Design of Specific Metal-Binding Sites

In the series of articles, an exploitation of the DFT/B3LYP calculations for the selection of highly specific combinations of simple functional groups representing amino acid (AA) side chains for each of the six divalent transition metal (TM) ions (Co²⁺, Ni²⁺, Cu²⁺, Zn²⁺, Cd²⁺, Hg²⁺) has been described. Finally, it resulted in the design of short peptide sequences (10-30 amino acids) with a potential of highly specific binding of TM ions. Several of these have been tested experimentally and specific binding has been observed (ITC and mass-spec experiments). Attempts based on the automatic generation of all possible combinations of AA side chains and implementation of QM/MM methods into the scoring function is planned to further enhance the observed selectivity in metal binding.

Reaction Mechanisms of Metalloproteins

Currently, there are four ongoing projects in the theoretical chemistry group aiming towards a better understanding of the details of reaction mechanisms of various classes of metalloenzymes at the atomic or electronic level. Multicopper oxidases are enzymes that couple four one electron oxidations of a reducing substrate with one four-electron reduction of dioxygen to two water molecules. The latter reaction takes place in a trinuclear copper cluster, an unprecedented bonding arrangement depicted on the illustrative figure. Therein, the putative reaction mechanism, proposed recently as a major result of an extensive QM/MM study, with part of it then being confirmed experimentally can be seen as well. Having settled the structural arrangement of the observed intermediates, their spectroscopic parameters have been calculated, which can be directly compared to experimental data. In the future work, we plan to study the protonation and oxidation states of partially re-oxidized enzyme and the barrier of O₂ cleavage. All of this can provide us with an unambiguous picture of the reaction mechanism pertinent to this class of enzymes. Glutamate carboxypeptidase II is an enzyme, which plays an important role in the regulation of the levels of one of the neurotransmitters, N-Ac-Asp-Glu (NAAG), in synapses. It is therefore a desirable pharmaceutical target and the regulation of its activity would likely help in postponing the onset of various neurodegenerative processes, such as Alzheimer’s, Parkinson’s, or Huntington diseases.

A recently published crystal structure of the enzyme has demonstrated the existence of a dinuclear zinc site and has prompted us to study the binding mode of the substrate (NAAG), the structures of various intermediates that occur along the reaction pathway (which is a hydrolytic cleavage of the peptide bond). In close collaboration with the experimental group of Jan Konvalinka, the calculations have already provided invaluable information complementing the experiments, such as mutation studies carried out in the collaborator’s lab. Superoxide dismutases (SOD) also represent a highly interesting class of redox active enzymes that catalyze the disproportionation of poisonous superoxide radical into a hydrogen peroxide and dioxygen, thus preventing an oxidative damage of cells. Following our early in vacuo and QM/MM/X-ray (so-called quantum refinement) studies, attempts are being made to model the complete reaction cycle of Mn-SOD (and Fe-SOD) by QM/MM methods, including the calculations of redox potentials that may explain the specificity of the enzyme (e.g., why Mn-SOD does not work with an iron cation, and vice versa). Desaturases are enzymes catalyzing the desaturation of double bonds in longer fatty acids. In this way, organisms can regulate the permeability and viscosity of membranes, which is an important factor in self-protective processes. Recently available crystal structure prompted several spectroscopic studies, carried out in the lab of Prof. Edward Solomon, Stanford University. Despite the enormous experimental efforts, the reaction mechanism is still unknown and calculations can provide many missing pieces of information. The computational efforts in the applicant’s lab are proposed to be carried out in this direction, in close collaboration with the above laboratory.

Method Development

Method development will be directed into the further improvement of QM/MM coupling schemes, including the promising method of Ryde and coworkers - quantum refinement - which locally improves the experimentally determined crystal structures, while not deteriorating the overall R-factors. In the future, the full set of methods taking us all the way from the raw crystallographic data towards the accurate values of experimentally observable quantities can be accomplished. This also includes (a) QM(+D)/MM coupling scheme (i.e., inclusion of the empirical dispersion parameters into the QM part, which is usually represented by density functional theory), (b) the implementation of the second derivatives into Ulf’s Ryde QM/MM code ComQum.