Subscribe to eTOC

Evidence Points to Amyloid Beta as the Key to Alzheimer Disease


Dr. Dennis Selkoe provides an overview of research progress in understanding the role of amyloid beta in the pathophysiology of Alzheimer disease.

TORONTO—The amyloid beta (Abeta) peptide lies at the heart of amyloid plaques in Alzheimer disease (AD) and is central to disease pathogenesis, according to Dennis Selkoe, MD, professor of neurological sciences at Harvard Medical School in Boston.

In the George C. Cotzias Lecture here at the AAN annual meeting, Dr. Selkoe, one of the original architects of the Abeta hypothesis of AD, offered an update about what is known about the molecular mechanisms through which Abeta impairs memory.

“The hypothesis is that Alzheimer disease is a syndrome caused by imbalance of production and clearance of the Abeta protein, which leads to its gradual accumulation in brain circuits involved in memory,” Dr. Selkoe said.

By the end of the decade, he said, Alzheimer disease research will be focused sharply on Abeta, assessing the risk for its accumulation, measuring its levels in the blood and CNS, and beginning early treatment to prevent its build-up and enhance its clearance. Such strategies will become central to the prevention and treatment of AD, he predicted.

The Abeta peptide is formed by cleavage off of amyloid precursor protein (APP), catalyzed by a protease, presenilin, he explained. “The strongest evidence for the hypothesis comes from genetics. All of the mutations currently known that cause autosomal dominant Alzheimer disease occur either in the substrate APP, or the protease, presenilin, that produces Abeta,” he said.

Presenilin actually has a much more important substrate than APP, Dr. Selkoe noted, since it also cleaves the protein notch, central to a developmental signaling pathway. While animals can live without APP, they can't live without notch. The long-term accumulation of Abeta, he suggested, may simply be the byproduct of a secondary function of an enzyme whose central importance lies elsewhere. “This may help explain how Alzheimer disease arose,” he said.

The three dominant mutations — one in APP, two in different forms of presenilin — all increase the production of Abeta. But the most common risk factor for the disease, the e4 allele of the apolipoprotein (APOE) gene, does not, at least not directly. Instead, it increases the density of Abeta plaques and vascular deposits.

“We know there is an increased level of Abeta as a result of inheritance of APOE4, but we don't know the precise mechanism. This is one of the most important unsolved issues in the field,” Dr. Selkoe said, “in terms of exactly how APOE4 inheritance changes the molecular milieu in the brain, and leads to Alzheimer disease.” And many patients don't possess even this risk factor. “We don't know why most of our patients have elevated Abeta,” he added.

How does an increase in Abeta cause disease? Abeta monomers can link to form soluble dimers or trimers (collectively called oligomers), which may coalesce to form plaques, the pathological hallmark of the disease.

While plaques were originally thought to be the problem, they are increasingly seen as a partial and ultimately unsuccessful mitigation strategy by the neuron attempting to cope with toxic oligomers.

“Soluble Abeta is detected selectively in the cortices of patients with clinically and pathologically typical Alzheimer disease,” he said. When Abeta isolated from brains of patients with Alzheimer disease or other forms of dementia is inoculated into animal brain, it is the more soluble oligomers that appear to cause the most Alzheimer-like degeneration. “Soluble Abeta seems to correlate better with the disease we call Alzheimer's.”

The same conclusion can be drawn from experiments in mouse brain, which have begun to reveal the direct effects of Abeta on memory. Applying Abeta to hippocampal slices from healthy mice leads to impairment of long-term potentiation (LTP), the molecular mechanism underlying synaptic plasticity and, ultimately, learning and memory. The same soluble Abeta oligomers facilitate long-term depression, reducing basal synaptic transmission.

In live rats, injection of soluble oligomers through a cannula into the brain impairs their ability to learn in a standard avoidance task. “It's the soluble species we believe are generally causing functional problems.” Exactly how they do so is still unclear. At the membrane, they bind to a still unidentified “molecule X,” and secondarily perturb a variety of receptors, including those for AMPA (alpha-amino-3-hydroxyl-5-methyl-4-isoxazole-propionate), NMDA (N-methyl-D-aspartic acid), and insulin.


DR. DENNIS SELKOE predicted that by 2020, “or a little later, Alzheimerologists” will assess risk in individuals starting at age 50, including cognitive and gene screening. “Thats essentially possible now,” he said.

They also ultimately lead to disruption of the microtubule cytoskeleton and formation of neurofibrillary tangles, the second pathologic hallmark of the disease.


There are several therapeutic approaches that derive from this growing understanding of Abeta, including inhibiting proteases, preventing oligomer formation, inhibiting secondary neuroinflammation, and interfering with the toxic response in neurons, “but these may be too little too late,” Dr. Selkoe said.

Instead, the goal of therapy should be reduce Abeta directly, he said. One approach is by vaccination, a strategy that worked in mice, and looked initially promising in human trials, with clearance of accumulated protein in patient brains. However, treatment led to a T-cell mediated autoimmune response to Abeta in 6 percent of patients, forcing a halt to the trial.

An alternative approach is passive Abeta immunodepletion, and this strategy is currently being tested in humans with a monoclonal antibody, bapineuzumab. Phase II results were announced two years ago, but the outcomes were “very mixed, not convincing by any means,” he said. The intent-to-treat analysis showed no significant improvement, but a completer's analysis — of all patients who received all scheduled doses—indicated that treatment could slow worsening of cognitive abilities. “But these results are not enough to convince any of us that immunotherapy has had an effect, and we will need to wait for the results from the phase III trial,” to see if the improvements are more robust, Dr. Selkoe said.

Despite the equivocal results from these trials, Dr. Selkoe is bullish on the future of AD treatment focused on Abeta. By 2020, “or a little later,” he predicted, “Alzheimerologists” will assess risk in individuals starting at age 50, including cognitive and gene screening. “That's essentially possible now.” For those at highest risk, neuroimaging would be used to scan for Abeta accumulation, accompanied by serum or CSF testing of circulating Abeta. The outcome would be a numerical Alzheimer's risk score. “That's not so far in the future.”

Based on the risk score, enzyme inhibitors or aggregation blockers could be prescribed “years before symptom onset,” or for those with higher Abeta burden, antibody or vaccine therapy would be instituted. “This is where I believe this field is going,” he concluded.

The ultimate test of the Abeta hypothesis will come from clinical trials, commented Stefan Pulst, MD, professor of neurology at the University of Utah. “I think the proof will be if getting rid of Abeta will actually halt or reverse decline.” There is also the question of how important tangles are, he said. “In Parkinson disease, we know that both alpha-synuclein and tau are important — it's a mixed picture.” The same could be true in Alzheimer disease, he said, adding that solving the Abeta problem would be only half the battle. •