Similarly, much excitement greeted the still somewhat controversial identification of γ-secretase, and inhibitors of this aspartyl protease have made their way into early clinical trials. Lately, however, γ-secretase has lost some of its luster as follow-up basic research identified additional substrates. Including notch and others, these targets raise concerns about side effects of γ-secretase inhibition and make designing a drug that hits one γ-secretase activity but spares the others a tricky business (see, for example, related news item.)

Will a similar fate befall BACE1? Unlikely, says Philip Wong of John's Hopkins University, Maryland. Today at the Neuroscience meeting, Wong, working with Don Price and others, presented data that appear to make BACE1 an even more appealing target. In one study, Wong asked why only brain is prone to develop amyloidosis when all cells of the body process AβPP. His group found that BACE1 is much more abundant in brain than in non-neural tissues. Studies in mice show that BACE1 is highly expressed in hippocampus, olfactory bulb and posterior cortex, areas affected in AD, but not in the cerebellum, a motor and balance-control region less damaged in AD. At the same time, high BACE1 was coupled with low levels of BACE2 (which acts more like α-secretase in that it limits Aβ generation), whereas low levels of BACE1 and high levels of BACE2 were seen in non-neural cells, such as fibroblasts.

Moreover, Wong's group has crossed a BACE1 knockout into PS1+APP transgenic mice to address the question whether partial BACE1 inhibition might be sufficient to slow plaque formation. The researchers presented data showing that a heterozygous BACE1 knockout, which still produces 50 percent of the normal BACE1 protein levels, led to marked reduction of plaques and soluble Aβ levels in this mouse model of amyloid deposition. The homozygous BACE1 knockout/PS1+APP transgenic mice had no detectable Aβ, Wong said in a press conference.

When localizing BACE1 in mouse hippocampal neurons, Wong et al. found it in the dystrophic terminal dendrites that surround mature amyloid plaques. Interestingly, Yong Shen of the Sun Health Research Institute in Sun City, Arizona, who recently succeeded in isolating and culturing living neurons from rapidly autopsied AD brains, also detected endogenous BACE1 in the terminals of those neurons.

In related news, M.J. Chiocco and Bruce Lamb of Case Western Reserve University in Cleveland, Ohio, report the development of a new tool to study the role of BACE1 and BACE2 in AβPP processing and Aβ deposition. These researchers created genomic-based BACE1 transgenic mice that overexpress human BACE1 mRNA and protein and crossed these with APP transgenic mice. These mice should grossly overproduce Aβ. Wong noted that some evidence indicates that BACE1 activity increases in humans with age, suggesting that whatever causes this increase could contribute to late-onset Alzheimer's disease.

Caveats remain. While the finding that BACE1 knockout mice are viable and appear to develop largely normal bodes well for drug discovery, BACE2 might still turn out to have essential substrates whose inhibition could be deleterious. This question awaits the generation of BACE1/BACE2 double knockout mice.-Gabrielle Strobel.

Reference:Wong PC et al. BACE1 protein is most abundant in neurons: a determinant of Ab amyloidosis. Soc Neuroscience 2001.

Cai H et al. b-secretase (BACE1) and modulation of Ab amyloidosis in Alzheimer's disease. Soc Neuroscience 2001. Abstract

Yan X, et al. Endogenous expression of b-secretase in isolated living neurons from Alzheimer's brains and APP transgenic mouse brains. Soc Neuroscience 2001.

Chiocco MG, Lamb BT. Soc Neurosci 2001.