Hydroxyl radical generation: the effect of bicarbonate, dioxygen and buffer concentration on pH-dependent chemiluminescence

Abstract

Enhanced chemiluminescence (CL) induced by HO^· was detected using the primary enhancers luminol (Lum), isoluminol (ILum) and lucigenin (Luc), in the absence of HCO₃^-, at pH 9.5, 10.0 and 10.5 but not between pH 4.0 and 9.0. This was confirmed using nine different HO^· generators. FeCl₃/NTA/H₂O₂ was the only HO^· generator that was able to generate singular HO^· which was obtained entirely from the Fenton reaction. However, this was so only at pH 10.0 since at all other pHs multiple ROS were produced. This was confirmed by the chemical detection of the fluorescent hydroxylated product of terephthalic acid in the absence of O₂. No HO^· Lum-CL pH optima coincided with the O₂^·- mediated Lum-CL optima found at pH 8.0 and 9.0. Bicarbonate had an enhancing effect on Lum-CL which was 70–2700% at pH 10.0 for the different HO^· generators. This was due to the conversion of the radicalelectron from HO^· to CO₃^·-, making CL detection more efficient since less HO^· were lost initially before detection.

Methyl-cypridine-luciferin analogue (MCLA) elicited CL in the pH range 4.0–10.0 with the same set of generators in the absence of HCO₃^-. The iron-containing generators had their different MCLA-CL optima at pH 4.5, 5.0 or 6.0, excluding those overlapping with the O₂^·--mediated CL optima. The two copper-containing generators had optima at the same pHs, viz. 7.0 and 10.5. Again, FeCl₃/NTA/H₂O₂ was the only HO^· generator able to produce singular HO^· by the Fenton reaction. However, whereas Lum-CL was able to detect singular HO^· only at pH 10.0, MCLA-CL detected it at pH 5.0 and 5.5. Therefore, MCLA is the most suitable CL enhancer for physiological assessments since it is the most sensitive enhancer and has HO^· CL optima nearer to physiological pH than the other probes. The HCO₃^- enhancement of MCLA-CL was even greater than that of Lum-CL, since increases of 114-fold and 37-fold, respectively, were obtained at these HO^·-specific pH optima for FeCl₃/NTA/H₂O₂. Therefore, bicarbonate concentration is as important a parameter as pH when the enhanced CL of a non-cellular system is determined.

Hydrogen peroxide was not able to elicit CL directly but, due to trace metal contamination, it produced artifactual CL due to HO^· formation. High H₂O₂ levels, which prevent spontaneous O₂^·- dismutation, helped to establish the overlapping pH optima of CL mediated by O₂^·- and HO^· which were artifactually produced either by O₂^·- via H₂O₂ and trace metals or by perferryl intermediates, respectively. Due to spontaneous dismutation to H₂O₂, only 22% of the O₂^·- produced by HX/XO could be detected by enhanced CL.