Silencing viruses with RNA

Our cells have a built-in mechanism for 'silencing' genes, called RNA interference. This capability has now been exploited to protect cells in culture dishes from HIV-1 and poliovirus.

Until recently, the two major ways to combat viral infections were vaccines and drugs that are targeted to specific viral enzymes or other viral proteins. Papers on pages 430 and 435 of this issue1,2, and another published earlier this month in Nature Medicine3, provide a significant new strategy — one that targets viral genomes and the messenger RNA molecules they encode.

Each of these papers makes use of a revolutionary new technology, RNA interference (RNAi), which prevents the expression of genes by using small RNA molecules, such as 'small interfering RNAs' (siRNAs). This technology in turn takes advantage of the fact that RNAi is a natural biological mechanism for silencing genes in most, if not all, cells of many living organisms, from plants to insects to mammals4. The idea is that RNAi prevents a gene from producing a functional protein by ensuring that the molecular intermediate — the messenger RNA (mRNA) copy of the gene — is destroyed.

How does it work? The key cellular components in RNAi are molecular complexes (the silencing complexes shown in Fig. 1) that contain double-stranded siRNAs of 21–23 base pairs in length, associated with a number of proteins, some of which remain to be identified. These complexes recognize their (single-stranded) mRNA targets by matching RNA sequences, and direct the cleavage and destruction of any mRNAs that perfectly match one of the two siRNA strands5,6,7. That prevents the mRNAs from being translated into protein. If even a single nucleotide differs between an siRNA and its target, the effect is greatly diminished, or even eliminated entirely. The required length of 21–23 nucleotides is short enough to ensure a low probability that RNAs other than the desired target will be destroyed; siRNAs that find no suitable target appear to remain inert within the cell. Following target recognition and destruction, the silencing complexes may act again on other target molecules.

Figure 1: RNA interference — a tool for cells as well as scientists.

The diagram shows how RNA interference is thought to occur in mammalian cells. Short double-stranded RNA molecules (dsRNAs) can be introduced into cells from the outside, or are produced within the cell nucleus from longer precursors. These precursors are typically RNAs that can fold back on themselves by base-pairing to form 'hairpin' structures, which are cleaved to generate shorter RNAs, containing 21–23 base pairs, that are then exported to the cytoplasm. In the cytoplasm, the short dsRNAs (called 'small interfering RNAs', or siRNAs) assemble with proteins into silencing complexes. Targets that perfectly match one strand of the dsRNA are cleaved and destroyed. The targets are messenger RNA copies of certain genes, such as those of viruses that have entered the cell. So RNA interference is useful to the cell, and new papers1,2,3,11,12 suggest that it could also be exploited by scientists to fight viral infections.

Silencing complexes can be assembled from siRNAs experimentally supplied from outside the cell, but can also be produced within the cell, using siRNAs processed from precursor RNAs expressed in the cell nucleus. Accordingly, several recent reports8,9,10 have shown that if short 'hairpin' RNAs (which fold back on themselves, generating double-stranded regions) are expressed in the nucleus, they can be trimmed by cellular enzymes to the correct size for RNAi, then assembled into silencing complexes that act outside the nucleus (Fig. 1).

RNAi is useful to cells because it enables them to control gene expression; for example, it might be an ancient defence mechanism that prevents viruses that have infected the cell from producing functional proteins. From a scientist's standpoint, because RNAi relies on nucleotide interactions, it can be tailored to almost any gene — whether cellular or viral — and so can be used to study the function of genes by knocking out their expression without actually destroying or mutating the genetic material1,2,3,11,12,13,14. And it has appeal for treating viral infections: it offers an attractive combination of exquisite specificity (with no apparent adverse side effects) and the exploitation of an underlying cell pathway that might have evolved naturally for a similar purpose.

But does it actually work as an antiviral weapon? The three new papers1,2,3 suggest that it does, at least in cells in culture dishes. Gitlin and colleagues1 report that specific siRNAs administered to human cells from the outside, like a drug, can enter them and protect them against infection by the rapidly multiplying poliovirus. Jacque and co-workers2, meanwhile, report similar results in their studies with the AIDS virus HIV-1. These authors further demonstrate that if the siRNAs are expressed from inside cells, rather than simply administered to the outside, the cells become largely immune to subsequent HIV-1 infection.

