Buttressing a new paradigm in protein folding: experimental tools to distinguish between downhill and multi-state folding mechanisms

Resumen

Many single-domain proteins fold in milliseconds or longer. However, the advent of fast folding kinetic techniques has permitted to identify many other proteins that fold in the order of (few) microseconds and thus very closely to the folding speed limit. This suggests that the proteins that fold in microsecond timescale either cross a marginal single free energy barrier, multiple very small barriers (multi-state), or no barrier at all (downhill). This results in the potential observation of broad complex unfolding transitions in these ultrafast folding proteins (in contrast to simple two-state behavior). Many of the ultrafast folding proteins have small size and fold into simple alpha helix-bundle topologies. Theoretical studies support the size scaling of protein folding barriers. Engrailed homeodomain, a 61-residue ¿-helical domain with a helix-turn-helix topology folds in microseconds and exhibits an apparently complex (un)folding process. The observed complexity in the (un)folding behavior of engrailed homeodomain rules out a simple two-state model, but the folding mechanism of this protein has been interpreted with a conventional three-state model. The current work aims to develop a set of experimental and analytical methods that can determine unambiguously whether an apparently complex folding process of a fast folding protein is downhill or multi-state using engrailed homeodomain as a model. A large-scale multiple probe approach that combines equilibrium, fast-folding measurement and single molecule measurements has been used to provide critical information to unravel the mechanistic details of the folding mechanism of this protein. Double perturbation measurement on engrailed, in which the protein was unfolded by both chemical denaturant and temperature, showed complex results. Multi-probe equilibrium thermal and chemical unfolding measurements on engrailed revealed differences in the melting temperature and chemical denaturation midpoints respectively. All these signatures conformed to downhill folding mechanism or existence of low-barrier(s). The estimated overall barrier height was ~ 0.5 RT near Tm, by globally fitting the entire equilibrium thermal unfolding data to Mean Field Model. Multi-probe temperature jump studies resulted in single exponential relaxations by infrared and non-exponential relaxations by fluorescence and probe-dependent kinetic amplitudes for the slow rates. This result could still be explained by a downhill behavior by globally fitting both the equilibrium and the kinetic data using the same model. Single molecule FRET measurements explored the transition path of engrailed near Cm and further confirmed the existence of downhill behavior with the estimated marginal barrier of < 1 RT. These results emphasize the importance of multi-probe measurements and appropriate utilization of statistical mechanical for analysis for fast-folding proteins.