Cardiovascular Implant Research Laboratory, Department of Bioengineering, Clemson University, Clemson, South Carolina, USA

Glycosaminoglycan-degrading Enzymes in Porcine Aortic Heart Valves: Implications for Bioprosthetic Heart Valve Degeneration Dan T. Simionescu, Joshua J. Lovekamp and Narendra R. Vyavahare Cardiovascular Implant Research Laboratory, Department of Bioengineering, Clemson University, Clemson, South Carolina, USA Background and aim of the study: Glutaraldehyde (GA)-fixed aortic valves used in heart valve replacement surgery have limited durability due to tissue degeneration and calcification. Despite their structural and functional importance, very little is known about the fate of glycosaminoglycans (GAGs) within the extracellular matrix of bioprosthetic heart valves. The study aim was to investigate the stability of GAGs in GA-fixed tissues and to identify enzymatic mechanisms that may be responsible for GAG degeneration. Methods: Porcine aortic valve cusps were fixed with GA and implanted subdermally in rats for 21 days. Fresh, fixed and explanted cusps were analyzed for GAG content by hexosamine determination, and GAG-degrading enzyme activity was evaluated using zymography. GAG classes in fresh cusps were also assessed by flurorophore-assisted carbohydrate electrophoresis. Fresh and GA-fixed cusps were also exposed in vitro to hyaluronidase and chondroitinase in order to test the susceptibility of cusp GAGs towards enzymatic degradation. Results: Native aortic cusps contained ~3.5% GAGs by dry weight, consisting of hyaluronic acid, chondroitin sulfate and dermatan sulfate. Significantly lower GAG levels were found in aortic cusps after fixation with GA, and even lower levels were found after subdermal implantation in rats. GAG levels in GA-fixed cusps were also significantly reduced by in-vitro incubation with hyaluronidase and chondroitinase. Novel GAG-degrading enzymes were detected in considerable levels in native cusps, in lower levels in GA-fixed cusps and significantly increased levels after subdermal implantation of GA-fixed cusps. Conclusion: The combined action of active GAGdegrading enzymes and the failure of GA to stabilize GAGs towards enzymatic digestion may contribute significantly to bioprosthetic heart valve degeneration and subsequent structural failure. The Journal of Heart Valve Disease 2003;12:217-225 Cardiac valves are specialized forms of cardiovascular connective tissues that are subjected to strenuous and highly complex mechanical forces (1). The compressive forces and shear stresses that act on the valvular cusps are managed by the central layer of the cusp (spongiosa), which is comprised largely of glycosaminoglycans (GAGs) and loosely arranged collagen fibers. GAGs are acidic, unbranched polysaccharides such as chondroitin sulfate (CS) and dermatan sulfate (DS) that occur in vascular connective tissues bound to core proteins, and as large polymers that lack protein cores, such as hyaluronic acid (HA) (2). Loss of GAGs from within the cusp structure has been noted in Address for correspondence: Narendra R. Vyavahare PhD, Cardiovascular Implant Research Laboratory, Department of Bioengineering, 501 Rhodes Engineering Research Center, Clemson University, Clemson, South Carolina, 29634, USA valves obtained from patients suffering from congenital defects, rheumatic fever and old age, all of which are associated with valve failure (3). Information available regarding GAG metabolism in cardiac tissue indicates that among cardiac tissues tested, cardiac valves exhibit the highest levels of GAG biosynthesis (4). Normal GAG degradation occurs at a very high rate, as it was shown that the half-life of HA in skin is about 12 h (5) and that approximately onethird of the total body HA is replaced each day (6). The high turnover rate of GAGs in connective tissues has prompted the study of GAG-degrading enzymes, a necessary metabolic component. Mammalian hyaluronidases are enzymes capable of degrading glycosidic bonds in both HA and CS, and have been described in a variety of tissues and body fluids, including blood (6). They are also secreted by fibroblasts in culture (7), but their existence has not been reported in cardiac valves. Copyright by ICR Publishers 2003

