Nitric Oxide as an Inflammatory Mediatorin Autoimmune MRL-Ipr/lIpr MiceJ. Brice WeinbergDepartment of Medicine, Division of Hematology-Oncology, VeteransAffairs and Duke University Medical Centers, Durham, North Carolina

Nitric oxide (NO) may exhibit proinflammatory features. *NO synthase type 2 (NOS2) isoverexpressed and 'NO overproduced in rodent models of induced inflammation. Blockage of'NO production by administration of NOS inhibitors prevents or reduces various types of inducedinflammation in mice and rats. We have shown that autoimmune MRL-lpr/1pr mice overexpress

NOS2 and overproduce 'NO in an age-dependent fashion that parallels expression of arthritis,glomerulonephritis, and vasculitis. Blocking 'NO production by oral administration of the NOSinhibitor M\-monomethyl-L-arginine reduced the arthritis, glomerulonephritis, and vasculitis, but itdid not modify serum anti-DNA antibody levels or glomerular deposition of immune complexes.When mice with genetically disrupted NOS2 were backcrossed to MRL-lpr/1pr mice, the resultant(-/-) mice expressed no NOS2 and produced no 'NO, the wild-type (+/+) mice overexpressedNOS2 and overproduced 'NO (in comparison to normal, control mice), and the heterozygous (+/-)mice expressed and produced intermediate levels. Nephritis and arthritis in the (-/-) mice were

comparable to that in MRL-lpr/lpr mice, but vasculitis was markedly decreased. Levels of anti-DNA antibodies were comparable in all mice, but lgG rheumatoid factor production was markedlyreduced in the (-/-) mice. These results of studies in MRL-lpr/lpr mice with genetically disruptedNOS2 highlight the heterogeneity and complexity of the role of NOS2 and 'NO in inflammation.- Environ Health Perspect 106(Suppl 5):1131-1137 (1998). http:llehpnetl.niehs.nih.govldocs/1998/Suppl-5/1 131-1137weinberg/abstract.html

Key words: nitric oxide, superoxide, peroxynitrite, knockout, arthritis, autoimmunity, glomeru-lonephritis, vasculitis, macrophage, monocyte

Overview

Inflammatory joint tissue in rheumatoidarthritis (RA) is characterized by infiltrationand accumulation of mononuclear phago-cytes, lymphocytes, and plasma cells, pro-liferation of synovial cells, and expressionof proinflammatory cytokines (1-3).Although some cytokines are undetectableor are present in only low levels in rheuma-toid synovia, synovial macrophages andfibroblasts in this disease are a good sourceof cytokines such as interferon (IFN)-a,

interleukin (IL)-1, tumor necrosis factor(TNF)-a, IL-6, IL-8, and granulocytemacrophage-colony-stimulating factor(4-7). In addition to these protein media-tors, arachidonic acid metabolites, reactiveoxygen species ([ROS], superoxide anionradical 02-, hydrogen peroxide, andhydroxyl radical), and reactive nitrogenspecies (nitric oxide ['NO], and relatedmolecules such as peroxynitrite [OONO-])likely contribute to the pathology. The

This paper is based on a presentation at the Second International Meeting on Oxygen/Nitrogen Radicals andCellular Injury held 7-10 September 1997 in Durham, North Carolina. Manuscript received at EHP 19 December1997; accepted 23 April 1998.

This work was funded in part by the Veterans Affairs Research Service, The James Swiger HematologyResearch Fund, and National Institutes of Health grant AR-39162.

Address correspondence to J.B. Weinberg, Veterans Affairs and Duke University Medical Centers, 508Fulton Street, Durham, NC 27705. Telephone: (919) 286-6833. Fax: (919) 286-6891. E-mail:brice@acpub.duke.edu

Abbreviations used: COX, cyclooxygenase; eNOS or NOS3, endothelial cell nitric oxide synthase; EPR,electron paramagnetic resonance; Hb, hemoglobin; IFN, interferon; IL, interleukin; iNOS or NOS2,inducible nitric oxide synthase; V-IRE, IFN-y response element; IRE-BP, iron response element-binding pro-tein; ISRE, IFN-a stimulatable response element; KO, knockout; LPS, lipopolysaccharide; Mb, myoglobin;NMMA, M-monomethyl-L-arginine; nNOS or NOS1, neural nitric oxide synthase; 'NO, nitric oxide; NOHb, nitro-sylhemoglobin; NOS, nitric oxide synthase; OA, osteoarthritis; OONO, peroxynitrite; 02-. superoxide anionradical; RA, rheumatoid arthritis; RF, rheumatoid factor; ROS, reactive oxygen species; SOD, superoxide dismu-tase; TfR, transferrin receptor; TGF-,B, transforming growth factor P; TNF, tumor necrosis factor.

combined actions of these mediators(along with certain inherent and inducedanti-inflammatory mediators) contributeto the eventual pathology-accumulationof inflammatory cells, modification of syn-ovial vascular cells, proliferation of syn-ovial fibroblasts, disruption of the generalsynovial architecture, and destruction ofcartilage and bone. NO, 02,- andOONO- appear to be central to theinflammatory process. Antioxidantenzymes such as catalase, superoxide dis-mutase (SOD), and glutathione peroxidasemay be critical as endogenous defensesagainst inflammation. Various cells in thejoint may participate in the inflammation.These include mononuclear phagocytesand chondrocytes.

NO and Nitric OxideSynthaseThe simple gas NO has multiple importantphysiologic and pathologic functions[(8-10) for reviews]. These include roles in(to mention only a few) host resistance totumors and microbes, regulation of bloodpressure and vascular tone, neuro-transmission, learning, and neurotoxicity,carcinogenesis, and control of cellulargrowth and differentiation (11,12). In thepresence of oxygen, NO rapidly (seconds)is converted to nitrite and nitrate, sub-stances that are generally not bioactive[(13) for review]. NO binds with highaffinity to iron in heme groups of proteinssuch as hemoglobin (Hb), myoglobin(Mb), and guanylyl cyclase; Hb and Mbare very effective quenchers of *NO action.NO also reacts with 02-- and SOD pro-longs NO life by eliminating 02---. Onreacting with 02'-, 'NO may formOONO-, a very toxic/reactive moleculethat may be the most important effectortoxic molecule when one thinks of 'NOtoxicity in oxygenated systems.

