An important RNA interference (RNAi) milestone—the silencing of a therapeutically relevant endogenous gene in an animal model by systemic administration of short interfering (si) RNAs—was reported today in Nature. Researchers at the biotech company Alnylam in Kulmbach, Germany, and in Cambridge, Massachusetts, silenced the gene for apolipoprotein B (ApoB) by conjugating siRNAs with cholesterol to help them gain entry to cells.

Delivery to the right place is the biggest issue in designing therapeutic siRNAs, particularly gaining entry to cells. The two predominant solutions are to use lipid complexes or viral vectors (see ARF related news story). In a study published earlier this year, Alnylam scientists reported in vitro success with a simpler approach: RNAi conjugated to cholesterol was able to gain entry to liver cells and silence a reporter gene (Lorenz, 2004). How this occurs is not clear, though the scientists noted in that earlier paper that "siRNA modified with lipophilic moieties may enhance siRNA uptake via a receptor-mediated mechanism or by an increased membrane permeability of the otherwise negatively charged RNA."

The other critical issue in making RNAi therapy viable is protecting the RNA from nucleases, either inside or outside cells. Nucleases in the bloodstream are a particular problem, since the ability to deliver siRNAs systemically would be a great advantage. Fortunately, in the current Alnylam study Hans-Peter Vornlocher, first author Jürgen Soutschek, and colleagues were able to take advantage of well-tested methods to stabilize oligonucleotides against nucleases—the use of phosphorothioate backbone and methylated sugars. Plus, the cholesterol conjugation appeared to provide added protection for the siRNAs against nucleases in blood.

The researchers chose to target ApoB, a major structural component of the low-density lipoprotein (LDL) cholesterol complex, the "bad" sort that contributes to coronary artery disease. ApoB is the ligand for the LDL receptor, and is expressed primarily in liver and jejunum. In their first in vivo experiments, Soutschek and colleagues administered cholesterol-conjugated ApoB-siRNA (chol-ApoB-siRNA) systemically to normal C57BL/6 mice. The researchers detected the siRNAs in liver, jejunum, and other tissues (though apparently not in brain), and one of their constructs lowered ApoB mRNA by 57 +/- 6 percent in liver and 73 +/- 10 percent in jejunum (P Perhaps most impressive was the fact that these changes were reflected in blood cholesterol and lipoprotein profiles, including a lowering of total cholesterol by 37 +/- 11 percent (P Supporting evidence came from a transgenic mouse model expressing a human ApoB variant. In this case, the chol-ApoB-siRNA significantly reduced both endogenous ApoB mRNA and the human transgenic ApoB mRNA in liver. The authors did not report on whether the siRNA was able to prevent the atherosclerosis seen in these animals when fed a high-fat diet.

In a News and Views commentary, John Rossi of the Beckman Research Institute of the City of Hope in Duarte, California, extols the simplicity of the siRNA construct. "The system did not require expensive lipid complexes or other macromolecular carriers, but merely a single cholesterol conjugate per RNA duplex," he writes. But Rossi does provide some of the usual warnings against undue optimism. Presumably, this therapy would have to be used by human patients for many years, and the long-term effects of siRNA would have to be investigated closely. Rossi also points out that the dosage used in the mice would require regular injections of gram quantity chol-ApoB-siRNAs in humans, perhaps a prohibitive expense.—Hakon Heimer


  1. Many groups have adopted siRNA or shRNA technologies as the method of choice for gene knockdown since it was first demonstrated that these techniques work for mammalian cells. Although there are previous demonstrations that one can inject siRNA duplexes and evoke systemic effects, the paper by Soutschek et al. improves things further by increasing efficiency of delivery as well as duplex stability. For those of us who work on neurodegenerative diseases, the question now is, “Will it work in brain?” Perhaps not: most of the cholesterol in the brain is synthesized locally and relatively small amounts are taken up through the blood-brain barrier. However, it seems likely that the general concept may be useful. By using molecules that are permeable to the blood-brain barrier, one could imagine peripherally administered siRNA complexes being delivered to the brain quite efficiently. Given that the biodistribution of many natural and artificial molecules is known, one could even design compounds that act as postcodes for different organs. The challenge then would be to develop sufficiently efficient siRNA molecules that could be safely used at lower doses to target gene expression in vivo.