218 Glycosaminoglycan degradation in valves Glutaraldehyde (GA)-fixed porcine aortic valves have been used extensively as heart valve replacements in the form of bioprosthetic heart valves (BHVs). A majority of BHVs fail clinically after about 10 years due to degeneration and pathologic calcification (1). The mechanisms that underlie these degenerative phenomena after implantation in human patients are beginning to unravel (8), but little is known about the stability of tissue GAGs. Loss of GAGs from GA-fixed porcine aortic cusps may occur as a result of both mechanical and biological events. More than 80% of GAGs were lost from GA-fixed BHVs after only 10 million cycles in an accelerated fatigue tester (the equivalent of six months of heart valve function) (9), suggesting that mechanical fatigue per se can induce GAG loss. BHVs implanted in a sheep circulatory model lost almost 90% of cusp GAGs after five months of implantation (9), and similar levels of GAG losses from implanted BHVs were reported in a clinical retrieval study (10). These data showed that GAG loss is a clinical reality, and that fatigue testing and animal implants can be used as models for GAG loss in BHVs. Since GAGs within the GA-fixed BHVs may participate in maintaining proper mechanical functions, as well as preventing calcium deposition, loss of GAGs may contribute significantly to BHV degeneration. In the present study, it was shown that the GAG content in aortic cusps decreases after GA fixation and declines further after subdermal implantation. Evidence was also provided to show that GAG depletion may be related to the activity of GAG-degrading enzymes and the lack of GAG stabilization by aldehyde fixation. Materials and methods Materials Glutaraldehyde was obtained from Polysciences, Inc. (Warrington, PA, USA). Electrophoresis chemicals and the MiniGel system apparatus were from BioRad (Hercules, CA, USA). Gel densitometry was performed using the LabImage Software version 2.62 from Kapelan Software GmbH, Germany. The bicinchoninic acid (BCA) protein assay kit was from Pierce Biotechnology Inc. (Rockfort, IL, USA),and the tissue Biohomogenizer from Biospec Products Inc. (Bartlesville, OK, USA). The atomic absorption spectrophotometer Model 3030 was from Perkin-Elmer (Norwalk, CT, USA). Ultrafree 15 Biomax 10K and Microcon YM 3 centrifugal concentrators were from Millipore Corporation (Bedford, MA, USA). Polyacrylamide monosaccharide 8 10 gels (MonoGels) and Tris-borate monosaccharide running buffer were from Glyko Inc. (Novato, CA, USA). Proteinase K, testicular hyaluronidase, chondroitinase ABC, hyaluronic acid, chondroitin sulfate, GAG standard disaccharides, J Heart Valve Dis aminoacridone and other chemicals were of highest purity available and obtained from Sigma (St. Louis, MO, USA). Porcine hearts were obtained from a local USDA-approved slaughterhouses, while experimental rats (Sprague-Dawley strain) were purchased from Harlan (Indianapolis, IN, USA). Preparation of fresh and GA-fixed porcine aortic valves Porcine hearts were rinsed in ice-cold saline and transported to the laboratory on ice. Whole aortic roots were separated and the aortic cusps dissected from each root for flurorophore-assisted carbohydrate electrophoresis (FACE) analysis, enzyme extraction and hexosamine assays. The time between tissue harvesting and enzyme extraction did not exceed 4 h. Whole aortic roots were also fixed at room temperature for seven days in a 0.6% GA solution prepared in 50 mm HEPES-buffered saline, ph 7.4, using 100 ml per root. The fixative was changed with fresh GA solution after the first 24 h of fixation. After fixation, the aortic cusps were dissected, rinsed in saline and processed for enzyme extraction, hexosamine assays and subdermal implantation, as described below. Analysis of GAG classes using FACE FACE was performed as described by Calabro and colleagues (11,12), with minor modifications. Briefly, 2-3 mg of fresh cusp samples were digested with Proteinase K followed by hyaluronidase