Nitric oxide and inflammatory arthritides

LifeSciences,Vol. 61, No. 21 pp. 2073-2081.1997 Copyright 0 1997Else&r Sciare Inc. Printedin the USA. All ri ts reserved OK%-320997 P17.00 + .oo ELSEVIER

MINIREVIEW NITRIC OXIDE AND INFLAMMATORY ARTHRITIDES

Nobuyuki Miyasaka’ and Yukio Hirata*

First Department of Internal Medicine’ and Second Department of Internal Medicine’, School of Medicine, Tokyo Medical & Dental University l-5-45, Yushima, Bunkyo-ku, Tokyo, Japan

(Received in final form August 19,1997)

Summary Nitric oxide (NO), first identified as endothelium-derived relaxing factor (EDRF), is a free radical synthesized from L-arginine by NO synthases @OS). NO plays vital roles in biological responses, including regulation of vascular tone, neurotransmission, anti-viral defense and immune response. There are two isoforms in NOS; constitutive NOS (cNOS) and inducible NOS (iNOS). Inflammatory cytokines such as interleukin-1 (IL-l), interferony (IFN-y), tumor necrosis factor-a (TNF-a) induce iNOS expression in various cells including macrophages. NO production is increased in inflammatory arthritides both in rodent models and human. Enhanced NO production is observed in various compartment in vivo but inflammatory synovium and cartilage are the major source of NO. The onset of arthritis in rodent models is successfi~lly blocked by the NOS inhibitor, NG-monomethyl arginine (L-m). These data suggest a possible involvement of NO in the pathogenesis and tissue destruction in arthritis, and the significance of up-regulated NO production is discussed. Key Words: nitric oxide, rheumatoid arthritis, joint destruction Introduction

Nitric oxide (NO) is a short-lived, gaseous free radical, synthesized from L-arginine by NO synthases (NOS) (1). NO is involved in various physiological events such as neurotransmission, endotheliumdependent relaxation of the blood vessels and cell-mediated immune response (2). NO has also been Address for Correspondence:Nobuyuki Miyasaka, M.D., Professorof Medicine, First Departmentof InternalMedicine, Tokyo Medical &Dental University, l-5-45, Yushima, Bunkyo-ku, Tokyo, Japan; Phone: +81-3-5803-5201, FAX: +81-35684-0057; e-mail: [email protected]

2074

Nitric Oxide and Arthritis

Vol.61,

No. 21, 1997

implicated as a mediator of inflammatory responses (3). There are at least two isoforms of NOS; oneis constitutive NOS (cNOS) and Ca2+/calmodulin dependent, and the other is cytokine-inducible NOS (iNOS) and Ca2+/cahnodulin independent (4). Inflammatory mediators such as interleukin- 1 (IL-l), interferon-y (EN-y), tumor necrosis factor-a (TNF-a), and lipopolysaccharide (LPS) stimulate iNOS expression in rodent macrophages in vitro. The stimulated macrophages produce large amounts of NO for prolonged time periods. NO production and/or iNOS expression are also induced by inflammatory mediators in a wide variety of mammalian cells, such as hepatocytes, endothelial cells, tibroblasts, chondrocytes, and neutrophils (5). Rheumatoid arthritis (RA) is a human chronic inflammatory disease characterized by synovitis in multiple joints (6). It also causes extraarticular manifestations such as interstitial pneumonitis, vasculitis and secondary amyloidosis resulting in increased mortality. Pathologically, it is characterized by intense infiltration of T cells and macrophages, marked hyperplasia of synovial lining cells, and neovascularization in the synovium. Infiltrating T cells are predominantly CD4+CD4SRO+ cells expressing various activation markers including HLA-DR molecules. Activation of monocytes/macrophages is also marked, as shown by abundant production in inflammatory cytokines, such as IL- 1, IL-6, TNF-or, and granulocyte/macrophage-colony stimulating factor (GM-CSF) (7), and in multiple low molecular bioactive substances including endothelin- 1, a potent vasoconstrictive peptide (8). Synovial macrophages surrounding T cell foci express costimulatory molecules such as CD86 (9) and CD40 (Sekine et al. submitted for publication), contributing to the sustained immune responses between infiltrating T cells and synovial cells in patients with RA. These perpetuating processes lead to progressive destruction of cartilage and bone, which result in impaired quality of life in arthritic patients. Farrell et al. ( 10) firstly showed increased NO production in both RA and osteoarthritis (OA) patients and suggested a possible role of NO in inflammatory responses seen in arthritis. Furthermore, augmented NO production and iNOS expression has been reported in several animal models for arthritis including arthritis induced by streptococcal cell-wall G-agrnents(1 l- 19). The onset of arthritis was blocked by the NOS inhibitor, No-monomethyl- L-arginine ( L-NMMA). These data suggest that NO might play an important role in the induction or maintenance of arthritis. In this article, we describe excessive NO production and iNOS expression in inflammatory arthritides both in rodents and human, and discuss their pathophysiological significance in inflammatory arthritides. . .

