Two new studies appear to lend strong support for the theory that an overabundance of the amyloid-beta protein is the root cause of Alzheimer’s disease.

A genetic study of 1,800 residents of Iceland discovered a rare mutation that apparently protects against the disease in those who carry it, and even staves off the disease in those genetically programmed to develop it – probably by blocking the formation of amyloid-beta.

A second study, published July 11 in the New England Journal of Medicine, found that a processed fragment of amyloid-beta, called amyloid-beta-42 (Abeta42), begins to accumulate in the brain 25 years before symptoms appear in people with a mutation in one of three genes that is known to cause early-onset Alzheimer’s disease, suggesting that amyloid deposition initiates the cascade of physiologic and cognitive changes that culminate in full-blown AD.

“Our data suggest that amyloid deposition will develop and be detectable in all persons with a mutation while they are still asymptomatic,” wrote Dr. Randall J. Bateman, professor of neurology at Washington University, St. Louis, and his colleagues. “These findings suggest that targeting amyloid-beta earlier in the course of the disease may provide better clinical outcomes than the treatment of mild to moderate dementia after substantial neuronal and synaptic loss has occurred” (N. Engl. J. Med. 2012 July 11 [doi: 10.1056/NEJMoa1202753]).

Dr. Kari Stefansson, chief executive officer of deCODE Genetics in Reykjavik, led the Icelandic study. He and his colleagues sought to discover any unusual genetic risk factors for AD in a group of about 1,800 subjects. The researchers focused on variants of the amyloid precursor protein gene (APP), because different enzymes can either split this protein into a harmless fragment or change it into Abeta42 – the form that aggregates into AD brain plaques.

In the initial group, Dr. Stefansson and his colleagues looked for every APP variant that occurred in more than one person, and then imputed these into a database of 72,000 Icelanders who had already been genotyped. A computer program estimated how many times each allele might occur in almost 300,000 relatives of these subjects (Nature 2012 July 11 [doi: 10.1038/nature11283]).

The researchers found a mutation (A673T) in APP that was significantly more common in those without AD than in those with it. The variant was rare, occurring in 0.14% of those with Alzheimer’s, 0.45% of overall population-based controls, 0.62% of controls aged 85 years or older, and 0.79% of controls older than 85 who were still cognitively intact. Nevertheless, the authors concluded, the presence of the variant exerted a protective effect against the disease.

This was even more apparent in risk analyses. Compared with the group of people with AD, the gene was four times more likely to appear in the general population, five times more likely to occur in those 85 years or older, and nearly eight times more common among a group of cognitively healthy subjects who were 85 or older.

The allele also appeared to protect against expected age-related cognitive decline in elderly people who did not have AD, the authors noted. In a comparison of 41 allele carriers and 3,700 noncarriers, all of whom were 80-100 years old, those with the protective allele had significantly better cognitive ability. The relationship remained strong even after removal of patients with AD from the noncarrier group.

“Removing known Alzheimer’s disease cases [from this analysis] suggests that the protective effect of A673T extends beyond the boundaries of the Alzheimer’s disease phenotype,” the investigators wrote.

The mutation’s benefit probably arises from its influence over APP fragmentation, the researchers said. Three enzymes – alpha-, beta-, and gamma-secretase – can cleave the APP protein. If alpha-secretase splits the protein, no further fragmentation occurs. When gamma-secretase splits APP, it sets the stage for beta-secretase (BACE1) to finish the job, producing Abeta42, the form most likely to aggregate into brain plaques.

In vitro experiments indicated that the allele blocked the effect of BACE1, reducing the production of Abeta42 by up to 50%. This suggests that a well-tolerated BACE1 inhibitor might have a place in any future Alzheimer’s treatment regimen.

The New England Journal of Medicine study examined Abeta42 levels and cognitive function in 128 subjects enrolled in the Dominantly Inherited Alzheimer Network (DIAN) study. The 6-year, cross-sectional, longitudinal study is examining members of families who carry a mutation in one of the three genes that affect amyloid processing, including APP. All subjects have at least one parent with a mutation known to cause early-onset AD.

The study group’s mean age at enrollment was 39 years. About half of the carriers are symptomatic, compared with just 2% of the noncarriers. Parents of these subjects developed AD at a mean age of 46 years.

Because the likelihood of developing the disease is so high in carriers, their cognitive and brain health can serve as a natural history of the development of AD. By assuming that each person will develop symptoms at or about the same age as did the affected parent, researchers can plot changes that might occur in the years – or even decades – before the disease manifests.

