Talan, Jamie
Back to topAgene associated with diabetes has been linked to Alzheimer disease (AD), a discovery that comes on the heels of a growing body of research in humans and in laboratory models showing a relationship between the two diseases. This association could lead to new avenues for treating both conditions.
This latest study, led by scientists at Mount Sinai School of Medicine and published in the Sept. 29 Journal of Neuroscience, attempts to follow the biochemical path from the diabetes gene to the protein pathology that plays a major role in damaging brain cells in AD. The thought is that in some patients diabetes could pave the way to AD.
“Alzheimer's and type 2 diabetes are complex diseases and the risks of both include high cholesterol, obesity, and vasculopathy,” said Sam Gandy, MD, PhD, Mount Sinai Professor of Alzheimer's Disease Research, Professor of Neurology and Psychiatry and associate director of the Mount Sinai Alzheimer's Disease Research Center, who led the study. “These factors are highly interrelated and the challenge has been finding a clear starting point to understand the relationship between the two diseases. This finding could open up new doors for more effective treatments for both diseases.”
The common disease gene is SorCS1 and it belongs to the sortilin family of vacuolar protein sorting-10 (Vps10) domain-containing proteins. Other laboratories have shown that another member of this genetic family, SorL1, is decreased in the brains of AD patients. A number of problems have been identified in AD involving intracellular protein trafficking including vacuolar protein shuttling and a complex known as the retromer, which is involved in protein movement between endosomes and the trans-Golgi network. Deficiencies in the retromer-sorting pathway have been linked to late-onset AD.
In a 2008 paper in the Proceedings of the National Academy of Sciences, investigators reported hippocampal-dependent memory and synaptic dysfunction, which was associated with elevated amyloid beta (Abeta) in retromer-deficient mouse models. In addition, in retromer-deficient flies expressing human wild-type amyloid precursor protein (APP) and human beta-site APP-cleaving enzyme (BACE), they observed neuronal loss and human Abeta aggregates.
The investigators began to see a relationship between these abnormalities in protein trafficking and the build-up of Abeta and APP, which lead to the death of neurons and the range of symptoms that develop in AD.
SorCS1 had also been linked to type 2 diabetes, and its role in cell trafficking prompted the Mount Sinai group, in collaboration with other centers, to look at its potential involvement in AD.
In the latest study, they reported that high levels of SorCS1 decrease Abeta production. Following the path in animal models, they found that SorCS1 bound to the vacuolar sorting protein (Vps35) and retromer, which in turn led to an accumulation of Abeta. Vps35 has also been shown to be low in the AD brain. Another protein from the Vps family is SorL1, which is known to modulate APP metabolism and has been genetically linked to AD.
Dr. Gandy and his colleagues conducted experiments to determine what exactly was going on that linked the two seemingly unrelated conditions. First, they overexpressed SorCS1 in cell cultures to see how it behaves in the presence of amyloid-peptide and the APP. There was a marked reduction in the cells' ability to generate Abeta.
“Our research has revealed a molecular mechanism in the gene SorCS1 that is shared by Alzheimer's and type 2 diabetes,” said Dr. Gandy. “Maybe diabetes and Alzheimer disease share a common genetic link that controls both insulin sensitivity and amyloid-beta levels?”
The team then created knockdown mice with reduced expression of SorCS1, which allowed them to test their hypothesis that SorSC1 deficiency could be a risk factor for AD. When they looked for changes in amyloid metabolism in the knockdown mice, they found that the animals had elevated Abeta in their brains. This fit perfectly with their data showing that when they over-expressed SorCS1 in cells in culture, the amyloid beta levels went down.
The finding was intriguing but the question immediately rose: “How does SorCS1 control amyloid-beta generation?” They knew from other recent findings that SorL1 had to bind Vps35 in order to exert control over Abeta. Dr. Gandy and Rachel Lane were already collaborating with James J. Lah, MD, professor of neurology at Emory University, who discovered the SorL1 deficiency in AD brain, and with Scott A. Small, MD, the Herbert Irving Professor in Neurology in the Division of Aging and Dementia in the Taub Institute at Columbia University Medical Center, who discovered the Vps35 deficiency in the AD brain.
When the team measured SorCS1 levels in the brains of the knockdown SorCS1 mice, they found that total Vps35 protein levels were decreased by 49 percent (p<0.009) and total SorL1 protein levels were decreased by 29 percent (p<0.003) in the brains of the female mice. They said that this is strong evidence that a dysfunction of SorCS1 may contribute to both the APP and Abeta disturbances that underlie AD and to the insulin-glucose disturbance that is at the root of diabetes.
Moreover, they said, the SorL1 and SorCS1 pathways probably converge on Vps35, placing this once obscure protein front and center in the pathogenesis of sporadic AD. Dr. Small first hit on Vps35 about seven years ago when he discovered using a highly novel neuroimaging technique that Vps35 is deficient in hippocampal regions that are the most vulnerable in AD. Vps35 is part of the retromer complex; these proteins are critical in intracellular recycling, Dr. Small noted.
“The story is coming together nicely,” said Dr. Gandy. “We have evidence that SorL1 and SorSC1 control the retromer, which is probably what controls amyloid beta.” The team is now crossing SorSC1 to APP transgenic animals to see if that makes Abeta levels even higher so that plaques and oligomers might form. Dr. Gandy thinks that SorSC1, SorL1, and Vps35 could form a “master coordinated regulatory circuit that could be a potential therapeutic target for both AD and diabetes.”
“This gives us a clue for how the trafficking of APP through neurons affects amyloid beta,” added Rudolph Tanzi, PhD, professor of neurology at Harvard Medical School and a co-author on the study.
The association between AD and diabetes is not new. But Stephen Salloway, MD, a professor of neurology at Brown Medical School and an AD investigator, told Neurology Today that the latest genetic link between these two conditions “is interesting and might explain how the two conditions are related.” He added that it is too early to tell how important the new finding will be but “it certainly gives us a bridge between the two conditions and it also may provide a new treatment target.”
“I wouldn't over-read it,” he said. “It's nice when people find new clues but the next step is to develop the finding so that it becomes more clinically meaningful.”
While insulin's biggest role is in the periphery, researchers have shown that there are insulin receptors in the brain. Insulin is used to help glucose get into the cells and is known to play a role in learning and memory. Dr. Salloway said that the only other place where amyloid gets deposited is in the pancreas, yet another key that the two conditions may share similar pathological processes.
“This interesting finding …may be able to help us understand what connections there are between the two conditions,” said Suzanne Craft, PhD, professor of psychiatry and behavioral sciences at the University of Washington School of Medicine. “The gene association was only seen in female animals and we are seeing similar genetic effects in our human studies. For women, diabetes may be a strong risk factor for AD. We don't understand how it works but the documentation of it does make it an important step.”
There is a lot more work needed to understand how the gene sets the stage for AD or diabetes and whether the whole story is pointing to APP or other proteins are involved. Said Allen D. Roses, MD, Jefferson-Pilot Professor of Neurobiology and Neurology and director of the Deane Drug Discovery Institute at Duke University, “I think the phenomenology is correct (in this paper) but I have the long term viewpoint that the interpretation on a causative role due to its effect of amyloid is a measure of result not cause.”
Lane RF, Raines SM, Gandy S, et al. Diabetes-associated regulates Alzheimer's amyloid-beta metabolism: Evidence for involvement of SorL1 and the retromer complex.
Muhammad A, Flores I, Small SA, et al. Retromer deficiency observed in Alzheimer's disease causes hippocampal dysfunction, neurodegeneration, and Aβ accumulation. USA 2008; 105(20): 7327–7332.