Portal de Periódicos da CAPES

RNA Interference: Big Applause for Silencing in Stockholm

2006; Cell Press; Volume: 127; Issue: 6 Linguagem: Inglês

10.1016/j.cell.2006.12.001

1097-4172

Phillip D. Zamore,

RNA regulation and disease

Eight years ago, Craig Mello, Andrew Fire, and their coworkers provided the first demonstration that double-stranded RNA (dsRNA) triggers the gene-silencing technique that we now call RNA interference (RNAi). For this landmark discovery, Mello and Fire are honored with this year's Nobel Prize in Physiology or Medicine. Eight years ago, Craig Mello, Andrew Fire, and their coworkers provided the first demonstration that double-stranded RNA (dsRNA) triggers the gene-silencing technique that we now call RNA interference (RNAi). For this landmark discovery, Mello and Fire are honored with this year's Nobel Prize in Physiology or Medicine. If fields of study have birthdays, then RNA interference (RNAi) turned eight on February 19th this year. Even by the standards of 21st century science, RNAi has been precocious. A PubMed search for “RNA interference” retrieves more than 7900 articles, all published after the landmark 1998 Nature paper by Craig Mello at the University of Massachusetts Medical School and Andy Fire, then at the Carnegie Institution and now at Stanford University School of Medicine, that launched the whole field (Fire et al., 1998Fire A. Xu S. Montgomery M.K. Kostas S.A. Driver S.E. Mello C.C. Nature. 1998; 391: 806-811Crossref PubMed Scopus (11126) Google Scholar; see Figure 1). In this paper, Mello, Fire, and their colleagues working in the worm Caenorhabditis elegans provided the first demonstration that RNA interference (RNAi), as the new gene-silencing technique had just been named, is triggered by double-stranded RNA (dsRNA). For this discovery they are honored with this year's Nobel Prize in Physiology or Medicine. The Fire and Mello discovery proved that RNAi is mechanistically distinct from antisense RNA-based strategies for inhibiting gene expression. More fundamentally, their 1998 paper, often referred to simply as the Fire and Mello paper, suggested that dsRNA—presumed by most biologists either to be inert or, to those studying mammals, to signal the presence of a viral infection—could repress the expression of a single gene. Such a role for RNA harked back to ideas from the 1960s that small pieces of RNA might bind to genes, turning them off, ideas abandoned after the discovery of transcription factors. Rereading the 1998 paper, one is struck by how many of the salient features of RNAi were identified by Fire and Mello: the ability of a few molecules of dsRNA to direct destruction of a much larger amount of the corresponding mRNA, suggesting a catalytic mechanism; the transmission of RNAi across generations; the power of RNAi to bring genetics to any organism whose genome sequence is known; the near universality of RNAi among eukaryotes. Fire and Mello transformed into testable science a “phenomenon”—a word scientists often use to imply that the observation might yet prove untrue. Many of us read their paper or heard Craig Mello speak about it at meetings and immediately changed direction. I was a bit late in discovering the paper, reading it only in March 1999; by May, RNAi was my consuming passion. Others in the field have similar stories; quite a few can tell you where they were the first time they heard of RNAi. Great papers give rise to whole fields when they not only report a discovery but also pose a challenge to the scientific community. Fire and Mello dared us to imagine that the technique of RNAi—which they taught was the introduction of dsRNA into the nematode C. elegans for the purpose of interfering with the expression of a single gene—was a clue to a wholly new regulatory mechanism. “The mechanisms underlying RNA interference,” they wrote in their landmark paper, “probably exist for a biological purpose” (Fire et al., 1998Fire A. Xu S. Montgomery M.K. Kostas S.A. Driver S.E. Mello C.C. Nature. 