Hot and Crowded: New Insights into the Dynamics of Thermophilic Enzymes from Multiscale Modeling

Abstract

Understanding how the amino acid sequence of a protein determines its three-dimensional structure, dynamics, and, ultimately, its biological function remains one of the most fundamental biophysical problems. Although this problem is unlikely to be solved in a single step, one way of breaking it down is offered to us by nature's diversity itself. Adaptation to extreme environmental conditions, e.g., high temperatures, led to the evolution of thermophilic enzymes that are not only stable at temperatures beyond 60°C, but also exhibit their maximum activity at elevated temperatures. Remarkably, these thermophilic enzymes still maintain a significant degree of homology to their mesophilic counterparts, enzymes that catalyze the same reactions in species that thrive at moderate temperatures. Studying mesophilic/thermophilic enzyme homologs from different species side-by-side enables an assessment of how moderate changes in sequence and structure alter the enzyme's properties toward higher thermostability and high activity at elevated temperatures. Such a comparison also provides additional insights into structure-function relationships in general. The requirements for thermostability and enzyme activity at elevated temperatures are still under debate. Thermophilic enzymes with an increased rigidity compared to their mesophilic counterparts are reported in the literature, although there are also reported examples of thermophilic enzymes with an increased flexibility. Recently, increased resilience, manifested as a weak temperature dependence of protein flexibility measured by elastic incoherent neutron scattering (EINS), has been proposed (1) as an additional characteristic of thermophilic adaptation. EINS is a versatile tool that allows quantification of the flexibility of a protein via mean-squared fluctuations of nonexchangeable hydrogen atoms on picosecond-nanosecond timescales. In powder samples in which translational and rotational diffusion of the molecule as a whole are suppressed, EINS probes internal motions, and hence protein flexibility, directly. However, accurate extraction of internal motions from EINS data on proteins in solution requires deconvolution of the contributions of internal dynamics and translational and rotational diffusion to the observed atomic motion. These problems have been addressed, e.g., via an effective diffusion term after comparison to measurements on hydrated protein powders (2) and by molecular dynamics (MD) simulation (3). When comparing the dynamics between proteins of similar mass, shape, and size, such as thermophilic/mesophilic pairs of enzymes, it is reasonable to assume very similar diffusive dynamics. Under such an assumption, differences in the overall dynamics can be assigned to differences in the internal motions/structural flexibility. However, under crowding conditions, e.g., at the high protein concentrations used in solution EINS experiments, as well as under physiological conditions in the cell, this assumption may break down due to specific or nonspecific interactions, which may have distinct effects on the diffusive motion of the compared proteins. Marcos et al. (4) push our understanding of this matter to a new level by combining Brownian dynamics of the crowded protein solution with atomistic MD simulations of the isolated protein. With this multiscale approach, they are able to cleanly separate the contributions to the scattering function and extracted mean-squared fluctuations from internal dynamics from the rotational and diffusive motion of the molecule as a whole. Marcos et al. (4) consider a pair of homologous thermophilic and mesophilic enzymes, malate dehydrogenase from Methanococcus jannaschii and lactate dehydrogenase from Oryctolagus cunniculus, which were studied previously using EINS experiments (1). Marcos et al. not only confirm the previously observed increased flexibility of the thermophilic protein compared to the mesophilic one, but they also find that distinct diffusional motion of the two proteins under crowding conditions contributes significantly to the experimentally observed differences in overall dynamics and their temperature dependence. Moreover, their simulations reveal that the greater number of charged residues on the surface of the thermophilic enzyme results in significant attractive interactions in the crowded solution, which effectively reduces the diffusional mobility. The lower dielectric constant of water at increasing temperature further increases the strength of these electrostatic interactions, and compensates the increase in diffusive motion with increasing temperature. The slower diffusive motion, in turn, compensates the increased internal motion in the thermophilic versus the mesophilic enzyme, and this translates into the higher resilience, namely the weaker increase in overall dynamics with increasing temperature, observed via the fluctuations of scattering hydrogen atoms for the thermophilic enzyme. The article by Marcos et al. (4) exposes a new fundamental difference between thermophilic enzymes and their mesophilic counterparts, which typically carry a less amount of charged residues on their solvent exposed surface (5). The authors speculate that the increased electrostatic interactions for thermophilic enzymes may actually result in diffusion properties at the enzymes' optimum temperature that are comparable to those of mesophilic enzymes at moderate temperatures. Since diffusion plays a crucial role in enzymatic catalysis, the tuning of diffusive motion to challenging environmental conditions might be a key aspect of the adaptation of extremophilic enzymes for optimal activity.