Subscribe to eTOC

BACE1 Provides a Tantalizing Target in the Fight Against Alzheimer Disease


Investigators discussed data affecting both opportunities and challenges of using BACE1 as a target therapy for Alzheimer disease.

CHICAGO—The enzyme, beta-site APP-cleaving enzyme (BACE1) makes the first cut in amyloid precursor protein (APP), which resides on the axons, dendrites and cell body of neurons. A second cut by another enzyme, gamma secretase, produces amyloid-beta-42, the toxic peptide that somehow causes dysfunction and the eventual death of neurons in people with Alzheimer disease (AD).

Inhibiting BACE1 seems like an obvious way to alleviate AD pathology, said Robert Vassar, PhD, professor of cell and molecular biology at Northwestern University's Feinberg School of Medicine, at a symposium here at the Society for Neuroscience annual meeting.


DR. ROBERT VASSAR: “We would predict that BACE1 inhibitor drugs would reduce amyloid beta production in AD, and its encouraging that theres at least one BACE-inhibitor drug in clinical trials. But because of the side-effects seen in BACE knockout mice, its possible that careful drug titration may be required in human beings.” Courtesy: Dr. Robert Vassar

Dr. Vassar and others discussed data affecting both opportunities and challenges of using BACE1 as a target therapy for AD. He noted that inhibiting BACE1 too much may promote schizophrenia-like symptoms, reduce myelin production, and damage voltage-gated sodium channels, thereby inhibiting the generation and propagation of action potentials.

“We would predict that BACE1 inhibitor drugs would reduce amyloid beta [Abeta] production in AD, and it's encouraging that there is at least one BACE-inhibitor drug in clinical trials,” Dr. Vassar said. “But because of the side-effects seen in BACE knockout mice, it's possible that careful drug titration may be required in human beings.”


Mice genetically altered so they produce no BACE show schizophrenia-like behaviors, including impaired cognition, according to Philip C. Wong, professor in the departments of pathology and neuroscience at the Johns Hopkins University School of Medicine.

These rodent analogues of schizophrenic symptoms were first identified in BACE knockout mice by Dr. Wong and his colleague, Alena Savonenko, MD, PhD, assistant professor of pathology and neurology at Johns Hopkins University School of Medicine They reported in the April 8, 2008, issue of the Proceedings of the National Academies of Science, for example, that BACE knockout mice displayed hyperactivity in response to novel situations, reflecting the psychomotor agitation of humans with schizophrenia. The mice also showed deficits in recognizing other mice, reflecting alterations in social behavior. Moreover, some of the symptoms were relieved by the antipsychotic drug, clozapine.

In addition, electrophysiological evidence, including impaired synaptic function at the mossy fiber synapses, and morphological evidence, including reduced spine density in hippocampal neurons, supported Dr. Wong's conclusion that the BACE1 knockout mice do indeed display evidence of schizophrenia-like behavior.

Dr. Wong suspects these symptoms result from the disruption of neuregulin 1 (NRG1) signaling due to insufficient BACE1, which cleaves NRG1. NRG1 also promotes the production of myelin, the fatty white insulation that wraps around axons and facilitates signaling among neurons. This could become a problem in the aging brain or following injury.

To minimize problems associated with inhibiting BACE1 activity, Dr. Wong and his colleagues tested the effect of a moderate decrease of BACE1. When AD-model mice were genetically altered to produce only half the normal amount of BACE1, the amyloid burden was decreased in young animals but not old ones, and they showed none of the detrimental effects seen in mice that produce no BACE1.

Also, by reducing BACE1 as well as gamma-secretase, the other enzyme that cleaves APP into toxic fragments, Dr. Wong and his colleagues reduced Abeta production in Alzheimer-like mice without any observable toxic side effects. However, large reductions in gamma secretase produced serious age-associated skin tumors and shortened the lifespan of the mice.

“Even reducing 50 percent of gamma secretase could lead to serious side effects,” Dr. Wong said. “You trade Abeta deposition for tumors and splenomegaly.”

Dr. Wong suggested that a “beta and gamma secretase inhibitor cocktail” that moderately inhibited both enzymes might reduce the production of toxic Abeta without producing side effects.


BACE1 (red) accumulates in dystrophic neurites around amyloid plaques (blue) in the brains of amyloid precursor protein transgenic mice. Astrocytes (green) do not express BACE 1. Similar images are observed in the human Alzheimer disease brain and suggest a role for BACE1 in the initiation or progression of the disease.

“Moderate reductions of both BACE and gamma secretase rescue memory deficits occurring in aged mutant APP/PS1 {presenilin 1] mice,” he said. “We do not see behavioral deficits or schizophrenia-like behaviors, and no evidence of skin cancers. This suggests that combination therapy of reducing both beta and gamma secretase moderately might be an effective strategy in terms of reducing amyloidosis without major side-effects.”

