THE JOURNAL Q 1992 by The American Society for Biocbemistq

OF BIOLOGICAL CHEMISTRY ’ and Molecular Biology , Inc

Vol. 267, No. 36, Issue of December 25, pp. 25722-25729,1992 Printed in U.S.A.

Tetrahydrobiopterin Synthesis AN ABSOLUTE REQUIREMENT FOR CYTOKINE-INDUCED NITRIC OXIDE GENERATION BY VASCULAR SMOOTH MUSCLE*

(Received for publication, June 2, 1992)

Steven S. Gross$ and Roberto Levi From the Department of Pharmacology, Cornell University Medical College, New York, New York 10021

Nitric oxide (NO) synthesis is induced in vascular smooth muscle cells by lipopolysaccharide (LPS) where i t appears to mediate a variety of vascular dysfunc- tions. In some cell types tetrahydrobiopterin (BH4) synthesis has also been found to be induced by cyto- kines. Because BH4 is a cofactor for NO synthase, we investigated whether BH4 synthesis is required for LPS-induced NO production in rat aortic smooth mus- cle cells (RASMC). The total biopterin content (BH4 and more oxidized states) of untreated RASMC was below our limit of detection. However, treatment with LPS caused a significant rise in biopterin levels and an induction of NO synthesis; both effects of LPS were markedly potentiated by interferon-y. 2,4-Diamino-6- hydroxypyrimidine (DAHP), a selective inhibitor of GTP cyclohydrolase I, the rate-limiting enzyme for de novo BH4 synthesis, completely abolished the elevated biopterin levels induced by LPS. DAHP also caused a concentration-dependent inhibition of LPS-induced NO synthesis. Inhibition of NO synthesis by DAHP was reversed by sepiapterin, an agent which circumvents the inhibition of biopterin synthesis by DAHP by serv- ing as a substrate for BH4 synthesis via the pterin salvage pathway. The reversal by sepiapterin was overcome by methotrexate, an inhibitor of the pterin salvage pathway. Sepiapterin, and to a lesser extent BH4, dose-dependently enhanced LPS-induced NO synthesis, indicating that BH4 concentration limits the rate of NO production by LPS-activated RASMC. Se- piapterin also caused LPS-induced NO synthesis to appear with an abbreviated lag period phase, suggest- ing that BH4 availability also limits the onset of NO synthesis. In contrast to the stimulation of LPS-in- duced NO synthesis, observed when sepiapterin was given alone, sepiapterin became a potent inhibitor of NO synthesis in the presence of methotrexate. This is attributable to a direct inhibitory action of sepiapterin on GTP cyclohydrolase I, an activity which is only revealed after blocking the metabolism of sepiapterin to BH4. Further studies with sepiapterin, methotrex- ate, and N-acetylserotonin (an inhibitor of the BH4

* This work was supported by Grants HL46403 (to S. S. G.) and HL34215 (to R. L.) from the National Institutes of Health. Results presented herein have been previously published in abstract form in the proceedings of the 1992 FASEB meeting held in Anaheim, CA and the Second International Symposium on Endothelium-derived Vasoactive Factors held in Basel, Switzerland. The costs of publica- tion of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertise- ment” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ T o whom correspondence should be addressed The William Harvey Research Inst., St. Bartholomew’s Medical College, Charter- house Square, London EC1 M 6BQ, Great Britain. Tel.: 071-982- 6083: Fax: 071-251-1685.

synthetic enzyme, sepiapterin reductase) indicated that the BH4 is synthesized in RASMC predominantly from GTP; however, a lesser amount may derive from pterin salvage. We demonstrate that BH4 synthesis is an absolute requirement for induction of NO synthesis by LPS in vascular smooth muscle. Our findings also suggest that pterin synthesis inhibitors may be useful for the therapy of endotoxin- and cytokine-induced shock.

Nitric oxide (NO)’ is a potent endogenous vasodilator which appears to be identical to the endothelium-derived relaxing factor originally described by Furchgott in 1980 (1-3). NO plays a pivotal role in the regulation of vascular tone and blood pressure (4,5). In addition to its physiological function, altered rates of NO synthesis are likely to be involved in a variety of vascular pathophysiological conditions. Indeed, overproduction of NO has been implicated as the basis for hypotension caused by bacterial endotoxin (LPS; 6, 7) and the cytokines: tumor necrosis factor (8), IL-1 (9), and IL-2 (10). While the source of vasoregulatory NO in normal phys- iology appears to be exclusively endothelial cells, NO synthe- sis can also be induced by LPS in endothelial cells (11, 12) and other vascular cell types, including smooth muscle (13- 16). Preliminary studies indicate that vascular smooth muscle cells may be the dominant site of NO overproduction in LPS- and cytokine-induced vascular shock (16). Thus, an under- standing of the biochemical events involved in NO synthesis induction in smooth muscle and elucidation of control points for induction and expression of this pathway should provide important new insights leading to more effective therapy of cytokine-mediated shock.

NO is synthesized from L-arginine, 02, and NADPH by an FAD- and FMN-containing enzyme, nitric oxide synthase (NOS; EC 1.14.23.-) (17, 18). This poorly understood bio- chemical reaction constitutes a five-electron oxidation of one of the equivalent guanidino-nitrogens of L-arginine (19). Con- stitutive and inducible isotypes of NOS have been distin- guished on the basis of calcium/calmodulin dependence (17, 20), inhibitor specificity (12, 21), immunogenicity (22), and substrate/cofactor requirements (23, 24). A previous distinc- tion between constitutive and inducible isotypes of NOS was

The abbreviations used are: NO, nitric oxide; NOS, nitric oxide synthase; LPS, lipopolysaccharide; IFN, interferon-y; IL-1, interleu- kin-la; BH4,6-(~-erythro-l, 2-dihydroxypropyl)-5,6,7,8-tetrahydrop- terin; BH2, 6-(~-erythro-l, 2-dihydroxypropyl)-7,8-dihydropterin; DAHP, 2,4-diamino-6-hydroxypyrimidine; MTX, methotrexate (am- ethopterin); SEP, sepiapterin; MTT, 3-(4,5-dimethylthiazol-2-y1)- 2,5-diphenyltetrazolium bromide; NMA, Nu-methyl-L-arginine; SMC, smooth muscle cell; RASMC, rat aortic smooth muscle cells; ACT D, actinomycin D; Mb, myoglobin; NAS, N-acetylserotonin.

25722

Tetrahydrobiopterin Synthesis and Nitric Oxide Generation 25723

a requirement of the inducible isotype for tetrahydrobiopterin (BH4,6-(~-erythro-1,2-dihydroxypropyl)-5,6,7,8-tetrahydrop- terin) (25, 26), a characteristic which appeared not to be shared by the constitutive enzyme (20). However, this differ- ence is no longer accepted in view of more recent findings that constitutive NOS from brain does contain a tightly bound reduced pterin cofactor and, furthermore, that exogenous BH4 potentiates brain NOS activity (27, 28). NOS is the most recent addition to a group of only four other enzymes (29) known to utilize a biopterin cofactor.

