iPSC-Induced Neurons with the V337M MAPT Mutation are Selectively Vulnerable to Caspase-Mediated Cleavage of Tau and Apoptotic Cell Death

Background: Tau post-translational modications (PTMs) are associated with progressive tau accumulation and neuronal loss in tauopathies, including forms of frontotemporal lobar degeneration (FTLD) and Alzheimer’s disease (AD). Tau proteolysis by caspases, including caspase-6, represents an understudied PTM that may increase neurotoxicity and tau self-aggregation. Methods: To elucidate the presence and temporal course of caspase activation, tau cleavage, and neuronal death, we generated two novel epitope (neoepitope) monoclonal antibodies (mAbs) against caspase-6 tau proteolytic sites. We evaluated tau cleavage and response to apoptotic stress in cortical neurons derived from induced pluripotent stem cells (iPSCs) with frontotemporal dementia (FTD)-causing V337M MAPT mutation. We tested the neuroprotective effect of caspase inhibitors in the induced neurons. We also demonstrated the presence of the tau neoepitopes in postmortem brains from an individual with FTD (V337M MAPT) and an individual with AD, compared to a healthy control. Results: FTLD V337M MAPT and AD postmortem brains showed positivity for both cleaved tau mAbs and active caspase-6. Relative to isogenic wild-type MAPT controls, V337M MAPT neurons cultured for 3 months showed a time-dependent increase in pathogenic tau in the form of caspase-cleaved tau and phosphorylated (p)-tau, and higher levels of tau oligomers. Accumulation of toxic tau species in V337M MAPT neurons was correlated with increased vulnerability to pro-apoptotic stress. Notably, this mutation-associated cell death was pharmacologically rescued by inhibition of effector caspases. Conclusions: Culturing iPSC-derived neurons for three months exposes age-related tau pathologies, including caspase-mediated cleavage, that are also observed in human postmortem brains with abnormal tau deposition. Neoepitope antibodies to caspase-cleaved

Caspases are proteolytic enzymes with well-de ned roles in cell death and in ammation.Apoptotic caspases are broadly classi ed as either initiator caspases that activate their effector counterparts or effector/executioner caspases that induce cell death [9][10][11].Accumulating evidence support a key pathogenic role of the effector caspase-6 in major diseases including AD, Huntington disease, and ischemic stroke [12][13][14][15].In AD, active caspase-6 is associated with three stages of cellular dysfunction, including in ammation [16], apoptosis, and tau cleavage [8, 17,18].We previously demonstrated a positive linear correlation between AD neuropathological stages and increasing levels of active caspase-6 and p-tau colocalization in the locus coeruleus and dorsal raphe nucleus, brain regions showing the earliest vulnerability to AD-tau pathology [15].Moreover, active caspase-6 was found to colocalize with NFT-positive neurons in AD patients' brains [15,17,19,20], promote axonal degeneration in several cellular and in vivo animal models of AD [21,22], and its levels inversely correlate with cognitive performance in aged individuals [23].While other caspases are likely also involved in neurodegenerative disease, caspase-6 inhibition may be su cient to reduce neuronal damage with a signi cant therapeutic index, since caspase-6 knockout animals are healthy and protected from AD pathology or proin ammatory stimuli [16,24].
Active caspase-6 may promote NFT formation by cleaving tau after aspartic acid (D) at multiple sites, including D421 near the C-terminus [8, 17,25].Tau cleaved at D421, which can be targeted by several caspases, has a stronger seeding propensity than full-length tau and is a component of the core of paired helical laments (PHFs) in AD [5,26,27].Caspase-6 cleaves tau at additional sites, including D13 [18] and D402 [23,28,29], which are also found to mediate tau aggregation and toxicity in vitro.Yet, these sites are underexplored compared to D421, mainly due to a lack of available monoclonal antibodies (mAbs).Currently, D13 is detected by the loss of N-terminal tau epitopes recognized by the antibodies Tau-12 and 5A6 [18], and D402 is detected using a polyclonal antibody [29].
To elucidate the role of caspase activation, caspase-6 tau cleavage, and neuronal death in tauopathies, we generated induced cortical neurons (iNs) from induced pluripotent stem cells (iPSCs) with the FTDcausing V337M MAPT mutation (tau V337M ) [30][31][32][33] and WT isogenic controls (tau WT ).iPSCs and their subsequent conversion to iNs by expression of neuronal transcription factors offer a clinically-relevant model of human tau pathogenesis in a disease-speci c genetic background [32,[34][35][36][37].We evaluated tau proteolysis using newly developed novel epitope (neoepitope) mAbs against caspase-cleaved tau at D13 and D402 sites in the iNs and human postmortem brains with tau accumulation.Our results demonstrate neoepitope mAb positivity in human brains from individuals with AD and FTLD-tau, a timedependent accumulation of caspase-cleaved tau, and increased vulnerability to acute apoptotic stress in the tau V337M iNs that is reversed by inhibition of effector caspases, including caspase-6.This study offers insights into the potential of caspase-6 and other caspases as targets for therapeutic intervention against tau pathology in FTLD and other tauopathies.

