sulfate. J. Biol. Chem., 278, 3063–3071. Seidler, D.G., Breuer, E., Grande-Allen, K.J., Hascall, V.C., and Kresse, H. (2002) Core protein dependence of epimerization of glucuronosyl residues in galactosaminoglycans. J. Biol. Chem., 277, 42409–42416. Silbert, J.E. and Sugumaran, G. (1995) Intracellular membranes in the synthesis, transport, and metabolism of proteoglycans. Biochim. Biophys. Acta, 1241, 371–384. Sugahara, K., Shibamoto, S., and Yamashina, I. (1985) Elevated chondroitin, 6-sulfotransferase activity in fetal bovine serum. FEBS Lett., 183, 43–46. Taylor, K.R., Rudisill, J.A., and Gallo, R.L. (2005) Structural and sequence motifs in Dermatan sulfate for promoting fibroblast growth factor-2 (FGF-2) and FGF-7 activity. J. Biol. Chem., 280, 5300–5306. Tiedemann, K., Malmstrom, A., and Westergren-Thorsson, G. (1997) Cytokine regulation of proteoglycan production in fibroblasts: separate and synergistic effects. Matrix Biol., 15, 469–478. Tiedemann, K., Larsson, T., Heinegard, D., and Malmstrom, A. (2001) The glucuronyl C5-epimerase activity is the limiting factor in the dermatan sulfate biosynthesis. Arch. Biochem. Biophys., 391, 65–71. Trowbridge, J.M. and Gallo, R.L. (2002) Dermatan sulfate: new functions from an old glycosaminoglycan. Glycobiology, 12, 117R–125R. Tufvesson, E., and Westergren-Thorsson, G. (2002) Tumour necrosis factoralpha interacts with biglycan and decorin. FEBS Lett., 530, 124–128. Westergren-Thorsson, G., Onnervik, P.O., Fransson, L.A., and Malmstrom, A. (1991) Proliferation of cultured fibroblasts is inhibited by L-iduronatecontaining glycosaminoglycans. J. Cell Physiol., 147, 523–530. Westergren-Thorsson, G., Schmidtchen, A., Sarnstrand, B., Fransson, L.A., and Malmstrom, A. (1992) Transforming growth factor-beta induces selective increase of proteoglycan production and changes in the copolymeric structure of dermatan sulphate in human skin fibroblasts. Eur. J. Biochem., 205, 277–286. Westergren-Thorsson, G., Persson, S., Isaksson, A., Onnervik, P.O., Malmstrom, A., and Fransson, L.A. (1993) L-iduronate-rich glycosaminoglycans inhibit growth of normal fibroblasts independently of serum or added growth factors. Exp. Cell Res., 206, 93–99. Westergren-Thorsson, G., Tufvesson, E., Eklund, E., and Malmstrom, A. (2002) The role of small proteoglycans in the formation of fibrosis. In Garg, H.G., Roughley, P.J., and Hales, C.A. (eds.), Proteoglycans in Lung Disease. Marcel Dekker, New York, pp.169–190. Yada, T., Gotoh, M., Sato, T., Shionyu, M., Go, M., Kaseyama, H., Iwasaki, H., Kikuchi, N., Kwon, Y.D., Togayachi, A. et al. (2003a) Chondroitin sulfate synthase-2. Molecular cloning and characterization of a novel human glycosyltransferase homologous to chondroitin sulfate glucuronyltransferase, which has dual enzymatic activities. J. Biol. Chem., 278, 30235–30247. Yada, T., Sato, T., Kaseyama, H., Gotoh, M., Iwasaki, H., Kikuchi, N., Kwon, Y.D., Togayachi, A., Kudo, T., Watanabe, H., et al. (2003b) Chondroitin sulfate synthase-3. Molecular cloning and characterization. J. Biol. Chem., 278, 39711–39725.

