. BACE1 deficiency rescues memory deficits and cholinergic dysfunction in a mouse model of Alzheimer's disease. Neuron. 2004 Jan 8;41(1):27-33. PubMed.


Please login to recommend the paper.


  1. This is exactly what we predicted (Dewachter and Van Leuven, 2002) or, as the Americans would put it, "what the doctor ordered…"

    The overall message of this study is loud and clear: BACE is hereby proven to be the favorite target for drug-makers. That message rings even more clearly because γ-secretase inhibitors now prove bad not only for the brain, as we predicted, as well (Dewachter et al., 2002), but also in vivo for the immune system, the intestine, and likely for any, or even all, biological subsystems in our complex bodies that depend on intramembranous proteolysis for essential functions (Wong et al., 2004).

    Some points remain somewhat worrying or startling in this study. First, why not use the "classic" cognitive test, i.e., the water maze? Second, why not measure "classic" LTP instead of the somewhat exotic cholinergic AHP? Third, and most important, is the lack of any data on APP biochemistry. I, for one, would want to know what happens to APPs-α and to the α-CTF; the discussion circumvents this issue. Indeed, the lowering of Aβ by BACE-deficiency is spectacular and likely to be a major contributor to the "rescue." Nevertheless, APPs-α is claimed by many to be beneficial for brain, but only circumstantially demonstrated to be essential for "neuronal well-being."

    Here was a chance to increase the experimental proof for that action.


    . Secretases as targets for the treatment of Alzheimer's disease: the prospects. Lancet Neurol. 2002 Nov;1(7):409-16. PubMed.

    . Neuronal deficiency of presenilin 1 inhibits amyloid plaque formation and corrects hippocampal long-term potentiation but not a cognitive defect of amyloid precursor protein [V717I] transgenic mice. J Neurosci. 2002 May 1;22(9):3445-53. PubMed.

    . Chronic treatment with the gamma-secretase inhibitor LY-411,575 inhibits beta-amyloid peptide production and alters lymphopoiesis and intestinal cell differentiation. J Biol Chem. 2004 Mar 26;279(13):12876-82. PubMed.

    View all comments by Fred Van Leuven
  2. By Dione Kobayashi and Karen Chen The paper by Disterhoft et al. reporting cognitive and cholinergic rescue with their BACE1 knockout mice on a mutant hAPP background is extremely exciting and is a strong validation of BACE inhibitory strategies for Alzheimer's disease therapeutic efforts. Other recent works over the past three years have also reported on the behavioral effects of complete genetic removal of BACE1 from mice (Harrison et al., 2003; Kobayashi et al., in SFN abstracts 2002, 2003). In addition to their novel cholinergic function results, Disterhoft et al. found mild phenotypes in exploration in BACE1 -/- mice similar to other groups, although there is some discrepancy regarding the ability of BACE1 deletion to rescue cognitive deficits due to overexpression of mutant hAPP.

    It must be emphasized that not only do the various BACE1 -/- mice differ in their hAPP mutations, level of APP overexpression, and thus their subsequent cognitive deficits, but these mice also have been subjected to widely divergent behavioral tasks. While the Y maze and social recognition tasks used by Disterhoft et al. are both nonaversive and rely on hippocampal function, these tasks also depend on the motivational state of the animals, which is admittedly impaired in spontaneous exploration.

    In studies done by our group, BACE1 -/- PDAPP mice were tested in an aversive serial spatial memory water maze paradigm that utilizes working memory (Chen et al., 2000). This modified water maze protocol was in effect designed to detect subtle spatial impairments in the PDAPP mouse that could be missed when using other spatial memory tasks, including the classic reference memory task described by Morris in 1981. While we do report significant progressive spatial memory deficits in our BACE1 -/- PDAPP mice, their impairments can also be viewed as subtle and specific. Thus, our seemingly opposing BACE1 -/- hAPP data may simply represent the dynamic range of changes in BACE1 -/- hAPP mice tested with different cognitive tasks. Alternatively, these results may be entirely independent, underlining the murkiness inherent in working with transgenic animal models.

