TAU fragments carry an apoptosis

TAU fragments carry an apoptosis-inducing activity, which is further potentiated when tnf alpha inhibitors are exposed to TAU fragments carrying the N279K mutation (Fasulo et al., 2005). In addition, overexpression of full-length TAU carrying the N279K or the V337M mutation resulted in induction of apoptosis after serum deprivation (Furukawa et al., 2000). These findings are consistent with our data demonstrating an increased vulnerability of FTDP-17 neurons toward oxidative stress induced by the mitochondrial complex I inhibitor rotenone. This response was accompanied by an increased expression of cleaved-CASPASE-3. Rotenone-induced cytotoxicity has been observed in an iPSC-based model of PD (Reinhardt et al., 2013b), which is characterized by an increased vulnerability of midbrain DA neurons as similarly seen in FTDP-17. Increased cell death of FTDP-17 neurons could, at least in part and in addition to increased TAU fragmentation, be related to a direct impact of the N279K and V337M mutations on mitochondrial function. Such an effect was recently described for P301L mutant TAU, which when overexpressed in a neuroblastoma cell line led to reduced mitochondrial complex I activity and to an increased vulnerability toward H2O2-mediated oxidative stress (Schulz et al., 2012). Stress responses were reversible, as the antioxidant coenzyme Q10 prevented cell death in challenged FTDP-17 neurons. Furthermore, GSK-3 inhibition in FTDP-17 neurons by CHIR99021 resulted in overall rescue of cell viability, consistent with a pro-apoptotic function of GSK-3 in the context of oxidative stress, as shown in SH-SY5Y cells (King and Jope, 2005). Furthermore, GSK-3-β showed increased activity in SH-SY5Y cells upon UPR activation (Song et al., 2002), which was readily observed in our FTDP-17 iPSC-derived neurons. These experiments provided proof of principle that an intervention is possible to prevent cell death in patients’ neurons at risk. At the same time, they highlight the unique potential of iPSCs to identify potentially beneficial compounds in FTDP-17, as human neurons can be produced efficiently and at high numbers for high-throughput drug screening purposes (Heilker et al., 2014). While UPR is initially a protective response mechanism toward misfolded proteins, its prolonged activation can lead to the onset of apoptosis and eventually to cell death. We observed an upregulation of BiP protein in FTDP-17 neurons, which was accompanied by an increased expression of p-PERK and the apoptosis-inducing protein PUMA. These findings indicated increased ER stress as a contributing factor to FTDP-17. In line with this observation, we also found UPR activation in the brains of patients with FTDP-17. UPR activation has been described in brains of patients with AD and in cases of FTDL-TAU (Hoozemans et al., 2009; Nijholt et al., 2012). In neurons of AD patients, markers of UPR activation appeared at early stages of disease development and were co-expressed with AT8 prior to NFT formation (Hoozemans et al., 2009). Furthermore, a recent whole-genome association study for PSP identified a small nuclear polymorphism in EIF2AK3, the gene encoding PERK, as a risk factor for the development of PSP (Höglinger et al., 2011). These studies and our present data on FTDP-17 neurons thus demonstrate that TAU pathology is closely linked to ER stress in tauopathies. The mechanisms, however, by which TAU interferes with ER function and UPR are still unclear, as TAU is located in the cytoplasm and not in the ER (Abisambra et al., 2013). Our whole-genome expression analysis identified dysregulated genes in FTDP-17 neurons. Downregulated genes included PAK3 and RIT2, which have been linked to mental retardation (Allen et al., 1998) and PD (Pankratz et al., 2012), respectively. Upregulated genes included LOC100128252 (ZNF667-AS1), a non-coding RNA transcript with currently unknown function, and the intracellular molecule MAGEH1, both of which were also highly expressed in the post-mortem brains of FTDP-17 patients. While further studies have to be performed to characterize the function of LOC100128252 in the context of FTDP-17, we could identify a beneficial role of MAGEH1 in patient-derived neurons. Our MAGEH1 knockdown experiments revealed increased susceptibility of FTDP-17 neurons to oxidative stress and demonstrated reduced neurite outgrowth. Accordingly, overexpression of MAGEH1 in healthy control neurons resulted in reduced vulnerability toward oxidative stress, suggesting that upregulation of MAGEH1 in FTDP-17 neurons represents a so far unknown endogenous neuroprotective mechanism. MAGEH1 can associate with the intracellular domain of the p75 receptor (Tcherpakov et al., 2002), which has been linked to cell survival and differentiation in neural progenitor cells (Casaccia-Bonnefil et al., 1999). It remains to be shown whether the beneficial effects of MAGEH1 in FTDP-17 neurons involve the p75 receptor and further studies are necessary to identify the mechanisms by which MAGEH1 promotes neuroprotection in FTDP-17 neurons.