NO Production in AnigLBLLLbrthrW Mode& Rats given streptococcal cell wall fragments (SCW) develop acute inflammation in the joint followed by chronic erosive arthritis, which shares many features with human RA. McCartney-Francis et al. (11) reported that inflamed synovium produced large amounts of NO and expressed increased levels of iNOS mRNA. Peripheral blood mononuclear cells (PBMC) isolated from SCW-induced arthritic rats released significantly increased amounts of NO, which was blocked by the specific NOS inhibitor, L-NMMA. Administration of L-NMMA markedly suppressed swelling and cellular infiltration in the joint. MRL-lpr/lpr mice, which spontaneously develop glomerulonephritis and arthritis, were successfully treated by orally administered L-NMMA (13). Increased expression of iNOS mRNA was noted in kidney and spleen from MRL-lpr/lpr mice, and peritoneal macrophages obtained from these mice showed an enhanced capacity to produce NO in the culture.

Vol. 61, No. 21, 195’7

Nitric Oxide and Arthritis

2075

Adjuvant-induced arthritis (AIA) and collagen-induced arthritis (CIA) are different experimental models for human RA. Injection of either Freund’s complete adjuvant or type II collagen induces arthritis, respectively. Enhanced NO levels and/or iNOS expression were identified and NOS inhibitors blocked the induction of arthritis in these models ( 12,14,16,17). Arthritis and weight loss was exacerbated by L-arginme, the substrate for NOS, and suppressed by the NOS inhibitor, No-nitroL-arginine methyl ester ( L-NAME), in a dose-dependent fashion (12). In contrast, D-aq#he did not affect the disease. Another study indicated that two selective iNOS inhibitors, aminoguanidine and N-iminoethyl- L-lysine, reduced plasma nitrite levels and arthritic changes seen in AIA simultaneously (17). Not many studies were carried out to identify the cells producing NO in the synovium. Stadler et al. (19) initially reported that articular chondrocytes synthesized NO in response to IL- 1 and LPS. A combination of IL-l, TNF-a and LPS induced maximal production of NO in their study. NO production from chondrocytes was significantly augmented with the addition of culture supematants from concanavalin-A-activated lymphocytes and was abrogated by L-NMMA (20). However, the number of chondrocytes present in synovium may not be enough to account for the levels of NO produced in the synovium. Stefanovic-Racic et al. (2 1) found that synovial fibroblasts produced NO in response to 11-l and TNF-a whereas transforming growth factor p (TGF-p) suppressed its production. In contrast, others reported that iNOS immtmoreactivity was localized in synovial macrophages (17). These studies suggest that synovial cells can also be the source of NO in the joint.

