Probing heme protein conformational equilibration rates with kinetic selection

Double-pulse flash photolysis experiments on solutions of carbonmonoxymyoglobin (MbCO) are used to determine the time scale for protein conformational averaging. The interconversion times for transitions between the ''open'' and ''closed'' subpopulations of MbCO are found to be 10(-6)-10(-4) s, depending on solvent composition and temperature. In aqueous solution at 273 K, the interconversion rate is found to be 1.4 x 10(6) s. Since the interconversion rate is comparable to or slower than the geminate rebinding rate, we describe the geminate phase of the kinetics as a superposition of contributions from the open and closed states. Although geminate kinetics remain intrinsically nonexponential for both open and closed states near room temperature, we find that substates within these two subpopulations interconvert more rapidly than the geminate rebinding. These observations cannot be explained by a superposition of contributions from a quasicontinuous conformational distribution (Steinbach et al., 1991) and are probably due to the long-time tail of the relaxation of the protein (Tian et al., 1992). Bimolecular rebinding takes place at a statistically averaged rate, since the interconversion and relaxation rates are faster than the bimolecular kinetics. The geminate and bimolecular kinetics are analyzed quantitatively as a function of pH using this approach and the spectroscopically determined populations of the open and dosed states. The analysis accounts for the observed kinetics and also successfully predicts the kinetic response observed in the double-pulse experiments. In aqueous solution at 273 K, the geminate amplitudes and rates are found to be I-g(0) = 32% and k(g)(0) = 1.3 x 10(7) s(-1) for the open state and I-g(1) = 9.3% and k(g)(1) = 1.4 x 10(6) s(-1) for the closed state. In 75% glycerol solution at 264 K, the dominant component of the geminate rebinding is characterized by I-g1(0) = 89% and k(g1)(0) = 3.1 x 10(6) s(-1) for the open state and I-g1(1) = 26% and k(g1)(1) = 3.1 x 10(6) s(-1) for the closed state. The fact that the interconversion rate is comparable to the geminate rate of the closed state in aqueous solution is consistent with the idea that the open state provides an important pathway for ligand escape from (or entry to) the heme pocket (Tian et al., 1993). The increased viscosity of 75% glycerol solution delays the closed --> open interconversion until the end of the geminate phase, which forces the ligand to find alternative pathways to the solution. This observation, in conjunction with the near equivalence of the geminate rates for the open and closed states in 75% glycerol solution, suggests that the solvent composition fundamentally alters the protein-ligand dynamics.