BH4 synthesis occurs via two distinct pathways: a de nouo synthetic pathway which uses GTP as a precursor and a salvage pathway for preexisting dihydropterins (see Fig. 1 and Ref. 29). GTP cyclohydrolase I (EC 3.5.4.16) is the first and rate-limiting enzyme for the de nouo pathway leading to synthesis of dihydroneopterin triphosphate. GTP cyclohydro- lase I is inhibited by reduced pterins (30, 31) and can be induced by interferon-y (IFN) in a variety of cell types in- cluding macrophages, lymphocytes, and fibroblasts (32, 33). Subsequent metabolism of dihydroneopterin triphosphate to BH4 occurs via tetrahydropterin intermediates, which have not all been unequivocally identified; required enzymes in- clude 6-pyruvoyltetrahydropterin synthase and sepiapterin reductase. While the dihydropterins, BH2 and sepiapterin, cannot be utilized by the de nouo synthetic pathway for BH4 synthesis, they are converted to BH4 via the pterin salvage pathway which utilizes dihydrofolate reductase (EC 1.5.1.3) for its final step (29).

BH4 oxidation to the quinonoid isoform of BH2 (Q-BH2) is thought to provide two of the five electrons necessary for the oxidation of arginine to NO and citrulline (25). By analogy with other biopterin-dependent enzymes, Q-BH2 would be expected to recycle to BH4 via the NAD(P)H-requiring en- zyme dihydropteridine reductase (EC 1.6.99.7) (29). BH4 re- cycling would thus provide for sustained NO synthesis with catalytic rather than substrate quantities of BH4.

The initial reports that NOS requires a biopterin cofactor were based on findings that BH4 stimulates NOS activity in partially purified enzyme preparations from cytosol of cyto- kine-activated macrophages (25, 26). BH4 levels did not ap-

cGMP -GTP 0

,/,,O,'d J GTP cyclohydrolase I

,' Dihydroneopterin , triphosphate f . J/ 6-pyruvoyl tetrahydropterin

8' NO* + citrulline ARG I

synthase

r"-4 Sepiapterin reductase

Q-BH2 BH4 \ v i h y d r o f o l a t e reductase

Dihydropteridine BH2 reductase 2

Sepiapterin FIG. 1. Enzymatic pathways for the synthesis of tetrahy-

drobiopterin (BH4) in eukaryotic cells and the proposed re- lationship to nitric oxide synthase. GTP is indicated to play a dual role in the biology of NO; it is both precursor to BH4 (a required cofactor of NOS) and precursor to cGMP (the mediator of vascular actions of NO). The dashed arrow indicates that L-arginine (ARG)- derived NO activates soluble guanylyl cyclase, thereby triggering the conversion of GTP to cGMP.

pear to limit or regulate NOS activity, however, because the concentration of BH4 in cytosol from nonactivated macro- phages was sufficient for maximal NOS activity (25). In any event, IFN, which markedly potentiates the induction of NO synthesis by LPS in macrophages and fibroblasts (191, also induces GTP cyclohydrolase I in these cell types (32).

Activation of the BH4 synthetic pathway in man is indi- cated by increased levels of plasma neopterin; the extent of activation has been useful in clinical diagnosis and prediction of therapeutic outcome of a variety of pathophysiological conditions (34). Although activation of the BH4 synthetic pathway is considered to be one of the best indicators of immune cell activation in man, and is of diagnostic and prognostic value (34), its function in immunoactivated cells is obscure. The recent addition of NOS to the list of known biopterin-dependent enzymes raises the possibility that the function of BH4 in immunoactivated cells is to support co- induced NOS activity. This view is consistent with findings that inhibition of BH4 synthesis blocks immunostimulant- induced NO production by fibroblasts and endothelial cells (12,47).

The present investigation addresses the role of BH4 syn- thesis in LPS-activated vascular smooth muscle cells. Using selective inhibitors of various steps in each of the two known pathways of BH4 synthesis, we demonstrate that induction of vascular smooth muscle cell NOS is strictly dependent on synthesis of BH4. Smooth muscle cell BH4 arises mainly from GTP via the de nouo synthetic pathway and, to a lesser extent, via pterin salvage; simultaneous inhibition of both pathways completely prevents the induction of NOS activity by LPS.

MATERIALS AND METHODS

Smooth Muscle Cell (SMC) Culture-Aortic SMC were cultured by explanting segments of the medial layer of aortae from adult male Fischer 344 rats. Aortae were removed aseptically and freed of adven- titial and endothelial cells by scraping both the lumenal and ablu- minal surfaces. Medial fragments (1-2 mm) were allowed to attach to dry Primaria 25-cm2 tissue culture flasks (Falcon, Oxnard, CA) which were kept moist with growth medium until cells emerged. Cultures were fed twice weekly with Medium 199 containing 10% fetal bovine serum, 25 mM HEPES, 2 mM L-glutamine, 40 pg/ml endothelial cell growth supplement (Biomedical Technologies, Stoughton, MA), and 10 pg/ml gentamycin (GIBCO). When primary cultures became confluent, they were passaged by trypsinization and the explants were discarded. For these studies, cells in passage 10-15 were seeded a t 20,00O/well in 96-well plates and were used a t conflu- ence. The cells exhibited a classical SMC phenotype, with hill and valley morphology, and stained positively for smooth muscle a-actin (35). Cell culture medium and reagents, unless otherwise noted, were from Whittaker Laboratories (Walkersville, MD).

Nitrite Assay-Confluent SMC in 96-well plates (density of 60-80 X 103/well) were incubated a t 37 "C in a humidified incubator in 200 pl of RPMI 1640 containing 10% bovine calf serum, 2.5 mM glutamine, and penicillin (80 units/ml), streptomycin (80 pg/ml), fungizone (2 pg/ml), and the desired agents for the indicated times. In studies which assessed the effect of drugs on induced nitrite production, drugs were added simultaneously with the inducing agents. Nitrite accumulation was measured after a 16-40-h incubation period. Cell viability, as assessed by measuring cell respiration with MTT [3-(4, 5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] was not significantly diminished by the highest concentrations of drugs tested: 4,5-diamino-6-hydroxypyridine (3 mM), sepiapterin (300 p ~ ) , meth- otrexate (30 & I ) , and tetrahydrobiopterin (100 p ~ ) . Nitrite was measured by addition of 100 p1 of cell culture medium to 100 pI of Greiss reagent (0.5% sulfanilamide and 0.05% naphthylethylenedi- amine dihydrochioride in 2.5% phosphoric acid), and A,,, was im- mediately measured using a microplate reader (Molecular Devices, Menlo Park, CA). Nitrite concentrations were determined by com- parison with standard solutions of sodium nitrite prepared in culture medium. Background nitrite levels in smooth muscle cell cultures not exposed to cytokines were subtracted from experimental values.