Urinary Nitrate Excretion Major source of urinary nitrate is endogenously synthesized NO if there is no excess dietary intake of nitrate. NO is rapidly oxidized to nitrite and nitrate, and excreted into the urine. Urinary nitrate excretion in R4 patients was 2.7-fold greater than that in healthy controls and was reduced by the administration of prednisone (21). However, the amount of urinary nitrate excretion did not correlate with values of CRP, erythrocyte sedimentation rate, joint count, or duration of morning stiffness. Urinary nitrate: creatinine ratios are also reported to be significantly elevated (average 3-fold elevation) in RA patients (22). These studies indicate excessive in viva NO synthesis in RA patients. Serum NO Levels Serum nitrite concentrations in patients with RA and osteoarthritis (OA) were significantly higher than in age- and sex-matched controls (10). Serum nitrite concentration in RA was higher than that in OA, which may reflect the intense inflammatory changes in RA. However, there was no significant correlation between sermn nitrite and CRP. This could also be explained by the timing of blood sampling, as nitrite has a short serum half life. Levels of 3-nitrotyrosine, which can be produced by NO-dependent oxidative damage, are also elevated in patients with active RA (23). In contrast, sera from OA patients and normal controls had no detectable 3-nitrotyrosine with HPLC analysis. A recent study using sensitive chemiluminescence assay indicated that serum NO levels were higher in patients with RA than in patients with OA or in healthy subjects (24). Furthermore, NO levels correlated with the disease activity when assessed by duration of morning stitiess, the number of tender or swollen joints, and CRP in the same study. NO levels also significantly correlated with TNF-a and IL-6 levels in RA sera. These studies suggest that up-regulated NO production in RA patients may reflect heightened inflammatory changes in the joint.

2076

NO Production

Nitric Oxide and Arthritis

Vol. 61,No.

21, 1997

from PBMC

A very recent report revealed that fresh PBMC from RA patients had increased iNOS activity and increased iNOS antigen compared with those from normal controls, thus indicating systemic activation for iNOS expression in RA patients (25). Stimulation of PBMC with IFN-y resulted in increased iNOS expression and nitrite/nitrate production in vitro. iNOS activity of PBMC significantly correlated with disease activity, as assessed by tender and swollen joint counts in this study. NO Concentration

in Synovial Fluid

Synovial fluid nitrate was significantly higher than serum nitrate in patients with RA, indicating local YO synthesis in rheumatoid synovium (IO). Nitrotyrosine was also increased in RA synovial fluid (23). Our group and others have found that nitrite concentration was increased in not only RA patients but also in OA and gouty arthritis patients; thus we speculate that increased NO production is not unique to RA but may reflect nonspecific inflammatory changes, though not specific immune responses in the lesion (10,26). NO Production

in Cartilage

iNOS expression was induced when human chodrocytes were stimulated with IL- 1p, TNF-a or endotoxin in a dose- and time-dependent manner (19). Human iNOS was subsequently cloned from human chondrocytes stimulated with IL- 1p (28). Others also demonstrated that NO was vigorously produced from human IL-l-stimulated chondrocytes indicating that the cartilage is a major source of NO in the joint (29). NO production from human chondrocytes required both RNA and protein synthesis, and was inhibited by L-NMMA, cycloheximide, dexamethoasone and hydrocortisone, but not by indomethacin. NO had an antiproliferative effect on IL-1 induced proliferation of synovial cells in vitro (30). Furthermore, Blanc0 et al. (31) showed that NO was the primary inducer of apoptosis in human chodrocytes. NO, exogenously generated from NO donors, efficiently induced apoptosis of chondrocytes, whereas oxygen radicals caused necrosis. Endogenously induced NO was not sufficient for chondrocyte apoptosis but quenching oxygen radicals resulted in the induction of apoptosis suggesting that the balance between NO and oxygen radicals determines the type of cell death in chondrocytes (3 1). We also found that both e.~ vivo culture of cartilage and in vitro cultured chondrocytes obtained from RA patients produce large amounts of NO (26). Increased iNOS mRNA was also demonstrated by both reverse-transcriptase polymerase chain reaction (RT-RCR) and in situ hybridization assay. NO Production

