Plasmin- and thrombin-accelerated shedding of syndecan-4 ectodomain generates cleavage sites at Lys114 - Arg115 and Lys129 - Val130 bonds

Annette Schmidt1, Frank Echtermeyer2, Anthony Alozie1, Kerstin Brands2

and Eckhart Buddecke1,2

Institute of Arteriosclerosis Research1, Department of Physiological Chemistry and Pathobiochemistry2, University Hospital Muenster, Germany

Running title: Protease mediated shedding of syndecan-4 Address correspondence to: Annette Schmidt, Institute of Arteriosclerosis Research,

Department of Molecular Cardiology, Domagkstr. 3, D-48149 Muenster, Germany, Phone: +49-251-835 8626, Fax: +49-251-835 8628, E-mail: annschm@uni-muenster.de

Syndecans are transmembranous heparan sulfate proteoglycans abundant in the surface of all adherent mammalian cells and involved in vital cellular functions. In this study we found the syndecans-1, -2 -3 and -4 to be constitutively expressed by human umbilical vein endothelial cells. The exposure of the ectodomains of syndecan-1 and –4 to the cell surface and their constitutive shedding into the extracellular compartment was measured by immunoassays. In the presence of plasmin and thrombin shedding is accelerated and was monitored by detection and identification of 35S-labelled proteoglycans. In order to elucidate the cleavage site of the syndecan ectodomains we used a cell-free in vitro system with enzyme and substrate as the only reactants. For this purpose we constructed recombinant fusion proteins of the syndecan-1 and -4 ectodomain together with maltose binding protein and EYFP as reporter proteins attached to the N- and C-terminus via oligopeptide linkers. After protease treatment of the fusion proteins the electrophoretically resolved split products were sequenced and cleavage sites of the ectodomain were identified. Plasmin generated cleavage sites at Lys114-Arg115 and Lys129-Val130 in the ectodomain of syndecan-4. In thrombin proteolysates of the syndecan-4 ectodomain also the cleavage site Lys114-Arg115

was identified. The cleavage sites for plasmin and thrombin within the syndecan-4 ectodomain are not present in the syndecan-1 ectodomain. Cleavage of the syndecan-1 fusion protein by thrombin occured only at a control cleavage site (R-G) introduced into the linker region connecting the ectodomain with the EYFP. Since both, plasmin and thrombin are involved in thrombogenic and thrombolytic processes in the course of the pathogenesis of arteriosclerosis the detachment of heparan sulfate-bearing ectodomains could be relevant for the

development of arteriosclerotic plaques and recruitment of mononuclear blood cells to the plaque. The family of syndecans form a group of transmembrane heparan sulfate proteoglycans (HSPG)1 that are composed of a core protein with covalently attached glycosaminoglycan chains. The four mammalian syndecan genes (syndecan-1, -2, -3, -4) have been cloned and sequenced and are expressed in most human cells and tissues (for review see 1,2) including human umbilical vein endothelial cells (3,4). The syndecans contain an N-terminal extracellular domain or ectodomain, a hydrophobic transmembrane domain and a short C-terminal cytoplasmic domain. The ectodomain bears near the N-terminus three consecutive consensus Ser-Gly sequences for heparan sulfate chain attachment and may also contain Ser-Gly sequences near the plasma membrane that serve as attachment sites for chondroitin sulfate side chains. The length of the ectodomains varies markedly among family members, while the length of the transmembrane and cytoplasmic domain is highly conserved (5). The function of syndecans include anchorage of cells to extracellular matrix components with associated heparan sulfate binding domains (6), maintainance of epithelial and endothelial morphology (1), binding to and modulation of the activity of heparan sulfate binding growth factors (7) and modulation of the activity of several proteases and their inhibitors (for review see 2,8) but they may also serve as signalling molecules (9,10) and as arterial counterparts for monocyte L-selectin in the vascular endothel (10). The syndecan ectodomains are released from the cell surface in a process commonly known as shedding (10-12). Shedding of the ectodomain occurs as part of the normal turnover and involves the activity of a not identified cell surface zinc metalloproteinase that is specifically inhibited by TIMP-3 (8). So far the site of

1

JBC Papers in Press. Published on August 8, 2005 as Manuscript M501903200

Copyright 2005 by The American Society for Biochemistry and Molecular Biology, Inc.

by guest on February 13, 2018http://w

ww

.jbc.org/D

ownloaded from

cleavage has not been identified. For syndecan-1 a cleavage site has been localized within nine amino acid adjacent to the extracellular face of the plasma membrane (13). Shedding of syndecan-1 and syndecan-4 can be regulated by various external stimuli and intracellular signalling pathways such as growth factors (14), cell stress (13) or soluble microbial pathogens (15,16). Shedding of syndecan-1 and syndecan-4 can be accelerated via receptor activation such as the thrombin receptor and EGF receptor family or by direct action of proteases. Subramanian et al. (14) described an accelerated shedding by plasmin and thrombin and the presence of the soluble ectodomains in fluids accumulating following injury and inflammation but no informations about the cleavage site of the ectodomains are available. In the present study we investigated the plasmin- and thrombin-mediated shedding of syndecan-1 and syndecan-4 in a cell-free in vitro system and identified in the ectodomain of syndecan-4 two cleavage sites for plasmin and one for thrombin. For this purpose recombinant fusion proteins were constructed that contained the ectodomain of syndecan-1 or syndecan-4 with reporter proteins attached to the N- and C-terminus via linker oligopeptides. The pathophysiological relevance of our findings for the pathogenesis of arteriosclerosis lays in the fact that syndecans are constitutively expressed by endothelial cells of the vascular system and that plasmin and thrombin are involved in thrombogenic and thrombolytic processes at the site of arteriosclerotic lesions.

