Main

Non-coding small RNAs in the nucleus have been proposed to provide a sequence-specific interface between a DNA sequence and its epigenetic state, presumably by their base-pairing with genomic DNA or nascent RNA6,7. Recent studies in the fission yeast have indicated that RNA-interference (RNAi)-mediated heterochromatin assembly occurs by means of a self-enforcing loop mechanism8,9,10. A central player of this loop is the RITS (RNAi-induced initiation of transcriptional gene silencing) complex, which contains a chromodomain-containing protein (Chp1), Argonaute 1 (Ago1), Tas3, and siRNAs9,10,11,12. Ago1 confers the sequence specificity by binding to siRNAs and recruits other chromatin proteins to initiate the heterochromatization9. Despite this progress, the role of non-coding small RNAs in epigenetic regulation in higher organisms remains largely unexplored.

To examine this role, we focus on Piwi and its interacting piRNAs in Drosophila. Piwi is an Ago/Piwi protein that was initially identified to be essential for stem-cell self-renewal13. Subsequently, it was implicated in heterochromatin formation, transposon silencing, and clustering of multiple copies of transgenes through the RNAi-mediated pathway1,2,3,14,15,16. Piwi interacts with piRNAs3,4,5, bears RNA cleavage activity4, and may participate in an ‘amplification cycle’ that accelerates piRNA biogenesis5,17.

To identify Piwi-interacting piRNAs systematically, we conducted immunoprecipitation to purify the Myc-Piwi complex from ovaries of adult flies carrying a fully functional myc–piwi transgene (Supplementary Fig. 1a)18. Small RNAs ranging from 18 to 32 nucleotides (nt) in length were specifically precipitated with the Myc–Piwi complex (Fig. 1a, in which 24–26-nt RNAs are visible). We recovered 19,048 candidate small RNA clones with perfect matches in the Drosophila melanogaster genome, which represent 13,299 unique Piwi-associated small RNAs. Of these, about 8.7% match known non-coding RNAs (Supplementary Fig. 1b). The remaining 12,903 small RNAs are Piwi-associated piRNAs, which show a gaussian distribution in size and have a peak at 24–26 nt (Fig. 1b). Of these, 55.2% contain U as the first 5′ nucleotide, a bias similar to that in mammalian piRNAs, whereas the second 5′ nucleotide shows a strong bias against U (Fig. 1c). A total of 10,792 piRNAs can be mapped to the assembled genome (Fig. 1d and Supplementary Figs 1c–4). Of these, 7,651 (59.3%) are mapped to transposons (9.7% of the assembled genome), especially LTR (long terminal repeat) and LINE (long interspersed nuclear element) types of retrotransposon (Fig. 1d). In contrast, Piwi-associated piRNAs are underrepresented in gene-coding sequences, intergenic regions, and simple or low-complexity repeats. Along chromosomes, Piwi-associated piRNAs are highly enriched in pericentromeric regions and subtelomeric regions, which contain a high density of transposons (Supplementary Figs 1c–4). These results are consistent with recent findings that Piwi subfamily proteins bind to transposon-derived piRNAs3,4,5,17 and echo the role of Piwi in the epigenetic regulation of transposons and tandem transgenes1,2,3,16.

Figure 1: Identification and characterization of Piwi-associated piRNAs.
figure 1

a, Small RNAs of 24–26 nt (arrow) and about 80 nt (open arrowhead) are specifically precipitated together with Myc–Piwi. IP, immunoprecipitation. b, Size distribution of Piwi-associated piRNAs. c, Nucleotide preference at 5′ and 3′ ends of Piwi-associated piRNAs. d, Percentages of Piwi-associated piRNAs in various annotated genomic regions. Orange, expected number; blue, cloned number. CDS, protein coding sequence; UTR, untranslated region.