1998; 391: 806-811Crossref PubMed Scopus (11126) Google Scholar). The challenge to find that biology and explain both how it worked and why it existed was irresistible. RNAi was not just useful technology, it was biology—unexplored, understandable-if-only-we-think-hard-enough biology. RNAi occurs when cells respond to a dsRNA trigger by destroying a corresponding target, an mRNA sharing some or all of the sequence of the dsRNA. RNAi, we now understand, is but one of a group of mechanistically related RNA-silencing phenomena, all of which have in common small RNAs, often derived from longer dsRNA precursors, that guide protein complexes to target genes or mRNA, whose expression they silence. Besides RNAi, RNA-silencing pathways include the microRNA (miRNA) pathways of plants and animals, in which small RNAs derived from their own genes repress tens or hundreds of target mRNAs to which they are partially complementary. Small RNAs can also guide modification of chromatin structure, promoting the assembly of heterochromatin. In the protozoan Tetrahymena, such an RNA-silencing pathway uses small RNAs to mark specific DNA sequences for deletion from the macronuclear genome. RNA silencing also includes the rasiRNA and piRNA pathways, the first small RNA pathways for which there are no known dsRNA precursors contributing to the production of small RNAs. Hints of the existence of RNA-silencing pathways first came from experiments in plants. Rich Jorgensen and colleagues engineered petunias carrying transgenic copies of the gene encoding chalcone synthase, the limiting enzyme in the synthesis of the flower's purple pigment. Their intent was to generate more intensely purple flowers, but often their efforts yielded plants with dramatic pigmentation patterns comprising tissue in which chalcone synthase was produced, flanking tissue in which both the transgenic and endogenous chalcone synthase genes were mysteriously turned off. They dubbed this phenomenon “cosuppression” (Napoli et al., 1990Napoli C. Lemieux C. Jorgensen R.A. Plant Cell. 1990; 2: 279-289PubMed Google Scholar, van der Krol et al., 1990van der Krol A.R. Mur L.A. Beld M. Mol J.N.M. Stuitji A.R. Plant Cell. 1990; 2: 291-299PubMed Google Scholar). Jim Birchler, working in Drosophila melanogaster, similarly reported that the more transgenic copies of alcohol dehydrogenase (adh) that were engineered into flies, the less adh mRNA was actually made. In the mid-1990s, David Baulcombe's laboratory discovered that plant viruses could also trigger sequence-specific silencing. Moreover, they showed that this virus-induced gene silencing (VIGS) was mechanistically related to the silencing triggered by transgenes; the link between the two silencing pathways was that both required transcription of RNA from the trigger locus, implicating RNA as the initiator of silencing (Ratcliff et al., 1997Ratcliff F.G. Harrison B.D. Baulcombe D.C. Science. 1997; 276: 1558-1560Crossref PubMed Scopus (622) Google Scholar). More recently, viral defense has been proposed to be the primary function of RNAi in both plants and flies. Presciently, Baulcombe suggested that dsRNA—formed by two RNA molecules annealing or one folding back on itself—might be recognized by plant cells as a trigger for the sequence-specific silencing of genes (Ratcliff et al., 1997Ratcliff F.G. Harrison B.D. Baulcombe D.C. Science. 1997; 276: 1558-1560Crossref PubMed Scopus (622) Google Scholar). Nine months later, Fire and Mello proved this to be true for RNAi in animals, and Waterhouse and colleagues subsequently showed the same for plants. In animals, the application of antisense technology to inhibit gene expression in the worm C. elegans led to the discovery of RNAi. The rationale was that injection of antisense RNA corresponding to a cellular mRNA should block translation of that mRNA through base-pairing, converting the mRNA to an untranslatable and presumably inert form. Ken Kemphues' laboratory, however, found that injection of the mRNA itself (sense RNA) also “interfered” with its own expression, an observation at odds with an antisense mechanism (Guo and Kemphues, 1995Guo S. Kemphues K.J. Cell. 1995; 81: 611-620Abstract Full Text PDF PubMed Scopus (865) Google Scholar). It was in this context—a series of paradoxical and intriguing reports that an exogenously provided RNA could interfere with expression of an identical endogenous mRNA—that Fire and Mello formulated their unifying hypothesis: eukaryotic cells perceive dsRNA as a sequence-specific signal to inhibit expression of the corresponding mRNA. Their idea resolved the paradox inherent in the experiments of Guo and Kemphues and provided a testable model