and chondroitinase ABC to yield unsaturated GAG disaccharides. Samples were fluorotagged with aminoacridone and separated on polyacrylamide monosaccharide gels in parallel with GAG disaccharide standards. Fluorescent bands were photographed under ultraviolet light, images were graphically inverted and the lanes analyzed using densitometry. Rat subdermal implantation Ten samples of GA-fixed aortic cusps were implanted subdermally in five juvenile male, 26-day-old rats (mean body weight 89 ± 2 g) as described previously for calcification studies (13). After 21 days, the rats were sacrificed, and the cusps explanted and processed for calcium and phosphorus determination, quantitation of GAGs and enzyme extraction as described below. The animal protocol was approved by the Animal Research Committee at Clemson University. NIH guidelines for the Care and Use of Laboratory Animals (NIH publication 85-23, Revised 1985) were observed. Calcium and phosphorus determination The analysis of calcium and phosphorus in explants was performed as described previously (14). Briefly,

J Heart Valve Dis explanted cusps were hydrolyzed in 6 M HCl; subsequently, calcium content was determined using atomic absorption spectrophotometry, and phosphorus content using the molybdate complexation assay (15). GAG quantitation by hexosamine analysis GAG quantitation, using standard curves constructed with 0-100 µg/ml D-glucosamine, was achieved by following the procedure of Blix (16). Extraction of soluble enzymes from aortic cusps Fresh cusps, GA-fixed aortic cusps and cusps explanted with their surrounding capsules were rinsed separately in saline, minced by hand, homogenized in an extraction buffer comprising 0.1% Triton X-100, 50 mm Tris, 100 mm NaCl and 0.02% sodium azide, ph 7.4. After 24 h of extraction at 4 C on a magnetic stirrer, suspensions were centrifuged and the supernatants used as cusp soluble extracts. Extracts obtained from GA-fixed tissues were further concentrated on Ultrafree 15 Biomax 10K filter devices according to the manufacturer s instructions. Protein concentrations in soluble extracts were assayed with the bicinchoninic acid (BCA) assay, according to the manufacturer s protocol. Glycosaminoglycan degradation in valves 219 Detection of GAG-degrading enzymes in tissue extracts Substrate gel electrophoresis (zymography) with HA and CS embedded in polyacrylamide gels was performed as described by Miura et al. (17). Samples consisting of 20 µg of protein per lane were applied separately to 10% polyacrylamide gels containing 170 µg HA/ml for HA zymography, and to 10% polyacrylamide gels containing 100 µg/ml of derivatized CS (18) for CS zymography. Testicular hyaluronidase (50 mu) and chondroitinase ABC (25 mu) were also applied to each lane of the gels as positive controls. After electrophoresis, gels were washed in 2.5% Triton X-100 and incubated for 20 h at 37 C in developing buffer (50 mm citric acid, 50 mm dibasic sodium phosphate, 8 g/l sodium chloride, 0.02% sodium azide, ph 3). For experiments where the optimal ph was investigated, gels were also developed at ph 5 (same buffer as above, ph adjusted to 5) and separately at ph 7 (50 mm Tris, 8 g/l sodium chloride, 0.02% sodium azide, ph 7). After 20 h in development buffer, gels were stained in Alcian blue. GAG-degrading enzyme activities appeared as clear bands on a blue-turquoise semitransparent background of undigested, Alcian blue-stained GAGs. Images were taken with a digital camera, graphically inverted, and the lanes were analyzed using densitometry. Figure 1: Analysis of glycosaminoglycan (GAG) types in native aortic cusps using flurorophore-assisted carbohydrate electrophoresis (FACE). Two different cusp samples (A and B) were subjected to FACE analysis. The disaccharide standard mixture (S) consisted of unsaturated disaccharides specific for hyaluronic acid (HA), non-sulfated chondroitin and/or dermatan (0-S), 6-sulfated chondroitin sulfate (6-S) and 4-sulfated chondroitin and/or dermatan sulfate (4-S). FACE bands were analyzed using densitometry; the percentile distribution of the four GAG disaccharides is shown graphically.

