The role of nitric oxide (NO) in inflammation represents one of the most studied yet controversial subjects in physiology. A number of reports have demonstrated that NO possesses potent anti-inflammatory properties, whereas an equally impressive number of studies suggest that NO may promote inflammation-induced cell and tissue dysfunction. The reasons for these apparent paradoxical observations are not entirely clear; however, we propose that understanding the physiological chemistry of NO and its metabolites will provide a blueprint by which one may distinguish the regulatory/anti-inflammatory properties of NO from its deleterious/proinflammatory effects. The physiological chemistry of NO is complex and encompasses numerous potential reactions. In an attempt to simplify the understanding of this chemistry, the physiological aspects of NO chemistry may be categorized into direct and indirect effects. This type of classification allows for consideration of timing, location, and rate of production of NO and the relevant targets likely to be affected. Direct effects are those reactions in which NO interacts directly with a biological molecule or target and are thought to occur under normal physiological conditions when the rates of NO production are low. Generally, these types of reactions may serve regulatory and/or anti-inflammatory functions. Indirect effects, on the other hand, are those reactions mediated by NO-derived intermediates such as reactive nitrogen oxide species derived from the reaction of NO with oxygen or superoxide and are produced when fluxes of NO are enhanced. We postulate that these types of reactions may predominate during times of active inflammation. Consideration of the physiological chemistry of NO and its metabolites will hopefully allow one to identify which of the many NO-dependent reactions are important in modulating the inflammatory response and may help in the design of new therapeutic strategies for the treatment of inflammatory tissue injury.
- free radicals
the landmark discovery that nitric oxide (NO; nitrogen monoxide) is synthesized by mammalian cells initiated a tremendous number of studies over the past 10 years, demonstrating that this diatomic free radical plays crucial roles in the homeostatic regulation of the cardiovascular, neuronal, and immune systems (7,14). Despite these important physiological functions, this oxidized metabolite of nitrogen is also a well-known toxic agent, being a constituent of air pollution and cigarette smoke. In addition to being an environmental toxin, endogenously formed NO is thought to promote a number of chronic inflammatory diseases such as arthritis, hepatitis, inflammatory bowel disease, septic and hemorrhagic shock, and certain autoimmune disorders (5, 6, 12, 15, 17). It is known, for example, that, in the presence of molecular oxygen (O2), NO can form reactive nitrogen oxide species that can damage DNA, inhibit a variety of enzymes, and initiate lipid peroxidation (27). The question then becomes: How can this molecule be so unstable in an aerobic environment, generating toxic intermediates, yet play such important roles in normal physiology?
The answer to this apparent paradox lies in understanding the physiological chemistry of NO, which may be defined as the fundamental chemistry involved in specific biological processes. Consideration of these chemical reactions provides a blueprint by which one may distinguish the regulatory processes and/or anti-inflammatory effects of NO from its potential toxic and/or proinflammatory properties (Fig. 1). This brief overview will discuss the basic physiological chemistry of NO and its metabolites, with particular attention given to those interactions that may modulate the inflammatory process.
