Dynamic Mechanical Compression Influences Nitric Oxide Production by Articular Chondrocytes Seeded in Agarose

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS ARTICLE NO.

251, 580 –585 (1998)

RC989520

Dynamic Mechanical Compression Influences Nitric Oxide Production by Articular Chondrocytes Seeded in Agarose David A. Lee,* Stephen P. Frean,† Peter Lees,† and Dan L. Bader‡ *IRC in Biomedical Materials, Institute of Orthopaedics, University College London Medical School, Brockley Hill, Stanmore, Middlesex HA7 4LP, United Kingdom; †Department of Veterinary Basic Sciences, Royal Veterinary College, Hawkeshead Campus, North Mymms, Hatfield, Hertfordshire AL9 7TA, United Kingdom; and ‡IRC in Biomedical Materials, Queen Mary and Westfield College, University of London, Mile End Road, London E1 4NS, United Kingdom

Received September 14, 1998

Nitric oxide (NO) has been implicated in the inhibition of cell proliferation in cytokine and lipopolysaccharide (LPS)-stimulated chondrocytes and is known to be influenced by physical forces in several tissues. In this study, a well-characterized model system utilizing bovine chondrocytes embedded in 3% agarose constructs has been used to investigate the effect of dynamic strain at 0.3, 1, or 3 Hz on NO production. LPS induced a significant increase in nitrite levels, which was reversed by both L-NAME and dexamethasone. Dynamic compressive strain produced a significant reduction in nitrite production. The effect was partially blocked by L-NAME but unaffected by dexamethasone. L-NAME also reversed dynamic compressioninduced stimulation of [3H]-thymidine incorporation. NO appears to be a constituent of mechanotransduction pathways which influence proliferation of bovine chondrocytes seeded within agarose constructs. The inhibitor experiments also infer that alterations in cNOS activity primarily determine the response. © 1998 Academic Press

Articular cartilage is subjected to repetitive mechanical loading during normal activity and it is well established that the chondrocytes within the tissue can detect and respond to this mechanical environment by alterations in metabolic activity (1– 8). This process is termed mechanotransduction and may be resolved into extracellular factors such as cell and matrix deformation, hydrostatic pressure, fluid flow, and streaming potentials and intracellular events involving a variety of secondary signaling pathways (9, 10). While the extracellular components of chondrocyte mechanotransduction have been extensively studied, using in vitro model systems capable of varying factors independently, the intracellular pathways and their specific 0006-291X/98 $25.00 Copyright © 1998 by Academic Press All rights of reproduction in any form reserved.

roles in induction of effector cell response is poorly understood. Nitric oxide (NO) is a gaseous free radical which acts as both an inter- and intra-cellular messenger molecule in many cell types (11). It is known to be a potent regulator of chondrocyte activity in response to lipopolysaccharide (LPS) and interleukin-1 (Il-1) stimulation (12–18). NO has been reported to inhibit proteoglycan synthesis (13) and cell proliferation (14) in chondrocytes, both being components of the effector cell response when chondrocytes are subjected to mechanical compression. NO may also influence matrix metalloproteinase activity (19), apoptosis (20) and cytoskeletal organization (21) in chondrocytes. It is produced by the conversion of L-arginine to L-citrulline and NO through a process catalyzed by nitric oxide synthase (NOS) and various co-factors (22). Three distinct isoforms of NOS have been identified. Two isoforms, collectively known as cNOS, are constitutively expressed by cells and their activity is regulated by calcium/calmodulin binding. The third isoform (iNOS) is induced in response to LPS and Il-1 stimulation and its activity is independent of calcium levels. The role of various NOS isoforms can be elucidated by the use of specific inhibitors. The activity of all NOS isoforms can be inhibited by structural analogs of L-arginine, although some antagonists such as NG-nitro-L-arginine methyl ester HCl (L-NAME) show isoform selectivity towards cNOS (23). Induction of iNOS can be inhibited by glucocorticoids such as dexamethasone (24). Several studies have demonstrated that physical forces can influence NO production in a number of cell types. Endothelial cells upregulate NO production in response to either fluid flow or mechanical stretch through an iNOS mediated pathway (25, 26). Similar findings have been reported for osteoblasts (27, 28). A recent study suggested that isolated articular chondrocytes in monolayer culture also upregulate NO production in response to fluid flow-induced shear stresses

