Molecular Dynamics Studies of the Structure of Collagen
Collagen degradation plays an important role in a number of important diseases such as arthritis, tumor metastasis and atherosclerosis. In order to decipher the structural determinants underlying successful collagenolysis, we have conducted computational analyses of the structure of collagen using molecular dynamics. These simulations predicted that segments of collagen exist in two conformation states, normal and vulnerable. The vulnerable state represents a partially unfolded state that allows cleavage of the molecule by collagenases. Using biophysical techniques, we are studying the nature of unfolded state and are creating mutants to understand the conformational constraints that regulate the structural state of collagen, native or vulnerable.

Immunomodulation by Collagen-like Peptides
Inflammation plays a crucial role in many connective tissue disorders.  While it is known that inflammation can promote collagen degradation, the effect that collagen-degradation products have on inflammation has not been thoroughly studied.  We studied the effect of specific amino acid sequences, which model different types of collagen-degradation products, on human peripheral blood monocytes (HPBMs).   We have discovered a number of collagen-like peptides that can modulate innate immunity.  We are in the process of elucidating the molecular mechanism underlying this interesting finding.

Modeling the Unfolded State of Tau Protein
Alzheimer’s disease (AD) is the most common form of dementia among older populations. The disease exhibits distinctive pathological hallmarks – extracellular aggregates of amyloid β peptide, known as amyloid plaques, and intracellular aggregates of tau protein, known as neurofibrillary tangles (NFT). The proteins found in these aggregates are not only disease markers, but are suspected to play a role in the disease process. The focus of our studies is tau protein, which is a natively unfolded, microtubule-associated protein. Specifically, our goal is to elucidate the molecular basis of tau dysfunction in Alzheimer’s disease and related tauopathies with the aid of computational models of tau in normal and disease states. Furthermore, these models are used to design and optimize peptide inhibitors to tau-induced neurotoxicity.

Modeling Physiologic Unfolding of Fibronectin with Steered Molecular Dynamics Simulations 
The extracellular matrix (ECM) facilitates cell adhesion. Integrins (proteins found on cell surfaces) bind to fibronectin (FN), an ECM protein, at specific sites containing the tripeptide Arginine-Glycine-Aspartate.  Fibrillogenesis requires mechanical tension, exerted by the cell, which likely leads to some degree of FN unfolding.  As the the precise mechanism of fibrillogenesis is unknown, we are exploring the mechanism of FN unfolding when bound to an integrin using detailed molecular simulations. 

Symbolic Analyses of Cardiovascular Signals

In collaboration with the data-driven medicine group, we are developing automatic techniques for analyzing large amounts of cardiovascular data.  In contrast to traditional medical expert systems, this technique incorporates no a priori knowledge about disease states and therefore facilitates the discovery of unexpected events that are difficult to predict.  The ultimate goal is to use these methods to uncover novel patterns with prognostic significance from databases containing large amounts of clinical information.