Mitochondrial network fragmentation initiates Purkinje cells degeneration via Ca2+ buffering impairment: dissecting the molecular events from AFG3L2 mutation to defective mitochondrial dynamics.

Spinocerebellar ataxia type 28 (SCA28) is a neurodegenerative disorder characterized by unbalanced standing, gait incoordination, nystagmus, ophthalmoparesis and pyramidal signs. Several disease-causing mutations have been identified in the AFG3L2 gene. The encoded protein, AFG3L2, coassembles with paraplegin into multimeric complexes, called the m-AAA proteases, in the inner mitochondrial membrane. These complexes are crucial components of the mitochondrial protein quality control system and regulate mitochondrial morphology. The haploinsufficient Afg312 mouse recapitulates the features of SCA28 patients, displaying motor incoordination due to dark degeneration of Purkinje cells (PC-DCD). This is a form of degeneration characterized by toxic levels of intracellular Ca2+and activation of calpains. Differently from other forms of SCA, where this phenomenon is associated to excitotoxicity, in the SCA28 mouse is unique since it originates from a mitochondrial dysfunction. We established, that Afg312-depleted mitochondria ineffectively buffer the evoked peaks of Ca2+.This enhances cytoplasmic Ca2+ levels in PC, thus triggering PC-DCD. We demonstrated that this defect is caused by the negative synergism between the mitochondrial network fragmentation and the altered trafficking of the organelles to PC dendrites. To determine the molecular mechanism that induces mitochondrial network fragmentation in the absence of AFG3L2, we studied the regulatory pathway involved in mitochondrial dynamics. Previous work demonstrated that mitochondrial morphology alteration, in the absence of AFG3L2, is caused by an increased OMA1-mediated OPA1 processing. Studying the regulatory pathway that drives OMA1 activity, we discovered that AFG3L2 is deeply involved in this pathway. Full length OMA1 protein is 60 kD, while the proposed active form is 40kD. We identified AFG3L2 as the most important protease involved in OMA1 60/40 kD processing although, additional unknown protease(s) contribute to alternative pathway during stress conditions. We found that the absence of the m-AAA protease induces a striking accumulation of the 60 kD band and a reduction of the 40kD. To clarify the role of AFG3L2 in OMA1 processing we performed co-immunoprecipitation experiments, which disclosed a physical-interaction between the two proteins. Interestingly, we were able to rescue OMA1 processing in Afg3l2 ko murine embryonic fibroblasts by overexpression of wild type AFG3L2, but not with the proteolytic inactive mutant (AFG3L2-E575Q). Moreover, we showed that OMA1 40kD form is a substrate of the the i-AAA protease demonstrating that the m-AAA proteases and the i-AAA proteases work coordinately in the regulation of OMA1. These data, besides disclosing the pathogenetic mechanism of SCA28, demonstrate for the first time the impact of defective mitochondrial Ca2+uptake on local Ca2+ signaling in a physiopathological condition of the nervous system. We demonstrated that the altered Ca2+ homeostasis is caused by the mitochondrial network fragmentation, via OPA1 over processing OMA1-mediated consequent to the absence of AFG3L2. Moreover, we demonstrated for the first time the fundamental role of the m-AAA and the i-AAA complexes in the regulation of OMA1 processing and confirming their fundamental role in mitochondrial dynamics.