Fig 2: KDM2A modulates genome stability upon IR(A-i) KDM2A knockdown sensitizes cells to IR leading to micronuclei formation. U2OS cells transfected with siCTL or siKDM2A were irradiated (2Gy) and fixed 24 hrs later. Cells were stained with DAPI and micronuclei observed by fluorescence microscopy. The percentage of cells harboring micronuclei before and after IR was quantified in each condition (right panel). Experiments were performed in triplicates. Asterisk indicates p<0.05, two-tailed paired Student t-test. Error bars indicate standard deviation. (A-ii) 53BP1 forms foci in micronuclei found in KDM2A knockdown cells. Cells transfected with siKDM2A were irradiated (2Gy) and, 24 hrs later, subjected to immunofluorescence with anti-53BP1(green) and DAPI (blue). (B) The zinc finger domain of KDM2A is crucial for the maintenance of genome stability. Endogenous KDM2A was downregulated in U2OS cells stably expressing wild type MYC-tagged KDM2A-WT, -DD or -ΔZF mutants by transfection with siUTR. Cells were irradiated (2Gy) or left untreated and the percentage of cells showing micronuclei was determined 24 hrs after IR treatment. Single (*) and double (**) asterisks indicate p<0.05, two-tailed paired Student t-test. Error bars indicate standard deviation. Experiments were performed in triplicates. (C) 53BP1 localization in mitotic cells upon IR. U2OS cells transfected with siCTL or siUTR and, stably expressing or not, wild type MYC-KDM2A (WT), DD or ΔZF mutants were irradiated (2Gy) and fixed 24 hrs later. Cells were subjected to immunofluorescence with anti-53BP1 (green), anti-Tubulin (red) and DAPI (blue).

Fig 3: Recruitment of KDM2A and 53BP1 to DSBs depends on its demethylase activity and intact zinc finger domain(A) Endogenous KDM2A is recruited to DSBs. Detectable KDM2A at the FokI-induced breaks upon CSK treatment. U2OS-DSB-reporter cells were treated or not with Shield-1 and 4-OHT to induce expression of ER-mCherry-LacI-FokI-DD (FokI) (red). Cells were immunostained with anti-γH2AX, anti-53BP1 (green, upper panel) and anti-KDM2A (green, lower panel). Immunofluorescence analyses were conducted using cells that were pre-extracted with CSK buffer. (B) The demethylase activity and zinc finger domain of KDM2A are necessary for its recruitment to DSBs. U2OS-DSB-reporter cells were transfected with siCTL or siUTR and plasmids encoding for MYC-tagged variants of KDM2A. Following transfection, cells were treated with Shield-1 and 4-OHT to induce expression of ER-mCherry-LacI-FokI-DD (FokI) (red). Immunofluorescence analyses were conducted with anti-MYC (green) and anti-γH2AX. (C) Focal accumulation of 53BP1 at DSBs is modulated by the demethylase activity and zinc finger domain of KDM2A. U2OS-DSB-reporter cells were transfected with siCTL or siUTR and plasmids encoding for MYC-tagged variants of KDM2A. Cells were treated with Shield-1 and 4-OHT to induce expression of ER-mCherry-LacI-FokI-DD (FokI) (red). Immunofluorescence analyses were conducted with anti-MYC and anti-53BP1 (green). The percentage of cells displaying colocalization between 53BP1 and FokI fusion protein was quantified in each condition (right panel). Experiments were replicated four times. Single (*), double (**) and triple (***) asterisks indicate p<0.05, two-tailed paired Student t-test. Error bars indicate standard deviation.

Fig 4: KDM2A modulates the recruitment of 53BP1 to DSBs and entry into mitosis upon IR(A) KDM2A modulates 53BP1 accumulation at DSBs generated by irradiation. U2OS cells transfected with control siRNA (siCTL) or siRNA targeting the 3’-UTR region of the KDM2A mRNA (siUTR) and stably expressing MYC-tagged KDM2A variants were irradiated (2Gy), fixed 1h later and subjected to immunofluorescence analysis with DAPI, anti-53BP1 (green) and anti-MYC (red). Quantification of immunofluorescence analyses (right panel) shows the average number of 53BP1 IRIF per cell in each condition. Experiments were performed in triplicate. Single (*), double (**) and triple (***) asterisks indicate p<0.05, two-tailed paired Student t-test. Error bars indicate standard deviation. (B) KDM2A depletion impairs 53BP1 recruitment to DNA breaks generated by FokI endonuclease. U2OS-DSB-reporter cells transfected with siCTL or siKDM2A were treated with Shield-1 and 4-OHT to promote expression of mCherry-LacI-FokI (FokI) (red). Immunofluorescence analyses were conducted with anti-53BP1(green) and anti-KDM2A. Quantification of immunofluorescence analyses (right panel) shows the percentage of cells displaying colocalization between 53BP1 and FokI. Asterisk indicates p<0.05, two-tailed paired Student t-test. Error bars indicate standard deviation. (C) KDM2A depletion results in faster progression into mitosis without proper repair of damaged DNA. U2OS cells transfected with siCTL or siKDM2A were left untreated or irradiated (2Gy) and harvested at the indicated time points after irradiation. The percentage of mitotic cells, as indicated by H3pSer10 staining, was determined by flow cytometry. Data shown is representative of experiments replicated three times.

Fig 5: KDM2A associates with 53BP1(A-i) KDM2A interacts with endogenous 53BP1. HEK293T extracts from cells transfected with an empty vector or FLAG-tagged KDM2A were subjected to immunoprecipitation with anti-FLAG beads. Interaction between KDM2A and 53BP1 was observed by immunoblotting with anti-53BP1. (A-ii) Deletion of the zinc finger domain (564-610) of KDM2A disrupts its interaction with 53BP1. Schematic of KDM2A protein structure (top panel). FLAG-tagged 53BP1 co-expressed with wild type MYC-tagged KDM2A (WT), demethylase-deficient KDM2A (DD) or KDM2A lacking the zinc finger domain (ΔZF) was immunoprecipitated with anti-FLAG beads and immunoblotted with anti-MYC and anti-FLAG. (B) The region comprising residues 1052-1302 of 53BP1 is necessary for its interaction with KDM2A. Schematic of full lenght 53BP1 and mutants (right panel). Lysates from cells expressing HA-tagged 53BP1 mutants and MYC-KDM2A-WT were immunoprecipitated with anti-HA and immunoblotted with anti-HA and anti-MYC. (C) 53BP1 harbors methylated lysine residues. Lysates from cells expressing FLAG-53BP1, co-expressing or not MYC-tagged KDM2A variants, were immunoprecipitated with anti-FLAG beads and immunoblotted with anti-MYC, anti-FLAG and pan-dimethyl lysine antibody (anti-Kme2).