Full size image

To investigate the potential function of Piwi and its interacting piRNAs in epigenetic regulation, we focused on the 20-nt 3R-TAS1 piRNA, which is uniquely mapped to 3R-TAS (Fig. 2a). We chose this piRNA for five reasons: first, TAS is genetically well characterized heterochromatin crucial for telomere function and genome integrity19,20; second, TASs share structural similarity between eukaryotes, implicating their functional conservation21; third, 3R-TAS is composed of tandem repeats highly homologous to LTR regions of the Invader4 retrotransposon (Supplementary Fig. 5), providing an opportunity to study the role of retrotransposon-originated piRNAs; fourth, 3R-TAS1 piRNA is uniquely mapped to 3R-TAS (Supplementary Fig. 5), which allows us to establish a one-to-one functional relationship between a specific piRNA and its genomic sequence; and last, we previously identified a P-element insertional mutation, P{w+,ry+}A4-4, inserted 128 base pairs (bp) downstream of the 3R-TAS1 piRNA coding sequence (Fig. 2a), as the strongest suppressor of piwi for its germline stem-cell phenotype22. One copy of the P{w+,ry+}A4-4 allele restores germline stem cells in about 70% of the homozygous piwi2 mutant females22. This suggests an important role for 3R-TAS in germline stem-cell maintenance that probably resides in niche cells where Piwi functions18.

Figure 2: The expression of 3R-TAS1 piRNA is Piwi-dependent and correlates with germline stem-cell maintenance.
figure 2

a, The organization of 3R-TAS. The black arrow indicates the position and direction of the unique mapped 3R-TAS1 piRNA. 100F is the polytene band name; this band contains the chromosome 3 right-arm TAS and telomere. b, RNase protection assay showing that the 3R-TAS1 piRNA can protect a 597-bp antisense probe that spans the piRNA coding region and covers 386–982 nt of the 984-bp repeat unit in the antisense direction. The 3R-TAS1 piRNA is relatively enriched in nuclear extracts (NE) over whole-cell extracts (WCE). M, 10-bp DNA marker. c, RNase protection assay showing that 3R-TAS1 piRNA is expressed in the wild-type adult fly, in both ovarian and extra-ovarian cells. Its expression is markedly decreased in piwi1 and piwi2 mutants and is drastically increased by the P{w+,ry+}A4-4 insertion (four separately maintained lines, namely 221, 516, R86-2 and R86-2-3, were checked). All lanes were loaded with equal amounts of total RNA. The copy number of piwi and P{w+,ry+}A4-4 is noted at the bottom for c and d. d, P{w+,ry+}A4-4 restores expression of the 3R-TAS1 piRNA in piwi mutants. All lanes were loaded with equal amounts of total RNA.

Full size image

We first used electrophoretic mobility-shift assays to confirm that Piwi binds directly to 3R-TAS1 piRNA (Supplementary Results and Supplementary Fig. 6). We then examined 3R-TAS1 piRNA expression, which was detected in wild-type cells in both the germline and soma and was enriched in nuclei, suggesting its nuclear function (Fig. 2b, c). However, it is not detectable in piwi mutants, suggesting that Piwi is required for its transcription and/or its processing/stability.

To distinguish between these two possibilities, we first examined whether P{w+,ry+}A4-4 affects 3R-TAS1 piRNA expression. We reasoned that P{w+,ry+}A4-4, being a 20-kilobase (kb) euchromatic sequence inserted at the heterochromatic 3R-TAS1 piRNA locus, is more likely to affect the local chromatin state and transcription of 3R-TAS1 than its processing and/or stability. As expected, this piRNA is significantly overexpressed in P{w+,ry+}A4-4 flies (Fig. 2c, d). In precise excision revertants, its expression level is reduced to that in the wild type (Supplementary Fig. 7a), indicating that the P-element insertion is the cause of its overexpression. Moreover, P{w+,ry+}A4-4 rescues its expression in piwi mutants (Fig. 2d). These observations favour a role for Piwi in regulating the transcription of its precursor. The function of Piwi in its processing or stability, if any, must be redundant and thus can be replaced by Ago3 and/or Aubergine in piwi mutants.

P{w+,ry+}A4-4 rescues both 3R-TAS1 piRNA expression and germline stem-cell function in piwi mutants, suggesting the involvement of the 3R-TAS1 piRNA and potentially other piRNAs from 3R-TAS for germline stem-cell maintenance. To assess the range of the P{w+,ry+}A4-4 effect in 3R-TAS, we examined the expression of two piRNAs closest to P{w+,ry+}A4-4 and 3R-TAS1 piRNA: the TAS2 piRNA on the centromeric side, only 54 bp from P{w+,ry+}A4-4, and the HeT-A1 piRNA mapped to the HeT-A retrotransposon on the telomeric side. Because neither piRNA is unique to 3R-TAS, our nuclease protection assays reflect the effects of P{w+,ry+}A4-4 on the total cellular levels of these piRNAs; only the total cellular level is functionally relevant. The TAS2 and HeT-A1 piRNAs are expressed at similar levels in the wild-type and piwi mutant flies, either with or without P{w+,ry+}A4-4 (Supplementary Fig. 7b, c). This suggests that, even if P{w+,ry+}A4-4 has a cis effect on the transcription of these piRNAs from 3R-TAS, this effect is insignificant on their cellular levels and therefore should not have a functional impact. 3R-TAS1 piRNA therefore provides a unique opportunity to assess the cis effect of P{w+,ry+}A4-4 on 3R-TAS and its effect on germline stem-cell maintenance.

