Syndecans , Heparan Sulfate Proteoglycans , Maintain the Proteolytic Balance of Acute Wound Fluids *

An imbalance between proteases and antiproteases is thought to play a role in the inflammatory injury that regulates wound healing. The activities of some proteases and antiproteases found in inflammatory fluids can be modified in vitro by heparin, a mast cell-derived glycosaminoglycan. Because syndecans, a family of cell surface heparan sulfate proteoglycans, are the major cellular source of heparin-like glycosaminoglycan, we asked whether syndecans modify protease activities in vivo. Syndecan-1 and syndecan-4 ectodomains are shed into acute human dermal wound fluids (Subramanian, S. V., Fitzgerald, M. L., and Bernfield, M. (1997) J. Biol. Chem. 272, 14713–14720). Moreover, purified syndecan-1 ectodomain binds cathepsin G (Kd 5 56 nM) and elastase (Kd 5 35 nM) tightly and reduces the affinity of these proteases for their physiological inhibitors. Purified syndecan-1 ectodomain protects cathepsin G from inhibition by a1-antichymotrypsin and squamous cell carcinoma antigen 2 and elastase from inhibition by a1-proteinase inhibitor by decreasing second order rate constants for protease-antiprotease associations (kass) by 3700-, 32-, and 60-fold, respectively. Both enzymatic degradation of heparan sulfate and immunodepletion of the syndecan-1 and -4 in wound fluid reduce these proteolytic activities in the fluid, indicating that the proteases in the wound environment are regulated by interactions with syndecan ectodomains. Thus, syndecans are shed into acute wound fluids, where they can modify the proteolytic balance of the fluid. This suggests a novel physiological role for these soluble heparan sulfate proteoglycans.

Multiple factors orchestrate the inflammatory response to tissue injury. These include proteases, antiproteases, cytokines, chemokines, and the growth factors derived from the plasma and cells associated with the injury, as well as from cells invading the injury site (1). Emigrating polymorphonuclear leukocytes release proteolytic enzymes into the injury site, including the most potent serine proteases, neutrophil elastase and cathepsin G (CatG). 1 These enzymes aid wound repair by digesting extracellular proteins, releasing growth factors from extracellular matrix, and remodeling the tissue (2)(3)(4). However, these enzymes can also destroy tissues when proteolysis is prolonged, inappropriate, or excessive (5,6).
Enormous local concentrations of proteases, estimated to be in the millimolar range for elastase and cathepsin G, are released into extracellular spaces during the leukocyte activation associated with tissue injury (7). Serine protease inhibitors (serpins), provide efficient control mechanisms to prevent undesirable extracellular protein degradation at the injury site (8). These antiproteases, mostly derived from plasma, share three principal properties: (i) they form 1:1 covalent complex with proteases, (ii) complex formation results in both inactivation of the protease and proteolytic cleavage of the serpin, and (iii) inhibition is essentially irreversible (9). The balance of proteases and antiproteases at the site of injury can regulate the extent of the inflammatory response during the repair process (10,11).
Dermal wound repair requires harmonious protease-antiprotease interactions or proteolytic balance. Excessive elastase action in the wound bed can account for endothelial damage (12), degradation of the epidermal/dermal junction (13), and the development of chronic skin ulcers (14). Physiological neutrophil elastase inhibitors include plasma-derived ␣ 2 -macroglobulin and, most importantly, ␣ 1 -proteinase inhibitor (also known as ␣ 1 -PI, ␣ 1 -antitrypsin, or ␣ 1 -AT). The importance of ␣ 1 -PI in regulating the response to tissue injury is emphasized by the extensive elastin and collagen fiber destruction leading to pulmonary emphysema in the lungs of individuals with congenital ␣ 1 -PI deficiency (5). The major physiological cathepsin G inhibitor is ␣ 1 -antichymotrypsin (␣ 1 -Achy), another plasma-derived serpin (15). Inherited ␣ 1 -Achy deficiency is pleiomorphic, but it is often associated with chronic active hepatitis and increased residual lung volumes (16). Another serpin that inhibits cathepsin G is the squamous cell carcinoma antigen 2 (SCCA2), a newly described product of skin and respiratory tract epithelia (17).