There are three forms of the enzymenitric oxide synthase (NOS) encoded bythree different genes. Neural NOS (nNOSor NOSI) and endothelial cell NOS(eNOS or NOS3) are constitutiveenzymes, demonstrating low-level, con-stant transcription of mRNA. The enzy-matic actions of NOS1 and NOS3 aremodulated by regulation of cytoplasmiccalcium levels, with agents inducingincreases in calcium (e.g., calciumionophores, ligands such as acetylcholine,or mechanical stress), with subsequentbinding to calmodulin and activation of

Environmental Health Perspectives * Vol 106, Supplement 5 * October 1998 1131

J.B. WEINBERG

the enzyme. Inducible NOS (iNOS orNOS2) can be regulated at multiple levels,but induction of mRNA transcription byagents such as cytokines or lipopolysaccha-ride (LPS) appears to be of major impor-tance [(14) for review]. The activity ofNOS2 is generally thought to be calciumindependent. Although NOS2 wasdescribed initially in mononuclear phago-cytes, it also is found in synoviocytes, chon-drocytes, smooth musde cells, hepatocytes,and others (8,9,15-17).

Regulation of NOS2 can occur atmultiple steps (14), including mRNAtranscription, stability, and translation. Atthe protein level, NOS may be regulatedby calmodulin binding, dimer formation(the functional enzyme exists as a dimer),substrate (L-arginine) depletion, substraterecycling (L-citrulline to L-arginine),tetrahydrobiopterin availability, end prod-uct inhibition ('NO interaction withNOS heme), phosphorylation, and sub-cellular localization. Important NOScofactors include FAD, FMN, NADPH,tetrahydrobiopterin, and calmodulin-cal-cium. For NOS2, calmodulin is tightlybound to protein, making it relativelyresistant to inhibition by calcium chela-tors. Activities of NOS can be influencedby tetrahydrobiopterin levels, andcytokines/LPS can enhance tetrahydrobio-pterin production (18,19). Heme is a crit-ical component of NOS; 'NO can act as afeedback inhibitor of NOS activity bybinding to the iron in heme (20,21).

Human MononuclearPhagocyte NOS2 Expressionand NO ProductionSeveral researchers have shown high-level*NO production by murine macrophages,but many have difficulty showing thathuman monocytes or macrophages pro-duce *NO in vitro [(22) for review].Because of this, some investigators havequestioned the roles of mononuclearphagocyte NOS2 and 'NO in human dis-ease. However, human hepatocytes,chondrocytes, and human DLD- 1 colontumor cells can express NOS2 mRNA andproduce high levels of 'NO after treatmentwith cytokines and LPS (15,16,23-25).Some have shown human monocyte 'NOproduction, but levels have been low whencompared to murine mononuclear phago-cytes [(22,26-28) for review]. In a study ofnormal human monocytes and peritonealmacrophages, we showed that these cellscan produce NOS2 mRNA, protein, and'NO, but levels were much less than that

of mouse macrophages (29). We examineda large array of culture conditions andcytokines. IFN-y and LPS showed onlyslight activity. Human alveolar macro-phages have NOS2 antigen (30,31), andpatients with tuberculosis have increasednumbers of alveolar macrophages thatexpress NOS2 antigen (32). Humans canbe induced to make high levels of *NO invivo via an NOS mechanism (33), but thecells responsible for *NO production invivo are not known. We have recentlynoted that children with mild malariaoverproduce 'NO and have mononuclearcells that express NOS2 antigen (34). Wehave also noted that RA patients haveoverexpression of NOS2 antigen in bloodmononuclear cells, and that their mono-cytes have enhanced *NO production invitro and enhanced responsiveness to treat-ment with IFN-y and LPS (35). Thus,human mononuclear phagocytes canexpress NOS2 and generate 'NO, andhuman NOS2 expression and *NO gener-ation correlate with the severity of certaindiseases, including the autoimmunedisease RA.

NO and Inflammation'NO has many actions appropriate for aproinflammatory agent. It is made bynumerous cell types in sites of inflamma-tion, and it increases blood flow and vascu-lar permeability. *NO has cell/tissuedestructive abilities, and it can induce cyclo-oxygenase (COX), cause pain, destroy cer-tain protease inhibitors, and enhanceproduction of IL-1 and TNF, and NADPHoxidase activity in myeloid cells (11,36).'NO production may result from theactions of several substances including cyto-kines, immune complexes, and bacterialproducts (Figure 1). Because 02-- mayinteract with 'NO to produce OONO-,coincident production of'NO and O2f- setsthe stage for a severe inflammatory state.In the joint, several cell types may pro-duce 'NO-macrophages, chondrocytes,endothelial cells, and possibly others.

Macrophages and chondrocytes are themost likely contributors. Human articularchondrocytes produce relatively high levelsof 'NO, and cytokines can augment thisproduction (15,37,38). Synovial fibroblastsin patients with arthritis also expressNOS2 and produce 'NO (39). Becauseseveral paths might lead to increased 'NOproduction, blocking only one pathway(e.g., blocking an IFN alone) might notfully blunt 'NO production, since alternatepaths could compensate (Figure 1).

Researchers have noted that in additionto proinflammatory effects, 'NO may alsobe anti-inflammatory. The double-edgedsword phenomenon in *NO biology thusapplies to inflammation as well as toother areas in *NO biology (8,40). *NOmodifies adhesiveness and chemotaxis ofpolymorphonuclear neutrophils andmonocytes (41,42), inhibits platelet aggre-gation and secretion (43), and inhibits cellproliferation [including lymphocytes (8)].

Arachidonic acid metabolites playimportant roles in inflammation, and COXinhibitors are drugs useful in the manage-ment of inflammatory disease (44). Thereis significant cross-talk between *NO andCOX. Eicosanoids can reduce NOS2expression and 'NO production (45-47),and *NO modulates prostaglandin E2formation (48). Stimuli that enhanceNOS2 and *NO formation also mayinduce COX2 expression, but the timecourse for induction differs (49-51).Arginine analogues such as NG_monomethyl-L-arginine (NMMA) may beanti-inflammatory by inhibiting bothCOX2 and NOS (52). Furthermore,aspirin, in high doses, inhibits bothcyclooxygenase and NOS2 (53).

'NO is important in animal models ofarthritis that mimic human RA. Theseinclude adjuvant arthritis, collagen-induced arthritis, and spontaneous arthritisin MRL mice [(36,38,54) for reviews].Likewise, 'NO participates in the

Microbial Immune OtherCytokines products complexes activators

Cellular NOS2 activation]X -NOS inhibition

02- NOX -'NO quenching

(OONO-, 'OH, and etc.)

Inflammation

Figure 1. Nitric oxide and inflammation. Cellular NOS2can be activated by a variety of cytokines (e.g., IFN-y,TNF), microbial products (e.g., endotoxin), immunecomplexes, and others. The formed 'NO can by itself,or through interaction with superoxide and other morereactive molecules, initiate and propagate inflamma-tion. NOS2 can be inhibited by various means (e.g.,i-arginine antagonists, tetrahydrobiopterin depletion,L-arginine depletion). 'NO effects can be quenched bymolecules (e.g., hemoglobin, hydroxocobalamin).

