The Impact of NAD Bioavailability on DNA Double-Strand Break Repair Capacity in Human Dermal Fibroblasts after Ionizing Radiation

Nicotinamide adenine dinucleotide (NAD) serves as a substrate for protein deacetylases sirtuins and poly(ADP-ribose) polymerases, which are involved in the regulation of DNA double-strand break (DSB) repair molecular machinery by various mechanisms. However, the impact of NAD bioavailability on DSB repair remains poorly characterized. Herein, using immunocytochemical analysis of γH2AX, a marker for DSB, we investigated the effect of the pharmacological modulation of NAD levels on DSB repair capacity in human dermal fibroblasts exposed to moderate doses of ionizing radiation (IR). We demonstrated that NAD boosting with nicotinamide riboside did not affect the efficiency of DSB elimination after the exposure of cells to IR at 1 Gy. Moreover, even after irradiation at 5 Gy, we did not observe any decrease in intracellular NAD content. We also showed that, when the NAD pool was almost completely depleted by inhibition of its biosynthesis from nicotinamide, cells were still able to eliminate IR-induced DSB, though the activation of ATM kinase, its colocalization with γH2AX and DSB repair capacity were reduced in comparison to cells with normal NAD levels. Our results suggest that NAD-dependent processes, such as protein deacetylation and ADP-ribosylation, are important but not indispensable for DSB repair induced by moderate doses of IR.


Introduction
Among various DNA lesions, double-strand breaks (DSB) are the most dangerous for the cell because they can lead to chromosome rearrangements, oncogenic transformation or cell death if not repaired [1]. DSBs are repaired by two major pathways: homologous recombination (HR) and non-homologous end joining (NHEJ) [2,3]. The key steps of HR include the generation of 3' single strands by resection of broken DNA ends with nucleases and RAD51-dependent invasion of 3' single-stranded DNA into a homologous duplex that is available only during S and G2 phases of the cell cycle [2]. In mammalian cells, NHEJ is the major pathway of DSB repair, acting by direct sealing of DSB ends in all phases of the cell cycle. NHEJ is mediated by DSB sensing heterodimer Ku70/Ku80, which recruits and activates the catalytic subunit of DNA-dependent protein kinase (DNA-PKcs) at DSB sites followed by ligation using DNA ligase IV/XRCC4/XLF complex [3]. The central player of DNA damage response to DSB is ataxia-telangiectasia mutated kinase (ATM). ATM is rapidly recruited to the site of DSB through interaction with DSB recognition complex MRE11/RAD50/NBS1 (MRN), which results in ATM activation by autophosphorylation at Ser1981 [4,5]. Activated ATM phosphorylates multiple downstream targets promoting DNA repair and activating cell-cycle arrest [6]. One of the earliest events of cellular response to DSB is ATM-dependent phosphorylation of histone H2AX at serine 139 (referred to as γH2AX) at the sites of DNA damage [7,8]. In addition to ATM, with and stabilizes DNA-PK at DSB sites [46] and recruits chromatin remodeler SNF2H, increasing the accessibility of DNA repair factors to damaged DNA [37]. In contrast to SIRT1 and SIRT6, SIRT7 is recruited to the sites of DSB with slow kinetics in a PARP1dependent manner [47]. SIRT7-dependent deacetylation of H3K18ac has been shown to be important for NHEJ promotion [48]. SIRT7 also deacetylates ATM kinase at K3016, and this step is a prerequisite for ATM dephosphorylation and deactivation at the late stage of DSB repair [49].
Protein deacetylation and ADP-ribosylation, implemented by sirtuins and PARPs in response to DSB induction, are supposed to be controlled by intracellular NAD levels since these reactions are accompanied by the cleavage of NAD to nicotinamide (Nam) and ADP-ribose. However, the impact of NAD bioavailability on DSB repair remains poorly characterized.
In this study, using γH2AX foci quantification as an indirect detection of DSB, we have examined how pharmacological modulation of NAD levels affects the efficiency of DSB repair in human dermal fibroblasts (HDF) after exposure to ionizing radiation (IR). We have demonstrated that stimulation of NAD biosynthesis by nicotinamide riboside (NR) significantly increases the level of intracellular NAD, but this does not influence the DSB repair capacity in HDF after exposure to moderate doses of IR. Moreover, we have not observed any depletion of the NAD pool during the DNA damage response induced by IR at a dose of 1 or 5 Gy. We also have shown that critical depletion of the NAD in HDF by inhibition of NAD biosynthesis from Nam impairs the IR-induced activation of ATM kinase and its colocalization with γH2AX and decreases DSB repair capacity.

