MicroRNA-Dependent Trans-Acting siRNA Production

See allHide authors and affiliations

Science's STKE  06 Sep 2005:
Vol. 2005, Issue 300, pp. pe43
DOI: 10.1126/stke.3002005pe43


Less than a year has elapsed between the discovery of trans-acting small interfering RNAs (tasiRNAs) in plants and the elucidation of the major steps of the corresponding pathway. During tasiRNA biogenesis, polyadenylated RNAs transcribed from non–protein-coding TAS genes are cleaved by a microRNA (miRNA)–programmed RNA-induced silencing complex. In contrast to classical miRNA targets, RDR6 and SGS3 convert one of the TAS RNA cleavage products into double-stranded RNA, which is subsequently processed, in a phase determined by the initial miRNA cleavage site, by DICER-LIKE 4 to generate a 21-nucleotide tasiRNA population. tasiRNAs guide endogenous mRNA cleavage through the action of AGO1 or, perhaps in some cases, AGO7. Some of the tasiRNA targets probably regulate the juvenile-to-adult phase transition, but the roles of other tasiRNA targets remain to be determined.

Small RNAs [21 to 24 nucleotides (nt) long] are important regulators of gene expression that act at the transcriptional or posttranscriptional level (13). lin-4 was the first endogenous small RNA identified through a genetic screen during early Caenorhabditis elegans development (4). A few years later, 21- to 24-nt RNAs corresponding to viruses and transgenes were identified in virus-infected plants and in plants carrying transcriptionally or posttranscriptionally silenced transgenes, suggesting that, like worms, plants may tame gene expression by producing homologous small RNAs (5, 6). Large-scale cloning of small RNAs from animals (710) and plants (11, 12) revealed microRNAs (miRNAs) and small interfering RNAs (siRNAs) as important endogenous riboregulators. miRNAs and siRNAs differ in their biogenesis. miRNAs derive from single-stranded RNAs (ssRNAs) that adopt a fold-back stem-loop structure (1). These partially folded molecules are processed into miRNA duplexes by the ribonuclease III (RNaseIII) proteins Drosha and Dicer in animals and DICER-LIKE (DCL) in plants. The mature miRNA associates with the RNA-induced silencing complex (RISC), which contains a member of the Argonaute family and catalyzes the cleavage or represses the translation of partially complementary mRNAs. In contrast, siRNAs are processed from long double-stranded RNAs (dsRNAs) that result from the conversion of ssRNAs by cellular or viral RNA-dependent RNA polymerases. Many siRNAs, deriving from both RNA strands, are processed from a long dsRNA by Dicer enzymes, whereas a single miRNA commonly is processed from a miRNA precursor. Until recently, the miRNA and siRNA pathways were considered separate pathways with distinct functions: miRNAs regulating endogenous mRNAs in trans and controlling development, and siRNAs regulating invading nucleic acids (viruses, transposons, and transgenes) in cis and preventing their proliferation and deleterious effects. A series of recent reports described tasiRNAs, a class of small RNAs that requires components of both the miRNA and siRNA pathways for their biogenesis, establishing a link between the miRNA and siRNA pathways (1315) (Fig. 1).

Fig. 1.

Model of miRNA and tasiRNA biogenesis and action. miRNAs, processed from partially double-stranded precursor RNAs in a DCL1-, HYL1-, and HEN1-dependent manner, direct the AGO1-dependent cleavage of partially complementary targets. These targets include mRNAs encoding proteins involved in the control of development, as well as DCL1 and AGO1 mRNAs (guided by miR162 and miR168, respectively) and the three TAS precursor RNAs (guided by miR173 or miR390). The TAS cleavage products generated by miRNA-directed cleavage are copied to dsRNA in an RDR6- and SGS3-dependent manner and are subsequently diced to tasiRNAs by DCL4. tasiRNAs direct the cleavage of partially complementary target mRNAs in an AGO1-dependent manner (TAS1- and TAS2-derived tasiRNAs), but AGO7 also may play a role in tasiRNA-directed target cleavage, because mRNA targets of TAS3-derived tasiRNAs (ARF3 and ARF4) accumulate in ago7 (also known as zip) mutants. tasiRNAs deriving from the minus RNA strand also have the potential to cleave TAS precursor RNAs and regulate their own production, similar to the regulation of the miRNA pathway through miR162- and miR168-mediated cleavage of DCL1 and AGO1 mRNAs (indicated in dashed lanes).

