. High beta-secretase activity elicits neurodegeneration in transgenic mice despite reductions in amyloid-beta levels: implications for the treatment of Alzheimer disease. J Biol Chem. 2005 Sep 23;280(38):32957-67. PubMed.


Please login to recommend the paper.


  1. A number of human (h) BACE1 (β-site APP cleaving enzyme) transgenic mice have been generated and described. While these papers emphasized changes in APP processing, and in one case, behavior, they did not shed light on potential neurodegenerative pathology in these mice. The recent paper from the laboratory of Masliah and colleagues fills that gap. They created transgenic mice expressing high levels of hBACE1 from the mThy1 promoter, and assessed APP processing, neuropathology, and neurological decline in the mice. It is a real plus that this group analyzed three groups of transgenic mice—hBACE1 and hAPP singly transgenic mice, and hBACE1/hAPP doubly transgenic mice—thereby enabling the researchers to compare the phenotype produced by each of these genotypes.

    Masliah and colleagues found that high levels of hBACE1 expression in mice led to profound changes in APP processing, brain pathology, and neurological state in these animals. They discovered that the high levels of hBACE1 activity in these transgenic mice caused an increase in the cerebral accumulation of APP C-terminal fragments (CTFs) but a reduction in Aβ. Others have obtained the opposite result: Their hBACE1 transgenic mice showed an increase in the production of Aβ. This seeming contradiction may be resolved by the finding of Lee et al. (ref 19 in Masliah paper) that moderate hBACE1 expression in transgenic mice enhanced Aβ deposition, while high hBACE1 expression inhibited it. Their evidence suggests that high levels of hBACE1 activity shift the locus of APP processing from the axon terminals, where γ-secretase activity predominates, to the neuronal perikarya early in the secretory pathway.

    The most interesting data obtained by first author Rockenstein et al. concerns the effect of high levels of hBACE1 activity on brain pathology and neurological state. The transgenic mice displayed shrinkage of pyramidal neurons in the CA3 region of the hippocampus and in layers two and three of the neocortex. Axons and dendrites of pyramidal neurons in the hBACE1 and hBACE1/hAPP transgenic mice were “diminished, disrupted, and vacuolyzed.” Similar neurodegeneration, but to a less extent and not in the CA3 region of the hippocampus, was displayed by hAPP transgenic mice. Immunoreactivity for the dendritic marker MAP2 was reduced in both hBACE1 and hAPP transgenic mice, and even more greatly reduced in hBACE1/hAPP mice. Significant neurological deficits also were observed in the hBACE1-overexpressing transgenic mice.

    The take-home lesson from this paper is that overproduction of APP CTFs due to high hBACE1 activity causes neurodegeneration and neurological decline in transgenic mice despite the fact that Aβ production is down. The authors speculate that the increase in APP CTFs coupled with a decrease in Aβ generation unmasks an alternative neurodegenerative pathway mediated by APP CTFs that is independent of Aβ, and that may be operative in Alzheimer disease (AD).

    Several questions remain. The authors show that the predominant APP CTF generated in the transgenic mice is not C99, the product of cleavage at the β site, but is C89, the product of cleavage at the β’ site, 10 amino acids downstream of the β site. This is curious, because C89 is normally a minor cleavage product of hAPP relative to C99. It raises the question of whether C89 or C99 (or both) is the culprit causing the neurodegeneration. APP CTF levels have been shown to be increased in affected regions of AD brain; it will be informative to determine the ratio of C89 to C99 in these regions and compare it to the C89:C99 ratio in control brain. If it turns out to be the case that C89 and C99 contribute equally to the neurodegeneration, the question arises as to what these two APP cleavage products have in common. As the authors point out, the C31 subfragment that is present in both of these cleavage products has been shown to be neurotoxic. It will be interesting, then, to determine whether CTFs derived from APP that is mutated at the C31 cleavage site can cause neurodegeneration. A final question concerns the level of C83, the C-terminal product of α-secretase cleavage, in these mice. The authors state in the abstract that the level not only of C89 but also of C83 is increased in the brains of the hBACE1 transgenic mice. It is not clear why high β-secretase activity would cause a parallel increase in α-secretase activity. It also is not clear why the authors state in their discussion that their data are in agreement with those who have shown decreased α-secretase cleavage products of APP in hBACE1 transgenic mice. This point of confusion needs to be cleared up. Otherwise, this is a provocative paper that is consistent with a number of previous studies demonstrating neurotoxicity of APP CTFs both in vitro and in vitro; the novel finding in this paper is that this CTF neurotoxicity occurs in the presence of a decline in the production of Aβ.

Make a Comment

To make a comment you must login or register.

This paper appears in the following:

Research Models

  1. mThy1-hAPP751 (TASD41)


  1. Rounding Second BACE: Another Route to Memory Loss?