Macroautophagy (autophagy) is a key catabolic pathway for the maintenance of proteostasis through constant digestion of selective cargoes. The selectivity of autophagy is mediated by autophagy receptors that recognize and recruit cargoes to autophagosomes. SQSTM1/p62 is a prototype autophagy receptor, which is commonly found in protein aggregates associated with major neurodegenerative diseases. While accumulation of SQSTM1 implicates a disturbance of selective autophagy pathway, the pathogenic mechanism that contributes to impaired autophagy degradation remains poorly characterized. Herein we show that amyotrophic lateral sclerosis (ALS) and frontotemporal lobar degeneration (FTLD)-linked mutations of TBK1 and SQSTM1 disrupt selective autophagy and cause neurotoxicity. Our data demonstrates that proteotoxic stress activates serine/threonine kinase TBK1, which coordinates with autophagy kinase ULK1 to promote concerted phosphorylation of autophagy receptor SQSTM1 at the UBA domain and activation of selective autophagy. In contrast, ALS-FTLD-linked mutations of TBK1 or SQSTM1 reduce SQSTM1 phosphorylation and compromise ubiquitinated cargo binding and clearance. Moreover, disease mutation SQSTM1 G427R abolishes phosphorylation of Ser351 and impairs KEAP1-SQSTM1 interaction, thus diminishing NFE2L2/Nrf2-targeted gene expression and increasing TARDBP/TDP-43 associated stress granule formation under oxidative stress. Furthermore, expression of SQSTM1 G427R in neurons impairs dendrite morphology and KEAP1-NFE2L2 signaling. Therefore, our results reveal a mechanism whereby pathogenic SQSTM1 mutants inhibit selective autophagy and disrupt NFE2L2 anti-oxidative stress response underlying the neurotoxicity in ALS-FTLD.

Friedreich's ataxia (FRDA) is an autosomal recessive disease caused by mutations that produce a deficiency in frataxin. Despite the importance of neurodegeneration in FRDA, little is known about the consequences of frataxin deficiency in neuronal cells. Here we describe a neuronal cell model for FRDA based on the use of lentiviral vectors that carry minigenes encoding frataxin-specific shRNAs that silence the expression of this gene. These lentivectors can knockdown frataxin expression in human neuroblastoma SH-SY5Y cells, which results in large-scale cell death in differentiated neuron-like cells but not in undifferentiated neuroblastoma cells. Frataxin-deficient neuron-like cells appear to die through apoptosis that is accompanied by up-regulation of p53, PUMA and Bax and activation of caspase-3. No significant autophagy is observed in frataxin-deficient neuron-like cells and the pharmacological activation of autophagy does not significantly increase neuronal cell death in response to the frataxin deficiency. Cell death triggered by frataxin knockdown can be impaired by interference with p53, caspase inhibitors and gene transfer of FXN. These results suggest that frataxin gene silencing in human neuron-like cells may constitute a useful cell model to characterize the molecular changes triggered by frataxin deficiency in neurons, as well as to search for therapies that may protect against neurodegeneration.