Riverside, CA 92521, USA SUMMARY

Transcription

1 The Plant Journal (2010) 64, doi: /j X x Transcriptional silencing induced by Arabidopsis T-DNA mutants is associated with 35S promoter sirnas and requires genes involved in sirna-mediated chromatin silencing Sizolwenkosi Mlotshwa 1,*, Gail J. Pruss 1, Zhihuan Gao 1,, Nomathamsanqa L. Mgutshini 1, Junjie Li 2,, Xuemei Chen 2, Lewis H. Bowman 1 and Vicki Vance 1,* 1 Department of Biological Sciences, University of South Carolina, Columbia, SC 29208, USA, and 2 Department of Botany and Plant Sciences and Institute of Integrative Genome Biology, University of California Riverside, Riverside, CA 92521, USA Received 15 June 2010; revised 27 August 2010; accepted 7 September 2010; published online 8 October * For correspondence (fax ; sizo@biol.sc.edu and vance@biol.sc.edu). Present address: Center for Plant Cell Biology, Department of Plant Pathology and Microbiology, University of California, Riverside, CA 92521, USA. Present address: Life Technologies Inc., 2130 Woodward Street Austin, TX 78744, USA. SUMMARY The utility of many T-DNA insertion mutant lines of Arabidopsis is compromised by their propensity to trigger transcriptional silencing of transgenes expressed from the CaMV 35S promoter. To try to circumvent this problem, we characterized the genetic requirements for maintenance of 35S promoter homology-dependent transcriptional gene silencing induced by the dcl3-1 (SALK_005512) T-DNA insertion mutant line. Surprisingly, even though DCL3 and RDR2 are known components of the sirna-dependent transcriptional gene silencing pathway, transcriptional gene silencing of a 35S promoter-driven GUS hairpin transgene did occur in plants homozygous for the dcl3-1 T-DNA insertion and was unaffected by loss of function of RDR2. However, the transcriptional gene silencing was alleviated in dcl2 dcl3 dcl4 triple mutant plants and also by mutations in AGO4, NRPD2, HEN1 and MOM1. Thus, some T-DNA insertion mutant lines induce 35S promoter homologydependent transcriptional silencing that requires neither DCL3 nor RDR2, but involves other genes known to function in sirna-dependent transcriptional silencing. Consistent with these results, we detected 35S promoter sirnas in dcl3-1 SALK line plants, suggesting that the 35S promoter homology-dependent silencing induced by some T-DNA insertion mutant lines is sirna-mediated. Keywords: transcriptional silencing, DCL3, 35S promoter, sirna, T-DNA insertion mutant, SALK line. INTRODUCTION Arabidopsis T-DNA insertion mutants are an invaluable resource for studies of gene function. However, unintended homology-dependent transcriptional gene silencing caused by the T-DNA insertions limits the utility of many of these mutants (Daxinger et al., 2008). Potential triggers of this transcriptional silencing are promoter sequences that are common to both the T-DNA insertions and the unlinked transgenes that are silenced by the T-DNA. The T-DNA insertions in SALK, GABI and FLAG mutant lines all carry a copy of the CaMV 35S promoter, and a study of SALK and GABI lines found that a high proportion of those lines induced transcriptional silencing of transgenes expressed from the 35S promoter (Daxinger et al., 2008). Transcriptional gene silencing (TGS) and post-transcriptional gene silencing (PTGS) were originally thought to be mechanistically unrelated. However, just as dsrna that contains sequences from the coding region of a gene triggers post-transcriptional silencing, dsrna that contains promoter sequences triggers transcriptional silencing (Matzke et al., 2004, 2009; Matzke and Birchler, 2005). The initial steps of these two processes are very similar, and members of the gene families involved in post-transcriptional silencing play corresponding roles in transcriptional silencing. The dsrna involved in both processes can arise directly, for example by transcription of an inverted repeat, or indirectly, by copying of a single stranded RNA template by RNA-dependent RNA polymerase (RDR). Silencing in both TGS and PTGS is directed by sirnas produced by cleavage of the dsrna trigger by RNAse III-like enzymes [termed Dicer-like (DCL) in Arabidopsis], and a single strand of sirna bound to a member of the Argonaute (AGO) protein family forms the core of the sequence-specific 699 The Plant Journal ª 2010 Blackwell Publishing Ltd

