RNA SILENCING AND N ONCODING RNA 37. S. Chen et al., Biosystems, in press. 43. D. W. Selinger et al., Nature Biotechnol. 18, 1262 48. A. E. Souza, T. Hermann, H. U. Go¬ringer, Nucleic Acids 38. R. J. Klein, Z. Misulovin, S. Eddy, Proc. Natl. Acad. Sci. (2000). Res. 25, 104 (1997). U.S.A., in press. 44. K. E. Plant, S. J. E. Routledge, N. J. Proudfoot, Mol. 49. R. T. Batey, R. P. Rambo, L. Lucast, B. Rha, J. A. 39. E. Rivas, R. J. Klein, T. A. Jones, S. R. Eddy, Curr. Biol. Cell. Biol. 21, 6507 (2001). Doudna, Science 287, 1232 (2000). 11, 1369 (2001). 45. X. Chen, A. M. Quinn, S. L. Wolin, Genes Dev. 14, 777 40. R. J. Carter, I. Dubchak, S. R. Holbrook, Nucleic Acids (2000). 50. I thank S. Altuvia, J. Brosius, S. Gottesman, K. M. Res. 29, 3928 (2001). 46. J. Gu, Y. Chen, R. Reddy, Nucleic Acids Res. 26, 160 Wassarman, and A. Zhang for helpful discussions and 41. J. Cavaille« et al., Proc. Natl. Acad. Sci. U.S.A. 97, (1998). comments on the manuscript. I also thank them and 14311 (2000). 47. D. A. Samarsky, M. J. Fournier, Nucleic Acids Res. 27, many other investigators for extensive discussion of 42. A. Hu¬ttenhofer et al., EMBO J. 20, 2943 (2001). 161 (1999). the nomenclature. VIEWPOINT RNA Silencing: The Genome’s Immune System Ronald H. A. Plasterk Genomes are databases sensitive to invasion by viruses. In recent years, a which allows the vertebrate immune system defense mechanism has been discovered, which turns out to be conserved to raise a massive immune response (11–14). among eukaryotes. The system can be compared to the immune system in several ways: It has specificity against foreign elements and the ability to The Function of RNA Silencing amplify and raise a massive response against an invading nucleic acid. The Neither nematodes nor flies normally en- latter property is beginning to be understood at the molecular level. counter highly concentrated double-stranded RNA (dsRNA) of identical sequence to one All genomes of complex organisms are po- cells raised against self antigens are excluded of their endogenous genes. Nevertheless, ge- tential targets of invasion by viruses and from the mature immune system. netic analysis indicates that the number of transposable elements. Forty-five percent of How does the genome recognize invaders genes required for gene silencing triggered by the human genome consists of remnants of and raise an overwhelming and specific “im- exogenous dsRNA is probably larger than 10 previous transposon/virus invasions and ele- mune response” against them? One strategy (15–18). What is the natural function of this ments that are still active to date: 21% long to suppress transposons may be the selective elaborate pathway? interspersed nuclear elements, 13% short in- methylation of transposon sequences in the The clearest picture is seen in plants, terspersed nuclear elements, 8% retroviruses, genome (1), although it has also been argued where PTGS and virus-induced gene silenc- and 3% DNA-transposons, as compared with that this phenomenon is a secondary effect of ing are recognized as mechanisms that pro- less than 2% that encodes (nontransposon) suppression (2). This will not be discussed tect against frequently occurring viral infec- proteins. A priori, one would expect that further, but see a recent review for more tions (6, 19). An advantage of this defense organisms need to fight off such invasions to information (3). In recent years, an RNA- system is that the defensive signal can spread, prevent the genome from being completely based silencing mechanism has emerged that such that inoculation in one area of a leaf can taken over by molecular invaders. The two is ancient, conserved among species from confer immunity on surrounding cells. A problems with which the organism is faced in different kingdoms (fungi, animals, and study in this issue shows that an animal virus protecting the integrity of the genome are plants), and very likely acts as the “immune also encodes a suppressor of RNA interfer- similar to those faced by the vertebrate im- system” of the genome. This system was ence (RNAi), supporting the notion that mune system: (i) how to recognize self from initially independently discovered and stud- RNAi may have an antiviral function in ani- nonself, and (ii) how to amplify an initial ied in different organisms before it was rec- mals as well (20). In nematodes, loss of response in a specific fashion. ognized that the underlying mechanisms are function of genes required for RNAi results The vertebrate immune system fights off at some level identical. Posttranscriptional in the activation of multiple transposable el- invaders using a two-step strategy: a large gene silencing (PTGS) and co-suppression in ements in the germline (15), indicating that repertoire of