Biology Reference
In-Depth Information
and influences gene expression. Because a transient intermediate of demeth-
ylation would not be expected to be detected at high levels, this proposition
could explain the relatively high steady-state levels of 5hmC measured in
various cell types. One possible role is that 5hmC modulates the binding
of methyl-CpG binding proteins. Interestingly, several members of this fam-
ily (MBD1, MBD2, MBD4) do not recognize 5hmC
in vitro
(
Jin, Kadam, &
Pfeifer, 2010
), which suggests that 5mC oxidation could serve to displace
MBD binding and relieve transcription silencing. Inversely, MBD3, another
member of the MBD family that was long known to have low-binding affin-
ity for 5mC, was proposed to bind to 5hmC-enriched sequences in mouse
ES cells (
Yildirim et al., 2011
). Knockdown of TET1 impairs MBD3 bind-
ing to its genomic targets, suggesting that MBD3 binding to chromatin is
5hmC dependent (
Yildirim et al., 2011
). Another study identified MeCP2
as a major 5hmC-binding protein in brain cells (
Mellen, Ayata, Dewell,
Kriaucionis, & Heintz, 2012
), although earlier studies reported that MeCP2
does not efficiently bind to 5hmC-containing DNA
in vitro
(
Valinluck et al.,
2004
). Finally, quantitative mass-spectrometry-based proteomics recently
identified many novel proteins that recognize 5hmC in mouse embryonic
and neuronal cells (
Spruijt et al., 2013
), which further suggests that
5hmC could affect gene expression states by modulating the binding of pro-
tein complexes to chromatin. In the future, it will be of great interest to
characterize these 5hmC “readers” and study the stability of 5hmC in the
cell and potential mechanisms that maintain 5hmC during DNA replication.
6. CONCLUSIONS
The research in the past years has led to tremendous progress in the
characterization of cytosine methylation patterns in mammalian genomes,
as well as the discovery that other forms of modified cytosines play important
roles in epigenetic reprogramming and signaling. Whole-genome approaches
give us a detailed view of DNA modifications at high resolution and reveal
that DNA methylation profiles are more dynamic than previously anticipated
and can be regulated upon a variety of stimuli. At CpG islands, DNA meth-
ylation is generally not the initiating event in gene silencing but reinforces
long-term silencing. Further studies are now needed to understand the role
of DNA methylation at other sequences (gene bodies and enhancers), solve
the cause/consequence debates, and elucidate the mechanisms that establish
or remove methylation marks at specific targets. This is important for a better
understanding of not only themolecular mechanisms of development, but also