Dr. Randall J. Bateman

For this analysis, Dr. Bateman and his colleagues examined cerebrospinal fluid (CSF), brain imaging, and cognitive data from 128 participants who had been tracked from the 2009 baseline enrollment through 2010.

There were significant differences between the carriers and noncarriers in every measure investigated, with signs developing up to 25 years before the expected onset of disease.

In cognitive measures, carriers scored significantly lower on the Mini-Mental State Exam starting at 5 years before expected symptom onset, and in delayed recall at 10 years. Noncarriers’ scores remained normal anywhere from 30 years before to 20 years after expected onset. Carriers showed increased atrophy of the bilateral hippocampi, occurring about 15 years before expected onset.

But the differences seen in amyloid deposition and CSF biomarkers most strongly support the amyloid hypotheses. There was no detectable Abeta42 accumulation in the brains of any of the noncarriers. Those who carried the mutation had significant amyloid deposition in the precuneus up to 15 years before the expected onset of AD symptoms. The amount of amyloid increased as subjects approached that time.

Similarly, CSF levels of Abeta42 in carriers decreased as the time of symptom onset approached, indicating that more of the protein was aggregating in the brain. The change was apparent by 10 years before symptom onset. This was accompanied by an increase in CSF tau, a protein that helps stabilize neurons. Its presence in CSF indicates neuronal dysfunction.

Based on these findings, the authors were able to construct a time line of changes that precede the expected time of symptom onset in people with these genetic variants.

CSF biomarkers are the first to change; 25 years before symptoms appear, Abeta42 begins to accumulate in the brain. By at least 15 years before symptoms, imaging will show Abeta42 plaques, and neurons begin to destabilize, increasing CSF tau. Cerebral hypometabolism appears about 10 years before the expected onset of symptoms.

“After these biologic changes, cognitive impairment can be detected, which culminates in clinical impairment and eventually dementia,” the investigators wrote. “These findings suggest that the diagnosis of clinical dementia is made late in the course of the biologic cascade of autosomal dominant Alzheimer’s disease.”

The results support the hypothesis that inherited AD and the more common sporadic AD may share a common pathophysiological cascade. If this is true, they noted, then the new ability to detect amyloid deposition with brain imaging agents could be much more useful than originally envisioned.

“If autosomal dominant Alzheimer’s disease is similar to late-onset Alzheimer’s disease, this finding suggests that Alzheimer’s dementia will eventually develop in persons with positive scans for amyloid deposition,” Dr. Bateman and his associates wrote.

The DIAN study is being funded by the National Institute on Aging. Dr. Bateman and his colleagues disclosed numerous relationships with drug companies conducting Alzheimer’s therapy research. The Icelandic study was sponsored by deCODE Genetics, which employs some of the authors of the study.

A New Era for Alzheimer’s?

Dr. Richard J. Caselli

When asked to comment on the study results, Dr. Richard J. Caselli, professor of neurology at the Mayo Clinic in Scottsdale, Ariz., said, “Dr. Stefansson and his colleagues show convincingly that individuals with the A673T mutation have a greatly reduced rate of Alzheimer’s development as well as – to a somewhat less-robust degree – slowed cognitive aging even in the absence of Alzheimer’s disease.

“The investigators show the mechanism to be reduced production of amyloidogenic peptides by beta-secretase 1 (BACE1) activity, compared with the wild-type version of the gene. In contrast, the pathogenic A673V variant dramatically increases the amyloidogenic BACE1 products. The authors have essentially discovered nature’s cure for AD, and we can only hope that its pharmacological replication will soon follow,” said Dr. Caselli, who is also the clinical core director of the Arizona Alzheimer’s Disease Center.

“The separate report by Dr. Bateman and his colleagues from the Dominantly Inherited Alzheimer Network is another remarkable work,” Dr. Caselli continued, “providing unprecedented insight into the sequence of events in the same cohort with regard to biomarkers, imaging, and cognition in advance of symptoms, and using all the tools now in our arsenal: brain imaging, cerebrospinal fluid biomarkers, brain metabolism, and cognitive measures.

“More than 2 decades before any noticeable symptoms, pathophysiological changes are at work in patients who are destined to develop AD. This information may now provide us objective, identifiable outcomes with which to assess new therapies,” said Dr. Caselli.

“Based upon the findings of Dr. Stefansson and his colleagues, it’s easy to imagine how a new BACE1 inhibitor could be applied to a preclinical Alzheimer’s population, with outcomes measured by the targets identified by Dr. Bateman and his coauthors – perhaps catapulting us into a new age of prevention therapy for AD,” he noted.