Modulation of synovial fibroblast plasminogen activator and plasminogen activator inhibitor production by protein kinase C

Modulation of synovial fibroblast plasminogen activator and plasminogen activator inhibitor production by protein kinase C

283 Biochimica et Biophysica Acta, 1097(1991) 283-288 © 1991 Elsevier Science Publishers B.V. All rights reserved 0925-4439/91/$03.50 BBADIS 61083 ...

679KB Sizes 0 Downloads 36 Views

283

Biochimica et Biophysica Acta, 1097(1991) 283-288 © 1991 Elsevier Science Publishers B.V. All rights reserved 0925-4439/91/$03.50

BBADIS 61083

Modulation of synovial fibroblast plasminogen activator and plasminogen activator inhibitor production by protein kinase C Joanne Uhl 1, Robert C. Newton 2, Janet L. Gross and Eugene Mochan 3

2

Waheed Rommi 3

I Department of Inflammation, Sterling Research Group, Rennselaer, N Y (U.S.A.), 2 Department of Medical Products, E.L du Pont de Nemours and Company, Glenolden, PA (U.S.A.) and 3 Departments of Family Medicine and Biochemistry, Unit,ersity of Medicine and Dentistry of New Jersey, School of Osteopathic Medicine, Camden, NY (U.S.A.)

(Received 8 April 1991) (Revised manuscript received 1 July 1991)

Key words: Synovialfibroblast; Protein kinase C; Plasminogen activator inhibitor

Phorbol myristate acetate (PMA) added to human synovial fibroblast cultures caused a dose-dependent increase in the production of plasminogen activator inhibitor-type I (PAI-1). In addition, PMA inhibited endogeneous and interleukin-I (IL-I) induced plasminogen activator (PA) activity, while increasing mRNA PAI-1 levels. Other protein kinase C (PKC) activators, mezerein and teleocidin B4, caused similar effects. The simultaneous addition of the PKC antagonists, H-7 or staurosporine, prevented the inhibition of PA activity by PMA. This study shows that activation of PKC inhibits PA and stimulates PAI production in human synovial fibroblasts. These results suggest that activation of PKC may play an important role in regulating increased PA production associated with joint destruction in rheumatoid arthrits (RA).

Introduction Plasminogen activator (PA) is a serine proteinase that specifically converts the zymogen, plasminogen, into the active proteinase plasmin. At least two types of human PA's are known to exist, tissue PA (tPA) and urokinase (uPA). These forms of PA can be distinguished by differences in molecular weights, affinity for fibrin and immunoreactivity [1,2]. The PA/plasmin proteolytic system has received considerable attention because of its participation in a wide variety of biological activities, which include both normal physiological functions as well as pathological conditions involving tissue destruction [3,4]. Measurement of PA activity reflects the combined action of both PA and inhibitors of PA. Three differ-

Abbreviations: PMA, phorbol myristate acetate; PA, plasminogen activator; PAl, plasminogen activator inhibitor; PKC, protein kinase C; IL, interleukin. Correspondence: J. Uhl, Sterling Research Group, 81 Columbia Turnpike, Rennselaer, NY 12144, U.S.A.

ent types of inhibitors of PA have been isolated from a number of body fluids, tissues and cell lines. Endothelial cells [5] and platelets [6] have been shown to produce an inhibitor designated as plasminogen activator inhibitor type 1 (PAl-l). Another inhibitor of PA, type 2 PAl, has been isolated from the placenta [7] and the histiocytic lymphoma cell line, U-937 [8]. Type 1 PAl and type 2 PAl have been shown to react with both uPA and tPA [9]. The third type of PAl is proteinase nexin I, which has been purified from human foreskin cells [10]. Proteinase nexin has a broader specificity and in addition to inhibiting uPA and tPA, has been shown to react with other trypsin-like serine proteinases [9]. A major factor involved in joint destruction in arthritic disorders, such as rheumatoid arthritis (RA), is increased production of neutral proteinases by an inflammed synovium. Results from a number of studies suggest that PA/plasmin may be one of the enzyme systems involved in the degradative events. These include: (a) elevated levels of PA in RA synovial fluid [11], (b) increased production of PA in interleukin-1 (IL-1) stimulated synovial cells (12); (c) plasmin degradation of cartilage proteoglycan [13]; and (d) plasmin

