Chem. Res. Toxicol. 2000, 13, 963-966
963
Communications Radical Dinitroalkane Dianions from the Nitration of Nitroalkanes by Peroxynitrite Ernst V. Arnold, D. Scott Bohle,* and Yangyan Hu Department of Chemistry, University of Wyoming, Laramie, Wyoming 82071-3838 Received July 11, 2000
In alkaline solutions at pH >10, peroxynitrite (ONOO-) rapidly and efficiently nitrates acinitroalkane anions, RCHdNO2- (R ) H, CH3, or CH3CH2), to give the radical dinitrodianions, RC(NO2)22-. These anions have been characterized by EPR and have multiplets based on 1:2:3:2:1 pentets consistent with a hyperfine coupling with two equivalent 14N nuclei of the nitro groups, and the following hyperfine coupling constants, in gauss: R ) H, aN ) 9.96 and aH ) 1.85; R ) CH3, aN ) 9.90 and aH ) 3.22; and R ) CH3CH2, aN ) 9.57 and aH ) 4.01. Nitration is attributed to the trapping of nitrogen dioxide, formed by ONO-OCO2- homolysis for the carbon dioxide adduct of peroxynitrite, by the aci-nitroalkane anion. In air, the radical dinitrodianions are oxidized to the monoanions, and the kinetics of its formation is readily followed by UV/vis spectroscopy of the rise in absorption at 380 nm. This report represents the first successful spin trapping of nitrogen dioxide formed from peroxynitrite, and the method may be a useful one for preparing geminal dinitroalkanes. Central to understanding nitric oxide’s immunological role (1) is tracing its fate once released by cytokineactivated macrophages (2). Among the cytokine inducible macrophage responses are the expression of inducible nitric oxide synthase and the formation of the phagosomal NADPH oxidase cluster (3, 4). These responses are correlated with different but overlapping time lines (5, 6) but are nevertheless often colocalized within a single cell, and lead to the release of nitric oxide and superoxide. These observations led to the hypothesis in 1990 that peroxynitrite, which results from the diffusion-controlled reaction of superoxide and nitric oxide, is an important derived cytotoxin in this system, and thus contributes to macrophage cell killing and inflammatory induced host cell injury (7, 8). This hypothesis has been highly contentious in that peroxynitrite has not been directly detected in vivo due to its short lifetime (τ1/2 ) 1.5 s) and that many of its reactions and decomposition pathways have until recently been poorly understood (6). There is now an emerging body of kinetic evidence which indicates that pernitrous acid (ONOOH, pKa ) 6.8) and the carbon dioxide adduct of peroxynitrite (ONOOCO2-) (9, 10) decay by radical pathways resulting from O-O bond homolysis (eqs 1 and 2).
While many groups have sought, with varying degrees of success, to trap the hydroxyl radicals resulting from
eq 1 (11-20), and two separate recent reports describe the direct flow EPR observation of carbonate radical formation from eq 2 (21, 22), there are no reports of spin trapping of the nitrogen dioxide radical which results from either homolytic pathway. In large measure, this is due to the low efficiency in the reaction of nitrogen dioxide with nitroxyl-based spin traps such as DMPO (23). In contrast, the aci-nitroalkane anions (RR′CdNO2-) rapidly form stable well-characterized spin adducts with both nitric oxide (24) and nitrogen dioxide (25) (eqs 3 and 4), and these have been thoroughly characterized by EPR spectroscopy (26-28). Formation of these adducts re-
quires high pH and the generation of the aci-anions (RCHdNO2-, where R is H, Me, or Et), and they are thus only suited for trapping radicals produced in alkaline solutions. In this report, we describe (1) new alkaline nitration reactions of peroxynitrite, (2) formal spin trapping of the nitrogen dioxide radical which results from carbon dioxide-catalyzed decomposition of peroxynitrite, and (3) a new efficient high-yield synthesis of geminal dinitroalkanes. Together, these results provide new
10.1021/tx0001474 CCC: $19.00 © 2000 American Chemical Society Published on Web 09/13/2000
964
Chem. Res. Toxicol., Vol. 13, No. 10, 2000
Communications
Figure 2. Relative EPR intensity as a function of time at H ) 3438 G, the field which corresponds to an intense signal from CH3C(NO2)22-•. The EPR sample was measured at ca. 18 °C, in an unstirred flat cell within the X-band cavity with the same conditions described in the legend of Figure 1.