Development of caspase-6 cleaved tau neoepitope monoclonal antibodies
To address the absence of mAbs against tau sites primarily cleaved by caspase-6, we generated neoepitope mAbs against cleaved tau at D402 (mAbD402; 1-402) and D13 (mAbD13; 14-441) (Fig. 1a).We generated the mAbs by immunizing 6-8-week-old wild type Balb/c and SJL mice (SLAC) with keyhole limpet hemocyanin (KLH) conjugated tau peptides using protocols approved by the ChemPartner IACUC committee.50 µg of each peptide was injected into the abdominal cavity of each mouse along with 0.25 mL Complete Freund's Adjuvant (Sigma).To enhance the immune response, 25 µg of KLH conjugated tau peptides were injected into the abdominal cavity of each mouse along with 0.25 mL Incomplete Freund's Adjuvant (IFA) two weeks after the rst immunization, and subsequent boosts were administered three weeks apart.Blood samples from each mouse were collected one week after each immunization.The antibody titer and speci city were determined by enzyme-linked immunosorbent assay (ELISA) analysis and immunoblotting (see Supplemental Experimental Procedures).
Mice with speci c immune response against tau peptides and proteins were selected for fusion and were given a nal boost by intraperitoneal injection of 100 µg of the corresponding immunogen.After four days, mice were sacri ced and single-cell suspensions of splenocytes were prepared in NH 4 OH at 1% (w/w), followed by centrifugation at 1000 rpm and washes with DMEM (Invitrogen).Viable splenocytes were fused with mouse myeloma cells SP2/0 (ATCC) at a ratio of 5:1 with high-e ciency electric fusion (BTX ECM200).Fused cells were re-suspended in DMEM with 20% FBS and hypoxanthine-aminopterinthymidine (HAT) medium (Invitrogen).14 days after cell fusion, hybridoma supernatants were collected and screened by ELISA.Clones with an OD450 nm > 1.0 were expanded in a 24-well plate containing DMEM with 10% heat-inactivated FBS, and supernatants were collected after 3 days of culture.The antibody isotypes were determined, and ELISA and western blot were used to test their ability to bind to tau.Clones that showed desired reactivity and speci city against tau were subjected to subcloning to produce stable monoclonal hybridoma cells.Sub-cloning was carried out by limited dilution in a 96-well plate with DMEM media containing 10% FBS.Clones with speci c Tau binding were further expanded in DMEM media containing 10% FBS for subsequent antibody production with Protein A puri cation.Hybridoma cells were cryopreserved for future production.Reactivity and speci city of puri ed mAb.D13 and mAb.D402 were con rmed by ELISA against tau proteins and peptides (Fig. S1) and western blot analysis against recombinant tau proteins (Fig. S2-a, b).
Cell lines iPSC-derived induced neurons (iNs) with heterozygous V337M MAPT mutation (tau V337M ) and WT isogenic controls (tau WT ) were generated as previously described [32,38].The Neurogenin 2 (Ngn2)integrated iPSC line was created from a tau WT human iPSC line (male; WTC11) [38,39].The Tet-ON 3Gcontrolled Ngn2 transgene was integrated into the AAVS1 locus of human iPSC lines through a TALEN nuclease pair [38].CRISPR/Cas9 gene editing by homologous recombination was used to introduce the tau V337M into the Ngn2-integrated iPSCs.Brie y, iPSCs were transfected with the Human Stem Cell Nucleofector Kit (Lonza) with sgRNA (5'-CTTGTGGGGTCA-TGGTTTACAGG-3') plasmid (Addgene, 68463), Cas9 plasmid, and donor DNA plasmid containing a neomycin-resistance cassette (adapted from Addgene, PL552).Transfected cells were selected with neomycin for one week.Neomycin-resistant clones were selected and veri ed by genomic PCR and DNA sequencing.After sequence validation of the tau V337M site, 1 mM Cre recombinase (Excellgen) was added to remove the neomycin-resistance cassette from gene-targeted iPSCs.Moreover, the top 10 potential off-target sites were sequenced and for each genotype, the correct clones were karyotyped (Cell Line Genetics) and expanded for subsequent experiments.iNs were differentiated using a two-step protocol as previously described [32,38].See Supplemental Experimental Procedures for detailed methods.