EXPERIMENTAL PROCEDURES

Materials- Sodium [35S]sulfate (carrier free, 0.8-1.5 TBq mg-1 sulfur) was obtained from ICN Biomedicals GmbH (Eschwege, EU). Heparan sulfate lyase (heparitinase, EC 4.2.2.8), chondroitin sulfate lyase (chondroitinase ABC, EC 4.2.2.4) were from Medac (Hamburg, EU). Plasmin (human plasma fibrinolysin, EC 3.4.21.7) 5.7 U/mg protein and thrombin (human plasma, factor IIa, EC 3.4.21.5) 2800 NIH U/mg protein were from SIGMA (Muenchen, EU). Monoclonal mouse anti-syndecan-1 (DL101), anti-syndecan-4 (5G9), polyclonal rabbit anti-syndecan-1 (H174) and anti-syndecan-4 (H140) were from Santa Cruz, polyclonal goat anti-mouse IgG HRP-conjugated (P-0447) from DakoCytomation and polyclonal goat anti-rabbit IgG HRP-conjugated from Vector Lab. All other chemicals were of analytical grade or the best grade available.

Cell culture- HUVEC (human umbilical vein endothelial cells) were harvested from fresh human umbilical cords as previously described (15) Primary isolated cells were cultured in gelatin-precoated tissue flasks at 37°C under 5% CO2/95% air. Culture medium consisted of RPMI 1640 (Life Technologies) and supplements as previously described (17) including 10 µg/mL ciprofloxacin (Bayer, EU). Cultures of the 2nd to 5th passage were used for the experiments. Immunoassays- Cell surface exposed syndecan ectodomains were identified according to a protocol described previously (17) for cell adhesion molecules. Native cells were exposed to monoclonal anti-syndecan-1 or -4 antibodies (1:200) which recognize the glycosylated form of syndecans. Plates were finally incubated with HRP-conjugated secondary antibodies (1:10,000). Sandwich immunoassays were designed for detection of the shed syndecan-1 and -4 ectodomains according to Rioux et al. (18) with the exception that mouse monoclonal antibodies were employed as capture, polyclonal rabbit anti-syndecans as secondary antibodies and HRP-conjugated goat anti-rabbit IgG (1:50,000) for colour development. The serum-free cell supernatant containing the shed ectodomains of syndecan-1 and -4 was freeze-dried, redissolved in a small volume, dialyzed and deglycosylated by heparitinase and chondroitin sulfate lyase for an enhanced binding of antibodies. For Western blot analysis syndecan-1 and -4 were collected from the cell supernatant by protein-A sepharose precoated with both polyclonal anti-syndecan-1 and anti-syndecan-4 antibodies that recognize the glycosylated form of syndecans. Thereafter the protein-A sepharose suspension was degraded by heparitinase and chondroitin sulfate lyase, boiled in SDS buffer and submitted to PAGE. The blotted deglycosylated syndecans were detected according to a standard protocol using monoclonal anti-syndecan-1 or anti-syndecan-4 antibodies. Construction of syndecan-1 and syndecan-4 ectodomain fusion proteins- The pMAL-c2 vector (New England Biolabs GmbH, Frankfurt /Main, EU) was used to express and purify fusion proteins produced from the cloned ectodomain of syndecan-1 and syndecan-4 with 5’linker to the maltose binding protein (MBP) DNA and 3’linker to the enhanced yellow fluorescent protein (EYFP) DNA. The coding sequence of the syndecan-1 and syndecan-4 ectodomains were amplified by PCR from

2

by guest on February 13, 2018http://w

ww

.jbc.org/D

ownloaded from

normal human fibroblast cDNA using primers containing a 5’EcoRI and a 3’XbaI site (syndecan-1) or a 5’EcoRI and 3’KpnI site (syndecan-4). The sequence coding for EYFP was amplified from the plasmid pIRES-EYFP (BD Biosciences/Clontech) using primers containing a 5’XbaI and 3’HindIII site. The syndecan-1 and EYFP PCR products were digested with XbaI, ligated and digested again with EcoRI and HindIII. The purified Syn-1-EYFP product was cloned into the EcoRI and HindIII sites of the pMal-c2 expression vector downstream of the MalE gene coding for MBP. For construction of the syndecan-4 fusion protein a different cloning protocol was used. First, the EYFP PCR product was digested with XbaI and HindIII and ligated into the respective sites of pUC19 vector (Invitrogen), which was designated as pUC19-EYFP. Second, the syndecan-4 PCR fragment was cloned via the EcoRI and KpnI sites upstream of EYFP into the pUC19-EYFP vector. Third, after digestion with EcoRI and HindIII the syndecan-4-EYFP construct was cloned downstream of MPB into the respective sites of the pMAL-c2 expression vector. After sequence verification the plasmids for MBP-Syn-1-EYFP and MBP-Syn-4-EYFP fusion products were transfected into E. coli BL21 plysS cells. Overexpression and purification of the MBP-fusion proteins were performed following the manufacturers protocol. Briefly E. coli BL21 plysS cells were grown until the OD600 reached a value of 0.7. Then expression was induced with IPTG (0.3 mM) for 3 h. Cells were harvested by centrifugation, resuspended in column buffer (20 mM Tris/HCl pH 7.4, 200 mM NaCl, 1 mM EDTA) and sonicated. Cell extracts were centrifuged and the supernatant was bound to a 5 mL amylose resin column. After extensive washing with column buffer the fusion proteins were eluted from the column with column buffer containing 10 mM maltose. The elution fractions were examined on SDS-PAGE to check for their purity. Fractions containing recombinant MBP-Syn-1-EYFP or MBP-Syn-4-EYFP protein were pooled and used for further digestions with plasmin or thrombin. Expression of syndecan-1 and -4 on

