Refolding of denatured thioredoxin observed by size-exclusion chromatography

Publisher Website
Google Scholar
Abstract
Molecular sieve chromatography can resolve interactive systems into populations having different effective hydrodynamic volumes. In this report, the advantages of such resolution to protein folding are illustrated by using moderate pressure to decrease analysis time and lowered temperature to slow down the kinetics of conformational change. A 300-mm Bio-Sil TSK-125 size-exclusion column was equilibrated with a series of different concentrations of guanidine hydrochloride at 2.degree. C in 50 mM phosphate buffer, pH 7.0. Samples of native Escherichia coli thioredoxin, denatured thioredoxin, or thioredoxin equilibrated with the column solvent were injected, and the effluent was monitored at 220 nm. Injection of equilibrated protein samples defined three denaturant concentration zones identical with those observed by spectral measurements: the native base-line zone where only compact protein is observed in the effluent profile; the transition zone in which both compact and denatured forms are observed in slow exchange; and the denatured base-line zone in which only denatured protein is observed. Unfolding was observed by injection of native protein into columns having isocratic denaturant concentrations in the transition and denatured base-line zones. Effluent profiles indicated a dynamic converison of compact to denatured protein with a time constant which appeared to decrease markedly with increasing denaturant concentration. Refolding was observed by injection of denatured protein into columns having isocratic concentrations in the transition and native base-line zones. As the denaturant concentration was decreased, the effluent profiles evidenced a persistent slow conversion of denatured to compact protein which was suddenly accelerated about midway in the native base-line zone. Separate multimixing measurements indicated that the slow refolding protein was generated in the denatured state and that the compact products of the accelerated refolding in the native base-line zone were less stable than the native protein. The observed effluent profiles were simulated by using a computer program based on the equations of Endo et al. [Endo, S., Saito, Y., and Wada, A. (1983) Anal. Biochem. 131, 108-120]. Such simulations indicate that population of partially folded conformations having an elution time intermediate between that of the native and denatured thioredoxin is not obligatory. All the observed profiles can be simulated by assuming that the denatured protein is dominated by molecules containing one or more nonnative configurational isomers. At relatively high denaturant concentrations in the native base-line zone, only the denatured molecules containing the native configurational isomers appear to be able to fold. However, at lower denaturant concentrations, the denatured molecules containing nonnative configurational isomers appear to be able to fold into stable compact conformations in which configurational isomerization to the native protein can subsequently occur.