Protein-DNA Structure, Function, Design

The Kahn laboratory is interested in the analysis, structure, function, and design of protein-DNA complexes, focusing on the 50-1000 bp length scale. This is the biologically-relevant domain of multi-protein DNA complexes, DNA looping, chromatin, and DNA topology, and a quantitative understanding of these effects will support accurate systems biology predictions. We study the shapes (bending, twisting, topology) and flexibility of protein-DNA complexes and DNA loops, the functional consequences of changes in shape, and the design and control of DNA and protein-DNA shape. We use molecular biology techniques like DNA ring closure, electrophoretic mobility shift assays, and footprinting to guide hypotheses, and then move on to characterization with biophysical chemistry techniques such as fluorescence resonance energy transfer (FRET), single-molecule FRET, and atomic force microscopy (AFM). We apply analytical, Monte Carlo, and rod mechanics simulation methods to cyclization and topology results. The work has been funded by NIH and by NSF. We also have more applied interests in the use of nucleic acids for diagnostics and therapeutics. In collaboration with Celadon Laboratories, we have used absorbance melting curves to define the hybridization thermodynamics of modified oligonucleotides (Locked Nucleic Acid), with an eye to improving probe design. In collaboration with James Mixson at Univ. Maryland, Baltimore, we have applied biophysical techniques to the packaging of siRNA and of DNA plasmids for gene therapy by small histidine/lysine rich branched peptides. Work has been funded by NIH and by the Maryland Industrial Partnerships (MIPs). Representative projects are described below. A complete publication list and comprehensive PowerPoint presentations are available at http://www.biochem.umd.edu/biochem/kahn.

The Lac Repressor is the archetypal DNA looping protein. DNA constructs designed to stabilize different putative DNA loops provided hyperstable DNA loops anchored by the repressor, but the loops can exist in at least two conformations that are distinguishable using topology as well as bulk and single-molecule FRET. This work has led to fruitful collaborations with single-molecule biophysicists interested in the control of DNA looping (English, Meiners), mechanical engineers interested in DNA as an elastic rod (Goyal, Perkins), and molecular biologists interested in the action of DNA bending proteins in vivo (Maher). We have recently mapped the energy landscape for different loops over a sequence space of designed curved DNA, demonstrating that rod mechanics predictions for relative loop stabilities are remarkably accurate but also illuminating a requirement for protein flexibility (see Haeusler et al., Nucleic Acids Research 2012). Currently we are most excited about extending our understanding of looping to the engineering of DNA and protein partners that form designed DNA loops and self-assembled protein-DNA nanostructures.

Locked Nucleic Acid (LNA) is the most useful modified nucleic acid due to its enhanced thermal stability and resistance to nucleases. We performed the first thorough characterization of the sequence-dependent hybridization thermodynamics of internally LNA-modified DNA using UV absorbance melting curves, in collaboration with Celadon Labs; see McTigue et al. (2004), Biochemistry 43, 5388-5405. Along with subsequent work funded by SBIR grants to Celadon, this database enables improved design of LNA-modified primers, probes, and therapeutics that exhibit high specificity and stability even for very short molecules; we have demonstrated their utility for PCR and sequencing.