To further confirm the effect of P-element insertion on the expression of its nearby piRNA, we examined the expression of the 2R-42AB-B1 piRNA uniquely mapped to the 42AB region (piRNA cluster 17; Supplementary Table 1 and Supplementary Fig. 7d). Like that of 3R-TAS1 piRNA, the expression of the 2R-42AB-B1 piRNA is markedly decreased in piwi mutants. However, a P-element insertion 170 bp downstream of the 2R-42AB-B1 piRNA-coding region decreases rather than increases the expression of this piRNA. Thus, P-element insertions at different sites exert divergent effects on the expression of their nearby piRNAs.

We then investigated whether the positive role of Piwi in regulating 3R-TAS1 piRNA expression is due to its direct effect on the epigenetic state of 3R-TAS. We first examined whether Piwi directly localizes to 3R-TAS(D), a 73-bp region spanning the 3R-TAS1 piRNA-coding sequence (D here stands for distal; Fig. 2a), by Piwi-chromatin immunoprecipitation (ChIP). Piwi is associated with the TAS repeats at a level 46-fold that of the housekeeping gene Rp49 (Fig. 3a). This specific association is further confirmed by Myc–Piwi ChIP with an anti-Myc antibody. Myc–Piwi is enriched in TAS(D) 16.6-fold over an intergenic region on chromosome 2 (Fig. 3b), and 11.4-fold and 11.5-fold over the succinate dehydrogenase B and actin88F genes, respectively (Supplementary Fig. 8). In contrast, Piwi is not associated with any of the five piRNA-poor genomic regions examined, or even with a proximal 3R-TAS sequence (TAS(P)) only 387 bp away (Figs 2a and 3b, and Supplementary Fig. 8). Piwi is therefore strongly associated with TAS(D), which transcribes the 3R-TAS1 piRNA.

Figure 3: Piwi protein is associated with 3R-TAS and is required for its euchromatic state.
figure 3

a, ChIP by anti-Piwi antibody or preimmune serum (preim.) in wild-type and piwi mutant followed by quantitative PCR reveals that Piwi is strongly associated with the 73-bp TAS(D) region. The relative enrichment is calculated by normalizing the quantity of 3R-TAS genomic DNA against the quantity of Rp49 genomic DNA. Grey columns, ry506 ; black columns, piwi2/piwi2;ry506 . b, ChIP by anti-Myc antibody in wild-type and myc–piwi transgenic flies shows that Myc–Piwi is specifically associated with the TAS(D) region but not with three regions 2–10 kb near (N) the representative piRNA clusters (no. 9, no. 11 and no. 16), or non-telomeric HeT-A and TART homologous regions on chromosome 4, or even TAS(P). The relative enrichment is the ratio of the enrichment in myc–piwi flies to that in w1118 flies. c, ChIP reveals that TAS(D) has both euchromatic and heterochromatic histone modifications. Relative enrichment was calculated by normalizing the quantity of TAS(D) DNA co-precipitated by various antibodies against that of control without antibody. n.c., negative control, containing no DNA, for quantitative PCR. d, Association of modified histones H3K4me2, H3K4me3, H3K9ac, H3K9me2, H3K9me3 and HP1 with TAS(D) was assayed by ChIP and quantitative PCR. The relative enrichment of modified histones was calculated by normalizing the quantity of TAS(D) DNA against the quantity of Actin5C. Grey columns, ry506 ; black columns, piwi2/piwi2;ry506 ; hatched columns, piwi2/piwi2;ry506,P{w+,ry+}A4-4. Each individual experiment was repeated at least three times. Error bars indicate s.d.

