Mutations in GBA (glucosylceramidase beta), which encodes the lysosomal enzyme glucocerebrosidase (GCase), are the strongest genetic risk factor for the neurodegenerative disorders Parkinson's disease (PD) and Lewy body dementia. Recent work has suggested that neuroinflammation may be an important factor in the risk conferred by GBA mutations. We therefore systematically tested the contributions of immune-related genes to neuropathology in a Drosophila model of GCase deficiency. We identified target immune factors via RNA-Seq and proteomics on heads from GCase-deficient flies, which revealed both increased abundance of humoral factors and increased macrophage activation. We then manipulated the identified immune factors and measured their effect on head protein aggregates, a hallmark of neurodegenerative disease. Genetic ablation of humoral (secreted) immune factors did not suppress the development of protein aggregation. By contrast, re-expressing Gba1b in activated macrophages suppressed head protein aggregation in Gba1b mutants and rescued their lifespan and behavioral deficits. Moreover, reducing the GCase substrate glucosylceramide in activated macrophages also ameliorated Gba1b mutant phenotypes. Taken together, our findings show that glucosylceramide accumulation due to GCase deficiency leads to macrophage activation, which in turn promotes the development of neuropathology.

Mutations in the glucosylceramidase beta (GBA) gene are strongly associated with neurodegenerative diseases marked by protein aggregation. GBA encodes the lysosomal enzyme glucocerebrosidase, which breaks down glucosylceramide. A common explanation for the link between GBA mutations and protein aggregation is that lysosomal accumulation of glucosylceramide causes impaired autophagy. We tested this hypothesis directly by measuring protein turnover and abundance in Drosophila mutants with deletions in the GBA ortholog Gba1b. Proteomic analyses revealed that known autophagy substrates, which had severely impaired turnover in autophagy-deficient Atg7 mutants, showed little to no overall slowing of turnover or increase in abundance in Gba1b mutants. Likewise, Gba1b mutants did not have the marked impairment of mitochondrial protein turnover seen in mitophagy-deficient parkin mutants. Proteasome activity, microautophagy, and endocytic degradation also appeared unaffected in Gba1b mutants. However, we found striking changes in the turnover and abundance of proteins associated with extracellular vesicles (EVs), which have been proposed as vehicles for the spread of protein aggregates in neurodegenerative disease. These changes were specific to Gba1b mutants and did not represent an acceleration of normal aging. Western blotting of isolated EVs confirmed the increased abundance of EV proteins in Gba1b mutants, and nanoparticle tracking analysis revealed that Gba1b mutants had six times as many EVs as controls. Genetic perturbations of EV production in Gba1b mutants suppressed protein aggregation, demonstrating that the increase in EV abundance contributed to the accumulation of protein aggregates. Together, our findings indicate that glucocerebrosidase deficiency causes pathogenic changes in EV metabolism and may promote the spread of protein aggregates through extracellular vesicles.