Novina et al.3 also show that siRNAs directed against HIV-1 have the potential to be useful treatments. In addition, these authors used siRNAs targeted against the cellular receptor for HIV-1 — CD4 — to reduce the ability of the virus to enter cells. These1,2,3 and other recent studies11,12 show that siRNAs can inhibit viral replication at several stages of infection, including the very early stages, when viruses are most vulnerable. Infection can also be blocked by targeting either viral genes or host genes that are involved in the viral life cycle3.

These results are exciting, and suggest that RNAi could be perfectly suited to many antiviral applications. But success is not automatically assured. One important factor is that not all viral RNA sequences are equally accessible to siRNAs: some sequences might be buried within secondary structures or within highly folded regions in target RNAs, whereas others might form tight complexes with proteins that obscure their recognition4. Optimal targets must be chosen not only rationally, but also by trial and error.

Another, and quite serious, issue is that viruses are notoriously sloppy in their replication, and produce mutated progeny molecules. Some of these naturally help the viruses escape immune surveillance or inhibition by drugs, but they might also prevent recognition by siRNAs. To overcome this obstacle, one might need to target viral RNA sequences that are conserved and normally invariant between different strains, or to simultaneously target several viral sequences. Finally, the question of how to deliver siRNAs to cells needs to be addressed. They can certainly be delivered efficiently to cells in culture, but methods must be improved before RNAi can be used in animals, let alone patients.

Although antiviral RNAi technology is not yet optimized, the phenomenon does appear to be both general and effective. In 1988, David Baltimore15 proposed the concept of "intracellular immunization", whereby one would express within cells inhibitory molecules (usually proteins) that can protect those cells from specific viral infections in the future. The promise of intracellular immunization now appears to be closer to reality — although, amazingly, through the use of small RNAs rather than peptides or proteins.


  1. 1

    Gitlin, L., Karelsky, S. & Andino, R. Nature 418, 430–434 (2002); advance online publication, 26 June 2002 (doi:10.1038/nature00873).

    ADS  CAS  Article  Google Scholar 

  2. 2

    Jacque, J.-M, Triques, K. & Stevenson, M. Nature 418, 435–438 (2002); advance online publication, 26 June 2002 (doi:10.1038/nature00896).

    ADS  CAS  Article  Google Scholar 

  3. 3

    Novina, C. D. et al. Nature Med. 8, 681–686 (2002).

    CAS  Article  Google Scholar 

  4. 4

    Sharp, P. A. Genes Dev. 15, 485–490 (2001).

    CAS  Article  Google Scholar 

  5. 5

    Parrish, S., Fleenor, J., Xu, S., Mello, C. & Fire, A. Mol. Cell 6, 1077–1087 (2000).

    CAS  Article  Google Scholar 

  6. 6

    Nykanen, A., Haley, B. & Zamore, P. D. Cell 107, 309–321 (2001).

    CAS  Article  Google Scholar 

  7. 7

    Zamore, P. D., Tuschl, T., Sharp, P. A. & Bartel, D. P. Cell 101, 25–33 (2000).

    CAS  Article  Google Scholar 

  8. 8

    Paddison, P. J., Caudy, A. A., Bernstein, E., Hannon, G. J. & Conklin, D. S. Genes Dev. 16, 948–958 (2002).

    CAS  Article  Google Scholar 

  9. 9

    Brummelkamp, T. R., Bernards, R. & Agami, R. Science 296, 550–553 (2002).

    ADS  CAS  Article  Google Scholar 

  10. 10

    Yu, J.-Y, DeRuiter, S. L. & Turner, D. L. Proc. Natl Acad. Sci. USA 99, 6047–6052 (2002).

    ADS  CAS  Article  Google Scholar 

  11. 11

    Bitko, V. & Barik, S. BMC Microbiol. 1, 34 (2001).

    CAS  Article  Google Scholar 

  12. 12

    Lee, N. S. et al. Nature Biotechnol. 20, 500–505 (2002).

    CAS  Article  Google Scholar 

  13. 13

    Elbashir, S. M. et al. Nature 411, 494–498 (2001).

    ADS  CAS  Article  Google Scholar 

  14. 14

    Harborth, J., Elbashir, S. M., Bechert, K., Tuschl, T. & Weber, K. J. Cell Sci. 114, 4557–4565 (2001).

    CAS  PubMed  PubMed Central  Google Scholar 

  15. 15

    Baltimore, D. Nature 335, 395–396 (1988).

    ADS  CAS  Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Gordon G. Carmichael.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Carmichael, G. Silencing viruses with RNA. Nature 418, 379–380 (2002).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.