in Synovial Tissue

Ex-vivo cultured rheumatoid synovium produced a significant amount of nitrite, and the addition of L-NMMA significantly inhibited NO production (Fig. 1) (26). Immunohistochemical studies and in situ hibridization analysis revealed that both iNOS protein and mRNA were predominantly expressed in synovial lining cells, endothelial cells, chondrocytes, and, to a lesser extent, the infiltrating mononuclear cells and synovial fibroblasts (Fig. 2). Most of the synovial lining cells and infiltrating mononuclear cells reacting with anti-iNOS antibodies express CD14 and HLA-DR in this study, which suggests that most of iNOS producers are type A synovial cells. Normal synovium expressed iNOS protein minimally in synovial lining cells and endothelial cells. Western blot analysis of rheumatoid synovial cells showed that iNOS antibodies reacted to a 130-kD band protein corresponding to the human iNOS protein. Reverse-transcriptase (RT)-PCR analysis also confirmed an abundant iNOS mRNA expression in both RA and OA synovium (Fig. 3). In vitro study revealed that NO production from freshly isolated synovial cells was up-regulated by stimulation with a combination of IL-1 p, TIW-a and LPS (Fig. 4). However, synovial fibroblast cell lines in long-term culture did not liberate NO even after stimulation. In this connection, others recently reported that

Vol. 61,No. 21, 1997

Nitric

L-NMMA

(-)

Oxide and Arthritis

(+)

synovium

2077

f-1

(+I

cartilage

FIG. 1 Ex vivo NO production by the synovium and the cartilage from RA patients and its inhibition by L-NMh4A.Fitly mg synovium or cartilage was cultured for 24 h in the presence or absence of 0.5 mM L-NMMA. The nitrite concentration in the culture medium was quantified by using Griess reaction.

most of the iNOS-producing synovial cells in both R4 and OA synovium are negative for nonspecific esterase (NSE) and CD68, which are thought to be macrophage markers, and a minor population reactive with anti-iNOS antibodies expressed both markers (32). These authors speculate that the majority of iNOS producing cells among synovial cells are synovial fibroblasts (synovial B cells). They also showed that synovial adherent cells stimulated with S -nitroso-acetyl penicillamine, a NO donor, produced high concentrations of TNF-a in the culture.

inir SgfCB

n

fN

Pr

Whether excessive NO production in inflammatory arthritides is pro-inflammatory or antiinflammatory remains to be clarified. Possible pro-inflammatory effects of NO include augmentation of vascular permeability in inflamed tissues (33), the generation of destructive free radicals such as peroxynitrite and hydroxyl radical (34), the induction of cycloxygenase, and inflammatory cytokines like TNF-a and IL-1 (32,35). NO induces the production of angiogenic cytokines (36) and activates matrix metalloprotease. NO may also be relevant to induce chondrocyte apoptosis (31). Experimentally-induced arthritis are reversed by NOS inhibitors (1 1- 13, 16- 17). Mice lacking iNOS gene show reduced nonspecific inflammatory response to carregeenin (37). These findings definitely establish the crucial role of NO as an inflammatory mediator. In contrast, NO and its derivative, Snitrosoglutathione, inhibit DNA synthesis of T cells in association with intracellular cGMP accumulation, which may explain suppressed T cell function in RA patients (38). NO also shows

2078

Nitric Oxide and Arthritis

Vol. 61, No. 21, 1997

FIG. 2 In situ hybridization in iNOS mRNA in rheumatoid synovium probed with antisense riboprobes. iNOS mRNA was detected in synovial lining cells (large arrow/tea&), endothelial cells (smaN arrowheads),

and a part of infiltrating cells and fibroblasts. cytoprotective effect through inactivation of oxygen radicals (39). NO may exert an antiinflammatory effect in this sense. However, it should be noted that low amounts of NO increase prostaglandin production and bone resorption, while high amounts inhibit both (19,40). NO may therefore have biphasic or opposite biological effects in the joints, depending on its local concentration. It may therefore be a ‘double-edged sword’ in inflammatory artbritides (41).

s33,9

hwcL,,,,,,,,~ 2?B8P~~46~00000 iNOS

-357

b.p.