In plants, the biogenesis of miRNAs requires the action of the RNaseIII DCL1, the dsRNA binding protein HYL1, and the dsRNA methylase HEN1 (16). miRNAs direct the cleavage of partially complementary target mRNAs through the Slicer activity of AGO1 (17, 18). Transgene-induced posttranscriptional gene silencing (PTGS) is mediated by siRNAs and requires AGO1 and HEN1 (19, 20) but not HYL1 and DCL1 (2123). In addition, transgene-induced PTGS requires the RNA-dependent RNA polymerase RDR6 (previously known as SDE1 or SGS2), the coiled-coil protein SGS3 (24, 25), the RNA helicase SDE3 (26), and the exonuclease WEX (27), which are not required for the miRNA pathway. ago1 and hen1 mutants exhibit severe developmental defects, because AGO1 and HEN1 act in the miRNA pathway (28, 29). In contrast, rdr6 and sgs3 exhibit subtle leaf developmental defects, indicating that they have a minor role in the control of development.

In an attempt to identify the molecular basis of the rdr6 and sgs3 phenotypes, Vazquez et al. (15) performed a differential screen between wild-type plants and rdr6 mutants and identified a non–protein-coding RNA (At2 g27400, now referred to as TAS1a) that accumulated in rdr6 mutants. siRNAs corresponding to both strands of the TAS1a locus were found in wild-type plants, but these siRNAs did not accumulate in ago1, dcl1, hen1, hyl1, rdr6, and sgs3 mutants. Remarkably, TAS1a-derived siRNAs occurred in 21-nt increments as if they were processed sequentially from a long dsRNA produced from the TAS1a RNA through the action of RDR6. These siRNAs obeyed the asymmetry rules; that is, only the strand with lowest base-pairing stability at its 5′ end accumulated (30, 31), suggesting that they entered into a RISC, a result confirmed by the coimmunoprecipitation of one of these siRNAs with AGO1 (17, 18). Vazquez et al. (15) showed that siRNAs from the TAS1a locus guide the cleavage of several endogenous mRNAs, indicating that they act in trans to regulate gene expression in a manner similar to miRNAs. Trans-acting siRNAs from the minus strand also have the potential to regulate the accumulation of the TAS1a precursor RNA through a feedback loop, analogous to the feedback regulation of the miRNA pathway through miR162- and miR168-guided cleavage of DCL1 and AGO1, respectively (29, 32).

Additional rdr6 and sgs3 alleles and mutants impaired in another member of the Argonaute family, AGO7 [also known as ZIPPY (33)] were identified in a screen for mutants impaired in the juvenile-to-adult phase transition (14). A transcriptome comparison of wild-type plants and ago7, rdr6, and sgs3 mutants identified several genes, including those encoding auxin-response factors ARF3 and ARF4, that were up-regulated in all three mutants. Other mRNAs were up-regulated in rdr6 and sgs3 but not in ago7, including a pentatricopeptide repeat (PPR) mRNA. This study independently identified one of the targets of TAS1a found by Vazquez et al. (15), and also identified the tasiRNAs that are produced from the TAS1a locus or the related TAS1b and TAS1c loci.

