Construction of a comprehensive genome map of Z-DNA segments as real-time indicators for genetic activities in human embryonic stem cells

Abstract

Biological information is not exclusively encoded in the nucleotide sequence of DNA. The molecular of double-stranded DNA is dynamic and functions also as a carrier of biological information. We are just now beginning to realize how DNA structural transitions as real-time indicators for genetic activities inside living cells might be involved in regulatory circuits and cellular memory. Experimental evidence indicates left-handed Z-DNA plays roles in DNA transactions such as transcription, chromatin remodeling, and recombination. During the three years funding period of this project, we developed a computational approach which revealed that sequences with high Z-DNA forming potential at moderate levels of DNA supercoiling are enriched in human promoter regions (Part A and B). However, the actual distribution of Z-DNA segments in genomes of mammalian cells has been elusive due to the unstable nature of Z-DNA and lack of specific probes. At the end of this funding period, we were able to present a first human genome map of most stable Z-DNA segments obtained with A549 tumor cells (Part A and B). We used the Z-DNA binding domain, Zα, of the RNA editing enzyme ADAR1 as probe in conjunction with a novel chromatin affinity precipitation strategy. By applying stringent selection criteria, we identified 186 genomic Z-DNA hotspots. Interestingly, 46 hotspots were located in centromeres of 13 human chromosomes. There was a very strong correlation between these hotspots and high densities of single nucleotide polymorphism. We predicted that most hotspots undergo structural transitions at high DNA superhelical densities. Our study indicates that the genetic instability and rapid evolution of human centromeres might, at least in part, be driven by Z-DNA segments. Contrary to in silico predictions, however, we did not find hotspots in promoter regions.