PHYSIOLOGICAL CHEMISTRY OF NO
The physiological chemistry of NO is a complex process encompassing numerous potential reactions. However, the chemistry is, in fact, the most important determinant for dictating which type of effect NO has on biological systems. One may begin to decipher this chemistry by understanding how different concentrations or different rates of production of NO may influence certain reactions in a particular chemical event. To help guide us through this maze of reactions, it is convenient to separate NO chemistry into direct and indirect effects (27) (Fig.2). Direct effects are those reactions in which NO interacts directly with a biological molecule or target, whereas indirect effects are those reactions mediated by NO-derived intermediates such as reactive nitrogen oxide species derived from the reaction of NO with O2 or superoxide ( ). The advantage of categorizing NO chemistry into these two types of reactions is that the chemical reactions can be separated on the basis of relative rates (i.e., fluxes) of NO production. For example, direct effects may occur at lower concentrations or fluxes of NO, whereas higher NO fluxes result in indirect effects. Cells containing constitutive isoforms of nitric oxide synthase (NOS; i.e., endothelial cell or neuronal NOS) generate relatively small fluxes of NO and thus one would predict that direct effects of NO will predominate. Indeed, under normal physiological conditions, cells produce small but significant amounts of NO and only trace amounts of reactive oxygen species dictating that direct NO chemistry will predominate in normal tissue. However, in tissues where the high-output, inducible isoform of NOS (iNOS) has been upregulated, as in the case of chronic inflammation, indirect effects such as nitrosation, nitration, and oxidation will prevail (Fig. 2). Therefore, understanding the timing, location, and rates of NO production provides a guideline to the type of chemistry and relevant targets most likely to be affected.
The direct interaction of NO with metal-containing proteins or with organic free radicals represents two of the best characterized direct effects of NO in biological systems.
Metal nitrosyl complexes.
The reaction of NO with certain metals to form nitrosyl complexes occurs in vivo primarily with iron-containing proteins (26). The most facile reaction of NO to form stable nitrosyl adducts is with proteins that contain a heme moiety. The most notable of these is the reaction of NO and guanylate cyclase, which stimulates the formation of cGMP from GTP (7, 14). Synthesis of cGMP has several regulatory/anti-inflammatory effects such as regulation of vascular tone, inhibition of platelet aggregation, and leukocyte-endothelial cell interactions. Yet, this same type of chemistry can also inhibit other metalloproteins such as cytochromeP-450, NOS, cytochrome oxidase, and catalase (26).
In addition, it has been shown that small amounts of NO stimulate the production of prostaglandins (22). It is thought that inhibits cyclooxygenase (COX) and possibly other enzymes involved in prostaglandin synthesis by reducing the heme iron to its inactive ferrous (Fe2+) state. The presence of small amounts of NO would prevent this reduction of the heme iron by interacting with and scavenging . This would maintain the COX-associated heme iron in the ferric (Fe3+) form, resulting in the formation of potentially protective/anti-inflammatory prostaglandins. It should be noted, however, that the sustained overproduction of iNOS-derived NO will actually inhibit COX activity, which may, in and of itself, promote rather than inhibit inflammation.
NO, metal complexes, and oxidative stress.
It is becoming increasingly apparent that iron or hemoprotein-catalyzed oxidative reactions may mediate some of the pathophysiology associated with acute and chronic inflammation (5, 8, 27). This oxidative stress may enhance endothelial cell or leukocyte adhesion molecule expression, injure cells, induce apoptosis, damage DNA, promote inflammatory mediator synthesis, and regulate gene expression (4, 5, 27).
Several studies suggest that NO may modulate iron-catalyzed oxidation reactions such as the -driven Fenton reaction, which produces powerful oxidants such as the hydroxyl radical (OH ⋅) and metalloxo complexes The site-specific generation of OH ⋅ via this type of reaction has been proposed to help explain why superoxide dismutase (SOD), catalase, and/or iron chelators are effective in attenuating inflammatory tissue injury (4, 5). Kanner et al. (9) showed that iron-catalyzed oxidation reactions are inhibited by NO. Furthermore, we and others (13, 21) have found that NO dramatically inhibits the -driven Fenton reaction in vitro. For example, generation of and H2O2in the presence of small amounts of a redox-active Fe3+ chelate produces substantial amounts of the OH ⋅, as measured by the hydroxylation of benzoic acid. Addition of varying fluxes of NO to this system dramatically inhibited hydroxylation of benzoic acid such that equimolar fluxes of NO inhibited the - and H2O2-dependent hydroxylation of benzoic acid by ∼80% (13). These data demonstrate that, depending on the fluxes of the different reactive species, NO may have remarkable antioxidant capabilities.