220 Glycosaminoglycan degradation in valves J Heart Valve Dis Detection of gelatinases in tissue extracts Gelatinolytic activities were analyzed using zymography on 10% polyacrylamide gels containing 1 mg/ml gelatin, as described previously (19). Lanes were loaded with extracts containing 15 µg of protein/lane alongside 5 mu bacterial collagenase and prestained protein molecular weight standards (17 to 218 kda). Gelatinase activities appeared as clear bands on a blue background of undigested stained gelatin. Digital images of gels were graphically inverted and the lanes analyzed using densitometry. Resistance of cusp tissue to GAG-degrading enzymes Fresh cusps and separately GA-fixed cusps were lyophilized, and 5-15 mg dry samples were suspended in 1 ml of a solution containing 150 U/ml hyaluronidase and 0.5 U/ml chondroitinase ABC in 0.1 M ammonium acetate buffer, ph 7. Samples were incubated at 37 C on a shaker at 800 r.p.m. for 48 h. The enzyme solution was removed, after which the tissue samples were rinsed three times with the same buffer, lyophilized and weighed. The dry cusp samples were assayed for levels of GAGs before and after digestion, using the hexosamine assay described above. GAG levels were expressed as µg hexosamines per 10 mg dry cusp tissue. Data analysis Results were expressed as means of six samples (± SEM), with each sample run in triplicate. Statistical analyses of the data were performed using Student s t-test, and p-values for significance were calculated. A p-value <0.02 was considered statistically significant. Figure 2: Decrease in aortic cusp GAG content after glutaraldehyde (GA) fixation and after subdermal implantation. Cusp tissue samples were analyzed for GAG content using the hexosamine (Hex) assay. Figure 3: Detection and characterization of GAGdegrading enzymes in aortic cusps using hyaluronic acid (HA) zymography. Top panel: HA-degrading enzymes extracted from fresh cusps were tested for optimal activity at ph 3, 5, and 7. Middle panel: Extracts obtained from fresh cusps (F), cusps fixed for 7 days in GA (G) and GA-fixed cusps obtained after 21 days of subdermal implantation (I) were subjected to HA-zymography at ph 3. Pure testicular hyaluronidase (H) developed at ph 7 served as an internal control. Protein standards (S) are shown at left with molecular weights indicated in kda. Bottom panel: Densitometry and statistical analysis of the HA-ase levels in cusp extracts. RDU: Relative densitometric units.

J Heart Valve Dis Glycosaminoglycan degradation in valves 221 Results GAGs in native porcine aortic cusps Native porcine aortic cusps contained a mean of 172.92 ± 7.67 µg hexosamine per 10 mg dry tissue, yielding a total amount of GAGs in porcine aortic cusps as ~3.45% of the dry mass. These values were similar to those reported by others for bovine aortic cusps, i.e. 3.5% (20) and 3.93% (4), but were apparently higher than those described for human aortic cusps of 1.6% (21) and 2.8% (22). In order to develop specific methods for the study of GAG degradation in cusp tissues, the primary target was to establish the major GAG classes present in native aortic cusps using the newly developed FACE method (11). Native aortic Figure 4: Detection and characterization of cusp GAGdegrading enzymes by chondroitin sulfate (CS) zymography. Top panel: Extracts obtained from fresh cusps were tested for optimal activity at ph 3, 5, and 7. Middle panel: Extracts obtained from fresh cusps (F), cusps fixed for 7 days in GA (G) and GA fixed cusps obtained after 21 days of subdermal implantation (I) were subjected to CSzymography at ph 3. Pure bacterial Chondroitinase ABC (C) developed at ph 7 served as an internal control. Protein standards (S) are shown at left, with molecular weights indicated in kda. Bottom panel: Densitometry and statistical analysis of the CS-ase levels in cusp extracts. RDU: Relative densitometric units. Figure 5: Detection of gelatinases in aortic cusps. Top panel: Extracts obtained from fresh cusps (F), cusps fixed for 7 days in GA (G) and GA-fixed cusps obtained after 21 days of subdermal implantation (I) were subjected to gelatin zymography. Pure bacterial collagenase (C) served as an internal control. Protein standards (S) are shown at left, with molecular weights indicated in kda. Bottom panel: Densitometry and statistical analysis of the cusp gelatinase levels in tissue extracts. RDU: Relative densitometric units.