Biopterin Assay-Total cellular biopterin (biopterin plus BH2 and

25724 Tetrahydrobiopterin Synthesis and Nitric Oxide Generation BH4)) was measured after acidic oxidation of reduced forms of biopterin with iodine as previously described (36). Briefly, confluent monolayers of SMC in 75-cm2 tissue culture flasks were washed twice with 10 ml of HBS (150 mM NaCI, 10 mM HEPES g/liter, 40 mM KCl, 1.5 mM CaC12, 0.5 mM Mg2S04, 10 mM glucose, pH 7.5) and suspended with a Teflon scraper (Costar, Cambridge MA) in an additional 5-ml volume of HBS. Cells were pelleted by centrifugation at 500 X g for 10 min and lysed in dHzO by three cycles of freeze- thaw (freezing in liquid nitrogen, thawing in a 37 "C water bath). Lysates were centrifuged at 4 "C for 5 min at 12,000 rpm. Biopterin in the supernatant was oxidized by treatment with 1% I, containing 2% KI in 1 N HCI for 1 h at 37 "C in the dark. After centrifugation for 5 min at 12,000 rpm, supernatants were treated with ascorbate (0.1 M) to remove residual free 12, and extracts were then neutralized with 1 N NaOH followed by 200 mM TRIS, pH 7.8. Biopterin was quantitated by Cls reverse phase HPLC using fluoresence detection and authentic biopterin as a standard (36).

Smooth Muscle Cell Respiration Assay-Rat aortic smooth muscle cells in 96-well microtiter plates were incubated at 37 "C in complete culture medium (RPMI 1640 containing 10% bovine calf serum, 2.5 mM glutamine, and 80 units/ml penicillin, 80 pg/ml streptomycin, and 2 pg/ml fungizone) supplemented with 0.2 mg/ml MTT. After 90 min, culture medium was removed by suction and cells were solubi- lized in 100 p l of dimethyl sulfoxide. The extent of reduction of MTT to formazan within cells was quantitated by measurement of A650 and taken as an indicator of cellular respiration.

Smooth Muscle Cell Cytosol Preparation-Confluent rat aortic smooth muscle cells in 75-cm2 culture flasks were washed twice with 5 ml of HBS. Cells were then removed with a Teflon scraper in 5 ml of HBS and pelleted at 500 X g for 10 min. The supernatant was removed and the cell pellet was lysed by three cycles of freeze-thaw (freezing in liquid nitrogen, thawing in a 37 'C water bath) in 1 ml per 10 culture flasks of a protease inhibitor mixture containing: 1 mM dithiothreitol, 0.1 mM phenylmethylsulfonyl fluoride, 5 pg/ml apro- tinin, 5 pg/ml pepstatin A, and 5 pg/ml chymostatin. Cell cytosols were prepared from lysates by centrifugation at 100,000 X g for 1 h at 4 "C. Cytosols were immediately aliquoted and frozen at -70 "C until used.

NO' Synthase Assay-Nitric oxide formation by smooth muscle cell cytosol was measured by a previously described kinetic 96-well microplate assay (12). The assay is based on the capture of NO by Fe2+-myoglobin (Mb) which is subsequently oxidized to Fe3+-myoglo- bin. The progress of heme oxidation was continuously measured in a kinetic microplate reader (Molecular Devices, Menlo Park, CA) as the rate of change in Am5-G5,,. Data points were collected from all 96 wells every 16 s for 20 min at 25 "C with shaking prior to each A measurement. The slope of the best fit regression line (Almin) was used to calculate the rate of NO' synthesis. All samples contained 10 pl of crude smooth muscle cell cytosol (0.7-2.5 mg protein/ml) and final concentrations of 20 p M Mb, 500 p M L-arginine, 500 p M NADPH, 10 p~ FAD, 10 p~ BH4, 0.1 unit/ml dihydropteridine reductase, and 80 mM Tris, pH 7.6.

Fe'+-Myoglobin Preparation-2 mM myoglobin (from horse skeletal muscle, Sigma) was reduced with an excess of sodium dithionite and immediately applied to a Sephadex G-25 column, followed by elution with 50 mM TRIS buffer, pH 7.6. Fe2+-Mb was aliquoted and stored a t -70 'C for up to 2 months prior to use.

Protein Assay-Protein was measured by the Bio-Rad dye-binding assay (Bio-Rad) using bovine serum albumin as a standard.

Chemicals-L-Nu-Methylarginine (NMA) was synthesized as pre- viously described (4). Sepiapterin, biopterin, and tetrahydrobiopterin were purchased from Dr. Schirks (Jonas, Switzerland). Rat recombi- nant interferon-y was from Amgen (Thousand Oaks, CA). Endotoxin (Escherichia coli, Olll:B4), arginase (from bovine liver), L-arginine, N"-nitro-L-arginine, and all other chemicals were purchased from Sigma.

RESULTS

Characterization of LPS Induction of the Arginine-NO Pathway-Bacterial endotoxin (LPS) activates rat aortic smooth cells (RASMC) to synthesize and release nitrite (Fig. 2 ) . LPS-induced nitrite synthesis increased as a function of LPS concentration and the duration of LPS exposure (Fig. 2 A ) . Characteristically, a lag phase of 6-8 h preceded induc- tion of nitrite synthesis, followed by a progressive increase in nitrite synthesis and a tapering by 48 h. As shown in Fig. l B ,

2.0 / P' I

0 12 24 36 48 TIME (hrs)

n

-h 4.2 (u + 50 ng/rnl IFN-y

8 2.4

g 1.2

k 7 - : : :::+-ut"-

FIG. 2. Activation of rat aortic smooth muscle cell nitrite production by LPS: concentration and time dependence and potentiation by interferon-? (IFN-7). Panel A, cumulative re- lease of nitrite into the cell culture medium as a function of time and concentration of LPS. Panel B, 24-h nitrite production elicited by LPS alone (CONTROL) and in the presence of 50 ng/ml IFN-y. Note that in the absence of LPS, IFN-y did not activate nitrite production (< 0.2 nmo1/24 h). All points represent the mean of four values k S.E. In cases where error bars are not apparent they are contained within the data symbols.

LPS-induced nitrite release is markedly potentiated by IFN. In the presence of 50 ng/ml IFN, the concentration-response curve for LPS-induced nitrite release is shifted to the left by more than 2 log units. In contrast, RASMC treated with IFN alone (at concentrations of up to 1 pg/ml) did not produce detectable levels of nitrite (<0.2 nmo1/24 h).

