Downloaded from symposium.cshlp.org on April 9, 2013 - Published by Cold Spring Harbor Laboratory Press
A Transcription Fork Model for Pol IV and Pol V−dependent
RNA-Directed DNA Methylation
C.S. Pikaard, J.R. Haag, O.M.F. Pontes, et al.
Cold Spring Harb Symp Quant Biol published online April 8, 2013
Access the most recent version at doi:10.1101/sqb.2013.77.014803
P<P
Email alerting
service
Published online April 8, 2013 in advance of the print volume.
Receive free email alerts when new articles cite this article - sign up in
the box at the top right corner of the article or click here
Advance online articles have not yet appeared in the print volume. Citations to Advance
online articles must include the digital object identifier (DOI) and date of initial
publication.
To subscribe to Cold Spring Harbor Symposia on Quantitative Biology go to:
http://symposium.cshlp.org/subscriptions
Copyright © 2012 Cold Spring Harbor Laboratory Press; all rights reserved
Downloaded from symposium.cshlp.org on April 9, 2013 - Published by Cold Spring Harbor Laboratory Press
A Transcription Fork Model for Pol IV and
Pol V – dependent RNA-Directed DNA Methylation
C.S. PIKAARD,1,2 J.R. HAAG,1,2 O.M.F. PONTES,3 T. BLEVINS,1,2 AND R. COCKLIN1
1
Department of Biology and Department of Molecular and Cellular Biochemistry,
Indiana University, Bloomington, Indiana 47405
2
Howard Hughes Medical Institute, Indiana University, Bloomington, Indiana 47405
3
Department of Biology, University of New Mexico, Albuquerque, New Mexico 87131
Correspondence: [email protected]
In Arabidopsis thaliana, nuclear multisubunit RNA polymerase IV (Pol IV) and RNA-DEPENDENT RNA POLYMERASE 2
(RDR2) are required for the biogenesis of 24-nucleotide small interfering RNAs (siRNAs) that direct DNA methylation and
transcriptional silencing at target loci transcribed by nuclear multisubunit RNA polymerase V (Pol V). Pol IV and RDR2
physically associate and RDR2’s polymerase activity in vitro is dependent on Pol IV. RDR2 transcription of nascent Pol IV
transcripts might result in discontinuous second strands, analogous to lagging-strand Okazaki fragments generated during
DNA replication. In vitro, Pol V is unable to displace nontemplate DNA during transcriptional elongation. This suggests a
need for DNA duplex unwinding by helper proteins, perhaps analogous to the helicase-mediated duplex unwinding that occurs
at replication forks to enable leading strand synthesis by DNA polymerase 1. A multiprotein complex (DRD1, DMS3, DMS11,
RDM1) known to enable Pol V transcription might facilitate duplex unwinding via ATP-dependent DNA translocase, singlestranded DNA binding, and cohesin-like strand capture activities. These considerations are discussed and incorporated into a
“transcription fork” model for Pol IV and Pol V–dependent RNA-directed DNA methylation.
In eukaryotes as diverse as fission yeast, plants, and
humans, short interfering RNAs (siRNAs) or piwi-interacting RNAs ( piRNAs) direct heterochromatic modification and transcriptional silencing as a way of taming
transposons and other repetitive elements (Zaratiegui
et al. 2007; Brosnan and Voinnet 2009; Moazed 2009;
Ishizu et al. 2012). The model plant Arabidopsis thaliana
has a diverse inventory of machinery for small RNA biogenesis and function, including 10 ARGONAUTE family
proteins (AGO1 through AGO10), six RNA-dependent
RNA polymerases (RDR1 through RDR6), four DICER
endonucleases (DCL1 through DCL4) and two plant-specific multisubunit DNA-dependent RNA polymerases,
abbreviated as Pol IV and Pol V (Brodersen and Voinnet
2006; Wassenegger and Krczal 2006; Mallory and Vaucheret 2010; Haag and Pikaard 2011). Transcriptional silencing of transposons and endogenous repeats is achieved
by an RNA-directed DNA methylation (RdDM) pathway
that involves 24-nucleotide siRNAs generated by Pol IV,
RDR2, and DCL3. These 24-nucleotide siRNAs are bound
by AGO4 and interact with long noncoding RNA transcripts of Pol V (Wierzbicki et al. 2008, 2009), or sometimes Pol II (Zheng et al. 2009; Stroud et al. 2013), to bring
about the cytosine methylation of corresponding DNA
sequences by the de novo DNA methyltransferase
DRM2 (DOMAINS REARRANGED METHYLTRANSFERASE 2) (Cao and Jacobsen 2002; Matzke et al. 2009;
Haag and Pikaard 2011; Zhang and Zhu 2011; Wierzbicki
2012).
A MECHANISTIC PARTNERSHIP BETWEEN
POL IV AND RDR2
Pol IV and RDR2 are required for virtually all 24-nucleotide siRNA biogenesis such that 24-nucleotide
siRNAs are essentially eliminated in Pol IV- or RDR2null mutants (Xie et al. 2004; Herr et al. 2005; Onodera
et al. 2005). Pol IV’s partnership with RDR2, as opposed
to other RNA-dependent RNA polymerases, can be explained by their physical interaction in vivo, as revealed
by mass spectrometry and coimmunoprecipitation studies
(Law et al. 2011; Haag et al. 2012). Affinity-purified
Pol IV –RDR2 complexes display RNA polymerase activity in vitro using templates comprised of DNA and
RNA oligonucleotides (Haag et al. 2012). Pol IV isolated
from wild-type or rdr2 mutant backgrounds is similarly
active in vitro, indicating that Pol IV does not require
RDR2 for its enzymatic activity. In contrast, RDR2 isolated from a pol IV – null mutant lacks detectable activity.