Another modulatory and possibly more physiologically relevant reaction mediated by NO is its interaction with higher oxidation states of hemoproteins such as hemoglobin, myoglobin, and certain cytochromes. The reaction of the ferric forms of these pentacoordinate hemoproteins (HP—Fe3+) with H2O2results in the formation of the ferryl pi cation radical species (+⋅HP—Fe4+O), a potent and relatively stable oxidizing agent (8) Addition of NO reduces+⋅HP—Fe4+O to its original HP—Fe3+ oxidation state The rereduction of these higher oxidation states of hemoproteins prevents the accumulation of potent oxidizing agents, which could otherwise lead to potentially deleterious and proinflammatory effects. These types of reactions may account for the role of NO protecting cells against oxidant-induced cytotoxicity (9,27).
NO and free radicals.
In addition to metal-nitrosyl formation, NO may react directly with high-energy free radicals such as carbon-, oxygen-, and nitrogen-centered radicals. Because NO has an unpaired electron, it will rapidly react with alkoxyl (RO ⋅) and alkyl hydroperoxyl (ROO ⋅) radicals at near diffusion-limited reaction rates (k = 2 × 109M−1 ⋅ s−1) (16). These types of reactions may be critical in modulating metal or enzyme-catalyzed lipid peroxidation (21). Peroxidation of polyunsaturated lipids is thought to be an important pathophysiological event involved in the development of tissue injury and dysfunction in a number of different inflammatory conditions. Recent studies demonstrate that NO inhibits lipid peroxidation and thus may be important in modulating the inflammatory response by inhibiting the formation of proinflammatory lipids (21).
The mechanisms by which NO may inhibit lipid peroxidation are not entirely clear; however, one possible mechanism relates to the ability of NO to terminate propagation of lipid peroxidation reactions. Lipid peroxidation is initiated by the formation of potent oxidants such as OH ⋅ or ferryl hemoproteins that produce lipid alkyl radicals (L ⋅) from polyunsaturated fatty acids (LH). This radical is then converted to a hydroperoxyl radical (LOO ⋅) via its interaction with O2. The LOO ⋅ can abstract an allylic hydrogen atom from another LH, thereby propagating the free radical reaction As noted above, NO will rapidly react with these LOO ⋅ and/or lipid alkoxyl radicals (LO ⋅), resulting in LOONO or LONO formation, respectively, which in turn leads to chain termination (21). This chain termination step thus limits the extent of lipid peroxidation and may attenuate the formation of proinflammatory lipid mediators. However, it is still not clear what effects LOONO and/or LONO formation have under physiological conditions. Because LOONO is similar in structure to the potent oxidant peroxynitrous acid (HOONO), it is possible that this NO-free radical adduct may decompose under certain conditions to yield additional radicals and/or oxidants. On the basis of the preceding discussions, one would predict that the indirect effects of NO would be involved primarily but not exclusively in regulatory, protective, and/or anti-inflammatory processes in vivo.
In contrast to the direct effects of NO, the indirect effects of NO are mediated by reactive nitrogen oxide species derived from the interactions of NO with O2 or (Fig. 2). Indeed, reactive nitrogen oxide species have been suggested as important mediators of the pathophysiology associated with a variety of different models of inflammation (5, 6, 15, 17, 27). The most prevalent reactive nitrogen oxide species produced in vivo are dinitrogen trioxide (N2O3) and peroxynitrite (ONOO−), both of which can induce two types of chemical stresses: nitrosative and oxidative (27). Although oxidation and nitrosation may occur under normal physiological conditions, these types of chemical stresses are generally thought to be associated with certain pathophysiological situations (e.g., inflammation), where iNOS has been upregulated (27). Nitrosative and oxidative stresses provide different chemical signatures in biological systems with effects that are essentially orthogonal to each other. These reactive nitrogen oxide species-mediated reactions provide a balance in biological systems and are the key to understanding the role of the indirect effects in biological systems.
