CRISPR-Cas9 genetic screen leads to the discovery of L-Moses, a KAT2B inhibitor that attenuates Tunicamycin-mediated neuronal cell death

www.nature.com/scientificreports OPEN CRISPR‑Cas9 genetic screen leads to the discovery of L‑Moses, a KAT2B inhibitor that attenuates Tunicamycin‑mediated neuronal cell death Sofia Pavlou 1,2*, Stefanie Foskolou 1,2, Nikolaos Patikas 1, Sarah F. Field 1,2, Evangelia K. Papachristou 3, Clive D’ Santos 3, Abigail R. Edwards 3, Kamal Kishore 3, Rizwan Ansari 1, Sandeep S. Rajan 1,2, Hugo J. R. Fernandes 1,2 & Emmanouil Metzakopian 1* Accumulation of aggregated and misfolded proteins, leading to endoplasmic reticulum stress and activation of the unfolded protein response, is a hallmark of several neurodegenerative disorders, including Alzheimer’s and Parkinson’s disease. Genetic screens are powerful tools that are proving invaluable in identifying novel modulators of disease associated processes. Here, we performed a lossof-function genetic screen using a human druggable genome library, followed by an arrayed-screen validation, in human iPSC-derived cortical neurons. We identified and genetically validated 13 genes, whose knockout was neuroprotective against Tunicamycin, a glycoprotein synthesis inhibitor widely used to induce endoplasmic reticulum stress. We also demonstrated that pharmacological inhibition of KAT2B, a lysine acetyltransferase identified by our genetic screens, by L-Moses, attenuates Tunicamycin-mediated neuronal cell death and activation of CHOP, a key pro-apoptotic member of the unfolded protein response in both cortical and dopaminergic neurons. Follow-up transcriptional analysis suggested that L-Moses provided neuroprotection by partly reversing the transcriptional changes caused by Tunicamycin. Finally, L-Moses treatment attenuated total protein levels affected by Tunicamycin, without affecting their acetylation profile. In summary, using an unbiased approach, we identified KAT2B and its inhibitor, L-Moses, as potential therapeutic targets for neurodegenerative diseases. A shared pathology between various neurodegenerative diseases is the accumulation of misfolded and aggregated proteins in the brain, leading to cellular dysfunction, loss of synaptic connectivity and eventually neuronal d eath1. However, distinct proteins, brain regions and neuronal subtypes are affected across different neurodegenerative diseases. For example, Alzheimer’s disease (AD) is associated with accumulation of misfolded and aggregated amyloid beta (Aβ) and tau1, with the neocortex and hippocampus being the most affected areas2. Parkinson’s disease (PD), on the other hand, is characterized by the presence of misfolded and aggregated alpha-synuclein1,3 and the dopaminergic neurons within the substantia nigra are the most vulnerable cell t ype2. Accumulation of misfolded proteins leads to endoplasmic reticulum (ER) stress4. To alleviate ER stress and restore homeostasis, cells undergo a rapid and highly coordinated signalling cascade, known as the unfolded protein response (UPR). The UPR activates apoptosis if the damage is irreversible. C/EBP-Homologous protein (CHOP, also known as DNA Damage Inducible Transcript 3—DDIT3) is a key pro-apoptotic gene, activated by UPR5. Emerging evidence indicate a key role for ER stress and UPR in the pathophysiology of protein misfolding disorders4,6,7. Therefore, identifying novel modulators of ER stress that dampen the UPR has great therapeutic potential. 1 UK Dementia Research Institute, Department of Clinical Neurosciences, Cambridge Biomedical Campus, University of Cambridge, Cambridge CB2 0AH, UK. 2Open Targets, Wellcome Genome Campus, Hinxton, Cambridge CB10 1SA, UK. 3Cancer Research UK Cambridge Institute, University of Cambridge, Li Ka Shing Centre, Robinson Way, Cambridge CB2 0RE, UK. *email: ; Scientific Reports | (2023) 13:3934 | https://doi.org/10.1038/s41598-023-31141-6 1 Vol.:(0123456789) www.nature.com/scientificreports/ CRISPR-Cas9 mediated genetic screens allow the systematic identification of modulators of a specific process. Here, we performed a loss-of-function genetic screen to identify genes involved in Tunicamycin (Tun)-mediated neuronal death of human iPSC-derived cortical neurons. Tun blocks the initial step of glycoprotein biosynthesis within the ER, resulting in accumulation of unfolded glycoproteins and subsequently ER s tress8 and is therefore widely used to induce ER stress in tissue culture systems. Here, we performed a genetic screen using a sgRNA library against the druggable genome9 to identify genes, whose loss-of-function attenuates Tun-mediated neuronal death. We validated our hits in an arrayed format and identified 13 genes, whose knock-out provided neuroprotection against Tun. L-Moses, a Lysine Acetyltransferase 2B (KAT2B) i nhibitor10, reduced neuronal cell death, ER stress and UPR activation in both cortical and dopaminergic neurons. To our knowledge, this is the first time L-Moses was shown to be neuroprotective and to have therapeutic potential for both AD and PD. Materials and methods Cell culture. NGN2-OPTi-OX was a kind donation by the Kotter lab at the University of Cambridge, where the original iPSC line was sourced from the University of Cambridge (https://hpscreg.eu/cell-line/CAMi0 14-A)11. A second NGN2-iPSC line12 was sourced by the Ward lab at NIH, as a kind donation. KOLF-2 were sourced by the HipSci consortium and the Wellcome Sanger Institute (https://www.hipsci.org/lines/#/lines/ HPSI0114i-kolf_2). Human iPSCs NGN2-OPTi-OX cells were maintained and cortical-like iNeurons were induced as described before13. Tun-treatment was performed on day 14 (d14). A second independent cell line, engineered in the Ward lab12, was used to validate the effect of L-Moses. Human iPSCs (KOLF-2 or NGN2-OPTi-OX) were differentiated into dopaminergic neurons as previously described14. Tun-treatment was performed on d32–d35 neurons. HEK293FT cells were cultured in DMEM/F12 supplemented with 10% FBS and 1% Pen/Strep. iPSC engineering. Cas9 was inserted upstream of GAPDH, by CRISPR-mediated homology-directed repair as previously described15. Several clones were tested, and one was selected for all experiments, based on its Cas9-editing efficiency. Editing efficiency was assessed by transducing d4 neurons with a lentiviral vector encoding BFP, GFP and a sgRNA for GFP. Neurons were dissociated on d8 and subjected to flow cytometry using CytoFLEX S (Beckman Coulter). Editing efficiency was calculated by comparing the percentage of B FP+/ + + GFP (edited) to B FP /GFP (total transduced) cells using FlowJo version 10.8.1 Software for macOS (BD Life Sciences). Tag-RFP was inserted downstream of CHOP, separated by a T2A as described previously15. The following sgRNA sequence was used: 5ʹ-UGCUCCCAAUUGUUCAUGCU-3ʹ (Merck). Several clones were genotyped by PCR (primers: TATCTTCATACATCACCACACCTGA and TTCTAAAACACATCAGAGATTGGGG) and Sanger sequencing. Validation experiments were performed with two homozygote (518/535) and two heterozygote (403/407) clonal lines. All other experiments were performed with b (...truncated)