Pol IV bearing point mutations in the catalytic site lacks
transcriptional activity but is not impaired in its ability to
associate with RDR2; in this case, RDR2 displays activity
in vitro. Collectively, these results suggest that RDR2’s
interaction with Pol IV, or a Pol IV – associated factor, is
required for RDR2 activity (Haag et al. 2012).
The physical basis for RDR2’s association with Pol IV
is unknown, but Pol IV transcripts or other RNAs do
not appear to mediate the interaction based on RDR2’s
association with catalytically dead Pol IV, as discussed
Copyright # 2012 Cold Spring Harbor Laboratory Press; all rights reserved; doi: 10.1101/sqb.2013.77.014803
Cold Spring Harbor Symposia on Quantitative Biology, Volume LXXVII
1
Downloaded from symposium.cshlp.org on April 9, 2013 - Published by Cold Spring Harbor Laboratory Press
2
PIKAARD ET AL.
above, or the failure of RNase treatment to disrupt Pol IV –
RDR2 association (Haag et al. 2012). Most likely, protein – protein interactions occur between RDR2 and one
or more Pol IV subunits or associated proteins. Other
RNA-dependent RNA polymerases (RdRPs) have been
shown to require association with host factors for activity.
For instance, structural studies of Qb virus RdRP have
shown that association with the bacterial host proteins
EF-Tu (elongation factor thermo unstable) and EF-Ts
(elongation factor thermo stable) (Blumenthal and Carmichael 1979) alters the structure of the RdRP, converting it
into an active form (Kidmose et al. 2010; Takeshita and
Tomita 2010).
Evidence that Pol IV and RDR2 interact, and that
RDR2 activity is dependent on Pol IV, suggests that their
activities are coupled for the channeled synthesis of double-stranded RNAs (dsRNAs) that are destined for DCL3
dicing. Consideration of how this mechanistic coupling
might be achieved begs the question of whether RDR2
begins transcribing Pol IV nascent transcripts before Pol
IV transcription termination. If so, a series of relatively
short antisense RNAs might be produced by RDR2, analogous to lagging strand Okazaki fragments that are generated during DNA replication (Fig. 1) (Zheng and Shen
2011). Alternatively, RDR2 might only transcribe complete Pol IV transcripts following their release from Pol
IV, with RDR2 initiating from the free 30 ends and generating full-length dsRNA products (Fig. 1). Both activities are plausible based on studies of the Neurospora
RdRP, QDE1 (QUELLING DEFECTIVE 1), which is
capable of end-to-end transcription of RNA templates
as well as internal primer-independent initiation (Aalto
et al. 2010). QDE1 even has robust activity using singlestranded DNA templates (Aalto et al. 2010), as does Arabidopsis RDR6 (Curaba and Chen 2008), demonstrating
the potential template versatility of single subunit RdRPs.
Despite evidence that RDR2 activity is Pol IV dependent in vitro, most RDR2 in interphase nuclei does not
colocalize with Pol IV. As shown in Figure 2 (see also
Pontes et al. 2006), Pol IV is typically detected as numerous puncta distributed throughout the nucleoplasm
but is absent from the nucleolus, which appears as a black
hole in nuclei stained with the DNA-binding fluorescent
dye DAPI. In contrast, RDR2 typically localizes within a
prominent crescent, or ring, along the inner perimeter of
the nucleolus, in addition to being present in the nucleoplasm (Fig. 2, top row, red signals). In most interphase leaf
nuclei (74%, n ¼ 501), there is no obvious overlap in the
Pol IV and RDR2 signals. However, in approximately one
quarter (26%, n ¼ 501) of the interphase nuclei—specifically, nuclei in which RDR2 tends to be more abundant in
the nucleoplasm relative to the nucleolus—overlap in the
Pol IV and RDR2 signals is observed. Taken together, the
mass spectrometry and immunoprecipitation results indicate that Pol IV and RDR2 can stably associate, but the
immunolocalization studies suggest that the enzymes are
not obligate partners. We speculate that the pool of RDR2
not associated with Pol IV is involved in dsRNA amplification and/or trafficking, possibly still in association with
auxiliary proteins or posttranslational modifications (such
as phosphorylation) acquired in a Pol IV –dependent manner. The plausibility of this speculation is again supported
by the known activities of Neurospora QDE1, which is
self-sufficient for production of initial siRNA precursor
transcripts as well as amplification of dsRNAs (Lee et al.
2010).
A protein implicated in the functioning of the Pol
IV –RDR2 complex is CLASSY1 (Smith et al. 2007),
a member of a large family of ATP-dependent nucleic
acid translocases with diverse functions, including nucleosome repositioning or eviction (e.g., SWI2/SNF2,
ISWI) (Clapier and Cairns 2009) or branch migration at
Holliday junctions during homologous recombination or
DNA repair (e.g., RAD54) (Mazin et al. 2010). CLSY1
physically associates with affinity-purified Pol IV – RDR2
complexes, as do several related CLSY family members
Pol IV
Chromosomal DNA
RDR2
5'
RDR2 initiation on nascent
Pol IV transcripts
RDR2 transcription of released
Pol IV transcripts
Pol IV
Pol IV
RDR2
5'
dsRNA
3'
5'
3'
5'
RDR2
5'
3'
5'
5'
3'-OH
5'
5'
3'
3'
5'
RDR2
5'
5'
3'
Figure 1. Potential outcomes of Pol IV and RDR2 coupling. Initiation of RDR2 transcription before Pol IV transcription termination
might result in RDR2’s production of short Okazaki-like RNA fragments. In contrast, RDR2 transcription of terminated and released
Pol IV transcripts would allow RDR2 to synthesize full-length complementary strands.
