Transcriptional Control of Gene Expression by MicroRNAs

Transcriptional Control of Gene Expression by MicroRNAs Basel Khraiwesh,1,4,5,6 M. Asif Arif,1,5 Gotelinde I. Seumel,1,3 Stephan Ossowski,2 Detlef Weigel,2 Ralf Reski,1,3,4,* and Wolfgang Frank1,3,* 1Plant

Biotechnology, Faculty of Biology, University of Freiburg, Schaenzlestrasse 1, 79104 Freiburg, Germany of Molecular Biology, Max Planck Institute for Developmental Biology, 72076 Tuebingen, Germany 3Freiburg Initiative for Systems Biology (FRISYS), Schaenzlestrasse 1, 79104 Freiburg, Germany 4Centre for Biological Signalling Studies, Albertstrasse 19, 79104 Freiburg, Germany 5These authors contributed equally to this work 6Present address: Department of Plant Systems Biology, Flanders Institute for Biotechnology and Department of Plant Biotechnology and Genetics, Ghent University, Technologiepark 927, 9052 Gent, Belgium *Correspondence: [email protected] (R.R.), [email protected] (W.F.) DOI 10.1016/j.cell.2009.12.023 2Department

SUMMARY

MicroRNAs (miRNAs) control gene expression in animals and plants. Like another class of small RNAs, siRNAs, they affect gene expression posttranscriptionally. While siRNAs in addition act in transcriptional gene silencing, a role of miRNAs in transcriptional regulation has been less clear. We show here that in moss Physcomitrella patens mutants without a DICER-LIKE1b gene, maturation of miRNAs is normal but cleavage of target RNAs is abolished and levels of these transcripts are drastically reduced. These mutants accumulate miRNA: target-RNA duplexes and show hypermethylation of the genes encoding target RNAs, leading to gene silencing. This pathway occurs also in the wild-type upon hormone treatment. We propose that initiation of epigenetic silencing by DNA methylation depends on the ratio of the miRNA and its target RNA.

INTRODUCTION Small RNAs (sRNAs) are potent regulators of posttranscriptional and transcriptional gene expression (Carthew and Sontheimer, 2009; Voinnet, 2009). In plants, miRNAs produced from hairpin-like precursor transcripts are also required for the biogenesis of trans-acting small interfering RNAs (ta-siRNAs). Both miRNAs and ta-siRNAs regulate mRNA stability and translation. Another class of sRNAs, siRNAs, processed from perfectly double-stranded RNA (dsRNA), posttranscriptionally silence transposons, viruses, and transgenes and are important for DNA methylation (Baulcombe, 2004; Matzke et al., 2007). Evidence for a similar function of microRNAs (miRNAs) in DNA methylation is limited. The biogenesis of sRNAs from dsRNA is catalyzed by Dicer proteins. The number of different Dicer proteins varies between organisms, reflecting different degrees of specialization. For example, in D. melanogaster, Dcr1

produces miRNAs from hairpin precursors, whereas Dcr2 generates siRNAs from dsRNA (Tomari and Zamore, 2005), while in C. elegans the only Dicer protein produces sRNAs from different dsRNA triggers (Duchaine et al., 2006). Besides their function in dicing dsRNA, animal Dicers together with accessory proteins act in RNA-induced silencing complex (RISC) or RISC loading complexes (Doi et al., 2003; Pham et al., 2004). Thus, Dicer proteins are also essential components in the executive phase of RNA interference (RNAi), indicating that miRNA/siRNA processing and target RNA cleavage are coupled. In the seed plant A. thaliana, four Dicer proteins (AtDCL1–4) act in different sRNA pathways (Gasciolli et al., 2005; Henderson et al., 2006). The maturation of miRNAs from imperfect RNA foldbacks relies on AtDCL1. In consequence, Atdcl1 mutants show reduced miRNA levels and increased target mRNA levels, resulting in multiple developmental defects (Golden et al., 2002; Park et al., 2002). AtDCL2 mediates the generation of siRNAs from exogenous RNA (Xie et al., 2004), AtDCL3 acts in the formation of heterochromatin-associated endogenous siRNAs (Herr et al., 2005; Xie et al., 2004), and AtDCL4 is needed for the formation of ta-siRNAs acting in cell-to-cell transmission of silencing signals (Dunoyer et al., 2005; Xie et al., 2005). A moss, Physcomitrella patens, likewise harbors four DCL proteins (Axtell et al., 2007). PpDCL1a and PpDCL1b are similar to AtDCL1. PpDCL3 and PpDCL4 are orthologs of AtDCL3 and AtDCL4, whereas an AtDCL2 ortholog is lacking. Like AtDCL1, the primary PpDCL1a transcript harbors an intronic miRNA precursor, suggesting a conserved autoregulatory control (Axtell et al., 2007). We created null mutants of PpDCL1a and PpDCL1b, respectively, by reverse genetics. Their analysis revealed significant functional differences between both proteins and suggest a feedback control of gene expression involving miRNAs and DNA methylation.

RESULTS Requirement of PpDCL1a for miRNA Biogenesis Utilizing efficient gene targeting in P. patens (Strepp et al., 1998), we generated two PpDCL1a null mutants (DPpDCL1a) (Figure S1

Cell 140, 111–122, January 8, 2010 ª2010 Elsevier Inc. 111

PpDCL1a mutant 1

A WT

0.5 cm

0.5 cm PpDCL1a mutant 1

B WT

100 µM

100 µM

C

PpDCL1a mutants WT 1 2

(A) Protonema of identical density from the WT and two DPpDCL1a mutants grown for 28 days on solid medium. (B) Protonema grown in liquid culture. (C) RT-PCR of miR156, 160, 166, and 390. (D) RT-PCR of ta-siRNAs pptA013298 (from PpTAS3) and pptA079444 (from PpTAS1). (E) Semiquantitative RT-PCR of miRNA targets in the WT and DPpDCL1a. Bars indicate standard errors (n = 3). See Figure S1 for further molecular analyses and Figure S2 for phenotypic comparison with DPpDCL1b mutants.

0.5 cm PpDCL1a mutant 2

100 µM

D WT miR156 miR160 miR166 miR390

Figure 1. Analysis of DPpDCL1a Mutants

PpDCL1a mutant 2

PpDCL1a mutants 2 1 ta-siRNA pptA013298 ta-siRNA pptA079444

Relative transcript levels

E 7 6 5

Two ta-siRNAs (pptA079444 from PpTAS1 and pptA013298 from PpTAS3) (Axtell et al., 2006) were present in the wild-type (WT) but absent in DPpDCL1a (Figure 1D), indicating that lack of miR390 abolishes ta-siRNA production. To test whether reduced miRNAs levels result in elevated levels of target mRNAs, as described for A. thaliana dcl1 mutants, we analyzed miR156 target PpSBP3 (Arazi et al., 2005), miR166 targets PpC3HDZIP1 and PpHB10 (Axtell et al., 2007; Floyd and Bowman, 2004), miR160 target PpARF (Fattash et al., 2007), and miR390 target PpTAS1 (Axtell et al., 2006). All target RNAs accumulated to higher levels in DPpDCL1a (Figure 1E), revealing that PpDCL1a is the major P. patens DCL protein for miRNA biogenesis and thus the functional equivalent of AtDCL1.

WT

Requirement of PpDCL1b for miRNA-Guided Target Cleavage PpDCL1a mutant 2 3 To test possible redundancies between the two DCL1 homologs in moss, we additionally gener2 ated four PpDCL1b null mutants (DPpDCL1b) 1 (Figure S2). These were affected in cell division, 0 growth polarity, cell size, cell shape, and differPpSBP3 PpARF PpC3 PpHB10 PpTAS1 HDZIP1 entiation (Figure 2A and Figure S3A). In contrast to DPpDCL1a, DPpDCL1b grew faster and available online). Loss of PpDCL1a resulted in developmental developed some malformed leafy stems (Figure 2B and disorders and abnormalities in cell size and shape. Growth of Figure S3B). Thus, albeit DPpDCL1a and DPpDCL1b exhibited the mutants was severely retarded on minimal medium but severe defects, their phenotypes differed (Figures S2G and slightly better when supplied with vitamins and glucose. In all S2H). conditions, DPpDCL1a were developmentally arrested at the The levels of six different miRNAs (miR156, 160, 166, 390, 535, filamentous protonema and did not form leafy stems (Figures and 538) (Arazi et al., 2005; Fattash et al., 2007) in DPpDCL1b 1A and 1B, and Figures S2G and S2H). were like in the WT (Figures 2C–2E), revealing that PpDCL1b is To test miRNA biogenesis in these barely growing mutants, not pivotal for miRNA maturation and that the strong mutant we analyzed the accumulation of miR156, 160, 166, and 390 phenotype does not result from abolished miRNAs. Subse(Arazi et al., 2005; Fattash et al., 2007) by RT-PCR (Varkonyi- quently, cleavage of the miRNA targets PpARF, PpC3HDZIP1, Gasic et al., 2007) and sequenced the products as a control. PpHB10, and PpSBP3 in the WT was verified by 50 RACE, Levels of miR156, 160, and 166 were drastically reduced and complementary DNA (cDNA) cloning, and sequencing (FigmiR390 was undetectable in DPpDCL1a (Figure 1C). In P. ure 3A). An mRNA (PpGNT1) (Koprivova et al., 2003) that is no patens, ta-siRNA precursors (PpTAS1–4 RNAs) are cleaved at miRNA target served as control (Figure 3A). Although DPpDCL1b two distinct miR390 target sites. Subsequently, dsRNAs are accumulated normal levels of miRNAs, miRNA targets were not generated and processed in a phased manner to generate cleaved (Figure 3A), revealing a surprising requirement of ta-siRNAs (Axtell et al., 2006; Talmor-Neiman et al., 2006). PpDCL1b for miRNA-guided RNA cleavage. 4

112 Cell 140, 111–122, January 8, 2010 ª2010 Elsevier Inc.

PpDCL1a mutant 1

PpDCL1b mutant 1

A WT

PpDCL1b mutant 1

B WT

100 µm

4d 200 µm

C WT

6d

PpDCL1b mutants 3 1 2 4

21nt

miR156

21nt

miR390

21nt

miR535

21nt

miR538

8d

Figure 2. Analysis of DPpDCL1b Mutants (A) Regeneration of protoplasts from the WT and DPpDCL1b mutant 1 over indicated time points. Scale bars represent 100 mm, or 500 mm for the 18-day-old and 8-week-old plants (B) Scanning electron micrographs of gametophores. (C) RNA blots with 30 mg total RNA from protonema probed for miR156, miR390, miR535, and miR538. An antisense probe for U6snRNA served as loading control. (D) RNA blot with 80 mg total RNA from protonema treated with 5 mM auxin (NAA) for 8 hr probed for miR160. (E) RNA blot with 80 mg total RNA from gametophores probed for miR166. EtBr staining served as a loading control at the bottom of (D) and (E). See Figure S2 for further molecular analyses, complete images of blots including size markers, and phenotypic comparison with DPpDCL1a mutants and Figure S3 for phenotypes of other DPpDCL1b mutants.

after cleavage of mRNAs. These siRNAs most likely target mRNAs at additional sites, thus generating the additional fragments observed by 50 RACE. We did neither detect siRNAs in miR160 DPpDCL1b or for a control mRNA (PpGNT1) in EtBr the WT (Figures 3C and 3D), revealing that generation of transitive siRNAs depends on PpDCL1b and is specific for RNAs subject to miR166 miRNA-directed cleavage. Biogenesis of these EtBr transitive siRNAs involves RdRP-dependent formation of dsRNA precursors as we could synthesize cDNA from the antisense strand of miRNA target transcripts (Figure 3C) and detected transitive siRNAs in sense and antisense orientation (Figure 3D). Thus, biogenesis of transitive siRNAs in P. patens differs from biogenesis of secondary siRNAs in C. elegans, as these occur in antisense polarity only because of unprimed de novo synthesis by RdRP (Pak and Fire, 2007; Sijen et al., 2007). U6snRNA

