20 Sep 2005

Among the strategies currently being pursued for treatment of Alzheimer disease (AD), inhibiting β-secretase would seem to be a front runner for three reasons. [Also called BACE, it is one of two proteases involved in clipping the Aβ peptide from amyloid-β precursor protein (AβPP)]. First, unlike γ-secretase, BACE does not appear to be essential because mice lacking the gene survive and breed (see ARF milestone paper Roberds et al., 2001). Second, inhibiting BACE would prevent production of Aβ as well as of AβPP C-terminal fragments, which have also been implicated in neurodegeneration (see ARF related news story). Third, proteases generally make decent drug targets. A recent advanced online publication in Nature Neuroscience reports that silencing the protease by using small interfering RNAs could become a viable alternative to standard drug therapy.

RNAi is beginning to look attractive as more conventional small molecule approaches continue wrestling with the structural difficulties posed by BACE (see ARF related news story). Eliezer Masliah at the University of California, San Diego, Inder Verma at The Salk Institute, La Jolla, and colleagues report that they can prevent AD-like neuropathology in mouse models of the disease by using small interfering RNAs (siRNAs) to silence BACE. Small interfering RNAs prevent protein synthesis by promoting degradation of the protein’s template, its messenger RNA. This means RNA interference can bypass the need to come up with an inhibitor for the protease.

A drawback to siRNA therapy is that the siRNAs can be difficult to deliver to their intended targets. Joint first authors Oded Singer, Robert Marr, and colleagues overcame this problem by using intracranial injection of lentiviral particles (see ARF related news story) to deliver the siRNAs to neurons in the hippocampus, an area of the brain that can be severely affected by AD pathology. To test the effectiveness of BACE silencing, the authors delivered the viral particles to mice harboring human AβPP with London (V717I) and Swedish (K670M/N671L) mutations (see Rockenstein et al., 2001). They found that levels of both C-terminal fragments and Aβ dropped by almost three- and twofold, respectively, in these mice compared to animals that received mock viral particles. The siRNA therapy also substantially reduced the numbers of amyloid plaques that accumulated in the brain tissue of these animals.

The gene silencing method appeared to prevent neurodegeneration, the authors claim, as microscopic analysis revealed more preserved dendrites and synapses in mice that received the lentivirus particles. This may explain why the mice did better in a water maze test of learning and memory, the authors report.

To date, there have been several examples of siRNAs rescuing deficits in mice, including neuronal deficits (see, for example, ARF related news story on the rescue of spinocerebellar ataxia in mice), though it remains unclear when the technology might be deemed safe enough to use in humans. Meanwhile, the field is advancing rapidly. Only a few years ago researchers were uncertain if neurons even had the wherewithal to carry out siRNA-mediated degradation of messenger RNA (see ARF related news story). Now, as Kenneth Kosik and Anna Krichevsky outline in a mini-review in the September 15 Neuron, microRNAs, which can both stabilize and degrade messenger RNA, are an indispensable part of the nervous system and probably play vital roles in such diverse processes as neuronal differentiation (see ARF related news story) and modulating protein synthesis at synapses.—Tom Fagan