transcriptional and translational level- Human umbilical vein endothelial cells express syndecan-1, -2, -3 and –4 in cell culture at the cell surface. Fig. 1A shows the cDNA transcripts of the corresponding mRNA. The exposure of syndecan to the cell surface was evidenced by immunoassays (see Experimental Procedures) using monoclonal antibodies directed against the ectodomain of syndecan-1 and -4. The values

Shedding of cell membrane- integrated/associated sulfated proteoglycans- 50 000 human umbilical vein endothelial cells were seeded in 35 mm diameter plastic dishes, cultured to confluence and labelled with 370 kBq (10 µCi) [35S]sulfate/mL medium for 48 h. Then, the cultures were washed three times with PBS and cells were incubated in 1 mL chase medium (SFM, Gibco) containing 0.5 mg BSA, 10 µg

ciprofloxacin/mL in HEPES 20 mM in the absence or presence of 15 mU plasmin or 15 U thrombin for 3-6 h. In pilot experiments the concentration of plasmin and thrombin was adjusted to maximum activity in our system. At the end of the experiment the cell free medium and 1 mL washing solution (PBS) were pooled (chase medium) and stored at 4°C until use. The pericellular compartment was obtained by trypsinization of the cells and centrifugation at 800 x g for 3 min. (trypsin pool). Chase medium and trypsin pool were used for further analysis. Reference values for the cell number/dish were obtained from parallel cultures. Chase-medium and trypsin pool were used in direct route or after digestion with chondroitinase ABC for 4 h at 37°C, lyophylized and submitted to a size exclusion chromatography on calibrated 25 mL Sephadex G 50 medium columns equilibrated in 1 M NaCl. The 35S-radioactivity of the V0 fraction represents the total amount of total 35S-proteoglycans or after chondroitinase ABC degradation the 35S-heparan sulfate containing proteoglycans. The chondroitin sulfate/dermatan sulfate containing proteoglycans are calculated as difference of total and heparan sulfate containing proteoglycans. Electrophoresis and microsequence analysis- SDS-PAGE was performed under reducing conditions using a 12% polyacrylamide separating gel and a 3.5% stacking gel. The ectodomains were treated (4 h, 37°C) with plasmin or thrombin at an enzyme/substrate ratio of 1:5 (w/w) in a total volume of 50 µL. The digest was diluted with concentrated sample buffer to the standard SDS concentration used for electrophoresis, boiled for 5 min. and separated by electrophoresis (MiniProtean II, BioRad), transferred to PVDF-membrane and the resolved fusion protein fragments were stained with ponceau red. Bands indicated in Fig. 4B and C were cut and subjected to N-terminal sequencing (10 cycles) on an Applied Biosystems 492 gas-phase protein sequencer.

RESULTS

3

by guest on February 13, 2018http://w

ww

.jbc.org/D

ownloaded from

given in Fig. 1B depend on different affinity of the antibodies and do not reflect differences of concentration. For detection of the shed ectodomain sandwich immunoassays for syndecan-1 and -4 were designed. Since quantified standards were not available the values are given in arbitrary units normalized for cell number. Western blot analysis demonstrates the deglycosylated protein core of syndecan-1 and -4. Fig. 1C shows bands in the order of 25-35 kDa but a correct molecular size cannot be predicted since the cleaving sites of the ectodomain generated under cell culture conditions are not known. Kinetics of constitutive and accelerated shedding of 35S-labelled proteoglycans- The shedding of sulfated proteoglycans was monitored after pulse labelling of the cells with [35S]sulfate for 48 h. The kinetics of the constitutive and protease-accelerated shedding is shown in Fig. 2A. After a hyperbolic increase of 35S-labelled material a nearly constant level is reached after 3-6 h. The constitutive shedding is accelerated in the presence of plasmin or thrombin 1.5-fold over the control values (p< 0.05). The 35S-labelled material that was shed into the medium includes the 35S-labelled heparan sulfate and chondroitin sulfate bearing ectodomain of syndecans but also chondroitin/dermatan sulfate (CS/DS)-containing proteoglycans exported by HUVEC under physiological conditions. The shedding process can be followed also by examination of the cell membrane integrated/associated 35S-labelled material after releasing it from the cell surface by trypsin treatment (Fig. 2B). In control cells (constitutive shedding) the zero time value 5490 cpm/105 cells (100%) decreases within 6 h to 3304 cpm/105 cells (60.1%). In the presence of thrombin a decrease to 40% and of plasmin to 33.8% was found (p<0.05). No significant difference between the effect of plasmin and thrombin could be observed. The ectodomains of all syndecans have attachment sites for heparan sulfate at the N-terminus. In addition the syndecans-1 and -3 posses two and syndecan-4 one membrane-proximal attachment sites for chondroitin sulfate. In order to evaluate the heparan sulfate-specific shedding in further experiments the 35S-labelled CS chains of syndecan and other CS and DS containing proteoglycans were eliminated by exhaustive degradation of the shed 35S-labelled material with chondroitin sulfate lyase (EC 4.2.2.4). Fig. 3A shows that more than 50% of total proteoglycans shed into the medium

