Valeo, Tom
Investigators found a variant in the amyloid precursor protein was far more common in elderly people who remained cognitively intact beyond the age of 85.
About two-dozen mutations of the gene for amyloid precursor protein (APP) have been linked to the early-onset form of Alzheimer's disease (AD), but now one rare mutation in the same gene has been shown to powerfully inhibit the disease process, and perhaps typical age-related cognitive decline as well.
The research, published online in Nature on July 11, was led by Kari Stefansson, MD, president, CEO, and co-founder of deCODE Genetics, the biopharmaceutical company based in Reykjavík, Iceland. The findings, according to the authors, affirm the amyloid cascade hypothesis, which attributes AD to the accumulation of toxic peptides cleaved from APP, and also provides proof of principle for the idea that inhibiting enzymes that cleave APP could protect against AD.
They further suggest that AD may represent an advanced form of normal brain aging. In a prepared statement Dr. Stefansson said that the findings imply that “general cognitive decline and late-onset Alzheimer's disease share biological pathways.” As a result, treatments for Alzheimer's may also benefit those who do not have the disease.
In an effort to find low-frequency variants in the APP gene that might contribute to AD, the deCODE team, working with scientists from the biotech company Genentech, examined genetic information from 1,795 people in Iceland, where the company is located. They found that an alanine-to-threonine substitution at position 673 of the APP gene (A673T) was far more common in elderly people who remained cognitively intact beyond the age of 85. (The mutation is extremely rare, found in less than 1 percent of the Scandinavian population, and only in about 1 in 5,000 in other ethnic groups.)
A subsequent comparison of 41 carriers of A673T with 3,673 non-carriers — all free of AD — showed that carriers of the gene displayed consistently better scores on a cognitive test administered three times a year to residents of Icelandic nursing homes.
“The fact that the cognitive function of non-carriers remained poorer than for carriers of A673T after removing known Alzheimer's disease cases suggests that the protective effect of A673T extends beyond the boundaries of the Alzheimer's disease phenotype,” the authors wrote.
The A673T mutation decreases cleavage of APP, reducing the production of toxic peptides by 40–50 percent, which makes it more protective than the apolipoprotein 2 gene, which facilitates removal of toxic amyloid beta (Abeta) fragments from the brain.
The mutation is located just two amino acids away from the site where beta secretase (BACE1) cleaves APP. Other mutations in that region have been shown to increase the risk of Alzheimer's dramatically. A 2010 paper in Acta Neuropathologica, for example, reported that brain tissue taken from family members homozygous for a valine substitution at position 673 of the APP gene (A673V) displayed severe and extensive Abeta accumulation and tau pathology. However, because the mutation is recessive, family members who carried only one allele appeared somewhat resistant to Alzheimer's.
The finding that the A673T mutation reduces the production of toxic Abeta by inhibiting cleavage of APP by secretase enzymes gives a welcome boost to that line of research, said Robert Vassar, PhD, of Northwestern University Feinberg School of Medicine in Chicago, who was not involved with the current study.
Dr. Vassar, professor of cell and molecular biology, discovered and spent years studying the beta-secretase, or BACE1 enzyme, which cleaves APP into toxic fragments.
“I think these findings will stimulate BACE1 research and get more companies interested,” said Dr. Vassar.
One problem with BACE1 inhibition is that the enzyme is also involved in the production of myelin and other brain processes. Mice bred to produce no BACE1 have reduced myelin and display an array of mild behavioral abnormalities, such as a type of hyperactivity analogous to the psychomotor agitation of human schizophrenics, and other alterations in behavior.
However, the deCODE study shows that people with the A673T gene still produce myelin, suggesting that a moderate inhibition of the BACE1 enzyme would be enough to delay or prevent Alzheimer's pathology, according to Dr. Vassar.
“In humans, if we inhibited BACE1 by 50 percent, we'd still have 50 percent left for conducting normal function,” he said. “Maybe only a 25–50 percent reduction in BACE1 activity might be effective in protecting against Alzheimer's.”
Finding an effective BACE1 inhibitor will require a better understanding of the enzyme's targets, and two recent papers published almost simultaneously have identified cleavage targets of BACE1.
In the EMBO Journal, a German team led by Stefan F. Lichtenthaler, PhD, developed a method they call secretome protein enrichment with click sugars (SPECS), which enabled them to identify 34 novel substrates of BACE1 in neurons. Four of them — seizure-protein 6, L1, CHL1, and contactin-2 — were validated in BACE1 knockout mice and in mice treated with a BACE1 inhibitor.