NO can react with O2 to yield reactive intermediates, which in turn may mediate additional nitrosative and/or oxidative reactions. The formation of these intermediates has been well studied in atmospheric chemistry due to their association with the deleterious effects of air pollution. It is well appreciated that the autoxidation of NO in an aqueous environment leads to the formation of reactive nitrogen oxide species such as N2O3, which is a potent nitrosating agent (25-27). However, due to the rapid reaction of NO with other biological substrates such as hemoproteins, this reaction has been suggested to occur too slowly to be of any consequence in vivo. Recent studies have shown, however, that hydrophobic environments, such as those found in the cellular membrane, can accelerate this reaction over 1,000 times (11). Subsequent studies by this same group have shown that the rate at which NO is scavenged by erythrocytes is approximately three orders of magnitude slower than the rate of the reaction between NO and purified oxyhemoglobin, suggesting that nitrosation reactions mediated by N2O3may in fact play important roles in vivo. This nitrosating agent has been shown to N- andS-nitrosate a variety of biological targets to yield potentially carcinogenic nitrosamines and nitrosothiol (RSNO) derivatives (25, 27). N2O3is primarily responsible for mediating nitrosative stress in vivo.
The chemistry of N2O3has been actively investigated for a number of years. This intermediate can nitrosate as well as oxidize different substrates (25, 27). Nitrosation of amines to yield nitrosamines most probably takes place only through the interaction of N2O3with amino compounds, thus providing the best indicator of nitrosative stress in the living organism. Indeed, a number of studies have shown that N-nitrosation does occur in vivo and may have important implications in the known association between chronic inflammation and malignant transformation (25, 27).
S-nitrosothiol adducts (RSNO) are also readily formed during the autoxidation of NO via the interaction between N2O3and certain thiols (26, 27). These complexes have been suggested to play an important role in NO transport, signal transduction pathways, and regulation of gene expression (23). Ignarro and co-workers showed that these adducts could stimulate the conversion of GTP to cGMP in guanylate cyclase and suggested that they are key intermediates in the action of various nitrovasodilating compounds such as sodium nitroprusside and nitroglycerin (reviewed in Ref. 7). In fact, it is thought that endothelium-derived relaxing factor may be an RSNO adduct. S-nitroso-albumin is thought to be the most abundant nitrosothiol in human plasma, with concentrations reported to be as high as 5 μM (23). Because many of the high- and low-molecular-weight RSNOs release NO either spontaneously or via metabolism, they are capable of mediating many of the biological functions of NO. In addition, Stamler and co-workers (24) have proposed that a specific cysteine residue located in the β-subunits of hemoglobin isS-nitrosated in the lung, with the NO group being subsequently released during the arterial-venous transit of red blood cells. They propose that this type of reaction may enhance blood flow during tissue oxygenation (24). The exact mechanism(s) by which hemoglobin is S-nitrosated in vivo and how S-nitrosohemoglobin may affect smooth muscle relaxation are unclear at the present time.
The formation of RSNO may also play an important role in leukocyte adhesion to the microvascular endothelium. For example, RSNOs are known to inhibit leukocyte adhesion to microvascular endothelial cells in vivo, presumably via the release of NO. However, it is also known that —SH groups are essential for normal leukocyte-endothelial cell adhesion (5). S-nitrosation of these critical SH groups on the surface of endothelial cells and/or polymorphonuclear neutrophils (PMNs) could decrease adhesion, thereby limiting leukocyte infiltration. Furthermore, the formation of endogenous antiadhesive RSNOs by NO-derived nitrosating agents may be inhibited by , suggesting that during inflammation enhanced production may promote PMN-endothelial cell adhesion (26). Indeed, it is well established that exogenous NO donors are very effective at inhibiting PMN adhesion in vivo (4, 5). In addition, because reduced glutathione (GSH) has such an unusually high affinity for N2O3, it is likely to play a critical role in inhibiting the toxicity of NO produced during times of enhanced NO production (27). Depletion of GSH by buthionine sulfoximine (BSO) renders cells considerably more susceptible to NO-mediated toxicity (27). Formation of RSNOs may, on the other hand, promote or perpetuate chronic inflammation. Lander and colleagues (10) have demonstrated thatS-nitrosation of one specific cysteine residue on the p21 Ras protein in lymphocytes is critical for guanine nucleotide exchange and downstream signaling resulting in the formation of proinflammatory cytokines such as tumor necrosis factor.