accounts for heparan sulfate in plasmin or thrombin treated cells suggesting that the protease accelerated shedding involves the membrane integrated syndecans. The cell-associated 35S-radioactivity obtained by and quantified after trypsin treatment of the cells is shown in Fig. 3B. The shedding activity of the serine proteases plasmin and thrombin can be completely inhibited by APMSF (4-amidinophenyl-methanesulfonyl fluoride) (not shown). Construction of ectodomain fusion proteins- To elucidate the protein cleavage site of syndecans generated in response to plasmin and thrombin treatment, cell-free in vitro experiments with recombinant ectodomains of syndecan-1 and -4 were performed. This system contained substrate (fusion protein) and enzyme (plasmin, thrombin) as the only reactants excluding the cellular shedding activities. As the cleavage site was suggested to be proximal to the C-terminus of the ectodomains split products with a free N-terminus were expected to be in the order of oligopeptides containing a few amino acids only. For a better detection of split products the syndecan ectodomains were linked with EYFP at the C-terminus by a decapeptide linker. The N-terminus of the ectodomain was tagged via an octapeptide linker with MBP to facilitate purification of the fusion protein. Plasmin cleaves the syndecan-4 fusion protein at the Lys114 - Arg115 and at the Lys129 - Val130 bond- SDS-PAGE of the affinity purified fusion protein revealed beside the expected MBP-Syn-4 ectodomain-EYFP (predicted molecular weight 92.5 kDa) fragments with lower molecular weight obviously due to a truncated synthesis. A control blot of the native syndecan-4 fusion protein is shown in Fig. 4A and gave a similar band pattern for syndecan-1 fusion protein (not shown). The fusion protein of the syndecan-4 ectodomain was used as substrate for plasmin. The proteolytic split products obtained after plasmin treatment were resolved by SDS-PAGE and the bands I-VI were excised and analyzed by gas-phase protein sequencing. Fig. 4B shows the band pattern of syndecan-4 fusion protein obtained after incubation with plasmin. Bands I-III showed the sequence MKTEEGKLVI corresponding to the N-terminal sequence of MBP, band IV gave ambiguous values. The bands V and VI revealed the N-terminal sequences RISPYEESE (band V) and VSMSSTVQG (band VI). These sequences indicate cleavage sites between Lys114-Arg115 and between Lys129-Val130 of the ectodomain of

4

by guest on February 13, 2018http://w

ww

.jbc.org/D

ownloaded from

syndecan-4 localized in a distance from the cell membrane of 33 and 17 amino acids. Thrombin cleaves the syndecan-4 fusion protein at the Lys114 - Arg115 bond and the bond between the C-terminal linker oligopeptide and EYFP. The degradation pattern of the syndecan-4 ectodomain fusion protein after degradation with thrombin is shown in Fig. 4C. The bands I-III had the same N-terminal sequence (MKTEEGKLVI) as the corresponding bands obtained by plasmin treatment. Sequence analysis of the band V gave an identical sequence as the corresponding plasmin-generated band (RISPVEESE). Thus, plasmin and thrombin recognize the same sequence (Lys-Arg) and deliver identical split products. The linker oligopeptide connecting the ectodomain and EYFP that had the C-terminal sequence RGAG served as a control for the proteolytic activity and specificity of thrombin. Thrombin is known to cleave the Arg-Gly bond of fibrinogen thereby converting fibrinogen to fibrin monomers. Thus, the second split product after thrombin-treatment (band VI) gave the expected sequence GAGMVSKGEE, containing the start sequence of EYFP (MVSKGEE) with the last 3 amino acids of the decapeptide linker at the N-terminus. Table 1 shows the amino acid sequence of the complete fusion protein of syndecan-4 and the position of the identified cleavage sites. The sequences cleaved in the syndecan-4 ectodomain are not present in syndecan-1 and therefore neither plasmin nor thrombin catalyzed cleavage of the ectodomain could be observed. However, plasmin-specific cleavage sites we found to exist in MBP (K-V, K-R) and EYFP (K-R) of the fusion proteins. The corresponding split products of MBP were not found within the analysed band I-VI and the molecular weight of the split products of EYFP were too low for detection in the used electrophoretic system.

