Vekshin has recently claimed [l] that the quenching of protein fluorescence observed on mixing horse-liver alcohol dehydrogenase with NAD' (or NAD+ plus an alcohol) is mainly due to conversion of NAD' to NADH with oxidation of traces of endogenous ethanol. The implication is that complexes containing " 3 ' have little or no quenching. Although traces of ethanol could produce artefacts, I will show that many published experiments do provide evidence for quenching in enzyme. NAD' or enzyme. NAD' alcohol complexes.Stopped-flow fluorimetry was used by Shore et al.[2] to demonstrate quenching by NAD'. In these experiments enzyme was mixed with NAD', or NAD' plus an alcohol, in the fluorimeter and detection of the fluorescence change started within a few milliseconds. A rapid phase of proteinfluorescence quenching was observed with a rate constant of 300 -SO0 s-', followed by two slower phases when an alcohol was present. This rate constant is faster than that for hydride transfer from ethanol to NAD ' (130 s-', [3]). Therefore the rapid phase of quenching preceded production of bound NADH and quenching must have occurred in the enzyme . NAD' or enzyme . NAD' . alcohol complex.Similar experiments were carried out with the carboxymethylated enzyme [4]. In this case hydride transfer is very slow (less than 0.2 s-', [S]) but protein fluorescence quenching was observed, with a rate constant of 270 s-', when the modified enzyme was mixed with NAD' and ethanol [4]. Again the enzyme . NAD' . ethanol complex must have quenched fluorescence.The quenching of protein fluorescence in mixtures of enzyme, NAD + and trifluoroethanol has been reported for native [6 -81 and carboxymethylated [4] alcohol dehydrogenase. Trifluoroethanol binds tightly to the enzyme [9, 101 but is not oxidised. Saturating concentrations of trifluoroethanol, as used in these experiments, would displace traces of ethanol from the enzyme. NADH would not be formed unless enzyme was mixed with NAD+ for some time before addition of trifluoroethanol. This was clearly not the case in titrations with NAD' [7, 81. In other experiments the order of mixing was not specified but it is unlikely that trifluoroethanol was added last in every experiment. Therefore these experiments conclusively show quenching of protein fluorescence in the enzyme. NAD' . trifluoroethanol complex.The possibility of artefacts due to traces of ethanol has been apparent for some time. An NADH recycling system (lactate dehydrogenase and pyruvate) was described by Shore et al. [ l l ] and used for fluorescence titrations on carboxymethylated enzyme [4] and relaxation kinetic studies of NAD' binding to the native enzyme [12]. Changes in protein fluorescence were observed in the presence of the recycling system. An alternative method for limiting production of NADH is addition of acetaldehyde to mixtures of alcohol dehydrogenase and NAD' [I 31; the protein fluorescence quenching was less than in the absence of acetaldehyde but was still significant. In fact Vekshin's own data (Fig. S of [l]) ...