Environmental Health Perspectives a Vol 106, Supplement 5 * October 19981132

'NO AND AUTOIMMUNITY

pathogenesis of spontaneous myositis inthe SJL mouse (55). Mice with genet-ically disrupted transforming growth fac-tor P1 (TGF-P) have severe multifocalinflammation, and they die after only 2 to3 weeks of life. These mice have overex-pression of NOS2 and overproduction of'NO (56). The overproduction of NO isinhibited by treatment with NMMA invivo. The absence of TGF-,B, an inhibitorof NOS2 transcription and translation(57), appears to cause a systemic lethalinflammatory condition that results from'NO overproduction.

'NO may be increased in synovial fluidand serum of patients with RA (58-60).Kaur and Halliwell (61) showed increasedlevels of nitrotyrosine (a product resultingfrom OONO- action) in serum and syn-ovial fluid from arthritis patients (61).Sakurai et al. (62) found that synovia frompatients with RA and osteoarthritis (OA)produce 'NO in vitro, and express NOS2mRNA (reverse transcriptase-polymerasechain reaction) and protein (immunoblotand immunohistology). The NOS2 wasassociated primarily with CD14+ cells(mononuclear phagocytes). Other investiga-tors have also found that synovial fibroblasts(as well as macrophages) from patients withRA and OA produced 'NO and expressedNOS2 (39). Ueki and co-workers (63)showed that RA patients have higher serum'NO than do osteoarthritis (OA) patients,that RA synovial fluid 'NO was muchhigher than serum 'NO, and that serum*NO levels correlated significantly with clin-ical parameters of disease activity (durationof morning stiffness, number of swollen andtender joints, and serum levels of C-reactiveprotein, TNF, and IL-6). We had earliernoted comparable associations (35).

The MRL-Ipr/lIpr Model ofAutoimmunityThese mice develop a disease that resembleshuman systemic lupus erythematosus andRA (64). They have autoantibodies,rheumatoid factor, arthritis, nephritis, andvasculitis, and they die prematurely fromdisease. They have lymphadenopathy andsplenomegaly, and a defect in apoptosiscaused by a mutated fas gene (an insert ofthe early transposable element resulting inaberrant splicing and premature termina-tion of transcription) (65). Fas protein is amembrane protein that, after ligation withan antibody or with its ligand FasL, causescell death by apoptosis. C3H-gld mice aresimilar to MRL-/pr/lpr mice in that theyhave generalized lymphadenopathy and a

defect in apoptosis. The gld defect is adefect in the fasL gene, that encodes theligand for Fas. Mice of both strains haveincreased numbers of CD4-, CD8-, andCD3+ cells. These cells are polyclonal andnonmalignant. Fas is essential for the acti-vation-induced death of mature T cells inthe periphery but not in the thymus. InMRL-lpr/lpr mice, failure of apoptosisresults in persistence of autoreactive T cellsthat help autoreactive B cells that are noteliminated. Genetic factors in addition tolpr are also important in determination ofdisease. MRL mice without the fas abnor-mality still develop an autoimmune disor-der (albeit less severe). Our work showedwhile enhanced NOS2 expression and'NO production are critical for diseasedevelopment, disease likely does not resultfrom an abnormality in the NOS2 gene perse (66). Because these mice have highserum and tissue levels of immune com-plexes and various cytokines (includingIFN-y, IL-12, and TNF), activation mayresult from these factors. We and othershave proposed potential final effectors ofdisease; these include arachidonic acidmetabolites, ROS, and 'NO (66-68).Humans with fas gene mutations, defectiveapoptosis, and autoimmune abnormalitieshave been recently described (69,70).

NO, OONO-, Inflammation,and ArthritisWe noted earlier that the macrophages ofMRL-/pr/lpr mice are activated in variousways, including enhanced ROS generation(67). We hypothesized that they mightalso overproduce 'NO (66), and wedemonstrated that they spontaneouslyexcrete 5 to 10 times more urinarynitrite/nitrate (stable oxidation products of*NO) than normal mice. They haveenhanced expression of NOS2 mRNA andprotein in macrophages, liver, kidney, andspleen. NOS2 maps to mouse chromosome1 1 (corresponding to human 17p2 1); thislocation is different from sites previouslydetermined to be linked to disease suscepti-bility [chromosome 19 (fas), 7 and 12],making it unlikely that a defect in theNOS2 gene is the cause of this autoim-mune disorder. Treatment of the micewith NMMA orally blocks the 'NO over-production, and it prevents or blunts thedevelopment of arthritis, nephritis, andvasculitis, indicating that 'NO is causallyrelated to the disease. Despite marked im-provement in renal function and histology,drug treatment does not reduce the serumlevels of anti-DNA antibody or alter

deposition of immune complexes. Otherworkers have corroborated our findings ofoverproduction of NO in these MRL-lpr/Ipr mice and extended the findings toanother strain of mice with autoimmunedisease (i.e., New Zealand white/NewZealand black) (71,72).

'NO binds to Hb tightly to formnitrosylhemoglobin (NOHb). This NOHbcan be accurately detected by electronparamagnetic resonance techniques(EPR). Nonhuman animals in septicshock and humans receiving intravenousnitroglycerin have detectable levels ofNOHb (73). We measured NOHb inwhole blood. MRL-lpr/lpr mice had statis-tically significant elevated levels of NOHbin blood (74). This increased with ageand paralleled the course of 'NO overpro-duction. We also examined the kidneysusing EPR and found that MRL kidneyshave a nitrosyl nonheme iron protein sig-nal, as well as some NOHb (probablyfrom blood trapped in the kidney). Thenitrosyl nonheme iron protein at g= 2.04may be an iron-sulfur cluster protein suchas one of the mitochondrial electrontransport enzymes.

Our studies showed that kidneys fromMRL mice have an increase in the amountof nitrotyrosine-containing proteins.Nitrotyrosine is formed as a consequenceof action of OONO- on tyrosine in pro-teins, and thus may serve as a marker of'NO and OONO- action in tissues(75-77). On immunoblots using specificantinitrotyrosine antibody, extracts fromkidneys from normal mice were essentiallynegative, whereas those from MRL micehad two major bands of immunoreactivity(A,= 60,000 and 48,000) and three minorbands (78). Reactivity was eliminated byomitting the primary antibody or by co-incubating the primary antibody reactionmixture with 10 mM nitrotyrosine. Theidentity of the nitrated proteins in the tis-sues from the diseased kidneys is presentlyunknown. 'NO and OONO- can reactwith numerous different proteins, andthese reactions can alter functions of some(9,10). In an attempt to identify one of thetarget proteins for 'NO and OONO-, wemeasured catalase activity in the mousekidneys. Catalase levels were diminished inMRL mice, and this decrease was pre-vented by treating them with NMMA invivo. In in vitro studies, we noted thatOONO- would inactivate purified catalase.This suggested that catalase is one of thetarget proteins inactivated by 'NO and/orOONO-. All this indicates that catalase

Environmental Health Perspectives * Vol 106, Supplement 5 * October 1998 1133

J.B. WEINBERG

can serve as a target for OONO--mediatedmodification, and that the modified proteinhas decreased activity. Thus, in the MRLmice, there is overproduction of oxidants(NO, 02-, and OONO-) with depletionof the antioxidant enzyme catalase. Thisproduces the setting of extreme oxidantstress. Evidence of increased lipid peroxida-tion and oxidant stress in these mice hasbeen reported (79).