222 Glycosaminoglycan degradation in valves cusp GAGs were found to consist of disaccharides characteristic for: (i) hyaluronic acid (26% of total GAGs); (ii) non-sulfated chondroitin and/or dermatan (19%); (iii) chondroitin and/or dermatan 4-sulfate (26%); and (iv) chondroitin and/or dermatan 6-sulfate (29%) (Fig. 1). The proportion of HA, CS and DS in human and bovine aortic cusps (but not porcine cusps) has been reported previously. For example, Baig (23) reported a high percentile ratio of HA (63%) for human aortic valves, while bovine valves exhibited only a 41% HA ratio (24). It was apparent therefore that the GAG concentration and distribution of different GAG classes differ across species. Loss of GAGs from porcine aortic cusps Following fixation in GA for 7 days, a small (12%) but statistically significant (p <0.02) decrease in cusp GAG content was observed (Fig. 2), suggesting that GAGs are subjected to a continuous degradation process and that GA cannot fully prevent GAG degeneration during in-vitro preparation of tissues for use as BHVs. After 21 days of subdermal implantation of GAfixed cusps, an additional decrease (35%) in cusp GAG content was found, indicating that GAG degeneration in BHVs occurs after implantation and that GA fixation does not render GAGs resistant to in-vivo degradation. In parallel with GAG loss, the 21-day subdermal implants also accumulated significant amounts of calcium (65.81 ± 3.43 µg/mg dry tissue) and phosphorus (44.52 ± 1.05 µg/mg dry tissue). These data indicate that GAG diminution accompanies cusp calcification. J Heart Valve Dis GAG-degrading enzymes in porcine aortic cusps Hyaluronidase (HA-ase) activity in native cusps was detected as a doublet band migrating at ~46 kda and ~53 kda (Fig. 3). The cusp HA-ase was most active at ph 3, with very faint but detectable activities at ph 5 and 7, suggesting a possible lysosomal origin of the enzymes. Cusp HA-ase activity decreased significantly (p <0.02) after GA fixation (lane G, Fig. 3), but increased (p <0.02) after subdermal implantation (lane I, Fig. 3). Chondroitinase (CH-ase) activity in native cusps was detected as a major band migrating at ~53 kda (Fig. 4), pointing to the possibility that the cusp enzyme is capable of degrading both GAGs. The optimal ph for cusp CH-ase was also 3, with very faint activities at ph 5 and 7. Cusp CH-ase activity decreased (p <0.02) after GA fixation (lane G, Fig. 4) and increased significantly (p <0.02) after subdermal implantation (lane I, Fig. 4). These data indicate that GAG-degrading enzymes may contribute to GAG degradation during the early phases of tissue preparation for use as BHVs, as well as after BHV implantation. Gelatinases in porcine aortic cusps Fresh cusps exhibited a series of active gelatinolytic bands migrating at ~58-69 kda (lane F, Fig. 5), and Figure 6: Resistance of cusp GAGs to in-vitro degradation by GAG-degrading enzymes. Fresh, and separately GA-fixed cusp samples, were exposed for 48 h to a mixture of pure HA-ase and CH-ase and analyzed for GAG content before (not digested) and after digestion. GAG masses are expressed as µg hexosamines (Hex) per 10 mg dry cusp tissue. Figure 7: GAG structure and degeneration in valves. Large proteoglycans (PGs) are comprised of multiple chondroitin sulfate chains (CS) connected to a protein core (tall cylinders), which in turn are attached to a long hyaluronic acid chain (HA) creating a three-dimensional structure. Small PGs associate with the surface of the collagen fiber and have shorter core proteins and only one or two dermatan sulfate chains (DS). Matrix degeneration may be induced by hyaluronidases and chondroitinases (short arrowheads), enzymes capable of degrading HA, CS and DS chains, as well as by proteases (long arrows) capable of degrading PG core proteins.