Full size image

We then characterized the epigenetic states of 3R-TAS(D) by ChIP. In wild-type flies, 3R-TAS(D) is associated with both euchromatic modification markers (H3K4me2, H3K4me3 and H3K9ac, the last of these being a transcriptional marker) and heterochromatic modification markers (H3K9me2, H3K9me3, H4K12ac and HP1), suggesting that 3R-TAS(D) might be under dynamic equilibrium between euchromatic transcription and heterochromatic silencing (Fig. 3c). Although Piwi has a global function in heterochromatic silencing (Supplementary Results and Supplementary Fig. 9), it has the opposite effect on 3R-TAS(D). In piwi mutants, TAS(D)-associated H3K9ac, H3K4me2 and H3K4me3 levels are decreased 4.5-fold, 3.3-fold and 7.0-fold, respectively (Fig. 3c). In contrast, TAS(D)-associated HP1, H3K9me2 and H3K9me3 are enriched 2.2-fold, 1.3-fold and 1.5-fold, respectively. Similarly, in the 172-bp repeat region, levels of the three euchromatic markers were decreased more than in the 73-bp region, HP1 was enriched 10.3-fold, even though H3K9me2 and H3K9me3 levels were decreased by 31% and 41%, respectively (Supplementary Fig. 10). These histone modification profiles are consistent with our finding that 3R-TAS1 piRNA is expressed in the wild-type fly but not in piwi mutants, indicating that Piwi promotes the euchromatic feature of TAS(D) chromatin.

To further test the positive epigenetic role of Piwi towards TAS(D), we examined the histone modification profile of this region in piwi2/piwi2;P{w+,ry+}A4-4/+ flies, in which 3R-TAS1 piRNA expression is restored. Indeed, we found that the euchromatic feature of TAS(D) is significantly restored (Fig. 3d). In comparison with the piwi2 mutant, the TAS-associated H3K9ac in piwi2 ;P{w+,ry+}A4-4 flies is increased 3.1-fold, reaching about 70% of the wild-type level. Correspondingly, TAS-associated HP1, H3K9me2 and H3K9me3 levels are decreased to those of the wild type. This significant restoring effect of P{w+,ry+}A4-4 is unlikely to be due to other cryptic effects of P{w+,ry+}A4-4 itself, because the same sequence inserted into other genomic sites does not suppress the piwi mutant phenotype22. Instead, it is likely that the insertion of P{w+,ry+}A4-4, as a 20-kb unique sequence, may affect heterochromatization in 3R-TAS. P{w+,ry+}A4-4 therefore rescues the germline stem-cell phenotype of the piwi mutant by restoring the euchromatic feature of 3R-TAS and the transcription of 3R-TAS1 piRNA.

The positive epigenetic role of Piwi in TAS(D) is also demonstrated by the expression of a reporter gene, white, in P{w+,ry+}A4-4 insertion. This white gene exhibits a typical telomere position effect (Fig. 4), suggesting that TAS is under heterochromatic influence23. Expectedly, loss of piwi function enhances telomere position effect in a dosage-sensitive manner, opposite to the known suppression effects of Polycomb group proteins (Fig. 4a)24,25. In P{w+,ry+}A4-4 stocks, the white gene was variably expressed in different ommatidia in piwi+/piwi+ flies, with an overall orange eye colour with spotty red ommatidia in the posterior region. In piwi+/piwi2 flies, the eye colour becomes lighter. In piwi2/piwi2 flies, the eye colour turns completely white in most ommatidia, leaving the posterior red ommatidia apparently unaffected by the piwi dosage. This indicates that piwi is required for the expression of white inserted in 3R-TAS. Eye pigment assay shows that white is expressed in piwi2 homozygotes at a level 5.0–7.8-fold lower than in piwi2 heterozygotes (Fig. 4b). Because piwi does not affect the expression level of the endogenous white gene on the X chromosome (Fig. 4b), piwi is probably required for the expression of the white reporter gene by promoting the active epigenetic state of 3R-TAS.

Figure 4: piwi is an enhancer of telomere position effect.
figure 4

a, The eye colours of four independently maintained w1118;P{w+,ry+}A4-4 strains (211, 516, R86-2 and R86-2-3) in wild-type, heterozygous and homozygous piwi2 backgrounds. b, Levels of eye pigmentation in Canton-S, piwi2/piwi2 , w;piwi2/piwi2 and w1118;P{w+,ry+}A4-4 strains (211, 516, R86-2 and R86-2-3) in both heterozygous and homozygous piwi2 backgrounds. Error bars indicate s.d.