284 activation of latent collagenase [14] and latent neutral proteoglycanase [15]. Little is currently known about the mechanistic events leading to increased synovial PA production. However, we have previously shown that synovial cell PA production can be mediated through the metabolism of PGE2 and cAMP [16]. The role of protein kinase C (PKC) in signal transduction and its effect on the modulation of metabolic processes has been studied in a number of cell systems [17]. Phorbol esters are known to bind and activate PKC [17] and have been shown to induce PA production in a number of cell types [18-22]. Phorbol esters have also been reported to enhance production of an inhibitor of PA [23-25]. In the present study we investigated the effect of PKC modulation on synovial fibroblast PA and PAI production. Using human synovial fibroblast cultures we found that activation of PKC causes increased PAI production with concomitant decreased PA activity. These results suggest that PKC activation may be involved in regulating synovial fibroblast PA. Materials and Methods

Chemicals. Phorbol myristate acetate (PMA) was purchased from Consolidated Midlands, Brewster, NY. Mezerein, 4-O-methyl-12-O-tetradecanoyl phorbol-1 3-acetate (4-O-methyl TPA) and 4-a-phorbol-12,13-didecanoate (4-a-PDD) were purchased from Sigma, St. Louis, MO. Staurosporine was purchased from Kyowa Hakko U.S.A. New York, NY and 1-5-isoquinolinesulfonyl-2-methylpiperazine dihydrochloride (H-7) was from Seikagaku America, St. Petersburg, FL. Teleocidin B4 was a generous gift from K. Shudo, University of Tokyo, Tokyo, Japan. Stock solutions of reagents were in ethanol or dimethyl sulfoxide and stored at - 2 0 ° C . The reagents were diluted at least 1000-fold in D M E M before adding to the cells. The final concentration of solvent added to the cells had no effect on the synovial fibroblast events measured in these experiments. Syno~ial fibroblast cultures. Non-rheumatoid synovium (from patients with osteoarthritis (OA)) and rheumatoid synovium were obtained with informed consent from patients undergoing reconstructive and restorative joint surgery. Primary cultures were established by enzymatic digestion of the tissue, as previously described [12]. It has been reported by Krane [26] that the primary synovial culture contains a variety of cell types, however, after trypsinization only the fibroblast type cells are observed in culture. The synovial fibroblasts were cultured in Dulbecco's minimum essential medium (DMEM) containing 10% fetal bovine serum (Hyclone, Logan, UT) and 50 /xg/ml gentamycin (Tri Bio Laboratories, State College, PA) at 37°C in an atmosphere of 5% CO 2, 95% air. For