Downloaded from symposium.cshlp.org on April 9, 2013 - Published by Cold Spring Harbor Laboratory Press
RNA-DIRECTED DNA METHYLATION
3
Figure 2. Localization patterns of Pol IV and RDR2 in interphase nuclei. Immunolocalization of RDR2 (red) and epitope-tagged
NRPD1-FLAG (green) that is known to be fully functional based on its ability to complement an nrpd1-null mutation. DNA is stained
with DAPI (blue). Overlapping RDR2 and NRPD1-FLAG signals are yellow. The frequency of nuclei displaying each pattern, among
501 nuclei examined, is shown to the right.
(Law et al. 2011). Immunolocalization of CLSY1 shows
that its nuclear distribution more closely resembles that of
RDR2 than of Pol IV. Moreover, RDR2 nuclear localization is significantly disrupted, and Pol IV localization only
partially disrupted, in clsy1 mutants (Smith et al. 2007).
Collectively, these observations suggest that CLSY1 has a
role at the interface between Pol IV and RDR2 as well as
possible downstream dsRNA amplification steps performed by RDR2 alone. Such an activity might be analogous to the partnership between the RecQ DNA helicase
QDE3 and the Neurospora RdRP QDE1 for the production of dsRNA precursors of siRNAs (Lee et al. 2010).
Whether other CLSY1-related family members that interact with Pol IV are partially redundant with CLSY1 is
unclear, because no biochemical activities have yet been
assigned to any member of the family.
MYSTERIES OF POL IV AND POL V
RECRUITMENT
How Pol IV and Pol V are recruited to their sites
of action is unknown. Chromatin immunoprecipitation,
combined with deep sequencing of associated DNA sequences, has recently revealed sites of Pol V occupancy in
the Arabidopsis genome (Wierzbicki et al. 2012; Zhong
et al. 2012). However, consensus sequences that might
point to Pol V promoter sequences have not yet emerged.
Analogous chromatin immunoprecipitation – sequencing
(ChIP-seq) studies of Pol IV – associated sequences have
not been reported, but analyses of the loci corresponding
to Pol IV – dependent siRNAs have not revealed consensus sequences that account for Pol IV involvement at these
sites.
One possibility is that Pol IV and Pol V transcription is
initiated from specific nucleic acid structures, such as R
loops formed by the hybridization of RNA to duplex DNA,
thus displacing one DNA strand. One clue in support of
this hypothesis is that in vitro, we have only been able to
detect Pol IV or Pol V transcription primed by an RNA
oligonucleotide hybridized to a DNA template (Haag et al.
2012). Whether this RNA priming requirement in vitro
reflects reality in vivo is unclear, but mislocalization of
Pols IV and V in nuclei treated with RNase is consistent
with RNA having some role in Pol IV and Pol V recruitment at specific sites (Pontes et al. 2006). Moreover, if
RNAs that prime Pol IV or Pol V transcription were derived
from Pol II transcription, this might explain Pol II’s involvement in Pol IV and Pol V chromatin association, as
detected by ChIP (Zheng et al. 2009).
Another possibility is that chromatin modifications
have a role in Pol IV or Pol V recruitment, as suggested
by the apparent feed-forward loop connecting cytosine
methylation to Pol IV –dependent siRNA biogenesis.
For instance, in drm2 mutants, as in other mutants disrupting DNA methylation or chromatin modification, Pol IV –
dependent siRNA levels are often reduced (Zilberman
et al. 2004; Onodera et al. 2005) even though these modifications are thought to occur downstream from siRNA
biogenesis in the RdDM pathway. A recent study has
also shown that proteins required for maintenance cytosine methylation positively influence de novo cytosine
methylation patterns attributable to DRM2 and the
RNA-directed DNA methylation pathway (Stroud et al.
2013).
Among the proteins whose mutation brings about a
reduced siRNA abundance phenotype, similar to mutants
affecting cytosine methylation or other steps downstream
from siRNA biogenesis, is SHH1/DTF1 (SAWADEE
HOMEODOMAIN HOMOLOG 1; DNA-BINDING
TRANSCRIPTION FACTOR 1), a protein that affects
siRNA levels and RdDM at a subset of Pol IV –dependent
loci (Law et al. 2011; Liu et al. 2011). Interestingly,
SHH1/DTF1 is not implicated in the downstream part
of the RdDM pathway but physically associates with
Pol IV. Because homeodomains typically confer nucleic
acid binding, it is possible that SHH1/DTF1 has a transcription-factor-like role in recruiting Pol IV to loci that
display specific patterns of DNA methylation or chromatin modification.
A POL V “TRANSCRIPTION FORK” MODEL
In vitro, Pol V will elongate an RNA primer hybridized
to a single-stranded DNA template (Haag et al. 2012).
However, if a nontemplate DNA oligonucleotide is annealed to the template just downstream from the RNA
Downloaded from symposium.cshlp.org on April 9, 2013 - Published by Cold Spring Harbor Laboratory Press
4
PIKAARD ET AL.
primer, no Pol V transcription has been observed in our
initial experiments. This suggests that Pol V transcription elongation is impeded by downstream duplex DNA
(Fig. 3). In contrast, Pol IV and Pol II are able to transcribe through dsDNA regions (Haag et al. 2012).
The potential ramifications of Pol V’s apparent need for
single-stranded DNA are intriguing in light of the composition and putative activities of a Pol V helper complex
consisting of DRD1 (DEFECTIVE IN RNA-DIRECTED
DNA METHYLATION 1) (Kanno et al. 2004), DMS3
(DEFECTIVE IN MERISTEM SILENCING 3) (Kanno
et al. 2008), and RDM1 (RNA-DIRECTED DNA METHYLATION 1) (Gao et al. 2010; Law et al. 2010). We
showed that DRD1, a putative DNA translocase/ATPase,
and DMS3, a protein that shares homology with the hinge
domain region of cohesins and condensins (Kanno et al.
