Journal of Cardio vascular Pharmacology
`17(Suppl. 3):Sl-S9 © 1991 Raven Press, Ltd. , New York
`
`The L-Arginine:Nitric Oxide Pathway
`
`S. Moncada, E. A. Higgs, H. F. Hodson, R. G. Knowles, P. Lopez-Jaramillo,
`T. McCall, R. M. J. Palmer, M. W. Radomski, D. D. Rees , and R. Schulz
`
`The Wellcome R esearch Laboratories, Kent, England
`
`Downloaded from http://journals.lww.com/cardiovascularpharm by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 05/20/2021
`
`Summary: Nitric oxide (NO) is synthesized from L-argi(cid:173)
`nine by an enzyme called the NO synthase. NO produc(cid:173)
`tion by vascular endothelial cells accounts for the bio(cid:173)
`logical properties of endothelium-derived relaxing factor
`and maintains a vasodilator tone. Platelets also produce
`NO , which modulates their aggregability. In the central
`nervous system NO mediates the effects of excitatory
`amino acids. In all these tissues the L-arginine: NO path(cid:173)
`way acts as a transduction mechanism for the soluble
`guanylate cyclase, for which NO is the endogenous stim(cid:173)
`ulator. NO is also released after immunological stimula-
`
`tion in macrophages, neutrophils , and other cells. NO
`thus released acts as part of the host defense mechanism,
`for it is cytotoxic or cytostatic for tumor cells and invasive
`organisms and may mediate other aspects of the immu(cid:173)
`nological response. Discovery of the L-arginine: NO path(cid:173)
`way has far reaching implications for the understanding
`of several areas in biology and indicates potential novel
`therapies for a number of diseases. Key Words: Nitric
`oxide-L-Arginine-Endothelium-derived relaxing fac(cid:173)
`tor.
`
`The demonstration in 1987 of the formation of ni(cid:173)
`tric oxide (NO) by vascular endothelial cells has
`opened up what can now be considered a new area
`of biological research. It soon became clear that NO
`explains the biological properties of the so-called
`endothelium-derived relaxing factor (EDRF). Fur(cid:173)
`thermore, NO is the molecule responsible for the
`stimulation of the soluble guanylate cyclase in a
`number of tissues such as the brain and is also a
`cytotoxic factor released by activated murine mac(cid:173)
`rophages. NO is synthesized from the amino acid
`L-arginine by an enzyme that has been named NO
`synthase. The L-arginine: NO pathway has also
`been identified or implicated in many other tissues,
`including platelets, adrenal glands, hepatocytes ,
`Kupffer cells, the lung, and several cell lines.
`
`THE VASCULAR ENDOTHELIUM
`EDRF (l) is a labile humoral substance that re(cid:173)
`laxes vascular smooth muscle and inhibits platelet
`aggregation (2,3) and adhesion (4) via stimulation of
`the soluble guanylate cyclase. The chemical nature
`ofEDRF has now been identified as NO. The phar(cid:173)
`macological properties of EDRF and NO are iden(cid:173)
`tical (for review , see ref. 5) , including their ability
`
`to increase guanosine-3 ' ,5'-cyclic monophosphate
`(cyclic GMP) levels in vascular tissue (6). Further(cid:173)
`more, NO is released from vascular endothelial cells
`in culture (7 ,8) and from vascular preparations (6,9-
`12) in amounts sufficient to explain the biological
`actions of EDRF.
`Porcine vascular endothelial cells in culture syn(cid:173)
`thesize NO from the terminal guanidino nitrogen
`atom(s) of L-arginine (13,14). This reaction is spe(cid:173)
`cific , as a number of analogues of L-arginine, in(cid:173)
`cluding its o enantiomer, are not substrates. In ad(cid:173)
`dition, one analogue, NG-monomethyl-L-arginine
`(L-NMMA), inhibits this synthesis in a dose-depen(cid:173)
`dent and enantiomerically specific manner (15). The
`coproduct of this reaction is L-citrulline, as its syn(cid:173)
`thesis from L-arginine by vascular' endothelial cell
`homogenates is also inhibited by L-NMMA (16).