    It will be interesting to watch this promising line of research develop, as I agree with the authors wholeheartedly that further study with other transgenic mouse lines, as well as other behavioral assessments, will be illuminating, particularly if a conditional BACE1 -/- is developed like that reported with presenilin-1 (Yu, 2001). In addition, the mild behavioral phenotypes consistently reported with BACE1 -/- alone should not be overlooked from a perspective of understanding the role of the constituents of the APP processing pathway in normal learning and memory.


    . BACE1 (beta-secretase) transgenic and knockout mice: identification of neurochemical deficits and behavioral changes. Mol Cell Neurosci. 2003 Nov;24(3):646-55. PubMed.

    . A learning deficit related to age and beta-amyloid plaques in a mouse model of Alzheimer's disease. Nature. 2000 Dec 21-28;408(6815):975-9. PubMed.

    . APP processing and synaptic plasticity in presenilin-1 conditional knockout mice. Neuron. 2001 Sep 13;31(5):713-26. PubMed.

    View all comments by Karen Chen
  3. This elegant report further supports BACE1 as a rational therapeutic target for treating the cerebral amyloidosis of Alzheimer's disease (AD). Ohno and colleagues eliminated BACE1 function in a mouse model of AD by crossing Tg2576 mice—which overexpress the Swedish mutant of amyloid precursor protein (APPSwe)—with BACE1 knockout mice. They found that genetic elimination of BACE1 blocked cerebral amyloid β-protein (Aβ) production, ameliorated the cognitive deficits of the APP-transgenic mice in hippocampal-based learning tasks, and improved the cholinergic electrophysiologic deficits in hippocampal slice preparations.

    The Tg2576 APPSwe mouse model of Alzheimer's disease develops cerebral amyloid deposits by the age of nine to 11 months [1]. In the current study, soluble Aβ in brain was increased by 25-fold relative to nontransgenic mice at four to six months—the age at which cognitive deficits and electrophysiologic deficits were detected. The findings that certain cognitive [2] and electrophysiologic deficits occur prior to cerebral amyloid deposition suggest that soluble species of Aβ adversely affect physiologic processes in the brain. The rapid amelioration of behavioral deficits in APP-transgenic mice by passive immunization with anti-Aβ antibodies [3,4] further supports soluble forms of Aβ as a functionally toxic species in these mice.

    Ohno and colleagues tested the effects of eliminating brain Aβ in the Tg2576 APPSwe mice by knockout of BACE1, which is the principle enzyme responsible for the β-secretase cleavage of APP, the initial step of Aβ synthesis. Overexpression of BACE1 accelerates Aβ production in mouse brain [5,6], and BACE1 protein levels and enzymatic activity are increased in the AD brain [7-9]. Targeting BACE1 therapeutically is potentially less toxic than reducing γ-secretase activity since BACE1 knockout mice do not exhibit overt neurological or medical problems [10]. (However, BACE1 null mice alone did exhibit impaired performance in a spatial working memory task, a deficit that resolved in the APPSwe/BACE1 null mice [1]). This paper demonstrates that behavioral and electrophysiologic deficits in APPSwe mice can be ameliorated by elimination of Aβ production in the APPSwe/BACE1 null mice, implicating Aβ (or, less likely, β-cleaved APP fragments) as the source of these deficits.

    While very encouraging for BACE1 therapeutic strategies, the usual caveats regarding translating behavioral tests in mice to the cognitive deficits in human AD apply. Cognitive deficits in AD are most consistently associated with the development of neuronal loss and neurofibrillary tangle formation in corticolimbic regions, pathological features which are not reproduced in the Tg2576 APPSwe mice. Prolonged exposure to high levels of toxic forms of Aβ may initiate distinct downstream effects in mice (behavioral deficits, cholinergic electrophysiological deficits) and humans (neuronal and synaptic loss, cholinergic deficits, neurofibrillary tangle formation). The amyloid hypothesis postulates that reducing Aβ would prevent these downstream effects; the results presented in this paper support this idea in the case of certain deficits in the APPSwe transgenic mice.