2008), are required for synthesis of Pol V transcripts in
vivo (Wierzbicki et al. 2008, 2009). DRD1 and DMS3
were subsequently found to interact with RDM1 to form
the DDR complex, named for the first initials of the three
activities (Law et al. 2010). The DDR complex apparently
interacts with Pol V to some extent, based on detection of
Pol V subunits in purified DDR fractions (Law et al. 2010)
and is required for Pol V’s stable association with chromatin (Wierzbicki et al. 2008, 2009; Zhong et al. 2012).
The RDM1 protein of the DDR complex has singlestranded DNA-binding activity in vitro (Gao et al. 2010)
that is noteworthy in light of Pol V’s apparent need for
single-stranded DNA templates. RDM1 also appears to
have functions apart from the DDR complex, based on
its physical interactions with both AGO4 and DRM2,
the de novo cytosine methyltransferase responsible for
RdDM (Gao et al. 2010). Recently, DMS11 (DEFECTIVE IN MERISTEM SILENCING 11, also known as
MORC 6) was identified as a GHKL (gyrase, Hsp90,
histidine kinase, MutL) ATPase (Lorkovic et al. 2012;
Moissiard et al. 2012) that interacts with DMS3 (Lorkovic
et al. 2012), a partnership that might reconstitute a cohesin-like activity capable of nucleic acid strand capture
(Bender 2012; Lorkovic et al. 2012).
The in vitro template requirements of Pol V, the singlestranded DNA-binding activity of RDM1, the potential
helicase activity of DRD1 as an ATP-dependent DNA
translocase, and the putative cohesin-like potential of
the DMS3-DMS11 partnership all factor into the speculative model shown in Figure 4 (see also the model of
Zhang and Zhu 2011). The model envisions that DNA
duplex unwinding occurs in advance of elongating Pol V,
thus providing Pol V with a single-stranded template. An
analogy is the duplex unwinding accomplished by the
MCM helicase complex and the single-stranded DNAbinding protein Replication Protein A, at replication
forks, thereby facilitating leading strand synthesis by
DNA polymerase 1 (Bochman and Schwacha 2009; Remus and Diffley 2009; Zheng and Shen 2011). In this
“transcription fork” model, we speculate that the ATPase
activity of DRD1 is used to power the translocation of the
protein along single-stranded DNA, thus unwinding the
duplex. The putative cohesin-like activity of DMS3DMS11, together with RDM1’s single-stranded DNAbinding activity, may help to stabilize the unwound state,
with DMS3-DMS11 possibly acting as a clamp that aids
DRD1 processivity or directionality. These speculative
DMS3-DMS11 and RDM1 functions are analogous to
the roles of PCNA (Proliferating Cell Nuclear Antigen)
and Replication Protein A, respectively, at the DNA
replication fork (Remus and Diffley 2009; Zheng and
Shen 2011). Because RDM1 has an apparent life apart
from the DDR complex, we speculate that RDM1 may
bind multiple sites at a Pol V – transcribed locus and might
be loaded onto DNA as one function of the DDR complex.
RDM1’s reported interactions with both AGO4 and DRM2
(Gao et al. 2010) then suggest that RDM1 mediates the
recruitment of DRM2 to AGO4, which binds Pol V transcripts by siRNA base pairing (Wierzbicki et al. 2008,
2009) and physical interaction with the carboxy-terminal
domain (CTD) of the Pol V largest subunit (El-Shami et al.
2007).
It is noteworthy that Pol V transcripts are the RNA
version of the nontemplate DNA strand; thus, siRNAs
that basepair with Pol V transcripts could also potentially
basepair with the nontemplate DNA strand. This invites
the speculation that AGO4-siRNA complexes might
switch from an initial association with Pol V transcripts
to a direct interaction with the nontemplate DNA strand,
potentially facilitated by AGO4 slicing (Qi et al. 2006) of
Transcription
elongation?
3′
RNA 5′
primer
3′
RNA 5′
primer
Pol V
5′
Template DNA
Yes
Pol V
5′
3′
Template DNA
Nontemplate DNA
No
Figure 3. Summary of Pol V template requirements in vitro. Pol V will elongate an RNA hybridized to a single-stranded DNA
template. However, no transcription has been observed in initial experiments if dsDNA is present downstream from the RNA primer,
suggesting that Pol V cannot displace the nontemplate strand during elongation.
Downloaded from symposium.cshlp.org on April 9, 2013 - Published by Cold Spring Harbor Laboratory Press
RNA-DIRECTED DNA METHYLATION
5
Figure 4. A model for Pol IV and Pol V– dependent RNA-directed DNA methylation. Pol IV transcription initiation may be
programmed, in part, by chromatin modifications and/or context-dependent activating proteins, such as SHH1. The CLSY1 putative
ATP-dependent nucleic acid translocase is proposed to act at the interface between Pol IV and RDR2 as well as in subsequent dsRNA
amplification by RDR2 alone. Coupling of Pol IV and RDR2 results in channeled production of gapped or complete dsRNAs (see Fig.
