In the past few years it has become evident that double stranded RNA (dsRNA), when introduced into a variety of organisms and cell lines, can mediate a form of post-transcriptional gene silencing called RNA interference (RNAi). This silencing takes place after the ribonuclease DICER cleaves the injected dsRNAs into 21-25 nt RNA species called small interfering RNAs, or siRNAs (1). These siRNAs then bind to the cognate mRNA and tag it for destruction.

Interestingly, similarly sized 25 nt single-stranded antisense RNA (asRNA) fragments have been implicated in co-suppression, a phenomenon in plants in which transgenes sometimes cause silencing of both the transgene and the homologous endogenous gene (2). It has been shown that these suppressive transgenes give rise to aberrant asRNAs. It has further been proposed that asRNAs bind to their target mRNA and act as primers for the synthesis of dsRNAs by an RNA-dependent RNA polymerase (RdRP) (3). The long dsRNA products may get cleaved into siRNAs, which may then cause the co-suppressive silencing effect.

In a paper published in the 25 January issue of Science, Tijsterman and colleagues show that asRNAs can induce gene silencing in Caenorhabditis elegans germ cells (4). Moreover they present data that are consistent with the above model. Interestingly, the authors also use genetic data to make a connection between co-suppression and asRNA-mediated silencing. Antisense RNA was injected into nematodes containing mutations in 4 genes known to be important for the initiation and/or maintenance of cosuppression and RNAi: mut-7, mut-14 ("mut" stands for mutator; these genes are implicated in the maintenance of RNAi), rde-1, and rde-4 ("rde" stands for RNAi deficient; these genes are apparently involved in the initiation of RNAi by long dsRNA). The authors found that silencing through asRNA is dependent on the mut-7 and mut-14, but not the rde-1 and rde-4 genes. These experiments suggest that co-suppression is mediated through generation of dsRNAs by transgene-encoded asRNA primers.

The Experimental Approach

Tijsterman and colleagues first asked whether siRNAs, sense RNAs (sRNAs), and asRNAs were capable of silencing genes in C. elegans germ cells. To answer this question, they injected long dsRNA (800 bp), siRNA (25 bp), asRNA (25 nt), or sRNA (25 nt) corresponding to the pos-1 gene into the worm gonadal syncytium (germ cells). The pos-1 mRNA is maternally provided and if suppressed, leads to embryonic lethality. Therefore the percentage of lethal embryos was used to measure silencing. The authors showed that while the long dsRNAs led to 100% lethality, the siRNAs unexpectedly caused death in only 5–10% of the embryos. To their surprise, the asRNAs caused silencing in a sizable 50% of the embryos.

Several pieces of evidence suggested that asRNAs were mediating silencing by acting as primers for the synthesis of long dsRNAs. First, antisense pos-1 DNA oligonucleotides or morpholinos did not cause silencing, suggesting that the asRNAs were not merely preventing the assembly or movement of the ribosome during translation. Moreover, when the 3' end of asRNAs were modified to prevent polymerization, the silencing ability of the asRNAs diminished dramatically, implying that the asRNAs were acting as primers for dsRNA synthesis. Additionally the asRNA-dependent silencing was not effective in dcr-1 mutant worms, suggesting that asRNA-mediated silencing was dependent on DICER and the RNAi machinery.

The authors then showed that asRNA-mediated silencing was dependent on MUT-7 and MUT-14, but not on the RDE-1 and RDE-4 proteins. Previous studies have indicated that RDE-1 and RDE-4 are involved in the initiation of RNAi in C. elegans by introduction of long dsRNA, whereas MUT-7 and MUT-14 are important at later steps in the RNAi pathway. Interestingly, silencing of pos-1 by asRNA was dramatically reduced in mut-7 and mut-14 mutant worms. However, this reduction was not seen in rde-1 and rde-4 mutant C. elegans. Interestingly the requirement for MUT-7 and MUT-14 but not RDE-1 and RDE-4 is similar to that of co-suppression. However in dsRNA-mediated RNAi, all 4 gene-products are needed.

The Remaining Questions

While these data provide some intriguing links between co-suppression, dsRNA- and asRNA-mediated silencing, they also raise some interesting questions. For example, how does silencing induced by asRNAs compare to that induced by siRNAs directed toward various target sites? If asRNAs mediate silencing by acting as primers for the synthesis of long dsRNAs, one would expect that a range of 'downstream' siRNAs would be generated. Therefore the efficiency of silencing by a single asRNA might be expected to be comparable to that of a full set of 'downstream' siRNAs.

If the asRNAs work via the RNAi pathway as suggested, why do they not require the RDE-1 and RDE-4 proteins? After introduction of pos-1 dsRNA into C. elegans germ cells, the authors were not able to detect pos-1 siRNAs in rde-1, rde-4, or mut-7 mutant worms by RNase protection assays. Yet they were able to detect pos-1 siRNAs after incubating long pos-1 dsRNAs with cell-free extracts from these 3 mutant worms. These experiments imply that RDE-1, RDE-4, and MUT-7 are involved in the stabilization of siRNAs but not in their generation. It is therefore possible that asRNA silencing is mediated through the formation of unstable siRNAs. However, that is contrary to the observation that asRNA-mediated silencing in C. elegans germ cells is more effective than that of siRNAs.

The authors also cloned the mut-14 gene by genetic mapping, and showed that it belongs to a family of RNA helicases. So far the function of this gene in asRNA-mediated silencing, co-suppression, or the RNAi pathway is not clear.

It is apparent that this is only the first step in elucidating a connection between these potentially related mechanisms. Undoubtedly, many more studies will soon follow.

References

1. Bernstein E, Caudy AA, Hammond SM, and Hannon GJ. (2001) Role for a bidentate ribonuclease in the initiation step of RNA interference. Nature 409:363-366.
2. Hamilton AJ, and Baulcombe DC. (1999) A species of small antisense RNA in posttranscriptional gene silencing in plants. Science 286:950-952.
3. Nishikura K. (2001) A short primer on RNAi: RNA-directed RNA polymerase acts as a key catalyst. Cell 107:415-418.
4. Tijsterman M, Kettling RF, Kristy OL, Titia S, and Plasterk RH. (2002) RNA helicase MUT-14-dependent gene silencing triggered in C. elegans by short antisense RNAs. Science 295:694-697.