Full size image

The above findings reveal the complexity of small RNA-mediated epigenetic regulation, namely that Piwi can exert opposite effects on different genomic regions. piRNAs may have a function in guiding Piwi to the target sites, yet the opposite effects of Piwi at different target sites might be mediated by the local chromatin context, which would render the selective binding of the Piwi–piRNA complex to different partners such as HP1a or JmJC domain-containing histone demethylases26. The activating effect of Piwi, 3R-TAS1 piRNA and possibly other piRNAs in 3R-TAS can be explained by a heterochromatin/euchromatin counterbalance model in which the repetitive nature of 3R-TAS by default is a substrate for heterochromatization. The heterochromatic state could be established and maintained by the Polycomb group proteins or an RNAi pathway mediated by Ago proteins, or yet another mechanism. However, the association of Piwi with 3R-TAS by means of the 3R-TAS1 piRNA or P{w+,ry+}A4-4 insertion as a roughly 20-kb unique sequence counteracts the heterochromatization (Supplementary Discussion and Supplementary Fig. 11). Our finding that Piwi is required for the epigenetic activation of the subtelomeric region starts to reveal a mechanism underlying epigenetic regulation and stem-cell maintenance.

Methods Summary

piRNA cloning, mapping and annotation

Myc–Piwi ribonucleoprotein complexes were immunoprecipitated from adult ovarian tissues with monoclonal anti-Myc antibody (9E10). Total RNA was prepared by using TRIzol. Small RNAs were gel-purified, cloned and sequenced as described27. Cloned small RNAs were mapped to the D. melanogaster genome assembly, version 5.1. Functional annotation was performed by in-house Perl scripts aiding BioPerl modules and Ensembl API.

RNase protection assay

PCR-amplified template DNAs were cloned into the pGEM-T vector (Promega). High-specific-activity probes were generated by in vitro transcription with the MaxiScript T7/Sp6 kit (Ambion) and [α-32P]UTP. TRIzol-extracted RNA (20 μg) was hybridized overnight with 2 × 105 c.p.m. of radioactive probe at 42 °C. Unpaired RNA was digested by an RNaseA/RNaseT1 mixture.

Chromatin immunoprecipitation and quantitative PCR

Nuclei from adult flies were isolated and crosslinked with 0.1% formaldehyde. Chromatin was fragmented by sonication in RIPA buffer. Antibodies were incubated overnight with nuclear extracts at 4 °C. Bead-bound DNA was eluted, reverse-crosslinked and precipitated. Quantitative PCR was conducted on a Roche LightCycler 2.0 system with the LightCycler DNA Master SYBR Green I. Sequences of primers are provided in Methods. Normalized enrichment values were calculated with a standard formula.

Online Methods

Drosophila stocks and cultures

All fly stocks were maintained at 20 °C. For the immunoprecipitation assay, a Myc–Piwi transgenic strain (G38-1A) and w1118 flies were used. For the RNase protection assay, w1118, w1118;piwi1/piwi1 , w1118;piwi2/piwi2 , w1118;P{w+,ry+}A4-4,ry506 (221|516|R86-2|R86-2#3) and w1118;piwi2/piwi2;P{w+,ry+}A4-4,ry506(221|516|R86-2|R86-2#3) flies were used. For the ChIP assay, w1118;ry506 , w1118;piwi2/piwi2 ;ry506 , w1118;piwi2/piwi2;P{w+,ry+}A4-4(221),ry506 /+,ry506 and G38-1A flies were used. For global analysis of histone PTMs, w1118 and w1118;piwi2/piwi2 flies were used. For the telomere position effect assay and the fly eye pigmentation assay, Canton-S, piwi2/piwi2 , w1118;piwi2/piwi2 , w1118;Sco/CyO;P{w+,ry+}A4-4,ry506(221|516|R86-2|R86-2#3), w1118;piwi2/CyO;P{w+,ry+}A4-4,ry506(221|516|R86-2|R86-2#3) and w1118;piwi2/piwi2;P{w+,ry+}A4-4,ry506(221|516|R86-2|R86-2#3) flies were used.