    . Correlative memory deficits, Abeta elevation, and amyloid plaques in transgenic mice. Science. 1996 Oct 4;274(5284):99-102. PubMed.

    . The relationship between Abeta and memory in the Tg2576 mouse model of Alzheimer's disease. J Neurosci. 2002 Mar 1;22(5):1858-67. PubMed.

    . Immunization reverses memory deficits without reducing brain Abeta burden in Alzheimer's disease model. Nat Neurosci. 2002 May;5(5):452-7. PubMed.

    . Reversible memory loss in a mouse transgenic model of Alzheimer's disease. J Neurosci. 2002 Aug 1;22(15):6331-5. PubMed.

    . Expression of human beta-secretase in the mouse brain increases the steady-state level of beta-amyloid. J Neurochem. 2002 Mar;80(5):799-806. PubMed.

    . Transgenic BACE expression in mouse neurons accelerates amyloid plaque pathology. J Neural Transm. 2004 Mar;111(3):413-25. PubMed.

    . Increased expression of the amyloid precursor beta-secretase in Alzheimer's disease. Ann Neurol. 2002 Jun;51(6):783-6. PubMed.

    . Beta-secretase protein and activity are increased in the neocortex in Alzheimer disease. Arch Neurol. 2002 Sep;59(9):1381-9. PubMed.

    . Elevated beta-secretase expression and enzymatic activity detected in sporadic Alzheimer disease. Nat Med. 2003 Jan;9(1):3-4. PubMed.

    . Mice deficient in BACE1, the Alzheimer's beta-secretase, have normal phenotype and abolished beta-amyloid generation. Nat Neurosci. 2001 Mar;4(3):231-2. PubMed.

    View all comments by Michael Irizarry
  4. Ohno et al. bred the BACE1 knockout (BACE1-/-) mice onto the Tg2576 APP transgenic background and tested the effects of BACE1 deficiency on Aβ production, behavioral performance, and cholinergic function at a young age (4-6 months), prior to amyloid plaque deposition. The authors showed, quite convincingly, that inhibition of Aβ production, as a result of BACE1 deficiency, rescued the behavioral deficit and cholinergic impairment present in Tg2576 transgenic mice.

    This result has several important implications: 1) It lends strong support to the amyloid hypothesis; 2) it strengthens the notion that BACE1 is a valid therapeutic target for AD intervention; and 3) it establishes that the behavioral abnormality seen in Tg2576 mice is caused by APP processing/Aβ production rather than APP overexpression.

    However, as the authors pointed out, BACE1 deficiency leads not only to inhibition of Aβ, but also to changes in other APP fragments (e.g., β-CTF). Therefore, a definitive link between Aβ and functional deficits cannot be established. In addition, the therapeutic potential of BACE1 inhibitors has to be interpreted with caution since BACE1 knockout mice alone show impaired performance in the Y maze test (Fig. 1B). Although this impairment is neutralized on the APP-overexpressing background, it represents an artificial condition as AD individuals do not overexpress APP. In this regard, it is prudent to test the BACE1 knockout and BACE1-/- Tg2576 animals in other behavioral paradigms, such as the Morris water maze or fear conditioning.

    View all comments by Hui Zheng
  5. This is an interesting paper, and these mice are very useful for addressing a number of issues.

    The behavioral paradigms the authors chose to use are not the strongest learning and memory tests available; more robust and better established hippocampal-dependent learning and memory paradigms, such as the Morris water maze and the contextual fear conditioning tests, might have been preferable. This could be the reason that in the spontaneous alternation Y maze, even the Tg2576 APP-transgenic mice did not perform that poorly compared to the control. More importantly, the Y maze results require more explanation: I wonder how the double-mutant mice could have behaved "normally" while BACE-/- and APP-Tg mice performed poorly (Figure 1B)? It seems a bit premature to conclude "rescue" from such results.

