|
Saturday June 5/Morning
CHROMATIN INHERITANCE & ASSEMBLY
Maarten Hoek
The epigenetic
regulation of transcriptional states ultimately converges on the
assembly and disassembly of higher-order chromatin structures. This
morning’s session centered on the machinery that acts to assemble
active and silent chromatin. Steve Henikoff (Fred Hutchinson Cancer
Research Center, US) in the previous session demonstrated that histone
H3.1 is exclusively deposited onto DNA during DNA replication whereas
the histone variant H3.3 deposits outside of S-phase onto
transcriptionally active DNA. These observations were extended by
Genevieve Almouzni (CNRS, France), who described recent experiments done
in collaboration with the group of Yoshihiro
Nakitani (Harvard University) in which the H3.1 and H3.3 histone
variants were double affinity tagged and purified. H3.1 was found in a
complex exclusively with the replication coupled chromatin assembly
factor CAF-1 whereas H3.3 bound the replication independent chromatin
assembly factor HirA. In both purifications only the tagged H3 variants
were recovered (as well as H4), even though purified tagged nucleosomes
contained endogenous untagged H3 in a 1:1 molar ratio. This implied that
soluble H3 and H4 exist as dimer, at least when in complex with histone
chaperones – a result that is at odds with years of traditional
histone biochemistry in which H3 and H4 always purify as a
heterotetramer under conditions of low ionic strength. Based on these
provocative findings, Almouzni proposed that the old H3/H4 tetramer
might split during replication and segregate semi-conservatively onto
each daughter strand, thus providing a way to replicate epigenetic marks
found on these histones.
The problem of
remembering epigenetic marks was also discussed in two talks on the fly
polycomb complex. Renato Paro (University of Heidelberg, Germany)
reported on elements in the Drosophila bithorax gene cluster that
confer transcriptional memory after an initial transcriptional
activation event. The polycomb group proteins are recruited to DNA
elements called PREs in this cluster and repress transcription.
Silencing can be countered by the establishment of transcription and
this active state is then retained throughout development. This is
accompanied by the accumulation of transcripts corresponding to the
memory element. He showed today that transcription of PREs is required
for the maintenance of transcriptional memory in a reporter system in
which the Fab7 PRE controlled expression of the mini-white gene. In the
presence of transcription through the PRE, the mini-white gene remained
on through development, but if transcription of the PRE was inhibited by
the insertion of a terminator between the promoter and the PRE,
silencing was restored to this locus. This appeared to require
transcription by a full-fledged RNA polymerase II because Paro mentioned
that transcription of the PRE by T7 polymerase was not sufficient for
the establishment of transcriptional memory. Inspired by the work of
Steve Henikoff, Paro suggested that the replacement of H3.1 with H3.3
during transcription might be the epigenetic mark that confers
transcriptional memory on the PRE.
Bob Kingston (Massachusets
General Hospital, USA) gave some insight into the nature of Polycomb
repressed DNA by using electron microscopy to probe the structure of
Polycomb bound chromatin. He showed that a reconstituted Polycomb group
complex was able to bind to nucleosomal templates and compact them in
vitro and that this worked most efficiently on nucleosomal arrays
containing three to four nucleosomes. Histone tails were not required
for compaction, which came as some surprise since other work discussed
by Paro implied that trimethylated histone tails might be important in
polycomb recruitment. Mutations that strongly affected the
transcriptional activity of the Psc1 component of the complex also
abolished Psc1 dependent compaction suggesting that the compaction
activity was physiologically relevant. Since mononucleosomes are
inefficient substrates for compaction, Kingston suggested that there may
be multiple nucleosome binding pockets in the complex.
Craig
Peterson (University of Massachusetts) also presented some biophysical
experiments to examine the effects of specific SIN mutations in the
nucleosome core on chromatin compaction. Many of the SIN mutations,
which relieve the need for the SWI/SNF chromatin remodeling machinery in
mitotic transcription, map to the histone core. A mutant R45C histone H4
that confers the SIN phenotype was incorporated into recombinant octamer
and assembled into phased nucleosome arrays. These arrays were compared
to wild type nucleosome arrays by analytical ultracentrifugation. Normal
chromatin compacted into a structure consistent with a 30nM chromatin
fiber in the presence of Mg++, but SIN mutant chromatin could
not compact. These results imply that the histone SIN mutations relieve
the need for SWI/SNF dependent remodeling in transcription by altering
the compaction of chromatin. Peterson also showed some interesting work
examining the requirement for SWI/SNF in activated transcription of the
Gal locus, which is repressed by the addition of glucose and activated
by the addition of galactose. When the locus was activated, then
repressed for 20 min, and again reactivated, induction kinetics for the
second activation were extremely rapid. In a swi2∆
mutant, however, the induction kinetics for the second activation
matched that of the first activation, implying that SWI/SNF is involved
in maintaining transcriptional memory as well. Peterson mentioned that
SWI/SNF remains bound to chromatin for at least 2h following induction,
and suggested that this might confer the transcriptional memory seen at
this locus.
The
mechanism of chromatin remodeling was discussed for the CHRAC complex by
Peter Becker. He showed the ISWI component of CHRAC could slide
nucleosomes along a DNA substrate by itself but that this activity was
enhanced an order of magnitude by the addition of the ACF subunit to
this reaction. ACF also suppressed the center to edge sliding activity
of ISWI and worked better at sliding a nucleosome from the edge to the
center of a substrate. He showed that the PHD fingers of ACF are
critical for the enhancing activity and that these domains bind
nucleosome. Thus, the PHD fingers may provide the anchor on nucleosomes
required for the sliding reaction. The p14 and p16 subunits of CHRAC
also enhance the sliding activity of CHRAC, and this may relate to their
weak DNA binding activity because mutations in p16 that enhance DNA
binding inhibit CHRAC activity. Becker suggested that p14 and p16 might
function as a “DNA chaperone” that might prime the DNA for movement
relative to the histone octamer.
Finally,
in a talk that was somewhat less related to chromatin structure, Tim
Bestor discussed the effect of deleting DNMT3L on the development of
male germ cells. DNMT3L is a catalytically inactive DNMT3 family member
that interacts with the de novo
DNA methyltransferase DNMT3a. DNMT3L deficiency resulted in meiotic
catastrophe in spermatocytes and in the transcription of the IAP
retrotransposon as well as LINE elements. This is the first mutation
known to reactivate LINE elements in the genome. Methylation of these
transposable elements was sharply reduced whereas methylation at the
imprinted H19 gene was reduced
by only 50% and the gene was not reactivated. Based on these results as
well as on a crystal structure of DNMT3L, Bestor suggested that this
protein participates in a homology search of the genome for dispersed
repeats – leading to their methylation perhaps through DNMT3a. In the
absence of DNMT3L, meiotic catastrophe might result either from high
transposase activity leading to double strand breaks or due to the
synapsis of non-allelic repeats.
Other
Dispatches
Symposium
69 Live
Symposia
Past (a bit of history and photographs from previous Symposia)
Online
Symposium Volumes (searchable database of past Symposia volumes and
currently received manuscripts) |
Maarten Hoek
(Stillman lab)
2
0
0
4
2
0
0
4
2
0
0
4
2
0
0
4
|