Western blot protein analysis
Western blot analysis was performed using the Mini-PROTEAN Tetra system (Biorad) and standard immunoblotting techniques.See Supplemental Experimental Procedures for detailed methods.
Cell assays and compound treatment iNs cultured in 96-well plates were treated with either vehicle (DMSO) or staurosporine (STS) and the pancaspase inhibitor z-VAD-fmk (Enzo Lifesciences) directly in the media.Following 6h treatment, caspase-3/7 levels were examined using Caspase-Glo 3/7 (Promega) according to the manufacturer's instructions.
An ELISA based on lamin A cleavage, a speci c substrate of active caspase-6, was used for detection and quanti cation of caspase-6 activity as previously described [40].Brie y, iNs were treated with either vehicle (DMSO) or staurosporine (STS) and the caspase-6 inhibitor z-VEID-fmk (Enzo Lifesciences) directly in the media for 48h.Next, cells were xed with 4% paraformaldehyde (Thermo Fisher Scienti c) in PBS for 15 min and blocked with 5% bovine serum albumin (Sigma) in PBS with 0.01% Triton X-100 for 1h at room temperature.Cells were then incubated with cleaved lamin A primary antibody (Cell Signaling, 2035, rabbit, 1:200) overnight at 4°C, followed by an HRP-conjugated secondary antibody incubation (Invitrogen) for 1h at room temperature (GE Healthcare, NA934V, donkey anti-rabbit, 1:1000) a 5 min incubation with chemiluminescent HRP substrate (Thermo Fisher Scienti c), and reading with a Spectramax microplate reader (Molecular Devices).iN cytotoxicity was measured using lactate dehydrogenase (LDH) release assay (Promega) following 48h treatment with vehicle (DMSO), STS, or STS/ z-VAD-fmk according to the manufacturer's instructions.Readings were performed with a Spectramax microplate reader, followed by western blot or immunocytochemistry analysis.
To examine the morphological effects of STS and the z-VAD-fmk treatment in the iNs, we performed automated quanti cation of neurite length using IN Cell Developer Toolbox analysis routines based on image stacks uploaded from the IN Cell Analyzer 6500 confocal imager (GE Healthcare).To accommodate the treatment-induced variation of neurite morphology during image acquisition, the Software Autofocus function was chosen to optimize the focal plane for each imaging eld.DAPI was used for nuclei detection and microtubule-associated protein 2 (MAP2) for the detection of the cytoplasm and neurites.Mean neurite length per cell was calculated by subtracting a binary image of the cell body from a binary image of the entire cell, including the neurites.Data were obtained from 3-7 replicate wells per condition and an average of 1.5-2K cells per well.For IF analysis, xed neurons were imaged using an IN Cell Analyzer 6500HS confocal imager (GE Healthcare).For further details see Supplemental Experimental Procedures.

