D. Lyn H. Williams

NO, nitrosonium ions, nitroxide ions, nitrosothiols and iron-nitrosyls in biology: a chemist's perspective

The multiplicity of biological functions thus far attributed to NO has led to suggestions that some effects might be mediated by other, related species instead. The radical nature of NO cannot account for its cytotoxicity, but its reaction with superoxide to form peroxynitite and highly reactive hydroxyl radicals may be important in this context. The ease with which NO can react with and destroy Fe-S clusters is also an important factor. Nitrosonium and nitroxide ions can be produced in vivo and will react under conditions that are physiologically relevant. Both could, in theory, serve in cell signalling or as cytotoxic agents. More direct experimental evidence for their involvement is needed before we can confidently assign them specific biological roles. In this article, Anthony Butler, Frederick Flitney and Lyn Williams discuss the chemistry of NO and related species.

Generation of nitric oxide from S-nitrosothiols using protein-bound Cu2+ sources

BACKGROUND We have recently shown that S-nitrosothiols (RSNOs) decompose in aqueous buffer to give nitric oxide, an important signalling molecule, and the corresponding disulphides. This occurs by reaction with Cu+ generated from Cu2+ (supplied as hydrated Cu2+) by thiolate reduction. To establish whether these reactions are feasible in vivo, we set out to determine whether Cu2+ bound to an amino acid, a tripeptide or to human serum albumin (HSA) could serve as a Cu+ source for generation of NO from S-nitrosothiols. RESULTS Experiments with Cu2+ bound to the tripeptide Gly-Gly-His or to two histidine molecules or to HSA showed that Cu+ was released (and trapped with neocuproine) when the copper source was treated with a thiol at pH 7.4. RSNO decomposition was achieved with all three copper sources, although not as rapidly as with added hydrated Cu2+. Decomposition was also catalyzed by ceruloplasmin. CONCLUSIONS These results show clearly that amino-acid- and protein-bound Cu2+ can be reduced by thiolate ion to Cu+, which will generate NO from RSNO species, thus providing a realistic model for these reactions in vivo.

Identification of Cu+ as the effective reagent in nitric oxide formation from S-nitrosothiols (RSNO)

Decomposition of S-nitrosothiols (RSNO) in aqueous solution at pH 7.4 is brought about by copper ions, either present as an impurity or specifically added. The primary products are nitric oxide and the disulfide. In the presence of the specific Cu+ chelator, neocuproine, reaction is progressively inhibited as the [neocuproine] is increased, the reaction eventually stopping completely. The characteristic UV–VIS spectrum of the Cu+ adduct can be obtained from the reaction solutions. This shows clearly that Cu+ and not Cu2+ is the effective catalyst. Two limiting kinetic conditions can be identified for a range of S-nitrosothiols at specific copper ion concentrations (a) a first-order dependence and (b) a zero-order dependence upon [RSNO]. Normally both situations also have a short induction period. This induction period can be removed by the addition of the corresponding thiol RSH. A mechanism is proposed in which Cu+ is formed by reduction of Cu2+ by thiolate anion via an intermediate, possibly RSCu+. Loss of nitric oxide from RSNO is then brought about by Cu+, probably via another intermediate in which Cu+ is bound to the nitrogen atom of the NO group and another electron-rich atom (such as nitrogen from an amino group, or oxygen from a carboxylate group) involving a six-membered ring. As well as NO this produces both RS– and Cu2+ which then are part of the cycle regenerating Cu+. Thiolate ion is oxidised to RS˙ which dimerizes to give the disulfide. Depending on the structure (and hence reactivity) of RSNO either Cu+ formation or its reaction with RSNO can be rate-limiting. Computer modelling of the reaction scheme allows the generation of absorbance time plots of the same forms as those generated experimentally, i.e. first- or zero-order, both with or without induction periods. We suggest that the thiolate ion necessary to bring about Cu2+ reduction is either present as a thiol impurity or is generated in small quantities by partial hydrolysis of the nitrosothiol, which results in an induction period. Addition of small quantities of thiol removes the induction period and leads to catalysis but larger quantities bring about a rate reduction by, it is suggested, complexation of the Cu2+. For two very unreactive substrates, S-nitrosoglutathione and S-nitroso-N-acetylcysteine very large induction periods were observed, typically three hours. This results, we suggest, from competitive re-oxidation of Cu+ to Cu2+ by the dissolved oxygen. Experiments carried out anaerobically confirm this, since there is then no induction period. Addition of hydrogen peroxide extends the induction period ever further. The results are discussed in terms of the biological properties of S-nitrosothiols which are related to nitric oxide release.

Haemoglobin: NO transporter, NO inactivator or NOne of the above?

The structural and functional characterization of haemoglobin (Hb) exceeds that of any other mammalian protein. Recently, the biological role attributed to Hb has been extended from the classical role in the transport and exchange of the respiratory gases O(2) and CO(2) to include a third gaseous molecule, nitric oxide (NO). It is postulated that Hb might be involved in the systemic transport and delivery of NO to tissues and in the facilitation of O(2) release. However, definitive evidence for these putative activities is yet to be produced and many questions remain. Here we describe the present status of these hypotheses and their strengths and weaknesses.

Metal ion catalysis in nitrosothiol (RSNO) decomposition

The decomposition of S-nitroso-N-acetyl D,L penicillamine (SNAP), an NO-donor drug, to give the disulfide and NO is catalysed by trace amounts of Cu2+ and Fe2+.

