Proteins and Enzymes Research

Chemistry Professor Robert Cukier uses theoretical and computational methods to study the functions of proteins and other complex many-body systems. "MSU's High Performance Computing Center plays an important role in this research by providing the resources for large-scale computing," Cukier states.

He uses the methods of Statistical Mechanics and Quantum Mechanics to create theories of an carry out simulations on: electron (ET), proton (PT) and proton coupled electron (PCET) transfer, proton translocation, large-scale domain motion of proteins, and enzyme reactions mechanisms.

This research relies heavily on code development. He has developed CUKMODY, a molecular dynamics (MD) code oriented towards simulating proteins. Accelerating the exploration of configuration space is an outstanding problem in MD simulation. His methods for overcoming barriers in the high dimensional configuration space of proteins are fully integrated into CUKMODY. The coupling of MD to quantum mechanics for his studies on proton transfer/transport is also integrated to CUKMODY.

Biological function relies on the chemical reactions of electron and proton transfer that take place in enzyme active centers and are strongly influenced by the geometry and energetics of the surrounding medium. The small masses of electrons and protons necessitates a quantum mechanical treatment and, as electrons and protons are charged species, they strongly interact with their surroundings, the many strongly polar amino acids in the surrounding proteins. It is now appreciated that a coupling of proton motion to electron transfer is basic to mechanisms of biological energy conversion. Chains of hydrogen bonded water and/or amino acid residues are essential to proton translocation, the movement of protons across membranes to create electrochemical gradients for energy transduction. He has developed methods to treat the protons involved in translocation quantum mechanically and have coupled them to the surrounding protein in order to simulate rates of this process. The methods have been applied to cytochrome c oxidase, an enzyme responsible for converting the energy from redox and oxygen chemistry to support proton translocation and create a transmembrane pH gradient used for ATP production.

Many proteins undergo large-scale conformational changes when they bind ligands responsible for their catalytic activity. For example, Adenylate Kinase (AK) will bind ATP and AMP and, as it does so, the protein closes over these substrates to form a catalytically active complex, with water excluded. Exploring the closing motion of proteins with conventional Molecular Dynamics is unrealistic on normal simulation time scales. Cukier develops methods that can accelerate the exploration of configuration space and permit sampling of the motion along the path from open to closed-like protein configurations. Applied to Adenylate Kinase, the potential of mean force along a closing reaction coordinate (RC) has a distinct minimum at an RC value that is close to the ligand bound geometry. Therefore, apo AK can readily fluctuate to conformations favorable for ligand binding.

Protein stability is based on a delicate balance between energetic and entropic factors. While folded proteins are essential for catalytic activity, the role of disordered states in proteins has become manifest more recently. Intrinsically disordered proteins (IDPs) that may sample a large conformational space by rapid interconversions among a large number of states are known to be essential for cell function. IDPs interacting with a folded partner protein in the act of binding can order the IDP to form the correct functional interface by decrease in overall free energy. Thus it is important to evaluate the entropic cost of ordering an IDP. Cukier has developed computational methods to evaluate this cost and find that it is a significant contributor to the overall free energy of folding a protein. The MD data required to sufficiently sample, to a statistically significant level, the many conformations that and IDP can take on is very challenging. It requires both enhanced sampling methods and very long trajectories.

The virtue of MD is that it is fully atomistic. However, because atom interactions take place on a femtosecond time scale while functional motion may take place on nanosecond to second time scales, the number of time steps that must be simulated is enormous. Cukier shares, "Without resources such as the MSU HPCC, this kind of research would not be feasible."