• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of embojLink to Publisher's site
EMBO J. Oct 7, 2009; 28(19): 2932–2944.
Published online Jul 30, 2009. doi:  10.1038/emboj.2009.220
PMCID: PMC2760103

Endo-siRNAs depend on a new isoform of loquacious and target artificially introduced, high-copy sequences


Colonization of genomes by a new selfish genetic element is detrimental to the host species and must lead to an efficient, repressive response. In vertebrates as well as in Drosophila, piRNAs repress transposons in the germ line, whereas endogenous siRNAs take on this role in somatic cells. We show that their biogenesis depends on a new isoform of the Drosophila TRBP homologue loquacious, which arises by alternative polyadenylation and is distinct from the one that functions during the biogenesis of miRNAs. For endo-siRNAs and piRNAs, it is unclear how an efficient response can be initiated de novo. Our experiments establish that the endo-siRNA pathway will target artificially introduced sequences without the need for a pre-existing template in the genome. This response is also triggered in transiently transfected cells, thus genomic integration is not essential. Deep sequencing showed that corresponding endo-siRNAs are generated throughout the sequence, but preferentially from transcribed regions. One strand of the dsRNA precursor can come from spliced mRNA, whereas the opposite strand derives from independent transcripts in antisense orientation.

Keywords: Drosophila melanogaster, endo-siRNAs, post-transcriptional regulation, RNA interference, transposon silencing


The appearance of a new selfish genetic element in an organism's genome is often accompanied by amplification and mobilization, leading to a highly mutagenic situation until the cells regain control. Most eukaryotic organisms carry a large proportion of transposons in their genome, indicating that the taming of selfish genetic elements was often not immediately successful (Orgel and Crick, 1980; Pace and Feschotte, 2007). In Drosophila, zebrafish, rat and mouse germ line cells, the piRNA system can efficiently repress transposons (for review, see Malone and Hannon, 2009), but only if there is a pre-existing, maternally contributed pool of piRNAs with corresponding sequence (Blumenstiel and Hartl, 2005; Brennecke et al, 2008). If not, the resulting progeny has severely impaired fertility due to frequent transposon mobilization. The necessary maternal pool of piRNAs to prevent this phenomenon (called hybrid dysgenesis) can be built up slowly over the course of several generations (for review, see Chambeyron and Bucheton, 2005). Although an amplification loop has been proposed that can explain the maintenance of corresponding sense and antisense piRNA populations, we currently do not understand how a piRNA response is initiated de novo (for review, see Aravin et al, 2007; Hartig et al, 2007; Malone and Hannon, 2009; O'Donnell and Boeke, 2007).

Even though germ cells are the ultimate target for selfish genetic elements, they are not the only cells that can be faced with a newly emerging transposon threat. Certain viruses, for example, can carry transposable elements in their genome, which will become integrated into the host cell genome on establishment of latency (van Oers and Vlak, 2007). Silencing of transposable elements is, therefore, also important in somatic cells to reduce the possibility of transition into the germ cell lineage (Chalvet et al, 1999; Pelisson et al, 2002). The recent discovery of endo-siRNAs points to one mechanism by which transposons can be silenced in the somatic cells of Drosophila (Chung et al, 2008; Czech et al, 2008; Ghildiyal et al, 2008; Kawamura et al, 2008), and a similar class of RNAs has been detected in mouse ES cells (Watanabe et al, 2006, 2008; Babiarz et al, 2008; Tam et al, 2008). The transposon-matching endo-siRNAs are comparable with piRNAs in the sense that they induce homology-dependent repression, but their biogenesis is clearly different: a roughly equal distribution of sense-matching and antisense-matching endo-siRNAs implies a double-stranded precursor. This is substantiated by the dependence of endo-siRNAs on the enzyme Dicer-2 (Dcr-2) that also processes dsRNA to induce ‘classical' RNA interference (for review, see Golden et al, 2008). In the case of mouse oocytes, a generation of dsRNA in trans from two distinct loci (e.g. a gene and a corresponding processed pseudogene) has been proposed (Tam et al, 2008; Watanabe et al, 2008). Despite these insights into endo-siRNA biogenesis, one fundamental question remains unsolved: can a newly appearing selfish genetic element be recognized without the need for a pre-existing pool of small RNAs and an inactive genetic remnant or pseudogene? And if yes, how are the double-stranded precursor molecules for endo-siRNAs generated?

Loqs normally functions together with Dcr-1 in miRNA biogenesis (Forstemann et al, 2005; Jiang et al, 2005; Saito et al, 2005; Park et al, 2007), whereas the usual partner for Dcr-2 is R2D2 (Liu et al, 2003). The dependence of endo-siRNAs on dcr-2 in combination with loqs, but not r2d2, was unexpected, as it clearly distinguishes the dsRNA-mediated response against transposons from the r2d2-dependent response that occurs after infection with an RNA virus (Wang et al, 2006). In addition, it suggested significant overlap between the miRNA and siRNA biogenesis pathways at the level of the processing step, whereas so far an exchange had only been described at the Argonaute loading step (Forstemann et al, 2007; Tomari et al, 2007). The complex could be verified biochemically (Czech et al, 2008), but it is unclear which of the Loqs protein isoforms is binding to Dcr-2. As only one isoform, Loqs-PB, is essential for miRNA biogenesis together with Dcr-1, the overlap between the miRNA and the endo-siRNA pathways may, in fact, be smaller than what was initially assumed. In an increasingly complex repertoire of small RNAs (Core et al, 2008; He et al, 2008; Preker et al, 2008; Seila et al, 2008; Fejes-Toth et al 2009; Lee et al, 2009; Taft et al, 2009), the importance of distinguishing each class during its biogenesis and of ensuring functional specificity is obvious. This is indeed an almost daunting task, as the various classes of small RNAs all share a common chemical structure and populate similar size ranges.

In this study, we show that biogenesis of endo-siRNAs depends on a new isoform of the dsRNA-binding domain protein Loquacious (also known as R3D1), which is distinct from the isoform acting in the miRNA pathway. We describe a cell-culture model in which an artificial plasmid sequence, integrated at high-copy number, has become subject to endo-siRNA-mediated repression. The artificial, as well as an endogenous transposon target for endo-siRNAs are silenced through a post-transcriptional mechanism, indicating that repression is direct, and deep sequencing analysis showed corresponding endo-siRNAs. The repressive response also occurred in transiently transfected cells. Thus, Drosophila endo-siRNAs can mount a de novo response without a need to integrate into the host cell genome or the need for any pre-existing sequence that can serve as a template for the production of small RNAs.


Production of Drosophila endo-siRNAs depends on a new isoform of loqs

The observation that endo-siRNA-mediated silencing depends on loqs, in combination with dcr-2, was a surprise because it suggested that a hybrid complex with components of the canonical miRNA and siRNA biogenesis pathways exists. However, this interpretation may have been an oversimplification because the loqs gene can give rise to, at least, three different mRNAs, each coding for a protein with distinct properties (Forstemann et al, 2005; Jiang et al, 2005; Saito et al, 2005). Only one isoform, called Loqs-PB, is an essential partner of Dcr-1 in the biogenesis of miRNAs (Jiang et al, 2005; Park et al, 2007). We carried out 3′ RACE experiments and detected a new mRNA variant of loqs (loqs-RD), in which an alternative polyadenylation in the third intron leads to a new protein isoform that lacks the 3rd dsRBD of Loqs-PB/PA and contains 22 amino acids of new protein sequence (Figure 1A and Supplementary Figure 1, Loqs-PD).

Figure 1
The Loqs-PD isoform is primarily responsible for endo-siRNA biogenesis. (A) Schematic diagram of the four loqs mRNA and protein variants currently known. Start codons are indicated by vertical green bars, stop codons by vertical red bars. The regions ...

