Protein folding: lessons learned and new frontiers

Abstract

In appropriate physiological milieux proteins spontaneously fold into their functional three-dimensional structures. The amino acid sequences of functional proteins contain all the information necessary to specify the folds. This remarkable observation has spawned research aimed at answering two major questions. (1) Of all the conceivable structures that a protein can adopt, why is the ensemble of native-like structures the most favorable? (2) What are the paths by which proteins manage to robustly and reproducibly fold into their native structures? Anfinsen's thermodynamic hypothesis has guided the pursuit of answers to the first question whereas Levinthal's paradox has influenced the development of models for protein folding dynamics.Decades of work have led to significant advances in the folding problem. Mean-field models have been developed to capture our current, coarse grain understanding of the driving forces for protein folding. These models are being used to predict three-dimensional protein structures from sequence and stability profiles as a function of thermodynamic and chemical perturbations. Impressive strides have also been made in the field of protein design, also known as the inverse folding problem, thereby testing our understanding of the determinants of the fold specificities of different sequences.Early work on protein folding pathways focused on the specific sequence of events that could lead to a simplification of the search process. However, unifying principles proved to be elusive. Proteins that show reversible two-state folding–unfolding transitions turned out to be a gift of natural selection. Focusing on these simple systems helped researchers to uncover general principles regarding the origins of cooperativity in protein folding thermodynamics and kinetics. On the theoretical front, concepts borrowed from polymer physics and the physics of spin glasses led to the development of a framework based on energy landscape theories. These theories predict that evolved sequences (functional proteins as opposed to random sequences) find their native folds by minimizing geometric (topological) frustration (i.e. avoiding entropic bottlenecks/kinetic traps). In some cases, following a dominant pathway is the optimal way to minimize frustration, whereas in extreme cases, proteins may fold without encountering bottlenecks. Experimental studies of two-state proteins led in turn to the development of quantitative descriptors that have allowed specific testing of theoretical predictions. These include methods such as phi value analysis to characterize transition state ensembles and descriptors that measure the effects of geometry/topology on folding rates. Interestingly, there exists a striking inverse correlation between the relative contact order (the distance in sequence space between spatially proximal contacts made in the native state) and the folding rates of several two-state proteins. The relative contact order provides a rough estimate of the net entropic cost associated with realizing the folded state, and theories have been developed to explain the observed correlation between the contact order and folding rates.Despite its maturity as a field, there are several areas that come under the rubric of protein folding that are just beginning to receive attention. For example, how do complications in vivo such as macromolecular crowding, confinement, the presence of cosolutes, membrane anchoring, and tethering to surfaces influence protein stabilities and folding dynamics? While we are accustomed to studying proteins at concentrations that are amenable to investigation via probes whose signal intensities grow with protein concentration, this does not make these readouts relevant to the in vivo setting. In cells, protein concentrations are tightly regulated and are likely to be orders of magnitude lower than what we are accustomed to using within in vitro experimental setups. Protein folding in vivo is a complex multi-scale dynamical problem when one considers the synergies between protein expression, spontaneous folding, chaperonin-assisted folding, protein targeting, the kinetics of post-translational modifications, protein degradation, and of course the drive to avoid aggregation. Further, there is growing recognition that cells not only tolerate but select for proteins that are intrinsically disordered. These proteins are essential for many crucial activities, and yet their inability to fold in isolation makes them prone to proteolytic processing and aggregation.In the series of papers that make up this special focus on protein folding in physical biology, leading researchers provide insights into diverse cross-sections of problems in protein folding. Barrick provides a concise review of what we have learned from the study of two-state folders and draws attention to how several unanswered questions are being approached using studies on large repeat proteins. Dissecting the contribution of hydration-mediated interactions to driving forces for protein folding and assembly has been extremely challenging. There is renewed interest in using hydrostatic pressure as a tool to access folding intermediates and decipher the role of partially hydrated states in folding, misfolding, and aggregation. Silva and Foguel review many of the nuances that have been uncovered by perturbing hydrostatic pressure as a thermodynamic parameter. As noted above, protein folding in vivo is expected to be considerably more complex than the folding of two-state proteins in dilute solutions. Lucent et al review the state-of-the-art in the development of quantitative theories to explain chaperonin-assisted folding in vivo. Additionally, they highlight unanswered questions pertaining to the processing of unfolded/misfolded proteins by the chaperone machinery. Zhuang et al present results that focus on the effects of surface tethering on transition state ensembles and folding mechanisms of a model two-state protein. Their results are important because several proteins in vivo fold while being anchored to membranes. Finally, several neurodegenerative and systemic diseases are associated with the aggregation of intrinsically disordered polypeptides. The search for cures in these debilitating and fatal diseases has focused attention on shared attributes in aggregation mechanisms of different proteins and the possibility of identifying druggable targets from mechanistic studies. Abedini and Raleigh review common features gleaned from mechanistic studies of the aggregation of several intrinsically disordered proteins. They propose that the population of helical intermediates and their stabilization via interactions with membranes might be an important route by which the process of aggregation leads to toxicity.The five papers that form this protein folding focus cover specific sub-topics within the larger field of protein folding. They address current questions and emphasize the importance of the growing and productive interface between the physical sciences and biology. We hope that these papers will stimulate much discussion and more importantly advances in the areas highlighted by the contributors.