The VWG hypothesis: Predicting chromatin higher order structure from the DNA sequence
Abstract
DNA has evolved in living organisms to function as chromatin, and the basic unit of chromatin is the nucleosome. It is known that the way nucleosomes are arrayed on the DNA molecule influences chromatin structure and function, and that nuclear factors, histone modifications and the replication machinery can determine where nucleosomes form. What is debated is whether there is any information in the DNA sequence that influences nucleosome array formation. This study addresses this question in three steps. First, experiments with a 4.5 kb fragment of the chicken ovalbumin gene established that the period-10 non-T, A/T, G (p-10 VWG) motif influences where nucleosomes form in chromatin reconstituted in vitro, tending to avoid p-10 VWG-poor regions. The nucleosome periodicity of this fragment was also predicted from the DNA sequence by Fourier analysis of the density of this motif. Small sequence alterations on the reconstituted chromatin of the gene confirmed that the observed periodicities were solely the result of the p-10 VWG signal. Second, our ability to predict nucleosome array periodicities solely from the DNA sequence was tested in mouse liver chromatin. Long-range array periodicities were predicted from the annotated sequence of the mouse genome. Probing the corresponding regions validated ten consecutive predictions (P = 10-7). Repeats as short as ∼20 bp less than the bulk repeat for mouse liver were found; the first time that such low values have been observed. Third, a plausible correlation was found between regions (∼100 kb range) of predicted ordered nucleosome arrays and regional recombination coldspots in the human genome (P = .01). Also, similar-size regions (∼300 kb range) of predicted ordered nucleosome arrays were found to contain significantly higher gene densities than in genome regions lacking these signals (P < .01). This offers a functional explanation for why this motif has been reported to be overly abundant in the human genome. This is the first time that evidence is provided that the DNA sequence is implicated in containing information that may influence chromatin higher order structures. It is reasonable to expect that distinct chromatin structures play a role in gene regulation and, therefore, can also be a cause of miss-regulation and disease.
Degree
Ph.D.
Advisors
Stein, Purdue University.
Subject Area
Molecular biology|Genetics
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.