580

Vol. 251, No. 2, 1998

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

(29). There are no reports in the literature, however, investigating the influence of mechanical compression on NO production by chondrocytes. In previous studies by the authors and other groups a well characterized and reproducible model system has been used, involving chondrocytes isolated from articular cartilage and embedded in agarose gel which prevents loss of chondrocytic phenotype (30 –33). Investigations of the effects of dynamic compressive strain on chondrocyte metabolism within the model system revealed that proteoglycan synthesis and cell proliferation were affected in a distinct and frequency dependent manner (33). The objectives of the present study are to determine whether dynamic compressive strain influences NO production, suggesting a role as a mechanotransduction mediator, and to investigate which NOS isoforms are involved. MATERIALS AND METHODS Preparation of chondrocyte/agarose constructs. Full depth slices of cartilage were removed from the metacarpalphalangeal joints of 18-month-old steers. The cartilage slices were diced finely, and incubated at 37°C for 1 h in Dulbecco’s minimal essential medium supplemented with 20% (v/v) fetal calf serum (DMEM 1 20% FCS, Gibco, Paisley, UK) 1 700 unit mL21 pronase (BDH Ltd., Poole, UK) and for 16 h at 37°C in DMEM 1 20% FCS 1 100 unit mL21 collagenase type 1a (Sigma, Poole, UK). The supernatant containing released chondrocytes was passed through a 70mm pore size sieve (Falcon, Oxford, UK), washed twice in DMEM 1 20% FCS and resuspended at 8 3 106 cells mL21. The chondrocyte suspension was added to an equal volume of 6% agarose (type VII, Sigma, Poole, UK) in EBSS to give a final concentration of 4 3 106 cells mL21 in 3% agarose. The agarose/chondrocyte suspension was plated in a Perspex mold and allowed to gel at 4°C for 20 min. Cylindrical constructs (5 mm in diameter and 5 mm in height) were cut from the gel and subsequently cultured in DMEM 1 20% FCS at 37°C/5% CO2. Activation and inhibition of NOS. Constructs were prepared as above and incubated, unstrained, for 48 h in DMEM 1 20% FCS supplemented with 10 mg mL21 LPS (E. coli serotype 055:B5, Sigma, Poole, UK), 1 mM NG-nitro-L-arginine methyl ester HCl (L-NAME, Sigma, Poole, UK) or 10 mM dexamethasone (Sigma, Poole, UK). At the end of the culture period, the culture medium was removed and frozen for subsequent analysis. The agarose/cell constructs were digested with 2.8 unit mL21 papain and 10 unit mL21 agarase (both Sigma, Poole, UK) as described previously (33). Cell-free constructs were used as controls. Application of mechanical compression. The cell strain apparatus (Dartec Ltd., Stourbridge, UK) detailed in a previous study (33) was employed in this study to apply unconfined uniaxial compressive strain to constructs using fluid impermeable platens. Constructs were located in the center of each well of a 24 well tissue culture plate (Coster, High Wycombe, UK) and the plate was mounted in the apparatus. One milliliter of DMEM 1 20% FCS 1 10 mCi mL21 35 SO4 (Amersham International, Amersham, UK) 1 1 mCi mL21 tritiated thymidine ([3H]-TdR, Amersham International, Amersham, UK) was introduced to each well. The constructs were subjected to a 15% compressive strain at separate frequencies of 0.3, 1, and 3 Hz at 37°C/5% CO2 for 48 h. Control constructs were unstrained but maintained within the strain apparatus. In a further experiment constructs were subjected to a 15% compressive strain at 1 Hz for 48 h in DMEM 1 20% FCS 1 10 mCi mL21 35 SO4 1 1 mCi mL21 [3H]-TdR supplemented with 1 mM L-NAME, 10