Figure 1. EPR spectrum generated by the reaction of peroxynitrite and (A) nitropropane (giso ) 2.05, aN ) 9.96 G, and aH ) 1.85 G), (B) nitroethane (giso ) 2.05, aN ) 9.90 G, and aH ) 3.22 G), and (C) nitromethane (giso ) 2.05, aN ) 9.57 G, and aH ) 4.01 G) at pH 10 with 0.2 M nitroalkane, 0.2 M peroxynitrite, and adventitious levels of carbonate. Spectra were obtained on a Bruker EMX X-band spectrometer in an aqueous flat cell at 22 °C.
evidence for radical products from the decomposition of peroxynitrite. When alkaline peroxynitrite solutions are treated with nitroalkanes at room temperature, there is rapid formation of the corresponding radical dinitroalkane dianions. As shown in Figure 1, these radicals possess the characteristic 1:2:3:2:1 intensity pattern due to coupling of the unpaired electron with two equivalent I ) 1 14N nuclei, and the electron is then further coupled to β-carbon protons to give the final observed patterns in their X-band EPR spectra. The hyperfine coupling constants for the radical dianions [RC(NO2)2]2-• where R is H, Me, or Et correspond to prior results (29, 30), for anions prepared by purging aci-nitroalkanes with nitrogen dioxide, UV radiation of nitrite in aci-nitromethane (25), or glucose reduction of the monoanion HC(NO2)2-• (29). The new nitropropane derivative where R is Et has g ) 2.005 and is characterized by the following hyperfine coupling constants: aN ) 9.96 G and aH ) 1.85 G. These radicals are not observed when the aci-nitroalkanes are treated with gross excesses of nitrite, nitrate, or hydrogen peroxide, or combinations thereof. Moreover, in the presence of a 10-fold excess of added 15NO2-, there is no 15N label incorporated into the radical dianion. This rules out a ter-Meer reaction mechanism, where geminal substitution occurs by way of one-electron abstraction from RHCdNO2- to give RHCdNO2• before nucleophilic addition of nitrite leads to the product. Clearly, the source of the second introduced nitro group is peroxynitrite. Figure 2 shows the time course for the intensity of the 3438 G band in the X-band EPR spectrum during the carbon dioxide-catalyzed nitration of nitroethane by ONOO-. Provided that air is rigorously excluded, the solutions of [RC(NO2)2]2-• are stable for prolonged periods. When this reaction is followed by UV-vis spectroscopy, there is a loss of peroxynitrite absorbance at 302
Figure 3. (A) Under anaerobic conditions, UV-visible spectra following the decomposition of 65 µM peroxynitrite in the presence of 2 mM nitroethane at 22 °C and pH 10 with adventitious levels of carbon dioxide. (B) Same reaction as described for panel A but performed after an oxygen purge of the solution.
nm, which under anaerobic conditions (Figure 3A) is the only observed change between 280 and 500 nm. However, under conditions of oxygen saturation, this change is accompanied by the isosbestic rise in a band at 380 nm (Figure 3B) due to the formation of the diamagnetic monoanion [EtC(NO2)2]-. Therefore, under anaerobic conditions, the monoanion is not formed, but under aerobic conditions, these reactions provide a facile method for preparing geminal dinitroalkanes. For example, the reaction between peroxynitrite and nitroethane has an unoptimized yield of 68% CH3C(NO2)2- based on nitroethane. Prior syntheses of geminal dinitroalkanes have relied on variations of the ter-Meer reaction and have employed silver nitrite or ferrocyanide salts, and in general return between 90 and 64% yields (31-33). The mechanism for the formation of [RC(NO2)2]2-• can either involve the well-known rapid direct coupling of RHCdNO2- with nitrogen dioxide (eq 3) or by a direct carbon dioxide-catalyzed reaction of ONOO- with RHCd NO2- (eqs 5 and 6).
The rate law for the loss of peroxynitrite under these conditions is first-order in peroxynitrite and independent
Communications
Chem. Res. Toxicol., Vol. 13, No. 10, 2000 965
of nitroalkane concentration (Figures S1 and S2 in the Supporting Information). For example, for mole ratios of nitroethane to peroxynitrite between 2:1 and 7300:1, the rate of reaction is uniform. Thus, decomposition of peroxynitrite is rate-determining and the subsequent nitration and oxidations steps are fast (34-37). While the nitrogen dioxide is efficiently spin trapped in this system, and leads to the observed stable radicals [RC(NO2)2]2-•, the carbonate radical has not been directly observed in this study. Although static solution methods have failed to give detectable EPR spectra of CO3-• (38), the recent use of flow EPR systems has allowed for the detection of weak signals with a giso of 2.013 which are attributed to this radical following ONOOCO2- decomposition (21, 22). Reported attempts to spin trap carbonate radicals with RHCdNO2- have been frustrated by the low efficiency of the trapping reaction, and the apparent instability of the radical RHC(OCO2-)NO2-•, which has EPR characteristics similar to those of RHC(O-)NO2-• (30). When high concentrations of peroxynitrite, carbonate, and nitroalkane are employed, we observe a transient-doubled 1:1:1 triplet EPR spectrum (g ) 2.005, aH ) 3.5 G, and aN ) 25.6 G), which corresponds to the presence of [MeHC(O)(NO2)]2-• or its carbonate ester. This spectrum disappears rapidly (