2 700 Sizolwenkosi Mlotshwa et al. effector complex that carries out silencing. RDR2, DCL3 and AGO4 are involved in transcriptional silencing, which blocks transcription via DNA methylation and histone modification. sirna-directed silencing appears to be a likely explanation for the 35S promoter homology-dependent transcriptional silencing seen in SALK and GABI lines. Surprisingly, however, dcl3 T-DNA insertion mutations (dcl3-1, SALK_005512; dcl3-4, GABI_327D02) transcriptionally silence 35S promoter-driven transgenes even when plants are homozygous for the dcl3 mutations (Daxinger et al., 2008). This observation is paradoxical, given the importance of DCL3 in sirna-mediated transcriptional silencing (Daxinger et al., 2009; Matzke et al., 2009), and raises the possibility that sirna-directed pathways are not involved in 35S promoter homology-dependent TGS induced by these T-DNA insertions. Indeed, only about one-third of the DNA methylation in Arabidopsis is associated with sirnas (Lister et al., 2008), suggesting that other mechanisms might be responsible for some cases of TGS. Therefore, we examined the genetic requirements for 35S promoter homologydependent transcriptional silencing induced by the dcl3-1 SALK line, in order to determine the mechanism responsible and to find ways to eliminate this undesirable side-effect of the T-DNA insertions in SALK, GABI and FLAG mutant lines. (a) (b) RESULTS AND DISCUSSION The dcl3-1 line (SALK_005512) is one of several T-DNA insertion lines that have been reported to induce TGS of 35S promoter-driven transgenes when crossed to transgenic lines. Induction of 35S promoter homology-dependent TGS occurs in the F 1 generation, which is hemizygous for the dcl3-1 T-DNA insertion, and is maintained in F 2 generation plants homozygous for dcl3-1 (Daxinger et al., 2008). We encountered this problem with the dcl3-1 SALK line when examining the roles of DCL genes in post-transcriptional silencing (Mlotshwa et al., 2008). These experiments used two kinds of post-transcriptionally silenced transgene. Line L1 carries a 35S promoter-driven GUS sense transgene that was designed to express high levels of GUS (Figure 1a diagram) but is silenced instead (Elmayan et al., 1998). Line (referred to as 306 below) carries a 35S promoterdriven GUS hairpin transgene (Beclin et al., 2002) that produces a self-complementary transcript (Figure 1b). Posttranscriptional silencing of the GUS transgene in each of these lines is indicated by the low levels of GUS mrna and the accumulation of GUS sirnas (Figure 1a, lanes 5 and 6, and Figure 1b, lanes 3 and 4). The main GUS mrna species that accumulates in line 306 corresponds to the loop portion of the hairpin (Mlotshwa et al., 2008), as has been shown for other 35S promoter-driven hairpin transcripts (Wang et al., 2008). We crossed lines L1 and 306 with the dcl3-1 SALK line, allowed the F 1 progeny to self, and analyzed F 2 plants that carried at least one copy of the GUS transgene and were homozygous for the dcl3-1 T-DNA insertion. These plants (c) Figure 1. Transcriptional silencing of 35S promoter-driven transgenes is triggered by the dcl3-1 T-DNA insertion. Accumulation of GUS mrna and antisense strand GUS sirnas in wild-type (wt) and homozygous dcl3-1 plants carrying the L1 or 306 GUS transgene was determined using RNA gel-blot analysis. Aerial tissue from three or four plants was pooled for RNA isolation, and each RNA sample is from separately pooled plants. The positions of nt RNA size markers are indicated. Ethidium bromide (EtBr)-stained rrna and the major RNA species in lowmolecular-weight RNA are shown as loading controls. (a) The diagram shows the coding region of the silenced GUS sense transgene in line L1, plus the location of the sense and antisense probes used. The position and length in base pairs (bp) of the stem of the 306 GUS hairpin construct are also shown. dcl3-1 RNA samples are from plants hemizygous (lanes 1 and 2) or homozygous (lanes 3 and 4) for the L1 transgene; the wildtype RNA samples (lanes 5 and 6) are from the homozygous L1 line. (b) The diagram shows the coding region of the line 306 hairpin transgene to illustrate the double-stranded configuration of the transcript. The locations of the hybridization probes are indicated. Probe 1 is the same as in (a). The wildtype RNA samples are from the homozygous 306 line. (c) Relative levels of transcription of the GUS transgene in wild-type (lane 1) and dcl3-1 mutant (lane 2) homozygous L1 GUS plants were determined in isolated nuclei. Nuclear transcripts were labeled with 32 P by run-off transcription, and then hybridized to slot blots loaded with plasmid DNA containing GUS, actin and puc19 empty vector sequences.