Several findings suggest that LPS-induced nitrite produc- tion by RASMC originates from arginine-derived nitric oxide. As shown in Fig. 3A, the quantity of nitrite produced by RASMC during a 24-h exposure to LPS (50 pg/ml) increases as a function of the concentration of L-arginine in the cell culture medium (E& = 20 p ~ ) . Moreover, RASMC treated for 24 h with a combination of LPS (50 pg/ml) and IFN (50 ng/ml) in arginase-containing culture medium (4.1 units/ml) were found to produce less than 2% of the nitrite observed when similarly treated in the absence of arginase. Prototypic nitric oxide synthase inhibitors, Nw-methyl- and N"-nitro-L- arginine (NMA and NNA, respectively) elicit concentration- dependent inhibitions of LPS/IFN-induced nitrite synthesis (Fig. 3B) ; ECSO concentrations for NMA and NNA were 30 and 100 pM versus an L-arginine concentration of 1.24 mM. This rank order of potency is similar to that which we have

Tetrahydrobiopterin Synthesis and Nitric Oxide Generation 25725

3 _ _ A.

’ + l W H N M A e/e e“.”.-e’-

0- : : ::+ . : . : . ... : . : . d

.o 1 0.1 1 .o [ARGININE] (mM)

[NG-SUBSTITUTEO ARGININE] 0 4 )

FIG. 3. LPS-induced nitrite production is arginine-depend- ent (panel A ) and inhibited by Nu-substituted arginine analogs (panel B ) . Confluent rat aortic smooth muscle cell cultures were pretreated with LPS (50 pg/ml) for 24 h; after this time, culture medium was replaced with fresh medium containing the desired concentrations of L-arginine and L-arginine analogs. In panel A , accumulation of nitrite in the cell culture medium during the subse- quent 16 h is expressed as mean concentration * S.E. ( n = 4). In panel B, nitrite is expressed as a percent of the concentration k S.E. obtained after 20 h in the absence of added inhibitor ( n = 4). Note that the concentration of arginine in the cell culture medium for the experiment depicted in panel B was 1.24 mM.

previously observed for the induced form of nitric oxide syn- thase in macrophages and endothelial cells (12,21). Inhibition of LPS/IFN-induced NO synthesis by NMA was reversible by high concentrations of L-arginine (Fig. 3A) , suggesting that this action of NMA is mediated specifically via compet- itive inhibition of the arginine-NO pathway.

To test the requirement of LPS/IFN-induced nitrite syn- thesis on mRNA synthesis, experiments were performed with actinomycin D (ACT D). Inclusion of ACT D (0.5 pg/ml) during a 15-h pretreatment period with LPS/IFN (50 ng/ml and 50 pg/ml, respectively), followed by washout and replace- ment with fresh medium, resulted in a >98% reduction in nitrite synthesis during the next 20 h. In contrast, when cells were induced with LPS/IFN for 15 h in the absence of ACT D, but ACT D was continuously present during the 20-h postinduction period, nitrite synthesis during these 20 h was completely unaffected. Thus, the action of ACT D is not mediated by nonspecific cytotoxicity. Furthermore, mRNA synthesis is required for induction of nitrite synthesis by LPS/IFN, but it is not required to sustain the activity of the NO synthetic pathway, once induced.

Induction of nitrite synthesis is coincident with the release of a factor which oxidizes Fez+-heme myoglobin (Mb) to met- myoglobin; Fez+-heme iron oxidation is a prototypic reaction of NO and one which has served as the basis for a useful kinetic assay for NO (37, 12). As shown in Fig. 4, RASMC that had been pretreated for 16 h with LPS/IFN in L-arginine- containing culture medium oxidize Mb at a rate 7-fold greater than control cells. While the basal rate of Mb oxidation observed with untreated RASMC is unaffected by addition of 1 mM NMA, the increased rate of Mb oxidation by LPS/IFN- treated cells is abolished. Moreover, LPS/IFN-activated Mb oxidation by RASMC is markedly diminished (-85%) by

UNTRFATED WS/IFN-TREATED W S / I F ? I - T R ~ ~ + ARGININE + ARGININE - ARGININE

FIG. 4. Fez+-myoglobin oxidation by LPS/IFN-activated rat aortic smooth muscle cells. Cell monolayers were grown to conflu- ence in 96-well plates and were either untreated or pretreated for 24 h with the combination of 50 pg/ml LPS and 50 ng/ml interferon-? (LPSIIFN). After this time, the cell culture medium was replaced with 200 p1 of fresh medium either containing or free of arginine (1.24 mM), in the absence or presence of Nu-methyl-L-arginine (NMA, 1 mM). Fez+-myoglobin (5 PM) was added to each well, and the rate of increase in A550.650 was monitored at 16-s intervals for 30 min in a kinetic microplate reader at 37 “C.

BAUL Lps LPS LPS IFN + IFN + IFN + LMHP

FIG. 5. Influence of immunostimulants on total biopterin content (oxidized plus reduced forms) of rat aortic smooth muscle cells. Biopterin was assayed in groups of confluent cells that were either untreated (basal), or treated for 12 h with LPS (30 pg/ ml), interferon-? (ZFN, 50 ng/ml), the combination of LPS and IFN, or the combination of LPS, IFN, and 2,4-diamino-6-hydroxypyrimi- dine (DAHP, 3 mM). Bars indicate mean values * S.E. of three to four replicate treatments, and data are expressed as a function of protein concentration. n.d. indicates no detectable biopterin (<0.2 ng/ml).

removal of L-arginine from the cell culture medium. Effect of Immunostimulants on the Biopterin Content of

RASMC-As shown in Fig. 5, basal levels of biopterin (re- duced and oxidized forms) in RASMC were below the limits of assay detection (c0.2 ng/mg protein). However, after treat- ment with 30 pg/ml LPS for 12 h, cellular biopterin increased to >1 ng/mg protein. While IFN alone did not have a detect- able effect on cellular biopterin, it caused a doubling of the increase observed with LPS. This LPS/IFN-induced increase in cellular biopterin was completely abolished when cells were concomitantly treated with 3 mM 2,4-diamino-6-hydroxypyr- idine (DAHP), a selective inhibitor of GTP cyclohydrolase I.

Effect of Tetrahydrobiopterin (BH4) Synthesis Modulators on Induction of Nitric Oxide Synthesis-In addition to block- ing BH4 synthesis, DAHP elicited a concentration-dependent inhibition of LPS/IFN-induced nitrite synthesis (E& = 1.6 mM; see inset to Fig. 6). The influence of DAHP (2 mM) on the time course of nitrite production by LPS/IFN-activated RASMC is shown in Fig. 6. DAHP inhibited nitrite produc- tion at all times studied. Inhibition by DAHP was completely overcome by co-administration of sepiapterin (SEP, 100 p ~ ) , an agent which is a substrate for BH4 synthesis via the dihydrofolate reductase-dependent pterin salvage pathway (see Fig. 1) and therefore would be expected to restore BH4 synthesis during GTP cyclohydrolase I blockade. In the pres-

25726 Tetrahydrobiopterin Synthesis and Nitric Oxide Generation 5 -

a . W

0 5 10 15 20 25 30 35 40

TIME AFTER LPS TREATMENT (hrs)

FIG. 6. 2,4-Diamino-6-hydroxypyrimidine (DAHP) inhib- its the activation of nitrite production elicited in rat aortic smooth muscle cells by LPS (50 pg/ml) in combination with interferon-y (50 ng/ml). Nitrite accumulation in the cell culture medium was assayed as a function of time after addition of LPS/IFN alone (CONTROL) or LPS/IFN in the presence oE DAHP (3 mM), DAHP plus sepiapterin (SEP, 100 pM), or DAHP plus sepiapterin and methotrexate (MTX, 10 p ~ ) . Note that DAHP inhibits induced nitrite synthesis by a mechanism which is reversed by sepiapterin and restored by MTX. Inset, concentration-response relationship for inhibition by DAHP of 24-h nitrite accumulation elicited by LPS/ IFN. All points represent mean values * S.E. ( n = 4).