Statistics
Statistical analyses and graphics were generated using R Statistical Software (version 3.6.1;R Foundation for Statistical Computing, Vienna, Austria).Groupwise differences were analyzed by comparing means (student's t-test, one or two-way ANOVA with post hoc Tukey tests) with the family-wise error rate set to 0.05 to correct for multiple comparisons.Data from at least three independent experiments are presented as mean ± SEM.
Here, we developed two neoepitope mAbs against D13 and D402 tau cleavage sites (Fig. 1a) using hybridoma technology.Neoepitope antibodies selectively bound to cleaved tau (1-402 or 14-441) but not full-length tau (1-441) as detected by ELISA and western blot analyses (Fig. S1; S2-a, b; See Supplemental Experimental Procedures).Antibody speci city was also con rmed by immuno uorescence (IF) and antigen competition assays where mAb.D402 and mAb.D13 primary antibodies were preincubated with the cognate peptides used as immunogens (Table S1), resulting in loss of antibody signal (Fig. S2-c).Together with TauC3, mAb.D402, and mAb.D13 neoepitope antibodies were used to identify caspasemediated pathological changes in postmortem brains and in tau V337M and tau WT iNs.

Tau pathological changes in postmortem brains of FTLD and AD patients
We explored whether tau pathological changes are present in human postmortem brains with tau accumulation, including a case with the same tau V337M mutation as in the iN model described below.We immunostained sections from the temporal cortex of an FTLD tau V337M carrier and an AD patient, both at end-stage disease, and a healthy elderly control using multiplex immunohistochemistry targeting active caspase-6 and caspase-cleaved tau markers (Fig. 1b).We detected active caspase-6 positivity in the cytoplasm and neurites in both disease cases (brown), but not in the healthy control even after prolonged incubation with the chromogen to con rm the absence of signal (Fig. 1b vii-ix).The FTLD tau V337M carrier was also positive for TauC3 and mAb.D402 (red) in the cytoplasm and neurites that partially co-occurred with active caspase-6.AD brain sections showed strong positivity for all caspase-cleaved tau markers, including mAb.D13 that was absent in FTLD.Such variation on tau pathological burden is expected among individuals with different tauopathies.Moreover, neurons in the AD sections showed partial overlap between cleaved-tau and active caspase-6 in cytoplasm and neurites.Taken together, our results show that neoepitope mAbs label pathological tau inclusions in AD and FTLD-tau and corroborate the presence of active caspase-6 in the human brain seen in earlier studies [15].