DISCUSSION

Fundamental data on enzymatic shedding of syndecans have been reported (1,2,5,8,9,13) but cleavage sites for plasmin and thrombin – two enzymes involved in thrombogenic and thrombolytic processes within the circulation – have not been described. The present studies demonstrate by a cell-free system that plasmin and thrombin can hydrolyse the Lys114-Arg115 and Lys129-Val130 peptide bonds of the ectodomain of syndecan-4 independently of the cellular shedding activities. The cleavage sites found are located within the juxtamembrane

domain of the syndecan-4 molecule in a distance of 33 and 17 amino acids to the cell membrane. These data identify and localize for the first time two cleavage sites of the syndecan-4 ectodomain. No cleavage of the syndecan-1 ectodomain was observed under the action of thrombin and plasmin. This is in apparent contrast to a report about an accelerated shedding of syndecan-1 by thrombin (14). However, the data of this report (14) are based on experiments with cultured endothelial cells where a direct proteolytic action of thrombin cannot be shown. However, these data (14) indicate a receptor-mediated proteolysis of syndecan-1 and syndecan-4 under cell culture conditions. Thus, it appears that the thrombin accelerated shedding requires the interaction of thrombin with the cellular thrombin receptor and that the thrombin effect can also be triggered by the thrombin receptor agonist peptide L-1-tosylamido-2-phenylethyl-chlormethyl ketone that is proteolytically inactive by itself. Moreover, accelerated shedding of syndecan-1 and -4 could also be achieved by PMA (phorbol 12-myristate 13-acetate), EGF (epidermal growth factor), HB-EGF (heparin binding EGF) and TNF-alpha (tumor necrosis factor). Thus, the authors conclude that at least activation of two receptor classes, namely the G-protein-linked thrombin receptor and the EGF-specific tyrosine kinase receptor regulate syndecan shedding by activation of a not known probably cell membrane-associated proteolytic system. As other growth factors such as VEGF (vascular endothelial growth factor), FGF-2 (basic fibroblast growth factor), PDGF (platelet derived growth factor) and TGF-beta (transforming growth factor) failed to accelerate syndecan shedding (14), the growth factor-induced shedding is considered to be selective. These findings suggest that distinct signalling pathways converge to activate a cell surface associated metalloproteinase that cleaves the syndecans by a common but unknown mechanism. Taken together our data and the findings of (14), we conclude that thrombin/plasmin accelerated shedding may result, at least partly, from a direct proteolytic cleavage of the syndecan-4 ectodomain, while the thrombin-mediated cleavage of syndecan-1 is restricted to an indirect receptor-triggered shedding. This is in accordance with differences of the primary structure of syndecan-4 and -1. The amino acid sequences we found susceptible for direct thrombin and plasmin cleavage of syndecan-4 are not present in syndecan-1. Furthermore,

5

by guest on February 13, 2018http://w

ww

.jbc.org/D

ownloaded from

shedding by plasmin and thrombin releases the ectodomain of syndecan-4 that still contains the complete set of attachment sites for 3 heparan sulfate chains (SG consensus sequences 39/40, 61/62 and 62/63) and for 2 chondroitin sulfate chains (95/96). These covalently linked glycosaminoglycan chains are absent in our fusion proteins and therefore, it remains to be clarified whether the native ectodomain bearing the complete set of heparan- and chondroitin sulfate chains would also be split off by plasmin or thrombin. Matrix metalloproteinases (MMPs) and membrane-type MMPs (MT1-MMPs) may also be involved in shedding and catabolic processes of syndecans. Thus, the ectodomain of the basic fibroblast growth factor receptor (FGFR-1) could be split off by MMP-1 (15). Matrilysin (MMP-7) was shown to mediate shedding of a syndecan-1/CXC chemokine (KC) complex (16). In addition syndecan-1 has been identified as a substrate of MT1-MMP that cleaves the Gly245-Leu246 bond of a recombinant syndecan-1 fusion protein (19). However, it has to be clarified whether MMPs and MT-MMPs are involved in constitutive shedding by themselves or whether they merely induce it. Our present knowledge of syndecan points to different functions of individual syndecan types. Thus, syndecan-4 is capable of binding PIP2 and activating protein kinase C and to confer signalling functions on syndecan-4 as bFGF coreceptor (20,21) whereas such a property is not known for syndecan-1. Furthermore, syndecan-4-specific functions concern the link to antithrombin III (22), the upregulation of IL-1ß in response to LPS (20) and the predominant expression of syndecan-4 in chronic venous ulcers where syndecan-1 is expressed in smaller quantities. These different functions are accentuated by a different sensitivity to the proteolytic activity of plasmin and thrombin that are able to cleave the ectodomain of syndecan-4 directly but not that of syndecan-1. The susceptibility of syndecan shedding to multiple effectors (13,14,18,23,24) that act directly or via receptors implies that the ectodomains of syndecans have physiological roles as soluble proteoglycans. Shed ectodomains retain their heparan sulfate chains and hence the ligand binding activity of their cell surface counterparts. The release of the ectodomain of syndecan-4 leads to a downregulation of its membrane-anchored coreceptor function but confers the ectodomain