Targeted Disruption of theGenes for NOS2Studies using homologous recombinationbetween incoming DNA and a chosentarget gene (gene targeting) in embryonicstem cells to make planned changes in themouse germ line have allowed productionof desired knockout (KO) mice. One canstudy the effects of the absence of genesand their products. KO models for thethree isoforms of NOS [NOS 1 (80),NOS2 (81-83), and NOS3 (84)] havebeen reported. Mice of each of these KOsdevelop normally, reproduce, and appeargrossly normal. The NOS1 KO mice havegastromegaly due to absence of *NO-releasing neurons in the pylorus of thestomach, and they are overly aggressive.NOS3 KO mice have hypertension (84).NOS2 KO mice do not elevate levels of*NO after immune stimulation or afterLPS injection, and their macrophages can-not make NOS2 protein or 'NO (81,83).They have enhanced susceptibility to dis-seminated infection with Leishmania andListeria monocytogenes, and anesthetizedmice may have increased lethality afterchallenge with high doses of LPS (espe-cially those mice who have previously beeninjected with Corynebacterium parvum).One group has shown that awake,untreated NOS2 KO mice are not resistantto LPS-induced death (82), suggesting thatother factors are operative in this complexmodel of shock. NOS2 KO mice havediminished paw swelling after injection ofcarrageenin, and their lymphocytes showincreased production of Th2 cytokinesafter stimulation in vitro (83). NOS2 KOmice have reduced resistance to infectionwith Leishmania major and Mycobacteriatuberculosis (83,85).

NOS2 Knockout MouseStudies in Autoimmune MiceWe have crossed the NOS2 KO micewith MRL mice, and these are to fourbackcrosses (N4). Because the embryonicstem cells are the 129 strain, the targetedmice are not genetically homogeneous,

but N4 mice have diluted out most of129 strain of the embryonic stem cells andexpress essentially the MRL background.MRL-lpr/lpr littermates homozygous fordisrupted NOS2 [(-I-)], heterozygous fordisrupted NOS2 [(+1-)], or no disruptionof NOS2 [(+/+)] were derived for thisstudy (86).

The (+/+) mice excreted large amountsof nitrite/nitrate, confirming our priorobservations (66). The (+/-) mice excretedintermediate levels, whereas (-/-) miceexcreted very low levels of nitrite/nitrate(comparable to those of normal BALB/cmice). Nitrite/nitrate levels in sera from20-week-old animals paralleled the urinarymeasures, with very low levels in the (-/-)mice, high levels in the (+/+) mice, andintermediate levels in the (+/-) mice.Levels of urinary and serum nitrite/nitratein (+/+) or (+/-) mice were significantlyhigher than those in the (-/-) mice (86).

To assess in vitro 'NO production bycells from mice of the three groups, we cul-tured peritoneal macrophages withoutadditives and with IFN-,y (50 U/ml) andLPS (10 ng/ml). Nitrite/nitrate levels weresignificantly lower in the tissue culturesupernatants of macrophages from the(-/-) mice than (+/+) mice or the (+/-)mice, both at baseline and following stimu-lation. Similarly, NOS2 enzyme activity, asmeasured by the conversion of L-arginineto L-citrulline, was significantly less in thecells from (-/-) mice than those from (+/-)or (+/+) mice. NOS activities were reducedby more than 90% by inclusion of 2 mMNMMA in the reaction mixtures. Thesestudies confirm lack of detectable NOSactivity in (-/-) mice (86).

Immunoblots were performed onprotein extracts from spleens, kidneys,liver, and peritoneal macrophages from themice using an anti-NOS2 antibody. Themacrophage cell line J774 (no treatment ortreated with LPS/IFN-y) served as control.As noted earlier, tissue from BALB/c micehad no NOS2. There was minimal or nodetectable NOS2 in the splenic, kidneyand macrophage protein extracts from(-/-) mice. Extracts from (+/+)and (+/-)mice contained NOS2 protein, with thoseof the (+/-) mice having approximatelyhalf the amount of the (+/+) mice (86).

Based on the effect of in vivoadministration ofNMMA on renal diseaseand arthritis in MRL-Ipr/ipr mice (66), wepredicted that (-/-) mice would developless renal disease and arthritis than mice ofthe other two groups. Twenty-four-hoururinary protein excretion at 20 weeks of

age was less in the (-/-) mice than in theother two groups (but not statisticallysignificantly different), suggesting a possi-ble difference in renal disease. Pathologicexamination, however, indicated that glo-merulonephritis in the (-/-) mice was sim-ilar in severity to that of the other twogroups. Proliferative glomerulonephritiswas present in all mice regardless ofNOS2genotype, with overall glomerular scoressimilar between the groups. Crescentic glo-merulonephritis and interstitial disease waspresent in a small number of mice in eachgroup. Synovitis was present in most of themice, with overall synovial scores similar inthe three groups. Synovial hypertrophy,synovial inflammation, and erosive diseasewere present to a similar degree in theMRL-Ipr/lpr mice regardless of NOS2genotype (86).

Infiltrates of lymphocytes andperivascular collection of lymphocytes occurin the kidneys of all /pr mice regardless ofgenetic background (64,87); true medium-to-large vessel vasculitis is found, however,only in MRL-lpr/ipr mice (87-89). Mono-nuclear cell infiltrates were present in thekidneys of all mice in this study. However,in contrast to glomerulonephritis and arth-ritis, there was a significant difference in thepresence and severity of medium-to-largevessel vasculitis, depending on NOS2 geno-type. Indeed, 4 of 6 (+/+) mice had prom-inent vasculitis ofmedium-to-large vessels inthe kidney, whereas only 1 of 9 of the (+/-)mice and 0 of 9 of (-/-) mice had medium-to-large vessel vasculitis. The incidence ofvasculitis in the (+/+) mice was similar tothat in 20-week-old female MRL-lpr/lprmice (80%). The difference in the occur-rence of vasculitis between the (+/+) miceand the (-/-) mice was statistically signifi-cant. Pathologic examination of the brain,liver, lymph nodes, spleen, and lung re-vealed similar mild lymphocytic infiltrationin all three groups (86).