J Heart Valve Dis Glycosaminoglycan degradation in valves 223 these activities were reduced (p <0.01) but still detectable after GA fixation (lane G, Fig. 5). After implantation for 21 days, gelatinase activities in GAfixed cusps increased significantly (p <0.01), revealing intense bands of 58-69 kda and additional bands migrating at 92 kda. These data suggest the presence of active gelatinases in cusp tissues. Susceptibility of cusp GAGs to in-vitro degradation by GAG-degrading enzymes Exposure of cusps to high concentrations of pure HA-ase and CH-ase (Fig. 6) led to a 70% loss of GAGs (p <0.001) from fresh tissues, and a 55% loss from GAfixed cusps (p <0.001). The difference in the GAG content of the fresh and GA-fixed cusps after digestion was small, but statistically significant (p <0.02), indicating that GA fixation renders cusp GAGs somewhat more resistant to the action of GAG-degrading enzymes. Discussion In the present study, it was shown that the GAG content in aortic cusps decreases after GA fixation and declines further after subdermal implantation. Evidence was also produced suggesting that GAG depletion might be due to the activity of GAG-degrading enzymes and may also be related to the lack of GAG stabilization as a result of aldehyde fixation. Mechanisms of GAG loss from porcine aortic cusps The present quantitative data for in-vitro GAG loss have been confirmed earlier by transmission electron microscopy studies conducted in the present authors laboratory, whereby a visible reduction in toluidine blue-stainable GAGs in aortic cusps after GA fixation was shown (25). In the present study, it was shown that enzymes capable of degrading GAGs are present in relatively large amounts in native cusps, indicating that these tissues may be endowed with an intense GAG metabolism. To the present authors knowledge, this is the first description of a HA-ase/CH-ase in porcine aortic cusp tissue. The nature and substrate specificity of these enzymes require further biochemical studies. These enzymes may be active during the initial BHV preparation steps (tissue harvesting, cleaning, dissection, mounting, which altogether could take up to 24 h) and may lead to pre-fixation GAG degradation. After 7 days of GA fixation, levels of cusp GAG-degrading activities were found to be significantly lower, but still detectable, suggesting that GAG degradation may persist (albeit at a lower rate) during the fixation and storage of BHVs. After implantation of BHVs, extrinsic sources of GAG-degrading enzymes could accelerate GAG degradation and contribute to BHV failure. Such sources may be infiltrating cells and serum-derived hyaluronidase (17). The present data also provide evidence for the activity of gelatinases in native, fixed and explanted cusps, suggesting that proteases may also participate in BHV degeneration, most likely by acting on the core protein of GAG-bearing proteoglycans (26). Figure 7 depicts diagrammatically the current understanding of GAG distribution in porcine cusps based on ultrastructural observations (25), together with the putative action of matrixdegrading enzymes. GAGs in GA-fixed cusps are vulnerable to enzymatic degradation Chemically cross-linked tissues are routinely subjected to a series of tests that ascertain the extent and efficacy of a particular cross-linker (27). In the present study it was shown that GA fixation endows the aortic cusp GAGs with little resistance towards GAGdegrading enzymes. These results may be explained in part by the fact that GAGs, and in particular HA, lack the amine functionalities required for proper crosslinking by GA (2). This test can be considered as an accelerated in-vitro GAG degradation system because within 48 h GAG deprivation in cusps was achieved corresponding to more than 21 days of subdermal implantation in rats. This test may prove useful for demonstrating the cross-linking efficacy of other chemical or physical agents that may react with functional groups in GAGs (such as carbodiimides, periodate-mediated fixation). GAG loss and calcification It was shown in the present study that calcification of GA-fixed cusps in the rat subdermal model was accompanied by a marked decrease in GAG content and a concomitant increase in GAG-degrading enzyme activities. Correlations between matrix degeneration and calcium deposition in BHVs have been long sought, and definitive proof has yet to be provided. Nevertheless, it is tempting to speculate that GAGdeprivation of cusp tissues facilitates the onset of calcium deposition and that the chronic removal of GAGs from the cusp spongiosa layer promotes further calcification. Others have shown that quantitative extraction of GAGs before GA fixation stimulated the calcification of GA-fixed pericardium in the rat subdermal model (28). Covalent binding of GAGs to valve tissue was shown to inhibit calcification (29). Finally, recent data have shown that periodate-mediated preservation of valve GAGs before fixation with GA reduced the valve s propensity towards calcification (30). Thus, it may be hypothesized that GAGs might act as inhibitors of cusp calcification, though validation of this concept requires further study.

