Neurodegeneration caused by LRRK2-G2019S requires Rab10 in select dopaminergic neurons

Inherited mutations in the LRRK2 protein are the commonest known cause of Parkinsons, but the molecular link from increased kinase activity to pathological neurodegeneration remains to be determined. In vitro (biochemical and cell culture) assays led to the hypothesis that several Rab GTPases might be LRRK2 substrates. Here we show that Rab10 potently modifies LRRK2-G2019S mediated electrophysiological responses in an in vivo screen, in which each Rab was overexpressed in Drosophila dopaminergic neurons. We therefore tested the effect of Rab10 loss of function on three LRRK2-G2019S phenotypes (vision, movement and sleep) that rely on dopaminergic circuits in both flies and mammals. The knock-out of Rab10 in vivo fully rescues the reduced responses induced by dopaminergic LRRK2-G2019S in visual and motor (reaching, proboscis extension) assays, but the sleep phenotype is unaffected. We show that Rab10 is expressed in dopaminergic (tyrosine hydroxylase positive) neurons controlling vision and proboscis movement, but undetectable in those controlling sleep, indicating that anatomical and physiological patterns of Rab10 are related. Our results support the idea that LRRK2 phosphorylates separate targets in distinct neurons and confirm that one degenerative pathway starts with Rab10. Although Rab3 is another putative substrate of LRRK2, it shows no synergy with G2019S and localises to a different subset of dopaminergic neurons from Rab10. We propose that variations in Rab expression may contribute to differences in the rate of neurodegeneration seen in different dopaminergic nuclei in Parkinsons.\n\nSignificance StatementA key question in Parkinsons is why dopamine neurons die particularly fast in some parts of the substantia nigra. We focused on the commonest Parkinsons-related mutation, LRRK2-G2019S. In vitro assays suggested that neurodegeneration may start by LRRK2-G2019S increasing phosphorylation of Rab10. We found Rab10 in fly dopamine neurons in visual and motor pathways, but not in the sleep system. Rab10 knock-out rescues G2019S-induced visual and movement degeneration, leaving sleep dysfunction unaffected. Thus, LRRK2 activates at least two pathways, one Rab10-dependent, leading to neurodegeneration in vivo. Rab3 is found in a different subset of dopaminergic neurons and shows no synergy with LRRK2-G2019S. We propose that variations in Rab expression contribute to differences in neurodegeneration seen in Parkinsons.