Although the activity of one class of serpins is accelerated by binding to heparin or other glycosaminoglycans (GAGs) (9,18), ␣ 1 -PI and ␣ 1 -Achy belong to the class of serpins that function independently of heparin and other GAGs. However, heparin can bind with high affinity to both neutrophil elastase and cathepsin G (19,20). This binding inhibits the enzymatic activities, but most importantly, it reduces the ability of the enzymes to interact with serpins (19,20). The heparin used clinically and in these studies is a pharmaceutical product derived from processing of the heparin proteoglycan within mast cells (21). The major physiological source of the heparin-like GAG, heparan sulfate, is found in proteoglycans within cells, at the cell surface and in the extracellular matrix (22).
Most cellular heparan sulfate derives from the syndecan family of cell surface proteoglycans. This family (currently known as syndecan 1-4 in mammals) consists of single transmembrane proteins containing conserved cytoplasmic and transmembrane domains and less well conserved extracellular domains (ectodomains), which bear variable numbers of GAG chains. All syndecans bear heparan sulfate, although syndecan-1 and -3 can also bear chondroitin sulfate. Syndecans bind many of the factors that orchestrate the inflammatory response to tissue injury as well as a variety of extracellular matrix components and adhesion molecules via their heparan sulfate chains and are individually expressed in distinct cell-, tissue-, and development-specific patterns (23).
Syndecan expression is highly regulated during wound repair. During cutaneous wound repair, keratinocytes migrating from the wound edge show loss of cell surface syndecan-1 (24). Concomitantly, syndecan-1 expression increases on the endothelial cells, and syndecan-4 expression increases on the dermal fibroblasts that form the granulation tissue (24,25), apparently due to inductive action of neutrophil-derived antimicrobial peptides (26). Syndecans on cell surfaces can be cleaved near the plasma membrane, which releases the now soluble intact proteoglycan ectodomains into the surrounding milieu (27). This shedding is accelerated by activation of protease (e.g. thrombin) and growth factor receptors (epidermal growth factor family members) and by the direct action of proteases (e.g. plasmin) involved in wound repair (27). Moreover, soluble syndecan-1 and -4 ectodomains are detected in acute dermal wound fluids (27). Although syndecan expression and shedding are highly regulated during the response to tissue injury, the role these processes play in this response is not clear.
A key aspect of the response to tissue injury is the establishment and maintenance of proteolytic balance at the wound site. The action of the major proteases, neutrophil elastase and cathepsin G, must be countered by their major inhibitors, ␣ 1 -PI and ␣ 1 -Achy, for normal wound repair to ensue. Loss of this balance can prevent normal repair, potentially leading to chronic wounds, which in the skin are difficult to treat satisfactorily (14). We postulated that because activities of the major proteases in acute wound fluids can be modified in vitro by heparin, soluble syndecan ectodomains could be involved in establishing and maintaining the proteolytic balance in wounds in vivo. We found syndecan-1 and -4 ectodomains in acute human dermal wound fluids. We also found that purified syndecan-1 ectodomain binds to both neutrophil elastase and cathepsin G, markedly reducing their affinity for serpins and thus protecting these enzymes from their physiological inhibitors. Moreover, both degradation of endogenous heparan sulfate and removal of syndecan-1 and -4 from wound fluids reduce proteolytic activities in the fluid. Thus, syndecan ectodomains maintain the proteolytic balance in acute wound fluids, a novel physiological role for soluble heparan sulfate proteoglycans.
For production of monoclonal antibodies specific to human syndecan-1, recombinant syndecan-1 was used for immunization of mice, and production of monoclonal antibody was by Maine Biotechnology Services, Inc. (Portland, ME). Mice were immunized and boosted with 100 g of recombinant syndecan-1. Out of 12 hybridoma clones, which produced antibodies reacting with human syndecan-1 fusion protein, only 1 (DL-101) reacted specifically with native human syndecan-1 ectodomain purified from conditioned media of human A431 cells. This was assessed as reactivity on Western blotting with a 300-kDa proteoglycan smear, which reduced to a 70-kDa core protein after treatment with heparitinase and chondroitinase ABC as described previously (30).