Characterization of iPSC-induced neurons
Considering that tau V337M causes FTD in humans, we generated heterozygous tau V337M iNs and isogenic tau WT controls to examine the disease mechanisms involving caspase activation and cleaved tau pathology in a clinically relevant cell culture model.Neuronal induction of a well-characterized human tau WT line was performed using TALEN-based integration of a doxycycline-inducible Ngn2 transgene into the AAVS1 safe harbor as previously described [32,38].CRISPR/Cas9 genome editing was used to introduce the tau V337M into tau WT iPSCs to study the mutation's effects in isolation from the donor's genetic background.Both iPSC groups had a normal karyotype (Fig. S3-a) and typical colony-type morphology (Fig. 2a).Moreover, genomic DNA sequencing of tau V337M and tau WT iPSCs con rmed the presence of the heterozygous tau V337M in exon 12 (Fig. S3-b), and homogeneous expression of the pluripotency markers NANOG, OCT4, and SOX2 (Fig. S3-c) in the iNs con rmed the absence of NGN2 expression leakage without the addition of doxycycline.Following doxycycline treatment, iNs exhibited neuron-like morphology in 5-7 days and mature neuronal morphology between 3-4 weeks (Fig. 2a).
One month-old iNs were positive for the microtubule-associated protein 2 (MAP2) neuronal marker in the cytoplasm and neurites, and for the deep and upper cortical nuclear markers CTIP2 and SATB2, respectively, as con rmed by IF (Fig. 2b).iNs were also positive for the glutamatergic marker vesicular glutamate transporter 1 (VGlut1) in the cytoplasm and neurites, an anticipated outcome of the NGN2 expression [32,41].Overall, our ndings demonstrate that at one-month post-differentiation, iNs exhibit neuron-speci c morphology and express cortical and glutamatergic markers.
Increased levels of pathological tau in the tau V337M neurons Since FTLD-tau is characterized by progressive accumulation of toxic tau species and neuronal loss, we examined the presence and temporal course of tau pathological accumulation of the tau V337M in the mutant iNs relative to WT isogenic controls.We compared iNs cultured from 1 to 3 months using western blot and a panel of tau antibodies, including total and oligomeric tau, 3R and 4R tau, and caspasecleaved tau (Fig. 3).
Protein analysis using the total tau antibody HT7 showed positivity for distinct molecular weight bands corresponding to separate tau isoforms (Fig. 3a).Tau antibody speci city was con rmed by the recombinant human tau ladder included here as an approximate guide of the tau isoform placement as it lacks the tau PTMs present in the cell lysates that could in uence tau molecular weight.In agreement with previous studies showing enriched 3R tau levels in early neuronal development [42,43], we detected predominantly 3R isoform expression in the iNs and minimal 4R tau levels, both in tau V337M and tau WT iNs (Fig. 3a).Antibody speci city for 3R and 4R tau was con rmed by the positivity of the respective isoform bands in the recombinant tau ladder.Total tau levels were comparable between the mutant and control iNs (Fig. S4a), suggesting that any changes in the amount of pathological tau between the two groups are likely MAPT mutation-dependent and not due to changes in the overall tau levels.
Oligomeric tau aggregates could represent highly toxic and pathologically signi cant tau species in tauopathies [44].We therefore investigated the presence and temporal changes of oligomeric and misfolded tau aggregates recognized by the antibody T18 [45] and non-denaturing conditions to preserve the original folded state of tau (Fig. 3b).To estimate the molecular weight of the native proteins, we used a protein standard for native electrophoresis stained with Coomassie Blue for band visualization.We detected elevated oligomeric tau levels in tau V337M iNs relative to controls at 2 and 3 months post differentiation.These results were reproduced in three independent experiments (Fig. S4-b-d).The protein standard revealed bands at 480-1,048 kD molecular weight, corresponding to tau oligomers, with no tau detected below that range.Conformational tau species identi ed by the MC1 antibody [46] represent one of the most commonly detectable pathological features of tauopathies.Using the same experimental approach, we probed iN lysates using native electrophoresis.We observed a similar band pattern to T18, with elevated levels of conformational tau present after two months of culture only in the mutant iNs (Fig. 3b, S4 b-d).Overall, our results indicate a time-dependent increase of tau toxic species in the form of oligomeric and conformationally-modi ed tau in the mutant iNs compared to controls at two months post differentiation, but not earlier.
Next, we examined the temporal changes in p-tau and caspase-cleaved tau levels in the iNs using the anti-phospho-tau mAb PHF1 (Ser396/404), an epitope that is phosphorylated early in tau inclusion formation in humans [47], and three caspase-cleaved tau mAbs, including cleavage sites primarily targeted by caspase-6, D402 and D13, and the D421 site cleaved by multiple caspases [8,17].Semiquantitative analyses of GAPDH-normalized band intensities revealed a 2.5-fold increase of PHF-1 levels in tau V337M relative to controls at three months post differentiation but not earlier (Fig. 3c).Similarly, caspase-cleaved tau levels showed a 2.5 to 3-fold increase of D421 and D402-positive bands and a 2-fold increase of the D13 in tau V337M iNs relative to controls at three months post differentiation.Again, these differences were absent in younger cells (Fig. 3d-f).Altogether, our results demonstrate a signi cantly higher accumulation of p-tau and caspase-cleaved tau in tau V337M iNs relative to controls that accumulated at three months post differentiation, but not at earlier stages.