experimental purposes, cells were plated after the second through seventh passages at a conccnlration ol 2 ' 105 cells/ml in 16 mm diameter 24 well plates (Costar, Cambridge, MA). Alter 2 days in culture, lhe media was removed from the confluent cells. The cells were washed twice in serum-free media and then incubated with the indicated PKC activators (agonists) and antagonists in serum-free media fl)r 10 rain. In some experiments the cells were then exposed to IL-I at a final concentration of 10 ng/ml. After 24 h incubation, the conditioned media was removed and stored in polypropylene tubes at - 7 0 ° C until assayed. Synovia[ fibroblast cultures from three different RA patients and two patients with OA were used for this study. The results obtained were consistent for each synovial fibroblast culture and were independent of passage state and source of tissue. IL-1 preparation. Recombinant IL-1 beta was prepared as described in detail elsewhere [27]. The recombinant IL-1 used in this study had a specific activity of 2. 107 units/rag. One unit of IL-I is defined as the amount of 1L-1 which generates half-maximal activity in the thymocyte proliferation assay. Measurement of PA actiLqty. PA activity was determined by a procedure involving hydrolysis of the chromogenic substrate, L~-valyl-leucyl-lysyl-p-nitroanilidc, S-2251 (Ortho Diagnostics, Raritan, N J)[28]. The assay was performed in 96 well flat bottom plates by adding 0.05 ml of sample to 0.2 ml of 0.1 M Tris-HCl (pH 8.3) containing 0.5% Triton X-100 and 40 /xg of plasminogen (Helena, Beaumont, TX). The plates were incubated at 37°C for 10 rain, followed by addition of 10/al of 10 2 M S-2251. At the end of the incubation period (4-24 h), the change in absorbance at 405 nm was recorded using a micro-ELISA reader. PA activity tk~r each sample was determined by employing uPA (Calbiochem, San Diego, CA) as a standard and was represented as mU PA activity/cell number. Background proteinase activity was corrected for by running the samples in the abence of plasminogen. Measurement of PAL1 and PA antigen. PAI-I and PA antigen levels were determined by double antibody sandwich ELISA using the Imubind PAI-I and u-PA ELISA kits (American Diagnostica, Greenwich, CT). Both the P A I d and PA antibodies used in the ELISA detect free protein as well as PA-PAI complexes [29]. RNA isolation and blot hybrization. Confluent synovial fibroblast cultures grown in 150 cm: flasks were exposed to IL-I (10 n g / m l ) a n d / o r PMA (10 riM) for 4 h in serum-free media. Cells were harvested and total cellular RNA isolated by guanidium isothiocynate extraction [30] followed by cesium chloride density gradient centrifigation [31]. RNA (20 ug) was run on formaldehyde gels and transferred onto nylon membranes (Nytran, Schleicher and Schull, Kennc, NH). The membranes were hybridized with ~2P-labeled CTP

285 cDNA probes ( = 3000Ci/nM, New England Nuclear, Boston, MA). The cDNA probes for human u-PA and glyceraldehyde-3-phosphate-dehydrogenase ( G A P D H ) were obtained from The American Type Culture Collection (Rockville, MD). The c D N A PAI-1 probe was a gift from Dr. T.C. Wun (Monsanto). The blots were prehybridized for 6 h at 4 2 ° C in 50% formamide, 5 x SSC (1 x SSC = 0.15 M NaC1 and 0.015 M sodium citrate), 5 x Denhardt's solution (1 X Denhardt's solution = 0.02% polyvinylpyrrolidone, 0.02% Ficoll, 0.02% BSA), 1 M Tris-HCl (pH 7.4) and 2.5 m g / m l denatured salmon sperm DNA. Blots were hybridized overnight at 42°C with denatured 32p-labeled CTP cDNA probes. All blots were then extensively washed with 0.1 x SSC followed by 0.1 x SSC containing 0.1% SDS at 42 °C. Blots were dried, exposed to Kodak XAR-5 film and film developed. Scanning densitometry on blots was performed using a Hoefer GS-300 densitometer.

Effect of PMA on endogeneous and IL-1 induced PA actiuity. Several studies have reported that phorbol esters enhance PA production [18-22]. In the present study human synovial fibroblasts were incubated for 24 h with various concentrations of PMA. The results in Fig. 1, however, show that PMA significantly inhibited both endogeneous PA activity ( P < 0.05) as well as IL-1 induced PA activity ( P < 0.01). The concentration of PMA required to inhibit 50% of IL-1 stimulated PA activity and constitutive PA activity was 5 riM. Cons±s10"

8-

6"

0

g

Effect of other PKC actiL,ators and inactit,e tumor promoters on endogeneous and 1L-I stimulated synovial fibroblast PA activity Synovial fibroblasts were incubated in serum-free media with PKC activators or inactive tumor promoters for 10 rain. IL-1 (10 n g / m l ) was added and the cells incubated for 24 h. The cell supernatants were removed and assayed for PA activity. Data points represent mean PA activity from triplicate samples ±S.E.M. Addition