[SEPUPTERIN] M) [YETHOTRDUTE] W)

FIG. 7. Concentration-response relationships showing the influence of sepiapterin and methotrexate on LPS/IFN-in- duced nitrite production by rat aortic smooth muscle cells. Nitrite accumulation in the cell culture medium was assayed after a 24-h exposure to the combination of LPS (50 pg/ml), interferon-y (50 ng/ml), and the desired agents. Data are expressed as the percent change from control (LPS/IFN alone) nitrite production. Left panel, effect of sepiapterin (open circles) and sepiapterin in combination with 10 p~ methotrexate (filled circles) on nitrite production. Right panel, effect of methotrexate (open circles) and methotrexate in combination with 100 p M sepiapterin (filled circles) on nitrite pro- duction. Points represent mean values * S.E. ( n = 4).

ence of methotrexate (10 p ~ , MTX), a selective inhibitor of dihydrofolate reductase, SEP no longer restored nitrite syn- thesis to DAHP-treated, LPS/IFN-activated, RASMC.

In the absence of other biopterin synthesis modulators, SEP elicited a concentration-dependent potentiation of LPS/ IFN-induced NO synthesis by RASMC (Fig. 7A, open circles). At a concentration of 300 p ~ , SEP more than doubled 24-h LPS/IFN-induced nitrite production. In contrast, when 10 p~ MTX was present, SEP elicited a dose-dependent and complete inhibition of induced NO synthesis (EC50 = 30 p ~ ; Fig. 7A, closed circles). These findings are consistent with prior reports demonstrating that in addition to sepiapterin being a substrate for BH4 synthesis via the pterin salvage pathway, it is also a potent inhibitor of GTP cyclohydrolase I (38, 39) and therefore a blocker of de no00 BH4 synthesis. Thus, when the metabolism of SEP to BH4 is blocked by MTX, inhibition of de novo BH4 synthesis by SEP is revealed.

MTX, alone, diminished LPS/IFN-induced NO synthesis t o a maximal extent ranging from 10 to 40% ( n = 4), depend- ing on the RASMC preparation studied. The maximal inhib- itory effect of MTX was observed even with 0.3 pM, the lowest

concentration studied (Fig. 7B, open circles). This potency of MTX is consistent with its established Ki for inhibition of dihydrofolate reductase in the nanomolar range (40). In the presence of 100 pM SEP, MTX (0.3-30 p ~ ) elicited a dose- dependent and complete inhibition of LPS/IFN-induced ni- trite synthesis with an EC5o < l pM (Fig. 7B, filled circles). The finding that BH4 synthesis becomes 100% dependent on dihydrofolate reductase-dependent pterin salvage in the pres- ence of SEP can be explained by SEP's inhibitory action on GTP cyclohydrolase I. Thus, when de nouo synthesis of BH4 is blocked by SEP, MTX-inhibitable pterin salvage is the only alternative route for cellular BH4 synthesis.

Media supplementation with the combination of 100 p~ hypoxanthine and 40 p~ thymidine did not alter the ability of MTX to reduce LPS/IFN-induced NO synthesis, indicating that inhibition of nucleotide synthesis does not contribute to the inhibitory effect of MTX on induced NO synthesis. More- over, inhibition of induced nitrite synthesis by 10 p~ MTX did not result from a generalized cytostatic action, because 40-h exposure did not markedly reduce RASMC viability as determined by MTT cell respiration assay (88.8 f 3.6% of control, n = 4). In contrast, a 40-h exposure to LPS/IFN (50 and 30 pg/ml, respectively) was found to abolish RASMC respiration, both in the absence and presence of 1 mM SEP (3.2 f 2.2% and 5.4 f 5.7% of control, respectively; n = 4). Consistent with the view that NO is the mediator of LPS/ IFN-induced respiratory inhibition in RASMC, we observed that 10 p~ MTX in the presence of 100 p~ SEP (which blocks LPS/IFN-induced NO synthesis, see Fig. 7 B ) signifi- cantly reduced LPS/IFN-induced cytotoxicity, whereas MTX alone afforded little protection (52.4 f 3.5% versus 9.9 k 5.7% of control, respectively).

Detailed study of the the time course of induction of NO synthesis by LPS/IFN revealed that the potentiation of NO synthesis caused by 100 p~ SEP was much more marked a t early time points (Fig. 8). While SEP enhanced NO produc- tion 2-fold at 24 h, production was enhanced by over 10-fold after only 6 h. Moreover, results shown in Fig. 8 demonstrate that, in the presence of SEP, induction of NO synthesis by LPS/IFN is shifted to earlier times and rises at a steeper rate. These findings suggest that BH4 limits both the onset and magnitude of LPS/IFN-induced NOS activity. Fig. 8 also shows that in the presence of 10 p~ MTX, SEP does not enhance the rate of induction of NO synthesis. Indeed, con- sistent with results depicted in Fig. 6, the combination of MTX and SEP was found to completely abolish induction of

TIME (hr)

FIG. 8. Modulators of tetrahydrobiopterin (BH4) synthesis on the time course of LPS/IFN-activated nitrite synthesis in rat aortic smooth muscle cells. Nitrite accumulation in the cell culture medium was assayed at the indicated times after addition of LPS (50 pg/ml), interferon-y (ZFN, 50 ng/ml), and the indicated agent(s). BH4 synthesis was either stimulated by treatment with sepiapterin (SEP, 100 p ~ ) or inhibited by treatment with the com- bination of sepiapterin and methotrexate (MTX, 10 p ~ ) . Points represent mean values f S.E. (n = 6).

Tetrahydrobiopterin Synthesis and Nitric Oxide Generation 25727

NO synthetic activity by LPS/IFN. In contrast with SEP, BH4 elicited only a small potentia-

tion of LPS/IFN-induced nitrite synthesis at the highest concentration studied (15% with 300 phi; see Fig. 9). It is apparent that BH4 can be utilized to support NOS activity, because BH4 elicited a concentration-dependent restoration of LPS/IFN-induced NO synthesis to cells in which de nouo BH4 synthesis has been blocked with DAHP (Fig. 9). How- ever, restoration of NO synthesis to DAHP-treated cells was not observed when BH4 was administered in combination with 10 FM MTX. This finding suggests that BH4 does not enter the cell as BH4 per se. Rather, BH4 being unstable, likely enters the cell as a dihydropterin, requiring reduction by dihydrofolate reductase back to BH4 for NOS utilization.