antibody-encoding genes is gen- plants (4, 5), as well as RNA-mediated virus they function to repress the spreading of erated from a limited set of gene segments by resistance in plants (6), RNA interference in transposons within the genome of subsequent combinatorial gene rearrangements, and this animals [first discovered in Caenorhabditis generations of worms. repertoire is stored in a distributed fashion elegans (7)], and silencing in fungi [“quell- Protection against viruses and transposons over large numbers of cells. After infection, ing” in Neurospora (8)] and algae (9) are all may be the natural function of the core of the clonal selection and expansion of a few of based on the same core mechanism. This RNAi pathway, but it does not explain all these cells results in an immune response conclusion is based on the discovery of com- aspects of what is now considered to be specifically directed to the immunogen. The mon mechanistic elements [such as the small RNAi. One of the most striking features of vertebrate immune system has solved the interfering RNAs (siRNAs) (10)] and of ho- RNAi in C. elegans is the systemic effect. specificity problem by initially generating a mology between genes required for this Injection of naked dsRNA into one region of more or less random repertoire, which, during mechanism in plants, animals, and fungi and the animal may affect gene expression else- a phase of early development, is limited by a algae. where, and dsRNA present in the lumen of filtering process, called tolerance induction: The precise mechanism of this group of the gut as part of the food is apparently taken phenomena, now referred to as RNA silenc- up and affects gene expression in progeny ing, is being rapidly unraveled. The aspect that arises in the gonads (21). In plants, graft- Hubrecht Laboratory, Centre for Biomedical Genetics, Uppsalalaan 8, 3584 CT Utrecht, Netherlands. E-mail: that I specifically address here is the equiva- ing experiments have shown immunity trav- [email protected] lent in RNA silencing of “clonal selection,” eling over 30 cm of stem tissue (22); this www.sciencemag.org SCIENCE VOL 296 17 MAY 2002 1263 RNA SILENCING AND N ONCODING RNA ability may add to the protective effect in able that all “good” genes share structural ed amplification) and thus initiate an RNA- case of repeated infections by a virus. This motifs in their mRNAs, possibly even in the directed RNA polymerization reaction (see systemic effect is not seen in all systems (e.g., interaction between the 5Ј and 3Ј termini, and Fig. 1). is not seen in Drosophila). In the case of C. proteins factors bound to them. mRNAs that elegans, it may be entirely coincidental that lack such features might be turned into Target-Dependent Amplification the RNA-silencing effect can be triggered by dsRNA by a specialized machinery. Several In the first step of this reaction, the mRNA is dsRNA in the food. C. elegans can take C. elegans mutants that are defective in trans- recognized by primary siRNAs. The hypo- precursors for nucleic acids from its food. poson silencing are not defective in RNAi thetical sequence of events is as follows: RNAi induced by feeding (21, 23) may take after administration of dsRNA, possibly re- dsRNA is cut into short siRNAs, presumably advantage of two distinct pathways, one vealing the hypothetical step that turns for- these are converted from dsRNA into ssRNA, whose natural function is to import nucleic eign mRNAs into dsRNA. and then two things can happen. These acids to be used as precursors for replication Plant mutants that are defective in trans- siRNAs (presumably bound to proteins) are and transcription, and the other that functions gene silencing are found not to be defective by themselves unstable, and are degraded, as the virus/transposon shield. in virus-induced silencing (18). They contain unless they recognize homologous target a mutation in an RNA-directed RNA poly- mRNA, present in the cell, and base pair to it. Self and Nonself merase (RdRP), and its likely role is to con- The evidence for this in C. elegans is three- Given the link of the genome’s “immune vert the single-stranded RNA (ssRNA) of the fold: (i) RNAi directed against a marker gene system” to RNAi, which is triggered by “foreign” transgene into dsRNA. Thus, for [green fluorescent protein (GFP)] does not dsRNA, one may ask how transposons or viruses the nonself feature could simply be result in detectable siRNAs in vivo unless the viruses induce dsRNA that corresponds to dsRNA, whereas for transgenes the nonself GFP gene is expressed in the target tissue their own sequence. In C. elegans, at least feature would be something that is recog- (25). (ii) Only the antisense strand of siRNAs three explanations have some plausibility.