PA activity ( m U / l O s cells)

Control IL-1 IL- 1 + mezerein (150 n M) Mezerein (150 nM) IL-1 + teleocidin B4 (200 nM) Teleocidin B4 (200 nM) IL-1 +4-o-methyl T P A (10 nM) IL-I + 4 a - P D D (10 nM)

0.9 ± 0.1 5.1 ±0.5 1.4 ± 0.4 0.5 ±0.1 2.9 _+0.5 0.5 _+0.1 6.2_+0.6 6.2+0.5

tent with other reports [19,22], time-course studies revealed maximum effects of PMA occurring between 12 and 24 h (data not shown).

Results

--I ILl 0 u~

TABLE I

4

< O.

0 0

10

20

PMA (nM) Fig. 1. Effect of P M A on e n d o g e n e o u s and IL-1 stimulated synovial fibroblast PA activity. Synovial fibroblasts were cultured as described in Materials and Methods. Confluent cells were incubated in serumfree D M E M with increasing concentrations of P M A for 10 min followed by the addition of IL-1 (10 n g / m l ) . After 24 h the conditioned media was removed and assayed for PA activity. Data is a summary of five independent experiments and data points represent mean PA activity from triplicate samples ± S.E.M. (e • IL-1 + PMA; © © P M A alone).

Effect of other activators of PKC on IL-1 stimulated PA production. To further investigate the effect of phorbol esters on PA activity, the effect of the other PKC activators, mezerein and teleocidin B4, was examined. Both mezerein and teleocidin B4 have been reported to be tumor promoters which can also activate PKC [17]. Mezerein and teleocidin B4 inhibited both constitutive and IL-1 induced PA activity (Table I). As a control, the effects of two analogs of 12-O-tetradecanoyl phorbol-13-acetate (TPA), which are inactive as both tumor promoters and stimulators of PKC, were examined. The results summarized in Table I show that the inactive analogs, 4-O-methyl TPA and 4-aPDD did not inhibit IL-1 induced PA activity in the synovial fibroblasts. Prer,ention of PMA inhibition of PA activity by PKC antagonists. Further support correlating the activation of PKC with decreased PA activity, comes from the findings that antagonists of PKC effectively prevent this inhibition. Table II shows that the addition of either staurosporine (20 nM) or H-7 (25 nM), both inhibitors of PKC activation [17], prevented the PMA inhibition of IL-I stimulated synovial fibroblast PA production. Effect of PMA on PAL1 production. The data in Fig. 1 show that PMA decreases PA activity and suggest that the inhibitory effect of PMA may be due to production of an inhibitor of PA. The results in Fig. 2 show that PMA increased the production of ELISAdetectable PAI-1 production by synovial fibroblasts in a dose-dependent manner. Half maximum stimulation occurred at approx. 5 nM. This concentration of PMA is the same concentration required to inhibit 50% of PA activity as shown in Fig. 1.

286 TABLE 11

•FABLE II1

Pret:ention of PMA inhibition of IL-1 induced PA actit'ity by PK(" activators

E[}Fct of PMA on production of synoHal t~hrohla.~t PA antigen

Synovial fibroblasts (cultured as described in Fig. 1) were incubated with staurosporine (21) nM) or H-7 (25 ~M) for 10 rain. PMA (lll nM) was added for an additional 10 min followed by the addition of IL-I (10 ng/ml). After 24 h the cell supernatants were removed and assayed tl~r PA activity. Values shown represent mean PA activity from triplicate samples + S.E.M. Addition

PA activity ( m U / I I I ~ cells)