D 21nt 18 d

E 21nt

8 weeks

Generation of Transitive siRNA Triggered by miRNA-Guided Transcript Cleavage In the WT, we detected 50 RACE products resulting from miRNAguided cleavage of target RNA, as well as various shorter and longer products (Figure 3A). Analogous to other plants, where targeting of transcripts with dsRNA-derived siRNAs or multiple miRNAs (Axtell et al., 2006; Howell et al., 2007; Vaistij et al., 2002) causes production of secondary siRNAs, the miRNA cleavage products detected here may serve as templates for synthesizing cRNA by RNA-dependent RNA polymerase (RdRP). Subsequently, dsRNAs may be processed into secondary siRNAs that spread the initial trigger (Figure 3B). In seed plants, this phenomenon, known as transitivity, was rarely observed in miRNA-based regulation of gene expression (Axtell et al., 2006; Howell et al., 2007) but may be more prevalent than suspected (Luo et al., 2009). In accordance with possible transitivity in P. patens, we could synthesize cDNAs from sense and antisense strands of miRNA target mRNAs (PpARF and PpC3HDZIP1), indicating the presence of dsRNA, in the WT but not in DPpDCL1b (Figure 3C). To determine whether transitive siRNAs were generated from these dsRNAs, we performed RNA blots with probes for sequences upstream and downstream of the miRNA targeting site. Small RNAs corresponding to sense and antisense strands of PpARF and PpC3HDZIP1 mRNAs were present in the WT but not in DPpDCL1b (Figure 3D and Figure S4). Thus, in P. patens, transitive siRNAs arise from regions upstream as well as downstream of the miRNA targeting motif

DNA Methylation of Genes Encoding miRNA Targets in DPpDCL1b A. thaliana ago1 and dcl1 mutants are defective in miRNAdirected cleavage or miRNA biogenesis, respectively, and thus overaccumulate miRNA target transcripts (Ronemus et al., 2006). Conversely, in DPpDCL1b, all miRNA targets analyzed had reduced levels (Figure 4A and Figure S5A), although transcripts were not cleaved. These unexpected findings hint at an epigenetic control of genes encoding miRNA targets. Since methylation of cytosine residues is associated with transcriptional silencing in eukaryotes (Bender, 2004), we performed methylation-specific PCR from five loci, four encoding miRNA targets and one encoding PpGNT1 (Figures 4B–4E, Figure S5B, and Figure S6A–S6E). Promoters of the five genes were unmethylated in the WT, whereas in DPpDCL1b the four promoters of miRNA targets were methylated, but not the promoter of PpGNT1 (Figure 4C). This was confirmed by sequencing the PCR products from the PpARF promoter, demonstrating no methylation in the WT and methylation of CpG residues in DPpDCL1b (Figure S6A). Taken together, our experiments


A

Figure 3. RNA Cleavage Products, Antisense Transcripts, and Transitive siRNAs

PpDCL1b mutants WT 1 2 3 4 6/9 5’ UGGCAUGCAGGGGGCCAGGCA 3’ 3’ ACCGUAUGUCCCUCGGUCCGU 5’

PpARF miR160

PpC3HDZIP1 5’ miR166 3’

7/9 CU U GG AUGAAGCCUGGUCCGG 3’ CC UACUUCGGACCAGGCU 5’ CC U

PpHB10 miR166

5’ 3’

7/9 CU U GG AUGAAGCCUGGUCCGG 3’ CC UACUUCGGACCAGGCU 5’ CC U

PpSBP3 miR156

U 8/9 5’ GUGCUC CUCUCUUCUGUCA 3’ 3’ CACGAG GAGAGAAGACAGU 5’ U

PpGNT1 (control)

B

5’

3’

5’

3’ 5’

5’

3’

3’ 5’ RdRP

3’ RdRP

transitive siRNAs 5’

3’

5’

3’ 5’

5’

C

PpARF KO KO WT 1 2

antisense cDNA PpC3HDZIP1 KO KO WT 1 2

antisense cDNA PpGNT1 KO KO WT 1 2

antisense cDNA

D

3’

3’

5’

3’

3’

5’

PpEF1 PpARF PpEF1 KO KO KO KO KO KO WT 1 2 WT 1 2 WT 1 2

control

sense cDNA

control

PpC3HDPpEF1 ZIP1 KO KO KO KO KO KO WT 1 2 WT 1 2 WT 1 2 PpEF1

control

sense cDNA

control

PpGNT1 PpEF1 PpEF1 KO KO KO KO KO KO WT 1 2 WT 1 2 WT 1 2

control

sense cDNA

control

Upstream Downstream Sense Sense Antisense Antisense KO KO KO KO KO KO KO KO WT 1 2 WT 1 2 WT 1 2 WT 1 2

21nt

PpARF

21nt

PpC3HDZIP1 U6 sRNA


(A) 50 RACE products of miRNA targets and a control transcript (PpGNT1) from the WT and DPpDCL1b. Arrows mark products of the size expected for cleavage products that were eluted, cloned, and sequenced. Numbers above miRNA:target alignments indicate sequenced RACE products with the corresponding 50 end. (B) Scheme for the generation of transitive siRNAs. DsRNA is synthesized from cleaved miRNA targets by RdRP, processed into transitive siRNAs that mediate cleavage of the target RNA upstream and downstream of the miRNA recognition motif. Black line, mRNA; gray box, miRNA binding site; curved line, miRNA. Arrows indicate primers for RT-PCR, with gray indicating primers for cDNA synthesis from antisense strand and black for sense strand. (C) RT-PCR products derived from antisense or sense-specific cDNAs from the WT and two DPpDCL1b mutants (KO1 and KO2). (D) Detection of sense and antisense transitive siRNAs derived from PpARF and PpC3HDZIP1 RNAs using hybridization probes targeting regions upstream and downstream of the miRNA binding sites. U6snRNA served as a control. See Figure S4 for complete images of blots including size markers.

show that targeted disruption of the PpDCL1b gene led to loss of miRNA-directed mRNA cleavage accompanied by specific epigenetic changes in genes encoding miRNA targets. While it is well known that siRNA pathways govern DNA methylation in A. thaliana, e.g., at repeat-associated loci (Cokus et al., 2008; Herr et al., 2005; Lister et al., 2008), only one study suggests a function of miRNAs in gene silencing: normally methylated DNA sequences downstream of the miRNA complementary motif became hypomethylated in plants with dominant alleles of PHB and PHV, encoding targets of miR165/166, while the promoters remained unmethylated (Bao et al., 2004). The dominant alleles alter the miRNA targeting motif, so that the mRNAs are no longer miRNA targets. Like AtPHB and AtPHV, the moss genes PpC3HDZIP1 and PpHB10 harbor an intron in the miRNA binding site (Figure S5B), whereas the miRNA targeting motif of PpARF is not disrupted. Similar to the promoters, PpC3HDZIP1 and PpARF sequences flanking the miRNA targeting motif as well as the disrupting intron in PpC3HDZIP1 were CpG methylated in DPpDCL1b but not in the WT (Figures 4D and 4E, and Figure S6B–S6E). To study whether this hypermethylation in DPpDCL1b leads to transcriptional silencing, we performed nuclear run-ons. When compared to the WT, transcription rates of genes encoding miRNA targets were reduced in DPpDCL1b, whereas transcription rates of the unmethylated PpGNT1 control gene were similar in the WT and mutants (Figure 4F and Figure S6F). Thus, we conclude that hypermethylation of genes encoding miRNA targets in DPpDCL1b downregulates their expression, hence accounting for the reduced levels of their mature RNAs. A possible scenario for these findings is that PpDCL1b is pivotal for a cleavage competent RISC and that in DPpDCL1b miRNAs are not loaded to RISC but to an RNA-induced transcriptional silencing (RITS) complex that induces methylation and subsequent downregulation of the corresponding genes. As the sequences encoding the future miRNA binding site are disrupted by introns in PpC3HDZIP1 and PpHB10 it is unlikely that their methylation in DPpDCL1b is initiated by miRNA:DNA hybrids. Instead, the miRNA-loaded RITS complex may interact with target mRNAs forming stable miRNA:mRNA duplexes.

A

B

PpDCL1b 3 2 4

WT 1

PpARF 100

7

6

6

9

8

15

12

100 31 29

20

20

100 42 37

28

25

MSP KO KO WT 1 2

WT

USP KO KO 1 2 PpARF

PpARF

PpC3HDZIP1 100 13

C

DNA Untreated + Bisulfite DNA KO KO KO KO WT 1 2 WT 1 2

PpC3HDZIP1

PpC3HDZIP1

PpHB10

PpHB10

PpHB10

PpSBP3 PpGNT1

PpSBP3

PpSBP3

PpGNT1

PpGNT1

100 100 100 102 102 PpEF1

D

Exon MSP KO KO WT 1 2

USP KO KO WT 1 2

E

Intron MSP KO KO WT 1 2

USP KO KO WT 1 2 upstream of miR160 binding site

Intron located in miR166 binding site

Intron upstream of miR166 binding site MSP KO KO WT 1 2

USP MSP KO KO KO KO WT 1 2 WT 1 2

Exon downstream of miR166 binding site USP MSP USP KO KO KO KO KO KO WT 1 2 WT 1 2 WT 1 2

downstream of miR160 binding site

Relative transcription rate

F

WT

PpDCL1b KO1

PpDCL1b KO2

1.2 1.0 0.6 0.6 0.4 0.2 0.0

PpGNT1 PpC3HD- PpHB10 PpSBP3 ZIP1

G

H

PpDCL1b mutants WT 1 2 PpARF PpC3HDZIP1 PpHB10 PpSBP3 PpEF1

PpEF1 KO KO WT 1 2

PpC3HDPpARF ZIP1 KO KO KO KO WT 1 2 WT 1 2

PpHB10 PpSBP3 KO KO KO KO WT 1 2 WT 1 2

PpARF

Figure 4. Expression of miRNA Target Genes, DNA Methylation, and miRNA:mRNA Duplexes (A) RNA blot analysis of genes encoding miRNA targets (PpARF, PpC3HDZIP1, PpHB10, PpSBP3) and control genes (PpGNT1 and PpEF1a) in the WT and DPpDCL1b mutants (1–4). Hybridization signals were normalized to PpEF1a and transcript levels relative to the WT are indicated. See Figure S5A for an independent expression analysis by qRT-PCR. (B) Specificity analysis of bisulfite PCR using primers specific for unmodified sequences. PCR was performed with untreated and bisulfite-treated genomic DNA of the WT and two DPpDCL1b mutants (KO1 and KO2). (C–E) PCR reactions with bisulfite-treated genomic DNA using methylation-specific (MSP) and unmethylation-specific (USP) primers. (C) Bisulfite PCR for promoters of genes encoding miRNA targets and the PpGNT1 control. (D) Bisulfite PCR analysis of PpARF sequences surrounding the miR160 targeting motif. See Figure S5B for gene structures and Figure S6A–S6E for nucleotide sequences of PCR products obtained from bisulfite PCR of the PpARF gene. (E) Bisulfite PCR analysis of PpC3HDZIP1 sequences upstream of, the intron disrupting, and downstream of the miR166 targeting motif. Arrows in (C)–(E) mark primer bands. (F) Relative transcription rates of genes encoding miRNA targets and of PpGNT1 in the WT and in DPpDCL1b determined by nuclear run-ons. Hybridization signals were normalized to PpEF1a and transcription rates in the WT were set to 1. See Figure S6F for images of hybridized membranes. (G) PCR products of miRNA targets using cDNA synthesized from the WT and two DPpDCL1b mutants (1 and 2) without exogenous primers. For the PpGNT1 control PCR products were detected from neither the WT nor DPpDCL1b (data not shown). A PpEF1a primer for cDNA synthesis was added as internal control to all reactions. (H) The same experiment performed with RNA samples that had been heated for 5 min to 95 C prior to cDNA synthesis. The control PpEF1a primer was added after cooling of the RNA samples.

Subsequently, these duplexes could guide the RITS complex to the corresponding genomic regions, initiate and spread DNA methylation. If stable miRNA:mRNA duplexes exist, they should prime cDNA synthesis without exogenous primers. In support of this scenario, we obtained RT-PCR products from unprimed cDNA synthesis for all miRNA targets examined but not for a control transcript in DPpDCL1b. No such products were obtained with RNA from the WT (Figure 4G). In addition, from DPpDCL1b no PCR products were obtained with primers located downstream of the miRNA targeting site. As a further control, we heated the

RNA samples prior to cDNA synthesis. This should lead to denaturation of miRNA:mRNA complexes and hence eliminate priming. Indeed, this setup prevented amplification of PCR products from DPpDCL1b (Figure 4H). These findings point at the presence of base-paired miRNA:mRNA duplexes in DPpDCL1b but not in the WT. To further scrutinize our hypotheses of transitivity and miRNAdependent DNA methylation, we analyzed the ta-siRNA pathway. After miR390-mediated cleavage of TAS precursors the RNA cleavage products are converted into dsRNA and further processed into ta-siRNAs (Axtell et al., 2006; Talmor-Neiman


A

Figure 5. The ta-siRNA Pathway

PpDCL1b mutants WT 1 2 3 4 PpTAS4 miR390-5’

(A) 50 RACE-PCRs from the WT and DPpDCL1b mutants (1–4) for miR390 target PpTAS4 and ta-siRNA target PpEREBP/ AP2. Arrows mark products of the size expected for cleavage products that were eluted from the gel, cloned, and sequenced. The number of sequenced RACE-PCR products with the corresponding 50 end is indicated above the alignment. (B) ta-siRNAs derived from PpTAS4 and transitive siRNAs derived from PpEREBP/AP2. U6snRNA served as a control. See Figure S7 for complete images of blots including size markers. (C) RNA blots for PpTAS4 and PpEREBP/AP2 transcripts. EtBr staining as loading control below. (D) Bisulfite PCR with methylation-specific (MSP) and unmethylation-specific (USP) primers for PpTAS4 and PpEREBP/ AP2. Arrow marks primer dimers.