1), which are diced into 24-nucleotide siRNAs by DCL3, methylated at their 30 ends by HEN1 (HUA ENHANCER 1) and loaded
primarily into AGO4. Pol V transcription is proposed to be facilitated by duplex unwinding catalyzed by DRD1’s ATP-dependent
translocation along single-stranded DNA, with the single-stranded DNA-binding activity of RDM1 and the putative cohesin-like
activities of DMS3-DMS11 having roles in generating or stabilizing the unwound state. AGO4 finds its sites of action by siRNA
basepairing to Pol V transcripts and physical interactions with the long C-terminal domain of the Pol V largest subunit (dark green tail),
as well as interactions with RDM1 bound to the extended Pol V transcription bubble. SPT5L/KTF1 bound to Pol V transcripts and to
the Pol V – transcribed DNA interacts with AGO4 to stabilize the complex. DRM2 is recruited by interaction with RDM1 and carries
out cytosine methylation in all sequence contexts, allowing subsequent recruitment of the H3K9 methyltransferases SUVH2 and
SUVH9 that bind methylated DNA via their SRA domains. Prior deacetylation of H3K9 by HDA6 (and/or related histone deacetylases) may be a prerequisite for H3K9 methylation by SUVH2, SUVH9, and SUVR2. Assembly of heterochromatin involves
nucleosome positioning mediated by the SWI/SNF chromatin remodeling complex, which is recruited by interactions with the
IDN2-IDP protein complex bound to the Pol V transcript.
Pol V transcripts to destabilize the siRNA –Pol V transcript interaction. Such a scenario might help to explain
the Pol V –dependent chromatin association of AGO4
(Wierzbicki et al. 2009; Rowley et al. 2011) and the specificity of RNA-directed methylation, whereby RNA triggers as short as 30 nucleotides are capable of bringing
about the methylation of the corresponding DNA sequence (Pelissier and Wassenegger 2000). However, there
is no definitive evidence for direct siRNA– DNA interactions. In fact, in Arabidopsis, the frequency at which a
cytosine is methylated is elevated for the DNA strand
that matches the siRNA rather than for the complementary
strand to which the siRNA could potentially basepair
(Lister et al. 2008). Interpretation of this latter observation is complicated by the fact that cytosine methylation
feeds back on siRNA production coupled with the observation of sequence bias among siRNAs (e.g., 24-nucleotide siRNAs are G rich and U poor; Lister et al. 2008)
that could favor the increased abundance of one strand
of a diced siRNA duplex over another. Recent studies
of 24-nucleotide microRNA (miRNA) – mediated DNA
methylation in rice have shown that methylation of both
strands of the DNA occurs, mostly within the miRNAhomologous region but also extending 30 base pairs
on either side of the miRNA site (Wu et al. 2010). Thus,
we imagine that the DNA strands, Pol V transcripts, and
siRNA-AGO4 complexes are all in close proximity to one
another, such that DRM2 can methylate either DNA
strand.
Multiple activities implicated in chromatin modification appear to be coordinated by Pol V and its transcripts.
The homolog of the yeast SPT5 (SUPPRESSOR OF TY
5), SPT5-LIKE (SPT5L, also known as KTF1) interacts
with Pol V and is important for RNA-directed DNA methylation (Bies-Etheve et al. 2009; Huang et al. 2009).
SPT5L/KTF1 interacts with AGO4, binds RNA, and is
thought to assist in the recruitment and/or stabilization of
AGO4 at Pol V– transcribed loci (He et al. 2009; Rowley
et al. 2011). The RNA-binding protein IDN2 (INVOLVED IN DE NOVO 2) (Ausin et al. 2009), also identified in other genetic screens as DMS10 or RDM12
(Zheng et al. 2010; Lorkovic et al. 2012), was recently
shown to dimerize, form a complex with paralogous
IDN2-like proteins (IDPs) (Xie et al. 2012; Zhang et al.
2012), and interact with Pol V transcripts and at least one
subunit of the SWI/SNF chromatin remodeling complex
(Zhu et al. 2012). The SWI/SNF complex is proposed to
help mobilize nucleosomes bearing repressive chromatin
Downloaded from symposium.cshlp.org on April 9, 2013 - Published by Cold Spring Harbor Laboratory Press
6
PIKAARD ET AL.
modifications to assemble more tightly packed heterochromatic structures (Zhu et al. 2012).
Among the activities responsible for RdDM-associated
repressive posttranslational histone modifications are
the Jumonji domain protein JMJ14, which demethylates
histone H3 lysine 4 (Lu et al. 2010; Searle et al. 2010),
thus removing a mark associated with active chromatin.
Likewise, HDA6 (HISTONE DEACETYLASE 6), a
broad specificity histone deacetylase involved in RdDM
at some loci (Murfett et al. 2001; Aufsatz et al. 2002;
Probst et al. 2004), can remove acetyl groups from multiple histone lysines (Earley et al. 2006). Deacetylation
of key positions, such as lysine 9 of histone H3 (H3K9),
is a prerequisite for their subsequent methylation.
Thus, HDA6 may facilitate the activities of SUVH2
and SUVH9 (SU[VAR]3-9 HOMOLOGOUS 2 and 9),
H3K9 methyltransferases with SRA domains that can
bind methylated cytosines in CG or CHH motifs, respectively (Johnson et al. 2008), thereby coupling histone
modification with cytosine methylation. A third H3K9
methyltransferase, SUVR2, has recently been found to
be important for RNA-directed DNA methylation (Stroud
et al. 2013). The SWI/SNF chromatin remodeling complex may catalyze nucleosome sliding to enable these
histone modifications, and their cross-talk with cytosine
methylation, while also facilitating changes in spacing
among nucleosomes that bear repressive modifications
(Zhu et al. 2012).
FUTURE PROSPECTS
Biochemical experiments to address the predictions and
speculations of the transcription fork model are sorely
needed. Understanding the basis for Pol IV and Pol V
recruitment to their sites of action is one priority, as is
the need to further explore the potential requirement for
RNA primers in Pol IV and Pol V transcription in vivo as
well as in vitro. Defining the full-length transcripts of Pol
IV and Pol V is needed to define their transcription units
and aid the search for potential regulatory elements near
the 50 and 30 ends of these transcription units. Experiments
to define the basis for Pol IV – RDR2 association are needed, as are experiments to determine whether RDR2 can
begin transcription on Pol IV – generated RNA templates
while Pol IV is still engaged in transcriptional elongation,
thus generating Okazaki-like RNA fragments. Alternatively, determining whether RDR2 requires the free
30 end of a released Pol IV transcript, or can use its own
transcripts to amplify dsRNA, is a priority. Defining activities of the DDR complex, including potential DNA
translocase/helicase activities and an ability to help Pol
V transcribe through dsDNA, should be enlightening.