Control IL-I IL- 1 + PMA IL- 1 + PMA + staurosporine lL-1 + P M A + H-7

0.9 + 0. I 7.2+0.5 1.3 + 0.2 7.8 4__0.6 6.9+0.7

Effect of PMA on PA antigen production. Several studies [18-22] have reported increased PA production by PMA, but the results presented show decreased PA activity (Fig. 1) with increased PAI production (Fig. 2) following PMA treatment. We therefore, determined whether an excess production of PAI could be masking detectable increased PA activity levels by examining the effect of PMA on PA antigen production. The effect of PMA on u-PA was examined since we previously showed that synovial fibroblasts produce urokinase-type PA [12]. The results in Table III show that PMA does not increase u-PA antigen and is capable of slightly decreasing endogeneous and IL-1 stimulated PA antigen production as measured by ELISA. Effect of PMA on u-PA and PAL1 mRNA levels. Consistent with activity and protein levels, Northern blots (Fig. 3) and densitometric scanning (Table IV) reveal that PMA does not increase basal or IL-1 stimu-

250 -

~" ,,,-I ,-I UJ m ~

200

150

c

,r|

1 O0 13.

0

5

10

15

20

PMA (nM) Fig. 2. Effect of PMA on synovial fibroblast PAI-I production. Synovial fibroblasts (cultured as described in Fig. 1) were incubated with increasing concentrations of PMA. After 24 h, the conditioned media was removed and assayed for PAI-1. Data points represent mean PAI-I production from triplicate samples + S.E.M.

Synovial fibroblast supernatants generated from the experiment in Table lI were assayed for PA antigen by ELISA as described m Materials and Methods. The samples assayed were from cells incu bated with 10 nM PMA a n d / o r 10 n g / m l IL I. Values represent mean of triplicate samples + S.E.M. Addition

PA antigen ( n g / m l )

Control 1L-1 PMA IL-I + P M A

0.7 + 0.06 1.6±0.2 0.5 + 0.05 1.0+l).l

lated u-PA levels (Fig. 3A), and increases PAI-1 mRNA levels (Fig. 3B). As a control, Fig. 3C shows that under these culture conditions, incubation of synovial fibroblasts with IL-1 or PMA had no significant effect on G A P D H mRNA levels. Discussion

The results of the present study provide experimental evidence for the modulation of synovial fibroblast PA and PAl by activation of PKC. This conclusion is based on the following observations: (a) PKC activators inhibited endogeneous and IL-1 stimulated PA activity; (b) PKC antagonists reversed the inhibition of PA by PMA; and (c) PMA induced synovial fibroblast PAI antigen and PAI m R N A production. Several studies have shown that PA production in biological systems is influenced by hormones, retinoids and calcium ionophores [1-2]. Protein kinase C activating agents, such as phorbol esters, have been shown to enhance PA secretion [18-22] as well as to increase PA gene expression [32] in a number of cell types. In the present study, we examined whether PA production in synovial fibroblasts could be modulated by activation of PKC. The results presented show that addition of PKC activators to synovial fibroblasts did not increase PA antigen production or m R N A expression but rather inhibited PA activity. These significant decreases in PA activity by PMA are most likely due to increased PAI-1 levels and not decreased PA production, since PMA only slightly decreased PA antigen and mRNA levels. It has been shown that prolonged incubation of cells with phorbol esters is capable of down regulating PKC [17]. It is therefore possible that the inhibitory effect of PMA on PA activity may result from a down regulation of PKC. To determine whether decreased PA activity was indeed linked to activation of PKC, the effect of antagonists of PKC was examined. The results in Table II show that the simultaneous addition of the PKC antagonists, H-7 or staurosporine, with PMA was capable of preventing the inhibitory effect of PMA. Therefore, these results demonstrate that activation of PKC

287 T A B L E IV 1

2

3

4

(AI

Densitometric scann&g of effect of PMA on u-PA and PAl-1 mRNA Synovial fibroblasts were incubated at 37°C for 4 h with IL-1 (10 n g / m l ) , P M A (10 nM) or a mixture of both. T h e ceils were collected and northern analysis performed as described in Materials and Methods. The reported values represent arbitrary units obtained from laser densitometric scanning of the autoradiographic films shown in Fig. 3.