When searching for targets of miRNAs, Allen et al. (13) discovered that miR173 targeted the TAS1a, TAS1b, and TAS1c precursor RNAs, as well as the RNA produced by an unrelated locus (TAS2), from which another cluster of 21-nt tasiRNAs derives. They also identified TAS3, a third tasiRNA-producing locus, as a target of miR390. TAS2-derived tasiRNAs targeted the PPR mRNA and TAS3-derived tasiRNAs targeted the ARF3 and ARF4 mRNAs, which were identified by Peragine et al. (14) in their screen for RDR6- and SGS3-regulated genes. An independent computational search comparing transcribed genomic regions to intergenic regions extended the numbers of tasiRNAs targets by identifying ARF2, as well as ARF3 and ARF4, as targets of TAS3-derived tasiRNAs (34). The work of Allen et al. (13) elucidated why the tasiRNA pathway requires components of the miRNA pathway and how the frame is set for tasiRNAs production from the TAS precursor RNAs. Indeed, TAS1a, TAS1b, TAS1c, TAS2, and TAS3 tasiRNA clusters begin at the exact position where the TAS precursor RNAs are cleaved by the corresponding miRNA. In addition, mutant TAS1 and TAS2 RNAs lacking complementarity to miR173 were unable to produce tasiRNAs. Therefore, Allen et al. (13) proposed a model in which after AGO1-DCL1-HEN1-HYL1–dependent miRNA-guided cleavage of the TAS precursor RNA, the 5′ (TAS3) or 3′ cleavage products (TAS1a, TAS1b, TAS1c, and TAS2) were converted in an RDR6-SGS3–dependent manner into dsRNA that then was cleaved sequentially by an unidentified RNaseIII (Fig. 1).

Gasciolli et al. (35) confirmed the requirement for miR173 to produce TAS1- and TAS2-derived tasiRNAs by showing that a mir173 mutant accumulated reduced levels of both miR173 and TAS1- and TAS2-derived tasiRNAs. They also identified DCL4 as the RNaseIII processor of tasiRNAs by showing that a dcl4 null mutant accumulated reduced levels of TAS1-, TAS2-, and TAS3-derived tasiRNAs, whereas miR173 and miR390 levels remained unchanged. In the absence of DCL4, DCL3 was able to produce some functional tasiRNAs, and thus dcl3,dcl4 double mutants exhibited molecular and phenotypic characteristics (reduced tasiRNA accumulation, increased tasiRNA target accumulation, and downward curling of leaf margins) that were most similar to rdr6 and sgs3 mutants.

The discovery of tasiRNAs has been rapidly followed by the elucidation of their biogenesis pathway, but many questions remain that constitute the next challenges.

1) Why are TAS precursor RNAs transformed into dsRNA by RDR6 after miR173- or miR390-guided cleavage, whereas the mRNAs targeted for cleavage by other miRNAs are not? Is this related to the nature of the miRNAs or of their targets, or to a particular compartmentalization of this process?

2) Why do tasiRNAs derive from only one of the two cleavage products (the 3′ cleavage product in the case of TAS1 and TAS2 and the 5′ cleavage product in the case of TAS3). Again, is this related to the nature of the miRNA or of its target TAS RNA?

3) Why does tasiRNA-guided target mRNA cleavage require either AGO1 (TAS1 and presumably TAS2) or AGO7 (TAS3)? Is this connected to the origin of the tasiRNAs from the 3′ (TAS1 and TAS2) or 5′ cleavage product (TAS3)?

4) What is the purpose of generating a cluster of tasiRNAs to regulate target mRNAs in a manner similar to miRNAs? Is it to coregulate multiple nonhomologous targets at the same time in cases where a single miRNA could not achieve such a broad regulation?

5) What is the role of tasiRNA targets that are not involved in the juvenile-to-adult phase transition? What are the targets of TAS1 and TAS2 tasiRNAs that do not depend on AGO7 for their regulation?

In a field of research moving as fast as the RNA silencing field, one can anticipate that answers to such questions will come in the very near future.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
View Abstract

Navigate This Article