`The enzyme that synthesizes NO from L-arginine
`has been studied in endothelial homogenates and
`characterized (16). This enzyme, which now is
`called NO synthase (17), is NADPH- and Ca2 + -de(cid:173)
`pendent (18,19). Furthermore, NO synthesis from
`endothelial cell cytosol is inhibited by calmodulin(cid:173)
`binding peptides and antagonists, an effect that is
`reversed by calmodulin, suggesting that the Ca2 + -
`
`Address correspondence and reprint requests to Dr. S. Moncada at The Wellcome Research Laboratories, Langley Court, Beckenham,
`Kent BR3 3BS, England.
`
`SJ
`
`Human Power of N Company
`EX1041
`Page 1 of 9
`
`

`

`S2
`
`S. MONCADA ET AL.
`
`dependent stimulation of NO synthase in endothe(cid:173)
`lial cells is mediated by calmodulin (20) .
`We have recently shown that small variations in
`extracellular Ca2 + concentrations around the phys(cid:173)
`iological range alter the production of both basal and
`stimulated NO by the vascular endothelium (21).
`Release of NO was maximal at the physiological
`Ca2 + concentration of l.25 mM and abolished in the
`absence of Ca 2 + . Reduced NO formation was ob(cid:173)
`served at either low or high Ca2 + concentrations.
`These results indicate that Ca2 + plays a crucial role
`in the modulation of vascular tone and suggest that
`small changes in extracellular Ca2 + levels may be
`relevant in some forms of hypertension, where low
`ionized Ca2 + concentrations have been observed in
`serum (22 ,23).
`In rabbit aortic rings , the NO synthase inhibitor
`L-NMMA induces a small but significant endothe(cid:173)
`lium-dependent contraction and inhibits the relax(cid:173)
`ation and the release of NO induced by acetylcho(cid:173)
`line (ACh) . L-Arginine , which on its own only
`induces a smalJ endothelium-dependent relaxation ,
`antagonizes all of the actions ofL-NMMA (24). Sim(cid:173)
`ilar results have been obtained in guinea pig pul(cid:173)
`monary artery rings (25) and in the coronary cir(cid:173)
`culation of the rabbit heart in vitro (26).
`in
`the anesthetized rabbit , L(cid:173)
`Furthermore ,
`NMMA induces a dose-dependent, long-lasting in(cid:173)
`crease in mean arterial blood pressure (BP) and
`inhibits the hypotensive action of ACh without
`affecting that of the endothelium-independent vaso(cid:173)
`dilator, glyceryl trinitrate (GTN) (27). These effects
`of L-NMMA are reversible by L-arginine. L-NMMA
`has subsequently been shown to cause a rise in BP
`in anesthetized guinea pigs (28) and rats (29-32). In
`one of these studies (31) , the vasodilation induced
`by ACh was accompanied by an increase in the uri(cid:173)
`nary excretion of cyclic GMP , both of which were
`prevented by L-NMMA. The effects of L-NMMA
`on BP were accompanied by a decrease in glomer(cid:173)
`ular filtration rate .
`The increase in BP induced by L-NMMA was also
`associated with a decrease in vascular conductance
`in the renal, mesenteric, carotid , and hindquarters
`vascular beds of conscious, chronically instru(cid:173)
`mented rats (33). Furthermore, these effects were
`sustained if the infusion of L-NMMA was continued
`for up to 6 h (34), indicating not only the critical role
`of NO in maintaining vascular patency in all of these
`beds, but also the fact that regulatory systems in the
`vasculature are unable to reaccommodate the flow
`toward pretreatment levels . In awake, chronically
`instrumented dogs , L-NMMA induced a dose-re(cid:173)
`lated, L-·arginine-reversible , constriction of the cor(cid:173)
`onary circulation together with a reduction in rest(cid:173)
`ing phasic coronary flow (35). Preliminary evidence
`suggests that the coronary vasodilation that follows
`vagal stimulation is also NO dependent (36) .
`Studies in which L-NMMA has been infused into
`
`J Cardio vasc Pharmacol, Vol. 17 (Suppl. 3), 1991
`
`the brachial artery or the dorsal veins of the hand
`in humans demonstrated that the vasodilation in(cid:173)
`duced by ACh or bradykinin, but not that induced
`by GTN , could be attenuated by the compound
`(37 ,38). Furthermore, L-NMMA induced direct
`vasoconstriction in the brachial artery, whereas it
`had no such direct effect on the hand veins . This
`suggests that in the arterial side of the circulation,
`but not the venous side, there is a continuous release
`of NO that maintains a dilator tone . That the arterial
`side of the circulation releases, in general, more NO
`than the venous side, is also suggested by the fact
`that ACh-induced dilatation in veins was rapidly
`transformed into constriction as the dose increased.
