Back to Table of Contents

42390-GB6
Simulating Linear and Nonlinear Optical Spectra in Condensed Phase Systems with a Mixed Molecular Dynamics and Quantum Mechanical Method

Brent P. Krueger, Hope College

Introduction

Fluorescence-detected resonance energy transfer (FRET) experiments are widely utilized to examine the structure and structural dynamics of polymer and biopolymer systems.  With a variety of modern fluorescence spectroscopies that can be coupled to microscopies, as well as the relative ease with which (two) fluorescence probe molecules can be incorporated into polymers, FRET has become relatively easy to employ in a qualitative fashion and has experienced a rebirth in the past decade or so.  However, quantitative analysis is hampered by the uncertainty surrounding a number of approximations that must be made.  Most famous of these is the  approximation, which requires that both probes isotropically sample all possible orientations.  Several others also lead to difficulty, including the assumption of no correlation between the probe-probe orientation factor () and the probe-probe distance, which we recently demonstrated (VanBeek et al., 2007, Biophys. J., 92, 4168–4178). 

The two approximations mentioned above are associated with the relative motions of the probe molecules.  Thus, if one could estimate the structural dynamics of the two fluorescent probes these difficult approximations would not be needed and the quantitative ability of FRET experiments would be improved.  We have spent much of the past year beginning molecular dynamics (MD) calculations that will be used to simulate the structural dynamics.  In addition, we have conducted quantum mechanical (QM) calculations that will explore the breakdown of another key FRET approximation—the ideal dipole approximation.  Finally, we also began experimental work, learning sample preparation and technical details and collecting our own data using single-molecule fluorescence techniques.  The computer simulations will mesh with fluorescence data to provide an estimate of the fluorescent probe motion, obviating the need for some approximations, and therefore improving the quantitative capability of FRET experiments.

Progress

The short-term goal for the past year (year 3 of the grant) was to use calculations to help identify what portion of the experimentally observed structural dynamics are due to the biopolymer of interest, due to the probe molecules, and due to the linkers that connect the probes to the biopolymer.  We are focusing on several small RNA constructs that are all derived from the hairpin ribozyme (HR)—a small catalytic RNA molecule that exhibits a large structural change during activation.  The HR has been widely studied using ensemble and single-molecule FRET experiments.  However, it is not known how much the probes themselves (and the linking moieties) contribute to the total system dynamics observed experimentally.

To examine the structural dynamics of the HR constructs we are using molecular dynamics (MD) simulations.  The force fields used in biomolecular MD simulation packages do not have parameters for the fluorescent dyes of interest.  So, this research began by developing MD parameters for the fluorescent molecules and the linkages that connect them to the nucleic acid constructs.  This research was performed by David Paul, the SUMR student supported by this grant.  David completed parameterization of several probe molecules as well as generic linkage moieties for connecting any of the probes to DNA or RNA.  These results allowed David to construct simulation systems matching those that we are using in the single-molecule experiments. 

Though MD simulations are not yet complete, David's experience in performing parameterization has been published as an “advanced tutorial” on the AMBER website.  AMBER is a MD package that is widely used by academics and industry for biomolecular simulations, receiving 90-100 citations per year.  Our tutorial is part of a suite of tutorials that assist both novice and experienced AMBER users and receive an average of over 50,000 hits per day.

      In addition to the MD work, we have also used QM calculations to examine the breakdown of the ideal dipole approximation (IDA).  Conducted in collaboration with Dr. Benedetta Mennucci's group at the University of Pisa, these calculations have involved undergraduate Lydia Hartsell.  Dr. Mennucci's group has developed a QM description of the interaction between molecules that drives the energy transfer necessary for FRET.  This development, based on my previous work with Dr. Graham Fleming and Dr. Greg Scholes, has allowed us to determine an accurate value for the interaction at any distance, and therefore to quantify breakdown of the IDA when the two fluorescent probes are close to each other.  A manuscript describing the result is currently preparation and will likely be submitted in October.

Summary

Over the three years of this award my group has completed investigation of a hybrid MD-QM method for examining solvation dynamics, has quantified the breakdown of the IDA, has completed parameterization of several fluorescent probe molecules, has begun MD simulation of RNA constructs for FRET experiments, and has begun single-molecule FRET experiments of these same systems.  This work has resulted in two publications, with one more near submission, and an AMBER tutorial.

Back to top