Structural Biology
Click on picture to see picture legend.
NMR Studies at LLNL
Structural Determination
Protein Structure Prediction Center
CASP: Critical Assessment for Protein Structure Prediction
Computational Biochemistry
Computational modeling, x-ray crystallography, and nuclear magnetic
resonance spectroscopy are being applied to analyze a variety of molecules,
including chemical mutagens that bind to--and damage--DNA, and proteins that
repair DNA damage, inactivate the genome, and replicate the DNA molecule.
Studies of small chemical mutagens such as PhIP, a heterocyclic amine produced
during the cooking of foods, employ computational and physical methods to
predict the 3D structure of the molecule and its highly reactive metabolites,
to identify the structure of the complex it forms with DNA, and to determine
how these and related molecules damage DNA and cause cancer. Various forms of
spectroscopy, computer modeling, and x-ray diffraction are being used to
identify the structures adopted by the two sperm nuclear proteins protamine 1
and 2 when they bind to DNA, to determine how they inactivate the entire genome
of the sperm cell, and to understand how defects in the process cause male
infertility and early fetal deaths. Computational methods (homologous modeling)
are being used to predict the structure of unusually thermostable enzymes, such
as the Stoffel fragment of Thermus aquaticus DNA polymersase. Through a
collaboration with Gladstone Institute investigators, we are determining
the crystal structures of proteins and protein variants important to
cardiovascular and Alzheimer's disease.
In the future, these techniques will be extended to the analysis of a series of
proteins that repair damaged DNA. This will allow us to begin investigating how
repair-related proteins interact with each other and function at the molecular
level. Once the DNA repair-related genes XPA, XRCC1 and ERCC2 have been
produced in sufficient quantity by over-expression in vitro, their protein
products will be crystallized and their structures determined by x-ray
diffraction. Computational methods will be employed in parallel to predict the
structures of the repair proteins, as well as other disease-related molecules.
As the genome efforts move into their sequencing phase, we will also begin to
develop computer codes that will allow investigators to search raw DNA
sequence, pick out exons using structural features characteristic of proteins,
and search for and identify families of proteins with common structural domains
or functions. This will be accomplished by incorporating protein structural
features and threading algorithms used for protein structure prediction into
codes that analyze all possible reading frames of the sequence.
We now have a 600 MHz NMR
instrument for studies on small molecule-DNA interactions. Our
emphasis will be on structure-function analysis of DNA repair proteins
or peptide fragments, interactions between peptides and DNA or small
molecules and DNA, and protein structure prediction. This project
couples the traditional disciplines of physical sciences and biology
and is a natural for Livermore given the multiple disciplines that
exist here including the Human Genome Center.
BBRP home page || Back
to Research In Progress
Web page maintained by BBRP Webmaster (BBRPWebmaster@humpty.llnl.gov).
UCRL-MI-117984
LLNL disclaimers