`Interestingly , the dilatation was attenuated by L(cid:173)
`NMMA while the constriction was enhanced, sug(cid:173)
`gesting that NO mediates, at least in part , the di(cid:173)
`lation and functionally antagonizes vasoconstrictor
`responses (38).
`ln the microcirculation of the hamster cheek
`pouch , intravenm,is L-NMMA caused a reduction in
`diameter of the arterioles but not the venules (39).
`The arterioles of this preparation dilated in response
`to NO, prostacyclin, ACh , and bradykinin (39,40) .
`Other L-arginine analogues have recently been de(cid:173)
`scribed as inhibitors of NO generation in vascular
`tissue (32 ,41-46) . NG-nitro-L-arginine (L-NN A), its
`methyl ester L-NAME , and N-iminoethyl-L-orni(cid:173)
`thine (L-NlO) all caused concentration-dependent
`inhibition of the Ca2
`+ -dependent NO synthase from
`porcine aorta and induced endothelium-dependent
`contraction and inhibition of endothelium-depen(cid:173)
`dent relaxation induced by ACh in aortic rings (32).
`Furthermore , they each induced a dose-dependent
`increase in mean systemic arterial blood pressure
`and significantly inhibited the hypotensive response
`to ACh and bradykinin in the anesthetized rat. In(cid:173)
`terestingly, L-NlO was approximately five times
`more potent than the other analogues, suggesting
`that there may be differences in uptake , distribu(cid:173)
`tion , or metabolism of these compounds (32). Some
`of these compounds are orally active, for L-NMMA
`and L-NAME, given by this route to Brattleboro
`rats, induced a sustained increase in blood pressure
`(46).
`All of these results clearly indicate that there is ,
`in the vasculature, a continuous utilization of L-ar(cid:173)
`ginine for the generation of NO, which plays a role
`in the maintenance of blood pressure. This release
`of NO can be enhanced further by endothelium-de(cid:173)
`pendent vasodilators. Furthermore, the marked rise
`in blood pressure obtained after inhibition of NO
`synthesis confirms the proposal that NO is the en(cid:173)
`dogenous nitrovasodilator (for review, see ref. 47)
`and suggests that a reduction in the synthesis of NO
`may contribute to the pathogenesis of hypertension.
`
`THE PLATELET
`We have recently shown that platelets also gen(cid:173)
`erate NO and that the L-arginine: NO pathway acts
`
`Page 2 of 9
`
`

`

`L-ARGININE:NITRIC OXIDE PATHWAY
`
`S3
`
`as a negativ~ feedback mechanism to regulate plate(cid:173)
`let aggregation (48). Aggregation induced by colla(cid:173)
`gen was accompanied by an increase in intraplatelet
`levels of cyclic GMP but not adenosine-3' ,5' -cyclic
`monophosphate (cyclic AMP). L-NMMA inhibited
`thi~ increase in cyclic GMP and enhanced aggre(cid:173)
`gation. Furthermore, L-arginine, which had no ef(cid:173)
`fect on basal levels of cyclic GMP enhanced the
`i~c~ease in cyclic GMP induced by ~ollagen and in(cid:173)
`h1b1ted aggregation.
`An increase in cyclic GMP in platelet cytosol was
`observed not only with sodium nitroprusside (a gen(cid:173)
`erator of NO) but also with L-arginine. The effect
`of L-arginine was enantiomer specific, inhibited by
`L-NMMA, and dependent on
`the presence of
`NADPH . In addition, measurements in the platelet
`cytosol demonstrated an L-arginine- and NADPH(cid:173)
`dependent formation of NO that was inhibited by
`L-NMMA, providing conclusive evidence for the
`existence of the L-arginine: NO pathway in the
`platelets. The formation of NO from L-arginine in
`platelet cytosol was dependent on the free Ca2 +
`concentration, showing that the NO synthase in
`platelets, as in the vascular endothelium, is Ca2+ -
`dependent.
