The paper addresses several open questions in Alzheimer's research. First, the normal function(s) of AβPP remain unknown, though knockout mice point to synaptic defects, gliosis, and neuromuscular damage (Zheng et al. 1995, Dawson et al. 1999). Second, it remains unclear where in the neuron Aβ is generated in vivo. Previous studies point to the endoplasmic reticulum, pre-Golgi and trans-Golgi compartment, or endosomes, but most were done in cell lines and few have been able to colocalize the substrate and two enzymes needed for Aβ generation.

In this study, Adeela Kamal et al. tried to make their point with several different sets of experiments. First, they compared levels of kinesin-1, AβPP, and its presumed cargo in sciatic nerves and corpus callosum from normal and AβPP knockout mice. Knockout mice had reduced levels of kinesin-1, Bace, Ps-1 in both peripheral and central nerve tracts. Also reduced were levels of the axonal protein Gap43, the synaptic marker synapsin-1, and the neurotrophin receptor TrkA, which may be additional cargo in vesicles transported via the APP-kinesin link, the authors suggest. The levels of these proteins were increased in dorsal root ganglia of AβbPP knockout mice, possibly reflecting grounded proteins that need AβPP for transport. The control proteins synaptotagmin and synaptophysin, which are transported by a different motor, were unaffected by the AβPP loss.

The authors then tied the sciatic nerve of wildtype and AβPP knockout mice and showed that kinesin-1, Bace, PS-1, synapsin-1, Gap-43, and TrkA markedly accumulated over time on the proximal side of the ligation in wildtype mice but did so less in AβPP-deficient mice. Immunofluorescence of the ligated nerves showed colocalized, overlapping staining of APP, BACE, and Ps-1.

Fractionation of sciatic nerve yielded an AβPP-, BACE-, and Ps-1-containing fraction that looked like vesicles in the electron microscope. Immunoprecipitation with APP-, kinesin-, or TrkA antibodies brought down a fraction containing kinesin-1, AβPP, BACE, and PS-1, but not synpathophysin, synaptotagmin, or ER markers. Next, the scientists fractionated mouse cortex and found two pools of kinesin-1, AβPP, BACE, and PS-1; one pool contained ER and Golgi markers but the other did not. Immunoprecipitation again brought down the same group of proteins.

Next, the authors studied whether these axonal membrane vesicles generate Ab. Using immunoblotting, they detected Aβ40 and 42 in sciatic nerve axon vesicle fractions, and found that both peptides accumulate in the proximal side of ligated sciatic nerves of control mice. The also observed an Aβ40 and 42 increase over six hours in sciatic nerve segments that were ligated in two places to exclude the possibility that Aβ have been made in the cell body and shipped down the axon. When warmed to 37 degrees, even the isolated membrane fractions generated some Aβ over six hours' time.

Finally, the authors wondered whether Aβ generation affected transport in some way. When extracting proteins from sciatic nerves double-ligated for six hours, they found increased soluble kinesin-1 and a soluble fragment that is probably the APP C-terminus, indicating that γ-secretase cleavage of AβPP in the axonal membrane vesicles releases its C-terminus and causes kinesin to detach from the vesicle.

"Our data confirm a role for AβPP as a kinesin-1 membrane receptor needed for BACE and Ps-1 transport in an axonal membrane compartment," the authors write, adding that any role for AβPP processing in normal transport remains speculative. One physiological role could be that AβPP cleavage ends kinesin-1 transport once the vesicle has arrived at the nerve terminal. An underlying disease process or axon damage could also trigger AβPP processing. The released C terminus could, in turn, transmit an injury signal to the cell body and initiate neuroprotective gene transcription, the authors speculate, (see related news item.)-Gabrielle Strobel.

Reference:Kamal A et al. Kinesin-mediated axonal transport of a membrane compartment containing beta-secretase and presenilin-1 requires APP. Nature. 2001 Dec 6;414(6864):643-8. Abstract