5/6 5’ GGCGUUAUCCCUCUUGAGCUG 3’ 3’ CCGCGAUAGGGAGGACUCGAA 5’

U A A 8/9 5’ G UGU UAUC CUCCUGAGCUA 3’ 3’ C GCG-AUAG GAGGACUCGAA 5’ C G C G 8/8 PpEREBP/AP2 5’ GAAGCA UCAUCACACCCUA 3’ ta-siRNA6(+)3’ CUUCGU AGUAGUGUGGGAU 5’ A G PpTAS4 miR390-3’

C

B Antisense KO KO WT 1 2

Sense KO KO WT 1 2

WT 1

21nt

PpTAS4

21nt

PpEREBP/AP2

PpDCL1b 2 3 4 PpTAS4 PpEREBP/AP2 rRNA

U6 sRNA PpEREBP/AP2

D PpTAS4 MSP KOKO WT 1 2

Coding Sequence

USP MSP KOKO KOKO WT 1 2 WT 1 2

USP KOKO WT 1 2

Promoter MSP USP KOKO KOKO WT 1 2 WT 1 2

et al., 2006). In P. patens, the mRNA encoding an EREBP/AP2 transcription factor is targeted by one of the ta-siRNAs derived from the TAS4 precursor (Talmor-Neiman et al., 2006). TAS4 RNA cleavage products were detected by 50 RACE-PCR in the WT but not in DPpDCL1b (Figure 5A), although miR390 was present in equal amounts in both (Figure 2C). Furthermore, ta-siRNAs of both sense and antisense orientation were present in the WT but undetectable in DPpDCL1b (Figure 5B and Figure S7), confirming the requirement of PpDCL1b for initiation of the ta-siRNA pathway. In agreement with our findings for other miRNA targets, TAS4 transcript levels were reduced (Figure 5C) and the TAS4 gene was methylated in DPpDCL1b only (Figure 5D). If, similar to miRNAs, ta-siRNA-mediated cleavage of target RNAs initiates the generation of transitive siRNAs, the lack of ta-siRNAs in DPpDCL1b should abolish both cleavage of EREBP/AP2 mRNA and generation of transitive siRNAs. Consistent with this, EREBP/AP2 mRNA was cleaved in the WT but not in DPpDCL1b (Figure 5A), and only the WT produced EREBP/ AP2 mRNA-derived siRNAs in sense and antisense orientation (Figure 5B and Figure S7). These observations indicate that siRNA-dependent amplification of target RNA degradation initially triggered by ta-siRNA- or miRNA-guided cleavage is a common phenomenon in P. patens. In addition, in contrast to direct miRNA targets, EREBP/AP2 mRNA levels should be elevated in DPpDCL1b, as stable ta-siRNA:mRNA duplexes that could initiate silencing of the corresponding gene should be absent. Indeed, EREBP/AP2 mRNA levels were increased in DPpDCL1b (Figure 5C) and the corresponding gene was methylated in neither the WT nor DPpDCL1b (Figure 5D).


Dependence of DNA Methylation on miRNA Levels We propose that miRNA:target-RNA duplexes lead to methylation and subsequent downregulation of the corresponding genes in DPpDCL1b. To gain more insight into this mechanism, we generated transgenic lines that express different levels of an artificial miRNA (amiRNA) targeting the control gene PpGNT1. An amiRNA against PpGNT1 was engineered into the A. thaliana miR319a precursor (Khraiwesh et al., 2008) and transfected into the WT and DPpDCL1b (Figure 6A). RNA blots confirmed precise maturation of amiR-GNT1 in the transformed lines, independent of expression level (Figure 6B and Figure S8D). The expected PpGNT1 mRNA cleavage products were present in the WT and absent in DPpDCL1b (Figure 6C). As a consequence, PpGNT1 transcript levels were reduced in amiR-GNT1 WT plants and even lower in the DPpDCL1b background, despite abolished amiRNA-directed cleavage of PpGNT1 RNA in the latter (Figure 6D). Consistent with our model of miRNA-dependent epigenetic silencing, the PpGNT1 promoter was methylated in DPpDCL1b (Figure 6E and Figure S8A). Beyond that, we observed methylation of this promoter in a WT line with high amiR-GNT1expression levels (line #2) but not in a line with low amiR-GNT1 levels (line #1) (Figure 6E and Figure S8A). Thus, methylation and subsequent silencing of miRNA target loci is not limited to DPpDCL1b but also occurs in the P. patens WT and is miRNA-dosage dependent. Therefore, we hypothesized that the ratio of a miRNA to its target RNA is crucial for induction of DNA methylation at the target locus. If miRNA concentrations exceed a threshold, the miRNA may either interact directly with its target and the duplex might then be recruited to a DNA methylation silencing complex or the excess miRNA might be loaded into an effector complex such as RITS, triggering duplex formation that directs DNA methylation. We obtained supporting evidence for the expected amiR-GNT1:PpGNT1-mRNA duplexes by cDNA synthesis without exogenous primers and subsequent PCR from DPpDCL1b and from a WT line with high levels of amiRNAs while this was impossible from a WT line accumulating moderate amounts of amiR-GNT1 (Figure 6F).

B amiRNA WT lines #1 #2

WT background

#3 #4 #5 #6

PpDCL1b KO1

WT + amiRNA #1 #2

WT

21nt

PpGNT1 amiRNA

U6 snRNA

E

PpDCL1b KO1 + KO2 + amiRNA amiRNA #3 #4 #5 #6

F MSP PpDCL1b WT + amiRNA KO1 KO2 #1 #2 #3 #5

PpGNT1 1.0

PpDCL1b + amiRNA KO1 KO2 WT #3 #5

amiRGNT1

PpDCL1b KO2 PpEF1

D

C

PpDCL1b WT + KO1 + KO2 + amiR amiR amiR WT #1 #2 #3 #4 #5 #6

WT +amiRNA#1

A

USP PpDCL1b WT + amiRNA KO1 KO2 #1 #2 #3 #5

WT + amiRNA #2 #1

6/8 C 5’ AA CGUCCUGAUUAUUUGGAG 3’ 3’ UU GCAGGACUAAUAAACCUU 5’ C

PpDCL1b KO1 KO2 #3 #5

WT + amiRNA #1 #2

PpDCL1b KO1 KO2 #3 #5 PpGNT1

0.22 0.11 0.022 0.017 0.017 0.017

PpEF1

rRNA Duplex RT-PCR

G

I Untr.

Untr.

H ABA 1h 2h 4h 6h 8h PpbHLH

ABA 4h

ABA 1h 2h 4h 6h 8h

Untr. PpbHLH C 3’ 8/10 UCCUCUCAAGUCUUUCUC AGGAGAGUUCAGAAAGAG 3’ AC U 5’ miR1026

miR1026

21nt 1.0 2.9 2.5 3.3

1.0 0.5 0.3 0.3 0.4 0.4

5’ CC

3.0 3.0

PpEF1

U6snRNA

ABA 4h 6h 8h

4h

ABA 6h 8h

Untr.

ABA 4h 6h 8h

K

USP

Untr.

Untr.

MSP

Untr.

PpCOR47

J

4h

ABA 6h 8h PpbHLH

PpbHLH intragenic PpbHLH Promoter PpGNT1 Promoter

Heating control

PpEF1 Duplex RT-PCR

Heating control

Figure 6. Lines Expressing amiR-GNT1 and Analysis of miR1026 Target PpbHLH (A) PCR-based identification of transgenic lines harboring a PpGNT1-amiRNA expression construct in the WT (lines #1 and #2), DPpDCL1b mutant 1 (lines #3 and #4), and DPpDCL1b mutant 2 (lines #5 and #6) backgrounds. PpEF1a served as a control. (B) Detection of amiR-GNT1 by RNA blot loaded with 50 mg total RNA. U6snRNA served as a control. See Figure S8D for complete images of blots including size markers. (C) Cleavage mapping of PpGNT1 in amiR-GNT1 lines by 50 RACE-PCR. The number of sequenced RACE-PCR products with the corresponding 50 end is indicated above the alignment. (D) RNA blot of WT and amiR-GNT1 lines probed for PpGNT1. Hybridization signals were normalized to rRNA. Levels relative to the WT are indicated. (E) Bisulfite PCR on genomic DNA from amiR-GNT1 lines using methylation-specific (MSP) and unmethylation-specific (USP) primers derived from the PpGNT1 promoter. See Figure S8A for nucleotide sequences of PCR products obtained from bisulfite PCR. (F) RT-PCR to detect amiR-GNT1:PpGNT1-mRNA duplexes using cDNA synthesized without exogenous primers. PCR was carried out with a primer pair upstream of the amiR-GNT1 target motif. Amplification controls as in Figures 4G and 4H. Arrows mark primer dimers. (G) RNA blots with 20 mg total RNA from untreated (Untr.) and ABA-treated WT using probes for PpbHLH, the loading control PpEF1a and PpCOR47, a known ABA-induced gene. PpbHLH levels were normalized to PpEF1a. Relative PpbHLH RNA levels compared to the WT are given. (H) RNA blot with 50 mg total RNA from untreated (Untr.) and ABA-treated WT. MiR1026 levels were normalized to U6snRNA. Numbers indicate miR1026 levels relative to the WT. See Figure S8D for complete images of blots including size markers. (I) 50 RACE-PCR for PpbHLH using RNA from untreated (Untr.) and WT treated for 4 hr with ABA. Arrows mark products of the size expected for cleavage products that were eluted from the gel, cloned, and sequenced. Numbers above miRNA:target alignments indicate sequenced RACE-PCR products with the corresponding 50 end. (J) Bisulfite PCR reactions on DNA from untreated (Untr.) and ABA-treated WT with methylation-specific (MSP) and unmethylation-specific (USP) primers targeting PpbHLH genomic sequences. PpGNT1 promoter served as control. Arrows mark primer dimers. See Figures S8B and S8C for nucleotide sequences of PCR products obtained from bisulfite PCR. (K) RT-PCR to detect miR1026:PpbHLH-mRNA duplexes using cDNA synthesized without exogenous primers. PCR was carried out with a primer pair upstream of the miR1026 binding site. Amplification controls as in Figure 4G and 4H. Arrows mark primer dimers.

Hormone-Dependent DNA Methylation So far, we have found evidence for miRNA-directed epigenetic silencing in a knockout mutant and in a transgenic WT that ectopically expresses high levels of an amiRNA. If the proposed mechanism of gene regulation is of general relevance, at least in

moss, it should be detectable in the nontransgenic WT under physiological conditions as well. The plant hormone abscisic acid (ABA) represses mRNA levels of a gene (PpbHLH) encoding a basic helix-loop-helix transcription factor (Richardt et al., 2009). The PpbHLH mRNA was


predicted as target of miR1026 (Axtell et al., 2007). ABA is a potent regulator of abiotic stress signaling pathways in plants including moss (Frank et al., 2005b). Northern blots confirmed downregulation of PpbHLH in response to ABA (Figure 6G). This effect correlated with an ABA-induced increase of miR1026 levels (Figure 6H and Figure S8D). MiR1026-mediated cleavage of PpbHLH transcripts was confirmed by 50 RACE (Figure 6I), revealing a regulatory circuit between ABA, PpbHLH, and miR1026. Further, upon ABA treatment, the PpbHLH gene was methylated at CpG sites (Figure 6J, and Figures S8B and S8C), mirroring the methylation patterns we found in DPpDCL1b and in the WT ectopically expressing high levels of amiRNA. The promoter of the control gene PpGNT1 was unmethylated in the nontransgenic WT regardless of ABA treatment (Figure 6J). However, under these physiological conditions, methylation of the PpbHLH gene was not quantitative, as unmethylationspecific primers allowed some, albeit inefficient, PCR amplifications from ABA-treated samples. These findings appear to support our model, as we propose that DNA methylation is initiated only if miRNA levels exceed a threshold. In line with this, we obtained evidence for stable miR1026:PpbHLH-mRNA duplexes by unprimed RT-PCR. Consistent with the DNA methylation status, such duplexes were only found in ABA-treated samples (Figure 6K). DISCUSSION We have shown that in P. patens epigenetic silencing of genes encoding miRNA target RNAs contributes to the control of gene expression. Although we initially discovered this phenomenon in DPpDCL1b, subsequent analyses of the miR1026/ PpbHLH regulon confirmed that this type of miRNA-dependent transcriptional control operates also in the WT under physiological conditions. Our studies suggest that PpDCL1a is the functional ortholog of AtDCL1 required for miRNA and ta-siRNA biogenesis. Although PpDCL1b shares a similar level of sequence identity, it has a different function, as its deletion does not affect miRNA biogenesis but abolishes miRNA-directed target cleavage. It is unlikely that PpDCL1b directly cleaves RNAs as AGO proteins in RISC are the catalytic enzymes in sRNA-dependent target cleavage (MacRae et al., 2008). Biochemical analysis of AGO1 complexes from A. thaliana dcl1-7, dcl2-1 and dcl3-1 mutants provided evidence for distinct functional properties. An AGO1 complex extracted from dcl1-7 mutants was not able to cleave RNA targets because of the lack of 21 nt RNA accumulation in this mutant. In contrast, cleavage of RNA targets was not affected in AGO1 complexes from dcl2-1 and dcl3-1 mutants (Qi et al., 2005). Furthermore, purification of AtAGO1 revealed an 160 kDa complex, most likely only consisting of AGO1 and associated sRNA (Baumberger and Baulcombe, 2005). Thus, there is so far no evidence for a role of plant DCL proteins in sRNA-mediated target cleavage. In contrast, studies in animals have shown that Dicer proteins are part of the RNA loading complex (RLC) that loads sRNAs into RISC. Human RLC comprises the proteins Ago2, Dicer, and TRBP, and the purified components assemble spontaneously in vitro without any cofactors. The reconstituted RLC is functional and once