`Interestingly, L-arginine did not increase the basal
`levels of cyclic GMP in unstimulated platelets but
`did so only when they were stimulated with colla(cid:173)
`gen. This shows that this NO synthase can utilize
`exogenous substrate once it is activated. Whether
`this is a function of the uptake of arginine into the
`cell or of the regulation of NO synthase by the sub(cid:173)
`strate is not yet known. In this context it is worth
`noting that L-arginine was 20- 30 times' less active
`as an inhibitor of platelet aggregation in blood than
`in washed platelets . This may be a reflection of the
`high concentrations (48-140 µM) (49,50) of L-argi(cid:173)
`nine in blood .
`Hemoglobin, a powerful inhibitor of the actions
`of NO, did not reverse the effects of L-arginine in
`washed platelets. However, it was the most effec(cid:173)
`tive inhibitor of the increase in cyclic GMP induced
`by L-arginine in platelet cytosol, showing that he(cid:173)
`moglobin did not penetrate the platelet membrane
`efficiently.
`Thus , platelet aggregation in vivo is likely to be
`regulated by intraplatelet NO, as well as by NO and
`prostacyclin released from vascular endothelium
`(39,51). The combined action of these two mediators
`could result in a synergistic suppression of [Ca2 + Ji
`elevation and subsequent powerful inhibition of
`platelet aggregation (51) . Interestingly , NO-induced
`inhibition of platelet adhesion to the endothelium
`and other structures of the vascular wall is not af(cid:173)
`fected by prostacyclin (4). Since the biochemical
`rationale for the synergy between agonists of aden(cid:173)
`yJ;ite and guanylate cyclases is believed to be de(cid:173)
`pendent on cyclic GMP-mediated inhibition of
`cyclic AMP hydrolysis (52), it is likely that platelet
`
`adhesion is controlled by cyclic GMP rather than
`cyclic AMP (4).
`
`THE BRAIN
`In 1977, NO was shown to stimulate the soluble
`guanylate cyclase in homogenates of mouse cere(cid:173)
`bral cortex (53). ln the same year, Deguchi showed
`that the soluble fraction of rat forebrain contained
`a low molecular weight substance that activated sol(cid:173)
`uble guanylate cyclase and whose action was inhib(cid:173)
`ited by hemoglobin (54}. Because the effects of the
`activator and of nitroso compounds were similar,
`but not additive, and both were inhibited by N(cid:173)
`~ethylhydroxylamine and hydroxylamine , this ac(cid:173)
`tivator was thought to be related to the nitrosamines
`or to act at the same allosteric site (55). Similar find(cid:173)
`ings were reported in the rat cerebellum (56). In
`1982, the endogenous activator of the soluble guan(cid:173)
`ylate cyclase in the brain was identified as L-arginine
`(57).
`These observations, together with the discovery
`of the L-arginine : NO pathway in the vascular en(cid:173)
`dothelium, led to the investigation of the existence
`of this pathway in the central nervous system to
`determine whether it accounted for L-arginine-de(cid:173)
`pendent stimulation of the soluble guanylate cy(cid:173)
`clase. Addition of L-arginine to rat synaptosomal
`cytosol in the presence of NADPH resulted in the
`formation of NO and citrulline accompanied by
`stimulation of soluble guanylate cyclase (58). Both
`of these processes were inhibited by hemoglobin
`and by L-NMMA, but not by L-canavanine. These
`data showed that the rat brain possesses the NO
`synthase. This enzyme was strikingly dependent on
`the free Ca2 + concentration (58) . The enzyme was
`essentially inactive at the resting free Ca2 + con(cid:173)
`centration in synaptosomes (approximately 80 nM)
`(59), whereas it was fully active at Ca2 + concen(cid:173)
`trations around 400 nM. Therefore, as in the vas(cid:173)
`cular endothel.ium and the platelet, increases in
`[Ca2 + li may constitute the physiological mechanism
`for stimulating the synthesis of NO.
`Garthwaite et al. had reported that stimulation of
`rat cerebellar cells with N-methyl-o-aspartate
`(NMDA) induced an elevation of cyclic GMP levels
`that was associated· with the release of an EDRF(cid:173)
`like material (60). Furthermore, the cells that re(cid:173)
`leased this EDRF-1-ike material in response to ex(cid:173)
`citatory stimulation were not the target cells in
`which cyclic GMP levels were elevated. The cyclic
`GMP response to NMDA stimulation was later
`shown to be enhanced by L-arginine and inhibited
`by L-NMMA in a manner that was reversed by ad(cid:173)
`ditional L-arginine (61,62), showing that this re(cid:173)
`~ponse was indeed mediated by NO. More recently,
`it has- been shown that L-NMMA administered in(cid:173)
`tracerebellarly in mice inhibits the increase in cyclic
`GMP induced by NMDA, quisqualate, kainate, bar-
`
`J Card;ovasc Pharmacol. Vol. 17 (Suppl . 3), 1991
`
`Page 3 of 9
`
`

`

`S4
`
`S. MONCADA ET AL.