mM dexamethasone or without supplements. Control constructs were unstrained but maintained within the strain apparatus. Biochemical analysis. Total nitrite was measured in the culture medium using the Griess assay (34). Nitrite levels in the agarase/ papain digests were also assessed in initial experiments. A volume of 100 mL of sample was added in a 1:1 ratio to a solution containing 0.5% (w/v) sulfanilamide (Sigma, Poole, UK) and 0.05% (w/v) naphthylethylenediamine dihydrochloride (Sigma, Poole, UK) in 5% (v/v) phosphoric acid (BDH, Poole, UK). Absorbance at 550 nm was measured immediately on a microplate reader. Absolute concentrations of nitrite in culture medium and agarase/papain digests were determined by comparison with standard solutions of sodium nitrite (Sigma, Poole, UK) freshly prepared in culture medium. [3H]-TdR incorporation was measured in the agarase/papain digest using the Multiscreen TCA precipitation method described previously (33). Incorporation of 35SO4 into newly synthesized proteoglycans was determined in both medium and agarase/papain digests using the Alcian blue precipitation method (35). Seventy-fivemicroliter aliquots of 50 mM sodium acetate, 0.5% (v/v) Triton X-100, pH 5.8, 50 mL of Alcian blue solution consisting of 0.2% (w/v) Alcian blue 8GX (BDH, Poole, UK) in 50 mM sodium acetate 1 85 mM magnesium chloride and 25 mL aliquots of culture medium or agarase/papain digest were added to individual wells of a multiscreen plate (0.45 mm pore filter, Millipore, Watford, UK). The plate was agitated for 1 h at room temperature, vacuum aspirated the wells were washed three times with 50 mM sodium acetate 1 100 mM sodium sulfate 1 50 mM magnesium chloride, pH 5.8. The filters were punched out into scintillation vials, agitated for 1 h in 0.5 mL of 4 M guanidine HCl in 4.3 M propan-2-ol and counted in 4 mL scintillation fluid (Emulsifier Safe, Packard, Pangbourne, UK) using a Tricarb 4000 series counter (Packard, Pangbourne, UK). Tenmicroliter aliquots of culture medium were added to scintillation vials, mixed with 4 mL scintillation fluid and counted to determine total 35SO4. Total DNA, determined using the Hoescht 33258 method (36), was used as a baseline for nitrite production, sulfate incorporation and [3H]-TdR incorporation. Statistical analysis. Unpaired student’s t-tests and one way ANOVA were used for statistical analysis. In all cases p # 0.05 was considered significant.

RESULTS Constructs incubated for 48 h in DMEM 1 20% FCS produced measurable quantities of nitrite. The concentration was 27.7 6 1.5 mM in the culture medium and 27.8 6 2.6 mM in the construct. These values suggest a basal rate of nitrite production of 0.198 6 0.018 nmol nitrite mg21 DNA h21 assuming a total system volume of 1.1 mL (1 mL culture medium 1 0.1 mL construct volume). In further experiments nitrite was assessed in the medium and adjusted to determine total nitrite in the system. No nitrite could be detected in the culture medium or agarase/papain digests prepared from cell free constructs cultured for 48 h. Figure 1 represents the rate of nitrite production in chondrocytes incubated in medium supplemented with 1 mM L-NAME, 10 mM dexamethasone or 10 mg mL21 LPS. L-NAME at 1 mM reduced basal nitrite production by approximately 84%. Dexamethasone at 10 mM, by contrast, reduced basal levels by only 10% and the reduction was not statistically significant. The addition of 10 mg mL21 LPS stimulated nitrite production rates

581

Vol. 251, No. 2, 1998

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

FIG. 1. Nitrite production by chondrocytes seeded in agarose constructs and cultured in DMEM 1 20% FCS (control) or DMEM 1 20% FCS supplemented with 1 mM L-NAME, 10 mM dexamethasone (Dex), 10 mg mL21 LPS or a combination of 10 mg mL21 LPS and either 1 mM L-NAME or 10 mM dexamethasone. Each value represents the mean and standard error for six replicates. Unpaired t-test results indicate differences between supplemented and control values as follows: *p , 0.05; and between LPS alone and LPS 1 L-NAME or dexamethasone as follows: #p , 0.05.

