DNA damage and transcriptional regulation in iPSC-derived neurons from Ataxia Telangiectasia patients

Ataxia Telangiectasia (A-T) is neurodegenerative syndrome caused by inherited mutations inactivating the ATM kinase, a master regulator of the DNA damage response (DDR). What makes neurons vulnerable to ATM loss remains unclear. In this study we assessed on human iPSC-derived neurons whether the abnormal accumulation of DNA-Topoisomerase 1 adducts (Top1ccs) found in A-T impairs transcription elongation, thus favoring neurodegeneration. Furthermore, whether neuronal activity-induced immediate early genes (IEGs), a process involving the formation of DNA breaks, is affected by ATM deficiency. We found that Top1cc trapping by CPT induces an ATM-dependent DDR as well as an ATM-independent induction of IEGs and repression especially of long genes. As revealed by nascent RNA sequencing, transcriptional elongation and recovery were found to proceed with the same rate, irrespective of gene length and ATM status. Neuronal activity induced by glutamate receptors stimulation, or membrane depolarization with KCl, triggered a DDR and expression of IEGs, the latter independent of ATM. In unperturbed A-T neurons a set of genes (FN1, DCN, RASGRF1, FZD1, EOMES, SHH, NR2E1) implicated in the development, maintenance and physiology of central nervous system was specifically downregulated, underscoring their potential involvement in the neurodegenerative process in A-T patients.

The parental WT1 hNPCs were employed to knockout the ATM gene using the CRISPR-GFP system, as detailed in Materials  DIV55 neurons were treated with KCl for 1hr and analysed by RNAseq. The numbers indicate the up-and down-regulated differentially expressed genes that are unique to each cell type or common between them. Specifically, a small set of genes that included NR4A3, FOS, NR4A1 and NPAS4, was strongly up-regulated after treatment in all three cell lines.

Suppl Fig 9: qRT-PCR analysis of FOS
Neurons at DIV55 were treated with KCl for 1hr and then analysed by qRT-PCR with primers specific for FOS and GAPDH (for normalization). Results are from two independent experiments and values +/-SD for FOS were normalised against those of GAPDH (** p<0.01).

Suppl Fig 10: Venn diagram of basal differentially expressed genes in WT1
compared to A-T1 and ATM-KO1 neurons.
Transcripts that are differentially expressed between WT1, A-T1 and ATM-KO1, identified by RNA-seq analysis of unstimulated neurons. Numbers indicate the differentially expressed genes that are unique or common between each group. FN1, DCN, RASGRF1, FZD1, EOMES, SHH, NR2E1 showed an ATM-dependence, being expressed in WT1 but not in A-T1 or ATM-KO1. Of these genes, RASGRF1, a Ca 2+ activated protein, is involved in spinogenesis of primary hippocampal cultures, while the nuclear receptor NR2E1 (also known as TLX) participates in synaptic plasticity and dendritic structure formation in the dentate gyrus [47] (Christie et al, 2006), as well as neurogenesis, learning and memory [48] (Murai et al, 2014).
Moreover, lower expression levels were detected in A-T1 and ATM-KO1 neurons for the T-box transcription factor EOMES, a key regulator of neurogenesis in the subventricular zone (SVZ) [49] (Arnold et al, 2008) and for the morphogenic factor SHH. Down-regulation of SHH could be particularly interesting since this gene is involved in cerebellum in the proliferation of granular cell precursors, differentiation of Bergmann glial cells and normal Purkinje neuron development [50, 51] (De Luca et al, 2016;Lee et al, 2010). Curiously, we noticed another A-T1 and ATM-KO1 common down-regulated GO Term, described as "negative regulation of cell death" (GO:0060548), in which we found the following genes: SHH, CTGF, NR2E1, CARD16. This could suggest that the absence of ATM might make neurons more apoptogenic. In addition to the shared GO terms, we focused on two categories associated with DDR induced by ionizing radiation, a finding in accordance with the marked radiosensitivity of A-T cells. Specifically, A-T1 neurons displayed two downregulated genes belonging to the cellular response to X-ray (GO:0071481) while ATM-KO1 neurons had lower expression levels for three genes associated with cellular response to γ-radiation (GO:0071480) (data not shown).