`
`maline, and pentylenetetrazole (63). Furthermore,
`the increases in cyclic GMP induced by kainate in
`rat cerebellar slices were also inhibited by L-NMMA
`(64).
`Interestingly, physiological Ca2 + levels , which
`are essential for the action of the NO synthase, were
`found to inhibit the brain-soluble guanylate cyclase
`(58,65). This has led to the suggestion that this could
`represent a control mechanism whereby guanylate
`cyclase is not activated in those cells stimulated to
`produce NO but only in the effector cells (58) .
`The NO synthase in the brain was further char(cid:173)
`acterized and shown to be inhibited competitively
`by L-NMMA, L-NNA, and L-NIO (66). Like the en(cid:173)
`dothelial cell and platelet enzyme , the brain NO syn(cid:173)
`thase required only NADPH as a cofactor. Bovine
`brain cytosol has also been shown to contain the
`NO synthase (67). This enzyme has now been pu(cid:173)
`rified from rat cerebelJum and shown to be calmod(cid:173)
`ulin-dependent (68). The purified enzyme migrates
`as a single 150-kDa band on SDS-PAGE and ap(cid:173)
`pears to be a monomer. The biological conse(cid:173)
`quences of stimulating the soluble guanylate cyclase
`in different parts of the brain are yet to be eluci(cid:173)
`dated .
`
`PHAGOCYTIC CELLS
`When mammals and humans were fed a low
`NO:J diet, endogenous synthesis of NO; was ob(cid:173)
`served (69). This process was originally believed to
`be a result of intestinal microbial metabolism but
`was later shown, in germ-free animals , to be mam(cid:173)
`malian in origin (70) . A marked increase in urinary
`N03 excretion was observed in human subjects
`with diarrhea and fever (71 ,72) . Furthermore, treat(cid:173)
`ment of rats with Escherichia coli lipopolysacchar(cid:173)
`ide (LPS) led to an increase in the urinary output
`ofN03 that correlated with the degree of fever (73) .
`Turpentine and carrageenin produced similar,
`though smaller, increases in urinary N03 in rats.
`These results suggested a correlation between im(cid:173)
`munostimulation and elevated N03 synthesis.
`Later work showed that blood levels and urinary
`excretion of N03 increase after exposure to LPS
`in LPS-sensitive mice (74) and that activated mouse
`peritoneal macrophages show increased No 2- and
`NO; production in vitro (74). Moreover, L-arginine
`is essential for the formation of NO; and NO:J that
`occurs after activation of the macrophage cell line
`RAW 264.7 with LPS and -y-interferon (IFN --y) (75) .
`The NOi and N03-
`formed are derived from the
`terminal guanidino nitrogen atom(s) of L-arginine.
`This reaction, which is independent of the respira(cid:173)
`tory burst, results in the formation of L-citrulline as
`a coproduct (75).
`An L-arginine-dependent pathway, distinct from
`arginase (76), has been shown to be responsible for
`the cytotoxic activities of macrophages (77). These
`
`J Cardiovasc Pharmacol, Vol . 17 (Suppl . 3), 1991
`
`cytotoxic activities, which include inhibition of mi(cid:173)
`tochondrial respiration, aconitase activity, and
`DNA synthesis, are thought to be mediated by in(cid:173)
`hibition of iron-containing enzymes in target cells.
`These activities, as well as the generation of
`N02- and NO:l, are inhibited by L-NMMA (77).
`Tumor necrosis
`factor
`increases macrophage
`NO:l production in response to IFN--y, and this ef(cid:173)
`fect is also inhibited by L-NMMA (78).
