Loss and gain of motor protein function cause microtubule bundle damage in Drosophila axons

Neurodegeneration often starts by atrophy of the cable-like nerve fibers (axons) that wire nervous systems. Maintaining axons requires supply via motor-protein-driven transport along uninterrupted bundles of microtubules. Functional loss of motor proteins, but surprisingly also their hyperactivation, links to conditions of axonal atrophy; in both cases the underlying mechanisms are little understood. To bridge this important knowledge gap, we carried out systematic studies using 40 different genetic tools to manipulate 19 context-related genes in one standardized Drosophila primary neuron system. Starting with transport motors, we found that downregulation in at least three of them-dynein heavy chain, the kinesin family member 5 (KIF5) ortholog kinesin heavy chain (Khc), and KIF1A ortholog Unc-104-caused disintegration of axonal microtubule bundles, which we refer to as "microtubule-curling"; this damages the essential highways for life-sustaining axonal transport. To understand this phenomenon, we focused on Khc's various subfunctions. We found that abolishing Khc-mediated mitochondrial and lysosomal transport affects the homeostasis of reactive oxygen species (ROS), which in turn triggers microtubule-curling in fly and mouse neurons alike. Taking the opposite approach by using conditions where Khc is hyperactive, we observed comparable microtubule-curling, triggered by an ROS-independent mechanism likely involving excessive mechanical force generation. To assess wider relevance of our findings, we studied Unc-104, its binding partner KIF-binding protein (KIFBP), and human KIF5A. These studies suggest that functional loss and hyperactivation of other transport motors also cause ROS-dependent and -independent microtubule-curling, which could therefore represent two fundamental pathways that link transport motors to microtubule bundle decay and neurodegeneration.