Catalysis by Cu2+ of nitric oxide release from S-nitrosothiols (RSNO)

The decomposition of a range of S-nitrosothiols (thionitrites) RSNO, based on cysteine derivatives, yields in water at pH 7.4 nitrite ion quantitatively. If oxygen is rigorously excluded then no nitrite ion is formed and nitric oxide can be detected using an NO-probe. The reaction is catalysed by trace quantities of Cu2+(there is often enough present in distilled water samples) and also to a lesser extent by Fe2+, but not by Zn2+, Cu2+, Mg2+, Ni2+, Co2+, Mn2+, Cr3+ or Fe3+. The rate equation (measuring the disappearance of the absorption at ca. 350 nm due to RSNO) was established as v=k[RSNO]·[Cu2+]+k′ over a range of [Cu2+] typically 5–50 µmol dm–3. The constant term k′ represents the component of the rate due to residual Cu2+ in the solvent and buffer components, together with the spontaneous thermal reaction. Decomposition can be virtually halted by the addition of EDTA. Reactions carried out in the presence of N-methylaniline gave a quantitative yield of N-methyl-N-nitrosoaniline, but a negligible yield when oxygen was rigorously excluded. Values of the second-order rate constant k were obtained for a range of S-nitrosothiols. Reactivity is highest for the S-nitrosothiols derived from cysteamine and penicillamine, when Cu2+ can be complexed both with the nitrogen atom of the nitroso group and the nitrogen atom of the amino group, via a six-membered ring intermediate. If there is no amino (or other electron donating group) present, reaction is very slow (as for RSNO derived from tert-butyl sulfide). N-Acetylation of the amino group reduces the reactivity drastically as does the introduction of another CH2 group in the chain. There is evidence of a significant gem-dimethyl effect. Kinetic results using the S-nitrosothiols derived from mercaptoacetic, thiolactic and thiomalic acids suggests that coordination can also occur via one of the oxygen atoms of the carboxylate group. EPR experiments which examined the Cu2+ signal showed no spectral change during the reaction suggesting that the mechanism does not involve oxidation and reduction with Cu2+⇄ Cu+ interconversion.

Transnitrosation between nitrosothiols and thiols

Transfer of the nitroso group from nitrosothiols to thiols occurs very readily in aqueous solution particularly at pH > ≈8. The results are consistent with attack by the thiolate anion at the nitroso nitrogen atom of the nitrosothiol. Results have been obtained for the reaction of S-nitroso-N-acetylpenicillamine (SNAP) with thioglycolic acid and also for the reaction of S-nitrosocysteine (SNCys) with thiomalic acid. Both reactions showed the same kinetic characteristics. The results are discussed in terms of transnitrosation reactions of nitroso compounds generally, and also in the case of nitrosothiols, in terms of possible in vivo transnitrosation and subsequent decomposition of a possibly more unstable nitrosothiol to yield nitric oxide; this may have implications for the mechanism of action of nitric oxide in a range of physiological processes.

The mechanism of nitric oxide formation from S-nitrosothiols (thionitrites)

S-Nitrosothiols (RSNO) are easily made by electrophilic nitrosation of thiols and are a convenient source of nitric oxide. Reaction occurs readily (in many cases) in aqueous buffer at pH 7.4 to give in addition the corresponding disulfide RSSR. If oxygen is not rigorously excluded from the solution, then the nitric oxide is converted quantitatively to nitrite ion, whereas in the absence of oxygen nitric oxide can be detected using a commercial NO-probe. Reaction, however, only occurs (apart from the photochemical pathway) if Cu2+ is present. There is often enough Cu2+ in the distilled water–buffer components to bring about reaction, but decomposition is halted if Cu2+ is complexed with EDTA. Experiments with the specific Cu+ chelator neocuproine however show that the true effective reagent is Cu+, formed by reduction of Cu2+ with thiolate ion. Kinetic experiments show that the most reactive nitrosothiols are those which can coordinate bidentately with Cu+, and there is a wide range of reactivity amongst the structures studied. Reactivity is crucially dependent on the concentrations of Cu2+ and RS–.Reaction also occurs, although somewhat more slowly, if the source of copper is the CuII complex with the tripeptide diglycyl-L-histidine (GGH) or as the CuII complex with human serum albumin (HSA). This allows the possibility that nitrosothiols could in principle generate nitric oxide in vivo using the naturally occurring sources of CuII.Rapid exchange of the NO-group in RSNO with thiols occurs, again in aqueous buffer at pH 7.4. This reaction has been established as a nucleophilic substitution reaction by the thiolate ion at the nitroso nitrogen atom.The implications of these results with regard to possible involvement of nitrosothiols in vivo are discussed.

Nitric oxide release from S-nitrosoglutathione (GSNO)

In the presence of Cu2+ (10–5 M) very little NO is generated from GSNO at ≡10–3 M at pH 7.4, whereas the reaction is quantitative at ≡10–6 M; this is explicable in terms of the complexation of Cu2+ by the product GSSG.

Decomposition of S-nitrosothiols: the effects of added thiols

The Cu+ (added as Cu2+) mediated decomposition of the five S-nitrosothiols, derived from penicillamine, cysteamine, thiomalic acid, N-acetylpenicillamine and cysteine have been examined kinetically in the presence of varying amounts of the corresponding thiols. Large differences in behaviour were found. In some cases, reactions were catalysed by added thiols, whereas in others, stability was conferred, often resulting in the appearance of quite large induction effects. The results are explained in terms of the dual function of the thiol (as thiolate), (a) as a reducing agent generating Cu+, and (b) as a complexing agent for Cu2+, when it is then less available for reduction. The balance of these two effects depends on the structure and concentration of the added thiol. The findings were supported by examining the two effects separately, using ascorbate as a reducing agent, and ethylenediaminetetraacetic acid as the complexing agent. For penicillamine, cysteine and thiomalic acid, the Cu2+ complexes were identified from their UV spectra, and their decomposition was followed kinetically.