Inhibiting BACE1 in mice may affect the repair of myelin that insulates axons, but this impairment is relatively mild and may affect only small numbers of elderly people who may suffer neuronal injury that can cause loss of myelin, according to Riqiang Yan, PhD, professor of neurology at the Lerner Research Institute of Cleveland Clinic.


One concern Dr. Yan expressed involved epileptic seizures. BACE1 knockout mice have more epileptic seizures and abnormally increased levels of neuronal excitability and hyperactive astrocytes — cells that form the blood-brain barrier and regulate nutrient and calcium distribution to neurons.

“Drugs used to inhibit BACE1 activity therefore need to be applied with caution because Alzheimer patients have a higher incident rate of epileptic seizures,” Dr. Yan said.

BACE1 also cuts a protein component of voltage-gated sodium channels — a change found in AD brains, according to Dora M. Kovacs, PhD, associate professor of neurology at Harvard Medical School, and associate neuroscientist at Massachusetts General Hospital. In her experimental models this results in decreased sodium channel function, which would presumably lead to disruptions in the generation and propagation of action potentials in humans.

In AD brains with high BACE1 levels, Dr. Kovacs found that the protein component of voltage-gated sodium channels was cut by BACE1. “Altered sodium channel function is connected with epilepsy,” she said, “but it may also accelerate the onset of AD.”


In a September paper published online ahead of the print edition of Neurobiology of Aging, George Bartzokis, MD, professor of psychiatry at the University of California-Los Angeles Institute for Neuroscience and Human Behavior, presented his “myelin model,” which attributes AD and other forms of neurodegeneration to the myelin repair process and its byproducts.

“The myelin model of the human brain proposes that the processes of myelin development, maintenance, and its eventual breakdown are essential to understanding our species' unique cognitive and behavioral trajectories through life,” he wrote. In the paper Dr. Bartzokis pointed out that the problems produced in mice by reduced myelination would undoubted be magnified in humans, who have proportionately more brain myelin than any other species.

“Researchers have found that when myelin was reduced in mice, the mice seemed pretty much OK,” said Dr. Bartzokis. “But these are mice. If you reduce the thickness of a mouse's myelin it probably won't make a huge difference, but for what we humans do with myelin it might be a huge problem because we are much more myelinated than mice.”

Signaling the need for myelination and its repair may be the most important function of BACE1, according to Dr. Bartzokis. “The maximal BACE1 level occurs at birth,” he said. “BACE1 declines in adults and increases again in old age as the need for myelin repair increases. BACE1 is even higher in degenerative diseases like AD, where myelin repair needs are higher than normal due to various inefficiencies in the repair process. I see the triggering of Abeta production by cutting APP as a secondary effect of BACE1 that, in the context of the myelin model, may actually have underappreciated beneficial effects. Thus, in humans, inhibiting BACE1 may produce more problems than it solves.”

Dr. Vassar expressed skepticism that the elevated BACE1 levels found in AD patients results from myelin repair. “In human AD and APP transgenic mice BACE1 accumulates in swollen axons near amyloid plaques in a pattern that may reflect an axonal transport problem induced by amyloid toxicity,” he said, “but we find little evidence that this relates to myelin repair.”

Also, the other problems observed in BACE1 knockout mice — reduced myelin, schizophrenia-like behaviors and sodium channel changes — may result only when BACE1 is inhibited during embryonic and post-natal development, “and may not occur when BACE1 is inhibited in the adult,” Dr. Vassar added.

The problems seen in BACE1 knockout mice might be “ a consequence of BACE1 inhibition during development, and may not occur when BACE1 is inhibited in the adult,” he said. “It is most important that BACE1-inhibitor drugs be pursued vigorously for this devastating neurodegenerative disease, but we need to be aware of all potential mechanism-based side effects of BACE1 inhibition in order to devise strategies to avoid them.”


• Vassar R, Kovacs DM, Wong PC. The beta-secretase enzyme BACE in health and Alzheimer's disease: regulation, cell biology, function, and therapeutic potential. J Neurosci 2009;29(14):12787–12794.
    • Savonenko AV, Melnikova T, Wong PC, et al. Alteration of BACE1-dependent NRG1/ErbB4 signaling and schizophrenia-like phenotypes in BACE1-null mice. Proc Natl Acad Sci USA 2008;105(14):5585–90. E-pub 2008 Apr 2.
      • Bartzokis G. Alzheimer's disease as homeostatic responses to age-related myelin breakdown. Neurobiol Aging 2009; E-pub 2009 Sep 21.