Much of the vascular and tissue injury observed in different models of inflammation have been shown to be inhibited by either SOD or NOS inhibitors, suggesting that both and NO are important mediators of tissue injury and organ dysfunction (4-6, 15). Because neither nor NO is a particularly potent oxidant or cytotoxin, it has been suggested that and NO may rapidly interact to produce the potent cytotoxic oxidants ONOO− and its conjugate acid ONOOH (1). Indeed, this hypothesis has generated an astonishing amount of interest because it has provided a biochemical rationale to account for the remarkable yet perplexing protective effects of intravenous administration of l-arginine analogs (NO synthase inhibitors) or SOD in models of inflammatory tissue injury. Numerous studies have been published describing the physicochemical and cytotoxic properties of preformed, chemically synthesized ONOO− (1, 20). Peroxynitrite in neutral solution is a powerful oxidant, oxidizing thiols or thioethers, nitrating tyrosine residues, nitrating and oxidizing guanosine, degrading carboydrates, initiating lipid peroxidation, and cleaving DNA. It is thought that the oxidant primarily responsible for these oxidative reactions is an electronically excited isomer of ONOOH that has both OH ⋅- and NO2-like properties (1, 20). Peroxynitrite exists in equilibrium with ONOOH, and in the absence of substrate this protonated form decomposes to yield nitrate ( ) via the intermediate formation of the excited isomer (ONOOH*) In addition to acting as a source of potentially toxic oxidants, this same reaction has been suggested to represent a major detoxification and anti-inflammatory pathway for the removal of without the concomitant formation of H2O2(4, 27). Only relatively recently have investigators attempted to quantitatively characterize the interaction between and NO under physiological conditions. We, as well as others, have attempted to systemically quantify the interaction between NO and and have made some rather surprising observations (13, 19, 21).
It has been demonstrated, for example, that simultaneous generation of equivalent amounts of and NO synergize to yield a potent oxidant (presumably ONOO−). We found that, when the flux of one radical exceeded the other, oxidant production was significantly reduced (13). We speculate that the decreased production of ONOO−/ONOOH in the presence of either excess NO or may be due to secondary chemical interactions occurring between NO or and ONOO−/ONOOH. Although the interaction between NO or and ONOO− is thermodynamically possible, recent work by Pfeiffer et al. (18) suggests that the rate constant for the interaction between NO and ONOO− may be too slow to compete with the spontaneous decomposition of ONOO−. It may be that NO reacts with and decomposes a specific isomer of HOONO, thereby effectively reducing steady-state concentrations of this potent oxidant (Wink and Grisham, unpublished observations).
The reaction rate constant for the interaction between and NO (6 × 109M−1 ⋅ s−1) has been used to suggest that ONOO−/ONOOH may be formed in vivo. One of the most important considerations for whether a chemical reaction occurs in vivo is the relative pseudo first-order rate constant and not just the rate constant itself. The extent of participation of a particular reaction is determined by the concentrations of the reactants as well as by the rate constant (27). The cellular concentrations of and NO under normal conditions have been estimated to be 0.1–1 nM and 0.1–1 μM, respectively (27). This suggests that the production of determines the location of the reaction between these radicals as well as the amount of ONOO− formed. The concentration and diffusion coefficient of NO determine whether the reaction occurs at any specific location. To illustrate this concept, it is useful to consider a model with two point sources of NO and (Fig.3). Simultaneous production of and NO by these two point sources would produce maximal nitrosative stress (i.e., NO-O2 interactions) near the source of NO. As NO diffuses away from its source, its concentration dilutes proportionally to distance, thereby enhancing the direct effects of NO as it diffuses from its source. As NO approaches the source of , it reacts to form ONOO−. However, as ONOO− diffuses into areas of or NO excess, it decomposes, thereby limiting ONOO−production (and its oxidative stress) to sites close to the source (Fig. 3).