    Without going into details of the physiology result, Figure 2C seems to show lower values in the double-mutant group compared to the control, as well. Nevertheless, it is interesting and welcome that the authors looked at the cholinergic input.

    Overall, from the data shown in the paper, it appears that the story is more complicated, and more thought-provoking, than a simple rescue story. BACE-/- mice appear to exhibit a memory deficit (due to loss of Aβ or other BACE
    substrates?) in one behavioral test but not the other, whereas BACE-/-xTg2586 bigenic mice appear to perform better than BACE-/- mice (due to ???). If this result can be confirmed by other behavioral tests, it warrants further investigation to characterize the underlying mechanism.

    Together with APP-transgenic mice crossed into the PS1 conditional KO background, the Ohno et al. mice stand to be great models to tease out the different contributions of various fragments of APP in brain function. This would require crossing the same APP transgenic line into either PS1 cKO or BACE null background. More importantly, a battery of robust learning and memory tests should be used to test these groups of mice together using identical protocols; only then we can compare results and make meaningful conclusions.

    View all comments by Jie Shen
  6. Previous studies showed that the deletion of BACE1 abolished the production of Aβ and BACE1 knockout mice are apparently normal. In this current paper, Ohno et al tested whether cognitive deficits occurring in the mutant APP mice (Tg2576) can be ameliorated in the absence of BACE1, results that have important implication for the potential therapeutic value of BACE1 in AD. Since developmental cognitive (as assessed by either social recognition task or spontaneous alternation in Y maze) and electrophysiological (hippocampal cholinergic dysfunction) abnormalities occur prior to Ab deposition in 4-6 months old Tg2576 mice, Ohno et al. elected to examine whether such memory deficits can be rescued in Tg2576 mice lacking BACE1. Their results demonstrating that the deletion of BACE1 prevented these early onset behavioral abnormalities strongly support their conclusion that increased levels of Aβ (as opposed to increased APP levels) causes the cognitive deficits occurring in Tg2576 animals.

    These authors also interpreted their findings to support the view that the inhibition of BACE1 has therapeutic value in reversing AD-associated cognitive deficits. However, it will be critical to test whether age-associated spatial memory deficits (as assessed by Morris water maze) occurring in aged Tg2576 mice can be rescued in the absence of BACE1. It is unclear at present whether Tg2576 mice lacking BACE1 will display age-associated spatial memory abnormalities; if they don't, such a positive outcome would strongly favor the idea that BACE1 inhibitors have the potential to ameliorate age-associated cognitive deficits occurring in AD.

    It is interesting to note that young (4-6 months of age) BACE1 knockout mice exhibit poor performance in spontaneous alternation in the Y maze, a deficit that is also observed in young Tg2576 mice. The reasons for this deficit are unclear. Because Tg2576 mice apparently can rescue this cognitive abnormality in BACE1 null mice and the levels of Aβ peptides in Tg2576;BACE1-/- mice were claimed to be similar to wild-type levels, these authors infer that the lack of Aβ in BACE1 null mice presumably is responsible for the behavioral deficits in the Y maze. This interpretation would raise the critical issue as to the origins of Ab peptides in the Tg2576;BACE1-/- mice. One uncertainty is whether the Aβ detected from brain extracts by the sandwich ELISA came from Aβ peptides derived from the processing of APP, or from APP-related fragments containing epitopes recognized by the antibodies used. That no Aβ peptides can be detected even when APP wild-type or APPSwe is expressed highly in BACE1-/- neurons, as previously demonstrated, would support the latter possibility.

    In this case, the lack of Aβ may not be sufficient to explain the behavioral deficits observed in these young BACE1 knockout mice. It is plausible that the lack of Aβ could affect other substrates of BACE1 to elicit the poor Y maze performance in the BACE1 null mice. However, because the behavioral abnormality observed in the BACE1 knockout mice can be rescued by the expression of APP in Tg2576 mice, this deficit is more likely related to the APP pathway rather than due directly to other BACE1 substrates. At any rate, further studies are necessary to clarify the behavioral abnormalities observed in the BACE1 null mice.