Likewise, determining whether extended transcription
bubbles are generated at sites of Pol V transcription, possibly resulting from persistent hybrid formation between
Pol V transcripts and the template DNA strand, could
provide important mechanistic insights. Through further
development of in vitro assays, the speculations inherent
to the model can be tested and conjecture ultimately supplanted by evidence.
ACKNOWLEDGMENTS
Pikaard laboratory studies of RNA-directed DNA methylation are supported by National Institutes of Health
(NIH) grant GM077590 and C.S.P.’s support as an Investigator of the Howard Hughes Medical Institute and the
Gordon & Betty Moore Foundation. J.R.H. and T.B. are
supported by funds from the Howard Hughes Medical
Institute. R.C. is (and T.B. was) supported by a Ruth
L. Kirschstein National Research Award from the NIH.
Opinions are those of the investigators and do not necessarily reflect the views of our sponsors.
REFERENCES
Aalto AP, Poranen MM, Grimes JM, Stuart DI, Bamford DH.
2010. In vitro activities of the multifunctional RNA silencing
polymerase QDE-1 of Neurospora crassa. J Biol Chem 285:
29367 – 29374.
Aufsatz W, Mette MF, Van Der Winden J, Matzke M, Matzke
AJ. 2002. HDA6, a putative histone deacetylase needed to
enhance DNA methylation induced by double-stranded
RNA. EMBO J 21: 6832 – 6841.
Ausin I, Mockler TC, Chory J, Jacobsen SE. 2009. IDN1 and
IDN2 are required for de novo DNA methylation in Arabidopsis thaliana. Nat Struct Mol Biol 16: 1325 – 1327.
Bender J. 2012. RNA-directed DNA methylation: Getting a grip
on mechanism. Curr Biol 22: R400– R401.
Bies-Etheve N, Pontier D, Lahmy S, Picart C, Vega D, Cooke R,
Lagrange T. 2009. RNA-directed DNA methylation requires
an AGO4-interacting member of the SPT5 elongation factor
family. EMBO Rep 10: 649 – 654.
Blumenthal T, Carmichael GG. 1979. RNA replication: Function and structure of Qbeta-replicase. Annu Rev Biochem 48:
525 – 548.
Bochman ML, Schwacha A. 2009. The Mcm complex: Unwinding the mechanism of a replicative helicase. Microbiol Mol
Biol Rev 73: 652 –683.
Brodersen P, Voinnet O. 2006. The diversity of RNA silencing
pathways in plants. Trends Genet 22: 268 – 280.
Brosnan CA, Voinnet O. 2009. The long and the short of noncoding RNAs. Curr Opin Cell Biol 21: 416 – 425.
Cao X, Jacobsen SE. 2002. Role of the Arabidopsis DRM Methyltransferases in De Novo DNA Methylation and Gene Silencing. Curr Biol 12: 1138– 1144.
Clapier CR, Cairns BR. 2009. The biology of chromatin remodeling complexes. Annu Rev Biochem 78: 273 – 304.
Curaba J, Chen X. 2008. Biochemical Activities of Arabidopsis
RNA-dependent RNA Polymerase 6. J Biol Chem 283:
3059 – 3066.
Earley K, Lawrence RJ, Pontes O, Reuther R, Enciso AJ, Silva
M, Neves N, Gross M, Viegas W, Pikaard CS. 2006. Erasure
of histone acetylation by Arabidopsis HDA6 mediates largescale gene silencing in nucleolar dominance. Genes Dev 20:
1283 – 1293.
El-Shami M, Pontier D, Lahmy S, Braun L, Picart C, Vega D,
Hakimi MA, Jacobsen SE, Cooke R, Lagrange T. 2007. Reiterated WG/GW motifs form functionally and evolutionarily
conserved ARGONAUTE-binding platforms in RNAi-related
components. Genes Dev 21: 2539 – 2544.
Gao Z, Liu HL, Daxinger L, Pontes O, He X, Qian W, Lin H, Xie
M, Lorkovic ZJ, Zhang S, et al. 2010. An RNA polymerase IIand AGO4-associated protein acts in RNA-directed DNA
methylation. Nature 465: 106 –109.
Haag JR, Pikaard CS. 2011. Multisubunit RNA polymerases IV
and V: Purveyors of non-coding RNA for plant gene silencing. Nat Rev Mol Cell Biol 12: 483 – 492.
Haag JR, Ream TS, Marasco M, Nicora CD, Norbeck AD, PasaTolic L, Pikaard CS. 2012. In vitro transcription activities
of Pol IV, Pol V, and RDR2 reveal coupling of Pol IV and
Downloaded from symposium.cshlp.org on April 9, 2013 - Published by Cold Spring Harbor Laboratory Press
RNA-DIRECTED DNA METHYLATION
RDR2 for dsRNA synthesis in plant RNA silencing. Mol Cell
48: 811 – 818.
He XJ, Hsu YF, Zhu S, Wierzbicki AT, Pontes O, Pikaard CS, Liu
HL, Wang CS, Jin H, Zhu JK. 2009. An effector of RNA-directed DNA methylation in Arabidopsis is an ARGONAUTE
4- and RNA-binding protein. Cell 137: 498 – 508.
Herr AJ, Jensen MB, Dalmay T, Baulcombe DC. 2005. RNA
polymerase IV directs silencing of endogenous DNA. Science
308: 118 – 120.
