Determination of S-nitrosoglutathione and other nitrosothiols by p-hydroxymercurybenzoate derivatization and reverse phase chromatography coupled with chemical vapor generation atomic fluorescence detection

Abstract

S-Nitrosoglutathione (GSNO) reacts with the organic mercurial probe, p-hydroxymercury benzoate (PHMB, HO–Hg–(C 6H 4)–COO −Na +) giving the complex GS–Hg(C 6H 4)COOH (GS–PHMB). This reaction has been studied by UV measurements at 334 nm also in the presence of ascorbic acid and the product of reaction, the GS–PHMB complex, characterized by Electrospray Ionization Mass Spectrometry (ESI-MS) and by Reversed Phase Chromatography (RPLC) coupled on-line and sequentially with a UV–visible diode array detector (DAD) followed by a cold vapor generation atomic fluorescence spectrometer (CVGAFS). The simultaneous presence of PHMB and ascorbate produced a synergistic effect on GSNO decomposition rate that can be observed only above a given concentration threshold of ascorbate (ascorbate/GSNO molar ratio ≥ 180). The results indicated that the formation of GS–PHMB, both in the presence and the absence of ascorbic acid, does not involve the formation of free thiolic species but it takes place through a more complex mechanism. The PHMB derivatives of GSH and GSNO obtained by the present method were found to be identical by ESI-MS. GSSG did not interfere because it was not reduced and derivatized to GS–PHMB. Once complexed by the alkylating agent N-ethylmaleimide (NEM), GSH did not interfere with the derivatization reaction. This ensured a good selectivity of the developed PHMB derivatization system for RSNO determination. Thus, we have optimized the operating conditions for the selective reaction of PHMB with GSNO and other nitrosothiols (RSNOs) in order to determinate RSNOs in human plasma. LODc for RSNOs in plasma ultrafiltrate was 30 nM (injected concentration, 50 μL loop), the DLR ranged between 0.08 and 50 μM and the CV% was 6.5% at 300 nM concentration level. Reduced and oxidized thiols spiked to plasma did not interfere with the measurement of RSNOs. We found that the sampling procedure was critical for the recovery of endogenous and spiked RSNOs. The ultrafiltrate samples of plasma of 8 healthy humans contained 1460 ± 310 CysNO, 1000 ± 330 nM HCysNO and 320 ± 60 nM GSNO if blood was sampled in a mixture NEM/ethylendiaminotetracetic acid (EDTA)/serine–borate complex (SBC), where serine–borate complex is a potent inhibitor of γ-glutamyltransferase, an enzyme involved in the conversion of GSNO into CysGlyNO. In the absence of SBC during the sampling of blood GSNO concentration found in the ultrafiltrate was lower (at level of the determination limit in plasma ultrafiltrate, i.e. 75 nM) and the peak of CysGlyNO appeared, which corresponded to a concentration of 200 ± 60 nM ( N = 4 blood samples).