“So far we know BACE1 cleaves APP, but it also cleaves other proteins,” said Dr. Lichtenthaler, senior scientist at the German Center for Neurodegenerative Diseases (DZNE), and the Technical University, Munich in Germany. “Our paper gives an idea of what kind of side-effects you can expect if you block beta-secretase with drugs. The Nature paper says, go ahead and inhibit BACE1. Our paper shows that when you develop drugs you must be aware that there are other sites cleaved by BACE1, so you must watch out for potential side-effects.”
In a paper published online June 14 in the Journal of Biological Chemistry, Bart De Strooper, MD, PhD, and colleagues in Belgium studied neural cell adhesion molecules L1 and CHL1, which are crucial for axonal guidance in the developing brain and the maintenance of neural circuits. Mutations in the L1 and CHL1 genes have been linked to mental retardation and schizophrenia.
The researchers found BACE1 cleavage sites at L1 and CHL1, and determined that BACE1 knockout mice as well as mice treated with a BACE1 inhibitor did not cleave these proteins as efficiently as control mice. Their results suggest that BACE1 cleavage of L1 and CHL1 could affect synaptic plasticity and learning.
“Our paper shows that BACE1 has several important substrates, and that clinical trials should look very carefully for side-effects,” said Dr. De Strooper, the Arthur Bax and Anna Vanluffelen chair for Alzheimer's disease research and co-director of the Leuven Institute for Neurodegenerative Diseases at the University of Leuven in Belgium. “On the other hand, it looks like beta-secretase can be targeted in a safer way than gamma secretase, which I think cuts up hydrophobic peptides in the membrane into shorter pieces until they become soluble.”
Drs. Lichtenthaler and De Strooper both consider the Stefansson paper in Nature a crucial contribution to Alzheimer's research. “It comes at a time when people are asking, is amyloid beta really so important?” Dr. Lichtenthaler said. “This is excellent proof that it is.”
Dr. De Strooper agrees that the findings will reinvigorate the search for treatments that curb the production of Abeta. “I don't think even the hard-core people believe the amyloid hypothesis is wrong, but many companies are starting to hesitate,” he said. “Here we have an experiment from Nature that supports the amyloid hypothesis. You could not ask for better evidence in my opinion.”
Thomas Bird, MD, professor of medicine, neurology, and medical genetics at the University of Washington in Seattle, called the Nature paper “fascinating.”
“I don't think it's of any immediate practical importance for clinical neurologists, but I think it emphasizes the importance of amyloid in Alzheimer's disease, and it highlights the importance of the BACE1 cleavage site as a region of interest for developing treatments or preventatives for the disease,” he said. Dr. Bird serves on the editorial advisory board of Neurology Today.
Efforts to develop secretase inhibitors that slow the production of toxic amyloid beta (Abeta) peptides have produced a long list of failures, primarily because the enzymes do much more than cleave the APP protein.
Semagacestat (LY450139), a gamma-secretase inhibitor developed by Eli Lilly and Elan, not only failed to slow progression of Alzheimer's, it also worsened measures of cognition and activities of daily living.
A BACE1 inhibitor known as CTS-21166, developed by CoMentis, Inc., went into phase I clinical trials “and we haven't heard from them since,” said Robert Vassar, PhD, associate professor of cell and molecular biology at Northwestern University Feinberg School of Medicine.
A BACE1 inhibitor from Lilly dubbed LY2811376 appeared safe and produced robust Abeta-lowering in humans, but subsequent experiments in rats revealed abnormalities in the pigment epithelium of their retinas, which brought development to a halt, even though the eye problems appeared to be unrelated to BACE1 inhibition.
“It was probably an off-target effect,” said Dr. Vassar. “They treated BACE1 knockout mice with this drug and the saw same phenotype in the retina. BACE1 inhibition is not impugned by that. I think cleaner drugs probably will work better.”
At the AAN April meeting in New Orleans, Merck presented phase I data on MK-8931, an oral BACE1 inhibitor. A randomized, double-blind, placebo-controlled single dose study involving 40 healthy adults produced reductions in Abeta of up to 92 percent in CSF, and appeared to be well tolerated.
And at the recent Alzheimer's Association International Conference in Vancouver several posters and talks described more recent progress on MK-8931 and other BACE1 inhibitors from Eisai and Lilly.
“The presentations looked encouraging,” Dr. Vassar said. “It's a really an exciting time for BACE1 inhibition.”
—Tom Valeo
• Jonsson T, Atwal JK, Stefansson K, et al. A mutation in APP protects against Alzheimer's disease and age-related cognitive decline. ; E-pub 2012 July 11.
• Giaccone G, Morbin M, Lichtenthaler SF, et al. Secretome protein enrichment identifies physiological BACE1 protease substrates in neurons. 2012;31(14):3157–3168.
• Zhou L, Barão S, De Strooper B, et al. The neural cell adhesion molecules L1 and CHL1 are cleaved by BACE1 in vivo. Jun 14.