new functions as soluble effector which can compete for ligands with the membrane integrated counterparts or other encounters as extensively reviewed (24-29). It is tempting to speculate, based on our data, that cleavage of the ectodomains of syndecans expressed by vascular endothelial and smooth muscle cells could determine the development and fate of arteriosclerotic plaques. Syndecan-4 has been shown to modulate events relevant to arteriosclerosis such as acute tissue repair (30), cell motility (31), focal adhesion formation (32) and matrix remodeling (33). Shedding of syndecans could lead to an accumulation of soluble HS-containing ectodomains that may interfere with the membrane bound syndecan-4 functioning as bFGF coreceptor (34). This would suppress the mitogenic activity of bFGF (35), abolish the lipoprotein lipase binding (36,37) and the anticoagulative properties of membrane integrated syndecans. Furthermore, the endothelial syndecans could interact via their heparan sulfate chains with L- and P-selectin expressed by mononuclear blood cells (10,38) and thereby modulate leukocyte rolling as well as binding and recruitment of leukocytes into arteriosclerotic plaques. Moreover, the soluble ectodomain can modify the activity of leukocyte-derived elastase and cathepsin G (39) or bind these proteases and reduce their interaction with the subendothelial fibrous cap of arteriosclerotic plaques. Chemokine binding of syndecans has a high impact for the regulation of inflammatory events. Syndecan-1 and/or –4 can generate a transendothelial IL-8 gradient by binding IL-8 via their HS chains (40) thereby increasing leukocyte/monocyte transmigration into arteriosclerotic lesions. Destroying the gradient by activated plasmin, which shed the syndecan-IL-8 complex (40) results in a reduced recruitment of mononuclear cells to arteriosclerotic plaques. The syndecan-4 turnover of rat aortic smooth muscle cells is regulated by oxidized linoleic acid and linolic acid, the major oxidable fatty acids in LDL (41). Both oxidized fatty acids induce a dose-dependent upregulation of syndecan-4 mRNA expression but simultaneously the oxidized linoleic acid (13-hydroperoxy-9,11-octadecadienoic acid) induced an accelerated shedding of syndecan-4 that would increase the amount of mobile HS-bearing ectodomains. As netto effect of syndecan overexpression and the associated accelerated shedding a variety of pro- and anti-atherogenic events could be created.

6

by guest on February 13, 2018http://w

ww

.jbc.org/D

ownloaded from

REFERENCES

1. Bernfield, M., Gotte, M., Park, P. W., Reizes, O., Fitzgerald, M. L., Lincecum, J., and Zako, M. (1999) Annu. Rev. Biochem. 68, 729-777

2. Jaakkola, P., Kainulainen, V., and Jalkanen, M. (2000) Syndecans, 115-145. In: Iozzo, R. V., editor. Proteoglycans, Marcel Dekker Inc., New York-Basel

3. Halden, Y., Rek, A., Atzenhofer, W., Szilak, L., Wabnig, A., and Kungl, AJ. (2004) Biochem. J. 377, 533-538

4. Mertens, G., Cassiman, JJ., Van den Berghe, V., Vermylen, J., and David, G. (1992) J. Biol. Chem. 267, 20435-20443

5. Horowitz, A., and Simons, M. (1998) J. Biol. Chem. 273, 10914-10918 6. Jackson, R. L., Busch, S. J., and Cardin, A. D. (1991) Physiol. Rev. 71, 481-539 7. Zimmermann, P., and David, G. (1999) FASEB J. 13 Suppl, S91-S100 8. Park, P. W., Reizes, O., and Bernfield, M. (2000) J. Biol. Chem. 275, 29923-29926 9. Rapraeger, A. C. (2000) J. Cell Biol. 149, 995-998 10. Giuffre, L., Cordey, A. S., Monai, N., Tardy, Y., Schapira, M., and Spertini, O. (1997) J. Cell

Biol. 136, 945-956 11. Jalkanen, M., Rapraeger, A., Saunders, S., and Bernfield, M. (1987) J. Cell Biol. 105, 3087-

3096 12. Yanagishita, M., and Hascall, V.C. (1992) J. Biol. Chem. 267, 9451-9454 13. Fitzgerald, M. L., Wang, Z., Park, P. W., Murphy, G., and Bernfield, M. (2000) J. Cell Biol.

148, 811-824 14. Subramanian, S.V., Fitzgerald, M.L., and Bernfield, M. (1997) J. Biol. Chem. 272, 14713-

14720 15. Levi, E., Fridman, R., Miao, H.Q., Ma, Y.S., Yayon, A., and Vlodavsky, I. (1996) Proc. Natl.

Acad. Sci. U.S.A 93, 7069-7074 16. Li, Q., Park, P. W., Wilson, C. L., and Parks, W. C. (2002) Cell 111, 635-646 17. Schmidt, A., Goepfert, C., Feitsma, K., and Buddecke, E. (2002) Atherosclerosis 164, 57-64 18. Rioux, V., Landry, RY., and Bensadoun A. (2002) J. Lipid Res. 43, 167-173 19. Endo, K., Takino, T., Miyamori, H., Kinsen, H., Yoshizaki, T., Furukawa, M., and Sato, H.