Our studies as well as those of Rubbo et al. (21) have important physiological implications. We propose that the relative fluxes of and NO may act to control the steady-state concentration of ONOO− in vivo. Situations and cell types that could produce the optimal amounts of and NO to generate ONOO− would be macrophages and possibly endothelial cells. Neutrophils on the other hand when fully activated would produce 100–1,000 times more than NO, suggesting that PMNs would produce little or no ONOO−/ONOOH and indeed this is exactly what we observe (Grisham and Jourd’heuil, unpublished data). The occurrence of this reaction also depends on the competing reaction of with SOD. For example, SOD reacts with at a similar rate constant to that of the reactions of with NO, thereby representing the most important competing reaction to consider. For 10% of the formed to be converted to ONOO−, the concentration of NO would have to be 0.4–5 μM. The intracellular concentration of SOD is thought to be between 4 and 10 μM, whereas the mitochondrial SOD is probably as high as 20 μM in the organelle in which most of the is produced. Therefore, the reaction of with NO, despite the high rate constant, is probably confined to specific sites, i.e., site-specific generation in much the same way as OH ⋅.
In addition to oxidation reactions, ONOO− has been demonstrated in vitro to nitrate certain phenolic compounds such as tyrosine to yield primarily 3-nitrotyrosine (3-NT). Indeed, nitration of certain tyrosine residues may exert dramatic physiological effects by altering kinase reactions and subsequent signaling pathways. Some investigators have also suggested that the presence of 3-NT in tissues obtained from inflamed tissues is evidence for the formation of ONOO− formation in vivo. Two recent studies suggest that this may not be necessarily the case. Eiserich et al. (2) have recently reported that 3-NT may be formed by activated human PMNs via the myeloperoxidase-catalyzed H2O2-dependent oxidation of and tyrosine without the intermediate formation of ONOO−. These data would suggest that 3-NT may not necessarily reflect the formation of ONOO− but may more likely indicate the production of NO-derived . A second study by Pfeiffer and Mayer (19) demonstrated that, when NO and are generated at equal rates, little or no significant nitrosation of tyrosine occurs. Indeed, they demonstrated that nitration was most efficient with an NO donor alone, and, when fluxes exceeded those of NO, 3-NT formation was undetectable. These investigators concluded that 3-NT formation may not necessarily arise from ONOO− formation in vivo.
NO has numerous potential reactions that may greatly influence a variety of physiological and pathophysiological processes. Consideration of the timing, location, and rates of production of NO and reactive nitrogen oxide species may help identify which of the many NO-dependent reactions are important in modulating the inflammatory response. A basic understanding of the physiological chemistry of NO may also help in the design of new therapeutic strategies for the treatment of inflammatory tissue injury.
We wish to thank F. Stephen Laroux for his helpful discussions and preparation of the figures.
Address for reprint requests: M. B. Grisham, Dept. of Molecular and Cellular Physiology, LSU Medical Center, PO Box 33932, Shreveport, LA 71130-3932.
↵* First in a series of invited articles on Nitric Oxide.
Some of this work was supported by a grant from National Institute of Diabetes and Digestive and Kidney Diseases (DK-43785), the Crohn’s and Colitis Foundation of America, and the Arthritis Center of Excellence at LSU Medical Center in Shreveport.
- Copyright © 1999 the American Physiological Society