Caspase inhibition is neuroprotective against stress-induced cytotoxicity in the tau V337M neurons
To test whether tau V337M iNs are more vulnerable to stressors relative to controls, we treated three-month tau V337M iNs with increasing concentrations of the wide-spectrum kinase inhibitor and apoptosis-inducer staurosporine (STS) for up to 48h.Treatment was followed by detection of cytotoxicity levels measured by lactate dehydrogenase (LDH) release (Fig. S5).We observed a 2-fold increase in cytotoxicity levels in the tau V337M iNs relative to tau WT controls, using 40 µM STS for 48h (Fig. S5-b), and selected this condition for further studies.Cytotoxicity was signi cantly reduced by treatment with the pan-caspase inhibitor z-VAD-fmk (300 µM for 48h or 4 x 75 µM in 12h intervals; Fig. S5-c), indicating that cell death occurred via apoptosis.
Based on these established conditions (Fig. S5), we exposed iNs to 40 µM STS for 48h and increasing doses of z-VAD-fmk (300 and 600 µM; Fig. 4).Tau V337M and tau WT iNs treated with vehicle (DMSO) or z-VAD-fmk (600 µM) in the absence of STS showed low LDH release (Fig. 4); hence, baseline levels of apoptosis were low for iNs cultured for three months.Following STS treatment, however, we observed almost a 5-fold increase in cytotoxicity in the tau V337M iNs compared to a 4-fold increase in control iNs.The mutant group was signi cantly more vulnerable to cytotoxicity following acute stress by STS (p < 0.001).Moreover, STS co-treatment with z-VAD-fmk signi cantly reversed cytotoxicity levels in the mutant iNs.Overall, our results demonstrated increased vulnerability to apoptotic cell death in tau V337M iNs that was ameliorated by caspase inhibition.
To assess speci c changes in the activity of caspase − 6 and − 3/7 between the tau V337M and tau WT iNs following stress exposure, we treated three-month tau V337M and control iNs with STS and measured caspase activity.Caspase-6 activity was measured by ELISA for cleaved lamin A, a selective substrate of caspase-6 over other caspases [48], while Caspases-3/7 activity was measured using a DEVDaminoluciferin substrate assay (Fig. 4).We observed a 4-fold increase in caspase-6 activity in the tau V337M iNs compared to a 1-fold increase in control iNs following 20 µM STS treatment for 48h (Fig. 5a), indicating that the mutant group had signi cantly higher caspase-6 activity relative to controls (p < 0.01).Notably, STS co-treatment with 10 µM of the caspase-6 inhibitor z-VEID-fmk signi cantly reversed caspase-6 activity levels in the mutant iNs (p < 0.001, z-VEID-FMK/STS vs STS alone).We also observed a 6-fold increase in active caspase-3/7 levels (Fig. 5b) in the tau V337M and tau WT treated with 40 µM STS for 6h; there was no signi cant difference between caspase-3/7 activity between tau V337M and tau WT iNs.Caspase activity was reversed to baseline levels after the addition of 300 µM of the pancaspase inhibitor z-VAD-fmk.Overall, these results demonstrate that STS treatment induced apoptotic cell death and a marked increase in caspase activity in the iNs that was suppressed by caspase inhibition.
Next, we performed western blot analyses using lysates of iNs treated with 40 µM STS and 300 µM z-VAD-fmk (Fig. 5c-d).Following STS treatment, we observed a statistically signi cant 2-fold increase in TauC3 binding in the tau V337M iNs treated with STS compared to vehicle-treated cells (p < 0.001).TauC3 binding was not increased in control iNs treated with STS.In line with the cytotoxicity data (Fig. 4), STS co-treatment with z-VAD-fmk in tau V337M iNs reversed TauC3 binding to baseline levels.