Affinity Co-electrophoresis (ACE) Analyses-NMuMG cells were labeled with radiosulfate, and [ 35 S]sulfate-labeled syndecan-1 ectodomain from conditioned medium was purified by DEAE and immunoaffinity chromatography (28). Binding of this ectodomain to elastase and cathepsin G was assessed by ACE as described elsewhere (31,32). Briefly, 1% (w/v) low melt agarose gels were cast containing distinct lanes with various concentrations of protease (indicated in Fig. 1). [ 35 S]Sulfate-labeled syndecan-1 ectodomain (12,500 cpm) was electrophoresed through these lanes. In competition assays, the same amount of syndecan-1 was mixed with 1 mg/ml chondroitin-6 sulfate or heparin prior to electrophoresis. The migration on syndecan-1 was detected on a PhosphoImager (Molecular Dynamics, Sunnyvale, CA). The pixel intensities were integrated and used to determine the migration distance of the major peak of 35 S-labeled syndecan-1 in each proteasecontaining lane. These mobilities were plotted as a function of ligand concentration and used to estimate the apparent K d values as described earlier (32).
Assays for Enzyme Inhibition-The amounts of proteases and serpins were calibrated by the method of Chase and Shaw (33). Trypsin was calibrated by using p-nitrophenyl-pЈ-guanidinobenzoate (Sigma), except that 100 mM Tris-HCl, pH 8.3, was used in place of sodium barbiturate buffer. The concentration of ␣ 1 -PI was standardized against calibrated trypsin. Elastase and cathepsin G were calibrated against the standardized ␣ 1 -PI. ␣ 1 -Achy was calibrated against the standardized cathepsin G. Reaction buffers were 50 mM Hepes, 150 mM NaCl, 5% N,N-dimethylformamide, pH 7.4, for cathepsin G, and 50 mM Tris, 150 mM NaCl, 0.1 mg/ml bovine serum albumin, pH 7.4, for elastase.
Enzyme inhibition was determined by mixing enzyme with increasing concentrations of syndecan-1 in the appropriate reaction buffer and incubating for 15 min at 25°C. The inhibitor was added, and residual enzyme activity was determined by adding the appropriate substrate and measuring hydrolysis at 405 nm with a UVmax microplate reader (Molecular Devices) or at 488 nm with a FluorImager 575 (Molecular Dynamics). The concentrations for cathepsin G assays were 34 nM cathepsin G, 34 nM ␣ 1 -Achy or SCCA2, and 3 mM Suc-AAPF-pNA. The concentrations for elastase assays were 34 nM elastase, 34 nM ␣ 1 -PI, and 5 M (CBZ-Ala-Ala-Ala-Ala) 2 -R110. In some experiments, syndecan-1 (up to 7.5 g as HS) was pretreated with heparinase (150 milliunits/ml), heparitinase (150 milliunits/ml each) or chondroitinase ABC (1.0 units/ ml) in 10 l of 50 mM Tris, 10 mM NaAc, pH 7.4, for 6 h at 37°C and boiled for 10 min.
Protease-Serpin Binding Stoichiometry-Constant concentrations of syndecan-1 ectodomain or heparin were preincubated with protease for 15 min at 25°C with increasing concentrations of serpin in the appropriate reaction buffer, and the residual enzyme activities were measured with an appropriate substrate. The concentrations for cathepsin G assays were 34 nM cathepsin G and 3 mM Suc-AAPF-pNA in cathepsin G reaction buffer. The concentrations for elastase assays were 34 nM elastase and 5 M (CBZ-Ala-Ala-Ala-Ala) 2 -R110 in elastase reaction buffer.
Determination of Rate Constant k ass for Enzyme-Inhibitor Association-The association rate constant for the interaction of serpins with free and syndecan-1-bound enzymes was determined under second order rate conditions (15). Equimolar amounts (34 nM) of enzyme with or without the syndecan-1 ectodomain and inhibitor were incubated at 25°C for varying periods of time. The reaction was quenched by the addition of substrate, for which the enzyme has higher affinity, and the release of pNA or rhodamine was measured. The residual enzyme activities were used to calculate the concentration of free enzyme E. Protease standard curves used for calculation of free enzyme were done in the presence and absence of syndecan-1 ectodomain. The rate of change in the amount of free enzyme over time is described as where the slope of the plot of reciprocal remaining free enzyme (1/E) over time (t) yields a second order rate constant (k ass ).