`Following the demonstration of the synthesis of
`NO from L-arginine by mammalian cells (13), it be(cid:173)
`came apparent that NO was the most likely inor(cid:173)
`ganic nitrogen oxide intermediate in the pathway of
`N02- and N03 synthesis in the macrophage. Three
`groups subsequently demonstrated that NO synthe(cid:173)
`sized from L-arginine was indeed the precursor of
`the No2- and NO:l in these cells (79- 81). The cy(cid:173)
`totoxic properties of NO were also demonstrated,
`for treatment of guinea pig hepatoma cells with NO
`gas under anaerobic culture conditions was shown
`to cause intracellular Iron loss, inhibition of mito(cid:173)
`chondrial respiration, inhibition of aconitase, inhi(cid:173)
`bition of DNA synthesis, and cytostasis in the tumor
`cells (80,82).
`The NO synthase in the macrophage differs from
`that in the endothelial cell , platelet, and brain in that
`it was not detectable in either macrophage cell lines
`or freshly elicited macrophages that had not been
`activated by an agent such as LPS , alone or in com(cid:173)
`bination with IFN--y (74,83,84), and it required pro(cid:173)
`tein synthesis for its expression (79). Studies of the
`NO synthase in LPS/IFN--y-activated RAW 264. 7
`cells showed that the enzymatic activity (which was
`not present in nonactivated cells) was cytosolic and
`was present in the 100,000 x g supernatant (79).
`The enzyme required L-arginine and NADPH, but
`not Ca2 +, and its activity was enhanced by Mg2
`+ ,
`although this cation was not essential for the for(cid:173)
`mation of NO. Similar results have been obtained
`in cytosols of the murine macrophage cell line 1774
`activated with LPS and IFN --y (85). NO generation
`from activated macrophage cytosol has also been
`shown to be dependent on the presence of tetra(cid:173)
`hydrobiopterin (86,87) and to be stimulated by tlavin
`adenine dinucleotide and reduced glutathione (88).
`Formation of No 2- and NO:l in the macrophage
`is inhibited not only by L-NMMA, but also by L(cid:173)
`canavanine (75) . This is a further difference between
`the NO synthase in the macrophage and that in the
`endothelial cell, platelet, and brain. More recently,
`L-NIO has been shown to be a potent, rapid in onset,
`and irreversible inhibitor of NO generation in ac(cid:173)
`tivated J774 cells (85). In contrast, L-NMMA was
`found to be slower in onset, and L-NAME and L(cid:173)
`NNA were considerably less potent and were fully
`reversible (85). These properties make L-NIO a po(cid:173)
`tentially useful tool to investigate the generation of
`NO in phagocytic cells .
`In 1988, glycogen-elicted rat peritoneal neutro-
`
`Page 4 of 9
`
`

`

`L-ARGININE:NITRIC OXIDE PATHWAY
`
`S5
`
`phils were found to release a factor that relaxed vas(cid:173)
`cular smooth m·uscle (89) and increased cyclic GMP
`levels in this tissue (90). Human neutrophils were
`subsequently shown to inhibit platelet aggregation
`(91), an effect accompanied by increased platelet
`cyclic GMP levels (92). The release of this antiag(cid:173)
`gregating activity from rat peritoneal and human
`neutrophils was inhibited by L-NMMA and by L(cid:173)
`canavanine (93,94). The inhibitory effects of both
`L-canavanine and L-NMMA required a preincuba(cid:173)
`tion time of 50 min, probably reflecting the slow
`uptake of these inhibitors into the neutrophils. The
`identity of this ED RF-like factor released from stim(cid:173)
`ulated human neutrophils was confirmed as NO
`using chemiluminescence (94,95).
`The release of NO from rat peritoneal neutrophils
`could also be inhibited by L-NIO (85). Unlike the
`effect of L-NMMA and L-canavanine, the effect of
`L-NlO was rapid in onset , reaching full inhibitory
`effect within 10 min, an observation similar to that
`!n J774 cells (85) . Thus, the uptake of this compound
`mto neutrophils differs significantly from that of L(cid:173)
`NMMA and L-canavanine and suggests that the up(cid:173)
`take of L-arginine analogues into cells is an active
`process that may differ from one cell to anot her.
`Furthermore, the maximum degree of inhibition of
`NO formation observed with L-NIO was signifi(cid:173)
`cantly greater than that observed with L-NMMA.