for the mechanism of RNAi in particular and RNA-silencing phenomena in general. With their 1998 paper, Fire and Mello inspired others to find RNAi in their favorite organism or to study it using their favorite techniques. By December of that year, Richard Carthew's laboratory had shown that dsRNA triggered a robust RNAi response in fly embryos (Kennerdell and Carthew, 1998Kennerdell J.R. Carthew R.W. Cell. 1998; 95: 1017-1026Abstract Full Text Full Text PDF PubMed Scopus (919) Google Scholar), a finding that led directly to the development of the first in vitro system for recapitulating RNAi in a cell extract (of Drosophila embryos) (Tuschl et al., 1999Tuschl T. Zamore P.D. Lehmann R. Bartel D.P. Sharp P.A. Genes Dev. 1999; 13: 3191-3197Crossref PubMed Scopus (676) Google Scholar). Within five years, aspects of RNA silencing could be studied in vitro for flies, mammals, and plants. The second great breakthrough, however, came in October 1999, from studies begun before the discovery that dsRNA triggered RNAi in animals. Baulcombe and his student, Andrew Hamilton, reported that in plants in which a specific gene was silenced, there was always accumulation of antisense RNAs ∼25 nucleotides long that were complementary to the silenced gene (Hamilton and Baulcombe, 1999Hamilton A.J. Baulcombe D.C. Science. 1999; 286: 950-952Crossref PubMed Scopus (2219) Google Scholar). We now call these molecules small interfering RNAs (siRNAs). The discovery of siRNAs ultimately allowed the application of RNAi to mammals, but their immediate impact was to elucidate the mechanisms underlying both RNAi and the related miRNA pathway. In vitro experiments in extracts from fly embryos and cultured cells soon demonstrated that siRNAs are produced by endonucleolytic processing of the dsRNA trigger by an enzyme called Dicer (Bernstein et al., 2001Bernstein E. Caudy A.A. Hammond S.M. Hannon G.J. Nature. 2001; 409: 363-366Crossref PubMed Scopus (3618) Google Scholar). MicroRNAs, too, were found to be produced by Dicer from double-stranded stem-loop structures within a single transcript, rather than from long dsRNA molecules (Bernstein et al., 2001Bernstein E. Caudy A.A. Hammond S.M. Hannon G.J. Nature. 2001; 409: 363-366Crossref PubMed Scopus (3618) Google Scholar, Grishok et al., 2001Grishok A. Pasquinelli A.E. Conte D. Li N. Parrish S. Ha I. Baillie D.L. Fire A. Ruvkun G. Mello C.C. Cell. 2001; 106: 23-34Abstract Full Text Full Text PDF PubMed Scopus (1455) Google Scholar, Hutvágner et al., 2001Hutvágner G. McLachlan J. Pasquinelli A.E. Balint É. Tuschl T. Zamore P.D. Science. 2001; 293: 834-838Crossref PubMed Scopus (2094) Google Scholar). The first miRNA, lin-4, had been discovered by Victor Ambros and coworkers in 1993 (Lee et al., 1993Lee R.C. Feinbaum R.L. Ambros V. Cell. 1993; 75: 843-854Abstract Full Text PDF PubMed Scopus (8936) Google Scholar); the second, let-7, by Gary Ruvkun and coworkers in 2000 (Reinhart et al., 2000Reinhart B.J. Slack F.J. Basson M. Pasquinelli A.E. Bettinger J.C. Rougvie A.E. Horvitz H.R. Ruvkun G. Nature. 2000; 403: 901-906Crossref PubMed Scopus (3565) Google Scholar). (Both lin-4 and let-7 were identified first in C. elegans.) Today, thousands of miRNAs have been identified in plants and animals, including at least 474 in humans. The discovery that both siRNAs and miRNAs were produced by the same enzyme reinforced the view that RNAi existed for a biological purpose. Small interfering RNAs and miRNAs are ∼22 nucleotide-long RNAs that act as guides—what Baulcombe called the “specificity determinant”—for Argonautes, a diverse family of proteins specialized for silencing gene expression. Argonaute proteins were first implicated in RNAi when Mello, Fire, and coworkers identified the RNAi-deficient 1 (rde-1) gene in a large-scale genetic screen for proteins required for RNAi in C. elegans (Tabara et al., 1999Tabara H. Sarkissian M. Kelly W.G. Fleenor J. Grishok A. Timmons L. Fire A. Mello C.C. Cell. 1999; 99: 123-132Abstract Full Text Full Text PDF PubMed Scopus (975) Google Scholar). Worms lacking rde-1 function cannot mount an RNAi response when exposed to dsRNA. The sequence of rde-1 revealed homologs in nearly every eukaryote, from fungi such as Neurospora crassa and Schizosaccharomyces pombe to plants, flies, and mammals. Although rde-1 mutant worms do not support RNAi, they appear