Nanosecond lasers were used to measure the rate of conformational changes in myoglobin after ligand dissociation at ambient temperatures. At low solvent viscosities the rate is independent of viscosity, but at high viscosities it depends on approximately the inverse first power of the viscosity. Kramers theory for unimolecular rate processes can be used to explain this result if the friction term is modified to include protein as well as solvent friction. The theory and experiment suggest that the dominant factor in markedly reducing the rate of conformational changes in myoglobin at low temperatures (less than 200 K) is the very high viscosity (greater than 10(7) centipoise) of the glycerol-water solvent. That is, at low temperatures conformational substates may not be "frozen" so much as "stuck."

Initiation of protein folding by light can dramaticafly improve the time resolution of kinetic studies. Here we present an example ofan optically triggered folding reaction by using nanosecond photodissociation of the heme-carbon monoxide complex of reduced cytochrome c. The optical trigger is based on the observation that under destabilizing conditions cytochrome c can be unfolded by preferential binding of carbon monoxide to the covalently attached heme group in the unfolded state. Photodissociation of the carbon monoxide thus triggers the folding reaction. We used time-resolved absorption spectroscopy to monitor binding at the heme. Before folding begins we observe transient binding of both nonnative and native ligands from the unfolded polypeptide on a microsecond time scale. Kinetic modeling suggests that the intramolecular binding of methionine-65 and -80 is faster than that of histidine-26 and -33, even though the histidines are doser to the heme. This optical trigger should provide a powerful method for studying chain collapse and secondary structure formation in cytochrome c without any limitations in time resolution.Protein folding generally occurs in two phases, one rapid and one slow. The rapid phase is the collapse of the unfolded polypeptide into a compact structure and the formation of secondary structure. The slow phase is the subsequent, often multistep rearrangement to the native conformation (1-7). Because of the limited time resolution (milliseconds) of conventional stopped-flow mixing experiments, the complete time course of protein folding has not yet been observed. A dramatic improvement in time resolution would result if protein folding could be initiated by light.Here we present an example of an optically triggered folding reaction. We take advantage of the fact that under destabilizing conditions cytochrome c can be unfolded by preferential binding of carbon monoxide (CO) to the covalently attached heme group in the unfolded state. The folding reaction can thus be triggered by photodissociating the CO complex. In this study, we used time-resolved absorption spectroscopy with nanosecond lasers to monitor binding events at the heme (8-11). Although rebinding of CO prevents the complete formation of the native conformation, the rapid recovery of the sample permits repetitive photolysis and therefore the acquisition of high signal/noise transient spectra for investigating submillisecond events. Simulations of the multiwavelength data with kinetic models were carried out to generate the spectra of intermediates, as well as the rate constants connecting them. We find that before folding begins there is transient binding of both nonnative and native ligands from the unfolded polypeptide chain on a microsecond time scale. MATERIALS AND METHODSTye VI horse heart cytochrome c from Sigma was purified by ion-exchange chromatography using carboxymethyl-cellulose (Whatman) (12), reduced with sodium dithionite anaerobically, and separated anaerobically from the sodium dithionite by gel filtrat...

Absorption spectroscopy with nanosecond time resolution shows that myoglobin undergoes conformational relaxation on the same time scale as geminate rebinding of carbon monoxide. Ligand rebinding following photodissociation of the heme-CO complex was measured from the amplitude of the average difference spectrum, while conformational changes were measured from changes in the detailed shape of the Soret spectra of the deoxyhemes. Experiments in which the solvent viscosity was varied between 1 and 300 cP and the temperature between 268 and 308 K were analyzed by fitting the multiwavelength kinetic data with both empirical and molecular models. Novel numerical techniques were employed in fitting the data, including the use of singular value decomposition to remove the effects of temperature and solvent on the spectra and of a Monte Carlo method to overcome the multiple minimum problem in searching parameter space. The molecular model is the minimal model that incorporates all of the major features of myoglobin kinetics at ambient temperatures, including a fast and slow rebinding conformation and two geminate states for each conformation. The results of fitting the kinetic data with this model indicate that the geminate-rebinding rates for the two conformations differ by at least a factor of 100. The differences between the spectra of the two conformations generated from the fits are similar to the differences between those of the R and T conformations of hemoglobin. In modeling the data, the dependence of the rates on temperature and viscosity was parametrized using a modification of Kramers theory which includes the contributions of both protein and solvent to the friction. The rate of the transition from the fast to the slow rebinding conformation is found to be inversely proportional to the viscosity when the viscosity exceeds about 30 cP and nearly viscosity independent at low viscosity. The viscosity dependence at high viscosities suggests that the two conformations differ by the global displacement of protein atoms on the proximal side of the heme observed by X-ray crystallography. We suggest that the conformational change observed in our experiments corresponds to the final portion of the nonexponential conformational relaxation recently observed by Anfinrud and co-workers, which begins on a picosecond time scale. Furthermore, extrapolation of our data to temperatures near that of the solvent glass transition suggests that this conformational relaxation may very well be the one postulated by Frauenfelder and co-workers to explain the decrease in the rate of geminate rebinding with increasing temperature above 180 K.