Proof of Principle for Shotgun DNA Mapping by Unzipping

Abstract

We are developing single-molecule methods for mapping protein-DNA interactions inside living cells by unzipping single chromatin fragments isolated from living cells. One avenue towards this capability involves unzipping random fragments that have been generated by site-specific restriction endonuclease digestion of whole genomic DNA or chromatin, a process we are calling shotgun DNA mapping or shotgun chromatin mapping. A key enabler of shotgun DNA mapping (SDM) will be the ability to assign the individual fragments to their specific sites in the genome, based on the sequence-dependent unzipping force of the underlying naked DNA sequence. We will present proof-of-principle results demonstrating the ability to match experimental data sets for pBR322 unzipping to the correct pBR322 sequence hidden in a library of approximately 3,000 yeast genome sequences arising from the known locations of XhoI recognition sites. We do so via an algorithm that scores the experimental data against simulated unzipping forces from a quasi-equilibrium model (Bockelmann, Essevaz-Roulet, & Heslot, 1997). Our next step is to perform SDM on yeast genomic DNA fragments produced by ligation of XhoI-digested DNA to unzipping constructs. Enhancements of the matching algorithm, data processing, and unzipping simulation will be discussed, along with studies of the robustness of the SDM method as a function of number of sites in genome and other parameters. In addition to the impact on our goal of single-molecule mapping of chromatin from living cells, SDM may have important applications in other areas of genomics, including high-throughput structural DNA mapping and genome-wide mapping of sequence-specific DNA binding proteins.