Isoform-specific knockdowns are possible if the corresponding mRNA contains a stretch of unique sequence that is amenable to RNAi. In the case of the loqs gene, it is possible to target the loqs-RC and loqs-RD RNA individually, the loqs-RB RNA together with the loqs-RC RNA, loqs-RA, loqs-RB and loqs-RC together through the common sequence towards the 3′ end and finally all four loqs isoforms simultaneously with dsRNA directed against the amino-terminus of Loqs. Detection of Loqs protein isoforms on immunoblots has shown three bands of distinct sizes, which had been assumed to correspond to Loqs-PB, Loqs-PA and Loqs-PC (Forstemann et al, 2005). Using our isoform-specific knockdown, we could show that the smallest of these bands predominantly contains the new Loqs-PD isoform (Figure 1B).

To determine which of the four Loqs protein isoforms is required for endo-siRNA biogenesis, we depleted individual Loqs protein isoforms and then measured the levels of an endo-siRNA derived from the long hairpin-forming gene, CG4068, by northern Blotting (Czech et al, 2008; Kawamura et al, 2008; Okamura et al, 2008b). Only depletion of the loqs-RD transcript correlated with a reduced production of the CG4068 ‘B' endo-siRNA (Figure 1C), whereas the biogenesis of the bantam miRNA was not impaired when loqs-RD was depleted. Depletion of loqs-RB+RC, as expected, led to an increased abundance of pre-bantam. Co-IP of Loqs with Dcr-2 has been described previously (Czech et al, 2008) but the antibody employed recognized all Loqs protein isoforms. Using cDNA constructs coding for only one of the splice variants, we could distinguish Loqs-PA, PB and PD in transfected S2 cells and determine the potential of each isoform to interact with Dcr-2 by co-immunoprecipitation. Epitope-tagged Loqs-PD, similar to tagged R2D2, was able to associate with Dcr-2 but not with Dcr-1, though the extent of Dcr-2 association varied between experiments (Supplementary Figure 2). This is consistent with the previous observation that on the level of the endogenous protein, the smallest Loqs isoform (identified as PD in this paper) does not co-immunoprecipitate with Dcr-1 (Forstemann et al, 2005). In contrast, tagged Loqs-PB and Loqs-PA can associate with Dcr-1, and also to some extent with Dcr-2. As only Loqs-PD is required for endo-siRNA generation, it seems that the Dcr-2–Loqs-PD complex is exclusively responsible for endo-siRNA generation.

Stably integrated transgenes can be subject to repression that depends on endo-siRNA biogenesis factors

To further validate the importance of the Loqs-PD isoform for endo-siRNA-dependent silencing, we attempted to generate a reporter system that closely resembles natural endo-siRNA targets and, at the same time, provides a convenient read-out. Our reasoning was that some of the canonical features of transposable elements, such as multicopy insertion and the formation of repetitive regions, are shared by transgenes that have integrated into the host cell genome after transfection and selection of stable cell culture lines. Indeed, Ago2-dependent repression of a stably integrated GFP expression plasmid in Drosophila cells has been described (Siomi et al, 2005). We examined a clonal cell line (called 63N1) that expresses GFP (Figure 2A) and tested the changes in GFP levels on impairment of known miRNA/siRNA pathway components. From the three RNaseIII enzymes Drosha, Dcr-1 and Dcr-2, only the latter seemed to be involved in repression of GFP. Depletion of the cytoplasmic dsRBD protein, Loqs, also resulted in a de-repression of GFP, whereas depletion of its homologue, R2D2, seemed to increase repression. Finally, Ago2 is the main effector protein mediating this response, although depletion of Ago1 also resulted in a slight de-repression (Figure 2B). These are precisely the genetic requirements of the endo-siRNA pathway in Drosophila (Czech et al, 2008; Kawamura et al, 2008; Okamura et al, 2008b). If the GFP transgene has caused an endo-siRNA response, then the new Loqs-PD isoform should be required for its repression. Using isoform-specific knockdowns, as described in Figure 1A, we observed that GFP fluorescence did not change upon a knock down of loqs-RA+RB+RC or loqs-RC alone. In contrast, specific targeting of loqs-RD as well as targeting loqs-RD and loqs-RC together led to an even stronger de-repression than knock down of all loqs variants (Figure 2C). Thus, Loqs-PD seems to be required for both hairpin-derived and repetitive element-derived endo-siRNAs.

Figure 2
Stably integrated transgenes are subject to endo-siRNA-mediated repression. (A) Our experimental system is based on transfection of a GFP expression construct together with an antibiotic resistance plasmid to allow for selection of stable transformants, ...

The 63N1 cell line has mounted a bona fide endo-siRNA response against the integrated transgene, but the extent to which this occurs varies between clones. In a previous publication, a cell line (called 63–6) derived independently from the same stock of parental S2 cells and transfected with the same expression plasmid showed only marginal response towards depletion of Dcr-2, Loqs and Ago-2 (Forstemann et al, 2007). We re-examined this cell line and could corroborate that there is a minor, but significant increase in GFP levels on depletion of Dcr-2 (1.2-fold, P<0,01) and loqs-RD (1.1-fold, P<0.02). One potential difference between the two cell lines is the number of plasmid copies that have integrated in the genome. Therefore, we compared the copy number with isolated genomic DNA by quantitative PCR. We amplified a region from the ubiquitin promotor that drives expression of GFP and compared the results with GAPDH and two other single-copy genes. Relative to GAPDH, the ubiquitin promotor was 8-fold more abundant in the cell line with the blunted response, whereas it was 42-fold more abundant in the 63N1 cell line (Figure 2E). This suggests that a high-copy number may favour the establishment of robust repression.

GFP mRNA is a direct target of the endo-siRNA response

Endogenous RNA targets of the endo-siRNA pathway become more abundant on depletion of Dcr-2 or Ago2. The simplest interpretation of these results is that endo-siRNAs—like siRNAs—induce the post-transcriptional degradation of mRNAs, and this capacity has indeed been shown (Chung et al, 2008; Czech et al, 2008; Ghildiyal et al, 2008; Kawamura et al, 2008; Okamura et al, 2008b). To prove that the GFP transgene is a direct target of the small RNAs, and to rule out that upregulation of GFP is an indirect consequence of endo-siRNA loss leading to increased transcription of the GFP-coding gene, we examined transcription and degradation rates of GFP mRNA in vivo by pulse labelling newly synthesized RNA (Johnson et al, 1991; Kenzelmann et al, 2007; Dolken et al, 2008). To verify the procedure (depicted in Figure 3A), we determined the ratio of the hsp70 mRNA, which has a half-life of 15–30 min in S2 cells (Petersen and Lindquist, 1988), to the stable rp49 mRNA in each RNA fraction. As expected, we found that relative to rp49, hsp70 mRNA was significantly more abundant in the newly synthesized RNA fraction than in unfractionated RNA (<1 h versus total RNA, Figure 3B). In addition, we examined the effect of our dcr-2 RNAi in comparison with our control RNAi against DsRed. As shown in Figure 3C, knock down reduced the dcr-2 mRNA levels by about twofold in total RNA (cDNA synthesis was primed with random hexamers, thus we potentially also detected mRNA degradation fragments), but it was much more pronounced (20-fold) in the flow-through fraction (>1 h old) of our RNA separation procedure. We did not observe any significant change of dcr-2 mRNA in the newly transcribed fraction. The 1 h time window for labelling seems, therefore, appropriate to separate transcriptional from post-transcriptional events.

Figure 3
GFP mRNA is a direct target of the endo-siRNA pathway. (A) Overview of the pulse labelling and purification method adapted from Dolken et al (2008). (B) To verify the RNA fractionation procedure, the ratio between the short-lived hsp70 and the long-lived ...

If de-repression of our GFP transgene occurred by a transcriptional mechanism, then inhibition of endo-siRNA biogenesis should cause an increased abundance of target mRNAs in the <1 h RNA fraction. However, if a post-transcriptional mechanism is responsible, the increase should be stronger in the >1 h RNA fraction. We observed a trend towards increased mRNA levels for our GFP transgene and the mdg1 transposon (an endogenous target of the endo-siRNA pathway) in the total RNA fraction on knock down of dcr-2. After fractionation, we could detect significant changes in the >1 h fraction for GFP and mdg1 (Figure 3D, P<0,05, two-tailed z-test, n=3). We did not detect any significant changes in the newly transcribed RNA fraction, indicating that the increased GFP levels caused by impaired endo-siRNA biogenesis are due to an increased stability of the GFP mRNA. As a positive control for changes in the rates of transcription, we used a UAS–GFP vector with and without co-expression of the Gal4 transcription factor. This resulted in an increase of GFP mRNA in all three RNA fractions (Supplementary Figure 3).