Stress-induced reduction of neurite length is rescued by caspase inhibition
The reduction in neurite length is a morphological indicator of compromised cell viability and neurotoxicity [34,49].To examine the phenotypic effects of STS and z-VAD-fmk treatment on neurites in mutant and control iNs, we used immunocytochemistry (ICC) with DAPI to label cell nuclei (blue), and MAP2 (red) to label cytoplasm and neurites (Fig. 6a).Cells were imaged and subjected to automated quanti cation of neurite length under the same treatment conditions as our cytotoxicity assay (Fig. 4).
We observed comparable mean neurite length in untreated iNs; neurite lengths were reduced upon treatment with 40 µM STS by 1.6-fold for tau V337M (p < 0.0001) and 1.3-fold (p < 0.0001) for tau WT (Fig. 6b).Co-treatment of STS with z-VAD-fmk partially restored neurite length and preserved MAP2positive processes in tau V337M iNs compared to STS-treated cells (p < 0.01).For tau WT iNs, neurite length in STS treated cells was also restored upon the addition of z-VAD-fmk (p < 0.05).These data further demonstrate that neurotoxicity following STS treatment in the iNs was signi cantly caspase-dependent.

Discussion
Active caspases can directly cleave tau, generating tau species with an enhanced propensity for selfaggregation and accumulation into neuronal cytoplasmic inclusions [8,18,29,50].Most of this evidence comes from studies using animal models and immortalized cell lines that do not necessarily recapitulate the clinical phenotype.Our study innovates by using (a) a disease-relevant, iPSC-derived human neuronal model with the FTLD-causing tau V337M , (b) neoepitope mAbs to detect caspase-6 cleaved tau species, and (c) pharmacological intervention against caspase activity.Different from most studies, we maintained tau V337M and tau WT iNs for up to three months to avoid capturing artefactual effects caused by age-dependent increases in expression of p-tau in developing neurons [42,51].We additionally see an age-dependent accumulation of pathogenic tau, including cleaved tau, p-tau, and oligomeric tau, in cells containing pathogenic tau V337M .Of particular interest, three-month tau V337M iNs showed a three-fold increase in caspase-cleaved D402 and D13 tau protein levels and increased p-tau and oligomeric tau species compared to controls.These mutant cells were also more sensitive to caspase-6 activation and apoptosis following acute treatment with STS.We furthermore con rmed the presence of active caspase-6, and D13 and D402 tau cleavages in postmortem tissue from patients with AD and MAPT V337M FTLD-tau.In contrast to the high levels of caspase-6 seen only in the tau V337M , activation of caspase-3 and − 7 was seen in both tau V337M and tau WT iNs after treatment with STS.The selective upregulation of active caspase-6 suggests that increased cell death following acute stress in the tau V337M iNs could be caspase-6 related.For example, previous studies in human primary neurons showed that caspase-6 but not caspase-3 is activated by serum deprivation and induces cell death following recombinant active caspase-6 and − 3 microinjection [52,53].
Our data are in alignment with earlier studies using iNs with MAPT mutations that show increased vulnerability to stressors in mature iNs cultured for at least two months [37].iNs with tau V337M MAPT mutation are selectively vulnerable to stressors, likely due to increased pathogenic tau levels and a lower capacity to survive additional stress relative to WT controls [34].Tau V337M iNs also showed tau cleavage at D421, tau detachment from microtubules, compromised axonal transport, and dysregulation of neuronal excitability [32,34].Moreover, transgenic mice expressing V337M human tau exhibited neurodegeneration and p-tau accumulation in the hippocampus, as well as behavioral abnormalities [54].However, little was known about cleaved tau in the iN model besides the D421 cleavage site.The mAb.D402 and mAb.D13 neoepitope antibodies facilitated the detection of tau species cleaved by caspase-6, a protease closely associated with NFTs and neuronal loss in the human AD brain, including CA1 of the hippocampus, entorhinal cortex, the olfactory bulb, and brainstem [15,23,28,29].Our results highlight and support previous suggestions of a central role of active caspase-6 in tauopathies.
Several of the differences detected between tau V337M and tau WT iNs only became apparent in iNs cultured for at least two or even three months.Western blot protein analysis of 1-3-month cultured iNs revealed a time-dependent upregulation of T18 oligomeric tau and MC1 conformational tau at two months in the tau V337M iNs relative to controls.P-tau levels recognized by the PHF-1 antibody increased in the tau V337M iNs relative to controls at three months, but not earlier.Physiological upregulation of p-tau in developing neurons is a known phenomenon in vitro and in human brains.Here, high levels of p-tau in tau WT iNs only started to subside at two months post-differentiation. Future studies using iNs would bene t from using more mature neurons, despite the technical di culties in maintaining neuronal lines in culture long-term.Some of our results were intriguing and unexpected.Despite observing similar morphological and biochemical features in our tau V337M iNs compared to other studies using tau V337M lines, not all previously described changes were reproduced in our cells.We did not detect tau V337M -speci c neurite length changes in early-differentiated iNs, as reported in another tau V337M model [34].This is likely due to different neuronal induction methodologies and resulting cell phenotypes.Although our results strongly suggest a signi cant role for caspase-6 relative to other effector caspases in tau truncation and neurotoxicity, we could not demonstrate with certainty which caspase(s) were responsible for tau cleavage in our cell model because the peptidic z-VEID-fmk inhibitor is only partially selective for caspase-6 [48] and z-VAD-fmk caspase inhibitors are nonselective.Our understanding of the mechanisms of action of caspase inhibitors would tremendously bene t from follow up studies using highly selective, nonpeptidic inhibitors of caspase − 6 that are currently under development.Moreover, exploration of apoptosis-related substrates targeted by caspases − 3 and − 6 in addition to tau, including nuclear and mitochondrial proteins, could further clarify the mechanisms of caspase-mediated neurotoxicity in our cell model.
In summary, our study demonstrates a time-dependent increase of tau cleavage and a pronounced vulnerability to stress and neuronal death in the tau V337M iNs that is pharmacologically reversed by caspase inhibition.Our data support a model in which tau V337M leads to activation of caspases, which in turn leads to increased vulnerability to toxic insult, including cleavage of tau to toxic or aggregation-prone tau species.While both tau V337M and tau WT iNs respond to STS by activating caspases, tau V337M iNs are more vulnerable to tau proteolysis and cell death after caspase activation.Future mechanistic studies of caspase-dependent pathways across MAPT genotypes in iN models could reveal viable therapeutic targets against tau pathogenesis in FTLD and other tauopathies.