224 Glycosaminoglycan degradation in valves Implications for long-term durability of BHVs The GAG-rich middle layer of the aortic valve cusp (spongiosa) is essential for correct mechanical functioning of valves during flexion, and it contributes greatly to the shear behavior of valve tissue, acting to lubricate the interface between the two outer valve layers (fibrosa and ventricularis) (31). GAGs may also play an inhibitory role in tissue calcification by chelating calcium ions and by their spatial involvement in the extracellular matrix and interference with hydroxyapatite nucleation (32). GAG deprivation may therefore contribute significantly to BHV degeneration affecting both the mechanical properties and the propensity of BHVs towards calcification. Study limitations and future perspectives Although hexosamine assays have been described as a reliable method for measuring amounts of GAGs in tissues (4), it is well known that some glycoproteins present in cusp tissues may also contain a small amount of hexosamines and hence any values obtained may slightly overestimate the actual amounts of GAGs. The optimal ph for the GAG-degrading enzymes extracted from cusps was found to be acidic (approximately 3), and only very low levels of enzymatic activities could be detected at physiological ph. A similar optimal ph was described for other GAGdegrading enzymes in mammalian tissues (6), though as yet it is not known what the role of these enzymes might be in vivo, and whether cells can create acidic microenvironments that would allow GAG-degrading enzymes to perform extracellularly. Further studies are required to establish the validity of the rat subdermal model as a relevant GAG degradation model, and additional experiments are required to understand the kinetics of GAG degradation in implants and in vitro. To the present authors knowledge, this is the first description of GAG-degrading enzymes in native aortic cusp tissues; consequently, their involvement in heart valve disease warrants further study. In conclusion, two aspects pertaining to BHV degeneration have been identified in the present study. First, active enzymes capable of degrading HA and CS are present in native porcine aortic cusps, and their activity persists after GA fixation as well as after implantation. Second, GAGs in GA-fixed aortic cusps are not efficiently stabilized by aldehyde fixation. The combined effects of degrading enzymes and the failure of GA to stabilize GAGs may contribute significantly to GAG depletion in cusp tissues and subsequently to BHV degeneration. J Heart Valve Dis Acknowledgements These studies were supported by NIH grant (HL 61652) and a Scientist Development Grant from AHA (to N.R.V.). No benefit of any kind will be received either directly or indirectly by the authors. References 1. Schoen FJ, Levy RJ. Founder s Award, 25th Annual Meeting of the Society for Biomaterials, perspectives. Providence, RI, April 28-May 2, 1999. Tissue heart valves: Current challenges and future research perspectives. J Biomed Mater Res 1999;47:439-465 2. Hascall VC, Hascall G. Proteoglycans. In: E. H (ed.). Cell Biology of Extracellular Matrix. Plenum Publishing Corporation, 1981:39-63. 3. Baig MM, Daicoff GR, Ayoub EM. Comparative studies of the acid mucopolysaccharide composition of rheumatic and normal heart valves in man. Circ Res 1978;42:271-275 4. Kanke Y, Mori Y, Bashey RI, Angrist AA. Biochemical study of cardiac valvular tissue. Biosynthesis in vitro of hexosamine-containing substances in bovine heart valve. Biochem J 1971;124:207-214 5. Laurent TC, Fraser JR. Hyaluronan. FASEB J 1992;6:2397-2404 6. Menzel EJ, Farr C. Hyaluronidase and its substrate hyaluronan: Biochemistry, biological activities and therapeutic uses. Cancer Lett 1998;131:3-11 7. Stair-Nawy S, Csoka AB, Stern R. Hyaluronidase expression in human skin fibroblasts. Biochem Biophys Res Commun 1999;266:268-273 8. Schoen FJ. Future directions in tissue heart valves: Impact of recent insights from biology and pathology. J Heart Valve Dis 1999;8:350-358 9. Vyavahare N, Ogle M, Schoen FJ, Zand R, Gloeckner DC, Sacks M, Levy RJ. Mechanisms of bioprosthetic heart valve failure: Fatigue causes collagen denaturation and