3 SALK line-induced transcriptional silencing 701 showed a complete absence of GUS mrna and sirna (Figure 1a, lanes 1 4, and Figure 1b, lanes 1 and 2). Lack of accumulation of both sirna and mrna is indicative of transcriptional silencing, and, in the case of the dcl3-1/l1 plants, we used nuclear run-off transcription assays to further confirm that transcriptional silencing had occurred. Labeled nuclei from L1 GUS wild-type transgenic plants produced a strong GUS hybridization signal (Figure 1c, lane 1), whereas nuclei from homozygous dcl3-1 L1 GUS plants did not (Figure 1c, lane 2), showing that the L1 GUS transgene is transcribed in the wild-type background, but not in plants that have the dcl3-1 T-DNA insertion. Because dcl3-1 is a null allele (Xie et al., 2004; and data not shown), maintenance of 35S promoter homology-dependent TGS in plants homozygous for the dcl3-1 T-DNA insertion suggests that either the process does not involve sirnas or that other DCL proteins can substitute for DCL3. To distinguish between these possibilities, we crossed the transcriptionally silenced dcl3-1/306 plants described above with dcl2-1 (SALK_064627) and dcl4-2 (Yoshikawa et al., 2005) mutants, and analyzed F plants that were homozygous for all three dcl mutations. We used line 306 in this and subsequent experiments instead of line L1 because reappearance of the GUS loop RNA provides a convenient marker for release of TGS. Although the dcl2-1 mutation, like dcl3-1, is a T-DNA insertion that can trigger TGS of 35S promoter-driven transgenes (Daxinger et al., 2008; Mlotshwa et al., 2008), the problem can be avoided in the case of dcl2-1 by using plants hemizygous for the transgene in question (Mlotshwa et al., 2008). The dcl2-1 dcl3-1 dcl4-2 mutant plants accumulated a low level of GUS loop RNA, indicating that maintenance of TGS is less effective in this background than in the dcl3-1 T-DNA insertion mutant background (Figure 2a, compare lanes 2-4 with lane 1). This result suggests that transcriptional silencing induced by the dcl3-1 T-DNA insertion mutation involves sirnas, and that sirnas produced by DCL2 or DCL4 or both can substitute for those produced by DCL3. Substitution of DCL2 and/or DCL4 function for that of DCL3 in TGS is consistent with the observation that AGO4, the Argonaute protein required for TGS, can bind 21 and 22 nt sirnas (Qi et al., 2006; Mi et al., 2008). Partial redundancy in DCL2, DCL3 and DCL4 function has previously been reported for small RNA-directed methylation of endogenous repeat loci (Henderson et al., 2006), but not of transgenes (Wang et al., 2008; Daxinger et al., 2009). To test whether 35S promoter sirnas accumulate in dcl3-1 plants, we performed RNA gel-blot analysis of the lowmolecular-weight fraction of RNA isolated from three plant lines containing the dcl3-1 T-DNA insertion. 35S promoterspecific sirnas were detected in all three lines (Figure 2b): line 306 plants that were hemizygous for dcl3-1 accumulated 24 nt 35S promoter sirnas (Figure 2b, lanes 6 and 7), consistent with the expected involvement of 24 nt sirnas in transcriptional silencing. In contrast, homozygous dcl3-1 plants with or without the 306 locus accumulated 21 and 22 nt 35S promoter sirnas (Figure 2b, lanes 4, 5, 8 and 9), consistent with sirna production by the other DCL proteins. No 35S promoter sirnas were detectable in wild-type nontransgenic or line 306 controls (Figure 2b, lanes 1 3). Together, our results suggest that the sirna-directed transcriptional silencing pathway is involved in maintenance of 35S promoter homology-dependent TGS induced by the dcl3-1 SALK line. To obtain further evidence concerning involvement