A second natural endo-siRNA target, the transposon 297, showed no significant upregulation in any of the three fractions. One possible reason for this may be that endo-siRNAs directed against 297 are exceptionally abundant (as determined by deep sequencing; reads antis. to GFP: 8405; reads antis. to mdg1: 4146; reads antis. to 297: 35083) and that our dcr-2 depletion may not have been efficient or long enough for these endo-siRNAs to decay.

Detection and analysis of transgene-derived endo-siRNAs

If GFP repression is due to mRNA targeting by endo-siRNAs, then small RNAs with complementarity to GFP should be present in the reporter cells. We isolated 18–30-nt long RNAs from the parental S2 cells as well as the stable GFP-expressing clone 63N1, then analysed them by deep sequencing. After removal of adapter sequences and length selection (20–24 nt size window), we obtained 1 098 002 reads for the parental cell library and 2 678 671 reads for the 63N1 cell library. Surprisingly, 359 845 reads from the parental cell library and 1 112 196 reads in the 63N1 library corresponded to miR-184. This exceeds by far the frequency with which miR-184 has been detected in other studies and we, therefore, disregarded these reads during further analysis, leaving 738 157 reads in the parental cell library and 1 566 475 reads in the 63N1 cell library.

After mapping to the sequence of our GFP expression construct, we found 16 424 corresponding reads in the 63N1 cell library and 167 reads in the parental cell library (see Table I). The size distribution of these reads showed a sharp peak at 21 nt (Supplementary Figure 4), consistent with the length preference of Ago2. When we analysed from which positions in our construct the corresponding reads were derived, we found that not only the GFP-coding region but also many other regions of the construct gave rise to small RNAs (Figure 4A). Overall, sense- and antisense-matching reads were equally represented (8019 sense reads versus 8405 antisense reads). This is an indication that the small RNAs directed against the construct derive from a double-stranded precursor (Czech et al, 2008; Ghildiyal et al, 2008; Kawamura et al, 2008; Okamura et al, 2008a). We also re-analysed the published deep sequencing data from the 63–6 cell line (Seitz et al, 2008) for corresponding small RNA reads. Similar to 63N1 cells, we found reads derived from many regions of the construct (Figure 4B), though the white gene produced many more endo-siRNAs than it did in the 63N1 cells. This could reflect differences between the cell lines (e.g. due to integration sites) or differences in the deep sequencing techniques (pyrosequencing versus sequencing by synthesis).

Figure 4
Deep sequencing of transgene-derived endo-siRNAs. S2 cell-derived small RNA reads were mapped to the sequence of the transfected GFP expression plasmid (horizontal axis; values above axis: quantification of sense reads, below: quantification of antisense ...
Table 1
Deep sequencing reads

Using the 63-6 cell dataset, we could show that the transgene-derived endo-siRNAs are indeed loaded into Ago2 complexes. If a small RNA resides in Drosophila Ago2, it will be 2′-O-methyl modified and become resistant to β-elimination at its 3′ end (Horwich et al, 2007). By comparing the read counts from a mock-treated RNA sample with a β-eliminated RNA sample, we could assess the extent to which the endo-siRNAs derived from our construct were loaded into Ago2 complexes (Supplementary Figure 5). We found 510 corresponding reads in the mock-treated small RNA population and 4167 reads in the β-eliminated reads (excluding the reads that derive from the dsRNA directed against GFP with which the cells had been treated in this study). This enrichment (8.17-fold) is identical to the 8.2-fold enrichment calculated for endo-siRNAs corresponding to transposable elements in this data set (Chung et al, 2008).

Transient transfection also leads to an endo-siRNA response

What is triggering the production of double-stranded RNA from the expression construct in the stable cell lines? One possibility is that a head-to-head arrangement of plasmid concatemers is produced upon integration and that read-through transcription yields double-stranded RNA, as has been proposed for natural transposon clusters (Siomi, 2008). Alternatively, it may be the high-copy number per se that is somehow triggering dsRNA production. To assess whether integration is a prerequisite for endo-siRNA generation, we examined a small RNA library generated from transiently transfected cells. The deep sequencing data were processed, as described before, and a total of 199 418 reads remained after the removal of adapter sequences and length selection. We found 759 reads corresponding to the plasmid sequence. After normalization to small RNAs derived from the CG4068 hairpin, this corresponds to a roughly fivefold lower abundance in the transiently transfected cells relative to the 63N1 cells (Table I). However, this comparison should be regarded as a rough estimate because normalization of deep sequencing libraries is not straightforward; furthermore, the total read number was considerably lower for the transfected cells than for the 63N1 cells. Nonetheless, the abundance of small RNAs in transiently transfected cells seems too high to be caused by rare cells within the population that have already integrated the plasmid in their genome.

To corroborate the fact that transient transfection can lead to endo-siRNA production, we analysed a published small RNA library derived from S2 cells that had been transfected with an expression plasmid for epitope-tagged Ago2. The library contains small RNAs isolated after immunoprecipitation of the tagged Ago2, among them siRNA-sized reads mapping to the transfected plasmid (G Hannon, personal communication). As both our GFP expression vector and the Ago2 expression construct are based on pCASPER-2, we determined the number of small RNA reads matching to the vector backbone to compare the occurrence between the libraries. The Ago2 IP library contained 50 564 small RNA reads matching pCASPER-2, indicating that in this case a substantial production of endo-siRNAs has occurred. Table I summarizes the read counts corresponding to the plasmid backbone as well as the endo-siRNA generating hairpin locus CG4068 for normalization. Though it is difficult to compare an analysis of all small RNAs with an analysis of only the Ago2-associated small RNAs, there seems to be a difference in the extent to which transient transfection has triggered the endo-siRNA response between the two experiments. We prefer not to speculate about potential causes for this, but simply want to conclude that integration into the host genome—though it may lead to a more efficient response—is not an absolute prerequisite for the generation of endo-siRNAs.

Spliced mRNA and antisense transcripts can each provide one strand for the duplex endo-siRNA precursor

A detailed analysis of the small RNA population can give insights into the nature of its precursor molecules. We combined all of the transgene-containing deep sequencing libraries discussed in the paper (63N1, pKF63 transf., 63-6 mock, 63-6 β-elim., tagged Ago2-IP) into one dataset, then analysed the endo-siRNA population derived from the mini-white gene present in all constructs. This marker gene is not derived from cDNA and thus still contains introns. Comparison of small RNA occurrence with the exon–intron structure showed that the exonic parts of the gene have a higher propensity to generate small RNAs than the intronic parts (depicted in Figure 5, top panel); whereas 38 620 reads were found from the entire mini-white gene (3360 nt), 34 504 reads mapped to the spliced white mRNA (2212 nt). As a 34% reduction of the sequence length reduced small RNA occurrence by only 10.1%, the intron sequences must be less likely to generate endo-siRNAs. The simplest explanation for this is that in most cases, spliced white mRNA is providing one strand of the endo-siRNA precursor.

Figure 5
The endo-siRNA population lacks the sequences that cross exon–exon junctions. Small RNA reads were pooled from five different libraries (63N1, pKF63 transf., 63-6 mock, 63-6 β-el., IP transfected Ago2) and the density of both sense and ...