N-Acetylserotonin (NAS) has been shown to block BH4 synthesis in cells by selectively inhibiting sepiapterin reduc- tase (41), an enzyme which catalyzes two consecutive NADPH-dependent keto-oxidation reactions, the final steps in BH4 synthesis from GTP (42). As shown in Fig. 10, NAS elicited a concentration-dependent inhibition of LPS/IFN- induced nitrite synthesis in RASMC; the maximal extent of inhibition was = 60% with an ECS0 of = 20 p ~ . When 10 WM MTX was simultaneously present, NAS caused a near-com- plete inhibition of LPS/IFN-induced nitrite synthesis, also with a 20 WM EC50.

DISCUSSION

Recent findings suggest that overproduction of NO within the blood vessel wall is a major cause of LPS- and cytokine-

0 1 10 1 0 0

[T~RAHYDROBIOPTERIN] bM)

FIG. 9. Tetrahydrobiopterin overcomes the inhibition by diamino-6-hydroxypyrimidine (DAHP) of LPS/lFN-induced nitrite synthesis in rat aortic smooth muscle cells. The effect of BH4 on 24-h nitrite accumulation was assessed on cells treated with LPS (50 pg/ml) and interferon-y (50 ng/mI) aIone (CONTROL), and on cells co-treated with DAHP (3 mM) or DAHP plus metho- trexate (MTX; 10 pM) . Data are expressed as the percent change from LPS/IFN-induced nitrite production in the absence of BH4. Points represent mean values k S.E. ( n = 4).

c e c 8 6 ?5

! K

_ e : : 0 1 10 1w

[N-ACE'VLSEROTONIN] b M )

FIG. 10. N- Acetylserotonin inhibits LPS/IFN-induced ni- trite synthesis by rat aortic smooth muscle cells: potentiation by (MTX). The effect of NAS on 24-h nitrite accumulation was assessed on cells treated with LPS (50 pg/ml) and interferon-y (50 ng/ml) alone (CONTROL), and in the presence of MTX (10 p ~ ) . Data are expressed as the percent change from LPS/IFN-induced nitrite production in the absence of NAS and MTX. Points represent mean values f S.E. (n = 4).

induced hypotension (6-10). Since it is likely that the prin- cipal source of LPS-induced NO synthesis in the endotoxic vessel wall is smooth muscle (15, 16), we have sought to elucidate mechanisms which contribute to this induction process. Results of the present study clearly indicate that BH4 synthesis is a necessary event for induction of NO synthesis in vascular smooth muscle.

We have measured the accumulation of nitrite, a stable oxidation product of the unstable free radical nitric oxide, to quantitate induction of NO synthesis. The validity of this measurement is supported by our findings that induced nitrite synthesis is: 1) arginine-dependent (Fig. 3 A ) , 2) potently inhibited by selective NOS inhibitors in an arginine-reversible fashion (Fig. 3, A and B ) , and 3) associated with the produc- tion of a factor which resembles NO, in that: ( a ) it oxidizes Fez+-myoglobin, ( b ) its synthesis requires arginine, and ( c ) is blocked by NMA (Fig. 4).

LPS-induced nitrite production by vascular smooth muscle cells occurs after a lag-phase of 6-8 h, during which time mRNA synthesis is required. While NO synthase is clearly one of the genes transcribed in this additional gene products may include secondary cytokines, cytokine receptors, and possibly synthetic enzymes for necessary cofactors such as BH4.

LPS-induced NO synthesis is markedly potentiated by IFN, both in terms of magnitude and LPS concentration required for half-maximal activation (see Fig. 2). Synergistic effects of IFN with IL-la and tumor necrosis factor-a for induction of smooth muscle nitrite synthesis have also been observed (9, 16). Since sepsis is a condition in which all of these cytokines are released in concert (43), potent stimuli abound during sepsis for induction of NO synthesis in vascular smooth muscle.

Synergistic actions between IFN and other immunomodu- lators for induction of NOS activity have been described for various cell types including: macrophages (44), endothelial cells (ll), hepatocytes (45), and fibroblasts (46). The mecha- nism of synergy between IFN and LPS for induction of NO synthesis is presently unknown. Previous studies demonstrat- ing induction of GTP cyclohydrolase I by IFN in vascular and circulating blood cells (32, 33) raise the possibility that syn- ergistic induction of NO synthesis by LPS and IFN may depend, at least in part, on the combined appearance of each of two factors which are required for NO synthesis: NOS and the enzymatic machinery for BH4 production.

In the present studies we have found that RASMC do not contain detectable quantities of BH4; however, BH4 accu- mulates upon treatment. with LPS. IFN potentiated the LPS- induced accumulation of BH4 by = 2-fold, similar to the degree of enhancement by IFN of LPS-induced NO synthesis. Interestingly, IFN on its own affected neither LPS-induced biopterin accumulation nor NO synthesis, therefore its action can be clearly attributed to synergism with LPS for both effects. The inability of IFN to enhance BH4 levels in RASMC came as a surprise in that it has been shown to do so in macrophages, lymphocytes, and fibroblasts (32, 33), where it acts by increasing activity of GTP cyclohydrolase I, the first and rate-limiting enzyme for de nouo synthesis of BH4. While it is likely that induction of GTP cyclohydrolase I by LPS/IFN is also the basis for elevated levels of BH4 in vascular smooth muscle, direct assay of GTP cyclohydrolase

* S. Gross, unpublished observation. 3The induction of vascular smooth muscle cell NO synthesis is

coincident with the appearance of a 120-125-kDa protein that is undetectable in noninduced cells and which purifies with NOS activ- ity.

25728 Tetrahydrobiopterin Synthesis and Nitric Oxide Generation

I activity will be required to confirm this possibility. Given our finding in the present study that BH4 content of RASMC limits LPS-induced NO synthesis, one contributant to the synergistic effect of LPS and IFN on induced NO synthesis is a synergistic elevation of BH4 levels. I t should be noted, however, that additional mechanisms, at the level of NO synthase transcription,' appear also to contribute to the syn- ergism between LPS and IFN for NO synthesis induction.

