CHICAGO—Cerebral blood volume mapping could bring clinical neurologists closer to the day when they can predict who will progress to late-onset Alzheimer’s disease. Investigators at Columbia University in New York City have demonstrated that knocking out a molecular trafficking complex highly expressed in the hippo­campus is associated with increased amyloid-beta, whereas overexpression of the complex leads to a decrease. Deposition of amyloid-beta is one of the central pathologic features of Alzheimer’s disease and the normal aging process in the decline of memory.

The complex, called retromer, traffics type 1 membrane proteins such as the amyloid precursor protein (APP) and the beta-site APP-cleaving enzyme (BACE) within a cell—namely, from the endosome to the trans-Golgi network. When retromer dysfunction occurs, it can result in an inability to traffic the proteins out of the endosome, causing a pathogenic buildup of this same molecular “cargo.”

The endosome is an organelle that is particularly relevant to Alz­heimer’s disease, as are such type 1 membrane proteins as APP and BACE. However, an even more important association involves the core component of retromer, a molecule called VPS35, which proved to be the best match among only five molecules found to conform to a spatial-temporal model of Alzheimer’s disease. Developed by the research group at Columbia, the evidence was derived via postmortem isolation of key subregions of the hippocampus—the entorhinal cortex and dentate gyrus—from the brains of patients with and without Alzheimer’s disease.

Earlier studies by the same group of investigators using cerebral blood volume mapping have shown significant reductions in the entorhinal cortex to be the primary defect––out of any seen among the subregions of the hippocampus––in patients with Alz­heimer’s disease compared to age-matched controls.

ISOLATING AN IMAGE

The findings on molecular mechanisms as well as those derived directly from MRI mapping are enabling researchers to “zoom in” on Alzheimer’s disease and the aging process, said Scott A. Small, MD, a principal investigator in the group at Columbia. “At this point, we can say that we really do have evidence to suggest that the retromer is related to late-onset Alzheimer’s,” he said at the 131st Annual Meeting of the American Neurological Association.

The evidence on the retromer validates, in Dr. Small’s view, many of the assumptions that went into his group’s decision to combine brain imaging with microarray technology. This effort of isolating molecules underlying various physiologic disorders of the brain is a process comparable to, as the subtitle of one of his journal articles puts it, “finding needles in molecular haystacks.”

“I’m happy to say that for the retromer, we have a confirmation at every level of analysis in our original Annals [of Neurology] paper,” he said. “We’ve shown that if you knock down retromer, you have an increase in amyloid-beta. If you overexpress retromer, you have a decrease in amyloid-beta compared to controls. Of course, we [also] want to see what’s happening in the true intact brain.”

MODEL BEHAVIOR

To achieve this last objective, Dr. Small and his colleagues have developed an experimental model of retromer dysfunction, involving mice that are heterozygote knockouts for another key component of retromer, VPS26. The mice are now being bred and phenotyped. “At the biochemical level, they have defects in VPS26 and VPS35, so biochemically phenocopy is what we’ll find in Alz­heimer’s tissue,” Dr. Small said. “They have clear hippocampal-dependent memory deficits, and they have cell sickness as measured by electrophysiology. What’s really quite appealing is that they have overproduction of murine wild-type amyloid-beta and amyloid-beta-42. Those of you in the field know that’s quite rare.” A final confirmation of the retromer comes from recent independent studies showing that polymorphisms in retromer-related molecules increase the risk for developing Alz­heimer’s disease.

The findings on the retromer generate a model suggesting that amyloid-β production can be increased in late-onset Alzheimer’s disease without molecular defects existing in the primary building blocks. “As you know, amyloid-beta is produced when APP is cleaved by BACE and by the gamma-secretase,” Dr. Small said. “That molecular pathway can account for early-onset Alzheimer’s, but it can’t account for late-onset Alzheimer’s. So what we suggest is that if you have molecular defects and sorting molecules and trafficking molecules that resulted in the increased clustering of these building blocks into the membrane compartment where they like to interact, that is a mechanism that will be sufficient to account for increased amyloid-beta.”

Near the end of his talk, Dr. Small showed images, found on the Web site of the Allen Brain Atlas, demonstrating that both APP and BACE are profusely distributed and the retromer is differentially expressed in the hippocampus, perhaps accounting for the region’s selectivity. “I’d like to back up my enthusiasm about the retromer,” he maintained. “Obviously, we think it’s interesting. We have a lot of projects in the lab currently. But obviously, Alzheimer’s disease is a complicated and complex disorder. Many molecular defects will probably contribute to late-onset Alzheimer’s. However, we do think that protein sorting is a general cell biological mechanism that might be uniquely vulnerable to Alz­heimer’s pathogenesis.”

NR