of the sirna-directed TGS pathway in SALK line-induced silencing of the 35S promoter, we examined the effect of other mutations in this pathway on accumulation of 306 GUS mrna in the dcl3-1/306 background. In addition to DCL3, RDR2 and AGO4, other genes in the pathway include Hua enhancer 1 (HEN1), which is required for sirna stability (Li et al., 2005), and nuclear RNA polymerase D2 (NRPD2). NRPD2 encodes the second largest subunit of two RNA polymerases, polymerase IV and polymerase V, which play roles in sirna biogenesis and RNA-directed DNA methylation, respectively (Matzke et al., 2009). We also examined the effect of mutating the Morpheus molecule 1 (MOM1) gene. MOM1 mutations release transcriptional silencing without altering DNA methylation (Amedeo et al., 2000), and MOM1 was originally thought to function in sirna-independent silencing (Habu et al., 2006). Recent studies, however, have linked MOM1 and the sirna-mediated TGS pathway (Numa et al., 2010; Yokthongwattana et al., 2010). To test for release of silencing, we analyzed F 2 progeny from crosses between our transcriptionally silenced dcl3-1/ 306 line, which is homozygous for dcl3-1, and lines containing known mutations in the above genes. The rdr2 mutant used is itself a SALK line (rdr2-2; SALK_ ). Homozygous rdr2 mutant F 2 progeny of a control cross between the rdr2-2 SALK line and line 306 induced transcriptional silencing of the GUS hairpin transgene (Figure 2c, lanes 6 and 7), showing that 35S homology-dependent TGS does not require RDR2, at least in the case of the rdr2-2 T-DNA insertion. TGS of the line 306 GUS transgene was also maintained in the homozygous rdr2-2 dcl3-1 background (Figure 2c, lanes 4 and 5), indicating that maintenance of TGS by at least one if not both of these T-DNA insertions requires neither RDR2 nor DCL3. Because RDR2 is not required for transcriptional silencing in the case of hairpin TGS constructs (Daxinger et al., 2009), our results suggest that the rdr2-2 T-DNA insertion and possibly also the dcl3-1 T-DNA insertion produce 35S promoter dsrna directly. Maintenance of 35S promoter homology-dependent TGS induced by the dcl3-1 T-DNA insertion was impaired but not eliminated by the other RNA silencing pathway mutations we examined and by mutation of MOM1. F 2 progeny that carry the line 306 GUS transgene and are homozygous for both dcl3-1 and for mutations in nrpd2 (Figure 2d, lanes

4 702 Sizolwenkosi Mlotshwa et al. (a) (e) (b) (f) (c) (g) (d) (h) Figure 2. Silencing pathway mutations impair maintenance of the 35S promoter homology-dependent transcriptional silencing triggered by the dcl3-1 T-DNA insertion. Accumulation of GUS mrna (a, c h) and antisense strand 35S promoter sirnas (b) in wild-type (wt) and mutant plants carrying the 306 GUS locus was determined using RNA gel-blot analysis. Plants were homozygous for the mutations shown except for samples from dcl3-1 hemizygous plants (c, lane 1; b, lanes 6 and 7). Ethidium bromide (EtBr)-stained rrna (a, c h) and the major RNA species in low-molecular-weight RNA (b) are shown as loading controls. All samples in (a) and (c h) are from plants that contain the 306 GUS transgene. The wt controls (last lanes in a, c h) are from the homozygous line 306 and were isolated from tissue pooled from three or four plants. Except as noted, other samples in (a, c h) are from individual F 2 progeny of crosses between thedcl3-1/306 line and the following lines: (a) dcl2-1 and dcl4-2, (c) rdr2-2 [the rdr2/306 controls (lanes 6 and 7) are from a direct cross between rdr2-2 and 306], (d) nrpd2a-2 [the nrpd2a-2/306 control (lane 6) is from pooled plants], (e) hen1-1 [the hen1-1/306 controls (lanes 9 and 10) are from a direct cross between hen1-1 and 306], (f) ago4-1 [the ago4-1/306 control (lane 6) is from pooled plants], (g) mom1-2 [the mom1-2/306 control (lane 6) is from pooled plants], and (h) Ler. (b) The Columbia (Col) control (lane 1) is from pooled, nontransgenic plants, and the line 306 controls (lanes 2 and 3) are from pooled homozygous line 306 plants. The remaining samples are from individual plants and are from the homozygous dcl3-1 SALK line (lanes 4 and 5) and from F 2 progeny of a cross between dcl3-1 and 306 (lanes 6 9). The positions of nt RNA size markers are indicated. 2 5), hen1 (Figure 2e, lanes 3 8), ago4 (Figure 2f, lanes 2 5) or mom1 (Figure 2g, lanes 2 5) all accumulated GUS loop RNA. Similar results were obtained when these mutant progeny were hemizygous for dcl3-1 (data not shown). Because the line 306 GUS transgene in hemizygous dcl3-1 F 2 progeny shows no release of TGS in the absence of other mutations (Figure 2c, lane 1), these results indicate that the impaired maintenance of TGS is due to the mutations in the other genes and does not require concurrent absence of DCL3 function. Although the nrpd2 mutant used in these crosses is a SALK line (SALK_046208), it does not induce TGS of the line 306 GUS transgene (Figure 2d, lane 6). The hen1 and ago4 mutants are in the Landsberg erecta (Ler) background, rather than Columbia, which was the background for all the other lines we used. Therefore, we crossed dcl3-1/306 with Ler as a control to check for the possibility that the alleviation of TGS observed in the hen1/dcl3/306 and ago4/dcl3/306 plants might be due to ecotype effects. TGS was fully maintained in the homozygous dcl3-1 F 2 progeny of this cross (Figure 2h, lanes 1 4), indicating that ecotype differences are not a concern. Thus, NRPD2, AGO4, HEN1 and MOM1 all appear to be required for complete maintenance of the 35S promoter homology-dependent TGS induced by the dcl3-1 T-DNA insertion. Our results for hen1-1 also demonstrate a role for HEN1 in inverted repeat PTGS in a system involving a 35S promoterdriven hairpin transgene. Increased levels of GUS loop and full-length hairpin transcript accumulate in hen1-1/306 compared to line 306 plants (Figure 2e, compare lanes 9 and 10 with lane 11), and the level of accumulation of GUS transcript in hen1-1/306 plants was similar to that seen when post-transcriptional silencing of the line 306 GUS

5 SALK line-induced transcriptional silencing 703 transgene is blocked by the HC-Pro viral suppressor of silencing (Figure 2e, compare lanes 9 and 10 with lane 1). This result contrasts with those of earlier studies using 35S promoter-driven hairpin transgenes to determine whether HEN1 is required for inverted repeat PTGS (Boutet et al., 2003; Li et al., 2005), but is in agreement with a more recent study that used a hairpin transgene driven by a moderately expressed phloem-limited promoter (Dunoyer et al., 2007). Together, our results support the involvement of sirnainduced transcriptional silencing in maintenance of the 35S promoter homology-dependent TGS induced by the dcl3-1 T-DNA insertion (Figure 3). It is not known why some SALK lines, such as dcl3-1, are potent inducers of 35S promoter homology-dependent transcriptional silencing while others are not. The propensity of individual T-DNA insertion lines to trigger 35S promoter homology-dependent transcriptional silencing is probably due to complex integration patterns of the T-DNA, at least some of which promote production of dsrna from the 35S promoter sequence on the T-DNA, and consequently give rise to 35S promoter sirnas. Although the dcl3-1 T-DNA insertion might not be