by 65% and the stimulation was reversed by either L-NAME or dexamethasone. Dynamic compressive strain at all three frequencies investigated (0.3 Hz, 1 Hz and 3 Hz) significantly reduced nitrite production compared to unstrained controls (Fig. 2). There was a statistically significant correlation between nitrite production and [3H]-TdR FIG. 3. Correlation between nitrite production and [3H]thymidine incorporation (A) and sulfate incorporation (B) by chondrocytes seeded in agarose constructs and subjected to 15% gross compressive strain at various frequencies for 48 h. The values have been normalized to unstrained control levels (100%). Each value represents the mean and standard error of at least 16 replicates from at least two separate experiments. Correlation coefficients r 5 0.97, p , 0.01 for [3H]-thymidine incorporation and r 5 0.22, p . 0.05 for sulfate incorporation.

FIG. 2. Nitrite production by chondrocytes seeded in agarose constructs and subjected to 15% gross compressive strain at various frequencies for 48 h. The values have been normalized to unstrained control levels (100%). Each value represents the mean and standard error of at least 16 replicates from at least two separate experiments. Unpaired student’s t-test results indicate differences from control values as follows: *p , 0.05.

incorporation (Fig. 3A, r 5 0.97, p , 0.05) but no correlation between nitrite production and sulfate incorporation (Fig. 3B, r 5 0.22, p . 0.05). The rates of nitrite production, [3H]-TdR incorporation and sulfate incorporation by chondrocytes in unstrained constructs maintained within the strain apparatus and supplemented with L-NAME or dexamethasone are presented in Table 1. Nitrite production by chondrocytes within unstrained constructs exhibited a similar pattern to that determined for constructs maintained in a conventional incubator (Fig. 1), except that the reduction in nitrite production in the presence of dexamethasone, although numerically small, was statistically significant. [3H]-TdR incorporation by

582

Vol. 251, No. 2, 1998

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS TABLE 1 3

Nitrite Production, [ H]-Thymidine Incorporation, and Sulphate Incorporation by Chondrocytes Seeded in Agarose Constructs and Subjected to 15% Dynamic Compressive Strain at 1 Hz for 48 Hours Unstrained

Nitrite production (pmole.mg21 DNA.hr21) [3H]-TdR incorporation (cpm.mg21 DNA.hr21) Sulphate incorporation (pmole.mg21 DNA.hr21)

Dynamic compressive strain (1 Hz)

No inhibitors

L-NAME

Dexamethasone

No inhibitors

L-NAME

Dexamethasone

142.0 6 8.1

21.6* 6 1.9

115.5* 6 12.2

109.0# 6 8.67

18.7*,# 6 2.4

83.8*,# 6 22.7

271.2 6 18.3

447.5* 6 23.8

265.4 6 20.0

387.8# 6 43.6

329.3*,# 6 17.6

350.9*,# 6 20.1

93.0 6 2.4

71.1* 6 6.0

76.8* 6 3.4

123.8# 6 4.5

98.0*,# 6 7.4

105.9*,# 6 3.2

Note. Culture medium was supplemented with 1 mM L-NAME, 10 mM dexamethasone or unsupplemented. Each value represents the mean 6 standard error of the mean for four replicates. Unpaired t-test results indicate differences between supplemented and unsupplemented values as follows: *p , 0.05; and between strained and unstrained values as follows: #p , 0.05.