Immunoprecipitation assay and cloning of piRNAs

Adult ovaries (100 pairs) were homogenized in an equal volume of ovary lysis buffer (20 mM HEPES pH 7.5, 100 mM KCl, 5 mM MgCl2, 0.1% SDS, 0.1% sodium deoxycholate, 1% Triton X-100, 1 mM dithiothreitol, 0.2 mM phenylmethylsulphonyl fluoride, 1 × Complete Mini, EDTA free Proteinase Inhibitor cocktail (Roche), 0.5 unit μl-1 RNAaseOUT (Invitrogen), 5% glycerol). Anti-Myc monoclonal antibody (9E10; Developmental Studies Hybridoma Bank at the University of Iowa) was added at 1:10 dilution to the precleared ovary lysate and incubated for 2 h at 4 °C. Protein G–Sepharose 4B beads were added and incubated for a further 1 h at 4 °C. Bead-bound RNAs were extracted with TRIzol (Invitrogen) and precipitated in the presence of 50 μg ml-1 GlycoBlue (Ambion). The co-precipitated RNA was 5′ labelled with [γ-32P]ATP by T4 polynucleotide kinase (NEB), purified on a Sephadex G-25 fine RNA spin column (Roche) and separated by 15% denaturing PAGE for detection. A PCR-amplified small RNA library (1 μg) was directly sequenced with a large-scale pyrosequencing method (454 Life Sciences).

Genome mapping and annotation of cloned small RNAs

Cloned small RNAs were mapped to both the D. melanogaster genome assembly, version 5.1, and annotated sequence databases to infer the likely genome origins. The mapping was performed with a standalone NCBI BLAST program (http://www.ncbi.nlm.nih.gov/BLAST/download.shtml), with sensitive parameters to identify perfect matches of short query sequences (no mismatch, insertion or deletion is allowed). In-house Perl scripts aiding BioPerl modules (http://www.bioperl.org) and Ensembl API (http://www.ensembl.org/info/software/core/core_tutorial.html) were used to automate the mapping, annotation and analysis procedure. A home-made non-coding RNA database was built by combining the available databases with batch-fetched entries with appropriate feature keys from Ensembl. The sequences and annotations of transposons and repetitive sequences in Drosophila melanogaster genome assembly, version 5.1, were identified by a standalone RepeatMasker program with a sensitive filter (http://www.repeatmasker.org).

A piRNA was annotated to a specific type of genomic sequence only if the sequence covered the full length of the piRNA. A piRNA cluster was defined as a genomic region containing more than 100 piRNA mapped sequences and not having any piRNA mapped sequence in the upstream 5-kb region or the downstream 5-kb region. The borders of a piRNA cluster were defined by the first nucleotide of the farthest upstream piRNA and the last nucleotide of the farthest downstream piRNA.

Chromatin immunoprecipitation and quantitative PCR

Nuclear extract was incubated at 4 °C for 2 h separately with the following antibodies: anti-Piwi (generated in the Lin laboratory against Piwi C-terminal peptide), 1:100 dilution; preimmune of anti-Piwi, 1:10 dilution; anti-Myc monoclonal antibody (9E10; Developmental Studies Hybridoma Bank, University of Iowa), 1:10 dilution; anti-dimethyl-histone3 K4 (Upstate), 1:100 dilution; anti-trimethyl-histone3 K4 (Upstate), 1:100 dilution; anti-dimethyl-histone3 K9 (Upstate), 1:50 dilution; anti-trimethyl-histone3 K9 (Upstate), 1:50 dilution; anti-acetyl-histone3 K9 (Upstate), 1:100 dilution; anti-acetyl-histone4 K12 (Upstate), 1:100 dilution; anti-dimethyl-histone3 K27 (Upstate), 1:50; and anti-HP1 (Covance), 1:50 dilution. Salmon-sperm DNA (50 μl)/Protein A/G-agarose beads (Upstate) was added and samples were incubated at 4 °C for 1 h. After washing, crosslinking was reversed at 65 °C overnight. DNA was purified from the beads by phenol/chloroform extraction and ethanol precipitation.