MRL-lpr/lpr mice are notable forautoantibody production (64,87). Wetherefore determined if MRL-lpr/lpr miceof various NOS2 genotypes displayedqualitative or quantitative differences inautoantibody production. Serum levels ofantibodies to single-strand or double-strand DNA did not differ in the variousmice. These results are comparable to thelack of effect of NMMA treatment onanti-DNA production we noted before inMRL-lpr/lpr mice (66). However, therewas a shift in the isotype of anti-DNAantibodies produced. The IgGl/IgG3 ratioof anti-DNA antibodies was higher in the

Environmental Health Perspectives * Vol 106, Supplement 5 * October 19981134

'NO AND AUTOIMMUNITY

(-/-) mice. In contrast to anti-DNAproduction, rheumatoid factor (RF)production differed among the MRL-lpr/Ilpr mice of various NOS2 genotypes.The (-/-) mice produced significantly lessIgG RF than did (+/+) mice. IgM RF levelswere also lower in the (-/-) mice, but levelswere not significantly different from thosein (+/+) mice. IgG3 RF activity [known tobe associated with small-vessel but notmedium-vessel vasculitis in MRL-lpr/lprmice (90)], was similar in the three groupsof mice. Total serum IgG and IgM weresimilar in the three groups, as were serumlevels of anti-Sm and anti-La antibodies.The different antibody levels in NOS2-dis-rupted mice may highlight the fact thatsmall amounts of NO can modulate B-cellfunction by modifying bcl-2 levels andapoptosis (91).

The mechanism(s) for the contrastingeffects on renal and synovial diseases of agenetically disrupted NOS2 as opposed topharmacological inhibition of NOS activ-ity with NMMA is not clear (Table 1). It isunlikely that there were compensatoryincreases in NOS1 or NOS3 in the (-/-)mice, since we noted very low total bodyNO production in the (-/-) mice.NMMA is an isoform-nonspecific NOS

Table 1. Nitric oxide and autoimmune disease in MRL-/pr/lprmice.a

NO production NOS2 expression Glomerulo-Mouse in vivob in vivoc nephritisd Arthritisd VasculitisdNormal (BALB/c) oe 0 O 0 0MRL-Ipr/lpr 4+ 4+ 3+ 3+ 3+MRL-/pr/lprtreated 0 NDf 0 0 0with NMMA

MRL-/pr/lpr-NOS2 (-I-) 0 0 3+ 3+ 0

aSummarized from Weinberg et al. (66) and Gilkeson et al. (86). bDetermined by urinary nitrate/nitrite excretion orplasma nitrate/nitrite. cDetermined by analyses of peritoneal macrophages, spleen, liver, and kidney tissues formRNA by Northern blot analysis and protein by immunoblot. dDetermined by histologic analyses of tissues.eDegree of positivity, with 0 being none and 4+ being the maximum. 'Not determined.

inhibitor, blocking all three isoforms of theNOS enzymes (92). Inhibiting all NOSisoforms (and hence potentially all NOproduction) with NMMA may be moreeffective in disease prevention than geneti-cally disrupting only NOS2. Alternativeinflammatory pathways may not be activewhen NO production is acutely blockedby NMMA; however, these pathwaysmight become active over time whenNOS2 is genetically disrupted and absentthe entire life of the animal.

ConclusionsThus, MRL-/pr/lpr mice spontaneouslyoverexpress NOS2 and overproduce NO

in parallel with the developmentof autoimmunity and inflammation.Inhibiting 'NO production in vivo by oraladministration ofNMMA from 10 to 20weeks of age prevents development of glo-merulonephritis, arthritis, and vasculitis.This indicates that 'NO is important in thepathogenesis of glomerulonephritis, arthri-tis, and vasculitis in these mice. However,in MRL-/pr/lpr mice with genetically dis-rupted NOS2, arthritis and glomeru-lonephritis are unaltered, whereas vasculitisis reduced. These studies highlight theheterogeneity and complexity of the rolesof NOS2 and 'NO in inflammation inMRL-Ipr/Ipr mice.

REFERENCES AND NOTES

1. Cush JJ, Lipsky PE. Cellular basis for rheumatoid inflammation.Clin Orthop Relat Res 265:9-22 (1991).

2. Harris ED Jr. Rheumatoid arthritis. Pathophysiology and impli-cations for therapy. N EnglJ Med 322:1277-1289 (1990).

3. Weyand CM, Goronzy JJ. Pathogenesis of rheumatoid arthritis.Adv Rheumatol 81:29-55 (1997).

4. Brennan FM, Feldmann M. Cytokines in autoimmunity. CurrOpin Immunol 4:754-759 (1992).

5. Alvaro-Gracia JM, Zvaifler NJ, Brown CB, Kaushansky K,Firestein GS. Cytokines in chronic inflammatory arthritis.VI:Analysis of the synovial cells involved in granulocyte-macrophage colony-stimulating factor production and geneexpression in rheumatoid arthritis and its regulation by IL-1 andtumor necrosis factor-alpha. J Immunol 146:3365-3371(1991).

6. Talal N. Interleukins, interferon and rheumatic disease. ClinRheum Dis 11:633-644 (1985).

7. Hooks JJ, Jordan GW, Cupps T, Moutsopoulos HM, Fauci AS,Notkins AL. Multiple interferons in the circulation of patientswith systemic lupus erythematosus and vasculitis. ArthritisRheum 25:396-400 (1982).

8. Moncada S, Higgs A. The L-arginine-nitric oxide pathway.N Engl J Med 329:2002-2012 (1993).

9. Nathan C. Nitric oxide as a secretory product of mammaliancells. FASEBJ 6:3051-3064 (1992).

10. Beckman JS, Crow JP. Pathological implications of nitric oxide,superoxide and peroxynitrite formation [Review]. Biochem SocTrans 21:330-334 (1993).

11. Magrinat G, Mason SN, Shami PJ, Weinberg JB. Nitric oxidemodulation of human leukemia cell differentiation and geneexpression. Blood 80:1880-1884 (1992).

12. Punjabi CJ, Laskin DL, Heck DE, Laskin JD. Production ofnitric oxide by murine bone marrow cells. Inverse correlationwith cellular proliferation. J Immunol 149:2179 (1992).

13. Stamler JS, Singel DJ, Loscalzo J. Biochemistry of nitric oxideand its redox-activated forms [Review]. Science 258:1898-1902(1992).

14. Nathan C, Xie Q-W. Regulation of biosynthesis of nitric oxide.J Biol Chem 269:13725-13728 (1994).

15. Palmer RM, Hickery MS, Charles IG, Moncada S, Bayliss MT.Induction of nitric oxide synthase in human chondrocytes.Biochem Biophys Res Commun 193:398-405 (1993).

16. Charles IG, Palmer RMJ, Hickery MS, Bayliss MT, Chubb AP,Hall VS, Moss DW, Moncada S. Cloning, characterization, andexpression of a cDNA encoding an inducible nitric oxide syn-thase from the human chondrocyte. Proc Natl Acad Sci USA90:11419-1 1423 (1993).