Conclusions
The mechanisms leading to tau pathology in frontotemporal dementia (FTD) and Alzheimer's disease (AD) remain elusive.Our data demonstrates that iN cultured for three months better model tau pathology than shorter culture times.Results with three-month old cultures of iNs bearing the V337M MAPT mutation demonstrate the deleterious role of caspase activation in tau pathology, the activation of caspase-6 in particular, and the therapeutic potential of caspase inhibitors against tau-related neurodegenerative pathology.

Declarations
Ethics approval and consent to participate Use of human brain tissue and iPSC-derived induced neurons from humans have been approved by the University of California San Francisco Institutional Review Board.

Consent for publication
Not applicable, the manuscript does not contain data from any individual person.(a) Expression of total tau (HT7) and tau isoforms (3R and 4R tau) in tauV337M and tauWT neurons cultured from 1 to 3 months.Antibody speci city was con rmed by a recombinant tau ladder containing all six tau isoforms (right panels); (b) Oligomeric (T18) and conformational (MC1) tau levels were detected under non-denaturing conditions for preserving protein conformation.The molecular weight of the proteins was estimated using a protein standard for native electrophoresis stained with Coomassie blue (far-left panel) for band visualization (see also

Supplementary Files
This is a list of supplementary les associated with this preprint.Click to download.

Figure 5 Stress
Figure 5