Ago2 is loaded with a miRNA it dissociates from the rest of the complex (MacRae et al., 2008). Similarly, Dcr-2 from D. melanogaster, which produces siRNA, acts in RISC assembly together with R2D2 by loading one of the two siRNA strands into RISC (Liu et al., 2003; Tomari et al., 2004). The C. elegans homolog of this protein, RDE-4, likewise interacts with Dicer (Tabara et al., 2002). Given this particular function of animal Dicers, P. patens DCL1b may equivalently act in loading miRNAs into RISC. Whatever the precise function of this protein is, our results revealed its pivotal role for miRNA-directed target cleavage. As we found no transitive siRNAs derived from miRNA and ta-siRNA targets in DPpDCL1b, it is tempting to speculate that PpDCL1b is involved in the generation of these siRNAs. A. thaliana DCL2 is essential for transitive silencing of transgenes (Mlotshwa et al., 2008; Moissiard et al., 2007). As no AtDCL2 homolog is encoded by the P. patens genome, PpDCL1b may function in siRNA production. Although we cannot exclude the possibility that PpDCL1b generates transitive siRNAs from double-stranded miRNA and ta-siRNA targets, we favor the scenario that miRNA-mediated cleavage of transcripts is essential for the onset of transitivity in P. patens and that the lack of transitive siRNAs in DPpDCL1b is caused by a lack of miRNA target cleavage. In A. thaliana dcl1 and ago1 mutants, which are affected in miRNA biogenesis or miRNA-directed target cleavage, respectively, levels of miRNA targets are elevated (Ronemus et al., 2006). Likewise, levels of miRNA targets are increased in DPpDCL1a because of the lack of miRNAs. In contrast, levels of miRNA targets are drastically reduced in DPpDCL1b, in spite of abolished target cleavage. We found methylation of the corresponding genes in these mutants, suggesting together with evidence from nuclear run-ons an epigenetic control at the transcriptional level. Small RNAs initiate transcriptional silencing of homologous sequences by methylation of cytosine residues at CpG, CpNpG, and CpHpH motifs or by histone modifications (Bender, 2004; Cao and Jacobsen, 2002). In all genes analyzed in DPpDCL1b, we solely detected methylation at CpG, although we cannot exclude that methylation also occurs in different sequence contexts in other regions. Moreover, we detected methylation of the genes encoding miRNA targets in introns, exons, and promoters, pointing to methylation that is able to spread over long distances. Although spreading of siRNA-directed DNA methylation into adjacent nonrepeated sequences is not common in A. thaliana, siRNA-mediated spreading of DNA methylation was found in the SUPPRESSOR OF drm1 drm2 cmt3 (SDC) locus, where methylation spreads beyond siRNA generating repeat regions present in the promoter (Henderson and Jacobsen, 2008). In A. thaliana, methylation can also spread in the PHV and PHB genes that are targets of miR165/166 (Bao et al., 2004). In both genes, the miR165/166 complementary motif is disrupted by an intron and the coding sequences were heavily methylated downstream of the miRNA complementary site in differentiated but not in undifferentiated cells of WT plants. Furthermore, methylation is reduced in phv-1d and phb-1d mutants, which have an altered miRNA recognition motif or a mutation in the intron splice donor sequence, suggesting that miR165/166 needs to bind

to nascent PHV and PHB transcripts to trigger gene silencing (Bao et al., 2004). Similarly, the P. patens genes PpC3HDZIP1 and PpHB10 are targeted by miR166 and its binding sites are only reconstituted after splicing of the primary transcripts. Both loci are hypermethylated in DPpDCL1b but not in the WT, suggesting that initiation of methylation upon defective target cleavage is not mediated by miR166 but involves interaction of miR166 with its target RNAs. We obtained evidence for the presence of stable duplexes of a miRNA and its target RNA in DPpDCL1b and propose that these duplexes guide a DNA modification complex. In A. thaliana RNA-directed DNA methylation (RdRM) by siRNAs requires RDR2, DCL3, and RNA PolIVa, which are all involved in siRNA biogenesis (Herr et al., 2005; Kanno et al., 2005; Onodera et al., 2005; Pontier et al., 2005; Xie et al., 2004), whereas AGO4, DRM2, DRD1, and RNA PolIVb are necessary for DNA methylation (Cao and Jacobsen, 2002; Kanno et al., 2005; Zilberman et al., 2004). P. patens encodes six AGO homologs (Axtell et al., 2007). Three of them cluster with AGO1 from A. thaliana and are under control of miR904, reminiscent of the autoregulatory control of AtAGO1 by miR168. Thus, it is likely that they are components of a P. patens miRNA-RISC. The other three moss AGO paralogs cluster with Arabidopsis AGO6, AGO4, AGO9, and AGO8 (supported by a bootstrap value of 99). Thus, it is unclear which of these P. patens proteins is functionally equivalent to AtAGO4. Furthermore, we cannot rule out functional redundancies between them. In fission yeast, RNA-directed heterochromatic gene silencing at centromeres relies on two complexes, the RITS complex with Ago1, Chp1, and TAS3, and the Argonaute siRNA chaperone complex (ARC) with Ago1, Arb1, and Arb2. However, these complexes are required to direct histone methylation, but do not trigger DNA methylation. Nevertheless, it was suggested that their action involves the recognition of nascent transcripts by RITS-bound siRNAs to promote recruitment of chromatinmodifying enzymes that implement silencing (Bu¨hler et al., 2006; Motamedi et al., 2004). So far, chromatin immunoprecipitation with subsequent unprimed RT-PCR was not successful in our hands, most probably because heat-induced reversal of crosslinking also denatures miRNA:RNA duplexes. Thus, it remains open whether the miRNA:target-RNA duplexes observed here form from nascent or from mature transcripts. We detected specific silencing of genes encoding miRNA targets not only in a targeted knockout mutant but also in the P. patens WT, where the expression of amiR-GNT1 caused methylation of PpGNT1. Methylation of this gene was dependent on high amiR-GNT1 abundance. Likewise, amiR-GNT1:PpGNT1mRNA duplexes were characteristic for lines with high amiRNA levels, supporting the model that miRNA:target-RNA duplexes are required for gene-specific DNA methylation. Finally, we found that the PpbHLH gene encoding a miR1026 target was methylated in response to ABA. As with amiR-GNT1 and its target, DNA methylation was miR1026 dosage-dependent and appeared to correlate with the formation of miR1026:PpbHLHmRNA duplexes. As ABA is a mediator of abiotic stress signaling, we suggest that miR1026-regulated silencing of PpbHLH is part of stress adaptation in the P. patens WT under physological conditions.

In plants, epigenetic changes in response to stress include DNA methylation, histone modifications, and chromatin remodeling (Boyko and Kovalchuk, 2008; Dyachenko et al., 2006; Henderson and Dean, 2004). In addition to that, our analysis of the miR1026:PpbHLH regulon suggests that miRNAs also act in the epigenetic control of stress-responsive genes. Taken together, we propose that epigenetic gene silencing can be triggered by duplexes of a miRNA and its target RNA, being either an mRNA or a primary TAS transcript. This transcriptional control of genes encoding miRNA target RNAs discovered in P. patens presents a new mechanism to affect the homeostasis of miRNA-regulated RNAs (Figure 7). Recently, two studies found an involvement of miRNAs in transcriptional gene silencing in mammals (Gonzalez et al., 2008; Kim et al., 2008) but the proposed mechanisms differ from the model presented here. The precursor of human miR-320 is encoded in antisense orientation proximal to the transcription start site of the POL3RD gene, suggesting a cis-regulatory function of miR-320. Evidence for this scenario was obtained from transfection of HEK293 cells with miR-320 which caused transcriptional silencing of POL3RD (Kim et al., 2008). Epigenetic changes such as DNA methylation or histone modifications were not analyzed (Kim et al., 2008). Gonzalez et al. (2008) reported on the role of mammalian miRNAs in repressing promoter activity, which requires promoter-overlapping transcription and the presence of miRNA seed-matches in the transcribed strand of the promoter. In addition to quite different underlying mechanisms of miRNA action, we note that these studies relied solely on ectopic expression of miRNAs in cell culture. Our model entails that the specific equilibrium of a cleavagecompetent RISC and a DNA-modifying RITS loaded with the same miRNA determines the relative contribution of both pathways to miRNA-mediated downregulation of gene expression in P. patens. In addition, siRNA-mediated transitivity as a major factor in amplifying the original miRNA- and ta-siRNA-directed cleavage signal appears to be more prevalent in this moss than in A. thaliana. It seems not unlikely that similar modifications and specializations of RNAi pathways will be found in other eukaryotes as well. EXPERIMENTAL PROCEDURES Plant Material Culture of P. patens, transformation, and molecular analyses were performed according to Frank et al. (2005a). For hormone treatment 10 mM (±)-cis-trans ABA was added to P. patens liquid cultures. DPpDCL1a mutants were propagated on full medium (Egener et al., 2002). Generation of DPpDCL1a and DPpDCL1b Mutants An nptII cassette was cloned into single restriction sites present in PpDCL1a and PpDCL1b, respectively. The gene disruption constructs were transfected into P. patens protoplasts and G418-resistant lines were analyzed by PCR to confirm precise integration into the targeted genes. Loss of PpDCL1a and PpDCL1b transcript, respectively, was confirmed by RT-PCR. The generation of DPpDCL1a and DPpDCL1b mutants is described in detail in the Extended Experimental Procedures and Figures S1 and S2. P. patens Expressing amiR-GNT1 The generation of an amiRNA targeting PpGNT1 was described previously (Khraiwesh et al., 2008). The amiRNA expression construct was transfected


Figure 7. Model for Posttranscriptional and Transcriptional Silencing of Genes Encoding miRNA Targets

Post-transcriptional gene silencing

RISC

PpDCL1b

RISC

PpDCL1b

target-RNA (mRNA or TAS-RNA)

+

AAAAA

Nucleus

AAAAA

mRNA or TAS-RNA cleavage products

RdRP

pre-miRNA

double-stranded mRNA cleavage products

double-stranded TAS-RNA cleavage products

transitive siRNAs

PpDCL1a

Cytosol

ta-siRNAs

Amplification of the miRNA signal

Targeting and cleavage of ta-siRNA targets

Low miRNA:target-RNA ratio

AAAAA

target-RNA cleavage

miRNA transitive siRNAs

RITS target-RNA gene (mRNA or TAS-RNA) CH3

CH3

AAAAA

CH3

RITS

RITS AAAAA

AAAAA

Epigenetic silencing AAAAA

Formation of miRNA:target-RNA duplexes

Formation of miRNA:target-RNA duplexes

High miRNA: target-RNA ratio

AAAAA

Whereas PpDCL1a acts in miRNA biogenesis, PpDCL1b is pivotal for a cleavage-competent RISC. At low miRNA:target-RNA ratios, gene expression is regulated posttranscriptionally via RISC. After loading of miRNAs into RISC, miRNA: target-RNA duplexes form on the basis of sequence complementarities, resulting in target RNA cleavage. Amplification of the miRNA signal by transitive siRNAs occurs. Targeted deletion of the PpDCL1b gene abolishes this pathway in the mutant and leads to gene silencing via RITS. In the WT, hormone treatment enhances miRNA levels and thus increases miRNA:target-RNA ratios. In addition to loading miRNA into RISC, miRNAs form duplexes with their target RNAs. Highly abundant miRNAs are either loaded into a RITS complex and subsequently interact with their target to form a duplex, or these duplexes are formed at first and then loaded into RITS. The miRNA:RNA duplexes bound by RITS initiate DNA methylation at complementary genomic loci. The RITS complex is able to spread CpG methylation over the entire locus.