1285


growth factor-beta, 1 up-regulates biglycan (PGI), and versican (large fibroblast PG) but down-regulates decorin (PGII) mRNA levels in human fibroblasts in culture. J. Biol. Chem., 266, 10608–10615. Kang, H.G., Evers, M.R., Xia, G., Baenziger, J.U., and Schachner, M. (2001) Molecular cloning and expression of an N-acetylgalactosamine-4-Osulfotransferase that transfers sulfate to terminal and non-terminal beta, 1,4-linked N-acetylgalactosamine. J. Biol. Chem., 276, 10861–10869. Kang, H.G., Evers, M.R., Xia, G., Baenziger, J.U., and Schachner, M. (2002) Molecular cloning and characterization of chondroitin-4-O-sulfotransferase-3. A novel member of the HNK-1 family of sulfotransferases. J. Biol. Chem., 277, 34766–34772. Karamanos, N.K., Syrokou, A., Vanky, P., Nurminen, M., and Hjerpe, A. (1994) Determination of 24 variously sulfated galactosaminoglycanand hyaluronan-derived disaccharides by high-performance liquid chromatography. Anal. Biochem., 221, 189–199. Kawashima, H., Hirose, M., Hirose, J., Nagakubo, D., Plaas, A.H., and Miyasaka, M. (2000) Binding of a large chondroitin sulfate/dermatan sulfate proteoglycan, versican, to L-selectin, P-selectin, and CD44. J. Biol. Chem., 275, 35448–35456. Kawashima, H., Atarashi, K., Hirose, M., Hirose, J., Yamada, S., Sugahara, K., and Miyasaka, M. (2002) Oversulfated chondroitin/dermatan sulfates containing GlcAbeta1/IdoAalpha1–3GalNAc (4,6-O-disulfate) interact with L- and P-selectin and chemokines. J. Biol. Chem., 277, 12921–12930. Kluppel, M., Vallis, K.A., and Wrana, J.L. (2002) A high-throughput induction gene trap approach defines C4ST as a target of BMP signaling. Mech. Dev., 118, 77–89. Little, P.J., Tannock, L., Olin, K.L., Chait, A., and Wight, T.N. (2002) Proteoglycans synthesized by arterial smooth muscle cells in the presence of transforming growth factor-beta1 exhibit increased binding to LDLs. Arterioscler. Thromb. Vasc. Biol., 22, 55–60. Lowry, O.H., Rosebrough, N.J., Farr, A.L., and Randall. R.J. (1951) Protein measurement with the Folin phenol reagent. J. Biol. Chem., 193, 265–275. Maimone, M.M. and Tollefsen, D.M. (1990) Structure of a dermatan sulfate hexasaccharide that binds to heparin cofactor II with high affinity. J. Biol. Chem., 265, 18263–18271. Malmstrom, A. (1981) Biosynthesis of dermatan sulphate. Loss of C-5 hydrogen during conversion of D-glucuronate to L-iduronate. Biochem. J., 198, 669–675. Malmstrom, A. (1984) Biosynthesis of dermatan sulfate. II. Substrate specificity of the C-5 uronosyl epimerase. J. Biol. Chem., 259, 161–165. Malmstrom, A. and Fransson, L.A. (1975) Biosynthesis of dermatan sulfate. I. Formation of L-iduronic acid residues. J. Biol. Chem., 250, 3419–3425. Malmstrom, A. and Aberg, L. (1982) Biosynthesis of dermatan sulphate. Assay and properties of the uronosyl C-5 epimerase. Biochem. J., 201, 489–493. Malmstrom, A., Carlstedt, I., Aberg, L., and Fransson, L.A. (1975) The copolymeric structure of dermatan sulphate produced by cultured human fibroblasts. Different distribution of iduronic acid and glucuronic acid-containing units in soluble and cell-associated glycans. Biochem. J., 151, 477–489. Mikami, T., Mizumoto, S., Kago, N., Kitagawa, H., and Sugahara, K. (2003) Specificities of three distinct human chondroitin/dermatan N-acetylgalactosamine, 4-O-sulfotransferases demonstrated using partially desulfated dermatan sulfate as an acceptor: implication of differential roles in dermatan sulfate biosynthesis. J. Biol. Chem., 278, 36115–36127. Miyoshi, E., Nishikawa, A., Ihara, Y., Saito, H., Uozumi, N., Hayashi, N., Fusamoto, H., Kamada, T., and Taniguchi, N. (1995) Transforming growth factor beta up-regulates expression of the N-acetylglucosaminyltransferase V gene in mouse melanoma cells. J. Biol. Chem., 270, 6216–6220. Pasonen-Seppanen, S., Karvinen, S., Torronen, K., Hyttinen, J.M., Jokela, T., Lammi, M.J., Tammi, M.I., and Tammi, R. (2003) EGF upregulates, whereas TGF-beta downregulates, the hyaluronan synthases Has2 and Has3 in organotypic keratinocyte cultures: correlations with epidermal proliferation and differentiation. J. Invest Dermatol., 120, 1038–1044.