Huang L, Jones AM, Searle I, Patel K, Vogler H, Hubner NC,
Baulcombe DC. 2009. An atypical RNA polymerase involved
in RNA silencing shares small subunits with RNA polymerase II. Nat Struct Mol Biol 16: 91– 93.
Ishizu H, Siomi H, Siomi MC. 2012. Biology of PIWI-interacting RNAs: New insights into biogenesis and function inside
and outside of germlines. Genes Dev 26: 2361 – 2373.
Johnson LM, Law JA, Khattar A, Henderson IR, Jacobsen SE.
2008. SRA-domain proteins required for DRM2-mediated de
novo DNA methylation. PLoS Genet 4: e1000280.
Kanno T, Mette MF, Kreil DP, Aufsatz W, Matzke M, Matzke
AJ. 2004. Involvement of putative SNF2 chromatin remodeling protein DRD1 in RNA-directed DNA methylation. Curr
Biol 14: 801 – 805.
Kanno T, Bucher E, Daxinger L, Huettel B, Bohmdorfer G,
Gregor W, Kreil DP, Matzke M, Matzke AJ. 2008. A structural-maintenance-of-chromosomes hinge domain-containing protein is required for RNA-directed DNA methylation.
Nat Genet 40: 670 – 675.
Kidmose RT, Vasiliev NN, Chetverin AB, Andersen GR, Knudsen CR. 2010. Structure of the Qbeta replicase, an RNAdependent RNA polymerase consisting of viral and host
proteins. Proc Natl Acad Sci 107: 10884 – 10889.
Law JA, Ausin I, Johnson LM, Vashisht AA, Zhu JK, Wohlschlegel JA, Jacobsen SE. 2010. A protein complex required
for polymerase V transcripts and RNA- directed DNA methylation in Arabidopsis. Curr Biol 20: 951– 956.
Law JA, Vashisht AA, Wohlschlegel JA, Jacobsen SE. 2011.
SHH1, a homeodomain protein required for DNA methylation,
as well as RDR2, RDM4, and chromatin remodeling factors,
associate with RNA polymerase IV. PLoS Genet 7: e1002195.
Lee HC, Aalto AP, Yang Q, Chang SS, Huang G, Fisher D, Cha J,
Poranen MM, Bamford DH, Liu Y. 2010. The DNA/RNAdependent RNA polymerase QDE-1 generates aberrant RNA
and dsRNA for RNAi in a process requiring replication protein
A and a DNA helicase. PLoS Biol 8. doi:pii: e1000496.10.1371/
journal.pbio.
Lister R, O’Malley RC, Tonti-Filippini J, Gregory BD, Berry
CC, Millar AH, Ecker JR. 2008. Highly integrated single-base
resolution maps of the epigenome in Arabidopsis. Cell 133:
523 – 536.
Liu J, Bai G, Zhang C, Chen W, Zhou J, Zhang S, Chen Q, Deng
X, He XJ, Zhu JK. 2011. An atypical component of RNAdirected DNA methylation machinery has both DNA methylation-dependent and -independent roles in locus-specific transcriptional gene silencing. Cell Res 21: 1691 – 1700.
Lorkovic ZJ, Naumann U, Matzke AJ, Matzke M. 2012. Involvement of a GHKL ATPase in RNA-directed DNA methylation
in Arabidopsis thaliana. Curr Biol 22: 933 – 938.
Lu F, Cui X, Zhang S, Liu C, Cao X. 2010. JMJ14 is an H3K4
demethylase regulating flowering time in Arabidopsis. Cell
Res 20: 387 –390.
Mallory A, Vaucheret H. 2010. Form, function, and regulation
of ARGONAUTE proteins. Plant Cell 22: 3879 – 3889.
Matzke M, Kanno T, Daxinger L, Huettel B, Matzke AJ. 2009.
RNA-mediated chromatin-based silencing in plants. Curr
Opin Cell Biol 21: 367 – 376.
Mazin AV, Mazina OM, Bugreev DV, Rossi MJ. 2010. Rad54,
the motor of homologous recombination. DNA Repair (Amst)
9: 286 – 302.
Moazed D. 2009. Small RNAs in transcriptional gene silencing
and genome defence. Nature 457: 413 – 420.
Moissiard G, Cokus SJ, Cary J, Feng S, Billi AC, Stroud H,
Husmann D, Zhan Y, Lajoie BR, McCord RP, et al. 2012.
7
MORC family ATPases required for heterochromatin condensation and gene silencing. Science 336: 1448– 1451.
Murfett J, Wang XJ, Hagen G, Guilfoyle TJ. 2001. Identification
of Arabidopsis histone deacetylase HDA6 mutants that affect
transgene expression. Plant Cell 13: 1047– 1061.
Onodera Y, Haag JR, Ream T, Nunes PC, Pontes O, Pikaard CS.
2005. Plant nuclear RNA polymerase IV mediates siRNA and
DNA methylation-dependent heterochromatin formation.
Cell 120: 613– 622.
Pelissier T, Wassenegger M. 2000. A DNA target of 30 bp
is sufficient for RNA-directed DNA methylation. RNA 6:
55– 65.
Pontes O, Li CF, Nunes PC, Haag J, Ream T, Vitins A, Jacobsen
SE, Pikaard CS. 2006. The Arabidopsis chromatin-modifying
nuclear siRNA pathway involves a nucleolar RNA processing
center. Cell 126: 79 – 92.
Probst AV, Fagard M, Proux F, Mourrain P, Boutet S, Earley K,
Lawrence RJ, Pikaard CS, Murfett J, Furner I, et al.
2004. Arabidopsis histone deacetylase HDA6 is required
for maintenance of transcriptional gene silencing and determines nuclear organization of rDNA repeats. Plant Cell 16:
1021– 1034.