Transcriptional gene silencing

into the P. patens WT and DPpDCL1b. The generation of amiR-GNT1-expressing lines is described in detail in the Extended Experimental Procedures. RT-PCR of Small RNAs RT-PCR analyses of miRNAs and ta-siRNAs was carried out as described (Varkonyi-Gasic et al., 2007). Primers for cDNA synthesis and subsequent PCR reactions are reported in Table S1. DNA Methylation Analysis DNA sequences were analyzed with the MethPrimer program (Li and Dahiya, 2002) to deduce methylation-specific (MSP) and unmethylation-specific (USP) primers for PCR of bisulfite-treated DNA. Two micrograms of genomic DNA were used for sodium bisulfite treatment with the EpiTect Bisulfite Kit (QIAGEN). DNA methylation analysis and primer sequences are reported in Table S1. Nuclear Run-on Assay Isolation of nuclei from 4 g protonema of the P. patens WT and two DPpDCL1b mutants and subsequent run-on transcription assays using 3 3 106 nuclei each were carried out according to Folta and Kaufman (2006). Radiolabeled RNA was hybridized to blots spotted with 3 mg of denatured gene-specific cDNA fragments of PpEF1a, PpGNT1, PpC3HDZIP1, PpHB10, PpSBP3, and PpARF and a negative control cDNA fragment of the bacterial zeocin gene (Sh ble), respectively. Blot hybridization was carried out at 65 C in 0.05 M sodium phosphate (pH 7.2), 1 mM EDTA, 63 SSC, 13 Denhardt’s, and 5% SDS. Blots were washed at 65 C twice with 23 SSC and 0.2% SDS and once with 13 SSC and 0.1% SDS. Hybridization signals were quantified with the Quantity One Software and normalized to PpEF1a. Oligonucleotides for amplification of cDNA fragments are reported in Table S1. Detection of miRNA:mRNA Duplexes CDNA was synthesized from 4 mg total RNA with Superscript III (Invitrogen) without adding primers, except a primer specific for the PpEF1a transcript to monitor cDNA synthesis efficiency. Gene-specific primers located upstream of miRNA binding sites that were used for RT-PCRs are reported in Table S1.


SUPPLEMENTAL INFORMATION Supplemental Information includes Extended Experimental Procedures, eight figures, and one table and can be found with this article online at doi:10.1016/j.cell.2009.12.023. ACKNOWLEDGMENTS This work was supported by Landesstiftung Baden-Wu¨rttemberg (P-LS-RNS/ 40 to D.W., W.F., and R.R.), the German Federal Ministry of Education and Research (FRISYS 0313921 to W.F. and R.R.), the Excellence Initiative of the German Federal and State Governments (EXC 294 to R.R.), the European Community FP6 IP SIROCCO (LSHG-CT-2006-037900 to D.W.), and the German Academic Exchange Service (DAAD to M.A.A.). We thank G. Gierga and R. Haas for technical assistance, K.M. Folta for advice on nuclear runons, and T. Laux, W.R. Hess, R. Baumeister, and P. Beyer for comments on the manuscript. Received: October 15, 2008 Revised: July 30, 2009 Accepted: November 28, 2009 Published: January 7, 2010 REFERENCES Arazi, T., Talmor-Neiman, M., Stav, R., Riese, M., Huijser, P., and Baulcombe, D.C. (2005). Cloning and characterization of micro-RNAs from moss. Plant J. 43, 837–848. Axtell, M.J., Jan, C., Rajagopalan, R., and Bartel, D.P. (2006). A two-hit trigger for siRNA biogenesis in plants. Cell 127, 565–577. Axtell, M.J., Snyder, J.A., and Bartel, D.P. (2007). Common functions for diverse small RNAs of land plants. Plant Cell 19, 1750–1769. Bao, N., Lye, K.W., and Barton, M.K. (2004). MicroRNA binding sites in Arabidopsis class III HD-ZIP mRNAs are required for methylation of the template chromosome. Dev. Cell 7, 653–662.

Baulcombe, D. (2004). RNA silencing in plants. Nature 431, 356–363. Baumberger, N., and Baulcombe, D.C. (2005). Arabidopsis ARGONAUTE1 is an RNA Slicer that selectively recruits microRNAs and short interfering RNAs. Proc. Natl. Acad. Sci. USA 102, 11928–11933. Bender, J. (2004). Chromatin-based silencing mechanisms. Curr. Opin. Plant Biol. 7, 521–526. Boyko, A., and Kovalchuk, I. (2008). Epigenetic control of plant stress response. Environ. Mol. Mutagen 49, 61–72. Bu¨hler, M., Verdel, A., and Moazed, D. (2006). Tethering RITS to a nascent transcript initiates RNAi- and heterochromatin-dependent gene silencing. Cell 125, 873–886. Cao, X., and Jacobsen, S.E. (2002). Locus-specific control of asymmetric and CpNpG methylation by the DRM and CMT3 methyltransferase genes. Proc. Natl. Acad. Sci. USA 99 (Suppl 4), 16491–16498. Carthew, R.W., and Sontheimer, E.J. (2009). Origins and Mechanisms of miRNAs and siRNAs. Cell 136, 642–655. Cokus, S.J., Feng, S., Zhang, X., Chen, Z., Merriman, B., Haudenschild, C.D., Pradhan, S., Nelson, S.F., Pellegrini, M., and Jacobsen, S.E. (2008). Shotgun bisulphite sequencing of the Arabidopsis genome reveals DNA methylation patterning. Nature 452, 215–219.

Henderson, I.R., and Jacobsen, S.E. (2008). Tandem repeats upstream of the Arabidopsis endogene SDC recruit non-CG DNA methylation and initiate siRNA spreading. Genes Dev. 22, 1597–1606. 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, 721–725. Herr, A.J., Jensen, M.B., Dalmay, T., and Baulcombe, D.C. (2005). RNA polymerase IV directs silencing of endogenous DNA. Science 308, 118–120. Howell, M.D., Fahlgren, N., Chapman, E.J., Cumbie, J.S., Sullivan, C.M., Givan, S.A., Kasschau, K.D., and Carrington, J.C. (2007). Genome-wide analysis of the RNA-DEPENDENT RNA POLYMERASE6/DICER-LIKE4 pathway in Arabidopsis reveals dependency on miRNA- and tasiRNA-directed targeting. Plant Cell 19, 926–942. Kanno, T., Huettel, B., Mette, M.F., Aufsatz, W., Jaligot, E., Daxinger, L., Kreil, D.P., Matzke, M., and Matzke, A.J. (2005). Atypical RNA polymerase subunits required for RNA-directed DNA methylation. Nat. Genet. 37, 761–765. Khraiwesh, B., Ossowski, S., Weigel, D., Reski, R., and Frank, W. (2008). Specific gene silencing by artificial MicroRNAs in Physcomitrella patens: an alternative to targeted gene knockouts. Plant Physiol. 148, 684–693.

Doi, N., Zenno, S., Ueda, R., Ohki-Hamazaki, H., Ui-Tei, K., and Saigo, K. (2003). Short-interfering-RNA-mediated gene silencing in mammalian cells requires Dicer and eIF2C translation initiation factors. Curr. Biol. 13, 41–46.

Kim, D.H., Saetrom, P., Snøve, O., Jr., and Rossi, J.J. (2008). MicroRNAdirected transcriptional gene silencing in mammalian cells. Proc. Natl. Acad. Sci. USA 105, 16230–16235.

Duchaine, T.F., Wohlschlegel, J.A., Kennedy, S., Bei, Y., Conte, D., Jr., Pang, K., Brownell, D.R., Harding, S., Mitani, S., Ruvkun, G., et al. (2006). Functional proteomics reveals the biochemical niche of C. elegans DCR-1 in multiple small-RNA-mediated pathways. Cell 124, 343–354.

Koprivova, A., Altmann, F., Gorr, G., Kopriva, S., Reski, R., and Decker, E.L. (2003). N-glycosylation in the moss Physcomitrella patens is organized similarly to that in higher plants. Plant Biol. 5, 582–591.

Dunoyer, P., Himber, C., and Voinnet, O. (2005). DICER-LIKE 4 is required for RNA interference and produces the 21-nucleotide small interfering RNA component of the plant cell-to-cell silencing signal. Nat. Genet 37, 1356–1360. Dyachenko, O.V., Zakharchenko, N.S., Shevchuk, T.V., Bohnert, H.J., Cushman, J.C., and Buryanov, Y.I. (2006). Effect of hypermethylation of CCWGG sequences in DNA of Mesembryanthemum crystallinum plants on their adaptation to salt stress. Biochemistry (Mosc.) 71, 461–465. Egener, T., Granado, J., Guitton, M.C., Hohe, A., Holtorf, H., Lucht, J.M., Rensing, S.A., Schlink, K., Schulte, J., Schween, G., et al. (2002). High frequency of phenotypic deviations in Physcomitrella patens plants transformed with a gene-disruption library. BMC Plant Biol. 2, 6. Fattash, I., Voss, B., Reski, R., Hess, W.R., and Frank, W. (2007). Evidence for the rapid expansion of microRNA-mediated regulation in early land plant evolution. BMC Plant Biol. 7, 13. Floyd, S.K., and Bowman, J.L. (2004). Gene regulation: ancient microRNA target sequences in plants. Nature 428, 485–486. Folta, K.M., and Kaufman, L.S. (2006). Isolation of Arabidopsis nuclei and measurement of gene transcription rates using nuclear run-on assays. Nat. Protoc. 1, 3094–3100. Frank, W., Decker, E.L., and Reski, R. (2005a). Molecular tools to study Physcomitrella patens. Plant Biol (Stuttg) 7, 220–227. Frank, W., Ratnadewi, D., and Reski, R. (2005b). Physcomitrella patens is highly tolerant against drought, salt and osmotic stress. Planta 220, 384–394. Gasciolli, V., Mallory, A.C., Bartel, D.P., and Vaucheret, H. (2005). Partially redundant functions of Arabidopsis DICER-like enzymes and a role for DCL4 in producing trans-acting siRNAs. Curr. Biol. 15, 1494–1500. Golden, T.A., Schauer, S.E., Lang, J.D., Pien, S., Mushegian, A.R., Grossniklaus, U., Meinke, D.W., and Ray, A. (2002). SHORT INTEGUMENTS1/SUSPENSOR1/ CARPEL FACTORY, a Dicer homolog, is a maternal effect gene required for embryo development in Arabidopsis. Plant Physiol. 130, 808–822. Gonzalez, S., Pisano, D.G., and Serrano, M. (2008). Mechanistic principles of chromatin remodeling guided by siRNAs and miRNAs. Cell Cycle 7, 2601–2608. Henderson, I.R., and Dean, C. (2004). Control of Arabidopsis flowering: the chill before the bloom. Development 131, 3829–3838.

Li, L.C., and Dahiya, R. (2002). MethPrimer: designing primers for methylation PCRs. Bioinformatics 18, 1427–1431. 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, 523–536. Liu, Q., Rand, T.A., Kalidas, S., Du, F., Kim, H.E., Smith, D.P., and Wang, X. (2003). R2D2, a bridge between the initiation and effector steps of the Drosophila RNAi pathway. Science 301, 1921–1925. Luo, Q.J., Samanta, M.P., Ko¨ksal, F., Janda, J., Galbraith, D.W., Richardson, C.R., Ou-Yang, F., and Rock, C.D. (2009). Evidence for antisense transcription associated with microRNA target mRNAs in Arabidopsis. PLoS Genet. 5, e1000457. MacRae, I.J., Ma, E., Zhou, M., Robinson, C.V., and Doudna, J.A. (2008). In vitro reconstitution of the human RISC-loading complex. Proc. Natl. Acad. Sci. USA 105, 512–517. Matzke, M., Kanno, T., Huettel, B., Daxinger, L., and Matzke, A.J. (2007). Targets of RNA-directed DNA methylation. Curr. Opin. Plant Biol. 10, 512–519. 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. Moissiard, G., Parizotto, E.A., Himber, C., and Voinnet, O. (2007). Transitivity in Arabidopsis can be primed, requires the redundant action of the antiviral Dicer-like 4 and Dicer-like 2, and is compromised by viral-encoded suppressor proteins. RNA 13, 1268–1278. Motamedi, M.R., Verdel, A., Colmenares, S.U., Gerber, S.A., Gygi, S.P., and Moazed, D. (2004). Two RNAi complexes, RITS and RDRC, physically interact and localize to noncoding centromeric RNAs. Cell 119, 789–802. 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, 613–622. Pak, J., and Fire, A. (2007). Distinct populations of primary and secondary effectors during RNAi in C. elegans. Science 315, 241–244. 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, 1484–1495.


