Deoxycytidine kinase augments ATM-Mediated DNA repair and contributes to radiation resistance

Abstract

Efficient and adequate generation of deoxyribonucleotides is critical to successful DNA repair. We show that ataxia telangiectasia mutated (ATM) integrates the DNA damage response with DNA metabolism by regulating the salvage of deoxyribonucleosides. Specifically, ATM phosphorylates and activates deoxycytidine kinase (dCK) at serine 74 in response to ionizing radiation (IR). Activation of dCK shifts its substrate specificity toward deoxycytidine, increases intracellular dCTP pools post IR, and enhances the rate of DNA repair. Mutation of a single serine 74 residue has profound effects on murine T and B lymphocyte development, suggesting that post-translational regulation of dCK may be important in maintaining genomic stability during hematopoiesis. Using [(18)F]-FAC, a dCK-specific positron emission tomography (PET) probe, we visualized and quantified dCK activation in tumor xenografts after IR, indicating that dCK activation could serve as a biomarker for ATM function and DNA damage response in vivo. In addition, dCK-deficient leukemia cell lines and murine embryonic fibroblasts exhibited increased sensitivity to IR, indicating that pharmacologic inhibition of dCK may be an effective radiosensitization strategy.

Figure 1. dCK is activated in an ATM-dependent manner after IR.

(A) In vitro cell uptake assay of [³H]-dC 2 hours after exposure of L1210-10K ± dCK (WT) to 3 Gy (*, P = 0.0002; N = 3) or 100 µM H₂0₂ (*, P = 0.0197; N = 3). (B) In vitro dCK kinase assay using L1210-10K ± dCK (WT) cell lysates, [³H]-dC as substrate and performed at indicated times after exposure to 3 Gy (*, P = 0.0002; N = 3) or 100 µM H₂0₂ (*, P<0.0001; N = 3). (C) Western blot of L1210-10K ± dCK (WT) cell lysates obtained at indicated times after 3 Gy irradiation. (D) Western blot of lysates from 10K+dCK (WT) cells treated for 1 hour with either DMSO vehicle, 10 µM ATM inhibitor (Ku55933) or 10 µM DNA-PKcs inhibitor (Nu7441) and irradiated with 3 Gy. (E) Western blot of lysates from CHOC6 (WT LCL) and DK0064 (ATR-defective) cell lines before IR or 2 hours after 3 Gy exposure. Lysates from DK0064 cells pretreated with 10 Gy exposure. Lysates from DK0064 cells pretreated with 10 µM ATM inhibitor Ku55933 are shown in the right panel. (F) In vitro dCK kinase assay using 10K+dCK (WT), CHOC6 and DK0064 cell lysates, [³H]-dC as substrate and performed where indicated after 1 hour pretreatment with inhibitors of ATM and DNA hour pretreatment with inhibitors of ATM and DNA-PKcs, and 2 hours after exposure to 3 Gy (*, 10K+dCK: P<0.0001; CHOC6: P = 0.001; DK0064: P = 0.0003; N = 3).

Figure 2. PET imaging of metabolic regulation during DNA damage response.

(A) Western blot of 10K ± dCK (WT, S74A, S74E) before or 2 hours after exposure to 3 Gy. (B) In vitro cell uptake assay of [³H]-dC after 3 Gy exposure of L1210 Gy exposure of L1210-10K ± dCK WT (*, P = 0.0004; N = 14) or dCK Ser⁷⁴ mutants. (C) In vitro dCK kinase assay using [³H]-dC and cell lysates from10K ± dCK WT (*, P<0.0001; N = 9), dCK S74A or dCK S74E, 2 hours after 3 Gy irradiation. (D) [¹⁸F]-FAC and [¹⁸F]-FDG microPET/CT scans of NOD-SCID mice with bilateral 10K ± dCK(WT, S74A, S74E) tumors after 3 Gy irradiation of right tumor. (E) Averaged ROI values of [¹⁸F]-FAC (top panel) or [¹⁸F]-FDG (bottom panel) uptake in irradiated (R, right) and untreated (L, left) tumors (*, P = 0.0267; N = 4 mice per group mice per group).

Figure 3. Changes in dCK substrate specificity and dCTP pool after IR.