chondrocytes within unstrained constructs was stimulated by L-NAME compared to unsupplemented cultures but unaffected by dexamethasone. Both L-NAME and dexamethasone induced a significant reduction in the rate of sulfate incorporation by chondrocytes within unstrained constructs compared to unsupplemented cultures. Absolute rates of nitrite production, [3H]-TdR incorporation and sulfate incorporation by chondrocytes subjected to 1 Hz dynamic compressive strain and supplemented with L-NAME or dexamethasone are presented in Table 1. Figure 4 indicates the percentage change in nitrite production (Fig. 4A), [3H]TdR incorporation (Fig. 4B) and sulfate incorporation (Fig. 4C) in constructs subjected to 1 Hz dynamic compressive strain compared to the unstrained controls. Compression-induced inhibition of nitrite production was significantly reduced by L-NAME but unaffected by dexamethasone. Compression induced stimulation of [3H]-TdR incorporation was reversed by L-NAME and significantly reduced by dexamethasone, while compression-induced stimulation of sulfate incorporation was unaffected by both NOS inhibitors. DISCUSSION Articular chondrocytes are known to produce NO constitutively or at an enhanced level in response to a variety of stimuli such as Il-1, LPS and fluid flow (12–21, 29). NO production can be assessed by the measurement of nitrite, a stable end production of NO (34). This method was used in the current study to assess NO production by chondrocytes seeded within agarose constructs. After 48 h in culture nitrite concentration was identical in the culture medium and construct, indicating that nitrite diffuses rapidly into the medium without specific retention in the construct. Thus nitrite levels were monitored in the medium alone in subsequent experiments. Measurable levels of nitrite were present in the culture medium from chondrocyte seeded constructs maintained in DMEM 1 20%

FCS, while no nitrite could be detected in the culture medium from cell free construct, indicating a significant basal cellular production of nitrite. LPS was used to activate chondrocyte NO synthesis in order to assess the efficacy of NOS inhibitors for use in compressive strain experiments and to determine the NOS isoforms involved in basal and activated NO production by bovine chondrocytes. L-NAME is a reversible competitive inhibitor of all NOS isoforms but has markedly differing inhibition Ki values for NOS isoforms (cNOS 5 0.5 mM, iNOS . 1000 mM, Calbiochem, Nottingham, UK product information). Thus, when used at 1 mM L-NAME exhibits partial selectivity toward inhibition of cNOS activity. By contrast, dexamethasone inhibits the expression of iNOS (Ki 5 0.005 mM, Calbiochem, Nottingham, UK product information) but has no effect on cNOS activity. L-NAME significantly reduced basal nitrite production whereas dexamethasone failed to induce a significant inhibition of nitrite production (Fig. 1). Basal nitrite production appears, therefore, to be primarily associated with the activity of cNOS activity and does not involve the induction of iNOS. Nitrite production in LPS stimulated cells appeared to be due to a combination of iNOS induction-dependent and the basal cNOS dependent pathways, since both inhibitors significantly reduced nitrite production in the presence of LPS. Dynamic mechanical compression clearly influences nitrite production by chondrocytes seeded in agarose (Fig. 2). While all three frequencies of dynamic strain significantly inhibited nitrite production, statistical analysis revealed a direct correlation between the frequency rate and the level of inhibition (ANOVA p , 0.05). Thus, the response has a frequency dependent component. There was a clear inverse correlation between dynamic compressive strain-induced alterations in nitrite production and [3H]-thymidine incorporation but none with sulfate incorporation (Fig. 3). This may indicate that NO is a component of mechanotransduction pathways influencing cell proliferation but not

583

Vol. 251, No. 2, 1998

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

FIG. 4. The percentage change from unstrained control values for nitrite production (A), [3H]-thymidine incorporation (B) and sulfate incorporation (C) by chondrocytes seeded in agarose constructs and subjected to 15% gross compressive strain at 1 Hz for 48 h. The constructs were cultured in DMEM 1 20% FCS or DMEM 1 20% FCS supplemented with 1 mM L-NAME or 10 mM dexamethasone. Each value represent the mean 6 standard error of the mean for four replicates. Unpaired t-test results indicate differences between supplemented and unsupplemented values as follows: *p , 0.05.