2.4 kb

Addition

Control IL-I PMA IL-1 + P M A

(B)

1

PAl- 1 (3.4 Kb)

(2.3 Kb)

(1.3 Kb)

GAPDH

0.16 0.26 0.10 0.17

1.45

1.36

4.05

2.55

1.83 2.02 1.73 1.78

2

i~i~i~+!/i~i~i!ii~!i!/~ ,~!~i ~:~ii i!/!~ii~

3.4 kb )P

2.3kb

(C)

u-PA (2.4 Kb)

4

1,3' kb

Fig. 3. Northern blot analysis of u-PA m R N A and PAl-1 m R N A in synovial fibroblasts incubated with PMA. Total R N A was isolated from synovial fibroblasts incubated with IL-1 (10 n g / m l ) , P M A (10 nM) or a mixture of both in serum-free media for 4 h. R N A (20 ng portions) were electrophoresed on formaldehyde gels, transferred to nylon m e m b r a n e s and hybridized with c D N A probes as described in Materials and Methods. The sizes of m R N A ' s calculated from their migration are indicated to the left. (A) u P A m R N A - lane 1 control, lane 2 IL-I, lane 3 PMA, lane 4 IL-1 + PMA; (B) PAI-1 m R N A lane 1 control, lane 2 PMA; (C) G A P D H m R N A - lane 1 control, lane 2 IL-1, lane 3 PMA, lane 4 IL-1 + PMA.

is capable of causing an inhibition of PA activity in synovial fibroblasts. Brinkeroff, et al. [33] demonstrated decreased synovial fibroblast proliferation and increased latent collagenase production by PMA, while our study indicated that PMA decreases PA. This data supports earlier studies [16,34], indicating that PA and collagenase production in synovial fibroblasts are not coordinately controlled. The biologic significance of the separate control mechanisms for these two enzymes remains to be determined, especially in light of reports that PA leads to activation of latent procollagenase. Our results also show that PKC activation in synovial fibroblasts leads to increased production of PAI-1. Although the significance of increased PAI production in synovial fibroblasts is not yet established, this study suggests that PKC activation may play an important role in modulating synovial fibroblast PA production. Whether the concentrations of PAI-I released from synovial fibroblasts can effectively prevent the proteolytic activity of PA found in rheumatoid joints, remains to be determined. However, the presence of human PKC activators in the synovium, such as platelet derived growth factor, may point to a physiological attempt to regulate the P A / P A I balance. In summary, the results of this study have linked PKC activation, leading to production of PAI, as one factor responsible for down regulating PA activity in synovial fibroblasts. Since PA biosynethesis has been associated with a variety of biological processes that involve tissue degradation, events that inhibit PA may play a regulatory role by controlling joint destruction. Studying the mechanistic features of PA production in synovial cells provide a better understanding of the complex cellular and biochemical interactions associated with articular cartilage damage in RA. Interruption of these destructive processes, therefore, could have important implications for pharmacologic control of events associated with rheumatoid cartilage destruction.