Additional Experimental procedures:
Karyotype analysis. hiPSCs were incubated overnight with 20ng/ml of the mitotic inhibitor Colcemid (Thermo Fisher Scientific), thereafter incubated in hypotonic solution (KCl; 1 h at 37 o C), and then fixed with Carnoy's fixative (3:1 methanol to acetic acid). Metaphase spreads were aged at room temperature for 5-7 days and banded with Wright stain.

Embryoid bodies generation and differentiation into cells of the 3 germ-layers.
hiPSCs colonies were incubated for 10 min at 37°C with 1mg/ml Collagenase type IV (Thermo Fisher Scientific) dissolved in DMEM F-12. After scraping, heterogeneous cell clumps were transferred into ultralow attachment six-well plates (Euroclone) and kept in a medium optimized for EBs growth (DMEM F-12, 20% KSR, 2 mM Lglutamine, 0.1 mM non essential aminoacids,100 units/mL penicillin, 100 µg/mL streptomycin, 1mM Sodium Pyruvate from LONZA, 1% N-2 Supplement (Thermo Fisher Scientific), 0.11mM β-ME (Sigma Aldrich). After 6 days of floating cell culture, EBs were transferred onto coverslips (13mm diameter) pre-coated with Geltrex TM LDEV Free Reduced Growth Factor Basement Membrane Matrix (Thermo Fisher Scientific) and kept for 7 days in a 3 germ layer differentiation medium (DMEM F-12, FBS 20%, 2 mM L-glutamine, 0.1 mM non essential aminoacids ,100 units/mL penicillin, 100 µg/mL streptomycin). Medium changes were performed 3 times per week for both floating and adherent cell cultures. At the end of the differentiation step, coverslips were fixed and immunostained with antibodies against the three germ layers (see Immunofluorescence section).

