67-6

24/93

532Collectively, the neuronal death associated with these three neurodegenerative disorders caused by PLA2G6 mutations is referred to as PLA2G6-asso-ciated neurodegeneration. The clinical phenotype of PARK14-linked Parkinson’s disease is levodo-pa-responsive dystonia parkinsonism with cogni-tive decline and frontotemporal lobar atrophy20). Moreover, postmortem examinations have revealed a marked Lewy body pathology in PARK14-linked Parkinson’s disease29). However, the detailed func-tion of PLA2G6 in Lewy body pathology remains unclear. To elucidate roles of PLA2G6 in Parkin-son’s disease, I and my colleagues generated a PLA2G6-knockout drosophila model30). PLA2G6-knockout drosophila developed motor and sleep dysfunction with dopaminergic neurodegeneration similar to patients with Parkinson’s disease. These phenotypes and neuronal degeneration were rescued by overexpression of human wild-type PLA2G6, but not PLA2G6 harboring the pathogenic p.A80T mutation. A digenic PLA2G6-knockout and α-synuclein-overexpression drosophila model exhibited aggregation of both α-synuclein and ubiquitin. In addition, alterations of lipid composi-tion were observed in the brains of PLA2G6-knockout drosophila. In this model, the ratio of acyl chain 18:0/18:X (X = 0, 1, and 2) to acyl chains 14:0/14:0 and 14:0/16:1 was significantly increased compared with wild-type drosophila brain. Because of alterations of lipid composition, the synaptic vesi-cles of PLA2G6-knockout model drosophila were smaller than those of wild-type drosophila, resulting in synaptic dysfunction. These alterations were improved by overexpression of human wild-type PLA2G6, but not PLA2G6 harboring the p.A80T mutation. Interestingly, administration of linoleic acid could normalize the abnormal lipid composi-tion of PLA2G-knockout drosophila to ameliorate phenotypes including motor and sleep dysfunction, dopaminergic neuronal degeneration, synaptic vesicle structure, and α-synuclein aggregation. These findings suggest that a lipid diet might prevent neuronal degeneration and α-synuclein aggregation. Therefore, a balanced lipid diet might lead to disease-modifying effects on Parkinson’s disease.Prosaposin and Parkinson’s diseaseGlycolipids are mainly metabolized in lysosomes, a membranous organelle containing approximately 40 types of hydrolases. All enzymes in lysosomes are acid hydrolases, and H+-ATPases present on the lysosomal membrane use the energy of ATP hydrolysis to take in and maintain protons at an acidic pH31). Complete dysfunction of lysosomal enzymes causes lysosomal storage disorders, known as lipidoses32). However, recent investiga-tions revealed that partial dysfunction of lysosomal enzymes may be associated with Parkinson’s disease and related disorders33). Complete dysfunc-tion of GBA (associated with Gaucher disease) impairs the storage of glucosylceramides, a substrate of glucocerebrosidase. Although enzyme replace-ment therapy can rescue the systemic dysfunction of patients with Gaucher disease, it cannot amelio-rate neurodegeneration because it does not cross the blood-brain barrier34). During long-term enzyme replacement treatment of patients with non-neuro-pathic Gaucher disease, several patients developed Parkinson’s disease35). Furthermore, a genetic asso-ciation study revealed that heterozygous mutation of GBA is a risk factor for Parkinson’s disease, but not Alzheimer’s disease21). Sidranskey and colleagues performed a 16-centers analysis of GBA mutations in Parkinson’s disease, which concluded that the odds ratio for any GBA mutation in Parkinson’s disease versus controls was 5.43 across centers22).As mentioned above, previous studies revealed that not only GBA, but also other glycolipid enzymes such as arylsufatase A23), acid sphyn-gomielinase24, 36, 37), and hexosaminidase38) can cause Parkinson’s disease and/or α-synuclein aggrega-tion. Thus, disturbances of glycolipid metabolic pathways play and important role in the patho-mechanisms of Parkinson’s disease. In lysosomes, four types of sphingolipid activator proteins (saposins A, B, C, and D) are needed for glycolipid enzymes to be fully active39). Saposins are generated from prosaposin following cleavage by cathepsin D in the lysosome. Each saposin interacts with a specific enzyme: saposins A, B, and C interact with galacto-sylceramidase, arylsulfatase A, and glucocerebrosi-dase, respectively; however, the function of saposin D function (which may activate acid ceramidase) remains unknown40). The clinical phenotype of complete saposin deficiency closely resembles that of diseases associated with the dysfunction of inter-acting enzymes. For example, saposin C deficiency