representative of all SALK, GABI and FLAG lines, we expect that many other such lines also produce 35S promoter sirnas and that mutations in the sirna-dependent TGS pathway impair maintenance of TGS in those cases as they do for the dcl3-1 line. EXPERIMENTAL PROCEDURES Arabidopsis lines and genotyping All transgenic and mutant Arabidopsis lines used have been described previously: L1 (Elmayan et al., 1998), (Beclin et al., 2002), P1/HC-Pro (Mlotshwa et al., 2005), dcl2-1 (SALK_064627; Xie et al., 2004), dcl3-1 (SALK_005512; Xie et al., 2004), dcl4-2 (Yoshikawa et al., 2005), ago4-1 (Zilberman et al., 2003), nrpd2a-2 (SALK_046208; Onodera et al., 2005), mom1-2 (SAIL_610_G01; Habu et al., 2006), hen1-1 (Park et al., 2002) and rdr2-2 (SALK_ ; Herr et al., 2005). All lines are in the Columbia (Col-0) ecotype, except for ago4-1 and hen1-1, which are in Landsberg erecta (Ler). Genotyping methods for homozygous dcl2-1, dcl4-2, ago4-1, nrpd2a-2 and rdr2-2 plants have been described previously (Zilberman et al., 2003; Herr et al., 2005; Onodera et al., 2005; Yoshikawa et al., 2005; Mlotshwa et al., 2008). For homozygous dcl3-1 PCR genotyping the T-DNA primer LBa1 (5 -TGGTTCACGTAGTGGGC- CATCG-3 ) was used with primers DCL3p1 (5 -CTGAATA TGGATAATAAGTTTGAGACATATC-3 ) and DCL3p2 (5 -GGACT- CAATGCAATATAGAGCTTTG-3 ). Primers DCL3p1 and DCL3p2 amplify a 400 bp product from the wild-type locus, and primers DCL3p1 and LBa1 amplify a 600 bp product from the dcl3-1 locus. For homozygous mom1-2 genotyping, the T-DNA primer LB3 (5 -TAGCATCTGAATTTCATAACCAATCTCGATACAC-3 ) was used with primers MOM1-F (5 -GCAACTGTAGCACATGCATCCAGCTC-3 ) and MOM1-R (5 -GATTGTCCACCACCTGCAGATGCAGG-3 ). Primers MOM1-F and MOM1-R amplify a 600 bp product from the wild-type locus, and primers MOM1-R and LB3 amplify a 400 bp product from the mom1-2 locus. Homozygous hen1-1 plants containing the line 306 GUS transgene were identified by their dwarfed phenotype. Genotyping for line L1 and 306 GUS transgenes was performed as described previously (Mlotshwa et al., 2008). RNA isolation, gel-blot analysis and run-off transcription Figure 3. Model for 35S promoter homology-dependent TGS induced by the dcl3-1 T-DNA insertion. The results of the present study are summarized in the context of the sirnadependent transcriptional silencing pathway. RDR2 was not required for the dcl3-1 SALK line-induced TGS, suggesting that a complex insertion of the T-DNA in this line promotes direct production of 35S promoter (35S-P) dsrna, either due to overlapping transcription or transcription of an inverted repeat. Detection of 24 nt 35S promoter sirnas in dcl3-1 hemizygous plants and 21 or 22 nt 35S promoter sirnas in dcl3-1 homozygous plants, plus partial loss of TGS in the dcl2 dcl3 dcl4 triple mutant, suggest that DCL3 is probably the enzyme responsible for sirna production in initiation of the TGS, but that DCL2, DCL4 and possibly DCL1 maintain it in the absence of DCL3. Partial loss of the TGS in the hen1 mutant suggests that maintenance involves HEN1 stabilization of the 35S promoter sirnas. Impaired maintenance in the ago4 mutant suggests that sirna-directed AGO4 methylation of 35S promoter DNA is also involved in maintenance. However, because NRPD2 is a subunit of both polymerases IV and V (POLIV and POLV), the impaired maintenance in the nrpd2 mutant does not indicate whether the POLV role in methylation or the POLIV role in amplification of sirna production or both are involved. Lastly, impaired maintenance in the mom1 mutant implicates MOM1-directed histone (H3K9) methylation in maintenance or reinforcement of the TGS. RNA was isolated from the F 2 progeny of all crosses. rdr2/306, ago4/ 