Pharmacological probes were used to assess the contribu- tion of BH4 synthetic pathways to induced NO production in LPS-activated vascular smooth muscle cells. Results strongly indicate that synthesis of BH4 is an absolute requirement for induction of NOS activity. A significant reduction in LPS/ IFN-induced smooth muscle NO synthesis was observed upon inhibition of the de novo pathway for BH4 synthesis from GTP, and to a lesser extent upon inhibition of BH4 synthesis via the pterin salvage pathway. In these experiments, de novo BH4 synthesis was selectively abolished with DAHP and NAS, agents previously shown to be selective inhibitors of GTP cyclohydrolase I (47) and SEP reductase (41), respec- tively. These agents attenuated induced NO synthesis to a maximal extent ranging from 60 to 90% of control values (see Figs. 6 and 10). This range of inhibitor potency observed with different preparations or passages of cultured smooth muscle cells may reflect differences in cell metabolic status or, more likely, arise from lot-to-lot differences in the pterin content of serum used for supplementation of the cell culture medium. Neither DAHP nor NAS inhibited smooth muscle cell respi- ration, thereby ruling out nonspecific cytotoxicity as a mech- anism of inhibition of induced NO synthesis. Moreover, in- hibition of NO synthesis by DAHP was circumvented by BH4 and by SEP, substrates for the pterin salvage pathway (see Figs. 7 and 9). The reversal of inhibition by DAHP with BH4 and SEP was not observed in the presence of MTX, an inhibitor of the terminal enzyme for dihydropterin salvage, dihydrofolate reductase. Thus, both SEP and BH44 require metabolism by dihydrofolate reductase to a product which can override the inhibition of induced NO synthesis by DAHP. These findings strongly suggest that the mechanism of NO synthesis inhibition by DAHP is specifically via block- ade of de nouo BH4 synthesis.

When administered alone, MTX maximally inhibited LPS/ IFN-induced NO synthesis by 10-40% of the control level. However, administration of MTX in combination with an inhibitor of de novo BH4 synthesis (DAHP, NAS, or sepiap- terin) completely abolished the ability of LPS/IFN to induce smooth muscle NO synthesis. I t is noteworhty that MTX, at concentrations up to 300 pM, does not attenuate NOS activity of crude LPS/IFN-induced vascular smooth muscle cytosol.* Thus, the mechanism by which MTX inhibits induced NOS activity appears to be specifically via inhibition of BH4 syn- thesis by the dihydrofolate reductase-dependent pterin sal- vage pathway. These findings suggest that the BH4 which is required for induced NOS activity arises predominantly from GTP via de nouo synthesis and, to a lesser extent, via pterin salvage. Whether this is also the case in vivo remains to be tested.

A major conclusion of the present studies is that LPS/IFN- induced NO synthesis by vascular smooth muscle cells is rate- limited by BH4 availability. Providing excess BH4, via ad-

BH4 is unstable at physiological conditions of pH, temperature, and oxygen tension, resulting in rapid oxidation to BH2 in the cell culture medium; thus, BH2 is the likely species taken up by cells when BH2 is administered. Dihydrofolate reductase is required for intracellular conversion of BH2 to BH4, explaining the MTX sensi- tivity of NO synthesis restoration by BH4.

ministration of the BH4 precursor SEP, or BH4 itself, in- creased LPS/IFN-induced NO synthesis to 200% and 117% of control values, respectively (see Figs. 7 and 9). While the mechanism of potentiation of LPS-induced NOS activity by SEP and BH4 is most likely via enhanced cofactor availabil- ity, we currently cannot rule out the possibility that these agents potentiate induced levels of NOS. The lower efficacy of BH4, relative to SEP, for potentiating induced NO synthe- sis probably reflects a rapid oxidation of BH4 in cell culture medium and a poor cellular uptake of the BH2 formed. It is noteworthy that the lag period which precedes the appearance of NO synthesis by LPS/IFN-treated smooth muscle cells is abbreviated in the presence of sepiapterin, indicating that NOS induction may actually occur earlier than is evident from measurement of nitrite accumulation in the cell culture medium (see Fig. 8).

The abolition of induced NO synthesis in smooth muscle cells treated with the combination of DAHP and MTX is clearly attributable to inhibition of the synthesis of biopterin cofactor required for NOS activity, rather than a diminished level of NOS enzyme. Indeed, cytosolic NOS activity in cells treated with LPS/IFN in the presence of DAHP/MTX (3 mM and 10 p ~ , respectively) is actually greater than that in cells treated for an equivalent time with LPS/IFN alone.5 More- over, crude cytosolic NOS activity in cells induced in the presence of these pterin synthesis inhibitors showed a much more marked dependence on added BH4 (>go%-dependent) than control induced cells (<20%-dependent).

Although the de nouo synthetic pathway for BH4 synthesis has been shown to be inducible by immunological stimuli in various cell types, the function of immunologically induced BH4 has remained elusive. The present findings are consist- ent with the view that a major function for this induced BH4 synthesis may be to support co-induced NOS (46). Taken a step further, it is likely that neopterin, a plasma metabolite arising from the de novo BH4 synthetic pathway, and one which has served as a clinical marker for immunoactivation in man (34), may actually be an indicator of the activation status of the arginine/NO pathway.

Much attention has been focused recently on the potential of arginine-based NOS inhibitors for therapy of clinical con- ditions which have been associated with NO overproduction, i.e. septic and cytokine-induced hypotension (6-10). There- fore, it is reasonable to consider biopterin synthesis inhibitors as a pharmacological strategy which may be employed to limit induced NO synthesis. Since neither DAHP (300 PM), MTX (10 FM), nor SEP (300 p ~ ) alter vasorelaxation to endothe- lium-dependent dilators in vitro over a period of at least 1-2 h,6 it is conceivable that biopterin synthesis inhibitors can be used to selectively abolish LPS- and cytokine-induced NO synthase activity. Although biopterin synthesis inhibitors ap- pear to have no acute inhibitory effect on the constitutive NOS of endothelial cells, chronic treatment has recently been shown to block endothelial NO synthesis (48). The apparent selectivity of biopterin synthesis inhibitors for induced NOS activity may reflect a requirement of induced NOS for de nouo BH4 synthesis, whereas constitutive NOS may already be replete with BH4. Thus, it is conceivable that toxicities which may arise from treatment with nonselective arginine-based NO synthase inhibitors (which inhibit the physiological con- stitutive production of NO in brain and endothelium in ad- dition to pathophysiological NO overproduction) may be avoided with biopterin synthesis inhibitors. We speculate that biopterin synthesis inhibitors will have an important place in

S. Gross, manuscript in preparation. A. Madera and S. Gross, unpublished observation.

Tetrahydrobiopterin Synthesis and Nitric Oxide Generation 25729

the treatment of conditions arising from NO overproduction.

Acknowledgments-We thank our colleagues Dr. Babette Weksler and Dr. David Hajjar for assistance with cell culture and Dr. Yang Xu for biopterin measurement. Expert technical assistance was pro- vided by George Lam.