Pham, J.W., Pellino, J.L., Lee, Y.S., Carthew, R.W., and Sontheimer, E.J. (2004). A Dicer-2-dependent 80s complex cleaves targeted mRNAs during RNAi in Drosophila. Cell 117, 83–94. Pontier, D., Yahubyan, G., Vega, D., Bulski, A., Saez-Vasquez, J., Hakimi, M.A., Lerbs-Mache, S., Colot, V., and Lagrange, T. (2005). Reinforcement of silencing at transposons and highly repeated sequences requires the concerted action of two distinct RNA polymerases IV in Arabidopsis. Genes Dev. 19, 2030–2040. Qi, Y., Denli, A.M., and Hannon, G.J. (2005). Biochemical specialization within Arabidopsis RNA silencing pathways. Mol. Cell 19, 421–428. Richardt, S., Timmerhaus, G., Lang, D., Qudeimat, E., Correˆa, L.G., Reski, R., Rensing, S.A., and Frank, W. (2010). Microarray analysis of the moss Physcomitrella patens reveals evolutionarily conserved transcriptional regulation of salt stress and abscisic acid signalling. Plant Mol. Biol 72, 27–45. Ronemus, M., Vaughn, M.W., and Martienssen, R.A. (2006). MicroRNAtargeted and small interfering RNA-mediated mRNA degradation is regulated by argonaute, dicer, and RNA-dependent RNA polymerase in Arabidopsis. Plant Cell 18, 1559–1574.

Talmor-Neiman, M., Stav, R., Klipcan, L., Buxdorf, K., Baulcombe, D.C., and Arazi, T. (2006). Identification of trans-acting siRNAs in moss and an RNA-dependent RNA polymerase required for their biogenesis. Plant J. 48, 511–521. Tomari, Y., and Zamore, P.D. (2005). Perspective: machines for RNAi. Genes Dev. 19, 517–529. Tomari, Y., Matranga, C., Haley, B., Martinez, N., and Zamore, P.D. (2004). A protein sensor for siRNA asymmetry. Science 306, 1377–1380. Vaistij, F.E., Jones, L., and Baulcombe, D.C. (2002). Spreading of RNA targeting and DNA methylation in RNA silencing requires transcription of the target gene and a putative RNA-dependent RNA polymerase. Plant Cell 14, 857–867. Varkonyi-Gasic, E., Wu, R., Wood, M., Walton, E.F., and Hellens, R.P. (2007). Protocol: a highly sensitive RT-PCR method for detection and quantification of microRNAs. Plant Methods 3, 12. Voinnet, O. (2009). Origin, biogenesis, and activity of plant microRNAs. Cell 136, 669–687.

Sijen, T., Steiner, F.A., Thijssen, K.L., and Plasterk, R.H. (2007). Secondary siRNAs result from unprimed RNA synthesis and form a distinct class. Science 315, 244–247.

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.

Strepp, R., Scholz, S., Kruse, S., Speth, V., and Reski, R. (1998). Plant nuclear gene knockout reveals a role in plastid division for the homolog of the bacterial cell division protein FtsZ, an ancestral tubulin. Proc. Natl. Acad. Sci. USA 95, 4368–4373.

Xie, Z., Allen, E., Wilken, A., and Carrington, J.C. (2005). DICER-LIKE 4 functions in trans-acting small interfering RNA biogenesis and vegetative phase change in Arabidopsis thaliana. Proc. Natl. Acad. Sci. USA 102, 12984–12989.

Tabara, H., Yigit, E., Siomi, H., and Mello, C.C. (2002). The dsRNA binding protein RDE-4 interacts with RDE-1, DCR-1, and a DExH-box helicase to direct RNAi in C. elegans. Cell 109, 861–871.

Zilberman, D., Cao, X., Johansen, L.K., Xie, Z., Carrington, J.C., and Jacobsen, S.E. (2004). Role of Arabidopsis ARGONAUTE4 in RNA-directed DNA methylation triggered by inverted repeats. Curr. Biol. 14, 1214–1220.


Supplemental Information EXTENDED EXPERIMENTAL PROCEDURES Isolation of PpDCL Full-Length cDNAs Partial cDNA sequences of P. patens DCL genes were initially identified by tblastn searches in P. patens EST sequences (Rensing et al., 2002) using A. thaliana DCL1-4 protein sequences (accession numbers P84634, Q9SP32, NP_189978, NP_566199) as queries. Corresponding cDNA clones were sequenced and used to obtain the full-length sequences. The cloning of full-length cDNA sequences was performed by 50 RACE-PCRs and RT-PCRs using primers derived from available P. patens genomic sequence data. To confirm that the amplicons were derived from the same cDNA all PCR and 50 RACE primers were selected to produce overlapping PCR fragments of already known sequence stretches. Generation and Molecular Analysis of PpDCL1a and PpDCL1b Knockout Lines For the generation of PpDCL1a and PpDCL1b knockout constructs we amplified a PpDCL1a genomic region with the primers 50 -CCAGTTGCGCATAAAGTTGA-30 and 50 -TCCAAGGCATCCAGAGAGTC-30 and a PpDCL1b cDNA region using the primers 50 -GCATTCCTGTGGAGTTTGATG-30 and 50 -ACCTTCCACACTTGGTGTGTG-30 . An nptII selection marker cassette was cloned into a single Eco72I restriction site present in the PpDCL1b cDNA fragment and into a single EcoRV restriction site of the PpDCL1a genomic fragment. The complete knockout cassettes were released from the vector prior to transformation. Primers used to identify DPpDCL1a transgenic lines were: 50 - TTATGTGGATTCAGTGCGCTTC-30 and 50 -CCATCGACTTAGCCAAACCAGT-30 . To confirm a precise 50 and 30 integration of the PpDCL1a knockout construct we used the primers 50 -TTTGCAGTTGACTGACCTCAAGA-30 and 50 -GCGGCTGAGTGGCTCCTTCA-30 (50 integration) and 50 -CCAAGGATCCCGGAAGAGGA-30 and 50 -AAATTATCGCGCGCGGTGTC-30 (30 integration). To confirm the loss of PpDCL1a transcript by RT-PCR the primers 50 TTGGTCCGTTGGAATACACA-30 and 50 - AATCTTTGTGCGCCTCTCAC-30 were used. Primers used for the screening of transgenic DPpDCL1b lines were: 50 -GCATTCCTGTGGAGTTTGATG-30 and 50 -ACCTTCCACACTTGGTGTGTG-30 . The same primers were used to confirm the loss of PpDCL1b transcript by RT-PCR. A second primer pair upstream of the integration site was used for RT-PCR: 50 -AGGATTGTTACTGCGGTGCA-30 and 50 -AAGCTCTGCACGCTCATAGC-30 . To confirm a precise 50 and 30 integration of the PpDCL1b knockout construct we used the primers 50 -TGCTACTCACTTCATGAACTG-30 and 50 -ACGTGACTCCCTTAATTCTCC-30 (50 integration) and 50 -CCCGCAATTATACATTTAATACG-30 and 50 -GCACCATGGCTGCAACAAAG-30 (30 integration). RT-PCR control primers for the PpEF1a control gene are reported Table S1. P. patens Lines Expressing an Artificial miRNA (amiRNA) Targeting PpGNT1 The amiR-GNT1 sequence was introduced into the A. thaliana miRNA319a precursor by overlapping PCR as described (Khraiwesh et al., 2008). The resulting construct was used for transfections of P. patens wild-type and DPpDCL1b mutant lines. To identify transgenic lines harboring the amiR-GNT1 expression construct PCR was performed using the primers 50 -TGATATCTCCACTGACGAAAGGG-30 and 50 -GGATCCCCCCATGGCGATGCCTTAAAT-30 . Detection of RNA Cleavage Products Synthesis of 50 RACE-ready cDNAs was carried out according to Zhu et al. (2001) with the BD Smart RACE cDNA Amplification Kit (Clontech). PCR reactions were performed using the UPM Primer-Mix in combination with gene-specific primers (reported in Table S1) derived from target RNAs. Cleavage products were excised from the gel, cloned and sequenced. Detection of Small RNAs by RNA Gel Blots Total RNA was separated in a 12% denaturing polyacrylamide gel containing 8.3 M urea in TBE buffer. The RNA was electroblotted onto nylon membranes for 1 hr at 400 mA. Radiolabeled probes were generated by end-labeling of DNA oligonucleotides complementary to miRNA, siRNA and ta-siRNA sequences and the U6snRNA control (listed in Table S1) with g32P-ATP using T4 polynucleotide kinase. Blot hybridization was carried out in 0.05 M sodium phosphate (pH 7.2), 1 mM EDTA, 6 x SSC, 1 x Denhardt’s, 5% SDS. Blots were washed 2-3 times with 2 x SSC, 0.2% SDS and one time with 1 x SSC, 0.1% SDS. Blots were hybridized and washed at temperatures 5 C below the Tm of the oligonucleotide. Sequences of oligonucleotides for the detection of small RNAs are listed in Table S1. Detection of Small RNAs by RT-PCR The RT-PCR analyses of the miR156, 160, 166, 390, and the ta-siRNAs pptA079444 (processed from the PpTAS1 gene) and pptA013298 (processed from the PpTAS3 gene) were carried out as described (Varkony-Gasic et al., 2007). Sequences of primers used for cDNA synthesis and subsequent PCR are reported in Table S1. Expression Analysis by Semiquantitative RT-PCR Semiquantitative RT-PCR with gene-specific primers for PpEF1a, PpSBP3, PpARF, PpC3HDZIP1, PpHB10 and PpTAS1 was performed from three independent biological replicates of the two DPpDCL1a mutants. PCR cycles were optimized for each gene to be in the exponential range not reaching the plateau phase. The optimum number of cycles for PpEF1a, PpSBP3, PpC3HDZIP1, PpARF,

Cell 140, 111–122, January 8, 2010 ª2010 Elsevier Inc. S1

PpHB10 and PpTAS1 were 23, 24, 28, 24, 26 and 24, respectively. PCR products were quantified with the Quantity One Software (Bio-Rad) and transcript levels were normalized to the constitutive control PpEF1a. Primer sequences are reported in Table S1. Expression Analysis by Quantitative Real Time PCR RNA samples from three biological replicates of P. patens wild-type and two DPpDCL1b mutant lines were treated with DNase I (Fermentas) and reverse-transcribed into first-strand cDNA by Taq Man Reverse Transcription Reagents (Applied Biosystems). Real time PCR was performed on Roche 480 Light Cycler using gene-specific primers and Light Cycler 480 SYBR Green I Master (Roche) according to the manufacturer’s instructions. Constitutively expressed PpEF1a was used as reference gene for normalization. Expression levels of PpC3HDZIP1, PpHB10, PpSBP3 and PpARF were calculated relative to transcript abundance in wild-type employing relative quantification with efficiency correction (Livak and Schmittgen, 2001). Primer sequences are reported in Table S1. RNA Gel Blots 20 mg of total RNA isolated from wild-type, DPpDCL1b mutants and transgenic lines expressing the amiR-GNT1, respectively, were separated in denaturing agarose gels and blotted onto nylon membranes. Hybridization probes for PpARF, PpC3HDZIP1, PpHB10, PpSBP3, PpGNT1, PpEF1a, PpTAS4 PpEREBP/AP2, PpbHLH, and PpCOR47 were amplified from wild-type cDNA (primers reported in Table S1). Hybridization signals were quantified with the Quantity One Software and normalized to the PpEF1a control. The ABA-responsive gene PpCOR47 (Frank et al., 2005b) was used to control the efficiency of the ABA treatments. DNA Methylation Analysis The cDNA sequences of PpC3HDZIP1 (DQ385516), PpHB10 (AB032182), PpARF (AR452951), PpSBP3 (AJ968318) and PpGNT1 (AJ429143) were used for BLASTN searches to identify corresponding genomic sequences from the P. patens whole-genomeshotgun (WGS) traces (accessible via www.ncbi.nlm.nih.gov/Traces/trace.cgi). The identified genomic sequences were clustered and assembled using the Paracel Transcript Assembler to determine the genomic exon/intron structure (Figure S5). The parameters for clustering threshold, overlap length and overlap identity were 100 nt, 80 nt and 90%, respectively. Primers to analyze the PpTAS4 genomic locus were derived from the reported PpTAS4 sequence (Talmor-Neiman et al., 2006). Primers for the analysis of the PpEREBP/AP2 and PpbHLH gene were derived from the corresponding gene model of the available P. patens genomic sequence (http:// genome.jgi-psf.org/Phypa1_1/Phypa1_1.home.html; gene model accession numbers Phypa1_129196 [PpEREBP/AP2] and Phypa1_209063 [PpbHLH]). The derived promoter, exon and intron regions were analyzed with the MethPrimer program (Li and Dahiya, 2002) to deduce methylation-specific (MSP) and unmethylation-specific primers (USP) for PCR analysis of bisulfite-treated DNA. Primer sequences are reported in Table S1. Detection of Sense and Antisense Transcripts cDNA from wild-type plants and DPpDCL1b mutants was synthesized from 4 mg total RNA with Superscript III (Invitrogen) using primers specific for sense and antisense transcripts, respectively. To monitor the efficiency of cDNA synthesis, primers specific for the PpEF1a sense transcript were added to each cDNA synthesis reaction. RT-PCRs were carried out with gene-specific primers. Primer sequences are reported in Table S1.

