RNA interference (RNAi) refers to a phenomenon in which introduction of double stranded RNA (dsRNA) molecules into a variety of cells and organisms (e.g. worms, fruit flies, fungi, and plants) leads to degradation of homologous mRNAs. The RNAi process can be divided into two steps. In the initiation step, long dsRNAs are processed into ~21-23 nt siRNAs by a nuclease called Dicer. In the effector step, the siRNAs assemble into a nuclease complex called RISC. The RISC complex then binds to homologous regions of cognate mRNAs, tagging them for destruction (1).

In view of the above mechanism, RNAi has generally been considered a post-transcriptional mode of gene silencing (PTGS). Several observations, however, have linked the RNAi pathway to transcriptional gene silencing (TGS). Several groups, for example, have shown that dsRNAs to plant-promoter regions can cause gene silencing--an effect that is mediated via DNA methylation (2, 3). More recently, it has been determined that a number of Polycomb proteins, which can bind to chromatin, are necessary for RNAi activity in worms (4).

The New Findings

In the 13 September 2002 issue of Science, Martienssen and colleagues add a new chapter to this story (5). The authors were working with the fission yeast Schizosaccharomyces pombe, which possess a number of RNAi genes. In fission yeast, transgenes that integrate into centromeric repeat regions are invariably silenced --a phenomenon that is referred to as "centromeric silencing". This centromeric silencing is mediated through methylation of the histone H3 protein and formation of a condensed heterochromatin.

In their paper, Martienssen and colleagues showed that mutations in three RNAi genes, Dicer, Argonaute, and RNA-dependent RNA polymerase, abolish centromeric silencing. Moreover, they showed that this de-repression is accompanied by loss of histone methylation at centromeric regions. These results suggest that in fission yeast, the RNAi machinery mediates centromeric silencing through histone methylation and formation of condensed heterochromatin.

The Model

The above findings, while interesting, raise a key question: how do the RNAi proteins mediate histone methylation? An important clue was revealed when the authors detected transcripts from both strands of certain centromeric repeat regions in the RNAi mutant yeasts--in contrast, transcripts from only one strand were detected in wildtype yeast. In a related paper, Bartel and colleagues cloned small RNA molecules (termed heterochromatin siRNAs) from fission yeast corresponding to the same or adjacent centromeric regions (6).

The above results suggest a model, which could help explain the mechanism of centromeric silencing. At the outset, one strand of certain centromeric regions (e.g. forward strand) is constitutively expressed, while the other (e.g. reverse strand) is transcribed intermittently. Transcription of both strands leads to formation of complementary RNA molecules, which could hybridize into dsRNAs. The dsRNAs would then get processed into siRNAs by the RNAi machinery. The siRNAs may then bind to methyl-transferases, guiding them to the homologous centromeric regions (this binding, however, has yet to be demonstrated). This would, in turn, result in histone methylation at the centromeric regions, formation of a condensed heterochromatin, and the spread of condensed heterochromatin to neighboring DNA regions.

But what is the biological role of centromeric silencing? The answer may lie in the connection between centromeric repeats and transposons. There is evidence to suggest that centromeric repeats in fission yeast have arisen from remnants of transposons. In addition, transposon immobilization is believed to be a key function of the RNAi pathway (1). Therefore, the involvement of RNAi in centromeric silencing may have originated from its role in transposon immobilization.

The Connection to Genomic Imprinting and DNA Rearrangements

It is interesting to note that two other biological events--genomic imprinting in mammals and genomic rearrangements in protozoa--are also mediated via non-coding RNA molecules and histone modifications. In genomic imprinting, expression of a mammalian gene occurs exclusively from one parental allele. As in centromere silencing, there is evidence to suggest that both non-coding RNA molecules and chromatin modifications play a role in this process (Also see Non-coding RNAs May Cause Silencing of Imprinted Genes). Currently the connection between the roles of non-coding RNAs and chromatin modifications in genomic imprinting is not clear. However in light of the above model, it is possible that the non-coding RNAs first cause chromatin modifications (e.g. histone methylation), which then leads to TGS of one parental allele.

More recently, it has been shown that ~28 nt siRNAs play a role in genomic rearrangements in the protozoa, Tetrahymena (8) (Also see Small RNAs May Cause Genomic Rearrangements in Tetrahymena). In addition, it has been shown that these DNA rearrangements require the binding of two chromo-domain proteins to methylated histones. Together, these results suggest that the ~28 nt siRNAs may also mediate genomic rearrangements via histone methylation (9).

The commonality between the mechanisms of these seemingly unrelated biological phenomena--centromeric silencing, genomic imprinting, and DNA rearrangements--point toward an interesting and unique connection between small RNA molecules and histone modifications. It is now becoming clear that the roles of small RNA molecules and chromatin condensation in various biological processes are more diverse and intertwined than anyone had previously imagined.

References

1. Zamore PD. (2002). Ancient pathways programmed by small RNAs. Science 296:1265-1269.
2. Mette MF, Aufsatz W, van der Winder J, Matzke MA, and Matzke AJM. (2000). Transcriptional silencing and promoter methylation triggered by double-stranded RNA. EMBO J. 19(19):5194-5201.
3. Jones L, Hamilton AJ, Voinnet O, Thomas CL, Maule AJ, and Baulcombe DC. (1999). RNA-DNA interactions and DNA methylation in post-transcriptional gene silencing. Plant Cell 11:2291-2301.
4. Dudley NR, Labbé J-C, and Goldstein B. (2002). Using RNA interference to identify genes required for RNA interference. Proc. Natl. Acad. Sci. USA 99(7): 4191-4196.
5. Volpe TA, Kidner C, Hall IM, Teng G, Grewal SI, and Martienssen RA. (2002). Regulation of heterochromatin silencing and histone H3 lysine-9 methylation by RNAi. Science 297:1833-1837.
6. Reinhart BJ and Bartel DP. (2002). Small RNAs correspond to centromeric heterochromatin repeats. Science 297:1831.
7. Allshire R. (2002). RNAi and heterochromatin-a hushed up affair. Science 297:18181819.
8. Mochizuki K, Fine NA, Fujisawa T, and Gorovsky MA. (2002). Analysis of a piwi-related gene implicates small RNAs in genome rearrangement. Cell 110:689-699.
9. Taverna SD, Coyne RS, and Allis CD. (2002). Methylation of histone H3 at lysine 9 targets programmed DNA elimination in Tetrahymena. Cell 110:701-711.

Related Articles

Overview of RNA Interference