(2003) J. Biol. Chem. 278, 40764-40770 20. Horowitz, A., Tkachenko, E., and Simons, M. (2002) J. Cell Biol. 157, 715-725 21. Carey, D. J. (1997) Biochem. J. 327, 1-16 22. Salmivirta, M., Lidholt, K., and Lindahl, U. (1996) FASEB J. 10, 1220-1279 23. Ishiguro, K., Kadomatsu, K., Kojima, T., Muramatsu, H., Iwase, M., Yoshikai, Y., Yanada,

M., Yamamoto, K., Matsushita, T., Nishimura, M., Kusugami, K., Saito, H., and Muramatsu, T. (2001) J. Biol. Chem. 276, 47483-47488

24. Park, P.W., Pier, G.B., Preston, M.J., Goldberger, O., Fitzgerald, M.L., and Bernfield, M. (2000) J. Biol. Chem. 275, 3057-3064

25. Aviezer, D., Levy, E., Safran, M., Svahn, C., Buddecke, E., Schmidt, A., David, G., Vlodavsky, I., and Yayon, A. (1994) J. Biol. Chem. 269, 114-121

26. Gotte, M. (2003) FASEB J. 17, 575-591 27. Kato, M., Wang, H., Kainulainen, V., Fitzgerald, M.L., Ledbetter, S., Ornitz, D.M., and

Bernfield, M. (1998) Nat. Med. 4, 691-697 28. Park, P.W., Foster, T.J., Nishi, E., Duncan, S.J., Klagsbrun, M., and Chen, Y. (2004) J. Biol.

Chem. 279, 251-258 29. Tumova, S., Woods, A., and Couchman, J.R. (2000) Int .J. Biochem. Cell Biol. 32, 269-288 30. Echtermeyer, F., Streit, M., Wilcox-Adelman, S., Saoncella, S., Denhez, F., Detmar, M., and

Goetinck, P. (2001) J. Clin. Invest 107, R9-R14 31. Cizmeci-Smith, G., Langan, E., Youkey, J., Showalter, L.J., and Carey, D.J. (1997)

Arterioscler. Thromb. Vasc. Biol. 17, 172-180 32. Echtermeyer, F., Baciu, P.C., Saoncella, S., Ge, Y., and Goetinck, P. F. (1999) J. Cell Sci. 112

( Pt 20), 3433-3441 33. Woods, A., and Couchman, J.R. (1994) Mol. Biol. Cell 5, 183-192 34. Schmidt, A., Lemming, G., Yoshida, K., and Buddecke, E. (1992) Eur. J. Cell Biol. 59, 322-

328

7

by guest on February 13, 2018http://w

ww

.jbc.org/D

ownloaded from

35. Schmidt, A., Yoshida, K., and Buddecke, E. (1992) J. Biol. Chem. 267, 19242-19247 36. Fuki, I.V., Kuhn, K.M., Lomazov, I.R., Rothman, V.L., Tuszynski, G.P., Iozzo, R.V.,

Swenson, T.L., Fisher, E.A., and Williams, K.J. (1997) J. Clin. Invest 100, 1611-1622 37. Mead, J.R., Cryer, A., and Ramji, D.P. (1999) FEBS Lett. 462, 1-6 38. Koenig, A., Norgard-Sumnicht, K., Linhardt, R., and Varki, A. (1998) J. Clin. Invest 101,

877-889 39. Kainulainen, V., Wang, H., Schick, C., and Bernfield, M. (1998) J. Biol. Chem. 273, 11563-

11569 40. Gotte, M., and Echtermeyer, F. (2003) The Scientific World Journal 3, 1327-1331 41. Houston, M., Julien, M.A., Parthasarathy, S., and Chaikof, E.L. (2005) Am. J. Physiol. Cell.

Physiol. 288, 458-466 FOOTNOTES 1The abbreviations used are: BSA, bovine serum albumin; HSPG, heparan sulfate proteoglycans; HUVEC, human umbilical vein endothelial cells; MBP, mannose binding protein; EYFP, enhanced yellow fluorescent protein; PBS, phosphate-buffered saline; SD, standard deviation; HS, heparan sulfate; CS, chondroitin sulfate; DS, dermatan sulfate; PMA, phorbol 12-myristate 13-acetate; EGF, epidermal growth factor; HB-EGF, heparin binding EGF; TNF, tumor necrosis factor; VEGF, vascular endothelial growth factor; FGF-2, basic fibroblast growth factor; TGF, transforming growth factor. ACKNOWLEDGEMENTS This work was financially supported by the Deutsche Forschungsgemeinschaft (SFB 492, project B12) to F.E. The authors would like to thank Michaela Tirre for skillful technical assistance.

