Biology Reference
In-Depth Information
a lineage-specific way after gastrulation and during late fetal and postnatal
development. For example, a comprehensive methylome mapping effort in
human tissues estimated that 18% of the covered regions show tissue-specific
methylation among 16 analyzed tissues (
Rakyan et al., 2008
). Mapping exper-
iments have also been performed in hematopoietic cells, which offer a con-
venient experimental system to study cellular differentiation
in vivo
,and
revealed widespread changes in DNA methylation associated with hemato-
poietic cellular commitment (
Calvanese et al., 2012; Hodges et al., 2011;
Ji et al., 2010
). Strikingly, DNA methylation profiles can even be used to seg-
regate hematopoietic cell lineages with a good accuracy (
Bock et al., 2012
).
Some of the tissue-specific DNA methylation occurs at CpG-rich promoters
of important developmental genes such as
Hox
or
Pax
genes (
Illingworth et al.,
2008
). Another interesting example is methylation of the
Elf5
gene, which
occurs earlier in development.
Elf5
encodes an Ets transcription factor crucial
for the development of the trophoblast lineage and its promoter is methylated
in embryonic cells but not the trophoblast lineage, which constitutes an epi-
genetic barrier between these earliest two cell lineages (
Ng et al., 2008
). These
observations suggest that promoter methylation could serve to stabilize lineage
decision by inducing stable repression of key developmental genes
(
Hemberger, Dean, & Reik, 2009
). Besides promoters, CpG islands residing
outside annotated promoters are more prone to methylation in tissues and
could provide additional positive or negative regulation of tissue-specific gene
expression in
cis
by regulating insulators or noncoding RNAs (
Deaton et al.,
2011; Illingworth et al., 2010; Maunakea et al., 2010
). Yet the majority of
tissue-specific DNA methylation occurs in CpG-poor sequences (
Nagae
et al., 2011; Rakyan et al., 2008
). In this case, because CpG-poor sequences
can be demethylated downstream of transcription factor binding (
Stadler et al.,
2011
;see
Section 3.1
), it cannot be considered a primary cause of lineage-
specific gene expression.
4.3. Does DNA methylation influence RNA processing?
One exciting recent discovery is that apart from regulating transcription,
chromatin epigenetic marks may also influence cotranscriptional RNA
processing events (
Luco et al., 2010
). In line with this idea, sharp transitions
inDNAmethylation levels were observed at exon-intron boundaries, as well
as correlations between exonmethylation and alternative splicing, suggesting
a possible role of DNA methylation in exon choice (
Anastasiadou, Malousi,
Maglaveras, & Kouidou, 2011; Chen, Feng, Joo, Jacobsen, & Pellegrini,