Qi Y, He X, Wang XJ, Kohany O, Jurka J, Hannon GJ. 2006.
Distinct catalytic and non-catalytic roles of ARGONAUTE4
in RNA-directed DNA methylation. Nature 443: 1008– 1012.
Remus D, Diffley JF. 2009. Eukaryotic DNA replication control: Lock and load, then fire. Curr Opin Cell Biol 21: 771 –
777.
Rowley MJ, Avrutsky MI, Sifuentes CJ, Pereira L, Wierzbicki
AT. 2011. Independent chromatin binding of ARGONAUTE4
and SPT5L/KTF1 mediates transcriptional gene silencing.
PLoS Genet 7: e1002120.
Searle IR, Pontes O, Melnyk CW, Smith LM, Baulcombe DC.
2010. JMJ14, a JmjC domain protein, is required for RNA
silencing and cell-to-cell movement of an RNA silencing
signal in Arabidopsis. Genes Dev 24: 986 – 991.
Smith LM, Pontes O, Searle I, Yelina N, Yousafzai FK, Herr AJ,
Pikaard CS, Baulcombe DC. 2007. An SNF2 protein associated with nuclear RNA silencing and the spread of a silencing
signal between cells in Arabidopsis. Plant Cell 19: 1507–
1521.
Stroud H, Greenberg MV, Feng S, Bernatavichute YV, Jacobsen
SE. 2013. Comprehensive analysis of silencing mutants reveals complex regulation of the Arabidopsis methylome. Cell
152: 352 – 364.
Takeshita D, Tomita K. 2010. Assembly of Qfbetag viral RNA
polymerase with host translational elongation factors EF-Tu
and -Ts. Proc Natl Acad Sci 107: 15733– 15738.
Wassenegger M, Krczal G. 2006. Nomenclature and functions of
RNA-directed RNA polymerases. Trends Plant Sci 11: 142 –
151.
Wierzbicki AT. 2012. The role of long non-coding RNA in
transcriptional gene silencing. Curr Opin Plant Biol 15:
517– 522.
Wierzbicki AT, Haag JR, Pikaard CS. 2008. Noncoding transcription by RNA polymerase Pol IVb/Pol V mediates transcriptional silencing of overlapping and adjacent genes. Cell
135: 635 – 648.
Wierzbicki AT, Ream TS, Haag JR, Pikaard CS. 2009. RNA
polymerase V transcription guides ARGONAUTE4 to chromatin. Nat Genet 41: 630 – 634.
Wierzbicki AT, Cocklin R, Mayampurath A, Lister R, Rowley
MJ, Gregory BD, Ecker JR, Tang H, Pikaard CS. 2012. Spatial and functional relationships among Pol V-associated loci,
Pol IV-dependent siRNAs, and cytosine methylation in the
Arabidopsis epigenome. Genes Dev 26: 1825 – 1836.
Wu L, Zhou H, Zhang Q, Zhang J, Ni F, Liu C, Qi Y. 2010. DNA
methylation mediated by a microRNA pathway. Mol Cell 38:
465– 475.
Xie Z, Johansen LK, Gustafson AM, Kasschau KD, Lellis AD,
Zilberman D, Jacobsen SE, Carrington JC. 2004. Genetic and
functional diversification of small RNA pathways in plants.
PLoS Biol 2: 642 – 652.
Downloaded from symposium.cshlp.org on April 9, 2013 - Published by Cold Spring Harbor Laboratory Press
8
PIKAARD ET AL.
Xie M, Ren G, Zhang C, Yu B. 2012. The DNA- and RNAbinding protein FACTOR of DNA METHYLATION 1 requires XH domain-mediated complex formation for its function in RNA-directed DNA methylation. Plant J 72: 491 –
500.
Zaratiegui M, Irvine DV, Martienssen RA. 2007. Noncoding
RNAs and gene silencing. Cell 128: 763 – 776.
Zhang H, Zhu JK. 2011. RNA-directed DNA methylation. Curr
Opin Plant Biol 14: 142 – 147.
Zhang CJ, Ning YQ, Zhang SW, Chen Q, Shao CR, Guo YW,
Zhou JX, Li L, Chen S, He XJ. 2012. IDN2 and its paralogs
form a complex required for RNA-directed DNA methylation.
PLoS Genet 8: e1002693.
Zheng L, Shen B. 2011. Okazaki fragment maturation: Nucleases take centre stage. J Mol Cell Biol 3: 23 – 30.
Zheng B, Wang Z, Li S, Yu B, Liu JY, Chen X. 2009. Intergenic
transcription by RNA polymerase II coordinates Pol IV and
Pol V in siRNA-directed transcriptional gene silencing in
Arabidopsis. Genes Dev 23: 2850– 2860.
Zheng Z, Xing Y, He XJ, Li W, Hu Y, Yadav SK, Oh J, Zhu JK.
2010. An SGS3-like protein functions in RNA-directed DNA
methylation and transcriptional gene silencing in Arabidopsis. Plant J 62: 92 –99.
Zhong X, Hale CJ, Law JA, Johnson LM, Feng S, Tu A, Jacobsen SE. 2012. DDR complex facilitates global association of
RNA polymerase V to promoters and evolutionarily young
transposons. Nat Struct Mol Biol 19: 870 – 875.
Zhu Y, Rowley MJ, Bohmdorfer G, Wierzbicki AT. 2012. A SWI/
SNF chromatin-remodeling complex acts in noncoding RNAmediated transcriptional silencing. Mol Cell 49: 298 – 309.
Zilberman D, Cao X, Johansen LK, Xie Z, Carrington JC, Jacobsen SE. 2004. Role of Arabidopsis ARGONAUTE4 in RNAdirected DNA methylation triggered by inverted repeats. Curr
Biol 14: 1214 – 1220.