Collection of Acute Wound Fluids-Acute human dermal wound fluids were collected from reduction mammoplasty patients (samples were kindly provided by Dr. E. Eriksson, Brigham and Women's Hospital, Boston, MA). Wound fluids were collected at 1-day intervals from sterile closed-suction drains routinely placed in the subcutaneous space after mammoplasty. After collection, fluids were centrifuged for 15 min at 200 ϫ g and 4°C to remove cells and further for 15 min at 3300 ϫ g and 4°C to remove debris. The supernatants were stored at Ϫ70°C until use. Porcine wound fluid was produced, collected, and processed as described previously (26,34).
Informed consent was obtained for all procedures, and the use of anonymous discarded material was approved by the Human Research Committee (protocol 95-7509-01, Brigham and Women's Hospital).
Assays for Enzyme Activities in Wound Fluids-Day 1 human wound fluids were treated with 150 milliunits/ml heparinase (Hase) and heparitinase (HSase) for 3 h at 37°C in 50 mM Hepes, 150 mM NaCl, pH 7.4, to degrade endogenous heparan sulfate in the fluid. Five min after adding heparin (up to 5 g/ml) to untreated or Hase/HSase-treated fluids, the chymotryptic and elastolytic activities in the samples were detected by adding the appropriate substrate (3 mM Suc-AAPF-pNA for chymotryptic and 5 M (CBZ-Ala-Ala-Ala-Ala) 2 -R110 for elastolytic activities). Hydrolysis was measured over time at 405 nm with a UVmax plate reader (Molecular Devices) or at 488 nm with a FluorImager 575 (Molecular Dynamics).

Syndecan-1 Ectodomain Binds Elastase and Cathepsin
G-Because heparin can protect elastase and cathepsin G against inhibition by certain plasma-derived serpins (19,20), we speculated that the soluble syndecan ectodomains in wound fluid might act similarly. [ 35 S]Sulfate-labeled syndecan-1 ectodomain purified from the conditioned media of NMuMG cells was incubated with nitrocellulose filters containing dots of purified human neutrophil cathepsin G and elastase, and serum-derived ␣ 1 -PI and ␣ 1 -Achy. The syndecan-1 ectodomain bound to cathepsin G and elastase at picomolar levels of protease, whereas no binding to the antiproteases was detected at 10-fold higher concentrations (data not shown). ACE (32) of [ 35 S]sulfate-labeled syndecan-1 ectodomain with cathepsin G and elastase confirmed this binding and yielded apparent K d values of 56 nM for cathepsin G and 35 nM for elastase (Fig. 1,  A and C). Heparin (1 mg/ml) completely abolished binding to the proteases, whereas chondroitin sulfate (1 mg/ml) had little or no effect, indicating that the binding is mainly due to the heparan sulfate chains on syndecan-1 (Fig. 1, B and D). The ACE profiles with both enzymes showed heterogeneity in syndecan-1 ectodomain binding at concentrations near the K d values (80 nM), suggesting that there are subfractions of the ectodomain that differ in their avidity for the proteases (data not shown).
Binding of Syndecan-1 Ectodomain to the Protease Reduces the Effect of Antiprotease-To determine whether the binding of the syndecan-1 ectodomain to the proteases affects their rate of interaction with a serpin, rate constants (k ass ) for these interactions were measured in the presence and absence of soluble syndecan-1 ectodomain (Table I). The protease and serpin form a 1:1 complex. Because neither heparin or the syndecan-1 ectodomain alters this stoichiometry (Fig. 2), the k ass were determined under second order conditions (15). Equimolar amounts (34 nM) of protease and serpin were incubated in the presence or absence of the syndecan-1 ectodomain at concentrations indicated in Table I. After various times, complex formation was quenched by adding substrate, and the remaining free enzyme activity was measured as described under "Experimental Procedures." The k ass for the interaction was calculated from linear regressions (Equation 1). The k ass for cathepsin G with ␣ 1 -antichymotrypsin decreased over 3700fold and with SCCA2 over 32-fold in the presence of the syndecan-1 ectodomain (Table I). The k ass for elastase with ␣ 1proteinase inhibitor was decreased 60-fold by the syndecan-1 ectodomain (Table I). For comparison, second order rate constants were also measured for protease-serpin complex formation in the presence of a heparin concentration equivalent to that of the syndecan-1 ectodomain HS (Table I). At the concentrations tested, the soluble syndecan-1 ectodomain decreased the rates of protease-serpin association at least as effectively as authentic heparin. The association rate for cathepsin G and ␣ 1 -antichymotrypsin was reduced to a significantly greater extent in the presence of soluble syndecan-1 ectodomain than in the presence of equivalent heparin concentration (Table I).