17. Stefanovic-Racic M, Stadler J, Georgescu HI, Evans CH. Nitricoxide synthesis and its regulation by rabbit synoviocytes.J Rheumatol 21:1892-1898 (1994).

18. Gross SS, Levi R. Tetrahydrobiopterin synthesis. An absoluterequirement for cytokine-induced nitric oxide generation by vas-cuLar smooth musde. J Biol Chem 267:25722-25729 (1992).

19. Rosenkranz-Weiss P, Sessa WC, Milstien S, Kaufman S, WatsonCA, Pober JA. Regulation of nitric oxide synthesis by proinflam-matory cytokines in human umbilical vein endothelial cells:elevations in tetrahydrobiopterin levels enhance endothelial nitricoxide synthase specific activity. J Clin Invest 93:2236-2243(1994).

20. Assreuy J, Cunha FQ, Liew FY, Moncada S. Feedback inhibi-tion of nitric oxide synthase activity by nitric oxide. BrJ Pharmacol 108:833-837 (1993).

Environmental Health Perspectives * Vol 106, Supplement 5 * October 1998 1135

J.B. WEINBERG

21. Rogers NE, Ignarro LJ. Constitutive nitric oxide synthase fromcerebellum is reversibly inhibited by nitric oxide formed fromL-arginine. Biochem Biophys Res Commun 189:242 (1992).

22. Denis M. Human monocytes/macrophages: NO or no NO? JLeukoc Biol 55:682-684 (1994).

23. Curran RD, Billiar TR, Stuehr DJ, Ochoa JB, Harbrecht BG,Flint SG, Simmons RL. Multiple cytokines are required toinduce hepatocyte nitric oxide production and inhibit totalprotein synthesis. Ann Surg 212:462-469 (1990).

24. Sherman PA, Laubach VE, Reep RB, Wood ER. Purificationand cloning of a cytokine-induced nitric oxide synthase from ahuman tumor cell line. Biochemistry 32:11600-11605 (1993).

25. Geller DA, Lowenstein CJ, Shapiro RA, Nussler AK, Di SilvioM, Wang SC, Nakayama DK, Simmons RL, Snyder SH,Billiar TR. Molecular doning and expression of inducible nitricoxide synthase from human hepatocytes. Proc Natl Acad SciUSA 90:3491-3495 (1993).

26. Demaria N, Colantoni A, Fagiuoli S, Liu GJ, Rogers BK,Farinati F, Vanthiel DH, Floyd RA. Association between reac-tive oxygen species and disease activity in chronic hepatitis C.Free Radic Biol Med 21:291-295 (1996).

27. Denis M. Tumor necrosis factor and granulocyte macrophage-colony stimulating factor stimulate Euman macrophages torestrict growth of virulent Mycobacterium avium and to killavirulent M aviunr. killing effector mechanism depends on thegeneration of reactive nitrogen intermediates. J Leukoc Biol49:380-387 (1991).

28. Paul-Eugene N, Mossalayi D, Sarfati M, Yamaoka K, AubryJP, Bonnefoy JY, Dugas B, Kolb JP. Evidence for a role of Fcepsilon RII/CD23 in the IL-4-induced nitric oxide productionby normal human mononuclear phagocytes. Cel Ilmmunol163:314-318 (1995).

29. Weinberg JB, Misukonis MA, Shami PJ, Mason SN, Sauls DL,Dittman WA, Wood ER, Smith GK, McDonald B, BachusKE, et al. Human mononuclear phagocyte inducible nitricoxide synthase (iNOS). Analysis of iNOS mRNA, iNOS pro-tein, biopterin, and nitric oxide production by blood mono-cytes and peritoneal macrophages. Blood 86:1184-1195(1995).

30. Kobzik L, Bredt DS, Lowenstein CJ, Drazen J, Gaston B,Sugarbaker D, Stamler JS. Nitric oxide synthase in human andrat lung: immunocytochemical and histochemical localization.Am J Respir Cell Mol Biol 9:371-377 (1993).

31. Tracey WR, Xue C, Klinghofer V, Barlow J, Pollock JS,Forstermann U, Johns RA. Immunochemical detection ofinducible NO synthase in human lung. Am J Physiol266:L722-727 (1994).

32. Nicholson S, Bonecinialmeida MDG, Silva LE Jr., Nathan C,Xie QW, Mumford R, Weidner JR, Calaycay J, Geng J,Boechat N, et al. Inducible nitric oxide synthase in pulmonaryalveolar macrophages from patients with tuberculosis. J ExpMed 183:2293-2302 (1996).

33. Hibbs JB Jr., Westenfelder C, Taintor R, Vavrin Z, Kablitz C,Baranowski RL, Ward JH, Menlove RL, McMurry MP,Kushner JP, Samlowski WE. Evidence for cytokine-induciblenitric oxide synthesis from 1-arginine in patients receiving inter-leukin-2 therapy. [Published erratum appears in J Clin Invest90(1):295 (1992)]. J Clin Invest 89:867-877 (1992).

34. Anstey NM, Weinberg JB, Hassanali M, Mwaikambo ED,Manyenga D, Misukonis MA, Arnelle DR, Hollis D,McDonald MI, Granger DL. Nitric oxide in Tanzanian chil-dren with malaria. Inverse relationship between malaria severityand nitric oxide production/nitric oxide synthase type 2 expres-sion. J Exp Med 184:557-567 (1996).

35. St. Clair EW, Wilkinson WE, Lang T, Sanders L, MisukonisMA, Gilkeson GS, Pisetsky DS, Granger DL, Weinberg JB.Increased expression of blood mononuclear cell nitric oxidesynthase type 2 in rheumatoid arthritis patients. J Exp Med184:1173-1178 (1996).

36. Clancy RM, Abramson SB. Nitric oxide-a novel mediator ofinflammation [Review]. Proc Soc Exp Biol Med 210:93-101(1995)-

37. Stadler J, Stefanovic-Racic M, Billiar TR, Curran RD,McIntyre LA, Georgescu HI, Simmons RL, Evans CH.Articular chondrocytes synthesize nitric oxide in response tocytokines and lipopolysaccharide. J Immunol 147:3915-3920(1991).

38. Evans CH, Stefanovic-Racic M, Lancaster JR Jr. Nitric oxideand its role in orthopaedic disease [Review]. Clin OrthopRelat Res 312:275-294 (1995).

39. Mclnnes LB, Leung BP, Wei XQ, Huang F-P, Sturrock RD,Kinninmonth A, Weidner J, Mumford R, Liew FY. Field M,Production of nitric oxide in the synovial membrane ofrheumatoid and osteoarthritis patients. J Exp Med184:1519-1524 (1996).