306, nrpd2/306, hen1/306 and mom1/306 controls were all obtained from a cross between the mutant line and the dcl3-1/306 line and a direct cross between the mutant and line 306. The level of accumulation of line 306 GUS mrna was about the same for each of these controls, independent of which type of cross the plants came from (data not shown). Total RNA isolation, gel-blot analysis and nuclear run-off transcription were performed as described previously (Mlotshwa et al., 2008). [a- 32 P]UTP-labeled RNA probes were used for the gel-blot analysis, and GUS antisense or sense polarity probes to detect mrnas and antisense sirnas, respectively, were prepared using an Ambion MAXIscript in vitro transcription kit (Ambion, as described previously (Mlotshwa et al., 2008). To detect 35S promoter sirnas, we prepared probes encompassing almost the entire 35S promoter sequence contained in the prok2 vector that was used to generate SALK lines. This strategy was necessary because we could not predict what part of the 35S promoter the sirnas would come from. To generate the probes, two 35S promoter DNA fragments corresponding to

6 704 Sizolwenkosi Mlotshwa et al. nucleotides and nucleotides in the prok2 vector were PCR-amplified, incorporating the minimal sequence TAATACGACTCACTATAGGG of the T7 promoter at their 5 ends. [a- 32 P]UTP-labeled RNA probes were separately transcribed from each template and mixed in order to probe for 35S promoter sirnas. In all cases, probes were hybridized to mrna blots at 68 C in Ambion ULTRAhyb buffer, or to sirna blots at 42 C in Ambion ULTRAhyb-oligo buffer. ACKNOWLEDGEMENTS We thank Scott Poethig (Department of Biology, University of Pennsylvania) for the dcl4-2 mutant and Herve Vaucheret (Institut Jean-Pierre Bourgin, INRA) for the L1 and lines. This work was supported by National Institutes of Health grant number GM , National Science Foundation grant numbers and to V.V. and L.H.B., and National Science Foundation grant number MCB to X.C. Note Added in Proof We have detected 24 nt 35S promoter sirnas in two additional T-DNA insertion lines (drb4-1, SALK_000736; dcl4-2, GABI_160G05) that were shown by Daxinger et al. (2008) to induce 35S promoter homology-dependent TGS. REFERENCES Amedeo, P., Habu, Y., Afsar, K., Scheid, O.M. and Paszkowski, J. (2000) Disruption of the plant gene MOM releases transcriptional silencing of methylated genes. Nature, 405, Beclin, C., Boutet, S., Waterhouse, P. and Vaucheret, H. (2002) A branched pathway for transgene-induced RNA silencing in plants. Curr. Biol. 12, Boutet, S., Vazquez, F., Liu, J., Beclin, C., Fagard, M., Gratias, A., Morel, J.B., Crete, P., Chen, X. and Vaucheret, H. (2003) Arabidopsis HEN1: a genetic link between endogenous mirna controlling development and sirna controlling transgene silencing and virus resistance. Curr. Biol. 13, Daxinger, L., Hunter, B., Sheikh, M., Jauvion, V., Gasciolli, V., Vaucheret, H., Matzke, M. and Furner, I. (2008) Unexpected silencing effects from T-DNA tags in Arabidopsis. Trends Plant Sci. 13, 4 6. Daxinger, L., Kanno, T., Bucher, E., van der Winden, J., Naumann, U., Matzke, A.J. and Matzke, M. (2009) A stepwise pathway for biogenesis of 24 nt secondary sirnas and spreading of DNA methylation. EMBO J. 28, Dunoyer, P., Himber, C., Ruiz-Ferrer, V., Alioua, A. and Voinnet, O. (2007) Intra- and intercellular RNA interference in Arabidopsis thaliana requires components of the microrna and heterochromatic silencing pathways. Nat. Genet. 