Soluble Syndecan-1 Ectodomain Modifies Protease Activities via Interactions of its Heparan Sulfate Chains-The effect of the purified syndecan-1 ectodomain on protease activity was assessed in the presence and absence of serpin. The ectodomain was preincubated with cathepsin G or elastase for 15 min and assayed for protease activity with or without equimolar concentrations of serpins (Fig. 3). Syndecan-1 ectodomain alone reduced cathepsin G activity in a concentration-dependent manner, reaching maximal inhibition (35%) at 2 g/ml as core protein (Fig. 3A). However, the syndecan-1 ectodomain markedly decreased the ability of both ␣ 1 -Achy and SCCA2 to inhibit cathepsin G activity (Fig. 3A). In the absence of ectodomain, these serpins completely inhibit the protease, but with increasing concentrations of ectodomain, their inhibitory activity is reduced and ultimately abolished (Fig. 3A). The syndecan-1 ectodomain was more effective in reducing cathepsin G inhibi-tion by SCCA2 (ED 50 ϭ 0.2 g/ml; Fig. 3A) than by ␣ 1 -Achy (ED 50 ϭ 0.5 g/ml; Fig. 3A). Analogous findings were obtained with elastase (Fig. 3B). The ectodomain alone reduced elastase activity, reaching maximal inhibition of approximately 40%. But the ectodomain also decreased the ability of ␣ 1 -PI to inhibit the protease (ED 50 ϭ 1.0 g/ml; Fig. 3B). These results indicate that syndecan-1 ectodomain alone reduces the activities of cathepsin G and elastase to an extent nearly identical to that observed with heparin (19,20). Moreover, binding of the syn-   decan-1 ectodomain to the proteases protects them from inhibition by serpins (Fig. 3).
To assess which GAG chains on the syndecan-1 ectodomain are responsible for its ability to protect the proteases from the antiproteases, syndecan-1 was pretreated with Hase (degrades heparin-like regions of HS chains), HSase (degrades low sulfated regions of HS chains), both Hase and HSase, or chondroitinase ABC (degrades chondroitin sulfate chains). Protease was preincubated for 15 min at 25°C with or without ectodomain that was untreated or had been pretreated with each of these enzymes and then mixed with antiprotease, and the residual enzyme activity was measured by adding the appropriate substrate. Hase and HSase preparations had no detectable elastase or cathepsin G activity, nor did they directly affect the activity of these proteases in test reactions. Progress curves show that Hase treatment alone reduced by about 30% and treatment with both Hase and HSase reduced by about 60% the ability of syndecan-1 to protect cathepsin G against antiproteases ␣ 1 -Achy and SCCA2 (Fig. 4, A and B), whereas chondroitinase ABC digestion had no effect (not shown). However, digestion of syndecan-1 HS chains with either Hase or HSase alone abolished the ability of syndecan-1 to protect elastase from inhibition by ␣ 1 -PI (Fig. 4C). Again, chondroitinase ABC digestion had no or little effect (data not shown). Heparan sulfate chains isolated from syndecan-1 and chondroitin sulfate-free syndecan-1 ectodomain were equally effective in protecting the proteases from inhibition by antiproteases (data not shown). Thus, the soluble syndecan-1 ectodomain protects cathepsin G and elastase against antiproteases via its heparan sulfate chains.