The following primer sets were used for quantitative PCR: for the TAS(D) region, 5′-gtgtctcatccatttcctttattcag-3′ (forward) and 5′-tggtcgtgttgatcggtacttg-3′ (reverse); for a 169-bp fragment of the 172-bp repeat region, 5′-gatcttcttacatttcccttcttcaac-3′ (forward) and 5′-cggcagaggcacgaacaac-3′ (reverse); for a 345-bp region in the proximal end of the 3R-TAS repeat, 5′-caacccaatcggacctcactt-3′ (forward) and 5′-gtgacgattaatacgaaaacttacaaac-3′ (reverse); for an 81-bp region about 2 kb upstream of the piRNA cluster no. 9, 5′-aaatgcagcaggcagcgcgaa-3′ (forward) and 5′-cctcaatatgtagagtagtgcgagtgactt-3′ (reverse); for an 86-bp region about 8 kb downstream of the piRNA cluster no. 11, 5′-gctcaagagtcctccagacaggtt-3′ (forward) and 5′-gcagtgatggtggtggcagtt-3′ (reverse); for a 128-bp region about 10 kb downstream of the piRNA cluster no. 16, 5′-agggttatgctaggttcttatgctgc-3′ (forward) and 5′-ggaaacgaataaacaaatgggtcaaca-3′ (reverse); for a 164-bp non-telomeric HeT-A homologous region on chromosome 4, 5′-agtcgatgttaaatccattccg-3′ (forward) and 5′-tgggttacttgtcctatgtgcc-3′ (reverse); for 129-bp non-telomeric TART homologous region on chromosome 4, 5′-tcaacatacgcagacgacact-3′ (forward) and 5′-agcatttactgagataccccatt-3′ (reverse); for a 97-bp region in the Rp49 gene, 5′-tcgagttgaactgcgttagtccgt-3′ (forward) and 5′-gccatttggcgaactttcacagga-3′ (reverse); for a 224-bp intergenic region [2R:11,327,115..11,327,338], 5′-ttgggaggcagggaaaggatg-3′ (forward) and 5′-aaggaggagaatggtaaggagatggat-3′ (reverse); for a 165-bp region in the Succinate Dehydrogenase B gene (SDHB), 5′-ccaggaaccaggtgagtggagtgc-3′ (forward) and 5′-gcgtgcttatctgtggcgtttctatc-3′ (reverse); for a 187-bp region in the Actin88F gene, 5′-aagctcttcaaaggcagcaaccag-3′ (forward) and 5′-aaatggccatgaaggatgagcacc-3′ (reverse); for 334-bp region in the Actin5C gene, 5′-tgcccgacggacaggtgat-3′ (forward) and 5′-tggaaggtggacagcgaagc-3′ (reverse). Using a standard formula, the quantities of target genomic regions precipitated by different antibodies were normalized against those of Rp49, the 2R intergenic region, Actin5C, SDHB and Actin88F. This ratio was further normalized against those from input DNA (without ChIP). The relative quantities of the TAS(D) region precipitated by different antibodies in Fig. 3c were normalized against the quantity of TAS(D) region precipitated by naked beads without any antibody. The difference in amplification efficiency between the various primer sets was determined by standard curves and was taken into account in the normalization calculation. The average values and standard deviations of the relative enrichments were used to compare the association of Piwi in different genomic regions and the histone modification profiles of the TAS region in different genotypes.

Drosophila eye pigmentation assay

Fifty 5–8-day-old fly heads were homogenized in 400 μl of acidified methanol (0.1% HCl in methanol) and vortex-mixed for 30 min at room temperature. After centrifugation for 5 min, 20 μl of 0.5% H2O2 was added to the supernatant. After centrifugation for 10 min, the absorbance at 480 nm was measured. Samples for each genotype were repeated three times. The average absorbance and standard deviations were used to quantify eye colours.

Electrophoretic mobility-shift assay

Piwi full-length complementary DNA with a 5′ Myc tag sequence was cloned into the pBlueScript KS vector (Stratagene). Myc–Piwi protein was translated in vitro with a TNT T7 Coupled Wheat Germ Extract System (Promega) and standard procedures. Mock in vitro translation was performed without pBS-KS:myc–piwi DNA template. Increasing amounts of concentrated Myc–Piwi and mock in vitro translation reactions were incubated at room temperature for 30 min with 5 fmol (about 104 c.p.m.) of 5′ end-labelled 3R-TAS1 piRNA probe or 3R-TAS1 piRNA probe with the 5′ U replaced with C, in the presence or absence of different amounts of unlabelled RNA oligonucleotides (10 and 50 fmol). The binding was done in 1 × binding buffer (20 mM HEPES pH 7.5, 150 mM NaCl, 2 mM MgCl2, 1 mM dithiothreitol, 10% glycerol, 15 nM yeast tRNA).