SUPPLEMENTAL REFERENCES Livak, K.J., and Schmittgen, T.D. (2001). Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25, 402–408. Rensing, S.A., Rombauts, S., Van de Peer, Y., and Reski, R. (2002). Moss transcriptome and beyond. Trends Plant Sci. 7, 535–538. Zhu, Y.Y., Machleder, E.M., Chenchik, A., Li, R., and Siebert, P.D. (2001). Reverse transcriptase template switching: a SMART approach for full-length cDNA library construction. Biotechniques 30, 892–897.

S2 Cell 140, 111–122, January 8, 2010 ª2010 Elsevier Inc.

A DEAD

HEL

DUF

B

PAZ

nosP

nptII

RNAse III

dsrm

nosT

EcoRV PpDCL1a gDNA DCL1 cDNA PpDCL1a nosP F2 PpDCL1a

nptII

nosT PpDCL1a

nptII

nosT PpDCL1a PpDCL1a

F1

F3

PpDCL1a nosP R2

C

*

*

D WT

R1

PpDCL1a mutants 1 2

WT Genomic PCR-Screen

E PpDCL1a PpEF1

R3

PpDCL1a mutants 1 2 WT 5’ integration 3’ integration

Figure S1. Generation of DPpDCL1a Mutants, Related to Figure 1 (A) Predicted domain structure of the P. patens DCL1a protein. DEAD: DEAD box helicase, HEL: helicase C, DUF: domain of unknown function, PAZ: PAZ domain, RNaseIII: ribonuclease III domain, dsrm: double-stranded RNA binding motif. (B) Scheme illustrating the generation of the DPpDCL1a mutants. The nptII cassette was cloned into a single EcoRV site of a PpDCL1a genomic DNA fragment (gDNA). Middle: Resulting PpDCL1a knockout construct. Bottom: Expected genomic structure of the PpDCL1a locus after integration of the PpDCL1a knockout construct by homologous recombination. Primers used for molecular analyses of the transgenic lines are indicated by arrows. White box: nptII cassette; gray boxes: PpDCL1a gDNA fragment; black boxes: genomic PpDCL1a locus. (C) PCR analysis of transgenic lines using genomic DNA performed with primers F1 and R1. Transgenic lines which failed to give rise to a PCR product are marked by asterisks. WT: wild-type control; the remaining samples were derived from transgenic lines. (D) Analysis of PpDCL1a mRNA expression. Top: RT-PCR studies from two DPpDCL1a mutants and wild-type (WT) with primers F1 and R1. Bottom: RT-PCR performed with PpEF1a control primers. (E) PCR analysis of DPpDCL1a mutants using genomic DNA to confirm 50 and 30 integration of the PpDCL1a knockout construct. PCR products obtained from PCR reactions with primers F2 and R2 (50 integration), F3 and R3 (30 integration). Sequences of primers used for the molecular analyses are reported in the Extended Experimental Procedures and Table S1.


A

HEL DUF

DEAD

RNAse III

PAZ

D

dsrm

G

PpDCL1b mutants 1

3

2

4

WT

WT

PpDCL1a mutant

PpDCL1b mutant

PpDCL1b PpEF1

B

nosP

nptII Eco72I

4550

E

5108

nptII

PpDCL1b mutants 1

PpDCL1b cDNA DCL1 cDNA PpDCL1b nosP

Standard growth medium

nosT

3

2

4

Supplemented medium (+ vitamins and glucose)

WT PpDCL1b PpEF1

nosT PpDCL1b

4 weeks F2 F1

F3

F4

PpDCL1b nosP

PpDCL1b

nptII

nosT PpDCL1b PpDCL1b

R4

R3

F

PpDCL1b mutants 1

R1 R2

2

3

Standard growth medium

4 5’ integration

C

WT

*

H Wild type PpDCL1a mutant PpDCL1b mutant

I

** *

Standard medium

Supplemented medium (+ vitamins and gulcose)

+++

7 weeks Gametophores

+++

+++

+++

+

Abnormalities in cell size, cell shape and polarity

-

++

Abnormalities in cell size, cell shape and polarity

Malformed gametophores

++

PpDCL1b mutants WT 1 2 3 4

Supplemented medium (+ vitamins and glucose)

Control PpEF1

Protonema

-

PpDCL1b mutants SM WT 1 2 3 4

3’ integration

Genomic PCR-Screen


100806050-

SM WT

1

PpDCL1b mutants 3 4 2

SM

40-

40-

30-

30-

WT

PpDCL1b mutants 3 4 1 2

SM


70605040-

4020-

30-

10-

10-

miR156

20-

30-

20-

miR166

2010-

10-

miR390

miR535

miR538

miR160

Figure S2. Generation of DPpDCL1b Mutants and Phenotypic Comparison of DPpDCL1a and DPpDCL1b Mutants, Related to Figure 1 and Figure 2 (A) Predicted domain structure of the P. patens DCL1b protein. DEAD: DEAD box helicase, HEL: helicase C, DUF: domain of unknown function, PAZ: PAZ domain, RNaseIII: ribonuclease III domain, dsrm: double-stranded RNA binding motif. (B) Scheme illustrating the generation of the DPpDCL1b mutants. The nptII cassette was cloned into a single Eco72I site of a PpDCL1b cDNA fragment. Numbers indicate nucleotide positions in the PpDCL1b cDNA. Middle: Resulting PpDCL1b knockout construct. Bottom: Expected genomic structure of the PpDCL1b locus after integration of the PpDCL1b knockout construct by homologous recombination. Primers used for molecular analyses of the transgenic lines are indicated by arrows. White box: nptII cassette; gray boxes: PpDCL1b cDNA fragment; black boxes: genomic PpDCL1b locus. (C) PCR analysis of transgenic lines using genomic DNA performed with primers F1 and R1. Transgenic lines which failed to give rise to a PCR product are marked by asterisks. WT: wild-type control; the remaining samples were derived from transgenic lines. (D) Analysis of PpDCL1b mRNA expression. Top: RT-PCR studies from four DPpDCL1b mutants and wild-type (WT) with primers F2 and R2. Bottom: RT-PCR performed with PpEF1a control primers. (E) Top: RT-PCR analysis of PpDCL1b expression in DPpDCL1b mutants and wild-type using primers F3 and R3 located upstream of the knockout construct integration site. Bottom: RT-PCR performed with PpEF1a control primers. (F) PCR analysis of DPpDCL1b mutants using genomic DNA to confirm 50 and 30 integration of the PpDCL1b knockout construct. PCR products obtained from PCR reactions with primers F2 and R4 (50 integration), F4 and R2 (30 integration) and PpEF1a control primers to confirm integrity of the used genomic DNA. Sequences of primers used for the molecular analyses are reported in the Extended Experimental Procedures and Table S1. (G) Filamentous protonema tissue from P. patens wild-type, DPpDCL1a and DPpDCL1b mutants was spotted in equal densities onto standard growth medium and medium supplemented with vitamins and glucose. Pictures of developing moss colonies were taken 4 weeks and 7 weeks after inoculation. Note that, in contrast to DPpDCL1b mutants, DPpDCL1a mutants are not able to propagate on standard growth medium. Furthermore, DPpDCL1a and DPpDCL1b mutants show distinct deviations in developmental progression. (H) Overview of phenotypic deviations of DPpDCL1a and DPpDCL1b mutants. (I) Full size RNA gel blots with corresponding RNA size markers (SM) shown in Figures 2C–2E.


PpDCL1b mutant 1

A WT

PpDCL1b mutant 2

PpDCL1b mutant 3

PpDCL1b mutant 4

4d

6d

8d

18 d

8 weeks

B

200 µm

100 µm

100 µm

90 µm

80 µm

Figure S3. Phenotypic Analysis of DPpDCL1b Mutants, Related to Figure 2 (A) Regeneration of protoplasts from wild-type plants (WT) and DPpDCL1b mutants was monitored at the indicated time points. Size bars 4 d, 6 d, and 8 d: 100 mm; 18 d, 8 weeks: 500 mm. (B) Electron micrographs of gametophores from wild-type and DPpDCL1b mutants.


Upstream

Downstream

SM

Antisense KO KO WT 1 2

Sense KO KO WT 1 2

Antisense KO KO WT 1 2

Sense KO KO WT 1 2

10080504030-

PpARF

2010-

90 70

1008060504030-

2010-

Figure S4. Related to Figure 3 Full size RNA gel blots with corresponding RNA size markers (SM) shown in Figure 3D.


PpC3HDZIP1

Relative expression level

A 1,2 1 0,8 0,6 0,4 0,2 0 PpHB10

PpC3HDZIP1

B 1

1315-1317

Intron 1

Intron 2

1893-2105

2731-2860

3959-3961

I2

I4

3041-3247

3673-3984

3000

I6

I8

I 10

4489-4641 5007-5135 5553-5735

4000

I 12

I 14

I 16

6208-6446

7039-7300

7807-8039

2329-2331

Stop codon 8181-8183

Promoter Region

8183 miRNA BS 3662-3672

2000 I1 2504-2839

Start codon 1 - 999

PpHB10 1

Stop codon

Intron 3 3749-3907

4194 2000

Start codon 1 - 1000

1

miRNA BS 3115-3135

Promoter Region 1000

PpC3 HDZIP1

PpSBP3

PpDCL1b KO2

Start codon 1 - 999

PpARF

PpARF

PpDCL1b KO1

WT

1393-1395

4000

I 3 3985-3894 3351-3581

I5

6000 I7

I9

8000

I 11

4193-4402 4796-4929 5243-5363 5919-6130

I 13

I 15

6588-6726

7407-7658

Stop codon

miRNA BS 2072-2092

4033-4035

4590


2000

3000

4000

Intron


PpSBP3 1

3280-3282

3282 1000

2000


PpGNT1 1

Stop codon

miRNA BS 2687-2706

1300-1302


Stop codon

1300-1302

2713-2715

Promoter Region

2715 1000

2000

Figure S5. Expression Analysis and Gene Models of PpARF, PpC3HDZIP1, PpHB10, PpSBP3 and PpGNT1, Related to Figure 4 (A) Expression analysis of miRNA target genes by quantitative real time PCR in wild-type (WT) and two DPpDCL1b mutant lines. Expression in wild-type was set to one, error bars indicate standard errors (n = 3). (B) For PpARF and PpC3HDZIP1, complete gene models including intron sequences were predicted by comparing available cDNA sequences with the P. patens genomic trace files. For PpHB10, PpSBP3 and PpGNT1, promoter regions were analyzed following the same strategy. For PpHB10 the 50 untranslated region was included based on the available full-length cDNA sequence, the intron within the miRNA binding site is depicted by an arrow. In the case of PpSBP3 and PpGNT1, cDNA sequences encompassing the open reading frame were available. Here, regions lying 300 nucleotides upstream of the start codon were used for promoter analysis. I 1 – I 16: Intron 1 to Intron 16; miRNA BS: miRNA binding site. The genomic nucleotide sequences were deposited in GenBank with the following accession numbers: BK006047 (PpHB10), BK006048 (PpC3HDZIP1), BK006049 (PpGNT1), BK006050 (PpSBP3) and BK006051 (PpARF).


