Biomedical Engineering Reference
In-Depth Information
Combination of ChIP and microarray or sequence methods, ChIP-ChIP (Buck
and Lieb, 2004), ChIP-PET (paired end ditags) (Loh et al., 2006), and ChIP-Seq
(short tag based sequencing) (Barski et al., 2007; Johnson et al., 2007; Mikkel-
sen et al., 2007; Robertson et al., 2007), are powerful high-throughput methods
for identifying gene regions bound by specific proteins. DNA regions bound by
Oct4, Nanog, and Sox2 were first identified in both hESCs and mESCs (Boyer
et al., 2005; Loh et al., 2006). The principal finding in the initial studies was the
remarkable extent of factor co-occupancy at promoter or other gene regions.
An extraordinary database of ChIP binding in hESCs and mESCs has been
generated recently. Core factors, Oct4, Sox2, and Nanog are also bound to their
own promoters. These studies indicate the existence of combinatorial occu-
pancy of target gene promoters by ''core'' factors within autoregulatory and
feed-forward transcriptional circuits. Kim et al. extended these findings further
by a modification of the ChIP-ChIP approach, termed bioChIP-ChIP (Kim
et al., 2008), that takes advantage of affinity capture of biotinylated proteins
bound to chromatin and subsequent hybridization to promoter arrays. By this
strategy, factors can be analyzed without the requirement for protein-specific
antibodies. Often, protein-specific antibodies fail to perform well in conven-
tional ChIP assays, sometimes for entirely obscure reasons. Kim et al. used the
bioChIP-ChIPmethod to identify the putative target promoters of ''core'' factors,
as well as the somatic cell reprogramming factors (Oct4, Sox2, Klf4, c-myc) and
others within the pluripotency network (Nanog, Dax1, Rex1, Zpf281, and
Nac1) (Wang et al., 2006). Remarkably, data showed that
800 gene promoters
are bound by four or more transcription factors (Fig. 2a). This finding high-
lighted greater combinatorial factor binding than previously appreciated. Of
the nine factors tested, seven factors (all except c-myc and Rex1) lie within a
subgroup sharing many targets and appear to be involved in both activation
and repression. Further analysis of the data revealed a striking correlation. The
promoters, which are active in ESCs, and then repressed upon differentiation,
tend to be occupied by
>
4 factors including Nanog, Sox2, Dax1, Nac1, Oct4,
and Klf4. On the other hand, promoters that are expressed upon differentiation
but silent in ECSx are generally occupied by few (
<
4) or a single factor. These
correlations are illustrated by the gene-set enrichment analyses shown in
Fig. 2b. Although the mechanisms accounting for these context-dependent
differences remain to be elucidated, these data demonstrate that the same
factors function both positively and negatively in transcription.
In addition to transcription factors, histone modifications, specifically
H3K4me3 (histone 3 lysine 4 trimethylation) and H3K27me3 (histone 3 lysine
27 trimethylation), signify important aspects of gene regulation in ESCs. c-myc
is almost exclusively bound to the promoters with the H3K4me3 signature
(Fig. 2c). This finding is consistent with the presence of the H3K4me3 signature
at active genes and c-myc functioning as a positive regulator. When other
factors (such as Nanog, Oct4 in Fig. 2d) are bound to promoters alone, they
do so in association with the H3K27me3 mark, which is correlated with gene
silencing. These observations are consistent with a model in which c-myc