If mRNA can provide the sense strand, then how is the antisense strand generated? In principle, three mechanisms can be envisioned: either the mRNA serves as a template for generation of the antisense strand in an RNA-dependent RNA polymerase (RdRP)-like manner; a processed pseudogene gives rise to antisense transcripts as described for mouse oocytes (Tam et al, 2008; Watanabe et al, 2008); or antisense transcription of the DNA is occurring and the two single-stranded RNAs form a duplex only after their production. The Drosophila genome contains few pseudogenes (Harrison et al, 2003) and a BLAST search at low stringency showed no potential pseudogene for white (data not shown). Thus we could test whether an RdRP activity generates the dsRNA precursor by examining endo-siRNAs covering the exon–exon junctions of the white gene: while an RdRP-like mechanism will produce double-stranded RNA along the entire transcript, antisense transcripts of the mini-white marker gene should not be spliced correspondingly and thus exon–exon junctions in the spliced sense strand cannot anneal with a contiguous antisense transcript. As a consequence, endo-siRNAs should only be produced from within an exon (i.e. not across an exon–exon junction). We, therefore, examined the small RNA density along the spliced white mRNA and compared this data with the location of the exon–exon junctions (Figure 4, bottom panel). In five out of five cases, the exon–exon junctions are preceded by a 15–20 nt region that is essentially devoid of endo-siRNA reads. This is consistent with a lack of endo-siRNAs that cross the exon–exon junction and, consequently, the idea that antisense transcription is providing the second strand to form the duplex precursor for endo-siRNAs.


We present an analysis that extends and refines our understanding of a small-RNA-mediated host defense system against selfish genetic elements in somatic cells. We have identified a new isoform of the Dicer partner Loquacious, thus explaining how endo-siRNA and miRNA biogenesis are kept distinct. Furthermore, we show that the endo-siRNA response can be initiated de novo and that artificial sequences can be targeted without a need for integration into the host genome.

Distinct Loqs isoforms separate the biogenesis routes for endo-siRNAs and miRNAs

The discovery of a new Loqs protein isoform as a central player in endo-siRNA biogenesis sheds light onto the question of how a single gene—loqs—can participate in the recognition of pre-miRNA structures together with Dcr-1 and dsRNA together with Dcr-2 (Czech et al, 2008; Okamura et al, 2008b): a different domain content and the Loqs-PD specific amino acids could well be responsible for the distinct properties. For example, the Loqs-related RDE-4 protein from Caenorhabditis elegans binds long double-stranded RNA cooperatively and shows a length-dependent decrease in binding affinity between a 650-nt fragment of dsRNA and an siRNA in vitro. This decrease is exacerbated by the deletion of the carboxy-terminus, starting at the end of the second dsRBD (Parker et al, 2008). Truncated RDE-4 does not form homodimers any more and cannot sustain the production of siRNAs in vitro (Parker et al, 2006). It is interesting to note that in humans, the loqs homologue TRBP also gives rise to, at least, three alternative splice variants, one of them lacking the third dsRBD and thus resembling Loqs-PD/PC (Haase et al, 2005).

The isoform-specific knockdown experiments in our study show that all three, presently known, Drosophila small RNA biogenesis pathways in somatic cells are defined by a distinct set of required factors and can be genetically separated: Dcr-1 and Loqs-PB for miRNAs (including miRtrons), R2D2 for siRNAs—even those derived from transgenic hairpin constructs (Forstemann et al, 2005)—and Loqs-PD for endo-siRNAs (Figure 6A). It should be noted, however, that this exclusively linear model is to some extent an oversimplification. Certain miRNAs can become incorporated into Ago2 complexes (Forstemann et al, 2007; Seitz et al, 2008), Loqs and Dcr-1 have a minor, but detectable, role in transgenic RNAi (Lee et al, 2004; Forstemann et al, 2005) and endo-siRNAs can persist to a certain extent in the absence of ago2 or loqs (Czech et al, 2008; Okamura et al, 2008b). It is also not known whether Loqs-PD is required for the processing of endo-siRNA precursors or for the Ago2 loading step, or for both. From the perspective of the Loqs protein isoforms, we now know that Loqs-PB participates in miRNA biogenesis (Forstemann et al, 2005; Jiang et al, 2005; Saito et al, 2005) and Loqs-PD participates in endo-siRNA biogenesis. This leaves the open question for the substrates of the Loqs-PA isoform.

Figure 6
Small RNA biogenesis pathways and generation of endo-siRNA precursors. (A) All somatic small RNA biogenesis pathways can be distinguished by, at least, one specific component. The model represents important biogenesis steps for miRNAs, endo-siRNAs and ...

Endo-siRNAs target multicopy sequences on the mRNA level

Using a nascent RNA labelling and purification strategy, we could show that endo-siRNAs directed against a transgene construct predominantly affect the degradation rate of the corresponding transcripts and do not impose transcriptional regulation. This post-transcriptional mechanism is fully consistent with previous experiments in which target-RNA cleavage mediated by endo-siRNA was directly detected in vitro (Kawamura et al, 2008; Okamura et al, 2008b) and in vivo (Czech et al, 2008). Our finding extends these observations to the statement that a post-transcriptional mechanism is sufficient to explain most—if not all—of the repression mediated by endo-siRNAs. A previous study of a GFP transgene expressed through multicopy plasmid insertion in Drosophila S2 cells concluded that on knock down of dcr-2 or ago2, the transgene mRNA has a longer half-life and carries a shorter poly-A tract (Siomi et al, 2005). As the depletion of dcr-2 or ago2 resulted in more abundant GFP mRNA, but with shorter poly-A tails, the silencing mechanism(s) employed by endo-siRNAs may not rely solely on homology-mediated cleavage but also involve other aspects of RNA metabolism.

At this point, we cannot exclude that endo-siRNAs cause a minor extent of transcriptional silencing that lies below our distinction limit. Furthermore, it is formally possible that transcriptional silencing may need a longer time to revert than what is achievable with RNAi in cell culture. However, a direct comparison showed overall very similar levels of transposon transcript accumulation between dcr-2 RNAi in cultured cells and dcr-2L811fsX mutant flies (Chung et al, 2008). In summary there is substantial support for a predominantly post-transcriptional silencing mechanism of the GFP transgene. At least one natural endo-siRNA target, the transposon mdg1, shares this predominance of post-transcriptional silencing and we speculate that this may generally be the case for endo-siRNA-mediated repression. This is not to say that there is no transcriptional silencing of transposable elements in Drosophila, but we propose that this mechanism depends on a distinct pathway.

The endo-siRNA pathway can initiate a de novo response against invading genetic elements

The analysis of small RNA libraries from various sources has led to the conclusion that endo-siRNAs derive from double-stranded precursors (Chung et al, 2008; Czech et al, 2008; Ghildiyal et al, 2008; Kawamura et al, 2008; Okamura et al, 2008a). The origin of this double-stranded RNA is obvious in the case of hairpin-forming sequences or regions with convergent transcription, though small RNA generation does not necessarily occur in all cases in which convergent transcription takes place (Okamura et al, 2008a). Our stably integrated GFP expression vector has allowed us to carry out fine mapping of endo-siRNA reads. Convergent transcription of a repetitive element, initiated by surrounding promotors or between neighbouring elements in a head-to-head arrangement, should result in a more or less even distribution of endo-siRNAs across the entire sequence. In contrast, the observation that endo-siRNAs are preferentially produced from transcribed and spliced mRNAs suggests an additional or alternative model in which distinct mechanisms generate each strand of the dsRNA precursor: the sense strand can be contributed by the normal mRNAs, whereas the opposite strand arises from antisense transcription. Alternatively, the mRNA could be converted to dsRNA by the action of an RdRP. Even though Drosophila does not have a gene coding for a canonical RdRP, an RdRP-like activity has been described (Lipardi et al, 2001, 2005; Nishikura, 2001; Wei et al, 2003), and in principle the mRNA-synthesizing enzyme RNA Pol II cannot only use DNA but also RNA as a template (Lehmann et al, 2007). Our finding that endo-siRNA sequences never cross an exon–exon junction rules out any RdRP-like mechanisms. Although we can exclude the involvement of a processed pseudogene in the case of white because it simply does not exist in the genome, the silencing of other natural endo-siRNA targets may be augmented by the presence of corresponding pseudogenes as described for mouse cells (Tam et al, 2008; Watanabe et al, 2008). Trans-acting sequences were also not required in the case of the artificial plasmid sequences giving rise to small RNAs, as it is impossible that a corresponding sequence is already present elsewhere in the genome. Together, our results lend support to the hypothesis that antisense-oriented transcripts arising from the repetitive element itself base pair with the sense transcript (Figure 6B).

