Finally, we are interested in finding other microarray experiments that
may be relevant to gene regulation by Ume6p. Therefore, we click on the
"Sort Experiments"
tab and are presented with the list of all PSAMs
similar to the display from the "Browse PSAMs" tab. We scroll down to
UME6_H2O2Hi and click on the affinity logo. We are then given a list
of all expression samples sorted by the absolute t-value of the
correlation between their expression values and the predicted promoter occupancies for
UME6_H2O2Hi. Of course, we see a number of the experiments
from the dataset that we just explored. We can investigate the
experimental design of any of the samples with which we are unfamiliar
by clicking on the "GEO" link next to the sample title, which takes
us to the NCBI GEO website [1]. After following several
of these links we find that many of these other experiments are
independent studies on sporulation, as could be expected. However,
there are other experimental designs among the top scoring experiment
samples. Of special interest is a segregant from a cross between a lab
strain and wild strain of S. cerevisiae, suggesting that there is
polymorphism that affects Ume6p regulation [15]. Also, of
great interest are a series of experiments that examined expression in
a variety of strains that had mutations eliminating one or more of four
acetylation sites on the N-terminal tail of histone H4
[16]. Ume6p is known to regulate expression of its target
genes by recruiting Sin3p and Rpd3p to hypoacetylate histones H3 and H4
[17]. If the lysines that get acetylated are absent from a
histone, it may affect the ability of Ume6p to regulate gene
expression. By examining the sorted list of experiments, we see that
the expression of Ume6p targets are significantly altered (E-value <= 10^-3) if at least three of the lysines are mutated. Dion et al. [16] report that the mutations of the lysines of histone
H4 has a largely non-specific effect on gene expression with over a
thousand genes affected. However, due to the strong correlation between
expression under histone H4 lysine mutations and predicted Ume6p
binding, being regulated by Ume6p must make genes particularly
sensitive to alterations in histone acetylation.
Barrett C. Foat
2007-09-17