40. Clancy RM, Leszczynska-Piziak J, Abramson SB. Nitric oxide,an endothelial cell relaxation factor, inhibits neutrophil super-oxide anion production via a direct action on the NADPHoxidase. J Clin Invest 90:1116-1121 (1992).

41. Darius H, Grodzinska L, Meyer J. The effects of the nitricoxide donors molsidomine and SIN-1 on human polymor-phonuclear leucocyte function in vitro and ex vivo. Eur J ClinPharmacol 43:629-633 (1992).

42. Kubes P, Suzuki M, Granger DN. Nitric oxide: an endoge-nous modulator of leukocyte adhesion. Proc NatI Acad SciUSA 88:4651-4655 (1991).

43. Stamler JS, Loscalzo J. The antiplatelet effects of organicnitrates and related nitroso compounds in vitro and in vivoand their relevance to cardiovascular disorders [seeComments]. J Am Coll Cardiol 18:1529-1536 (1991).

44. Harris ED Jr. Textbook of Rheumatology. 4th ed.Philadelphia:W.B. Saunders 1993.

45. Gaillard T, Mulsch A, Klein H, Decker K. Regulation byprostaglandin E2 of cytokine-elicited nitric oxide synthesis inrat liver macrophages. Biol Chem Hoppe-Seyler 373:897-902(1992).

46. Marotta P, Sautebin L, Di Rosa M. Modulation of the induc-tion of nitric oxide synthase by eicosanoids in the murinemacrophage cell line J774. Br J Pharmacol 107:640-641(1992).

47. Imai Y, Kolb H, Burkart V. Nitric oxide production frommacrophages is regulated by arachidonic acid metabolites.Biochem Biophys Res Commun 197:105-109 (1993).

48. Stadler J, Harbrecht BG, Di Silvio M, Curran RD, JordanML, Simmons RL, Billiar TR. Endogenous nitric oxideinhibits the synthesis of cyclooxygenase products and inter-leukin-6 by rat Kupffer cells. J Leukoc Biol 53:165-172(1993).

49. Salvemini D, Misko TP, Masferrer JL, Seibert K, Currie MG,Needleman P. Nitric oxide activates cyclooxygenase enzymes.Proc Natl Acad Sci USA 90:7240-7244 (1993).

50. Swierkosz TA, Mitchell JA, Warner TD, Botting RM, VaneJR. Co-induction of nitric oxide synthase and cyclo-oxygenase:interactions between nitric oxide and prostanoids. Br JPharmacol 114:1335-1342 (1995).

51. Vane JR, Mitchell JA, Appleton I, Tomlinson A, Bishop-Bailey D, Croxtall J, Willoughby DA. Inducible isoforms ofcyclooxygenase and nitric-oxide synthase in inflammation.Proc Natl Acad Sci USA 91:2046-2050 (1994).

52. Salvemini D, Manning PT, Zweifel BS, Seibert K, Connor J,Currie MG, Needleman P, Masferrer JL. Dual inhibition ofnitric oxide and prostaglandin production contributes to theanti-inflammatory properties of nitric oxide synthaseinhibitors. J Clin Invest 96:301-308 (1995).

53. Amin AR, Vyas P, Attur M, Leszczynska-Piziak J, Patel IR,Weissmann G, Abramson SB. The mode of action of aspirin-like drugs: effect on inducible nitric oxide synthase. Proc NatlAcad Sci USA 92:7926-7930 (1995).

54. Stefanovic-Racic M, Stadler J, Evans CH. Nitric oxide andarthritis. Arthritis Rheum 36:1036-1044 (1993).

55. Tamir S, Derojaswalker T, Gal A, Weller AH, Li XT, Fox JG,Wogan GN, Tannenbaum SR. Nitric oxide production inrelation to spontaneous B-cell lymphoma and myositis in SJLmice. Cancer Res 55:4391-4397 (1995).

1136 Environmental Health Perspectives * Vol 106, Supplement 5 * October 1998

NO AND AUTOIMMUNITY

56. Vodovotz Y, Geiser AG, Chesler L, Letterio J, Campbell A,Lucia MS, Sporn MB, Roberts AB. Spontaneously increased pro-duction of nitric oxide and aberrant expression of the induciblenitric oxide synthase in vivo in the transforming growth factorbeta-1 null mouse. J Exp Med 183:2337-2342 (1996).

57. Vodovotz Y, Bogdan C, Paik J, Xie QW, Nathan C. Mechanismsof suppression of macrophage nitric oxide release by transforminggrowth factor beta. J Exp Med 178:605-613 (1993).

58. Farrell AJ, Blake DR, Palmer RM, Moncada S. Increased con-centrations of nitrite in synovial fluid and serum samples suggestincreased nitric oxide synthesis in rheumatic diseases. AnnRheum Dis 51:1219-1222 (1992).

59. Stichtenoth DO, Fauler J, Zeidler H, Frolich JC. Urinary nitrateexcretion is increased in patients with rheumatoid arthritis andreduced by prednisolone. Ann Rheum Dis 54:820-824 (1995).

60. Grabowski PS, England AJ, Dykhuizen R, Copland M, BenjaminN, Reid DM, Ralston SH. Elevated nitric oxide production inrheumatoid arthritis-detection using the fasting urinarynitrate/creatinine ratio. Arthritis Rheum 39:643-647 (1996).

61. Kaur H, Halliwell B. Evidence for nitric oxide-mediated oxida-tive damage in chronic inflammation-nitrotyrosine in serumand synovial fluid from rheumatoid patients. FEBS Lett350:9-12 (1994).

62. Sakurai H, Kohsaka H, Liu MF, Higashiyama H, Hirata Y,Kanno K, Saito I, Miyasaka N. Nitric oxide production andinducible nitric oxide synthase expression in inflammatory arthri-tides. J Clin Invest 96:2357-2363 (1995).

63. Ueki Y, Miyake S, Tominaga Y, Eguchi K. Increased nitric oxidelevels in patients with rheumatoid arthritis. J Rheumatol23:230-236 (1996).

64. Cohen PL, Eisenberg RA. Lpr and gIa single gene models of sys-temic autoimmunity and lymphoproliferative disease. Ann RevImmunol 9:243-269 (1991).

65. Nagata S, Suda T. Fas and Fas ligand: Ipr and gld mutations.Immunol Today 16:39-43 (1995).

66. Weinberg JB, Granger DL, Pisetsky DS, Seldin MF, MisukonisMA, Mason SN, Pippen AM, Ruiz P, Wood ER, Gilkeson GS.The role of nitric oxide in the pathogenesis of spontaneousmurine autoimmune disease: increased nitric oxide productionand nitric oxide synthase expression in MRL-1pr/Ipr mice, andreduction of spontaneous glomerulonephritis and arthritis byorally administered NG-monomethyl-L-arginine. J Exp Med179:651-660 (1994).

