Fig 2: DNA Damage Promotes ER-mitochondria Contacts through EI24 and VDAC2. a U2OS cells stably expressing CFP-ER were treated with cpt, doxo or eto for 12 h, and subjected to PLA. Scale bar, 10 µm. b Quantification of the number of PLA-positive puncta/cell in a. c Representative EM images of U2OS cells treated with eto for 12 h. The boxed areas at the top are magnified at the bottom. The arrowheads indicate the ER-mitochondria contacts. ER endoplasmic reticulum; M mitochondrion. Scale bar, 500 nm. d Quantification of the ratio (%) of ER-mitochondria contacts (<30 nm) to the mitochondrial surface area and the average number of close contacts per mitochondrion in c. e U2OS cells stably expressing CFP-ER were transfected with either control (shControl) or EI24 shRNA (shEI24), treated with DMSO or eto for 12 h, probed with anti-TOM20 and anti-GFP antibodies, and subjected to PLA. Scale bar, 10 µm. f Quantification of the number of PLA-positive puncta/cell in e. g Representative EM images of wild-type (WT) and EI24-knockout (KO) U2OS cells treated with eto for 12 h. The boxed areas at the top are magnified at the bottom. Arrowheads indicate the ER-mitochondria contacts. ER, endoplasmic reticulum; M, mitochondrion. Scale bar, 500 nm. h Quantification of the ratio (%) of ER-mitochondria contacts (<30 nm) to the mitochondrial surface area and the average number of close contacts per mitochondrion in g. i Representative images of PLA results of U2OS cells stably expressing CFP-ER and transfected with either control or VDAC2 shRNA (shVDAC2), treated with DMSO or cpt for 12 h. Scale bar, 10 µm. j Quantification of the number of PLA-positive puncta/cell in i. k U2OS cells that stably expressed CFP-ER and overexpressed the vector, 3× Flag-EI24, 3× Flag-EI243m, 3× Flag-EI24?H1 or 3× Flag-EI24?H2 were subjected to PLA. Scale bar, 10 µm. l Quantification of the number of PLA-positive puncta/cell in k

Fig 3: EI24 Promotes ER-mitochondria Ca2+ Transfer. a Measurement of the mito-R-GECO1.2 fluorescence intensity, which indicates mitochondrial Ca2+, after histamine stimulation in the control (Vector), 3× Flag-EI24 or 3× Flag-EI24?H1 overexpressing HeLa cells that stably expressed mito-R-GECO1.2. Traces show the mean ratio (?F/F) of the increased fluorescence intensity (?F) to the background fluorescence intensity (F) of the cells. Shown are the average traces from >20 cells in each group. b Quantification of the mito-R-GECO1.2 peak intensity in a. c Measurement of the mito-R-GECO1.2 fluorescence intensity, which indicates mitochondrial Ca2+, after histamine stimulation in stable mito-R-GECO1.2-expressing HeLa cells co-transfected with 3× Flag-EI24 and control (shControl) or VDAC2 shRNA (shVDAC2). d Quantification of the mito- R-GECO1.2 peak intensity in c. e Representative images of HeLa cells stably expressing mito-R-GECO1.2 and treated with DMSO or 1 µM cpt for 10 h. Scale bar, 10 µm. f Fluorescence intensity of HeLa cells stably expressing mito-R-GECO1.2 and treated with DMSO or 1 µM cpt, with or without 10 µM KB-R7943 or 1 µM XeC for 10 h. g, h Fluorescence intensity of HeLa cells stably expressing mito-R-GECO1.2, transfected with control, EI24 (shEI24, g), or VDAC1/2 shRNAs (h), and treated with DMSO or 1 µM cpt for 10 h. i–k Western blot analysis of HEK293 cells co-treated with 50 nM cpt, 100 nM doxo or 1 µM eto and and 10 µM KB-R7943 (i),1 µM Xec (j) or 5 µM 2-APB (k) for 12 h. GAPDH served as a loading control. l Quantification of flow cytometry analysis for cells positive for Annexin V. Cells were treated with 100 nM cpt and 10 µM KB-R7943 for 16 h. m–o Western blot analysis of HEK293 cells transfected with either the vector or 3× Flag-EI24 and treated with 50 nM cpt, and 10 µM KB-R7943 (m),1 µM Xec (n) or 5 µM 2-APB (o) for 12 h

Fig 4: EI24 Promotes ER-mitochondria Contacts by VDAC2. a Western blot (WB) analysis of subcellular fractions from mouse liver. H homogenate; Mc crude mitochondrial fraction; Mp pure mitochondrial fraction; ER endoplasmic reticulum; MAM mitochondria-associated membrane; C cytosol. RTN4 Reticulon 4; ATL3 Atlastin 3; Mfn2 mitofusin-2. b U2OS cells that stably expressed CFP-ER and overexpressed the vector, 3× Flag-EI24 or 3× Flag-EI24?C were subjected to PLA (red) with anti-TOM20 and anti-GFP antibodies. DNA was stained with DAPI (blue). Scale bar, 10 µm. c Quantification of the number of PLA-positive puncta/cell in b. d Representative EM images of U2OS cells transfected with the vector, 3× Flag-EI24 or 3× Flag-EI24?C. Boxed areas at the top are magnified at the bottom, and the arrowheads indicate the ER-mitochondria contacts (<30 nm). ER endoplasmic reticulum; M mitochondrion. Scale bar, 500 nm. e Quantification of the ratio (%) of ER-mitochondria contacts (<30 nm) to the mitochondrial surface area (left) and the average number of close contacts per mitochondrion (right) in d (n = 118–311 mitochondria). f U2OS cells that stably expressed CFP-ER were co-transfected with 3× Flag-EI24 and control (shControl) or VDAC2 shRNA (shVDAC2) and subjected to PLA. Scale bar, 10 µm. g Quantification of the number of PLA-positive puncta/cell in f

Fig 5: DNA-PKcs interacts with ANT2 and VDAC2 at mitochondria HeLa cells expressing Flag-tagged ANT1-3 were subjected to Co-IP using anti-Flag antibody, then analyzed by western blot as indicated.Co-IP was performed with mitochondrial fraction. Co-IP of endogenous ANT2 with anti-DNA-PKcs antibody (clone 25–4), but not control IgG.Proximity ligation assay (PLA) analysis showed that DNA-PKcs interacts with Flag-ANT2, but not Flag-ANT1 or ANT3.PLA analyses were performed by using anti-DNA-PKcs and anti-ANT2 antibodies. Results showed interactions between endogenous ANT2 and DNA-PKcs in human skin fibroblasts (HSF), HeLa cells, human HCT116 cells, and human M059K cells. No PLA signal was detected in DNA-PKcs–deficient HCT116 PKcs-/- and M059J cells.HeLa cells expressing Flag-tagged VDAC1-3 were subjected to Co-IP using anti-Flag antibody and analyzed by western blot as indicated.PLA analysis show that the interaction between DNA-PKcs and Flag-ANT2 was disrupted by knockdown of VDAC2, but not VDAC1 or VDAC3. Quantification shown on the right. ***P < 0.001. Data information: The bar graphs were generated from three independent biological replicates analyses; data are represented as mean ± SD. Statistical significance was established using a two-way analysis of variance (ANOVA, *P < 0.05; **P < 0.01; ***P < 0.001).Co-IP confirmed that knockdown of VDAC2 by siRNA interrupts the interaction between Flag-ANT2 and DNA-PKcs. Source data are available online for this figure.