proteoglycan synthesis. These data support a previous studies which suggested that proteoglycan synthesis by bovine chondrocytes is not influenced by nitric oxide (15). These findings additionally support the hypothesis that chondrocyte proliferation and proteoglycan synthesis are regulated by distinct mechanotransduction signaling mechanisms as has been reported previously (33). The addition of L-NAME to constructs subjected to dynamic compressive strain at 1 Hz significantly reduced compressive strain-induced inhibition of nitrite production while dexamethasone failed to block the effect. These data suggest that the pathway by which mechanical compression alters nitric oxide production is independent of induction of iNOS and is probably due to alterations in the activity of cNOS. L-NAME prevented compressive strain-induced stimulation of [3H]-thymidine incorporation within chondrocytes sub-

jected to 1 Hz dynamic compressive strain. These findings confirm a role for nitric oxide in mechanotransduction pathways which influence cell proliferation within the chondrocyte/agarose system, determined primarily by alterations in cNOS activity. Interestingly, dexamethasone reduced compressive straininduced stimulation of [3H]-thymidine incorporation suggesting that iNOS induction dependent pathways cannot be precluded and may play a secondary role. Dexamethasone, however, has wide-ranging effects on chondrocyte metabolism and proliferation through pathways independent of NO. Neither inhibitor was able to significantly alter the level of compressive strain induced stimulation of sulfate incorporation. It would appear, therefore, that nitric oxide is not a component of mechanotransduction pathways which influence proteoglycan synthesis within the chondrocyte/ agarose system.

584

Vol. 251, No. 2, 1998

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

In summary, this study demonstrates for the first time that dynamic compressive strains influence NO production by chondrocytes. NO appears to be a constituent of mechanotransduction pathways which influence cell proliferation rather than proteoglycan synthesis. The inhibitor experiments also infer that alterations in cNOS activity primarily determine the response. Activity of cNOS is controlled by binding of calcium-calmodulin complex suggesting a role for calcium-sensitive stretch-activated ion channels, know to be present in chondrocytes, as upstream regulators of the effect (22, 37). ACKNOWLEDGMENTS This work was funded by the Engineering and Physical Sciences Research Council, the Biotechnology and Biological Sciences Research Council, and the Horserace Betting Levy Board.

REFERENCES 1. Palmoski, M. J., Perricone, E., and Brandt, K. D. (1979) Arthritis Rheum. 22, 508 –517. 2. Kiviranta, I., Tammi, M., Jurvelin, J., Saamanen, A-M., and Helminen, H. J. (1988) J. Orthop. Res. 6, 188 –195. 3. Palmoski, M. L., and Brandt, K. D. (1984) Arthritis Rheum. 27, 675– 681. 4. Gray, M. L., Pizzanelli, A. M., Lee, R. C., Grodzinsky, A. J., and Swann, D. A. (1989) Biochim. Biophys. Acta 991, 415– 425. 5. Sah, R. L. Y., Kim, Y. J., Doong, J. Y. H., Grodzinsky, A. J., Plaas, A. H. K., and Sandy, J. D. (1989) J. Orthop. Res. 7, 619 – 636. 6. Parkkinen, J. J., Lammi, M. J., Helminen, H. J., and Tammi, M. (1992) J. Orthop. Res. 10, 610 – 620. 7. Burton-Wurster, N., Vernier-Singer, M., Farquar, T., and Lust, G. (1993) J. Orthop. Res. 11, 717–729. 8. Kim, Y. J., Sah, R. L. Y., Grodzinsky, A. J., Plaas, A. H. K., and Sandy, J. D. (1994) Arch. Biochem. Biophys. 311, 1–12. 9. Gray, M. L., Pizzanelli, A. M., Grodzinsky, A. J., and Lee, R. C. (1988) J. Orthop. Res. 6, 777–792. 10. Urban, J. P. (1994) Br. J. Rheumatol. 33, 901–908. 11. Moncada, S., Palmer, R. M. J., and Higgs, E. A. (1991) Pharmacol. Rev. 43, 109 –142. 12. Stadler, J., Stefanovic-Racic, M., Billiar, T. R., Curran, R. D., McIntyre, L. A., Georgescu, H. I., Simmons, R. L., and Evans, C. H. (1991) J. Immunol. 147, 3915–3920. 13. Taskiran, D., Stefanovic-Racic, M., Georgescu, H., and Evans, C. (1994) Biochem. Biophys. Res. Commun. 200, 142–148. 14. Blanco, F. J., and Lotz, M. (1995) Exp. Cell Res. 218, 319 –325.