to be otherwise normal. In contrast, plants with mutations in the RDE-1 homolog, ARGONAUTE-1 (AGO-1), identified in Arabidopsis thaliana in a genetic screen, develop spike-like leaves instead of flat blades with distinct top and bottom surfaces. We now know that plant AGO-1 binds to miRNAs, whose function is required to specify the distinct developmental fates of the cells that form the top and bottom of a leaf. The Drosophila RDE-1 homolog, Piwi, is required for the maintenance of germline stem cells. The Piwi protein binds to repeat-associated siRNAs (rasiRNAs)—small RNAs corresponding to transposons and other selfish genetic elements—that block expression of RNA from these molecular parasites. The discovery that RDE-1 defines a large clade of related proteins immediately suggested that RNAi-like mechanisms play vital roles in regulating gene expression in both plants and animals. The last five years have seen an avalanche of data reinforcing this view. In C. elegans, RNAi initiated by exogenously supplied dsRNA can spread from cell to cell. In fact, RNAi can be inherited in worms, the silent state of a gene transmitted from mother to child. Silencing in plants can also spread cell-to-cell, with silent root stock triggering silencing of the same gene in the uppermost leaves of a previously nonsilenced graft. Such spreading reflects the transport of dsRNA into cells, its amplification by RNA-dependent RNA polymerases, and the export of some form of dsRNA, perhaps siRNAs, from the first cell to a second. Sadly, this remarkable and bizarre aspect of RNAi, recognized in the 1998 Fire and Mello paper, has not yet been observed in flies or mammals. Of course, one cannot overstate the importance of RNAi as a tool. Even the most hardcore biochemists now practice genetics of a sort using RNAi. RNAi-based whole-genome screens have become routine for C. elegans and cultured Drosophila cells and are becoming routine for cultured mammalian cells. Thanks to RNAi, the effect on a molecular pathway, cellular morphology, or organismal phenotype of the loss of function of every known gene can be determined in a matter of weeks. Beyond the original insight that dsRNA is a potent and specific trigger of RNAi, technological breakthroughs make possible such high-throughput RNAi screens. In C. elegans, Fire and Mello and their coworkers showed that soaking worms in dsRNA, or better yet feeding them bacteria expressing dsRNA, triggers RNAi. For flies, Jack Dixon showed that cultured S2 cells spontaneously internalize dsRNA from the culture medium, leading to a sequence-specific RNAi response. More spectacular still was the discovery by Thomas Tuschl and colleagues that siRNAs could be chemically synthesized and administered to cultured mammalian cells, eliciting a sequence-specific RNAi response lasting days and bypassing the non-sequence-specific responses mammalian cells mount when exposed to longer dsRNA (Elbashir et al., 2001Elbashir S.M. Harborth J. Lendeckel W. Yalcin A. Weber K. Tuschl T. Nature. 2001; 411: 494-498Crossref PubMed Scopus (7896) Google Scholar). The adaptation of RNAi technology to mammals has also inspired the hope that RNAi triggered by siRNAs might form the basis for new drugs capable of silencing viral or human genes that cause disease. Both young biotechnology companies and old pharmaceutical firms are actively pursuing this idea. In mice and monkeys, intravenous injection of chemically modified siRNAs can elicit long-lasting, sequence-specific silencing of the corresponding mRNA in several different tissues, and the initial results from human trials have been encouraging that siRNAs are well tolerated in people. What began as the desire of Fire and Mello to explain an antisense experiment that worked too well produced a revolutionary insight that has forever changed modern biology. Like the discovery of the structure of DNA, the realization that cells respond to dsRNA by silencing the corresponding gene has changed our view of gene regulation and the organization of DNA into chromosomes. Like the invention of PCR, the discovery of RNAi provided a transformative new technology that brings the power of genetics to the bewildering riches of genome sequences. And like the invention of the CAT scan and MRI before it, RNAi promises to transform modern medicine. Craig and Andy: thank you!