TORONTO—Since its discovery in 1990, RNA interference (RNAi) has moved quickly from obscure laboratory curiosity to genuine scientific breakthrough. Its ability to silence virtually any gene gives it the potential to treat such intractable conditions as Huntington’s disease and generalized dystonia. While its development as a therapeutic tool is still in its early stages, it may eventually become as important a tool in the clinic as it already has in the lab. The mechanisms and potential neurologic applications of RNAi were presented by Henry Paulson, MD, PhD, at the 129th Annual Meeting of the American Neurological Association.

“RNA interference is not like your typical pharmacologic treatment,” said Dr. Paulson. “It’s a technique that allows you to design a molecule that is specific for individual genes, that reduces expression of that gene product.” This makes it potentially ideal to treat diseases due to dominantly inherited genes, such as Huntington’s disease and DYT1 dystonia, along with some forms of Alzheimer’s disease and amyotrophic lateral sclerosis, he noted. Dr. Paulson is an Associate Professor of Neurology at the University of Iowa’s Carver College of Medicine in Iowa City.

A NATURAL PHENOMENON

RNAi is a natural phenomenon, found in organisms from yeasts to plants to mammals. Its normal function appears to be twofold: to protect against viruses and other exogenous gene sources, and to regulate gene expression. Both functions rely on the same machinery, which detects double-stranded RNA molecules in the cell, and prevents their translation into protein. “RNAi essentially shoots the messenger,” said Dr. Paulson, by either destroying the double-stranded RNA or silencing it without destruction.

A key event in the process of RNAi is detection and cleavage of double-stranded RNA by an enzyme appropriately named Dicer, in a multiprotein complex called RISC (RNA-induced silencing complex). The resulting small RNA fragments then serve as templates for the detection and cleavage of additional RNAs with the same sequence. This is the basis for the therapeutic application of RNAi: By introducing a double-stranded RNA that matches the gene to be silenced, one can effectively prevent protein synthesis from that gene for an extended period of time.

Dr. Paulson explained that RNAi has revolutionized investigation of gene activity in the lab. Simply by introducing the appropriately sequenced double-stranded RNA into a cell culture or animal model, it is possible to “knock down” production from any gene of interest. “It’s now very clear that the most potent way to silence a gene is through a small double-stranded intermediate,” he said. The first demonstration of RNAi in mammals was in 2001, and since then, the field has exploded. In that seminal study, the nuclear envelope gene lamin was silenced. “Not only was it extremely potent but there were very few of the nonspecific effects seen with previous antisense technology,” he elaborated, referring to a related but entirely artificial technique. “The double-stranded RNA used like this does a heck of a lot better job.”

When RNA itself is directly introduced to the cell, however, it is eventually degraded. This may be useful in the lab for short-term studies, but for disease therapy, longer-term effects are desired. The alternative is delivering DNA that codes for the RNA, via a viral vector. “The advantage is that you have sustained expression within the cell of the double-stranded RNA you want,” said Dr. Paulson. “Can this work in the brain? Yes.”

SUCCESS AND CHALLENGES IN POLYGLUTAMINE DISEASES

The most successful demonstration to date has been in spinocerebellar ataxia type 1, due to an expanded CAG repeat. Working with Beverly Davidson, PhD, and her group—also from the University of Iowa—Dr. Paulson’s group targeted a non-CAG region of the gene and showed in a cell culture model that ataxin-1 protein production can be reduced by RNAi to about 30% of normal levels. Importantly, the same success was seen in a spinocerebellar ataxia type 1 mouse model study by Xia et al, in which the double-stranded RNA was delivered via an adeno-associated viral vector. The virus efficiently transduced Purkinje cells in the cerebellum, in which the virally delivered RNAi led to lower ataxin-1 levels and the absence of proteinaceous inclusions, a hallmark of disease. Purkinje cells in the cerebellar molecular layer remained healthy rather than degenerating, and the mice improved on the performance of a motor task. An important concern is whether co-opting the RISC machinery for other than normal purposes will have adverse consequences. In this experiment, at least, no such effects were seen in treated mice. “RNAi delivered by virus seemed to be tolerated,” said Dr. Paulson.