FIGURE LEGENDS Figure 1. Evidence of syndecans on transcriptional and translational level. (A) Total RNA was isolated from confluent cultures (HUVEC). 2 µg of RNA were reversed transcribed into cDNA and submitted to PCR using syndecan specific forward and reverse oligonucleotide primers. PCR products were visualized by agarosegel electrophoresis. The bands are representative for five experiments. (B) Exposure of syndecan-1 and -4 to the cell surface. Confluent HUVEC cultures were incubated with monoclonal mouse anti-syndecan-1 or -4 antibodies and with HRP-conjugated goat anti-mouse IgG. Syndecans shed into the cell supernatant were detected by sandwich immunoassays designed for syndecan-1 and -4 (see Experimental Procedures). Since quantified standards were not available values are expressed in arbitrary units normalized for cell number. (C) Western blot analysis of deglycosylated shed syndecans. Syndecans of the cell supernatant were collected by binding to protein A sepharose precoated with a combination of polyclonal rabbit anti-syndecan-1 and -4 antibodies, degraded with heparitinase and chondroitin lyase and submitted to PAGE. The blotted syndecans were made visible by monoclonal anti-syndecan-1 or -4 antibodies and HRP-conjugated anti-mouse IgG. Figure 2. Shedding kinetics of cell membrane integrated/associated 35S-labelled proteoglycans. Confluent endothelial cells prelabelled with inorganic [35S]sulfate were washed exhaustively and incubated in serum free medium at 37°C in the presence or absence of plasmin or thrombin. (A) At the specified time intervals the amount of 35S-labelled proteoglycans shed into the medium was determined. (B) At the specified time intervals the amount of 35S-labelled cell membrane-integrated/associated material obtained by trypsinization was determined. Figure 3. Accelerated shedding of 35S-labelled heparan sulfate proteoglycans. Confluent HUVEC prelabelled with inorganic [35S]sulfate were washed exhaustively and incubated in serum free medium at 37°C in the presence or absence of plasmin or thrombin. (A) After 6 h the 35S radioactivity incorporated into the total and HS-containing proteoglycans of the medium was determined. (B) After 6 h the 35S radioactivity incorporated into the total and HS-containing cell membrane- integrated/associated proteoglycans was determined. Figure 4. SDS-PAGE of syndecan-4 ectodomain proteolysates. (A) 10 µg syndecan-4 fusion protein was submitted to PAGE, transferred to PVDF membrane and stained by ponceau red (B). 60 µg syndecan-4 fusion protein was incubated with plasmin in an enzyme/substrate ratio of 1:5 in a total volume of 50 µL. The electrophoretically resolved lysate was transferred to PVDF membranes and the

8

by guest on February 13, 2018http://w

ww

.jbc.org/D

ownloaded from

ponceau red stained bands (I-VI) were excised for gas-phase protein sequencing. (C) Same conditions as in (B) but incubation with thrombin. Tab. 1. Position of protease cleavage sites within the syndecan-4 ectodomain. The maltose binding protein (N-terminal sequence underlined) and EYFP were linked to the syndecan-4 ectodomain via oligopeptide linker. The plasmin and thrombin cleavage sites are indicated by arrows. The sequences registered by the gas-phase protein sequencer are underlined, the oligopeptide linker are in small bold letters.

9

by guest on February 13, 2018http://w

ww

.jbc.org/D

ownloaded from

Figure 1 A

Figure 1B

Figure 1C

10

by guest on February 13, 2018http://w

ww

.jbc.org/D

ownloaded from

Figure 2 A

B

11

by guest on February 13, 2018http://w

ww

.jbc.org/D

ownloaded from

Figure 3 A

B

12

by guest on February 13, 2018http://w

ww

.jbc.org/D

ownloaded from

Figure 4

B

A

C

13

by guest on February 13, 2018http://w

ww

.jbc.org/D

ownloaded from

Tab. 1 MKTEEGKLVIWINGDKGYNGLAEVGKKFEKDTGIKVTVEHPDKLEEKFPQVAATGDGPDIIFWAHDRFGGYAQSGLLAEITPDKAFQDKLYPFTWDAVRYNGKLIAYPIAVEALSLIYNKDLLPNPPKTWEEIPALDKELKAKGKSALMFNLQEPYFTWPLIAADGGYAFKYENGKYDIKDVGVDNAGAKAGLTFLVDLIKNKHMNADTDYSIAEAAFNKGETAMTINGPWAWSNIDTSKVNYGVTVLPTFKGQPSKPFVGVLSAGINAASPNKELAKEFLENYLLTDEGLEAVNKDKPLGAVALKSYEEELAKDPRIAATMENAQKGEIMPNIPQMSAFWYAVRTAVINAASGRQTVDEALKDAQTNSSSNNNNNNNNNNLGIEGRISEFMAPARLFALLLLFVGGVAESIRETEVIDPQDLLEGRYFSGALPDDEDVVGPGQESDDFELSGSGDLDDLEDSMIGPEVVHPLVPLDNHIPERAGSGSQVPTEPKKLE ▼ PV and TV ▼ PVI ▼ TVI ENEVIPKRISPVEESEDVSNKVSMSSTVQGSNIFERTEWTRGSSRGAGMVSKGEELFTGVVPILVELDGDVNGHKFSVSGEGEGDATYGKLTLKFICTTGKLPVPWPTLVTTFGYGLQCFARYPDHMKQHDFFKSAMPEGYVQERTIFFKDDGNYKTRAEVKFEGDTLVNRIELKGIDFKEDGNILGHKLEYNYNSHNVYIMADKQKNGIKVNFKIRHNIEDGSVQLADHYQQNTPIGDGPVLLPDNHYLSYQSALSKDPNEKRDHMVLLEFVTAAGITLGMDELYK.

14

by guest on February 13, 2018http://w

ww

.jbc.org/D

ownloaded from

BuddeckeAnnette Schmidt, Frank Echtermeyer, Anthony Alozie, Kerstin Brands and Eckhart

bonds130 - Val129 and Lys115 - Arg114cleavage sites at LysPlasmin-and thrombin-accelerated shedding of syndecan-4 ectodomain generates

published online August 8, 2005J. Biol. Chem. 

  10.1074/jbc.M501903200Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

by guest on February 13, 2018http://w

ww

.jbc.org/D

ownloaded from