  2. Immediately following the description that RNA interference (RNAi) works in mammalian cells, it was recognized that such specific inhibitors of gene expression held tremendous potential as drugs (1). A wealth of papers describing the inhibition of disease-causing or disease-associated genes soon followed (reviewed in 2,3). As scientists sought to extend these successes to animal models, a predictable but nevertheless difficult obstacle arose: unmodified small interfering RNAs were not easily delivered to the relevant tissues in vivo. Even delivery to relatively accessible organs such as the liver required techniques not amenable to clinical application in humans. Now Soutschek et al. present evidence that small interfering RNA (siRNA), delivered via a clinically acceptable route (IV), can inhibit gene expression with predictable and physiologically relevant results in vivo (4). The large and frequent doses of siRNA required, coupled with the potentially confounding technique of administering a cholesterol-containing drug for a lipid disorder, make it unlikely that this gene-specific approach will replace conventional drugs for hypercholesterolemia in the near future. But the broader implication of this study is that siRNA can be delivered to cells in vivo with simple modifications of the RNA backbone and coupling of the siRNA to cholesterol. For diseases in tissues accessible to intravenous administration, siRNA therapeutics may soon be a reality.

    What does this study mean for brain diseases? Clearly the CNS is less accessible than the liver or gut. The blood-brain barrier would probably exclude siRNA delivered intravenously, though cholesterol (or other lipid) conjugation might improve uptake by neurons if injected into the cerebrospinal fluid or brain parenchyma. On the bright side, effective siRNAs against neurodegenerative disease genes such as APP, tau, BACE1 and SOD1, to name a few, have already been developed and tested in cell culture (5-7). Further, a notable in vivo success was recently achieved when a mouse model of spinocerebellar ataxia type 1 (in which neurodegeneration is caused by polyglutamine expansion) was treated by viral delivery of siRNA. In this model, viral delivery to a minority of Purkinje cells led to a therapeutic benefit (8). Current viral vectors must be directly injected into the brain and may prove to be inefficient for diseases with more widespread pathology such as Alzheimer disease, but this growing body of work shows that in vitro validated siRNAs can indeed work in vivo if delivered to the right neurons.

    Additional and complementary modes of delivery may be crucial to widespread application of RNAi-based therapy to brain diseases. More hope is offered this week by a report demonstrating that conjugation of the peptide Penetratin1 to siRNA results in efficient delivery to primary neurons in culture (9). If it proves generally true that siRNA delivery can be enhanced by the addition of carrier molecules, this may open the door to tissue-specific targeting of gene-specific drugs. What we currently do not know is whether siRNAs will prove useful for chronic, slowly progressive conditions such as neurodegenerative diseases in which treatment duration will need to be long-term. Nonetheless, the tools provided by these two new studies will allow rigorous testing of the efficacy and safety of siRNA in animal models. They constitute a significant step toward turning siRNAs into viable drugs.


    . Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature. 2001 May 24;411(6836):494-8. PubMed.

    . Therapeutic potential of RNA interference. N Engl J Med. 2004 Oct 21;351(17):1772-7. PubMed.

    . Molecular medicine for the brain: silencing of disease genes with RNA interference. Lancet Neurol. 2004 Mar;3(3):145-9. PubMed.

    . Therapeutic silencing of an endogenous gene by systemic administration of modified siRNAs. Nature. 2004 Nov 11;432(7014):173-8. PubMed.

    . Targeting Alzheimer's disease genes with RNA interference: an efficient strategy for silencing mutant alleles. Nucleic Acids Res. 2004;32(2):661-8. PubMed.

    . BACE1 suppression by RNA interference in primary cortical neurons. J Biol Chem. 2004 Jan 16;279(3):1942-9. PubMed.

    . Selective silencing by RNAi of a dominant allele that causes amyotrophic lateral sclerosis. Aging Cell. 2003 Aug;2(4):209-17. PubMed.

    . RNAi suppresses polyglutamine-induced neurodegeneration in a model of spinocerebellar ataxia. Nat Med. 2004 Aug;10(8):816-20. PubMed.

    . Highly efficient small interfering RNA delivery to primary mammalian neurons induces MicroRNA-like effects before mRNA degradation. J Neurosci. 2004 Nov 10;24(45):10040-6. PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. RNA Interference in Vivo—Viral Vectors Deliver

Paper Citations

  1. . Steroid and lipid conjugates of siRNAs to enhance cellular uptake and gene silencing in liver cells. Bioorg Med Chem Lett. 2004 Oct 4;14(19):4975-7. PubMed.

Further Reading

Primary Papers

  1. . Therapeutic silencing of an endogenous gene by systemic administration of modified siRNAs. Nature. 2004 Nov 11;432(7014):173-8. PubMed.