We then asked whether these syndecan ectodomains could modify the balance between the endogenous proteases and their inhibitors in the wound fluid. To demonstrate this balance, we added heparin to wound fluids before and after de-grading the endogenous heparan sulfate in the fluid. Direct addition of as little as 2 g/ml heparin increased elastolytic activity, suggesting that heparin shifts the balance in favor of the protease (Fig. 6A). Degrading the endogenous heparan sulfate with the Hase/HSase mixture, as in Fig. 4, markedly reduced elastolytic activity (Fig. 6B). This enzymatic degradation similarly affected chymotryptic activity in pig wound fluids (data not shown). Adding heparin to the enzyme-treated sample shifted the balance in favor of the protease and nearly reversed the loss of activity (Fig. 6B). Thus, the activity of proteases and their inhibitors is balanced in wound fluid, and this balance can be shifted by the endogenous heparan sulfate in the fluid. To evaluate whether the syndecan ectodomains comprised the active heparan sulfate in the fluid, wound fluids were treated with monoclonal antibodies against the syndecan-1 and -4 ectodomains or, as a control, with equal concentrations of mouse IgG, and the elastolytic activity was measured. Removal of syndecan-1 and -4 ectodomains from each of 5 wound fluids reduced elastolytic activity 33-60% (Fig. 7). Adding heparin to immunodepleted samples shifted the balance in favor of the proteases and nearly reversed the loss of activity, identical to that observed in Fig. 6B (data not shown). These data indicate that soluble syndecan ectodomains maintain the proteolytic balance of wound fluids but do not complex with all the protease in the fluid. DISCUSSION In this study, we provide new insights into the regulation of protease-antiprotease balance during tissue injury. We show that syndecan-1 and -4, cell surface heparan sulfate proteogly- cans, are shed into wound fluids as soluble ectodomains. These acute dermal wound fluids contain the neutrophil-derived proteases cathepsin G and elastase, as well as the plasma-derived antiproteases ␣ 1 -PI and ␣ 1 -Achy. Both degradation of endogenous HS in the wound fluid and immunodepletion of syndecan-1 and -4 from the wound fluids alter the protease-antiprotease balance of the fluid. In vitro studies show that this balance results from binding of the HS chains on the ectodomains to cathepsin G and elastase. This interaction reduces the ability of the antiprotease to inhibit protease activity. These results indicate that syndecan-1 and -4 are shed into inflammatory fluids, where they modify the proteolytic balance of the fluids. The findings also suggest a novel physiological role for soluble heparan sulfate proteoglycans and new approaches to modulate the protease balance of inflammatory fluids.
Soluble Syndecan Ectodomains as Heparin-like Mediators at the Wound Site-Wound repair requires precise temporal and spatial regulation of a panoply of effectors, including chemo-kines, growth factors, extracellular components, cell adhesion proteins, proteases, and antiproteases. Many of these proteins bind heparin and heparan sulfate under physiological conditions and with high affinities (35). During repair of skin injury, cellular expression of syndecan-1 and -4 is altered (24,25), and cell surface syndecan-1 and -4 are converted to soluble molecules by juxtamembrane cleavage of their extracellular domains (ectodomains), a process known as shedding (27). Recent studies have shown that syndecan shedding is a highly regulated process that is stimulated by certain agents released at the site of tissue injury (27). Shedding instantly converts a cell surface proteoglycan into a soluble effector.
The functions of these soluble ectodomains are not clear. Syndecans on cell surfaces can act as co-receptors for heparinbinding growth factors; notably, the action of FGF-2 requires a growth factor-heparan sulfate proteoglycan-FGFR1 complex (36). However, because the soluble ectodomains retain all their HS, they can bind the same ligands as the cell surface syndecans, enabling them to be potential inhibitors of these ligand interactions. On the other hand, the soluble ectodomains place HS chains containing heparin-like domains into the wound environment. These chains can interact with heparin-binding proteins and peptides involved in the repair.
The inflammatory phase of tissue repair is characterized by plasma exudation and the involvement of neutrophils that produce and secrete the matrix remodeling enzymes elastase and cathepsin G. Although heparin binds and accelerates activity of some serpins (9), heparin does not interact with the serpins that regulate these enzymes. Rather, the enzymes themselves bind heparin, which reduces their affinity for the serpin and protects them from inhibition (19,20).
Our results indicate that the HS chains on the soluble syndecan ectodomains mimic this action of heparin and thus regulate the activity of the neutrophil-derived proteases in the wound environment (Fig. 4). Indeed, the syndecan-1 ectodomain HS chains are at least as effective in decreasing the protease-antiprotease interaction as an equal concentration of heparin (Table I). The binding affinities of the syndecan-1 ectodomain for the protease approximate that for heparin ( Fig.  1). As previously observed with heparin (19,20), the interaction between the syndecan-1 ectodomain and free protease inhibits the protease activity ( Fig. 3) but does not alter the stoichiometry of protease binding to the serpin (Fig. 2).
Proteolytic Balance in Wound Repair-Proteolysis is important for fibrinolysis, growth factor mobilization and activation, cell migration into the wound site, reepithelialization, angiogenesis, and extracellular matrix degradation (37). An imbalance of proteolytic activity disrupts normal wound repair and capillary morphogenesis (38,39). If the soluble ectodomains also act like heparin to accelerate the activity of heparin-activatable serpins (viz. antithrombin III, protease nexin I, plasminogen activator inhibitor-1, and others), the ectodomain could regulate several aspects of proteolysis during wound repair.