Generation of neural precursor cells (hNPCs) from hiPSCs. Proliferating neural
precursor cells (NPCs) were obtained as reported (1S). Essentially, hiPSCs clones were, the day before neural induction, split to get a 20% confluent cell culture; after that mTESR1 medium was replaced with the PSC Neural Induction Medium containing Neurobasal medium, 2% Neural Induction Supplement 50X (all from Thermo Fisher Scientific), 100 units/mL penicillin, and 100 µg/mL streptomycin, and cells cultured for 7 days. Induced cells were subsequently dissociated with Accutase solution (Carlo Erba Reagents), counted and seeded on a Geltrex pre-coated dish at 1x10 5 cells/cm 2 in the Neural Expansion Medium (50% Neurobasal medium, 50% Advanced DMEM F-12, 2% Neural Induction Supplement 50X, 100 units/mL penicillin, 100 µg/mL streptomycin). The medium was supplemented overnight with 1% RevitaCell Supplement, to improve cell viability. After induction, hNPCs were split with Accutase solution and used for terminal differentiation for no longer than 25 passages. During initial passages, the hNPCs were selectively isolated from the cultures containing non-neural cells through a differential gentle detachment with Terminal neural differentiation. After six passages, hNPCs were seeded at a 5x10 4 cells/cm 2 in a 2 fold concentrated Geltrex pre-coated plate and cultured with a neural differentiation medium (NDM) containing: Neurobasal medium, 2% B27 supplementserum free, 2mM GlutaMAX supplement (Thermo Fisher Scientific), 10ng/µl brainderived neurotrophic factor, 10ng/µl glial cell-derived neurotrophic factor (both from Immunotools), 200µM L-ascorbic Acid (Merck), 0.1 mM non essential aminoacids, 100 units/mL penicillin, 100 µg/mL streptomycin. During the differentiation, which was carried out for up to 55 days in vitro (DIV), 75% of the spent NDM was replaced with fresh NDM three times/week. phosphocreatine-tris, 2% ATP-Na 2 , 0.2%mM GTP-Na 2 and 0.1%leupeptin, pH 7.2 with KOH. Pipette resistance was between 3 and 4%MΩ. Cell capacitance and series resistance errors were carefully compensated (∼85%) throughout the experiment. The remaining linear capacity and leakage currents were eliminated online using a P/4 subtraction paradigm. Total voltage-gated currents were elicited by applying 66 mslong depolarizing voltage steps from -70 to +50 mV (10 mV increments), from a holding potential of -70 mV; signals were filtered at 10%kHz and sampled at 200%kHz.
Discharges of action potentials were evoked by the injection of 2.5 s-long depolarizing current pulses of increasing amplitude from the resting potential maintained at -70 mV; signals were filtered at 10%kHz and sampled at 20%kHz. Spontaneous excitatory and inhibitory postsynaptic currents (sEPSCs, sIPSCs) were recorded at -50 and 0 mV, respectively; signals were filtered at 3 kHz and sampled at 10 kHz.
Western blots and immunofluorescence. Western blot analysis were performed as reported [16] Briefly, cells detached with Accutase were washed and lysed in 20-100µl of lysis buffer (0.125 M Tris-HCl pH 6.8, 5% SDS) containing proteases and phosphatases inhibitors. Following sonication, centrifugation and quantification samples were loaded (30-50µg) on Novex NuPAGE precast gels, electrophoresed and blotted on PVDF membranes (Merck). After blocking in 4% nonfat dry milk , membranes were incubated with primary antibodies overnight at 16°C using the X-BLOT P100 System (Isenet, Milano, Italy), then for 1hr with horseradish peroxidaseconjugated secondary antibodies (GE Healthcare). Proteins were detected by chemiluminescence and signals quantitated by densitometric analysis using the Coverslips were incubated with primary antibodies diluted in 5% NGS for 3 hours at RT or overnight at 4°C, washed 3x with PBS and then incubated for 45 min at RT with appropriate dilutions of Alexa Fluor555 and 488 conjugated secondary antibodies (Thermo Fisher Scientific). Following washing with PBS, coverslips were incubated for 10 min with 0.35µg/ml of DAPI nuclear stain, mounted onto microscope slides using Prolong Gold antifade reagent (Thermo Fisher Scientific) and analyzed on a Nikon Eclipse E1000 fluorescence microscope equipped with a DS-U3 CCD digital camera. To increase the quality of Tra1-81 and SSEA4 IF staining, iPSCs dissociated into single cells were cytocentrifuged (500 rpm, 5 min) onto microscope glass slides with a Shandon Cytospin 2 Centrifuge (Block Scientific, Inc.), air-dried overnight and then fixed with PFA fixation.

53BP1 foci immunodetection.
Neurons grown on coverslips were treated with genotoxic agents, then fixed with 2% PFA for 20 min, washed and treated with a permeabilization solution (20mM HEPES buffer, 50mM NaCl, 3mM MgCl2, 300mM sucrose, 0.5%Triton X-100, pH 7.6) for 5 minutes. After blocking with 5% BSA (Sigma Aldrich) and 0.1% Tween-20 for 10min, coverslips were incubated with the 53BP1 antibody in 2% BSA for 45 min at RT, then extensively washed and incubated for 45min at RT with Alexa-Fluor555 secondary antibody diluted in 1% BSA. Following washing and counterstaining with DAPI, coverslips were mounted as above. Hybridization and washing were performed according to Agilent's standard protocols.
Microarray images were acquired using an Agilent DNA microarray scanner and raw data were generated using the Feature Extraction Software v10.7.3 involving automatic grid positioning, intensity extraction (signal and background). Raw data were preprocessed using the limma package [4S]. Briefly, after background correction with the normexp method, raw data were normalized with quantile method and log2 transformed. Spots detected in at least two samples according to the gIsPosAndSignif and gIsWellAboveBG flags were kept for further analyses. Finally the dataset was collapsed at the gene level calculating the mean expression of probes mapping on the same gene. Data were deposited in the Gene Expression Omnibus repository (GSE108605). Differentially expressed genes were identified using the linear model approach implemented in the limma package. Multiple-testing correction was performed using the Benjamini-Hochberg false discovery rate (FDR) [4S] . Genes with FDR < 0.05 and absolute fold-change ≥ 2 were considered significant. Pre-