Endogenous TDP-43 localized to stress granules can subsequently form protein aggregates

Abstract

TDP-43 proteinopathies are characterized by loss of nuclear TDP-43 and accumulation of the protein in the cytosol as ubiquitinated protein aggregates. These protein aggregates may have an important role in subsequent neuronal degeneration in motor neuron disease, frontotemporal dementia and potentially other neurodegenerative diseases. Although the cellular mechanisms driving the abnormal accumulation of TDP-43 are not understood, recent studies have shown that an early change to TDP-43 metabolism in disease may be accumulation in cytosolic RNA stress granules (SGs). However, it is unclear whether the TDP-43 in these SGs progresses to become irreversible protein aggregates as observed in patients. We have shown recently that paraquat-treated cells are a useful model for examining TDP-43 SG localization. In this study, we used the paraquat model to examine if endogenous TDP-43 in SGs can progress to more stable protein aggregates. We found that after treatment of HeLa cells overnight with paraquat, TDP-43 co-localized to SGs together with the ubiquitous SG marker, human antigen R (HuR). However, after a further incubation in paraquat-free, conditioned medium for 6h, HuR-positive SGs were rarely detected yet TDP-43 positive aggregates remained present. The majority of these TDP-43 aggregates were positive for ubiquitin. Further evidence for persistence of TDP-43 aggregates was obtained by treating cultures with cycloheximide after paraquat treatment. Cycloheximide abolished nearly all cytosolic HuR aggregation (SGs) but large TDP-43-positive aggregates remained. Finally, we showed that addition of ERK and JNK inhibitors together with paraquat blocked TDP-43-positive SG formation, while treatment with inhibitors after 24h paraquat exposure failed to reverse the TDP-43 accumulation. This failure was most likely due to the addition of inhibitors after maximal activation of the kinases at 4h post-paraquat treatment. These findings provide strong evidence that once endogenous TDP-43 accumulates in SGs, it has the potential to progress to stable protein aggregates as observed in neurons in TDP-43 proteinopathies. This may provide a therapeutic opportunity to inhibit the transition of TDP-43 from SG protein to aggregate.