0 - actin

-595

b.p.

FIG. 3 Southern blot hybridization of iNOS RT-PCR products amplified from synovium of RA and OA patients. (Lanes 104) Noninflamed synovium from patients with trauma, (Lanes 5-9) RA synovium, (Lanes 10-14) OA synovium. The iNOS mRNA were increased in inflamed synovium compared with noninflamed synovium.

Vol. 61, No. 21, 1997

Nitric Oxide and Arthritis

2079

Synoviocvteg control IL-1 p IL-1 P+TNFa+LPS Chondrocvteq control IL-1p IL-1 P+TNFa+LPS r--

0

2b

1’0

36

Nitrite [ pM ] FIG. 4 NO production from isolated rheumatoid synovial cells and cultured chondrocytes in the presence of IL1@(1 @ml) alone or a combination of II-1 p (1 @ml), TNF-a (10 U/ml) and LPS (10 @ml) for 48h. The nitrite in the media increased,25fold by stimulation with a combination of the 11-l p, TNF-a, and LPS. Data are shown as means + SD from triplicate cultures.

NO synthesis is under transcriptional, post-transcriptional and translational controls (5). Especially cytokines such as IL- 1, TNF-a and IFN-y induce NO production, whereas TGF- p and IL- 10 suppress it. These cytokines are abundantly elaborated from inflammatory synovium over a prolonged period of time (7) and the amount of NO present in inflamed joints could therefore be a net effect of the inflammatory cytokines. The significance of NO synthesis in rodent macrophages has been well established, although it is not certain whether an analogy can be extended to human monocytes/macrophages. Thus far, there are several evidences for human monocytes/macrophages to synthesize NO. Human macrophages exposed to TIW-a, GM-CSF, and selected avirulent strains of Mjcobucferium avium produce NO (42). F’uhnonary alveolar macrophages activated by Pneumocystis carinii are also shown to elaborate NO (43). However, expression and regulation of iNOS in human monocy-teslmacrophages has remained unknown, and high output NO production from human macrophages has not been observed. Our data showed that human CD14+ synovial cells expressed iNOS and produced NO in inflammatory synovium (26). Thus, it is likely that monocytes/macrophages produce NO in selective situations, perhaps in response to most complete stimuli. It needs to be added, however, that chondrocytes and possibly synovial tibroblasts also produce a significant amount of NO in inflammatory synovium. It also has to be pointed out that RA and OA synovium express similar amounts of iNOS mRNA (26). However, it is noted that OA joints have slight to moderate inflammatory changes in limited areas of the synovium in association of the destruction of cartilage. In addition, Farrell et al. (10) also have reported increased NO production both in sera and synovial fluids of OA patients. It will therefore be interesting to clarify NO producing cells in OA. NO was originally identified as the endothelium-derived relaxing factor (EDRF), which regulates vascular tone (44). We have already reported increased levels of endothelin- 1, a potent vasoconstrictor, in inflammatory arthritides (8). Endothelin-1 was shown to be mainly produced by type A synovial cells. Endothelins promote the synthesis of NO through type B receptors in vitro (45), thus generating a

Nitric Oxide and Arthritis

2080

possible

positive

feedback

regulation

in inflammatory

Vol. 61, No. 21, 1997

artbritides.

vasoconstrictive and vasodilative substances are up-regulated

It is also intriguing

in the inflammatory

to

find that both

synovium.

In conclusion, NO is introduced as a new member of mediatory overexpressed in inflammatory synovium. Although NO plays a pivotal role in inflammation, whether modulation of NO synthesis and/or quenching its effect is “YES” or “NO” for successful therapeutic intercvention for human arthritides including RA remains to be solved because of its dual nature.

References 1.

C. NATHAN,

2.

S.MONCADA and A. HIGGS, New Engl. J. Med, 329 2002-20 12 ( 1993). H.H.H.W. SCHMIDTt and U. WALTER, Cell 28 919-925 (1992).

3.

FASEB J. 63051-3064

(1991).

4.