15. Stefanovic-Racic, M., Morales, T. I., Taskiran, D. McIntyre, L. A., and Evans, C. H. (1996) J. Immunol. 156, 1213–1220. 16. Murrel, G. A. C., Dolan, M. M., Jang, D., Szabo, C., Warren, R. F., and Hannafin, J. A. (1996) J. Bone Joint Surg. 78A, 265–274. 17. Bird, J. L., Wells, T., Platt, D., and Bayliss, M. T. (1997) Biochem. Biophys. Res. Commun. 238, 81– 85. 18. Frean, S. P., Bryant, C. E., Froling, I. L., Elliott, J., and Lees, P. (1997) Equine Vet. J. 29, 98 –102. 19. Sasaki, K., Hattori, T., Fujisawa, T., Takahashi, K., Inoue, H., and Takigawa, M. (1998) J. Biochem. 123, 431– 439. 20. Blanco, F. J., Ochs, R. L., Schwarz, H., and Lotz, M. (1995) Am. J. Pathol. 146, 75– 85. 21. Frenkel, S. R., Clancy, R. M., Ricci, J. L., DiCesare, P., Rediske, J. J., and Abramson, S. B. (1996) Arthritis Rheum. 39, 1905– 1912. 22. Marletta, M. A. (1993) J. Biol. Chem. 268, 12231–12235. 23. Abu-Soud, H. M., Feldman, P. L., Clark, P., and Struehr, D. J. (1994) J. Biol. Chem. 269, 32318 –32323. 24. Radomski, M. W., Palmer, R. M., and Moncada, S. (1990) Proc. Natl. Acad. Sci. USA 87, 10043–10047. 25. Noris, M., Morigi, M., Donadelli, R., Aiello, S., Foppolo, M., Todeschini, M., Orisio, S., Remuzzi, G., and Remuzzi, A. (1995) Circ. Res. 76, 536 –543. 26. Hishikawa, K., and Luscher, T. F. (1997) Circulation 96, 3610 – 3616. 27. Johnson, D. L., McAllister T. N., and Frangos, J. A. (1996) Am. J. Physiol. 271, E205–E208. 28. Pitsillides, A. A., Rawlinson, S. C., Suswillo, R. F., Bourrin, S., Zaman, G., and Lanyon, L. E. (1995) FASEB J. 9, 1614 –1622. 29. Das, P., Schurman, D. J., and Lane-Smith, R. (1997) J. Orthop. Res. 15, 87–93. 30. Buschmann, M. D., Gluzband, Y. A., Grodzinsky, A. J., and Hunziker, E. B. (1995) J. Cell Sci. 108, 1497–1508. 31. Lee, D. A., and Bader, D. L. (1995)In Vitro Cell Dev. Biol. Animal 31, 828 – 835. 32. Knight, M., Lee, D. A., and Bader, D. L. (1996) Cell. Eng. 1, 97–102. 33. Lee, D. A., and Bader, D. L. (1997) J. Orthop. Res. 15, 181–188. 34. Gross, S. S., Jaffee, E. A., Levi, R., and Kilbourn, R. G. (1991) Biochem. Biophys. Res. Commun. 178, 823– 829. 35. Masuda, K., Shirota, H., and Thonar, J-M. A. (1994) Anal. Biochem. 217, 167–175. 36. Rao, J., and Otto, W. R. (1992) Anal. Biochem. 207, 186 –192. 37. Guilak, F., Donahue, H. J., Zell, R., Grande, D., McLeod, K. J., and Rubin, C. T. (1994) in Cell Mechanics and Cellular Engineering (Mow, V. C., Hochmuth, R. M., Guilak, F., and TranSon-Tay, R., Eds.), pp. 380 –397, Springer, New York.

585