    These results raise the possibility that the therapeutic inhibition of BACE1 may not necessarily be free of mechanism-based toxicities, particularly if age-associated behavioral abnormalities are observed in BACE1-deficient mice. Thus, it would be critical to determine whether age-associated spatial memory deficits are observed in aged BACE1 null animals. Furthermore, the behavioral analysis of conditional BACE1 knockout mice should further clarify this important issue.

    In summary, while the paper by Ohno et al provides compelling evidence that early onset cognitive and cholinergic abnormalities occurring in young Tg2576 mice can be rescued by deleting BACE1, critical issues regarding the impact of BACE1 on age-associated cognitive deficits in aged mice remain to be established, as these results will have important implications for the development of potential therapeutics designed to inhibit BACE1 and ameliorate Aβ amyloidosis in AD.

    View all comments by Philip Wong
  7. This new report by Ohno et al. demonstrates further that BACE1, an
    enzyme involved in the maturation of the APP precursor and the
    generation of amyloid peptides, is a potential therapeutic target
    toward the treatment of Alzheimer's disease. In mice in which the BACE1
    gene was deleted, the overexpression of the human APP-695 Swedish
    familial mutation failed to result in memory deficits and altered
    cholinergic functions. Hence, the expression of BACE1 resulting in the
    production of pathogenic amyloid peptides is apparently key to inducing
    cognitive and neurochemical deficits in the model studied. Together,
    these data suggest that BACE1 inhibitors could prove useful in the
    treatment of AD by reducing the production of amyloid peptides and
    ensuing cholinergic deficits and learning impairments. Of course, the
    safety of such inhibitors would have to be established, but data
    obtained in the mouse model are promising. Moreover, these data link
    some of the key features of the AD brain, including amyloid peptides,
    cholinergic dysfunction, and memory deficits. It would now be of interest
    to investigate the effects of clinically used acetylcholinesterase
    inhibitors in this animal model, as well as subtype selective
    nicotinic and muscarinic receptor ligands.

    View all comments by Remi Quirion
  8. Jeon et al. (1) suspect that the pyrogallol moiety on C-2 and/or C-3 of the catechin skeleton is responsible for the increased inhibition of BACE1 by these green tea catechins.

    It seems of interest that Bain et al. (2) have found that epigallocatechin-3-gallate inhibits DYRK1A, one the the genes considered responsible for the mental retardation of Down's syndrome.

    Could we expect that therapeutic intervention with epigallocatechin 3-gallate may be beneficial for those with Down's syndrome?

    Basi et al. (3) find that BACE2 suppresses Abeta production in cells that also express BACE1.

    Motonaga et al. (4) report increased BACE2 levels in those with Down's syndrome with Alzheimer's-type pathology and suggest that BACE2 is involved in this neuropathology.

    May there be reason to expect that the increased BACE2 may actually be beneficial?


    . Green tea catechins as a BACE1 (beta-secretase) inhibitor. Bioorg Med Chem Lett. 2003 Nov 17;13(22):3905-8. PubMed.

    . The specificities of protein kinase inhibitors: an update. Biochem J. 2003 Apr 1;371(Pt 1):199-204. PubMed.

    . Antagonistic effects of beta-site amyloid precursor protein-cleaving enzymes 1 and 2 on beta-amyloid peptide production in cells. J Biol Chem. 2003 Aug 22;278(34):31512-20. PubMed.

    . Elevated expression of beta-site amyloid precursor protein cleaving enzyme 2 in brains of patients with Down syndrome. Neurosci Lett. 2002 Jun 21;326(1):64-6. PubMed.

    View all comments by Mary Reid

Make a Comment

To make a comment you must login or register.

This paper appears in the following:


  1. Target BACE: Better Than Ever?
  2. 2008 MetLife Award for De Strooper, Vassar, Wong