`The presence of the NO synthase in these cells has
`recently been confirmed (85). The enzyme has the
`same characteristics as that in the J774 cells-it is
`cytosolic and Ca2 +-independent. Interestingly, L(cid:173)
`NAME, which is a potent inhibitor of endothelial
`NO synthesis (32,44), did not affect the generation
`of NO by the neutrophil (85 ,96). However, both L(cid:173)
`NNA and L-NAME were weak inh ibitors of NO
`synthase isolated from the neutrophil (85).
`The biological significance of NO production by
`neutrophils and whether the NO synthase is induced
`in these cells during differentiation remai n to be elu(cid:173)
`cidated. At present , there is no evidence that neu(cid:173)
`trophil -derived NO plays a role in the cytotoxic ac(cid:173)
`tivity of these cells . The release of NO may be a
`mechanism for controlling the levels of 02 , which
`is also released by stimulated neutrophils (93) .
`Thus, at low degrees of cell activation, such as dur(cid:173)
`ing margination and emigration, release of NO may
`predominate , whereas at higher degrees of activa(cid:173)
`tion the balance changes in favor of 0 2 . NO and
`02 are known to interact rapidly to form, under
`some conditions, the peroxynitrite an ion, which de(cid:173)
`composes once protonated into the potent oxidants
`OH- and N02· (97). It has been suggested that the
`protective action of superoxide dismutase in ische(cid:173)
`mic tissue may be due in part to the prevention of
`the formation of peroxynitrite and hence of these
`highly toxic radicals (97) . However, at present,
`there is no evidence for the formation of the per(cid:173)
`oxynitrite radical or its decomposition into OH· and
`
`N02· in vivo, and therefore it remains to be estab(cid:173)
`lished whether the interaction between NO and
`02 leads to an increased toxicity or is a neutralizing
`mechanism for two toxic radicals.
`
`OTHER CELLS AND TISSUES
`An NO synthase in the adrenal gland has been
`identified and partially characterized (98). Cyclic
`GMP levels in the soluble fraction of homogenates
`of rat adrenal glands were stimulated by sodium ni(cid:173)
`troprusside (SNP) and by S-nitroso-N-acetylpeni(cid:173)
`cillamine (SNAP), known activators of soluble
`guanylate cyclase, demonstrating the presence of
`this enzyme in this preparation. The soluble guan(cid:173)
`ylate cyclase from both the cortex and medulla was
`also stimulated in the presence of L-arginine and this
`stimulation was inhibited by hemoglobin. Further(cid:173)
`more, the stimulation by L-arginine was inhibited by
`L-NMMA, L-NNA, L-NAME, L-NIO , and L-cana(cid:173)
`vanine and was associated with the formation of
`both N02 and citrulline. Thus, an L-arginine: NO
`synthase is present in both the cortex and the me(cid:173)
`dulla of the adrenal gland . This NO synthase is also
`NADPH- and Ca2 +-dependent , like that in .the vas(cid:173)
`cular endothelium, the platelet, and the brain.
`The functional importance of this pathway in reg(cid:173)
`ulating adrenal cortex and medulla function is not
`clear. Cyclic GMP has been implicated in both cate(cid:173)
`cholamine secretion (99-101) and steroidogenesis
`( 102). It is therefore likely that the L-arginine : NO
`pathway plays a modulatory role in the control of
`the synthesis and secretion of hormones by the ad(cid:173)
`renal gland.
`Evidence for the presence of the L-arginine: NO
`pathway has also been found in some cell lines in(cid:173)
`cluding mouse neuroblastoma cells
`(NlE-115)
`(103, 104) and porcine kidney epithelial ceHs (LLC(cid:173)
`PK 1) stimulated with oxytocin (l 04) or with vaso(cid:173)
`pressin (105). Rat mast cells have also been shown
`to produce an NO-like substance that modulates the
`release of histamine (106). The recent finding that
`NO is released upon electrical stimulation of the
`nonadrenergic noncholinergic (NANC) nerve end(cid:173)
`ings in the isolated canine ileocolonic junction ex(cid:173)
`tends the presence of the L-arginine: NO pathway
`to peripheral nervous tissue ( 107).
`Rat Kupffer cells cocultured with hepatocytes
`and . stimulated with LPS induced a significant
`suppression of hepatocyte total protein synthesis,
`but only when L-arginine was present in the medium
`(108). This effect, which required an induction pe(cid:173)
`riod of several hours, was associated with formation
`of N02, NOi, and citrulline both in the Kupffer
`cells and the hepatocytes. It was subsequently
`shown that the supernatant from activated Kupffer
`cells induced the formation of NO in the hepato(cid:173)
`cytes, an effect that was blocked by L-NMMA (109).