A

WT KO1+2 BT

GGACAATCTTTGGATATGTGCCAGCGTATCTTGTGATCGTGGTTCTTAAGGGTCGAGTGCTTAGCTCCTCATCCTCATGCTTAGGTCTGGAAATATGTAAAAGGG GGATAATTTTTGGATATGTGTTAGCGTATTTTGTGATTGTGGTTTTTAAGGGTCGAGTGTTTAGTTTTTTATTTTTATGTTTAGGTTTGGAAATATGTAAAAGGG 10/10 10/10

WT KO1+2 BT

GACGTAATGACAACACGAAGCTTATAAAAACTCAAAGCT GACGTAATGATAATACGAAGTTTATAAAAATTTAAAGTT 9/10 10/10

B

WT KO1+2 BT

TTTCTGGCTGATACAGAAACAGACGAGGTATTTGCTCGTATTTGCCTGCAGCCTGAGATTGGCTCCTCCGCTCAGGATTTAACAGATGATTCTCTTGCGTCTCCG TTTTTGGTTGATATAGAAATAGACGAGGTATTTGTTTGTATTTGTTTGTAGTTTGAGATTGGTTTTTTCGTTTAGGATTTAATAGATGATTTTTTTGCGTTTTTG 10/10 10/10 10/10

WT KO1+2 BT

CCTCTAGAGAAACCAGCTTCATTTGCCAAAACGCTCACTCAAAGTGATGCAAACAACGGTGGAGGCTTTTCAATACCTC TTTTTAGAGAAATTAGTTTTATTTGTTAAAATGTTTATTTAAAGTGATGTAAATAACGGTGGAGGTTTTTTAATATTTC 10/10 10/10

C

WT KO1+2 BT

TAGCTCATAACCCTCTCACAGGACGTAATGGGGGTGACAACATGCTAACAGAATTGCACGGTAAAGGAAAACTGTACTAGGCATGTTATATGGGAATTCGGATCG TAGTTTATAATTTTTTTATAGGACGTAATGGGGGTGATAATATGTTAATAGAATTGTACGGTAAAGGAAAATTGTATTAGGTATGTTATATGGGAATTCGGATCG 9/10 10/10 1/10 10/10

WT KO1+2 BT

CTTCTTGCAATTAAACACGCTAGCGCCGTTTGGTGCCAATGTTATTCTGG TTTTTTGTAATTAAATACGTTAGTGTCGTTTGGTGTTAATGTTATTTTGG 10/10 10/10

D

WT KO1+2 BT

TCAAAGCAATCAATGTGCTAATGAACGTGCTGCCTGCATTGTCTGCAATGACTATGTACTGCGTAGTCAGTGGAAATAGGTGGTTGATTATGAGATTTTGGTTGT TTAAAGTAATTAATGTGTTAATGAACGTGTTGTTTGTATTGTTTGTAATGATTATGTATTGCGTAGTTAGTGGAAATAGGTGGTTGATTATGAGATTTTGGTTGT 9/10 10/10

WT KO1+2 BT

GCAGGTCACTTGGGATGAGCCGGACCTATTGCAGGGAGTGAATCGTGTAAGCCCATGGCAGTTAGAGCTTGTGGCGACACTTCCTATGCAGCTGCCC GTAGGTTATTTGGGATGAGTCGGATTTATTGTAGGGAGTGAATTGTGTAAGTTTATGGTAGTTAGAGTTTGTGGCGATATTTTTTATGTAGTTGTTT 10/10 9/10

E

WT KO1+2 BT

TTCACAATGTGCTCTTCAAGCTTCGTTGTATGCTAAACTCTACTGCAATACTGCTACGGCGTGCCTTTTCTTTTTTCGAAGTATGATATGATGACCTAATGGTCT TTTATAATGTGTTTTTTAAGTTTCGTTGTATGTTAAATTTTATTGTAATATTGTTACGGCGTGTTTTTTTTTTTTTCGAAGTATGATATGATGATTTAATGGTTT 9/10 10/10 10/10 1/10

WT KO1+2 BT

TTTTGAATACGACAGGAATTCCATGGAGACAG TTTTGAATACGATAGGAATTTTATGGAGATAG 10/10

F

Figure S6. DNA Methylation Analysis of Promoter and Intragenic Regions of the PpARF Gene in P. patens Wild-Type and Two DPpDCL1b Mutants, and Nuclear Run-on Assay, Related to Figure 4 Nucleotide sequences of PCR products obtained from methylation-specific PCRs are aligned. (A) Promoter region. (B) Exon 1 upstream of the miR160 binding site. (C) Exon 4 downstream of the miR160 binding site. (D) Intron 2 upstream of the miR160 binding site. (E) Intron 3 downstream of the miR160 binding site. WT: Wild-type nucleotide sequence of the analyzed region; CpG residues are highlighted in green. KO1+2 BT: Sequences of PCR products obtained with MSP primers from bisulfite-treated DNA from two DPpDCL1b mutants. Cytosine residues of CpG dinucleotides which are methlyated in the DPpDCL1b mutants are indicated in red. Cytosine residues of CpG dinucleotides which are not methylated are highlighted in yellow. Cytosine to thymine conversions are highlighted in bold and are underlined. Five independent clones from each DPpDCL1b mutant line were sequenced. The numbers below methylated cytosine residues (in red) indicate the frequency of detected methylation in the sequenced clones. (F) Membranes with dot-blotted cDNA fragments of the miRNA target genes PpC3HDZIP1, PpHB10, PpSBP3, PpARF, the control genes PpEF1a, PpGNT1, and the negative control bacterial zeocin resistance gene Sh ble. The membranes were hybridized with radiolabeled RNA obtained from nuclear run-on assays from P. patens wild-type and two DPpDCL1b mutant lines.


Antisense SM

WT

KO KO 1 2

Sense

Antisense

KO KO WT 1 2

KO 1

WT

KO 2

Sense WT

KO KO 1 2

1009080706050-

40-

30-

20-

10-

PpTAS4

PpEREBP/AP2

Figure S7. Related to Figure 5 Full size RNA gel blots with corresponding RNA size markers (SM) shown in Figure 5B.


A

WT TTTATCTCTAAATTCTTAGACAACGTCATTCAAAATAAGTTTTAAAACAGCGACTAGTCATAAAATACGTATTTACACACTTGTATATGATGTACCATAGACG WT1+amiRNA BT TTTATTTTTAAATTTTTAGATAATGTTATTTAAAATAAGTTTTAAAATAGTGATTAGTTATAAAATATGTATTTATATATTTGTATATGATGTATTATAGATG WT2+amiRNA BT TTTATTTTTAAATTTTTAGATAACGTTATTTAAAATAAGTTTTAAAATAGCGATTAGTTATAAAATACGTATTTATATATTTGTATATGATGTATTATAGACG 4/5 KO1+amiRNA BT TTTATTTTTAAATTTTTAGATAACGTTATTTAAAATAAGTTTTAAAATAGCGATTAGTTATAAAATACGTATTTATATATTTGTATATGATGTATTATAGACG KO2+amiRNA BT TTTATTTTTAAATTTTTAGATAACGTTATTTAAAATAAGTTTTAAAATAGCGATTAGTTATAAAATACGTATTTATATATTTGTATATGATGTATTATAGACG 4/5 WT GTAACCGTACATATTTGCCGACACCCTGCAATTAATAGAGTTCGAATATCCCCGCCGCGTTCAAGTCGCCT WT1+amiRNA BT GTAATTGTATATATTTGTTGACATTTTGTAATTAATAGAGTTTGAATATTTTTGTTGTGTTTAAGTTGTTT WT2+amiRNA BT GTAATTGTATATATTTGTTGACATTTTGTAATTAATAGAGTTCGAATATTTTCGTCGCGTTTAAGTCGTTT 4/5 KO1+amiRNA BT GTAATCGTATATATTTGTTGACATTTTGTAATTAATAGAGTTTGAATATTTTCGTCGCGTTTAAGTCGTTT 1/5 4/5 KO2+amiRNA BT GTAATTGTATATATTTGTCGACATTTTGTAATTAATAGAGTTTGAATATTTTCGTCGCGTTTAAGTCGTTT 3/5 4/5

B

WT WT/Con. BT WT/ABA BT WT WT/Con. BT WT/ABA BT

C

WT WT/Con. BT WT/ABA BT WT WT/Con. BT WT/ABA BT WT WT/Con. BT WT/ABA BT

ATCTTTCAAATTCCGCTTCCCTCGCACATGACTAAATCTAACATAATTTCTAAACTGCTGATTTTGTCCCATCGGTGTGTCAGGAAAGACTTCGACTCGTCAG ATTTTTTAAATTTTGTTTTTTTTGTATATGATTAAATTTAATATAATTTTTAAATTGTTGATTTTGTTTTATTGGTGTGTTAGGAAAGATTTTGATTTGTTAG ATTTTTTAAATTTCGTTTTTTTCGTATATGATTAAATTTAATATAATTTTTAAATTGTTGATTTTGTTTTATCGGTGTGTTAGGAAAGATTTCGATTCGTTAG 1/5 4/5 CTGTAACTTCAGTTTCGAAATCCCAGCTTGGACAGAACTTCGTTTTATCTAGACGGAGGTCACCAGGACTGGTAAC TTGTAATTTTAGTTTTGAAATTTTAGTTTGGATAGAATTTTGTTTTATTTAGATGGAGGTTATTAGGATTGGTAAT TTGTAATTTTAGTTTCGAAATTTTAGTTTGGATAGAATTTCGTTTTATTTAGACGGAGGTTATTAGGATTGGTAAT 1/5 4/5 GCGAAAAACCTCATGGCTGAGCGCAGGCGCCGCAAAAAACTCAACGATCGCCTGTACACGCTACGGTCTGTAGTTCCTAAGATTACAAAGGTGCTTCCAAACT GTGAAAAATTTTATGGTTGAGTGTAGGTGTTGTAAAAAATTTAATGATTGTTTGTATATGTTATGGTTTGTAGTTTTTAAGATTATAAAGGTGTTTTTAAATT GCGAAAAATTTTATGGTTGAGCGTAGGCGTCGTAAAAAATTTAATGATCGTTTGTATACGTTACGGTTTGTAGTTTTTAAGATTATAAAGGTGTTTTTAAATT 1/5 4/5 CTATCTTTGAACATGTTGCCCGCCTCGATTGCTGAATTGCACATCATGTATGTTGAGATGTCCACTTACGAATCAGTGGGGTGTGGAGTACAGATGGATAGAG TTATTTTTGAATATGTTGTTTGTTTTGATTGTTGAATTGTATATTATGTATGTTGAGATGTTTATTTATGAATTAGTGGGGTGTGGAGTATAGATGGATAGAG TTATTTTTGAATATGTTGTTCGTTTCGATTGTTGAATTGTATATTATGTATGTTGAGATGTTTATTTACGAATTAGTGGGGTGTGGAGTATAGATGGATAGAG 1/5 4/5 CCTCCATATTGGGGGATGCGATTGAGTACCTAAAGGAGCTCCTGCAACGCATCAATGAAATCCATAACGAACTGGAAGCAGCAAAGCTGGA TTTTTATATTGGGGGATGTGATTGAGTATTTAAAGGAGTTTTTGTAATGTATTAATGAAATTTATAATGAATTGGAAGTAGTAAAGTTGGA TTTTTATATTGGGGGATGCGATTGAGTATTTAAAGGAGTTTTTGTAACGTATTAATGAAATTTATAACGAATTGGAAGTAGTAAAGTTGGA 1/5

D

Figure S8. DNA Methylation Analysis and Full Size RNA Gel Blots with Corresponding RNA Size Markers, Related to Figure 6 (A) DNA methylation analysis of the PpGNT1 promoter region in lines expressing the amiR-GNT1. Nucleotide sequences of PCR products obtained from methylation-specific PCRs are aligned. WT: Wild-type nucleotide sequence of the analyzed region; CpG residues are highlighted in green. WT1 + amiRNA BT: Sequences of PCR products obtained with USP primers from bisulfite-treated DNA from wild-type line 1 expressing the amiR-GNT1 at low levels. WT2 + amiRNA BT: Sequences of PCR products obtained with MSP primers from bisulfite-treated DNA from wild-type line 2 expressing the amiR-GNT1 at high levels. KO1 + amiRNA BT and KO2 + amiRNA BT: Sequences of PCR products obtained with MSP primers from bisulfite-treated DNA from DPpDCL1b mutants expressing the amiR-GNT1. Cytosine residues of CpG dinucleotides which are methlyated are indicated in red. Cytosine residues of CpG dinucleotides which are not methylated are highlighted in yellow. Cytosine to thymine conversions are highlighted in bold and are underlined. Five independent clones from each PCR product were sequenced. The numbers below methylated cytosine residues (in red) indicate the frequency of detected methylation in the sequenced clones. No numbers are presented when sequences of independent clones were identical. (B) and (C) DNA methylation analysis of promoter and intragenic regions of PpbHLH in untreated and ABA-treated P. patens wild-type. Nucleotide sequences of PCR products obtained from methylation-specific PCRs are aligned. (A) Promoter region of PpbHLH. (B) Coding Sequence of PpbHLH (intron sequences are marked in blue). WT: Wild-type nucleotide sequence of the analyzed region. WT/Con. BT: Sequences of PCR products obtained with USP primers from bisulfite-treated DNA from untreated wild-type. WT/ABA BT: Sequences of PCR products obtained with MSP primers from bisulfite-treated DNA from ABA-treated wild-type. Five independent clones from each PCR product were sequenced. Color code and labeling as in (A). (D) RNA gel blots with corresponding RNA size markers (SM) shown in Figure 6B and 6H.
