Concocting a Knock-Out Punch for HIV-1

RISCY BUSINESS:Courtesy of Roger J. Pomerantz and the Center for Human Virology and Biodefense ©2003 John Wiley & Sons, Ltd.Dicer cleaves exogenous or endogenous double-stranded RNA (dsRNA) into 21–25-nucleotide small interfering RNAs (siRNA). The siRNAs then form into ATP-containing RNA-induced silencing complexes (RISCs), which in combination with helicase lead to siRNA unwinding. Unwound siRNAs bind target RNAs and prime synthesis of new dsRNA by RNA-dependent RNA polymerase (R

Mar 29, 2004
Aileen Constans

Courtesy of Roger J. Pomerantz and the Center for Human Virology and Biodefense ©2003 John Wiley & Sons, Ltd.

Dicer cleaves exogenous or endogenous double-stranded RNA (dsRNA) into 21–25-nucleotide small interfering RNAs (siRNA). The siRNAs then form into ATP-containing RNA-induced silencing complexes (RISCs), which in combination with helicase lead to siRNA unwinding. Unwound siRNAs bind target RNAs and prime synthesis of new dsRNA by RNA-dependent RNA polymerase (RdRP). The newly synthesized RNA is then degraded by Dicer or related enzymes. In Drosophila and plants, this becomes a catalytic system leading to increased production of siRNAs and amplification of the RNAi effect. (R.S. Dave, R.J. Pomerantz, Rev Med Virol, 13:373–85, 2003.) across the world, researchers

Since the technique hit lab benches across the world researchers have assessed the specificity and power of gene silencing through RNA interference. RNAi has found use both as a research tool and as a potential therapeutic agent. In RNAi, an enzyme called Dicer processes long, double-stranded RNA (dsRNA) transcripts into smaller (21–25 nucleotide) dsRNAs, which target homologous messenger RNAs for degradation.

Researchers noted the phenomenon in plants more than a decade ago, and more recently it was rediscovered and refined in nematodes. In 2001, Tom Tuschl's laboratory at Rockefeller University published evidence in mammals that RNAi, mediated by short interfering RNAs (siRNAs) roughly 21 nucleotides in length, could be used to knock down gene expression.1 Within a year, a handful of laboratories used this tool to target cellular and viral proteins necessary for human HIV-1 infection. Some of these studies have since become Hot Papers.

Data derived from the Science Watch/Hot Papers database and the Web of Science (Thomson ISI) show that Hot Papers are cited 50 to 100 times more often than the average paper of the same type and age. J.M. Jacque et al., "Modulation of HIV-1 replication by RNA interference," Nature, 418:435–8, 2002. (Cited in 115 papers) N.S. Lee et al., "Expression of small interfering RNAs targeted against HIV-1 rev transcripts in human cells," Nat Biotechnol, 20:500–5, 2002. (Cited in 155 papers) C.D. Novina et al., "siRNA-directed inhibition of HIV-1 infection," Nat Med, 8:681–6, 2002. (Cited in 116 papers)

"All three papers basically demonstrate, to varying degrees, that it might be possible to use these small RNAs to inhibit HIV replication ... in the future, in a patient," says Phillip Sharp at Massa-chusetts Institute of Technology.

Though a simple plan, complications arose quickly. HIV, a particularly slippery fish, evolves resistance to most known treatments. Further, questions remain about what aspects of the HIV (or host) life cycle would best serve as targets, and how such therapies should be administered. Nevertheless, these articles raise a provocative possibility.

Mario Stevenson's group at the University of Massachusetts Medical School, Worcester, applied dsRNAs to primary lymphocytes, the main target for HIV. They showed that the viral RNA genome, which is introduced into the cell upon infection, could be targeted with this approach.2 "If you can get at that part of the viral life cycle when it first comes into the cell, you can prevent its establishment [and] essentially sterilize the cell from infection," he says.

Likewise, John Rossi and colleagues at the City of Hope National Medical Center, Duarte, Calif., used DNA-based vectors to generate siRNAs targeted against the HIV-1 rev sequence; they demonstrated up to four logs of inhibition of HIV expression.3 Yet, even with powerful repression, viral resistance would likely become a problem.

Thus researchers tried a combination approach that attacked both the virus and the cellular mechanisms it usurps. In this issue's third Hot Paper, Sharp and Judy Lieberman of Harvard Medical School introduced siRNAs directly into cells to target both the coreceptor CD4 and the HIV-1 gag gene.4 They found that CD4 silencing inhibited viral entry in primary T cells and that gag silencing likewise knocked down viral transcripts.