C. NATHAN

5. 6.

C. NATHAN and Q-W. XIE, J. Biol. Chem. 269 13725-13728 M. FELDMANN, F.M. BRENNAN and R.N. MAINI, Cell, a

and Q-W. XIE, Cell, 78 915-918 (1994).

7.

G.S. FIRESTEIN,

8.

N. MIYASAKA, Y. HIRATA, K. ANDO, K. SATO, H. MORITA, TOMITA and F. MARUMO, Arthritis Rheum. 35 397-400 (1992).

9. to.

J.M. ALVARO-GRACIA

(1994). 307-3 10 (1996).

M.F. LIU, H. KOHSAKA,

H. SAKURAI,

and R. MAKI, J. Immunol. 144 3347-3353

M. AZUMA,

Arthritis Rheum. 39 110-l 14 (1996). A.J. FARRELL, D.R. BLAKE, R.M.J. PALMER

M. SHICHIRI,

K. OKUMURA,

and S. MONCADA,

(1990). K. KANNO,

K.

I. SAITO and N. MIYASAKA, Ann. Rheum. Dis. 51 I2 19- 1222

(1992). Il.

N. MCCARTNEYOFRANCIS, and S.M. WAHL,

12. 13.

J.B. ALLEN,

J. Exp. Med. 178 749-754

D.E. MIZEL, J.E. ALBINA,

A. IALENTI, S. MONCADA and M.D. ROSA, Br. J. Pharmacol. 110 701-706 (1993). J.B. WEINBERG, D.L. GRANGER, D.S. PISETSKY, M.F. SELDIN, M.A. MISKONIS, MASON,

A.M.

PIPPEN, P. RUIZ, E.R. WOOD

and G.S. GILKESON,

14.

(1994). Y. OYANAGUI,

15.

D.O. StICHTENOTH,

16.

J.C. FROELICH, Ann. Rheum. Dis. &j 547-549 (1994). M. STEFANOCVIC-RACIC, K. MEYERS, C. MESCHTER, C.H. EVANS, Arthritis Rheum. 37 1062-1069 (1994).

17. 18.

Q-W. XIE, C.F. NATHAN

(I 993).

Life Sci. u

PI, 285-289 (1994).

F-M. GUTZKI,

D. TSIKAS,

N. SELVE,

S.M. BOEGER,

J.W. COFFEY,

R.H. BOEGER

R.A. HOFFMAN

J.R. CONNER, P.T. MANNING, S.L. SETTLE, W.M. MOORE, G.M. JEROME, F.S. TJOENG and M.G. CURRIE, Eur. J. Pharmacol. 273 15-24 (1995). G. W. CANNON, M.M. GRIFFITHS,

S.N.

J. Exp. Med. 179 65 l-660

S.J. OPPENSHAW,

J.B. HIBBS,

Arthritis Rheum. 39 1677-1684

JR., J.R. HOIDAL,

and and

R.K. WEBBER,

T.P. HUECKSTEADT

and

(1996).

21.

M. STEFANOVIC-RACIC, T.R. BILLIAR, R.D. CURRAN, L.A. MCINTYRE, H.I. GEORGESCU, R.L. SIMMONS and C.H. EVANS, J. Immunol. 147 3915-3920 (1991). M. STEFANOVIC-RACIC, J. STADLER, H.I. GEORGESCU and C.H. EVANS, J. Rheumatol. 21 1892-1898 (1994). D.O. StICHTENOTH, J. FAULER, H. ZEIDLER and J.C. FROELICH, Ann. Rheum. Dis. 54 820-824

22.

(1995). P.S. GRABOWSKI,

19. 20.

23. 24. 25. 26. 27.