288 Acknowledgements The authors would like to thank Jean Williams and Lynne Brophy for their technical expertise, Dr. Elaine Merisko for her helpful comments and Ix)is Boulden for the preparation of the manuscript. References 1 Saksela, O. (1985) Biochim. Biophys. Acta 823, 35-65. 2 Blast, F., Vassalli, J.D. and Dano, K. (1987) J. Cell Biol. 104, 8111-8114. 3 Dano, K., Andreasen, P.A., GrondahI-Hansen, J., Kristensen, P., Nielsen, L.S. and Skriver, L. 11985) Adv. Cancer Res. 44, 139 266. 4 Mullins, D.G. and Rohrlich, S.T. (1983) Biochim. Biophys. Acta 695, 177 214. 5 Van Mourik, J.A., Lawrence, D.A. and Loskutoff, D.J. 11984) J. Biol. Chem. 259, 14914-14921. 6 Erickson, L.A., Ginsberg, M.H. and Loskutoff, D.J. 11984) J. Clin. lnvest. 74, 1465-1472. 7 Astedt, B., Lecander, I., Brodin, T., Lundblad, A. and Low, K. 11985) Thromb. Haemostasis. 53, 122-125. 8 Kruithof, E.K.O., Vassalli, J.D., Schleuning, W.D., Manaliano, R.J. and Bachmann, F. (1986)J. Biol. Chem. 261, 11207-11213. 9 Sprengers, E.D. and Kluft, C. 11987) Blood 69, 381 387. 10 Scott, R.W. and Baker, J.B. (1983) J. Biol. Chem. 258, 10439111444. 1l Mochan, E. and Uh[, J. (1984)J. Rheum. 11, 123-128. 12 Mochan, E., Uhl, J., and Newton, R. (1986) J. Rheum. 13, 15-19. 13 Mochan E. and Keler, T. (1984) Biochim. Biophys. Acta 8/11/, 312 315. 14 Werb, Z., Mainardi, C.L., Vatar, C.A. and Harris, E.D., Jr. (19771 N. Eng. J. Med. 296, 1017-1023. 15 Vaes, G., Eeckhout, Y., Lenaers-Claeys, G., Francois-Gillet, C. and Druetz, J.E. 11978) Biochem J. 172, 261-274. 16 Mochan, E., Uhl, J. and Newton, R. (1986) Arth. Rheum. 29, 1078 1084.

17 Nishizuka, Y.(198~)Science 233,3115 312. 18 Gross, J.L., Moscatclli, D., Jaffe, I:;.A. and Rifkin, I).13. II~S2).I Cell Biol. 95, 974-981. 19 Ashino-Fuse, It., Opdenakker, G., Fuse, A. and Billiau, A. (19~41 Proc. Soc. Exp. Biol. Med, 176, 109 118. 211 Crutchley, D.J. and Maynard, J.R. (19831 Biochim. Biophys. Acre 762, 76 85. 21 Tilly, J i . and Johnson, A.L. 119901 Endocrinology 12~. 2117t~ 21187. 22 Band, V., Karlan, B.Y., Zurawski, Jr., V.R. and Liulefield. B.A. (1989) J. Cell. Physiol. 138, 106-114. 23 Genton, C., Kruithof, E.K.O. and Schleuning, W.D. (19871 J. Cell Biol. 104, 705-712. 24 Rehemtulla, A., Gates. J. and Hart, D.A. (19871 Comp. Biochem. Physiol. 88B, 277-283. 25 Mayer, M., Lund, L.R., Riccio, A., Skouv, ,1., Nielsen, L.S., Staccy, S.N., Dano, K. and Andreasen, P.A. (1988)J. Biol. Chem. 263, 15688-15693. 26 Krane, S.M. (1981) Ann. Rheum. Dis. 41), 433-448. 27 ttuang, J., Newton, R.C., Pezzella, K., Covington, M., Tamblyn, T.. Rutlege, S.J., Kelley, M. and Lin, Y. (1987) Mol. Biol. Med. 4, 169-181. 28 Wun, T.C., Schleuning, W.D. and Reich, D. 11982) J. Biol. Chem. 257, 3276-3283. 29 Binnema, D.J., Van lersel, J.J.g. and Dooijewaard, G. (It186) Thromb. Res. 43, 569-577. 30 Maniatis, T., Fritsch, E.F. and Sambrook, J. (19821 Molecular Cloning: A Laboratory Manual. Cold Spring Harbor, NY, Cold Spring Harbor Laboratory. 31 Glisin. V., Crvenjqakav. R. and Byus. C. (1974) Biochem. 13, 2633. 32 Bell. S.M., Brackenbury, R.W., Leslie, N.D. and Degen, J.[.. 119901 J. Biol. Chem. 265, 1333-1338. 33 Brinckerhoff, C.E., McMillan, R.M., Fahey, J.V. and tiarris, E.D., Jr. (1979) Arth. Rheum. 22, 1109-1116. 34 Golds, E.E., Ciosek. C.P., Jr. and Hamilton J.A. 11983) Arth. Rheum. 26, 15-21.