|
Thursday June 3/Evening
SMALL RNAS I
Patrick Paddison
Double-stranded RNA (dsRNA) induced
gene silencing or RNA interference (RNAi) underlies many
homology-dependent epigenetic silencing phenomena in eukaryotes,
including cosuppression, virus-induced gene silencing, transgene-induced
silencing, quelling, etc. This session was devoted to the role of the
RNAi pathway in gene and viral silencing and heterochromatin formation
as well some of the nuts and bolts of the uptake and process of dsRNA
triggers of gene silencing. Here
I will present some of the background and
highlights of talks given by David Baulcombe, Robin Allshire,
Shiv Grewal, and David Bartel.
Two
of the hallmarks of RNAi in both C. elegans and plants are
“transitive” silencing and systemic transport of the silencing
effect throughout the organism.
Transitive silencing occurs when sequences of the RNA target,
which are not included in the original dsRNA trigger, are incorporated
as part of RNAi triggers (ie, siRNAs) in subsequent rounds of silencing.
Transitive silencing is likely part of an “amplification”
step of the initial silencing triggers and is mediated by RNA-dependent
RNA polymerases (RdRPs) which may use early siRNAs produced as primers.
Both transitive RNAi and systemic transport of siRNA likely act
in concert as a natural cell-based immunity against viral and genetic
parasites in nematodes and plants.
Genetic analysis of genes required for both transitive and
systemic silencing has suggested roles for several RdRPs, including
rrf-1 and ego-1 in C.elegans and SDE1/SGS2 in plants.
David
Baulcombe presented his group’s work on one RdRP in particular,
SDE1/SGS2/RDR6. Using RNAi
to silence RDR6 in grafting experiments and virus-induced silencing
experiments in tobacco, Baulcombe and colleagues showed that RDR6 is
likely required for the receipt and/or amplification of the silencing
signal, but not its production, during the spreading of RNAi throughout
the plant. Additionally
Baulcombe presented genetic screens in plants for mutants which either
enhance transgene induced silencing or, alternately, result in partial
loss of transgene-induced gene silencing. Interestingly, in all
“enhanced silencing” mutants arising from the screen, the
enhancement phenotype could be suppressed by removing RDR6, suggesting
that these mutants may act through enhancing an amplification step.
Recently, a related screen in C. elegans carried out in
Gary Ruvkun’s lab for RNAi enhancers led to the discovery of a gene
coding for an siRNA-nuclease; mutants in the siRNAase gene produced more
stable and sustained gene silencing.
Characterization
of three partial loss of transgene silencing mutants, SDE4 (silencing
defective4), RDR2 (rna dependent rna polymerase2), and AGO4
(argonaute4), showed that these mutants have extensive loss of SINE
element silencing at the chromatin level.
Baulcombe suggested that the partial loss of transgene silencing
in these mutants is an indication that transcriptional gene silencing (TGS)
reinforces post-transcriptional gene silencing (PTGS) of transgenes.
Baulcombe ended by suggesting that the RNAi pathways responsible
for TGS and PTGS are
functionally independent, containing separate Dicer, Ago, and RdRP
paralogs. Interestingly,
the Dicer required for TGS (presumably Dicer-like 3) produces 24-26nt
siRNAs rather than the usual 21-22nt siRNAs found in most organisms.
Robin
Allshire and Shiv Grewal each presented their groups’ work on RNAi-mediated
heterochromatin formation and maintenance in fission yeast.
S. pombe harbors homologues of Dicer (dcr1), Argonaute
(ago1), and an RdRP (rdp1).
All are required for proper kinetochore function and silencing of
the mating-type region and transposons as well as hairpin-mediated
silencing of endogenous genes (eg, ura4).
The RNAi pathway in S. pombe targets histone modification
to nucleate heterochromatin formation, acting in concert with Clr4, a
lysine 9 histone H3 methyltransferase (ortholog of Suv39), and Swi6, a
chromodomain protein (ortholog of HP1). Allshire and colleagues have previously
demonstrated that both Clr4 and Swi6 are required for the silencing of
centromeric repeats and proper segegration of sister chromatids during
mitosis. Recently, the Grewal, Allshire and Martienssen groups
have demonstrated that the RNAi mutants in S. pombe also sport
similar phenotypes.
One
looming question, as Grewal stated in his talk, is how to link the
production of siRNAs, a true hallmark of RNAi-related phenomena, to the
targeting of euchromatin and nucleation of heterochromatin.
The Grewal and Allshire groups have taken complementary
approaches, modeling chromatin silencing using cenH region in the
mating-type region or a short hairpin RNA expressed from a plasmid
targeting the ura4 gene, respectively, and, of course, the power of
fission yeast genetics, to begin dissecting the question of just how
dsRNA and siRNAs nucleate the formation of heterochromatin in fission
yeast.
Endogenously
expressed small hairpin RNAs regulate gene expression through the RNAi
pathway during C. elegans
development. These small
hairpin RNAs (~70nt) are processed into a 21-22nt mature form by Dicer
and then used to seek out mRNA targets of similar sequence (often via
imperfect base-pairing interactions).
For the two prototypes of this family, C.
elegans lin-4 and let-7, silencing occurs at the level of protein
synthesis. The first small
hairpin RNAs were dubbed small temporal RNAs (stRNAs), owing to their
role in developmental timing. More
recently, dozens of orphan hairpins (ie, triggers without identified
mRNA targets) have been identified in C. elegans, Drosophila,
plants, mouse, and humans, which are collectively referred to as
microRNAs (miRNAs).
MiRNAs
are among the most abundant gene-regulatory elements in multicellular
eukaryotes, constituting almost 1% of the predicted genes in worms,
flies, plants and humans. David
Bartel estimates that there are approximately 200-255 human, 110-220 C.
elegans, and 92 Arabidopsis miRNA genes.
In Arabidopsis, miRNAs can be broken down into ~22 gene
families with ~83 unique targets, which consist of 63 predicted
transcription factors, 5 F-box proteins, and interestingly 2 genes in
the RNAi pathway, Dicer-like and Ago1.
Many, if not most, of characterized plant miRNAs result is
cleavage of their cognate target, rather than just translational
repression, as is the case for the archetypal miRNAs, let-7 and lin-41,
from C. elegans and predicted to be the case for most metazoan miRNAs.
Interestingly, Bartel presented data showing that in mammals one
miRNA, miR-196, which maps to homeobox clusters in mammals, also results
in the cleavage of its target, the HOXB8 gene, during mouse
embryogenesis.
Bartel also presented computational
predictions of mammalian miRNA targets.
He suggested more than 400 target genes by identifying mRNAs with
matched pairing to the 5’ region of the miRNA, multiple target gene
hits and evolutionary
conservation in human, mouse, rat and puffer fish. Although mammalian miRNA targets were enriched for genes
involved in transcriptional regulation, unlike miRNA targets from Arabidopsis,
most gene targets fell into other functional categories.
Bartel
ended with what he calls the “micromanager” model miRNA-dependent
regulation of gene expression. In
this model miRNAs serve to dampen the expression of thousands of mRNAs
to allow customized expression of genes in different cell types.
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) |
Patrick Paddison
(Hannon lab)
2
0
0
4
2
0
0
4
2
0
0
4
2
0
0
4
|