As a consequence, antisense transcription is one of the key factors needed for an endo-siRNA response. One possibility to generate antisense RNA is the pervasive transcription of intergenic regions that has been detected in Drosophila and other species (for review, see Kapranov et al, 2007). This ‘noise' in transcriptional initiation can make a substantial contribution to the cellular RNA pool (Struhl, 2007) and essentially leads to a low level of transcription along most of the genome. In Saccharomyces cerevisiae, most promotors were even shown to generate a low level of transcription in the antisense direction relative to their targets (Neil et al, 2009). If these antisense transcription events are proportional to the copy number and—in the case of single-copy genes—below a threshold to allow significant production of dsRNA with their cognate mRNA transcripts (a reasonable assumption given the low abundance of antisense transcripts), then sequences with a high-copy number can be identified by the endo-siRNA system simply based on the amount of antisense transcripts produced. In other words, the products of RNA polymerases reading in the antisense direction may serve to integrate over multiple occurrences and thus ‘count' the copy number.

Genetic studies can provide an estimate of where the response threshold lies in Drosophila: at least six copies of an adh-white transgene were necessary to induce silencing and the production of 21–25 nt small RNAs (Pal-Bhadra et al, 2002). Furthermore, a putative self-regulation mechanism was proposed that restricts the copy number of active I-elements to 10–15 copies several generations after introduction (Vaury et al, 1993). Interestingly, resistance against the I-element transposon could be induced by the artificial introduction of partial I-element sequences. The efficiency increased with higher copy number of the I-element fragment and with transcription of the I-element sequence, but the orientation of transcription (sense versus antisense) did not matter (Jensen et al, 1999a, 1999b). Altogether, this correlates well with our finding that about eight integrated plasmid copies lead to an endo-siRNA response, which is just above our detection limit.

Do the endo-siRNA and the piRNA silencing systems interact?

In the germ line, the piRNA system is responsible for repressing mobile genetic elements but it is unclear how de novo initiation of a piRNA response occurs; the fact that piRNAs must be inherited maternally through the oocyte cytoplasm to control transposable elements argues that such a response cannot be initiated easily and requires several generations to establish (Jensen et al, 1999b; Blumenstiel and Hartl, 2005; Brennecke et al, 2008). Analysis of endo-siRNA and piRNA origin suggests that in certain cases the same genomic region gives rise to both endo-siRNAs and piRNAs (Chung et al, 2008; Czech et al, 2008; Ghildiyal et al, 2008; Kawamura et al, 2008), notably the flamenco cluster. The gypsy element can become mobilized in a cycle that involves generation of proviral elements in somatic cells, followed by their transfer into female germ cells and integration into the F1 genome (Chalvet et al, 1999; Pelisson et al, 2002). This ‘lifestyle' is paralleled in the repression by both piRNAs and endo-siRNAs (Vagin et al, 2006; Brennecke et al, 2007; Gunawardane et al, 2007; Chung et al, 2008; Czech et al, 2008; Ghildiyal et al, 2008; Kawamura et al, 2008). Ghildiyal et al (2008) detected piRNA-sized small RNAs in ago2 null mutant somatic tissue, whereas Czech et al (2008) found endo-siRNAs also in the gonads. The discovery that repression of flamenco targets in the somatic follicle cells does not depend on a ping-pong mechanism between Piwi family proteins showed a distinct biogenesis pathway for piRNAs targeting this class of transposable elements (Li et al, 2009; Malone et al, 2009). Our results cannot contribute further mechanistic information on a potential connection between piRNAs and endo-siRNAs, but we have shown that, at least, the endo-siRNA pathway has the capacity to mount a de novo response against any invading selfish genetic element in somatic cells. Thus, one hypothetical way to initiate a piRNA-dependent repression might be through the intermediate of an endo-siRNA response.

Materials and methods

Cell culture, RNAi and FACS analysis

Drosophila Schneider 2 (S2) cells were cultured in Schneider's medium (Bio&Sell, Nürnberg, Germany) containing 10% fetal calf serum (Thermo), transfected and treated to induce RNAi essentially as described (Shah and Forstemann, 2008). The dsRNA triggers were either previously described (Forstemann et al, 2005) or the following oligonucleotides were used to obtain new gene fragments: for dsloqs5′UTR 5′-CGTAATACGACTCACTATAGGGCAACCACAAATATCAGT-3′ and 5′-CGTAATACGACTCACTATAGGTTGCACGGTTTTC GGGAG-3′; for dsloqs3′UTR 5′-CGTAATACGACTCACTATAGGGTGCAGCCAACTGAATAGCA-3′ and 5′-CGTAATACGACTCACTATAGGCCTTCGCAAACTAGCACGTAG-3′; for dsGFP 5′-CGTAATACGACTCACTATAGGATGGTGAGCAAGGGCGAGGAGCTG-3′ and 5′-CGTAATACGACTCACTATAGGTTACTTGTACAGCTCGTCCATG-3′; for dsloqsRC 5′-CGTAATACGACTCACTATAGGAACGAATCTGTAAAGCACCT-3′ and 5′-CGTAATACGACTCACTATAGGCTGTAAATAAGAGCGCAAAG-3′; for dsloqsRB+RC 5′-CGTAATACGACTCACTATAGGAACGAATCTGTAAAGCACCT-3′ and 5′-CGTAATACGACTCACTATAGGCTGTAACTTAAGCAGTTTTTTGCC-3′; for dsloqsRD 5′-CGTAATACGACTCACTATGTGAGTATCATTCAAGACATCGATC-3′ and 5′-CGTAATACGACTCACTATAGGTAAGGTGTAAGCATTATGTTAATT-3′; for dsloqsRC+RD 5′-CGTAATACGACTCACTATGTGAGTATCATTCAAGACATCGATC-3′ and 5′-CGTAATACGACTCACTATAGGCTGTAAATAAGAGCGCAAAG-3′; for dsDsRed 5′-CGTAATACGACTCACTATAGGAGGACGGCTGCTTCATCTAC-3′ and 5′-CGTAATACGACTCACTATAGGTGGTGTAGTCCTCGTTGTGG-3′. Cells were seeded at 0.5 × 106 cells/ml and 20 μg/ml dsRNA was added to the medium. On day 2, the cells were split 1:5 into a fresh culture dish and the dsRNA treatment was repeated. On day 6, GFP fluorescence was quantified in a Becton Dickinson FACSCalibur flow cytometer. The pKF63 expression construct was used as published (Forstemann et al, 2007).

Sample preparation and bioinformatic analysis of deep sequencing data

Small RNAs were isolated and libraries for Solexa sequencing were prepared essentially as described by Czech et al (2008), a detailed protocol is available on request. Solexa sequencing was carried out at Fasteris (Plan-Les-Ouates, Switzerland). For previously published deep sequencing data, the FASTA files were retrieved from NCBI GEO (GSM239041, GSM239050, GSM239051, GSM239052, GSM280089) and mapped onto the target sequences using BOWTIE (http://bowtie-bio.sourceforge.net/) with the option –v0 to force selection of only perfectly matching sequences. Pre-processing of sequences and analysis of the BOWTIE output files were done using PERL scripts (available on request). The sequence data are available at NCBI GEO under the accession number GSE16958.

