Brent A Gregersen - Research Interests

Brent A Gregersen
Research Associate
Department of Chemistry
Tel. (612) 625-6317
Fax. (612) 626-7541
greg0148@tc.umn.edu

Department of Chemistry
P176 and P180 Kolthoff Hall
University of Minnesota
225 Pleasant Street SE
Minneapolis, MN 55455-0431
USA

Research Interests

An overall goal of this research group is to gain a better understanding of phosphate hydrolysis - an important reaction motif in biological chemistry through the use of various computational methodologies. Of particular interest is the study of the molecular mechanisms whereby RNA can catalyze fairly complicated reactions. One such catalytic RNA molecule of particular interest being the Hammerhead ribozyme (a self-cleaving RNA enzyme found in the tobacco ringspot virus). The analysis of ``thio effects'' has been used to experimentally to probe the mechanistic details of RNA catalysis. Unfortunately, the mechanistic interpretation of the results is at times ambiguous. Theoretical studies can provide valuable aid in the mechanistic interpretation of the observed thio effects.
It would be preferable to employ a high-level ab initio quantum model in quantum mechanical/molecular mechanical studies of RNA catalysis; however, reliable simulations of ribozyme-catalyzed reactions require consideration of a large quantum region and extensive configurational sampling that often necessitates the use of more efficient (e.g., semiempirical) quantum models.

Ab initio Database

Our group has invested a significant amount of time collecting data to add to our Quantum Chemical database for RNA catalysis (QCRNA). The database was designed using Perl and MySQL, and currently contains over 1500 optimized structures representing minima, transition states and in some instances, non-stationary points in-between. These structures form the basis for over 160 "reactions" related to RNA/DNA and phosphate hydrolysis. These structures and reations are used to parameterize our next generation semi-empirical models.

Semi-empirical method validation/development

To date, the MNDO/d semiempirical Hamiltonian has performed reasonably well in reproducing experimentally observed thio effects in the di-anionic transphosphorylation reaction mechanism. This corresponds experimentally to a reaction under basic conditions (completely deprotonated). In the case of neutral or acidic reactions in solution, or ribozyme-catalyzed reactions at optimal pH, the MNDO/d Hamiltonian cannot be reliably applied since there are known deficiencies of this model for proton transfer and hydrogen bonding.
The development of a new semiempirical Hamiltonian for investigation of the mono-anionic or neutral reaction mechanisms must be undertaken to overcome this problem. A theoretical framework has already been developed in the York group, but it has yet to be fully parameterized against high level density-functional calculations. This parameterization effort will be on-going, and the most promising parameter sets must be validated with QM/MM simulations. The ab initio calculations stored within our QCRNA database serve as a useful dataset in the parameterization of these new and/or semi-empirical models.

Smooth COSMO implementation into MD and semi-empirical codes

The smooth conductor-like screening model (developed in York, J. Phys. Chem. A, 103, 11060-11079(1999)) has many advantages over the original COSMO formulation. The introduction of a Gaussian basis for representation of the surface charge density and inclusion of a 'switching' region greatly improves the usefulness of the model when performing geometry optimizations or any other calculation where a potential with continuous first derivatives with respect to nuclear coordinates is required. We have implemented the smooth COSMO method into local versions of CHARMM and MNDO97.

Efficient modeling of the macromolecular environment in QM/MM simulations

The variational electrostatic projection method (VEP) is based off of the same basic principals of electrostatics as the conductor-like screening model. The VEP method introduces a smooth Gaussian basis on a discretized surface (constructed using Lebedev angular quadrature) for expansion of the multipolar charge distribution of a solvated macromolecule that surrounds the active/stochastic region of interest. A variational mapping procedure is then used to obtain the electrostatic interaction between particle distributions on opposite sides of the surface. Significant improvement over the basic VEP method (2 orders of magnitude average improvement in the force errors) are realized by using the variational electrostatic projection method in conjunction with a reverse variational mapping procedure (VEP-RVM).

Background

I received my Bachelor of Science degree in Biochemistry in 2000 from the Chemistry Department at the University of California, San Diego. While at UCSD, I worked for Dr. McCammon and Dr. Carlson using ab initio calculations to examine a deprotonation step in D-Ala:D-Ala ligase of the ddlb gene of E. coli (DdlB).

Publications

	"High-order discretization schemes for biochemical applications of boundary element solvation and variational electrostatic projection methods", Brent A. Gregersen and Darrin M. York, J. Chem. Phys., in press.
	"A smooth solvation method for d-orbital semiempirical calculations of biological reactions I: derivation, implementation and testing", Jana Khandogin, Brent A. Gregersen, Walter Thiel and Darrin M. York, J. Phys. Chem. B, in press.
	"A smooth solvation method for d-orbital semiempirical calculations of biological reactions II: application to phosphoryl transfer reactions.", Brent A. Gregersen, Jana Khandogin, Walter Thiel and Darrin M. York, J. Phys. Chem. B, in press.
	"Variational electrostatic projection (VEP) methods for efficient modeling of the macromolecular electrostatic and solvation environment in activated dynamics simulations", Brent A. Gregersen and Darrin M. York, J. Phys. Chem. B, 109, 536-556 (2005). [Abstract][PDF]
	"Hybrid QM/MM study of thio effects in transphosphorylation reactions: the role of solvation", Brent A. Gregersen, Xabier Lopez and Darrin M. York, J. Am. Chem. Soc., 126, 7504-7513 (2004). [Abstract][PDF]
	"Pseudorotation of natural and chemically modified biological phosphoranes: implications for RNA catalysis", Carlos Silva López, Olalla Nieto Faza, Brent A. Gregersen, Xabier Lopez, Angel R. de Lera, and Darrin M. York, Chem. Phys. Chem., (Communication) 5, 1045-1049 (2004). [Abstract][PDF]
	"Hybrid QM/MM study of thio effects in transphosphorylation reactions", Brent A. Gregersen, Xabier Lopez and Darrin M. York, J. Am. Chem. Soc. (Communication) 125, 7178-7179 (2003). [Abstract][PDF]

CV
My current CV