REFERENCES

2. Ignarro, L. J., Buga, G. M., Wood, K. S., Byrns, R. E., and Chaudhuri, G. 1. Furchgott, R. F., and Zawadski, J. V. (1980) Nature 288,373-376

4. Aisaka. K.. Gross. S. S.. Griffith. 0. W.. and Levi. R. (1989) Biochem. 3. Palmer, R., Ferrige, A., and Moncada, S. (1987) Nature 327, 524-526

(1987) Proc. Natl. Acad. Sei. U. S. A. 84,9265-9269

~~~~~ ~~~ , ~~I ~~

Biophys. Res. Cdmmun. 160, "1-886 '

5. Rees, D., Palmer R., and Moncada, S. (1989) Proc. Natl. Acad. Sci. U. S. A.

6. Kilbourn, R. G., Jubran, A., Gross, S. S., Griffith, 0. W., Levi, R., Adams, 86,3375 3378

J., and Lodato, R. F. (1990) Biochem. Biophys. Res. Commun. 172,1132-

, . .

11111 7. Thiemmerman, C., and Vane J. R. (1990) Eur. J. Pharmacol. 182, 591-

8. Kilboum, R. G., Gross, S. S., Jubran, A,, Adams, J., Griffith, 0. W., Levi, 595

R2-md Lodato, R. F. (1990) Proc. Natl. Acad. Sci. U. S. A. 87, 3629-

9.

10.

11. 12.

13. 14.

Kilbourn, R. G., Gross, S. S., Lodato, R. F., Adams, J., Levi, R., Miller, L. L., Lachman, L. B., and Griffith, 0. (1992) J. Natl. Cancer Inst. 84, 1008-1016

Kilbourn, R. G., Owen-Schaub, L., Gross, S. S., Griffith, 0. W., and Logothetis, C. (1992) in Biology of Nitric O d e (Moncada, S., Marletta,

Kilbourn, R. G., and Belloni, P. (1990) J. Natl. Cancer Inst. 82, 772-776 M., and Hibbs, J., eds) pp. 236-242, London, England, Portland Press

Gross, S. S., Jaffe, E. A,, Levi, R., and Kilbourn, R. G. (1991) Biochem.

Busse, R., and Mulsch, A. (1990) FEBS Lett. 275.87-90 Beasley, D., Schwartz, J. H., and Brenner, B. M. (1991) J. Clin. Inuest. 87,

J0.U

Biophys. Res. Commun. 178,823-829

15. Schini, V. B., Junquero, D. C., Scott-Burden, T., and Vanhoutte, P. M.

16. Gross, S. S., Madera, A. M., Griffith, 0. W., and Levi, R. (1991) FASEB J.

17. Stuehr D. J., Kwon, N. S., Nathan, C. F., Griffith, 0. W., Feldman, P. L.,

18. Mayer, B., John, M., Heinzel, B., Werner, E. R., Wachter, H., Schultz, G.,

19. Stuehr, D. J., and Griffith, 0. W. (1992) Adu. Enzymol. Relot. Areas Mol.

602-608

(1991) Biochem. Biophys. Res. Commun. 176,114-121

6, A1728

and Wiseman, J. (1991) J. Biol. Chem. 266,6259-6263

and Bohme, E. (1991) FEBS Lett. 288, 187-191

Biol. 66, 287-346

20.

21.

22. 23.

24. 25.

Bredt, D. S, and Snyder, S. H. (1990) Proc. Natl. Acad. Sci. U. S. A. 87,

Gross, S. S., Stuehr, D. J , Aisaka, K., Jaffe, E. A., Levi, R., and Griffith, 0.

Bredt, D. S, Hwang, M., and Snyder, S. H. (1991) Nature 347, 768-770 Moncada, S., Palmer, R. M. J., and Higgs, E. A. (1989) Biochem. Phurmacol.

Nathan, C. F., and Stuehr, D. J. (1990) J. Natl. Cancer Inst . 82 , 726-728 Kwon. N.. Nathan. C. F.. and Stuehr. D. (1989) J. Biol. Chem. 264.20496-

682-685

W. (1990) Biochem. Biophys. Res. Commun. 170,96-103

38, 1709-1715

20501 '

, ,

26. Tayeh, M. A,, and Marletta, M. A. (1989) J. Biol. Chem. 264,19654-19658 27. Mayer, B., John, M., and Bohme, E (1990) FEBS Lett. 277,215-219 28. Schmidt, H. H. H. W., Pollack, J. S., Nakane, M., Gorsky, L. D., Foster-

mann, U., and Murad, F. (1991) Proc. Natl. Acad. Sci. U. 5'. A. 88, 365-

29. Nichol, C., Smith, G., and Duch, D. (1985) Annu. Reu. Biochem. 64, 729- 369

30. Ballahsene, Z., Dhondt, J.-L., and Farriaux J.-P. (1984) Biochem. J. 217, 764

31. Shen, R., Alam, A,, and Zhang, Y. (1988) Biochim. Biophys. Acta 966, 9- 59-65

32. Werner, E. R., Werner-Felmeyer, G., Fuchs D., Hausen, A,, Ziehnegger, G., 15

33. Zieeler. I.. Schott. K.. LChbert. M. F.. Schwulera. U.. and Bacher. A. (1990) and Wachter, H. (1989) Biochem. J. 262,861-866

Bk.'Chem. 265, 17026-17030 '

, , , . .

34. Wachter, H., Fuchs, D., Hausen, A., Reibnegger, G., and Werner, E. R.

35. Bussolati, G. (1984) in Advances in Immunohistochemistry (Delellis, R. A.,

36. Fukushima, T., and Nixon, J. C. (1980) Anal. Biochem. 102, 176-188 37. Feelisch, M., and Noack, E. A. (1987) Eur. J . Pharmacol. 139, 19-30 38. Shen, R-S., Alam, A,, and Zhang, Y. (1988) Biochim. Biophys. Acta 965,

39. Jacobson, B. K., and Mannos, R. E. (1989) Biochem. J. 260, 135-141 40. Osborn. M. J.. Freeman. M.. and Huennekens. F. M. (1958) Proc. SOC. EXD.

(1989) Adu. Clin. Chem. 27,81-141

ed) pp. 325-337, Masson Publishing, New York

9-15

Biol. 'ked. 97,429-431

1099-1107 41. Milstien, S., and Kaufman S. (1985) Biochem. Biophys. Res. Commun. 28,

42. Milstien, S., and Kaufman S. (1989) J. Biol. Chem. 264,8066-8073 43. Dinerello, C., and Savage, N. (1989) Crit. Reu. Immunol. 9, 1-20 44. Stuehr, D. J., and Marletta, M. A. (1987) J. Immunol. 139,518-525 45. Curran, R. D., Billiar, T. R., Stuehr, D. J., Ochoa, J. B., Harhrecht, B. G.,

Flint, S. G., and Simmons, R. L. (1990) Ann. Surg. 212,462-469 46. Wener-Felmayer, G., Werner, E. R., Fuchs, D., Hausen, A,, Reibnegger, G.,

and Wachter, H. (1990) J. Ex Med. 172,1599-1607 47. Gal, E. M., Nelson, J. M., and {herman, A. D. (1978) Neurochem. Res. 3,

64-1111

48. Schmidt, K., Werner, E. R., Mayer, B., Wachter, H., and Kukovetz, W. R. VY V"

(1992) Biochem. J. 281,297-300