J. STADLER,

A.J. ENGLAND,

R. DYKHUZEN,

M. COPLAND,

N. BENJAMIN,

D.M. REID

and S.H. RALSTON, Arthritis Rheum. 39 643-647 (1996). H. KAUR and B. HALLIWELL, FEBS Letters, 350 9- I2 (1994). Y. UEKI, S. MIYAKE, Y. TOMINAGA and K. EGUCHI, J. Rheumatol. 23 230-236 (1996). E.W. ST. CLAIR, W.E. WILKINSON, T. LANG, L. SANDERS, M.A. MISUKONIS, G.S. GILKESON, D.S. PISETSKY, D.L. GRANGER and J.B. WEINBERG, J. Exp. Med. 184 1173-l 178 (1996).

H. SAKURAI, H., H. KOHSAKA, M-F. LIU, H. HIGISHIYAMA, Y. HIRATA, K. KANNO, 1. SAITO and N. MIYASAKA, J. Clin. Invest. 96 2357-2363 (1995). R.M.J. PALMER, M.S. HICKERY, I.G. CHARLES, S. MONCADA and M.T. BAYLISS, Biochem. Biophys. Res. Comm. 193 298-405 (1993).

Vol. 61, No. 21, 1997

28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45.

Nitric Oxide and Arthritis

2081

I.G. CHARLES, R.M.J. PALMER, MS. HICKERY, M.T. BAYLISS, A.P. CHUBB, V.S. HALL, D.W. MOSS and S. MONCADA, Proc. Natl. Acad. Sci. USA, 90 11419-l 1423 (1993). J-J. REDISKE’, C.F. KOEHNE, B. ZHANG and M. LOTZ, Osteoarthritis Cartil. 2 199-206 (1994). F. BLANC0 and M. LOTZ, FASEB J, 8 A365 (1994). F.J. BLANCO, R.L. OCHS, H. SCHWARZ and M. LOTZ, Amer. J. Pathol. 146 75-85 (1995). I.B. MCINNES, B.P. LEUNG, M. FIELD, X.Q. WEI, F-P. HUANG, R.D. STURROCK, A. KINNINMONTH J. WEIDNER, R. MUMFORD and F.Y. LIEW, J. Exp. Med. & 15 19- 1524 (1996). W.G. MAYHAN, Inflammation, 16 295-305 (1992). J.S. BECKMAN, T.W. BECKMAN, J. CHEN, P.A. MARSHALL and B.A. FREEMAN, Proc. Natl. Acad. Sci. USA, a 1620-1624 (1990). H.M. LANDER, P. SEHAJPAL, D.M. LEVINE and A. NOVOGRODSKY, J. lmmunol. m 15091516 (1993). S.J. LEIBOVICH, P.J. POLVERlNl, P.J., T.W. JONG, L.A. HARLOW and A.E. KOCH, Proc. Natl. Acad. Sci. USA, 914190-4194 (1994). X. WEI, I.G. CHARLES, A. SMITH, J. URE, G-j. FENGg, F-p. HUANG, D. XU, W. MULLER, S. MONCADA and F.Y. LIEW, Nature, 375 408-411 (1995). P.F. MERRYMAN, R.M. CLANCY, X.Y. HE and S.B. ABRAMSON, Arthritis Rheum. u 1414-1321 (1993). G.M. RUBANYI, E.H. HO, E.H. CANTOR, WC. LUMMA and L.H. BOTHELO, Biochem. Biophys. Res. Comm. 1811392-1397 (1991). S.H. RALSTON, L-p. HO, M.H. HELFRICH, P.S. GRABOWSKI, P.W. JOHNSTON and N. BENJAMIN, J. Bone Min. Res. 19 1040-1049 (1995). M. STEFANOVIC-RACIC, J. STADLER and C.H. EVANS, Arthritis Rheum. u 1036-1044 (1993). M. DENNIS, J. Leukocyte Biol. 49 380-387 (1991). M.P. SHERMAN, M.L. LORO, V.Z. WONG and D.P. TASHKIN, J. Protozool. 38:234-236 (1991). R.F. FURCHGOTT, Acta Physiol. Stand. m 257-270 (1990). Y. HIRATA, T. EMORl, S. EGUCHI, K. KANNO, T. IMAI, K. OHTA and F. MARUMO, J. Clin. Invest. 91 1367-1373 (1993).