Measurement of nitric oxide levels in the red cell: validation of tri-iodide-based chemiluminescence with acid-sulfanilamide pretreatment.

The tri-iodide-based chemiluminescence assay is the most widely used methodology for the detection of S-nitrosothiols (RSNOs) in biological samples. Because of the low RSNO levels detected in a number of biological compartments using this assay, criticism has been raised that this method underestimates the true values in biological samples. This claim is based on the beliefs that (i) acidified sulfanilamide pretreatment, required to remove nitrite, leads to RSNO degradation and (ii) that there is auto-capture of released NO by heme in the reaction vessel. Because our laboratories have used this assay extensively without ever encountering evidence that corroborated these claims, we sought to experimentally address these issues using several independent techniques. We find that RSNOs of glutathione, cysteine, albumin, and hemoglobin are stable in acidified sulfanilamide as determined by the tri-iodide method, copper/cysteine assay, Griess-Saville assay and spectrophotometric analysis. Quantitatively there was no difference in S-nitroso-hemoglobin (SNOHb) or S-nitroso-albumin (SNOAlb) using the tri-iodide method and a recently described modified assay using a ferricyanide-enhanced reaction mix at biologically relevant NO:heme ratios. Levels of SNOHb detected in human blood ranged from 20-100 nM with no arterial-venous gradient. We further find that 90% of the total NO-related signal in blood is caused by erythrocytic nitrite, which may partly be bound to hemoglobin. We conclude that all claims made thus far that the tri-iodide assay underestimates RSNO levels are unsubstantiated and that this assay remains the "gold standard" for sensitive and specific measurement of RSNOs in biological matrices.