Different mechanisms link gain and loss of kinesin functions to axonal degeneration

Axons are the slender, often meter-long projections of neurons that form the biological cables wiring our bodies. Most of these delicate structures must survive for an organism’s lifetime, meaning up to a century in humans. Long-term maintenance and sustained functionality of axons requires motor protein-driven transport distributing life-sustaining materials and organelles to places of need. It seems therefore plausible that loss of motor function would cause axon degeneration; however, also gain-of-function conditions were linked to disorders including motor neuron disease or spastic paraplegia. To understand this phenomenon, we studied ∼40 genetic manipulations of motor proteins, cargo linkers and regulators of reactive oxygen species in one standardised Drosophila primary neuron system. Using axonal microtubule bundle organisation as a relevant readout reflecting the state of axon integrity, we found that losses of Dynein heavy chain, KIF1A/Unc-104 and KIF5/Kinesin heavy chain (Khc) all cause bundle disintegration in the form of chaotically curled microtubules. Detailed functional studies of Khc and its adaptor proteins revealed that losses of mitochondrial or lysosomal transport cause ROS dyshomeostasis, which is a microtubule-curl-inducing condition in fly and mouse neurons alike. We find that hyper-activated Khc induces the same microtubule curling phenotype, not through ROS but likely more directly through enhanced mechanical forces. Studies with loss of Unc-104 or KIFBP and expression of an ALS-linked mutant form of the human Khc orthologue KIF5A suggest that loss or hyperactivation of different types of transport motors cause MT curling as a shared feature. We discuss a model which can explain our findings and their relevance for understanding motor-linked neurodegeneration.