A modified technique for the measurement of sulfhydryl groups oxidized by reactive oxygen intermediates

This paper suggests a simple modification of the Ellman procedure when used to measure accurate changes in sulfhydryl (-SH) content induced by reactive oxygen intermediates (ROI). This modification became necessary when we found that the standard technique did not produce time invariant results in the presence of ROI-generating systems. Cysteine (cys; 20–100 μM) in 20 mM imidazole buffer (pH 7.0) containing 1.0 mM EDTA was reacted with excess (0.2 mM) 5,5′-dithiobis(2-nitrobenzoic acid), DTNB. The absorbance of the product (p-nitrothiophenol anion) was recorded at 412 nm (A₄₁₂). This A₄₁₂ was stable for 60 min and gave a linear relationship with cys concentrations used. ROI were generated either by 0.01 U xanthine oxidase (XO) + 0.01–1.0 mM hypoxanthine (HX), 0.01–1.0 mM H₂O₂, or H₂O₂ + 100 μM FeSO₄. In the presence of ROI, A₄₁₂ decreased with time and its rate of decrease was dependent upon the concentration of components of the ROI-generating system. This time-dependent decrease in A₄₁₂ was prevented completely by the addition of 100 U of catalase (CAT). Therefore, we modified the DTNB method as follows: -SH groups were reacted with ROI for 30 min; this was followed by the addition of 100 U of CAT to scavenge the excess unreacted ROI before the addition of DTNB to generate the product. Using this modification the ROI-induced decrease in A₄₁₂ was stable with time and was linearly related to the cys concentration. We further tested the modified procedure using metallothionein (MT) as a substrate for the ROI-induced changes in -SH content. MT, at concentrations of 2.5, 5.0, and 7.5 μM, was treated with XO + 100 μM HX. Using the modified procedure, an average decrease (as compared to the untreated control) of 15, 22, and 33 μM in -SH content was observed consistently at the respective MT concentrations. However, without the modification in the procedure, these average decrease were 20, 38, and 51 μM, respectively and continued to further increase with time. These discrepancies could give rise to errors ranging from 28 to 35% or higher in determination of the ROI-induced decrease in the -SH groups of MT. This data suggests that scavenging the unreacted H₂O₂ with C prior to the addition of DTNB to the assay mixture gives a stable and accurate estimate of the ROI-induced oxidative damage to -SH groups.