67. Dang-Vu AP, Pisetsky DS, Weinberg JB. Functional alterationsof macrophages in autoimmune MRL-lpr/lpr mice. J Immunol138:1757-1761 (1987).

68. Spurney RF, Fan PY, Ruiz P, Sanfilippo F, Pisetsky DS, CoffmanTM. Thromboxane receptor blockade reduces renal injury inmurine lupus nephritis. Kidney Internat 41:973-982 (1992).

69. Fisher GH, Rosenberg FJ, Straus SE, Dale JK, Middleton LA,Lin AY, Strober W, Lenardo MJ, Puck JM. Dominant interfer-ing Fas gene mutations impair apoptosis in a human autoim-mune lymphoproliferative syndrome. Cell 81:935-946 (1995).

70. Cantin R, Fortin JF, Lamontagne G, Tremblay M. The acquisi-tion of host-derived major histocompatibility complex class IIglycoproteins by human immunodeficiency virus type 1 acceler-ates the process of virus entry and infection in human T-lym-phoid cells. Blood 90:1091-1100 (1997).

71. Huang FP, Feng GJ, Lindop G, Stott DI, Liew FY. The role ofinterleukin 12 and nitric oxide in the development of sponta-neous autoimmune disease in MRL:MP-lpr:lpr mice. J Exp Med183:1447-1459 (1996).

72. Oates JC, Ruiz P, Alexander A, Pippen AMM, Gilkeson GS.Effect of late modulation of nitric oxide production on murinelupus. Clin Immunol Immunopath 83:86-92 (1997).

73. Henry Y, Lepoivre M, Drapier JC, Ducrocq C, Boucher JL,Guissani A. EPR characterization of molecular targets for NO inmammalian cells and organelles. FASEB J 7:1124-1134 (1993).

74. Weinberg JB, Gilkeson GS, Mason RP, Chamulitrat W.Nitrosylation of blood hemoglobin and renal nonheme proteinsin autoimmune MRL-lpr/lpr mice. Free Radic Biol Med24:191-196 (1998).

75. Beckman JS, Ye YZ, Anderson PG, Chen J, Accavitti MA,Tarpey MM, White CR. Extensive nitration of protein tyrosinesin human atherosclerosis detected by immunohistochemistry.Biol Chem Hoppe-Seyler 375:81-88 (1994).

76. Kooy NW, Royall JA, Ye YZ, Kelly DR, Beckman JS. Evidencefor in vivo peroxynitrite production in human acute lung injury.Am J Respir Crit Care Med 151:1250-1254 (1995).

77. Haddad IY, Pataki G, Hu P, Galliani C, Beckman JS, MatalonS. Quantitation of nitrotyrosine levels in lung sections ofpatients and animals with acute lung injury. J Clin Invest94:2407-2413 (1994).

78. Privalle CT, Keng T, Gilkeson GS, Weinberg JB. The role ofnitric oxide and peroxynitrite in the pathogenesis of sponta-neous murine autoimmune disease. In: The Biology of NitricOxide (Stamler J, Gross SS, Moncada S, eds). Lon[on:PortlandPress, 1996;20.

79. Venkatraman JT, Chandrasekar B, Kim JD, Fernandes G.Genotype effects on the antioxidant enzymes activity andmRNA expression in liver and kidney tissues of autoimmune-prone MRL/MpJ-lpr/lpr mice. Biochim Biophys Acta1213:167-175 (1994).

80. Huang PL, Dawson TM, Bredt DS, Snyder SH, Fishman MC.Targeted disruption of the neuronal nitric oxide synthase gene.Cel[75:1273-1286 (1993).

81. Macmicking JD, Nathan C, Hom G, Chartrain N, Fletcher DS,Trumbauer M, Stevens K, Xie QW, Sokol K, Hutchinson N,Chen H, Mudgett JS. Altered responses to bacterial infectionand endotoxic shock in mice lacking inducible nitric oxide syn-thase. Cell 81:641-650 (1995).

82. Laubach VE, Shesely EG, Smithies 0, Sherman PA. Mice lack-ing inducible nitric oxide synthase are not resistant tolipopolysaccharide-induced death. Proc Natl Acad Sci USA92:10688-10692 (1995).

83. Wei XQ, Charles IG, Smith A, Ure J, Feng CJ, Huang FP, XuDM, Muller W, Moncada S, Liew FY. Altered immuneresponses in mice lacking inducible nitric oxide synthase.Nature 375:408-411 (1995).

84. Huang PL, Huang ZH, Mashimo H, Bloch KD, MoskowitzMA, Bevan JA, Fishman MC. Hypertension in mice lacking thegene for endothelial nitric oxide synthase. Nature 377:239-242(1995).

85. Macmicking JD, North RJ, Lacourse R, Mudgett JS, Shah SK,Nathan CF. Identification of nitric oxide synthase as a protec-tive locus against tuberculosis. Proc Natl Acad Sci USA94:5243-5248 (1997).

86. Gilkeson GS, Mudgett JS, Seldin MF, Ruiz P, Alexander AA,Misukonis MA, Pisetsky DS, Weinberg JB. Clinical and sero-logic manifestations of autoimmune disease In MRL-lpr/lprmice lacking nitric oxide synthase type 2. J Exp Med186:365-373 (1997).

87. Hang L, Theofilopoulos AN, Dixon FJ. A spontaneous rheuma-toid arthritis-like disease in MRL-l mice. J Exp Med155:1690-1701 (1982).

88. Nose M, Nishimura M, Kyogoku M. Analysis of granulomatousarteritis in MRL/Mp autoimmune disease mice bearing lympho-proliferative genes. The use of mouse genetics to dissociate thedevelopment of arteritis and glomerulonephritis. Am J Pathol135:271-280 (1989).

89. Tarkowski A, Jonsson R, Sanchez R, Klareskog L, KoopmanWJ. Features of renal vasculitis in autoimmune MRL lpr//prmice: phenotypes and functional properties of infiltrating ce[ls.Clin Exp Immunol 72:91-97 (1988).

90. Takahashi 5, Nose M, Sasaki J, Yamamoto T, Kyogoku M.IgG3 production in MRLIIpr mice is responsible for develop-ment of lupus nephritis. J Immunol 147:515-519 (1991).

91. Genaro AM, Hortelano S, Alvarez A, Martineza C, Bosca L.Splenic B lymphocyte programmed cell death is prevented bynitric oxide release through mechanisms involving sustainedBcl-2 levels. J Clin Invest 95:1884-1890 (1995).

92. Southan GJ, Szabo C. Selective pharmacological inhibition ofdistinct nitric oxide synthase isoforms. Biochem Pharmacol51:383-394 (1996).

Environmental Health Perspectives * Vol 106, Supplement 5 * October 1998 1137