Summary
Luke Czapla was a graduate student at Rutgers University working in the laboratory of Professor Wilma K. Olson, currently at Uppsala University. Some of the information on this page is out-of-date. My research interests are primarily in the statistical mechanics of DNA and understanding the three-dimensional structure of the double helix and its conformational changes induced by thermal fluctuations, asymmetric charge distribution, and bound proteins, such as those involved in the packaging of bacterial chromosomes (nucleoid proteins such as HU) and stabilizing transcription repression or initiation loops (HU as well as eukaryotic HMG-like proteins). I am also interested in the interplay between folded chromatin and proteins which direct unfolding or bind the periphery of nucleosomal DNA.
DNA bending proteins and chromosome structure
Monte Carlo simulations have been our primary technique of modeling the effects of DNA-binding proteins and their role on the structure of small DNA chains. The ring-closure technique, extended to understanding circularization as well as protein anchoring (such as binding to the lac repressor, working with the laboratory of David Swigon), has been an important tool in understanding the looping free energy and how these proteins facilitate DNA looping. We have also been working with the laboratory of Jim Maher in understanding the cyclization of DNA in the presence of HMG-like proteins. This research has yielded an alternative technique to classical methods like the Shimada-Yamakawa theory of helical worm-like chains and the more recent Zhang-Crothers theory of sequence-dependent DNA looping, while being more general and accurate in estimating these free energies. Furthermore, the topology (writhe, twist) of these small circles modeled with proteins such as HU give insight into the larger-scale packaging of a bacterial chromosome.
Sequence-dependent DNA structure
Analysis of protein-bound DNA structures in the Protein Databank (hosted at Rutgers) reveals varying degrees of flexibility for different sequences. Work in our lab with fellow group members Dr. Andrew Colasanti (now graduated) and Guohui Zheng have shown that different dinucleotide steps (for example, AA/TT vs. AC/GT) have varying degrees of flexibility. Using our Monte Carlo technique, we have found reason why certain sequences might have much higher ring-closure propensities (measured as J factors) than other sequences, or values predicted by classical theory of an isotropic worm-like chain. We are able to account for anomously high experimental J factors (Jonathan Widom, 2004) for 94-115 bp minicircles without invoking concepts of sharp bending or single-stranded bubble formation. Moreover, features of DNA fluctuation in the real deformations found in these PDB structures, such as coupling between DNA bending and twisting, which are omitted in classical theories, explain how the phasing of helical repeat influences the J factor.