Ultrarapid mixing experiments shed new light on the characteristics of the initial conformational ensemble during the folding of ribonuclease A

Abstract

The earliest folding events in single-tryptophan mutants of RNase A were investigated by fluorescence measurements by using a combination of stopped-flow and continuous-flow mixing experiments covering the time range from 70 micros to 10 s. An ultrarapid double-jump mixing protocol was used to study refolding from an unfolded ensemble containing only native proline isomers. The continuous-flow measurements revealed a series of kinetic events on the submillisecond time scale that account for the burst-phase signal observed in previous stopped-flow experiments. An initial increase in fluorescence within the 70-micros dead time of the continuous-flow experiment is consistent with a relatively nonspecific collapse of the polypeptide chain whereas a subsequent decrease in fluorescence with a time constant of approximately 80 micros is indicative of a more specific structural event. These rapid conformational changes are not observed if RNase A is allowed to equilibrate under denaturing conditions, resulting in formation of nonnative proline isomers. Thus, contrary to previous expectations, the isomerization state of proline peptide bonds can have a major impact on the structural events during early stages of folding.

Fig. 1.

Ribbon diagram of the structure of WT RNase A (12). Helices are labeled H1-H3; β-strands are labeled S1-S7. Also shown are the side chain of Tyr-92 (site of Trp substitution) and the four disulfide bonds (brackets).

Fig. 2.

Folding kinetics of Y92W RNase A. Shown are fluorescence changes during double-jump refolding of GuHCl-denatured Y92W RNase A at various GuHCl concentrations in CF (submilliseconds) and SF (milliseconds to seconds) experiments. The fluorescence intensity, normalized to the transiently unfolded condition (i.e., the 4.2 M GuHCl, pH 2, condition is 1.0), is given in arbitrary units. The solid lines represent a sum of fitted exponentials. The arrow on the ordinate represents the value for WdmPN at 0.7 M GuHCl that is referenced to the unfolded condition (4.2 M GuHCl, pH2). The burst phase at 0.7 M GuHCl is the difference between the value represented by the arrow and the initial value of the observed signal of the 0.7 M trace. The dead time of the CF measurements was 70 μs. The low GuHCl-concentration range is expanded in the Inset.

Fig. 5.

Submillisecond folding kinetics of the Y92W and Y92Wdes proteins, and the respective trace of the WdmPN peptide. Fluorescence change during refolding of GuHCl-denatured proteins and WdmPN at 0.7 M GuHCl in CF measurements. The fluorescence is normalized to a relative value of 1.0 for the corresponding unfolded conditions (2.9 M GuHCl for Y92Wdes RNase A in double-jump experiments and for WdmPN and 4.2 M GuHCl for Y92W in both single and double-jump experiments). Note that the unfolding conditions are different in the double-jump experiments with Y92W RNase A (Fig. 2) and Y92Wdes (here). Therefore, the fluorescence values of the WdmPN peptide that are scaled relative to their respective unfolding conditions are different in Figs. 2 and 5. The arrow indicates the fluorescence of the WPN-peptide, which is a control for the initial value expected for refolding from the equilibrated unfolded state, U. Traces are: thin solid line, WdmPN-peptide; open squares, U_vf of Y92Wdes RNase A; filled squares, U of Y92W RNase A. The dead time in these experiments is 120 μs. For easier comparison of the kinetics of the proteins under different conditions, the 0.7 M GuHCl trace for U_vf of Y92W RNase A (open circles), truncated at 120 μs, are also shown from Fig. 2. The kinetic traces are reproducible to within a 1–2% of the reference signal.

Fig. 3.

Rates (a) and amplitudes (b) of the various kinetic phases that are fitted to the traces obtained by the CF and SF measurements shown in Fig. 2, as a function of the final GuHCl concentrations. The error bars in b indicate standard fitting errors for amplitudes; except for the initial amplitude at 0.7 M GuHCl, all errors are <5%. The error bars for the rates (not shown) are comparable to or smaller than the size of the symbols.

Fig. 4.

CF fluorescence traces measured in double-jump refolding experiments on Y92W RNase A at a final concentration of 0.7 M GuHCl (a) and 2.5 M GuHCl (b). Residuals obtained by double-exponential and single exponential-functions are shown above each trace.