Integrated In Vivo and In Silico Studies of Cotranslational Protein Folding as a Function of Translation Rate

Abstract

In order for a protein to exhibit its proper function, it must first fold correctly. Most of what we know about protein folding comes from in vitro studies of small, single-domain proteins. However, larger and more complex proteins often fail to fold efficiently in vitro, yet fold to high yield in cells. One striking difference between protein folding in vitro and in vivo is how folding begins. In vitro, protein folding begins from an enormous ensemble of random conformations of full length polypeptide chains, while in vivo a protein can begin to fold as it is synthesized by the ribosome. Vectorial appearance of the nascent protein chain during translation is a universal feature of every protein in the proteome, yet its effect on protein folding is poorly understood. Our lab has created YKB, a protein construct with two mutually exclusive native states (YK-B and Y-KB), to study the effects of translation on folding mechanisms. Synonymous mutations to YKB alter the ratio of the YK-B and Y-KB native structures, yet have identical protein sequences. From this we can infer that vectorial appearance affects YKB folding, but the ratios provide little molecular-level detail. We have developed a coarse-grain computational model to simulate YKB cotranslational folding. This model accurately reproduces experimental results for YKB refolding in vitro and co-translational folding in vivo. These simulations provide us with details that cannot be learned through experiments, including specific molecular trajectories and the effects of altering translation rate on the YKB folding mechanism. Integrating information gleaned from simulations with targeted experimental studies will enable unprecedented insight into in vivo folding mechanisms, particularly the effects of varying translation rate on cotranslational folding and overall folding efficiency.