Biology Reference
In-Depth Information
replication-independent active demethylation involving DNA repair
(
Fig. 2.6
). Recent studies have shown that the DNA glycosylase TDG (thy-
mine DNA glycosylase) can efficiently remove 5caC and 5fC fromDNA (
He
et al., 2011; Maiti & Drohat, 2011
), which suggests that 5caC or 5fC could
trigger removal by TDG followed by base excision repair (BER). TDG, how-
ever, seems to have no activity on 5hmC (
Maiti & Drohat, 2011
). In support
of a role of TDG in epigenetic reprogramming,
Tdg
/
mice show embry-
onic lethality and hypermethylation of a number of developmental genes
(
Cortazar et al., 2011; Cortellino et al., 2011
). Another possible route initiated
by the TET proteins is deamination of 5hmC by the AID/APOBEC family of
deaminases, thereby creating 5hmU that could be removed by BER-mediated
DNA repair (
Fig. 2.6
). This scenario is supported by recent findings showing
that DNA glycosylases TDG and SMUG1 (single-strand-selective mono-
functional uracil-DNA glycosylase 1) can excise 5hmU from DNA
(
Cortellino et al., 2011; Guo, Su, Zhong, Ming, & Song, 2011
). It has also
been shown that AID/APOBEC deaminases use 5hmC as a preferential sub-
strate and promote 5hmC-dependant demethylation of a reporter construct in
combination with BER enzymes in HEK293 cultured cells (
Guo, Su, et al.,
2011
). In the same study, the authors also provide experimental evidence that
TET1 and AID/APOBEC deaminases participate in neuronal activity-
induced demethylation of endogenous targets (
Guo, Su, et al., 2011
). The
physiological relevance of this pathway still needs to be clarified as these deam-
inases could also act directly on 5mC and trigger demethylation independent
of TET activity (
Fig. 2.6
). In addition,
Aid
/
mice are viable and only show
a moderate defect in DNA methylation erasure in PGCs (
Popp et al., 2010
),
indicating that AID-mediated deamination is only responsible for a small part
of demethylation
in vivo
. Finally, one last scenario is that a deformylase or
decarboxylase directly converts 5fC or 5caC to C (
Fig. 2.6
). In summary,
experimental evidence point toward a variety of putative TET-dependent
demethylation pathways. Additional investigations are needed to determine
their relative contribution to physiological demethylation processes, which
is currently difficult because the precise kinetics of the modified cytosine forms
cannot be determined as traditional bisulfite sequencing does not discriminate
between 5mC and 5hmC, or 5caC and C.
5.2. Putative roles in gene regulation
In addition to being an intermediate of demethylation, 5hmC might also
function as a stable signal that modulates the binding of chromatin effectors