At first glance, the outcome of these experiments seems obvious: RNA interference is an RNA silencing mechanism; HIV-1 needs RNA to replicate. EMBO Journal rejected a related study submitted earlier by Bryan Cullen and colleagues at Duke University Medical School, Cullen says, because the reviewers called the results predictable. The work was subsequently published in the Journal of Virology.5

The Hot Paper authors similarly weren't astonished by the result. "At that time, it was becoming clear that the RNA-interference machinery could be educated to attack any RNA, whether it was viral or cellular. So it wasn't really a surprise that we could do this," says Stevenson. The ease, efficiency, and potency of the technique did amaze, though, he says.

But researchers recognized a critical limitation in viral resistance: Long-term therapy targeting viral genes alone would likely be impossible. "A single point mutation is enough to make the virus resistant to RNA interference," says Cullen.



Courtesy Carl D. Novina, Helen Cargill and Philip A. Sharp

People possessing a mutant form of CCR5 that is deficient in HIV binding have normal immune function but are resistant to HIV infection. Due to a high HIV mutation rate and concerns about generating virus escape mutants, cellular targets are preferable to viral targets. Thus, reducing CCR5 expression by RNA interference is a potential application for therapeutic gene silencing.

Indeed, Ben Berkhout's group at the University of Amsterdam demonstrated HIV escape using retroviral transduction, which stably introduces siRNAs targeting the HIV-1 Nef gene in T cells.6 They discovered several mutants that resisted RNAi, all of which contained Nef mutations. "The escape actually demonstrates that the initial pressure that was put on the virus was really potent ... it really is a perfect demonstration of the specificity of the RNA interference," says Berkhout. If multiple targets are used, escape would be more difficult, he adds, making an effective HIV therapy "doable."

Other laboratories followed up on CD4 targeting by going after the corecep-tor CCR5; humans who are homozygous for a defect in this gene are naturally resistant to HIV infection.78 Sharp speculates that if the technology will work as a therapy, it will do so by silencing cellular genes. "Not CD4, but more likely CCR5, whose expression may not be critical for normal immune function."

But this may not be the answer, either. "You can certainly imagine that HIV-1 can mutate and utilize different coreceptors within a host," says virologist Irvin Chen, director of the AIDS Institute at the University of California, Los Angeles. Chen adds, however, that such mutations would not likely impede the effectiveness of siRNA targeted against CCR5, especially if used in conjunction with other therapies.

Still, Roger Pomerantz of the Division of Infectious Diseases, Thomas Jefferson University, Philadelphia, says his own studies have shown that RNAi is not as robust an inhibitor of HIV-1 as previously thought. "It depends on the small interfering RNA that you pick, the phase of the life cycle, [and] how you set up your system." Pomerantz adds that with gene therapy in general, viruses can overcome the treatment through resistance, or even by simply overwhelming the system.


Even if such problems are overcome, the problem of delivery to cells remains. "In order to use [siRNAs] as a drug, you have to figure out a way either to introduce these small duplex RNAs, to get them through the plasma membrane of cells, or you have to express them, or express precursors of them, in cells," says Lieberman. Some approaches currently under investigation include expression of small hairpin RNAs from a viral vector, which then can be processed into siRNAs in vivo, and direct injection of modified siRNAs into cells.

Rossi's group put the construct they had used in their 2002 hot paper into a lentiviral vector to transduce siRNAs into hematopoietic progenitor cells.9 "We've gone a step further," Rossi says. His group took primary progenitor cells from donors, and modified them so that maturing macrophages, T cells, and monocytes would all be highly resistant to HIV infection.

Nevertheless, little is known about possible side effects of RNAi-based therapy. "We don't really understand enough about the complex mechanisms that are involved in RNAi, siRNAs, microRNAs, and what their normal role is in cells," says Chen. "I think the danger is that if we move too quickly, we could start interrupting normal pathways in cells that can cause unwarranted side effects."

While a growing number of scientists question the future of RNAi as an antiviral therapy, some point to new evidence that RNAi may be an ancestral defense against dsRNA-containing viruses. But none downplays the technique's importance as a tool to study viruses such as HIV. Pomerantz cites the CCR5-inhibition studies as an example of how this technique can be used to tease out aspects of the viral life cycle. Says Cullen: "It's having a dramatic effect on the molecular biology of human cells. You can really ask for the first time, what does this protein actually do in the cell, and you could never do that before in a direct way."

Aileen Constans can be contacted at