The rates of reaction of the sulfhydryl groups of human hemoglobin.

The hypothesis that the combination of a ligand with hemoglobin is associated with changes in the conformation of the molecule (l-3) is at present substantiated by considerable evidence (summarized by Rossi Fanelli, Antonini, and Caputo (4), and by Benesch and Benesch (5). It is, however, generally assumed that the changes in conformation which take place on combination with a ligand are independent of the type of ligand (4). This assumption is derived from the following data: (a) the uniformity of behavior in the reaction with bromthymol blue of the osy, carbonmonoxy, ferri, ferrifluoride, and ferricyanide derivatives of hemoglobin, as compared to deoxyhemoglobin (6) ; (b) the identical behavior of the sedimentation coefficients for carbonmonoxyand oxyhemoglobin as a function of pH (4) ; (c) the similarity of the equilibrium reactions (and associated phenomena) of hemoglobin with O2 and CO (7-9); (d) the identical behavior of carbonmonoxyand oxyhemoglobin on ion exchange chromatography;’ (e) the immunochemical identity between these liganded forms of hemoglobin (11) ; (f) the evidence that oxy-, carbonmonoxy-, and ferrihemoglobin crystals are isomorphous (12). On the other hand, the relative rate constants for the different steps in the reaction of sheep hemoglobin with 02 are different from those in the react.ion with CO (13-15). Furthermore, Hughes (16) has found that methylmercury is bound three times more tightly to bovine oxyhemoglobin than to carbonmonoxyhemoglobin, and there also appear to be slight differences in the solubility properties between human oxyand carbonmonoxyhemoglobin (12). Finally, Cecil (17) has shown that the velocities of reaction of the unreactive -SH groups of ferrihemoglobin, hemoglobin, and carbonmonox3;hemoglobin with heavy metal reagents decrease in that order. This is not the expected sequence. Since the reactive cysteine residue of normal adult human hemoglobin (cysteine residue 93/3) (10) is located next to the histidine residue which is directly linked to the iron atom of the heme group (18) and consequently might be expected to reflect small conformational changes dependent on the type of ligand, I have studied the kinetics of reaction of this cysteine residue with several reagents. On the basis of the differences in the velocities of reaction of the reactive -SH groups of human oxyhemoglobin and carbonmonoxyhemoglobin, I present in this paper the argument