Labelling and purification of newly synthesized RNA, qRT–PCR

S2 cells were treated with dsRNA directed against dcr-2 or DsRed in 10 -cm culture dishes as described above, except that the dsRNA concentration was 3 μg/ml. The general approach for RNA labelling, biotinylation and purification has been previously described by Dolken et al (2008), the final concentration of 4-thio-uridine was 100 μM during the 60 min incubation step (applied in Schneider's medium with 10% fetal calf serum). RNA was extracted with TRIzol (Invitrogen) and 100 μg of total RNA was used for biotinylation. Reverse transcription was primed with random hexamers (Eurofins-MWG) and carried out on 100 ng of RNA with Superscript II (Invitrogen). qRT–PCR was carried out on an ABI Prism 7000 sequence detection system (Applied Biosystems) and expression was quantified with the 2−(ΔΔCt) method (Livak and Schmittgen, 2001). Primers used for amplification are: hsp70 s 5′-AGGACTTTGACAACCGGCTA-3′ and hsp70 as 5′-ACAGTGCGTCAATCTCGATG-3′; dcr-2 s 5′-GTTCCGCTTTGGTCAACAAT-3′ and dcr-2 as 5′-GGCTGAACATCAGCTTCCTC-3′; white s 5′-CTA ATATCCTGCGCCAGCTC-3′ and white as 5′-ACGGAACCATGAGAGGTACG-3′; GFP s 5′-ACGTAAACGGCCACAAGTTC-3′ and GFP as 5′-AAGTCGTGCTGCTTCATGTG-3′; gapdh s 5′-CTTCTTCAGCGACACCCATT-3′ and gapdh as 5′-ACCGAACTCGTTGTCGTACC-3′; primers for mdg1, 297 and rp49 were as previously described by Chung et al (2008).

3′ RACE PCR analysis of the loqs-RD variant

RNA from S2 cells was prepared using the miRNeasy mini kit (Qiagen, Hilden, Germany). An amount equal to 1 μg of total RNA was reverse transcribed (Superscript II, Invitrogen) using the oligo(dT) primer from the GeneRACER Kit (Invitrogen) 5′-GCTGTCAACGATACGCTACGTAACGGCATGACAGTG (T)18-3′. PCR was carried out using a primer at the end of loqs exon 3 (5′-CAAGGATCCAATGCCACAGGCGGAGGAGAT-3′) and the 3′-end reverse primer from the GeneRACER kit (5′-GCTGTCAACGATACGCTACGTAACG-3′) using a 5:1 mix of Taq and Pfu (Fermentas) polymerases and a 1:20 diluted cDNA template at 55°C annealing temperature.

Copy number determination by qPCR

Quantitative PCR analysis was carried out with SYBR green mixes (Dynamo-Flash, Finnzymes) using 10 ng of genomic DNA as template. The primers for GAPDH were as described above; for the ubiquitin promotor: 5′-GCCGGTAGAGAAGACAGTGC-3′ and 5′-ACTGACTTGACCGGCTGAAT-3′; for CG5599: 5′-CTCCCGGTACTAACGTTCCA-3′ and 5′TTGCATCAACTGGGTCATGT-3′; for CG1673: 5′-ATGAACATGAACCGCATGAA-3′ and 5′-GGCTGAGGATCGTGTAGAGC-3′.

Protein extraction, co-immunoprecipitation and western blotting

The myc-tagged constructs for Loqs-PB and Loqs-PA have been described previously by Forstemann et al (2005), the construct for myc-tagged Loqs-PD was made analogously with 5′-AGCGGATCCATGGAACAAAAACTTATTTCTGAAGAAGACTTGGAA CAAAAACTTATTTCTGAAGAAGACTTGGCCAAGAACACCATGGACCAGGAG-3′ as sense and 5′-CAAAGCGGCCGCTTAGATCTTGATGAACTC-3′ as the antisense primer. Myc-tagged R2D2 was amplified from cDNA with the primers 5′-AGCGGATCCATGGAACAAAAACTTATTTCTGAAGAAGACTTG GAACAAAAACTTATTTCTGAAGAAGACTTGGCCCTTGAACTCATGGATAACAAG-3′ and 5′-CAAAGCGGCCGCTTATACGCATTAAATCAA-3′ and inserted into the same expression vector as the other plasmids (Forstemann et al, 2005).

Cells were collected and washed twice in PBS (Invitrogen). The pellet was re-suspended in lysis buffer (30 mM HEPES, 100 mM KOAc, 2 mM MgCl2, 1 mM fresh DTT, 1% (v/v) Triton X-100 (Sigma), 2 × protease inhibitor cocktail (complete without EDTA, Roche Molecular Biochemicals; Basel, Switzerland)) and frozen in liquid nitrogen. Samples were thawed on ice and cell debris was pelleted in a refrigerated microcentrifuge at 15 000 g (Eppendorf). Protein concentrations were determined by Bradford Assay (Bio-Rad, Hercules, CA, USA).

For immunoprecipitation, 1 mg total protein was incubated with 50 μl α-myc affinity agarose (A7470, Sigma; washed three times in 1 ml of lysis buffer) for 2 h at 4°C on an overhead rotator. Flow through and beads were separated by spin columns (MoBiTec, Göttingen, Germany) and washed four times with 500 μl lysis buffer. Bound proteins were eluted by applying 15 μl 1 × SDS sample buffer and heating to 95°C for 5 min. Western blotting was carried out as previously described by Forstemann et al (2007); α-Loqs antibody (Miyoshi et al, 2009) was diluted to 1:4000; α-Dcr-2 antibody (Abcam ab4732-100) was diluted to 1:1000.

RNA extraction and northern blotting

RNA was extracted with TRIzol and probes for the detection of bantam miRNA and 2S rRNA were used as described (Forstemann et al, 2005). We used a DNA antisense probe for the detection of CG4068 B endo-siRNA (Okamura et al, 2008b).

Supplementary Material

Supplementary Figure 1–5

Review Process File


We thank L Dölken and B Rädle for help with the nascent RNA labeling procedure; M Bayer, G Meister, D Martin, P Zamore and M Siomi for critical comments on the paper. This study was supported by DFG grant FO-360/2, Sonderforschungsbereich SFB646, BMBF grant NGFN-plus 01GS0801 and an ‘LMU excellent' grant awarded to KF; JVH is the recipient of a Boehringer-Ingelheim Fonds PhD Scholarship and KF is the recipient of a Career Development Award from the International Human Frontier Science Program Organization.


The authors declare that they have no conflict of interest.