glycosaminoglycan loss. J Biomed Mater Res 1999;46:44-50 10. Mako WJ, Calabro A, Ratliff NB, Vesely I. Loss of glycosaminoglycans (GAGs) from implanted bioprosthetic heart valves. Circulation 1997;I-155:863 11. Calabro A, Benavides M, Tammi M, Hascall VC, Midura RJ. Microanalysis of enzyme digests of hyaluronan and chondroitin/dermatan sulfate by fluorophore-assisted carbohydrate electrophoresis (FACE). Glycobiology 2000;10:273-281 12. Grande-Allen KJ, Griffin BP, Calabro A, Ratliff NB, Cosgrove DM, III, Vesely I. Myxomatous mitral valve chordae. II: Selective elevation of glycosaminoglycan content. J Heart Valve Dis 2001;10:325-332; discussion 332-333 13. Vyavahare NR, Hirsch D, Lerner E, Baskin JZ, Zand

J Heart Valve Dis Glycosaminoglycan degradation in valves 225 R, Schoen FJ, Levy RJ. Prevention of calcification of glutaraldehyde-crosslinked porcine aortic cusps by ethanol preincubation: Mechanistic studies of protein structure and water-biomaterial relationships. J Biomed Mater Res 1998;40:577-585 14. Vyavahare NR, Jones PL, Hirsch D, Schoen FJ, Levy RJ. Prevention of glutaraldehyde-fixed bioprosthetic heart valve calcification by alcohol pretreatment: Further mechanistic studies. J Heart Valve Dis 2000;9:561-566 15. Chen P, Toribara T, Warner H. Microdetermination of phosphorus. Anal Chem 1956;28:1756 16. Blix G. The determination of hexosamines according to Elson and Morgan. Acta Chem Scand 1948;2:467-473 17. Miura RO, Yamagata S, Miura Y, Harada T, Yamagata T. Analysis of glycosaminoglycandegrading enzymes by substrate gel electrophoresis (zymography). Anal Biochem 1995;225:333-340 18. Raja RH, LeBoeuf RD, Stone GW, Weigel PH. Preparation of alkylamine and 125I-radiolabeled derivatives of hyaluronic acid uniquely modified at the reducing end. Anal Biochem 1984;139:168-177 19. Vyavahare N, Jones PL, Tallapragada S, Levy RJ. Inhibition of matrix metalloproteinase activity attenuates tenascin-c production and calcification of implanted purified elastin in rats. Am J Pathol 2000;157:885-893 20. Moretti A, Whitehouse MW. Changes in the mucopolysaccharide composition of bovine heart valves with age. Biochem J 1963;87:396-402 21. Murata K. Acidic glycosaminoglycans in human heart valves. J Mol Cell Cardiol 1981;13:281-292 22. Torii S, Bashey RI, Nakao K. Acid mucopolysaccharide composition of human-heart valve. Biochim Biophys Acta 1965;101:285-291 23. Baig MM. Acid mucopolysaccharides of congenitally defective, rheumatic, and normal human aortic valves. Am J Pathol 1979;96:771-780 24. Mori Y, Honda A. Glycosaminoglycans and proteoglycans of cardiac tissue: With special reference to cardiac valves. In: R.S. V, editor. Glycosaminoglycans and proteoglycans in physiological and pathological processes of body systems. S. Karger, Basel, Switzerland, 1982:187-198. 25. Simionescu D, Lovekamp J, Vyavahare N. Degeneration of bioprosthetic heart valve cusp and wall tissues is initiated during tissue preparation: An ultrastructural study. J Heart Valve Dis 2003;12:225-233. 26. Simionescu A, Simionescu D, Deac R. Biochemical pathways of tissue degeneration in bioprosthetic cardiac valves. The role of matrix metalloproteinases. Am Soc Artif Intern Organs J 1996;42:M561- M567 27. Nimni ME, Cheung D, Strates B, Kodama M, Sheikh K. Chemically modified collagen: A natural biomaterial for tissue replacement. J Biomed Mater Res 1987;21:741-771 28. Jorge-Herrero E, Fernandez P, Gutierrez M, Castillo-Olivares JL. Study of the calcification of bovine pericardium: Analysis of the implication of lipids and proteoglycans. Biomaterials 1991;12:683-689 29. Strates B, Lian JB, Nimni ME. Calcification in cardiovascular tissues and bioprostheses. Nimni ME (ed.). Collagen 1989; volume III. Biotechnology. CRC Press, Boca Raton FL, 1989:274-292 30. Lovekamp J, Vyavahare N. Periodate-mediated glycosaminoglycan stabilization in bioprosthetic heart valves. J Biomed Mater Res 2001;56:478-486 31. Talman EA, Boughner DR. Glutaraldehyde fixation alters the internal shear properties of porcine aortic heart valve tissue. Ann Thorac Surg 1995;60(2 suppl.):s369-s373 32. Scott JE, Haigh M. Proteoglycan-type I collagen fibril interactions in bone and non-calcifying connective tissues. Biosci Rep 1985;5:71-81