Delcroix’ poster is the latest chapter in a continuing story Mobley’s group has been developing over the past few years in an effort to explain the role of neurotrophic factors in neurodegeneration. The story begins with the realization that almost all Down patients eventually develop Alzheimer pathology, with the exception of a few who happen to miss those bits of their triplicated chromosome 21 that harbor the APP gene. People with Down's inherit three copies of chromosome 21, which carries the gene for APP, among many others. Indeed, this developmental disease has played a historic role in helping AD researchers nail APP as the first known AD gene, and it is considered a natural human model of AD. The physiological function of APP, however, remains enigmatic today.

Mobley’s group employed the Ts65DN mouse, a model for Down Syndrome that is trisomic for a segment of mouse chromosome 16 that houses 140 genes, including APP, and that is quite homologous to the corresponding stretch of human chromosome 21. Like aging people with Down's, and like AD patients, Ts65DN mice lose cholinergic neurons that project into the hippocampus as they age. Earlier, the researchers reported that these neurons wither because they are unable to transport the life-sustaining protein nerve growth factor back to the cell body after their nerve terminals have picked it up in the hippocampus (Cooper et al, 2001).

To find out which genes are behind this transport block, Delcroix worked with collaborators at Stanford and elsewhere to examine retrograde NGF transport in other mouse models. First, the researchers narrowed down the number of candidates with a second mouse strain that carries a smaller chromosome 16 triplication of 93 genes excluding APP. These mice have neither significant axonal transport defects, nor do their cholinergic neurons degenerate. The researchers confirmed with Q-PCR from laser-captured bits of hippocampal tissue that the first strain overexpresses APP but the second does not. The clincher came when the scientists bred the Ts65DN mice to heterozygous APP-deficient mice. Some of the offspring were triploid for all the genes of the chromosome 16 segment, but were diploid for APP. This restored axonal transport and prevented neurodegeneration.

Moreover, Delcroix and colleagues asked directly whether APP overexpression was sufficient to stall NGF transport. They assessed retrograde NGF transport in mice that express low levels of full-length human APP from a yeast artificial chromosome (Lamb et al., 1999), and found that the mice indeed had a transport reduction, though not enough to cause neurodegeneration.

So far the story is clear-cut, but it quickly gets murky, said Delcroix. Which form of APP impairs NGF transport, then? And does APP processing have something to do with it? It turns out that NGF transport in various AD mouse models varies in ways pointing to the β-secretase cleavage product C99. For one thing, Aβ levels did not influence NGF transport. For another, mice overexpressing the Swedish mutation of APP, as well as mice overexpressing APP and PS1, had decreased NGF transport, but mice overexpressing only PS1 instead saw their NGF transport increase. This could perhaps reflect the ready degradation of C99 by overly abundant PS1, removing it before it can make a mess of axonal transport, Delcroix speculated. This question needs to be sorted out.

Even so, the data so far suggest that APP somehow interacts with the retrograde axonal transport mechanism in the neurons that are among the earliest to degenerate in Alzheimer disease. Numerous other studies also point to derailed axonal transport as an early step in the pathogenesis of neurodegenerative diseases (see, for example, ARF related news story). APP is thought to function in axonal transport, though it is linked more strongly to anterograde transport (see ARF related news story). Finally, it’s worth noting that this research represents a case in point for a broader hypothesis about endosome signaling. It holds that transport on endosomes—in this case of NGF, its receptor TrkA, and related signaling proteins—subserves important signaling function, and that problems with this pathway are among the earliest signs of a neuron in distress (see Delcroix et al., 2004).—Gabrielle Strobel.