39, Elmayan, T., Balzergue, S., Beon, F. et al. (1998) Arabidopsis mutants impaired in cosuppression. Plant Cell, 10, Habu, Y., Mathieu, O., Tariq, M., Probst, A.V., Smathajitt, C., Zhu, T. and Paszkowski, J. (2006) Epigenetic regulation of transcription in intermediate heterochromatin. EMBO Rep. 7, Henderson, I.R., Zhang, X., Lu, C., Johnson, L., Meyers, B.C., Green, P.J. and Jacobsen, S.E. (2006) Dissecting Arabidopsis thaliana DICER function in small RNA processing, gene silencing and DNA methylation patterning. Nat. Genet. 38, Herr, A.J., Jensen, M.B., Dalmay, T. and Baulcombe, D.C. (2005) RNA polymerase IV directs silencing of endogenous DNA. Science, 308, Li, J., Yang, Z., Yu, B., Liu, J. and Chen, X. (2005) Methylation protects mirnas and sirnas from a 3 -end uridylation activity in Arabidopsis. Curr. Biol. 15, Lister, R., O Malley, R.C., Tonti-Filippini, J., Gregory, B.D., Berry, C.C., Millar, A.H. and Ecker, J.R. (2008) Highly integrated single-base resolution maps of the epigenome in Arabidopsis. Cell, 133, Matzke, M.A. and Birchler, J.A. (2005) RNAi-mediated pathways in the nucleus. Nat. Rev. Genet. 6, Matzke, M., Aufsatz, W., Kanno, T., Daxinger, L., Papp, I., Mette, M.F. and Matzke, A.J. (2004) Genetic analysis of RNA-mediated transcriptional gene silencing. Biochim. Biophys. Acta, 1677, Matzke, M., Kanno, T., Daxinger, L., Huettel, B. and Matzke, A.J. (2009) RNAmediated chromatin-based silencing in plants. Curr. Opin. Cell Biol. 21, Mi, S., Cai, T., Hu, Y. et al. (2008) Sorting of small RNAs into Arabidopsis argonaute complexes is directed by the 5 terminal nucleotide. Cell, 133, Mlotshwa, S., Schauer, S.E., Smith, T.H., Mallory, A.C., Herr, J.M. Jr, Roth, B., Merchant, D.S., Ray, A., Bowman, L.H. and Vance, V.B. (2005) Ectopic DICER-LIKE1 expression in P1/HC-Pro Arabidopsis rescues phenotypic anomalies but not defects in microrna and silencing pathways. Plant Cell, 17, Mlotshwa, S., Pruss, G.J., Peragine, A., Endres, M.W., Li, J., Chen, X., Poethig, R.S., Bowman, L.H. and Vance, V. (2008) DICER-LIKE2 plays a primary role in transitive silencing of transgenes in Arabidopsis. PLoS ONE, 3, e1755. Numa, H., Kim, J.M., Matsui, A. et al. (2010) Transduction of RNA-directed DNA methylation signals to repressive histone marks in Arabidopsis thaliana. EMBO J. 29, Onodera, Y., Haag, J.R., Ream, T., Nunes, P.C., Pontes, O. and Pikaard, C.S. (2005) Plant nuclear RNA polymerase IV mediates sirna and DNA methylation-dependent heterochromatin formation. Cell, 120, Park, W., Li, J., Song, R., Messing, J. and Chen, X. (2002) CARPEL FACTORY, a Dicer homolog, and HEN1, a novel protein, act in microrna metabolism in Arabidopsis thaliana. Curr. Biol. 12, Qi, Y., He, X., Wang, X.J., Kohany, O., Jurka, J. and Hannon, G.J. (2006) Distinct catalytic and non-catalytic roles of ARGONAUTE4 in RNA-directed DNA methylation. Nature, 443, Wang, M.B., Helliwell, C.A., Wu, L.M., Waterhouse, P.M., Peacock, W.J. and Dennis, E.S. (2008) Hairpin RNAs derived from RNA polymerase II and polymerase III promoter-directed transgenes are processed differently in plants. RNA, 14, Xie, Z., Johansen, L.K., Gustafson, A.M., Kasschau, K.D., Lellis, A.D., Zilberman, D., Jacobsen, S.E. and Carrington, J.C. (2004) Genetic and functional diversification of small RNA pathways in plants. PLoS Biol. 2, e104. Yokthongwattana, C., Bucher, E., Caikovski, M., Vaillant, I., Nicolet, J., Mittelsten Scheid, O. and Paszkowski, J. (2010) MOM1 and Pol-IV/V interactions regulate the intensity and specificity of transcriptional gene silencing. EMBO J. 29, Yoshikawa, M., Peragine, A., Park, M.Y. and Poethig, R.S. (2005) A pathway for the biogenesis of trans-acting sirnas in Arabidopsis. Genes Dev. 19, Zilberman, D., Cao, X. and Jacobsen, S.E. (2003) ARGONAUTE4 control of locus-specific sirna accumulation and DNA and histone methylation. Science, 299,