Now the Brady lab has provided an answer to that question. In a paper just out online in Nature Neuroscience, they report that the polyQ expanded androgen receptor (the cause of spinal bulbar muscular atrophy) inhibits fast axonal transport by activating the c-jun N-terminal kinase (JNK). JNK then phosphorylates the kinesin heavy chain subunit, inhibiting its binding to microtubules. JNK inhibitors blocked the effects of the polyQ expanded AR on axonal transport in an extruded squid axon preparation, and also reversed inhibition of neurite outgrowth by the pathogenic protein in human cells. The results identify a new pathway by which polyQ expanded proteins can cause neurodegeneration, and suggest that JNK may be a promising target for treatment of spinal and bulbar muscular atrophy (SBMA). On top of that, Brady told Alzforum that his group has a manuscript in preparation showing similar effects with the polyQ expanded Huntington disease protein, huntingtin.

The new work follows up on Gyorgyi Szebenyi’s report that both polyQ-AR and polyQ-huntingtin inhibited axonal transport in an isolated squid axoplasm preparation. First authors Gerardo Morfini and Gustavo Pigino systematically examined the basis for this inhibition. First, they showed that polyQ-AR did not bind directly to kinesin 1, as has been suggested. The pathogenic protein did not affect the membrane association of kinesin 1 in cells, but did reduce the binding of kinesin to microtubules by an indirect mechanism. SH-SY5Y cells expressing polyQ-AR had increased phosphorylation of kinesin 1 heavy chain. To identify the kinase responsible, the researchers turned to the squid axoplasm vesicle transport assay, where they added polyQ-AR along with a variety of kinase inhibitors. Inhibitors of the stress activated protein kinase (SAPK) family, which includes the p38 map kinases and JNK, was able to block the effects of polyQ-AR, while GSK3 inhibitors and the phosphatase inhibitor okadaic acid had no effect. In SH-SY5Y cells, polyQ-AR expression inhibits neurite outgrowth, and this effect was also reversed by the SAPK inhibitor.

To narrow the field down even more, the researchers showed that polyQ-AR expression induced JNK activity, but not p38 map kinase. In the squid axon, they showed that JNK inhibitors, or a JNK interacting protein (JIP) peptide fragment, could block the polyQ-AR inhibition of transport. Finally, they showed that purified JNK could phosphorylate kinesin 1 heavy chain in vitro, and addition of active JNK to the squid axoplasm recapitulated the effects of polyQ-AR.

Is the slowing of axonal transport the primary lesion in SBMA and other polyQ expansion diseases? Brady and colleagues argue that it may very well be, given the unique dependence of neurons on axonal transport. A slight decrease in transport could allow neurons to function normally for some time before eventually succumbing, they suggest, and account for the delayed onset of many neurodegenerative disorders. Also, a selective vulnerability of neurons would explain why only certain populations die while all cells in the body express the mutated protein.

The idea that dysregulation of protein phosphorylation can lead to axonal transport defects and ultimately neurodegeneration echoes Brady’s previous observations in an Alzheimer disease model. There, a presenilin 1 mutation inhibited axonal transport via activation of glycogen synthase kinase. In mice carrying the M146V FAD mutant, increased activation of GSK3β kinase resulted in phosphorylation of kinesin light chain and lowered its binding to vesicle cargoes. The mice showed signs of compromised transport in the form of lower levels of synaptic vesicles and fewer mitochondria in nerve endings (Pigino et al., 2003). Of course, in AD there is also tau, which comes with its own set of troubles for kinesin-mediated transport and microtubules.—Pat McCaffrey.

Reference:
Morfini G, Pigino G, Szebenyi G, You Y, Pollema S, Brady ST. JNK mediates pathogenic effects of polyglutamine-expanded androgen receptor on fast axonal transport. Nat Neurosci. 2006 Jul;9(7):907-16. Epub 2006 Jun 4. Abstract