  • Aravin AA, Hannon GJ, Brennecke J (2007) The Piwi-piRNA pathway provides an adaptive defense in the transposon arms race. Science 318: 761–764 [PubMed]
  • Babiarz JE, Ruby JG, Wang Y, Bartel DP, Blelloch R (2008) Mouse ES cells express endogenous shRNAs, siRNAs, and other Microprocessor-independent, Dicer-dependent small RNAs. Genes Dev 22: 2773–2785 [PMC free article] [PubMed]
  • Blumenstiel JP, Hartl DL (2005) Evidence for maternally transmitted small interfering RNA in the repression of transposition in Drosophila virilis. Proc Natl Acad Sci USA 102: 15965–15970 [PMC free article] [PubMed]
  • Brennecke J, Aravin AA, Stark A, Dus M, Kellis M, Sachidanandam R, Hannon GJ (2007) Discrete small RNA-generating loci as master regulators of transposon activity in Drosophila. Cell 128: 1089–1103 [PubMed]
  • Brennecke J, Malone CD, Aravin AA, Sachidanandam R, Stark A, Hannon GJ (2008) An epigenetic role for maternally inherited piRNAs in transposon silencing. Science 322: 1387–1392 [PMC free article] [PubMed]
  • Chalvet F, Teysset L, Terzian C, Prud'homme N, Santamaria P, Bucheton A, Pelisson A (1999) Proviral amplification of the Gypsy endogenous retrovirus of Drosophila melanogaster involves env-independent invasion of the female germline. EMBO J 18: 2659–2669 [PMC free article] [PubMed]
  • Chambeyron S, Bucheton A (2005) I elements in Drosophila: in vivo retrotransposition and regulation. Cytogenet Genome Res 110: 215–222 [PubMed]
  • Chung WJ, Okamura K, Martin R, Lai EC (2008) Endogenous RNA interference provides a somatic defense against Drosophila transposons. Curr Biol 18: 795–802 [PMC free article] [PubMed]
  • Core LJ, Waterfall JJ, Lis JT (2008) Nascent RNA sequencing reveals widespread pausing and divergent initiation at human promoters. Science 322: 1845–1848 [PMC free article] [PubMed]
  • Czech B, Malone CD, Zhou R, Stark A, Schlingeheyde C, Dus M, Perrimon N, Kellis M, Wohlschlegel JA, Sachidanandam R, Hannon GJ, Brennecke J (2008) An endogenous small interfering RNA pathway in Drosophila. Nature 453: 798–802 [PMC free article] [PubMed]
  • Dolken L, Ruzsics Z, Radle B, Friedel CC, Zimmer R, Mages J, Hoffmann R, Dickinson P, Forster T, Ghazal P, Koszinowski UH (2008) High-resolution gene expression profiling for simultaneous kinetic parameter analysis of RNA synthesis and decay. RNA 14: 1959–1972 [PMC free article] [PubMed]
  • Fejes-Toth KSV, Sachidanandam R, Assaf G, Hannon GJ, Kapranov P, Foissac S, Willingham AT, Duttagupta R, Dumais E, Gingeras T (2009) Post-transcriptional processing generates a diversity of 5′-modified long and short RNAs. Nature 457: 1028–1032 [PMC free article] [PubMed]
  • Forstemann K, Horwich MD, Wee L, Tomari Y, Zamore PD (2007) Drosophila microRNAs are sorted into functionally distinct Argonaute complexes after production by dicer-1. Cell 130: 287–297 [PMC free article] [PubMed]
  • Forstemann K, Tomari Y, Du T, Vagin VV, Denli AM, Bratu DP, Klattenhoff C, Theurkauf WE, Zamore PD (2005) Normal microRNA maturation and germ-line stem cell maintenance requires Loquacious, a double-stranded RNA-binding domain protein. PLoS Biol 3: e236. [PMC free article] [PubMed]
  • Ghildiyal M, Seitz H, Horwich MD, Li C, Du T, Lee S, Xu J, Kittler EL, Zapp ML, Weng Z, Zamore PD (2008) Endogenous siRNAs derived from transposons and mRNAs in Drosophila somatic cells. Science 320: 1077–1081 [PMC free article] [PubMed]
  • Golden DE, Gerbasi VR, Sontheimer EJ (2008) An inside job for siRNAs. Mol Cell 31: 309–312 [PMC free article] [PubMed]
  • Gunawardane LS, Saito K, Nishida KM, Miyoshi K, Kawamura Y, Nagami T, Siomi H, Siomi MC (2007) A slicer-mediated mechanism for repeat-associated siRNA 5′ end formation in Drosophila. Science 315: 1587–1590 [PubMed]
  • Haase AD, Jaskiewicz L, Zhang H, Laine S, Sack R, Gatignol A, Filipowicz W (2005) TRBP, a regulator of cellular PKR and HIV-1 virus expression, interacts with Dicer and functions in RNA silencing. EMBO Rep 6: 961–967 [PMC free article] [PubMed]
  • Harrison PM, Milburn D, Zhang Z, Bertone P, Gerstein M (2003) Identification of pseudogenes in the Drosophila melanogaster genome. Nucleic Acids Res 31: 1033–1037 [PMC free article] [PubMed]
  • Hartig JV, Tomari Y, Forstemann K (2007) piRNAs—the ancient hunters of genome invaders. Genes Dev 21: 1707–1713 [PubMed]
  • He Y, Vogelstein B, Velculescu VE, Papadopoulos N, Kinzler KW (2008) The antisense transcriptomes of human cells. Science 322: 1855–1857 [PMC free article] [PubMed]
  • Horwich MD, Li C, Matranga C, Vagin V, Farley G, Wang P, Zamore PD (2007) The Drosophila RNA methyltransferase, DmHen1, modifies germline piRNAs and single-stranded siRNAs in RISC. Curr Biol 17: 1265–1272 [PubMed]
  • Jensen S, Gassama MP, Heidmann T (1999a) Cosuppression of I transposon activity in Drosophila by I-containing sense and antisense transgenes. Genetics 153: 1767–1774 [PMC free article] [PubMed]
  • Jensen S, Gassama MP, Heidmann T (1999b) Taming of transposable elements by homology-dependent gene silencing. Nat Genet 21: 209–212 [PubMed]
  • Jiang F, Ye X, Liu X, Fincher L, McKearin D, Liu Q (2005) Dicer-1 and R3D1-L catalyze microRNA maturation in Drosophila. Genes Dev 19: 1674–1679 [PMC free article] [PubMed]
  • Johnson TR, Rudin SD, Blossey BK, Ilan J (1991) Newly synthesized RNA: simultaneous measurement in intact cells of transcription rates and RNA stability of insulin-like growth factor I, actin, and albumin in growth hormone-stimulated hepatocytes. Proc Natl Acad Sci USA 88: 5287–5291 [PMC free article] [PubMed]
  • Kapranov P, Willingham AT, Gingeras TR (2007) Genome-wide transcription and the implications for genomic organization. Nat Rev Genet 8: 413–423 [PubMed]
  • Kawamura Y, Saito K, Kin T, Ono Y, Asai K, Sunohara T, Okada TN, Siomi MC, Siomi H (2008) Drosophila endogenous small RNAs bind to Argonaute 2 in somatic cells. Nature 453: 793–797 [PubMed]
  • Kenzelmann M, Maertens S, Hergenhahn M, Kueffer S, Hotz-Wagenblatt A, Li L, Wang S, Ittrich C, Lemberger T, Arribas R, Jonnakuty S, Hollstein MC, Schmid W, Gretz N, Grone HJ, Schutz G (2007) Microarray analysis of newly synthesized RNA in cells and animals. Proc Natl Acad Sci USA 104: 6164–6169 [PMC free article] [PubMed]
  • Lee H-C, Chang S-S, Choudhary S, Aalto AP, Maiti M, Bamford DH, Liu Y (2009) qiRNA is a new type of small interfering RNA induced by DNA damage. Nature 459: 274–277 [PMC free article] [PubMed]
  • Lee YS, Nakahara K, Pham JW, Kim K, He Z, Sontheimer EJ, Carthew RW (2004) Distinct roles for Drosophila Dicer-1 and Dicer-2 in the siRNA/miRNA silencing pathways. Cell 117: 69–81 [PubMed]
  • Lehmann E, Brueckner F, Cramer P (2007) Molecular basis of RNA-dependent RNA polymerase II activity. Nature 450: 445–449 [PubMed]
  • Li C, Vagin VV, Lee S, Xu J, Ma S, Xi H, Seitz H, Horwich MD, Syrzycka M, Honda BM, Kittler EL, Zapp ML, Klattenhoff C, Schulz N, Theurkauf WE, Weng Z, Zamore PD (2009) Collapse of germline piRNAs in the absence of Argonaute3 reveals somatic piRNAs in flies. Cell 137: 509–521 [PMC free article] [PubMed]
  • Lipardi C, Baek HJ, Wei Q, Paterson BM (2005) Analysis of short interfering RNA function in RNA interference by using Drosophila embryo extracts and Schneider cells. Methods Enzymol 392: 351–371 [PubMed]