(A) Western blot of purified 6-His-tagged dCK (WT or Ser⁷⁴ mutants) from 10K cells, and corresponding total cell lysates. (B) In vitro kinase assay on purified 6-His-tagged dCK (WT, S74A, S74E) using either [³H]-dC, [³H]-dG or [³H]-dA as a substrate before or after 3 Gy exposure (*, P = 0.0025; N = 2). (C) Intracellular dCTP pools before and hourly after 3 Gy exposure of 10K ± dCK (WT, S74A, S74E) cells (*, 10K vs. WT P = 0.0006; 10K vs. S74E P = 0.032; N = 6 for each time point).

Figure 4. dCK influences the rate of IR-induced DNA repair.

(A) Fluorescent images of DAPI and anti-pS¹³⁹ γH2A.X stained 10K ± dCK (WT) cells before or 0.5 and 5 hours after exposure to 3 Gy. (B) Number of γH2A.X foci per cell nucleus (*, P<0.0001; N = 10) calculated from (A). (C) FACS analysis of pS¹³⁹ γH2A.X positive 10K cells (indicated as % of total cell number) with or without dCK (WT, S74A, S74E) after 3 Gy irradiation. (D) Percentage of γH2A.X positive cells remaining over time after 3 Gy exposure (*, P = 0.01; N = 3 per time point per time point), obtained as shown in (C). (E) Recombination efficiency (RE) of digested plasmid via homologous recombination (HR) and non-homologous end joining (NHEJ) 24 hours after nucleofection of 3 Gy irradiated 10K ± dCK (WT) cells divided by the RE of the untreated cells (*, P = 0.0078, N = 10).

Figure 5. dCK absence sensitizes 10K cells and MEFs to IR.

(A) Clonogenic survival assay of MEFs from dCK KO and dCK WT mice (*, P = 0.0383, N = 2). (B) Schematic of in vivo vivo IR treatment schedule (FDG refers to PET imaging). (C) Volume measurements of in vivo vivo untreated control and irradiated tumor model based on (L×W²)/2 formula (N = 4 mice per group). (D) [¹⁸F]-FDG microPET/CT scans of CB17 SCID mice treated as shown in (B) or untreated control. LN – lymph node. (E) Quantification of [¹⁸F]-FDG uptake in irradiated and untreated 10K ± dCK in vivo vivo tumor model (*, 10K ± IR at day 4: P = 0.0181; irradiated 10K vs. 10K+dCK at day 4: P = 0.0272; N = 4 mice per group).

Figure 6. DCK Ser⁷⁴ influences T cell development.

(A) Western blot of CD45.2⁺ dCK KO cells (YFP⁻) carrying one of three dCK isoforms (WT, S74A, S74E; YFP⁺). Cells were isolated from recipient spleens and dCK expression level compared to endogenous dCK of CD45.1⁺ B6.SJL recipient. (B) In vitro dCK kinase assay using [³H]-dC and lysates of dCK KO donor CD45.2⁺ (dCK⁺YFP⁺ or dCK⁻YFP⁻) cells FACS purified from spleens of B6.SJL recipients (N = 3 mice per group). (C) Cellularity of B6.SJL thymi repopulated with dCK KO donor cells carrying different dCK isoforms (CD45.2⁺YFP⁺; N = 3). (D) FACS analysis of dCK KO ± dCK (WT, S74A, S74E) thymocytes. Bottom panels show pSer¹³⁹ γH2A.X stained DN3 thymocytes. (E-F) Averaged percentages of single live donor CD45.2⁺ thymocytes (± dCK/YFP) at different stages of development (N = 3). (G) Averaged percentages of DN3 thymocytes positive for pSer¹³⁹ γH2A.X (N = 3).

Figure 7. DCK Ser⁷⁴ influences B cell development.

(A) FACS analysis of CD45.2⁺ donor dCK KO ± YFP/dCK (WT, S74A, S74E) B cell development in the bone marrow of B6.SJL recipients. FACS plots are representative examples. Bottom panels show γH2A.X stained B–C Hardy fractions. (B) Averaged percentages of single live donor CD45.2⁺B220⁺IgM⁻ (± dCK/YFP) B cells at different stages of development (N = 3). (C) Averaged percentages of Hardy B–C fraction B cells positive for pSer¹³⁹ γH2A.X (N = 3).