Proteases in wounds co-exist with their physiological inhibitors, and thus their activity is finely regulated to provide optimal activity for repair. This activity results from a balance, involving enzyme production and activation counterpoised by enzyme degradation and inhibition. The involvement of syndecan ectodomains in regulating proteolytic balance could explain several observations, including the variability of elastase activity and the inconsistency of fibronectin degradation in wound fluids (11). Our finding that both HS degradation and immunodepletion of syndecan ectodomains reduce the proteolytic activity of acute wound fluids (Figs. 6 and 7) indicates that FIG. 6. Elastolytic activity of wound fluid is balanced by endogenous heparan sulfate. A, endogenous elastolytic activity in human acute dermal wound fluids was assessed before (Ⅺ) and after the addition of heparin (q, 2 g/ml; f, 5 g/ml) by adding the the synthetic substrate (CBZ-Ala-Ala-Ala-Ala) 2 -R110 and measuring hydrolysis over time. B, human wound fluids were treated with 150 milliunits/ml of Hase and HSase for 3 h at 37°C in 50 mM Hepes, 150 mM NaCl, pH 7.4, or with buffer alone. Elastolytic activity was measured in untreated (Ⅺ) and Hase/HSase-treated fluids before (OE) and 5 min after (f) the addition of 5 g/ml heparin. The Hase/HSase mixture itself contained no detectable elastolytic activity, nor did it directly effect the activity of proteases in test reactions. Each of three acute (day 1) wound fluids tested showed reduced elastolytic activity after HS degradation, which was reversed after heparin addition.
FIG. 7. Depletion of syndecan-1 and -4 from human acute dermal wound fluids reduces their elastolytic activity. Human acute wound fluid was incubated with purified monoclonal antibodies against human syndecan-1 (MCA-681 and DL-101) and syndecan-4 (5G9 and 8C7) (q) or with equal concentrations of mouse IgG (Ⅺ). Following immunoprecipitations, the supernatants were tested for elastolytic activity by adding the synthetic substrate (CBZ-Ala-Ala-Ala-Ala) 2 -R110 and measuring the hydrolysis over time. Each of five acute (day 1) wound fluids tested showed reduced endogenous elastolytic activity after immunodepletion of syndecan-1 and -4 (33-60% lower rate). Activity in a representative experiment is shown. these soluble proteoglycans contribute to balanced proteolytic activity in the wound environment.
Alterations in proteolytic balance are thought to be one reason why acute wounds do not heal properly and become chronic (11,40). The high levels of proteolytic activity in chronic wound fluid have led to the proposal that misregulated proteases contribute to the inability of chronic wounds to heal even when treated with exogenous matrix and growth factors (41,42). Whether alterations in the levels of syndecan ectodomains could lead to loss of proteolytic balance and thus to development of chronic wounds needs investigation.
Abnormalities in Proteolytic Balance-Optimal proteolytic activity is needed for normal wound repair. Formation of the fibrin-rich provisional matrix produced after tissue injury is an initial step in the repair process. Once the fibrin clot has formed, migrating keratinocytes at the wound edge and emigrating neutrophils produce a variety of serine proteases and matrix metalloproteases to degrade this matrix and close the wound. An imbalance of proteases and serpins contributes to chronic inflammatory conditions, such as rheumatoid arthritis, pulmonary fibrosis, emphysema, and the development of vascular plaques of atherosclerosis and amyloid plaques in the central nervous system in Alzheimer's disease (9). Whether or not syndecans have a role in regulating proteolytic activities in these events is not known, but in light of our data, this possibility seems worth investigating.
We have found that syndecan-1 and -4 ectodomains act within human acute wound fluids to maintain proteolytic balance. Although no evidence so far exists, the ectodomains of other heparan sulfate proteoglycans that can be shed, such as glypican-1 and CD-44 (Refs. 43 and 29, respectively), might act similarly. Altered proteolytic balance in the wound environment has the potential to interfere with therapeutic procedures ranging from growth factor application to skin grafting. Thus, syndecan expression and shedding should be considered in evaluating and attempting to modify the response to tissue injury.