  • Lipardi C, Wei Q, Paterson BM (2001) RNAi as random degradative PCR: siRNA primers convert mRNA into dsRNAs that are degraded to generate new siRNAs. Cell 107: 297–307 [PubMed]
  • Liu Q, Rand TA, Kalidas S, Du F, Kim HE, Smith DP, Wang X (2003) R2D2, a bridge between the initiation and effector steps of the Drosophila RNAi pathway. Science 301: 1921–1925 [PubMed]
  • Livak KJ, Schmittgen TD (2001) Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) method. Methods 25: 402–408 [PubMed]
  • Malone CD, Brennecke J, Dus M, Stark A, McCombie WR, Sachidanandam R, Hannon GJ (2009) Specialized piRNA pathways act in germline and somatic tissues of the Drosophila ovary. Cell 137: 522–535 [PMC free article] [PubMed]
  • Malone CD, Hannon GJ (2009) Small RNAs as guardians of the genome. Cell 136: 656–668 [PMC free article] [PubMed]
  • Miyoshi K, Okada TN, Siomi H, Siomi MC (2009) Characterization of the miRNA–RISC loading complex and miRNA–RISC formed in the Drosophila miRNA pathway. RNA 15: 1282–1291 [PMC free article] [PubMed]
  • Neil H, Malabat C, d'Aubenton-Carafa Y, Xu Z, Steinmetz LM, Jacquier A (2009) Widespread bidirectional promoters are the major source of cryptic transcripts in yeast. Nature 457: 1038–1042 [PubMed]
  • Nishikura K (2001) A short primer on RNAi: RNA-directed RNA polymerase acts as a key catalyst. Cell 107: 415–418 [PubMed]
  • O'Donnell KA, Boeke JD (2007) Mighty Piwis defend the germline against genome intruders. Cell 129: 37–44 [PMC free article] [PubMed]
  • Okamura K, Balla S, Martin R, Liu N, Lai EC (2008a) Two distinct mechanisms generate endogenous siRNAs from bidirectional transcription in Drosophila melanogaster. Nat Struct Mol Biol 15: 581–590 [PMC free article] [PubMed]
  • Okamura K, Chung WJ, Ruby JG, Guo H, Bartel DP, Lai EC (2008b) The Drosophila hairpin RNA pathway generates endogenous short interfering RNAs. Nature 453: 803–806 [PMC free article] [PubMed]
  • Orgel LE, Crick FH (1980) Selfish DNA: the ultimate parasite. Nature 284: 604–607 [PubMed]
  • Pace JK II, Feschotte C (2007) The evolutionary history of human DNA transposons: evidence for intense activity in the primate lineage. Genome Res 17: 422–432 [PMC free article] [PubMed]
  • Pal-Bhadra M, Bhadra U, Birchler JA (2002) RNAi related mechanisms affect both transcriptional and posttranscriptional transgene silencing in Drosophila. Mol Cell 9: 315–327 [PubMed]
  • Park JK, Liu X, Strauss TJ, McKearin DM, Liu Q (2007) The miRNA pathway intrinsically controls self-renewal of Drosophila germline stem cells. Curr Biol 17: 533–538 [PubMed]
  • Parker GS, Eckert DM, Bass BL (2006) RDE-4 preferentially binds long dsRNA and its dimerization is necessary for cleavage of dsRNA to siRNA. RNA 12: 807–818 [PMC free article] [PubMed]
  • Parker GS, Maity TS, Bass BL (2008) dsRNA binding properties of RDE-4 and TRBP reflect their distinct roles in RNAi. J Mol Biol 384: 967–979 [PMC free article] [PubMed]
  • Pelisson A, Mejlumian L, Robert V, Terzian C, Bucheton A (2002) Drosophila germline invasion by the endogenous retrovirus gypsy: involvement of the viral env gene. Insect Biochem Mol Biol 32: 1249–1256 [PubMed]
  • Petersen R, Lindquist S (1988) The Drosophila hsp70 message is rapidly degraded at normal temperatures and stabilized by heat shock. Gene 72: 161–168 [PubMed]
  • Pham JW, Pellino JL, Lee YS, Carthew RW, Sontheimer EJ (2004) A Dicer-2-dependent 80S complex cleaves targeted mRNAs during RNAi in Drosophila. Cell 117: 83–94 [PubMed]
  • Preker P, Nielsen J, Kammler S, Lykke-Andersen S, Christensen MS, Mapendano CK, Schierup MH, Jensen TH (2008) RNA exosome depletion reveals transcription upstream of active human promoters. Science 322: 1851–1854 [PubMed]
  • Saito K, Ishizuka A, Siomi H, Siomi MC (2005) Processing of pre-microRNAs by the Dicer-1-Loquacious complex in Drosophila cells. PLoS Biol 3: e235. [PMC free article] [PubMed]
  • Seila AC, Calabrese JM, Levine SS, Yeo GW, Rahl PB, Flynn RA, Young RA, Sharp PA (2008) Divergent transcription from active promoters. Science 322: 1849–1851 [PMC free article] [PubMed]
  • Seitz H, Ghildiyal M, Zamore PD (2008) Argonaute loading improves the 5′ precision of both microRNAs and their miRNA strands in flies. Curr Biol 18: 147–151 [PMC free article] [PubMed]
  • Shah C, Forstemann K (2008) Monitoring miRNA-mediated silencing in Drosophila melanogaster S2-cells. Biochim Biophys Acta 1779: 766–772 [PubMed]
  • Siomi H (2008) Transposable elements, RNA silencing, and their impacts on the genome throughout evolution. Uirusu 58: 55–60 [PubMed]
  • Siomi MC, Tsukumo H, Ishizuka A, Nagami T, Siomi H (2005) A potential link between transgene silencing and poly(A) tails. RNA 11: 1004–1011 [PMC free article] [PubMed]
  • Struhl K (2007) Transcriptional noise and the fidelity of initiation by RNA polymerase II. Nat Struct Mol Biol 14: 103–105 [PubMed]
  • Taft RJ, Glazov EA, Cloonan N, Simons C, Stephen S, Faulkner GJ, Lassmann T, Forrest AR, Grimmond SM, Schroder K, Irvine K, Arakawa T, Nakamura M, Kubosaki A, Hayashida K, Kawazu C, Murata M, Nishiyori H, Fukuda S, Kawai J et al. (2009) Tiny RNAs associated with transcription start sites in animals. Nat Genet 41: 572–578 [PubMed]
  • Tam OH, Aravin AA, Stein P, Girard A, Murchison EP, Cheloufi S, Hodges E, Anger M, Sachidanandam R, Schultz RM, Hannon GJ (2008) Pseudogene-derived small interfering RNAs regulate gene expression in mouse oocytes. Nature 453: 534–538 [PMC free article] [PubMed]
  • Tomari Y, Du T, Zamore PD (2007) Sorting of Drosophila small silencing RNAs. Cell 130: 299–308 [PMC free article] [PubMed]
  • Tomari Y, Matranga C, Haley B, Martinez N, Zamore PD (2004) A protein sensor for siRNA asymmetry. Science 306: 1377–1380 [PubMed]
  • Vagin VV, Sigova A, Li C, Seitz H, Gvozdev V, Zamore PD (2006) A distinct small RNA pathway silences selfish genetic elements in the germline. Science 313: 320–324 [PubMed]
  • van Oers MM, Vlak JM (2007) Baculovirus genomics. Curr Drug Targets 8: 1051–1068 [PubMed]
  • Vaury C, Pelisson A, Abad P, Bucheton A (1993) Properties of transgenic strains of Drosophila melanogaster containing I transposable elements from Drosophila teissieri. Genet Res 61: 81–90 [PubMed]
  • Wang XH, Aliyari R, Li WX, Li HW, Kim K, Carthew R, Atkinson P, Ding SW (2006) RNA interference directs innate immunity against viruses in adult Drosophila. Science 312: 452–454 [PMC free article] [PubMed]
  • Watanabe T, Takeda A, Tsukiyama T, Mise K, Okuno T, Sasaki H, Minami N, Imai H (2006) Identification and characterization of two novel classes of small RNAs in the mouse germline: retrotransposon-derived siRNAs in oocytes and germline small RNAs in testes. Genes Dev 20: 1732–1743 [PMC free article] [PubMed]
  • Watanabe T, Totoki Y, Toyoda A, Kaneda M, Kuramochi-Miyagawa S, Obata Y, Chiba H, Kohara Y, Kono T, Nakano T, Surani MA, Sakaki Y, Sasaki H (2008) Endogenous siRNAs from naturally formed dsRNAs regulate transcripts in mouse oocytes. Nature 453: 539–543 [PubMed]
  • Wei Q, Lipardi C, Paterson BM (2003) Analysis of the 3(′)-hydroxyl group in Drosophila siRNA function. Methods 30: 337–347 [PubMed]

Articles from The EMBO Journal are provided here courtesy of The European Molecular Biology Organization
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...