In the spinocerebellar ataxia type 1 model, however, the target was a human gene in a mouse, which had no corresponding normal allele that might be unintentionally silenced by the treatment. Can mutant alleles be silenced without affecting protein production from the normal allele? Dr. Paulson’s group explored this for the entire class of polyglutamine disease, all nine of which are caused by CAG repeat expansions that encode abnormally long stretches of polyglutamine in the disease protein. First they tagged the mutant polyglutamine gene with a green fluorescent signal, and the normal gene with a red one. RNAi directed against messenger RNA for the green signal virtually eliminated it, without affecting expression of the red signal. The next step was to target the polyglutamine repeat itself, to see if the technique could discriminate between the different lengths of repeats found in the mutant and normal genes. Unfortunately, silencing the one silenced the other.

“The bottom line here is that we don’t believe you can target the CAG mutation directly without affecting the normal gene, which also has a CAG repeat,” Dr. Paulson said. “Since there are dozens of genes in the human genome that express CAG repeats, including some that code for very important proteins, this kind of approach would be hazardous because of unavoidable off-target effects.” In addition, the normal Huntington disease protein, huntington, is itself essential, further complicating the choice of target for this polyglutamine disease.

An alternative approach is to target polymorphic differences elsewhere in the gene. These small sequence differences between the mutant and normal alleles are not implicated in disease but might serve as a target sequence in the mutant allele that allows one to discriminate it from the normal allele. This has been tried in spinocerebellar ataxia type 3, also known as Machado-Joseph disease (MJD). In the MJD gene, there is a single nucleotide polymorphism, apart from but tightly linked to the CAG/ polyglutamine expansion, that can be either a G or C. In about 70% of mutant alleles, the single nucleotide polymorphism is a C, while in most normal alleles it is a G.

When Dr. Paulson’s group targeted the small region containing the variant C, they achieved allele-specific suppression. “This may be where the money is,” Dr. Paulson said. “It does a very nice job of specifically suppressing the pathogenic allele, at least in cellular models.”

FUTURE DISEASE TARGETS

DYT1 dystonia might be an ideal target for RNAi, Dr. Paulson said. This is because virtually every patient has the same mutation—deletion of a GAG triplet—and the disease acts through a dominant mechanism. Initial cell culture experiments are promising, and the group is moving into mouse models. “It’s only a first small step,” he said, but perhaps one with promise for a disorder in which there are few other effective treatments.

Alzheimer’s disease may also be an attractive disease to treat with RNAi, if the right target can be found. Presenilins that are responsible for gamma secretase activity are probably not that target, since they are essential. The genes encoding beta-secretase, BACE1, and amyloid precursor protein might be good candidates, Dr. Paulson speculated. Early experiments in cell-based models are promising. Whether this tactic will have relevance to the large number of patients with sporadic disease remains to be seen, but some progress might be made on understanding pathogenesis with this approach by focusing on the role of amyloid precursor protein in the disease.

“There are many challenges to come,” he cautioned, the most prominent of them being delivery to the target cells. Others include achieving sustained expression, minimizing potential toxic effects, and determining whether long-term treatment leads to subtle genomic changes from persistent alteration of gene output. “It’s fair to say we do not yet fully understand the normal roles played by this machinery in the cell, and it may well include regulating chromatin structure.” To date, no evidence in animal models suggests this will be a problem, but with only a few years of sustained research with this system, it seems likely that RNAi still holds some surprises for those who are studying it.

NR

—Richard Robinson

Suggested Reading
Elbashir SM, Harborth J, Lendeckel W, et al. Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature. 2001;411:494-498.
Harborth J, Elbashir SM, Bechert K, et al. Identification of essential genes in cultured mammalian cells using small interfering RNAs. J Cell Sci. 2001;114(pt 24):4557-4565.
Starr DA, Hermann GJ, Malone CJ, et al. unc-83 encodes a novel component of the nuclear envelope and is essential for proper nuclear migration. Development. 2001;128:5039-5050.
Xia H, Mao Q, Eliason SL, et al. RNAi suppresses polyglutamine-induced neurodegeneration in a model of spinocerebellar ataxia. Nat Med. 2004;10:816-820.

Return to table of contents