Spastic paraplegia related loss of Kinesin-1 function causes developmental defects and synapse degeneration in a Drosophila model

Abstract

Autosomal dominant mutations in the gene coding for the neuronal Kinesin-1 isoform KIF5A cause spastic paraplegia type 10 (SPG10). The mechanism behind Kinesin-1 mutation leading to slow progressing spasticity and weakness of the lower limbs of patients is not clear and there is no causative treatment. One of the point mutations in KIF5A is KhcN256S. In vitro studies show that KhcN256S acts dominant negative over wild type Khc [Ebbing et al. 2008]. Mouse and zebrafish animal models with loss of Khc function hint at predominantly neurodevelopmental defects [Hurd and Saxton 1996; Xia et al. 2003; Karle et al. 2012; Nakajima et al. 2012; Campbell et al. 2014]. In contrast human SPG10 is clinically classified as a neurodegenerative disorder. The fly model used in this study is characterized by low-level ectopic overexpression of the Drosophila melanogaster orthologue of KIF5A (kinesin heavy chain, khc) bearing the SPG10 point mutation (khcN262S). As KhcN262S acts in a dominant negative manner, overexpression of mutant Khc resembles more closely the situation in heterozygous SPG10 patients. Using this milder affected model the parallel occurrence of neurodevelopmental defects and neurodegeneration could be shown for the first time. Larvae expressing khcN262S build less new synapses in a given time window than controls. In addition to this neurodevelopmental defect, neuromuscular junctions (NMJs) of khcN262S expressing larvae show overgrowth including enhanced transport of active zone (AZ) precursor vesicles that also lead to increased Brp intensity, a marker for AZs, at the NMJ. But in parallel neurodegenerative signs such as accumulation of neuronal membranes inside boutons and disorganized microtubule (MT) cytoskeleton are detectable. In larvae expressing khcN262S, the impaired axonal transport of mitochondria leads to a reduced number of mitochondria at the NMJ. This could explain the behavioral defects, characteristic for fly models of axonal transport deficits [Hurd and Saxton 1996; Gindhart et al. 1998; Martin et al. 1999; Bowman et al. 2000]. These results show in greater detail the progression of paralysis in the posterior part of larvae because the chronology of emerging pathological characteristics of a Drosophila model of SPG10 could be described. To further study the way of degeneration in a model of loss of functional Khc, NMJs of khc-/- larvae were analyzed. They grow much slower than controls and die before pupation. NMJs are characterized by smaller size compared to controls and show frequent occurrence of accumulated neuronal membranes in NMJs as well as in axons. Cargos like Brp or synaptic vesicle marker VGlut accumulated in- and outside of axonal swellings and were diminished at the NMJ. The NMJ was especially marked by fragmentation of neuronal membranes and MTs. These did not retract from the terminal bouton as described for classical neuronal retraction [Eaton et al. 2002], but dispersed in central regions of the NMJ, separating parts of the NMJ from the innervating axon. Therefore, they show a disassembly of the NMJ, which is different from classical neuronal retraction and resembles more a Wallerian-like deconstruction of neuronmuscle innervation.