Several research groups, including the authors, have found increases in BACE levels and even proteolytic activity in the brains of either sporadic or familial AD patients (Holsinger et al., 2002; Fukumoto et al., 2004; Yang et al., 2003; see also ARF related news story). In the current article, Li and colleagues sought to evaluate this increase in BACE protein relative to Aβ load—something they could not show in their previous study because in it, they did not separate out different Aβ species (Yang et al., 2003).

As in the previous study, the authors this time studied autopsy tissue frozen to -80 degrees less than three hours after death. At that time point, there are still many living cells, including some neurons. The assumption is that they are evaluating BACE levels and activity in a state close to that of the living brain. The authors found that elevations in BACE protein (on Western blot) in frontal cortex of six AD patients relative to six controls, was accompanied by elevations in Aβ (42 and other forms of Aβ, but not 1-40). An ELISA assay on a group of 20 AD cases and 20 controls showed increases over control of 28 percent in temporal cortex and 30 percent in hippocampus, with no increases in cerebellum.

Contrary to some earlier studies, Northern blots showed concurrent increases in BACE mRNA in four of six cases relative to controls. In the two AD cases that did not show this increase, the authors detected an increase with an RNase protection assay. The authors also found that BACE proteolytic activity was increased by 23 percent in temporal cortex relative to control. This increase was significantly correlated with both plaque load and length of disease in AD patients. The proteolytic activity should presumably be accompanied by an increase in the BACE cleavage product C99, and the researchers indeed found more C99 fragments immunoprecipitated from AD than control brain despite equal APP levels between the two groups.

The researchers also looked for disease-linked mutations or polymorphisms in the BACE gene, but found none. In a separate experiment in cells transfected with AD-causing mutations (APPsw or PS1), the researchers determined that neither mutation increased BACE expression directly in vitro.

This study leaves open the question of why BACE is elevated in AD brain. There have been suggestions from in-vitro studies that cytokines upregulate BACE, perhaps via nuclear receptor peroxisome proliferator-activated receptor-γ (PPARγ) (see ARF related news item), and that the sphingolipid intermediate ceramide stabilizes BACE, boosting Aβ levels (see ARF related news item). Li, Shen, and colleagues also offer the suggestion that an unknown process stabilizes BACE mRNA long enough to allow more production of BACE.—Hakon Heimer.

Reference:
Li R, Lindholm K, Yang LB, Yue X, Citron M, Yan R, Beach T, Sue L, Sabbagh M, Cai H, Wong P, Price D, Shen Y. Amyloid {beta} peptide load is correlated with increased {beta}-secretase activity in sporadic Alzheimer's disease patients. Proc Natl Acad Sci U S A. 2004 Feb 20 [Epub ahead of print] Abstract