University of Central Florida Undergraduate Research Journal - Computational Analysis of Broad Complex Zinc-Finger Transcription Factors
US tab

Computational Analysis of Broad Complex
Zinc-Finger Transcription Factors

By: Barbara Mascareno-Shaw
Mentor: Dr. Thomas Selby

Methodology

"The polypeptide backbone of each protein was fixed, allowing only the side chains to adjust to the environment.  This interaction was rationalized by the fact that proteins with a high degree of homology typically have a similar overall fold, but side chain interactions can vary.  "

The von Kalm laboratory at the Department of Biology at the University of Central Florida provided 4 amino acid sequences from fruit flies to initiate the computational studies of the zinc-finger DNA proteins:  Z1, Z2, Z3, and Z4.  The sequences were aligned against the NCBI database (National Center for Biotechnology Information) supported by National Institute of Health (NIH).  A search for homologous proteins within the protein structure database (PDB) showed that two crystal structures of zinc-finger DNA proteins were available for modeling studies (Berman, Westbrook, Feng, Gilliland, Bhat, Weissing, Shindyalov, & Bourne, 2000)  These crystal structures have the PDB codes 2DRP and 1A1H.  The first crystal structure is known as Tramtrack Protein with DNA from the fruit fly and corresponds to the 2DRP PDB code (Fairall et al., 1993).  The second crystal structure is called Zif268 Variant Finger-DNA Complex from Mus musculus (Elrod-Erickson et al., 1998) and corresponds to the 1A1H PDB code.  The next step was to use the experimentally determined crystal structures, 2DRP and 1A1H, as templates to model the 3D structures of the Z1-Z4 sequences.  The modeling was completed using the Swiss PDB-Viewer v3.7 software (Guex and Peitsch, 1997).  Once these models were obtained, the energy minimization calculations between the DNA substrate and protein backbone were tabulated using the molecular mechanics (MM3) force field parameters within the BioMedCACHE software (Group, 2002).  The polypeptide backbone of each protein was fixed, allowing only the side chains to adjust to the environment.  This interaction was rationalized by the fact that proteins with a high degree of homology typically have a similar overall fold, but side chain interactions can vary.   The side chain interactions also allowed simplification of the energy minimization calculations.    Next, the modeled structures were visually compared against the templates using ViewerPro 4.0 (Molecular Simulations, 2000) to observe any differences in side chain locations.  All sequences were treated in the same manner.  Within the modeled structures, mutations were performed using the Swiss PDB-Viewer to assess the interaction energy differences between the mutant and wild type zinc-finger proteins.  Once again, the energy minimization calculations were tabulated using BioMedCACHE and the differences between energies were compared.  Once the mutant proteins were created, they were docked against the DNA substrate to evaluate any energy differences. 

The energy calculations are based primarily on equation 1, which shows the total energy of a molecule based on the summation of the stretching, bending, torsional, and non-bonded interactions.  Equations 2-5 below describe various components of the total energy;

The is based on Hooke’s law (equation 2), where is the parameter that shows how stiff a bond can be and ro is the equilibrium length (NIH, 2002).  Each of these parameters are established for specific atom types; for example, C-C, C=O, or N-H bonds.  The is represented in equation 3, where is the parameter that predicts the angle of stiffness and qo is the equilibrium angle (NIH, 2002).  The larger the value of k for either the or , the greater the energy required to dissociate the molecule.  The is shown in equation 4, where A represents the amplitude of the wave, t is the rotational angle axis, f is the shift of the wave, and n determines the period of the wave (NIH, 2002).  It is used primarily as a correction value to bring the ET to a minimum optimum value.  The Enon-bonded is the sum of all the interactions between atoms and is represented in equation 5.  The first term is the van der Waals interactions, where A represents the proximity between atoms and B represents the maximum distance between atoms.  The second term is the electrostatic term associated with coulombic forces.  The energies are calculated using force fields, which is the mathematical interpretation of the data for the many types of atoms (NIH, 2002).  The calculation requires that the atom type be specified (Leach, 1996), which allows the force field parameters to distinguish atoms with respect to hybridization and geometry.  For instance, the sp3 hybridized carbon, which has a tetrahedral geometry, can be differentiated from the sp hybridized carbon, which has a linear geometry (Leach, 1996). 

The usefulness of these energy calculations is found in comparing the difference between energies of the macromolecules to understand the interaction energy (IE) for a specific complex as shown in equation 6

In this equation ECOMPLEX represents the total energy (found through equation 1) for the DNA-zinc-finger complex.  EPROTEIN  represents the total energy of the zinc-finger protein alone and EDNA represents the energy of the DNA alone.  By subtracting the sum of the two energies from the total energy of the complex, the relative energy of interaction can be determined.  The term “relative” is used to describe the energy components in this case, due to fact that solvation effects, and other types of entropy factors, are not used explicitly in the calculations.  When comparing different proteins, such as template vs. homology model or homology model vs. mutant, the interaction difference is described as ΔIE.

Results >>