`These data have led to the suggestion that Kupffer
`
`J Cardiovas c Plrarmaco/, Vol. 17 (Suppl. 3), 199.I
`
`Page 5 of 9
`
`

`

`S6
`
`S. MONCADA ET AL.
`
`ceUs activated by septic stimuli or other inflam(cid:173)
`matory states respond by forming NO themselves
`and by inducing NO formation in neighboring hep(cid:173)
`atocytes. A major effect of this NO is cytotoxic
`since it suppresses hepatocyte protein synthesis
`(109) . Induction by endotoxin of NO synthase in
`hepatocytes and in the lung has also been demon(cid:173)
`strated in vivo (110).
`An NO synthase similar to that found in the
`macrophage and the neutrophil has also been iden(cid:173)
`tified in some other nonphagocytic cell lines. EMT-
`6 cells, a spontaneous murine mammary adenocar(cid:173)
`cinoma cell line, when exposed to medium condi(cid:173)
`tioned by activated macrophages ,
`synthesize
`NOi, N03, and L-citrulline (I 11). This effect is de(cid:173)
`pendent on the presence of L-arginine and is accom(cid:173)
`panied by inhibition of aconitase and of DNA rep(cid:173)
`lication, together with the release of iron from the
`cells. Cycloheximide prevents all of these re(cid:173)
`sponses, showing that protein synthesis is required
`for the NO synthase to be induced (l 11). The same
`pattern of response can be induced in EMT-6 cells
`by IFN--y in combination with LPS. tumor necrosis
`factor (TNF), or interleukin- I (IL- I), where it is as(cid:173)
`sociated with prevention of cell growth (112). Treat(cid:173)
`ment of EMT-6 cells with IFN --y and LPS stopped
`not only their own proliferation but also inhibited
`DNA synthesis in other mouse, rat , and human
`tumor cell lines (I 1.3). Murine fibroblasts stimulated
`with IFN--y + IL-I or IFN--y + TNFa or IFN--y +
`IL-I + TNFa have also been shown to generate
`NO (Hibbs et al. , personal com munication). These
`observations provide further support for the role of
`NO as mediator of the cytostatic actions of lym(cid:173)
`phokines .
`
`CONCLUSIONS
`The discovery of the generation of NO by mam(cid:173)
`malian tissues and the elucidation of some of its
`biological roles has, in the last few years, thrown
`new light onto many areas of research. These range
`from the mediation of endothelium-dependent re(cid:173)
`laxation as a general adaptive mechanism of the car(cid:173)
`diovascular system to its role as the transduction
`mechanism for the stimulation of soluble guanylate
`cyclase in the central nervous system and many
`other tissues (114).
`With regard to NO as an immunologically re(cid:173)
`leased molecule, thus far we are only aware of its
`role as a cytotoxic-cytostatic agent. However, it is
`likely that this NO has other functions including the
`mediation of pathological increases in blood flow ,
`modulation of platelet and white cell behavior, and
`other aspects of the immunological response. It is
`likely that the inducible enzyme, which was origi(cid:173)
`nally thought to be present only in phagocytic cells,
`is more widely distributed and plays a role as a gen(cid:173)
`eral defense mechanism.
`
`J Cardiovasc Pharmacol. Vol. n (Supp l. 3), 1991
`
`The implications of the synthesis of NO in terms
`of potential novel treatments for different diseases
`needs to be analyzed and developed. It can , how(cid:173)
`ever, be predicted that, as with other fundamental
`biological discoveries, this will find its application
`in novel therapies.
`
`REFERENCES
`
`I. Furchgott RF . The role of endothelium in the responses of
`vascular smooth muscle to drugs. Annu Re v Pharma col
`Toxicol 1984;24: 175-97.
`2. Azuma H. Ishikawa M, Sekizaki S . Endothelium-dependent
`inhibition of platelet aggregation. Br J Pharmacol 1986;
`88:411-5.
`3. Radomski M , Palmer RMJ . Moncada S. Comparative phar(cid:173)
`macology of endothe lium-derived relaxing factor, nitric
`oxide and prostacyclin in platelets . Br J Pharmacol 1987;
`92: 181-7 .