Search

We describe a protocol to perform empirical valence bond (EVB) simulations using GROMACS software. EVB is a fast and reliable method that allows one to determine the reaction free-energy profiles in complex systems, such as enzymes, by employing classical force fields to represent a chemical reaction. Therefore, running EVB simulations is basically as fast as any classical molecular dynamics simulation, and the method uses standard free-energy calculations to map the free-energy change along a given reaction path. To exemplify and validate our EVB implementation, we replicated two cases of our earlier enzyme simulations. One of these addresses the decomposition of the activation free energy into its enthalpic and entropic components, and the other is focused on calculating the overall catalytic effect of the enzyme compared to the same reaction in water. These two examples give virtually identical results to those obtained with programs that were specifically designed for EVB simulations and show that the GROMACS implementation is robust and can be used for very large systems.

Cold-adapted enzymes are characterized both by a higher catalytic activity at low temperatures and by having their temperature optimum down-shifted, compared to mesophilic orthologs. In several cases, the optimum does not coincide with the onset of protein melting but reflects some other type of inactivation. In the psychrophilic a-amylase from an Antarctic bacterium, the inactivation is thought to originate from a specific enzyme-substrate interaction that breaks around room temperature. Here, we report a computational redesign of this enzyme aimed at shifting its temperature optimum upward. A set of mutations designed to stabilize the enzyme-substrate interaction were predicted by computer simulations of the catalytic reaction at different temperatures. The predictions were verified by kinetic experiments and crystal structures of the redesigned a-amylase, showing that the temperature optimum is indeed markedly shifted upward and that the critical surface loop controlling the temperature dependence approaches the target conformation observed in a mesophilic ortholog.

The structural origin of enzyme cold-adaptation has been the subject of considerable research efforts in recent years. Comparative studies of orthologous mesophilic-psychrophilic enzyme pairs found in nature are an obvious strategy for solving this problem, but they often suffer from relatively low sequence identity of the enzyme pairs. Small bacterial lipases adapted to distinctly different temperatures appear to provide an excellent model system for these types of studies, as they may show a very high degree of sequence conservation. Here, we report the first crystal structures of lipase A from the psychrophilic bacterium Bacillus pumilus, which confirm the high structural similarity to the mesophilic Bacillus subtilis enzyme, as indicated by their 81% sequence identity. We further employ extensive QM/MM calculations to delineate the catalytic reaction path and its energetics. The computational prediction of a rate-limiting deacylation step of the enzymatic ester hydrolysis reaction is verified by stopped-flow experiments, and steady-state kinetics confirms the psychrophilic nature of the B. pumilus enzyme. These results provide a useful benchmark for examining the structural basis of cold-adaptation and should now make it possible to disentangle the effects of the 34 mutations between the two enzymes on catalytic properties and thermal stability.

The structural principles of enzyme cold adaptation are of fundamental interest both for understanding protein evolution and for biotechnological applications. It has become clear in recent years that structural flexibility plays a major role in tuning enzyme activity at low temperatures, which is reflected by characteristic changes in the thermodynamic activation parameters for psychrophilic enzymes, compared to those of mesophilic and thermophilic ones. Hence, increased flexibility of the enzyme surface has been shown to lead to a lower enthalpy and a more negative entropy of activation, which leads to higher activity in the cold. This immediately raises the question of how enzyme oligomerization affects the temperature dependence of catalysis. Here, we address this issue by computer simulations of the catalytic reaction of a cold-adapted bacterial short chain dehydrogenase in different oligomeric states. Reaction free energy profiles are calculated at different temperatures for the tetrameric, dimeric, and monomeric states of the enzyme, and activation parameters are obtained from the corresponding computational Arrhenius plots. The results show that the activation free energy, enthalpy, and entropy are remarkably insensitive to the oligomeric state, leading to the conclusion that assembly of the subunit interfaces does not compromise cold adaptation, even though the mobilities of interfacial residues are indeed affected.

It has been suggested that heat capacity changes in enzyme catalysis may be the underlying reason for temperature optima that are not related to unfolding of the enzyme. If this were to be a common phenomenon, it would have major implications for our interpretation of enzyme kinetics. In most cases, the support for the possible existence of a nonzero (negative) activation heat capacity, however, only relies on fitting such a kinetic model to experimental data. It is therefore of fundamental interest to try to use computer simulations to address this issue. One way is simply to calculate the temperature dependence of the activation free energy and determine whether the relationship is linear or not. An alternative approach is to calculate the absolute heat capacities of the reactant and transition states from plain molecular dynamics simulations using either the temperature derivative or fluctuation formula for the enthalpy. Here, we examine these different approaches for a designer enzyme with a temperature optimum that is not caused by unfolding. Benchmark calculations for the heat capacity of liquid water are first carried out using different thermostats. It is shown that the derivative formula for the heat capacity is generally the most robust and insensitive to the thermostat used and its parameters. The enzyme calculations using this method give results in agreement with direct calculations of activation free energies and show no sign of a negative activation heat capacity. We also provide a simple scheme for the calculation of binding heat capacity changes, which is of clear interest in ligand design, and demonstrate it for substrate binding to the designer enzyme. Neither in that case do the simulations predict any negative heat capacity change.