Biology Reference
In-Depth Information
strategy, it is possible that genomic clustering of
Hox
genes is mostly used to
efficiently maintain their repressed states, via 3D compartmentalization,
rather than to help coordinating their transcriptional activation, as may be
the case in vertebrates (see below). This “repressive” function of
Hox
clus-
tering in flies may explain why these genes are still clustered in different sub-
species where the homeotic complexes have been reorganized (
Fig. 4.2
).
7. A 3D CHROMATIN TIMER FOR VERTEBRATE
COLLINEARITY?
The 3D conformations of mammalian
Hox
clusters have been deter-
mined by 4C during the implementation of both spatial and quantitative col-
linearities. In E10.5 embryonic brain cells, the interaction profiles revealed
that inactive
Hox
clusters are organized as distinct local 3D compartments
(
Fig. 4.7
A;
Noordermeer et al., 2011
). The genomic limits of these com-
partments strikingly coincide with the domains of H3K27me3 modifica-
tions. As in
Drosophila
, the efficient maintenance of
Hox
gene repression
appears to be accompanied by a 3D compartmentalization mechanism,
which may thus be an ancestral feature of bilateria embryogenesis. Little sub-
structure is observed within these inactive compartments, suggesting that the
chromatin displays a random 3D organization. However, the mammalian
CTCF insulator protein (
Herold, Bartkuhn, & Renkawitz, 2012
), may pro-
vide
some scaffolding for local 3D compartmentalization. Mammalian
HOX
clusters contain series of CTCF binding sites, due to the unusually high con-
centration of GC islands, with the majority of them being located in
between the (posterior) 5
0
-located genes. 3D modeling, using a dataset from
cells where
Hox
clusters are inactive, indicated that CTCF binding sites are
found in close spatial proximity. As CTCF sites are numerous at the 5
0
extremity of the clusters, 3D compartments may partially nucleate around
these sites (
Ferraiuolo et al., 2010
).
The activation of distinct subsets of
Hox
genes, at different AP domains,
is paralleled by a dynamic 3D reorganization of chromatin, different from
what is observed in
Drosophila
. In E10.5 cells obtained from the developing
primary AP axis, the inactive 5
0
-located
Hox
genes remain organized in a
local 3D compartment of restricted size, expectedly matching the
H3K27me3 domain. In addition, active genes located at more 3
0
positions
and marked by H3K4me3 are also organized in a discrete local 3D compart-
ment, which is distinctly separated from the negative domain (
Fig. 4.7
B;
Noordermeer et al., 2011
). At a later developmental stage, the
HoxC
cluster