Cell Culture and Ionizing Radiation Treatment of Cells
Human dermal fibroblasts (HDF) (purchased from Pokrovsky Stem Cell Bank, St. Petersburg, Russia) were cultured in Minimum Essential Medium (MEM) supplemented with 10% fetal bovine serum (FBS), 100 U/mL penicillin, 100 µg/mL streptomycin, and 2 mM glutamine. The cells were cultured at 37 • C in a humidified atmosphere of 5% CO 2 . NR (150 µM) or FK866 (2 µM) were added to the culture medium as indicated. Cells were exposed to ionizing radiation (IR) at 1 or 5 Gy using X-ray irradiator RAP-150/300-14 (Promrentgen, Moscow, Russia) with a Cu/Al filter.

Flow Cytometry
In total, 2.5 × 10 5 HDF cells were plated in 6-well plates. After 1, 2, 3, 4, and 7 days, cells were trypsinized and stained with 50 µg/mL propidium iodide for the estimation of the number of dead and viable cells. Flow cytometry was performed using a CytoFLEX in-

NAD Quantification
Intracellular NAD content was determined by a colorimetric enzymatic assay using a NAD/NADH Quantification Kit (Sigma-Aldrich, St. Louis, MO, USA, Cat. No. MAK037). A total of 2.5-3.0 × 10 5 cells/well were seeded in 12-well plates. The next day, the cells were lysed directly in wells with 200 µL extraction buffer provided in NAD/NADH Kit, and further procedures were performed according to the manufacturer's protocol. NAD concentration was obtained by normalizing the measured NAD content to the protein amount in the sample and was expressed as pmol/µg protein. Protein concentration was determined using Pierce TM BCA Protein Assay Kit (Thermo Scientific, Rockford, IL, USA).

MTT-Assay
Cell metabolic activity was determined using a 3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide (MTT) assay kit (Rosmedbio, St Petersburg, Russia). The cells were seeded in 24-well plates (0.3 × 10 5 cells/well) and incubated for 1-7 days in a CO 2 incubator in a humidified atmosphere with 5% CO 2 at 37 • C. An MTT solution in PBS was added to the growth medium with a final concentration of 0.5 mg/mL, and cells were incubated for 3 h in a CO 2 incubator. After that, the growth medium was discarded, and 1 mL DMSO was added to each well. The plate was incubated with shaking at 42 • C for 30 min until the complete solubilization of purple formazan crystals. Spectrophotometric absorbance of samples was measured at 570 nm wavelength using Multiskan TM FC Microplate Photometer (Thermo Fisher Scientific, Waltham, MA, USA).

Immunostaining of γH2AX and Ki-67 Combined with the Detection of S-Phases after EdU Incorporation
Cells were grown on 18 × 18 mm glass coverslips placed in Petri dishes. After IR, the cells were incubated for 1-24 h in a growth medium. In addition, 5-ethynyl-2 -deoxyuridine (EdU) (10 µM) was added 20 min before fixation for labeling of S-phase cells. The cells were fixed with 4% formaldehyde at +4 • C, rinsed with PBS, permeabilized with 0.5% Triton X-100 in PBS for 15 min with shaking, rinsed with 3% BSA in PBS, and then, coppercatalyzed click reaction of EdU with Alexa Fluor 647 picolyl azide was performed in accordance with the Click-iT Plus EdU imaging kit manufacturer's recommendations. After blocking in 1% Blocking Reagent (Roche, Mannheim, Germany) in PBS with 0.02% Tween 20 for 30 min at 37 • C, immunofluorescence staining for detection of γH2AX, the marker of DSB repair, and Ki-67, the marker of cell proliferation [50,51], was performed. The dilution of antibodies was carried out in 0.5% Blocking Reagent in PBS with the addition of 0.02% Tween 20. All incubations with antibodies were performed at 37 • C. Between subsequent incubations, slides were washed with shaking for 30 min in PBS supplemented with 0.1% Tween 20. Cells were incubated for 1 h with the following primary antibodies: rabbit anti-γH2AX (1:100) combined with monoclonal mouse anti-Ki-67 (1:50), and 40 min with secondary antibodies: Alexa Fluor 488-conjugated goat anti-rabbit IgG and Alexa Fluor 568-conjugated goat anti-mouse IgG (1:400). DNA was counterstained for 10 min at room temperature with 0.5 µg/mL 4 ,6-diamidino-2-phenylindole (DAPI) (Sigma-Aldrich, St. Louis, USA) in PBS, and samples were mounted in a Citifluor antifade solution AF1 (Science Services, Munich, Germany).

Immunostaining of γH2AX and Phospho-ATM Combined with the Detection of S-Phases after EdU Incorporation
For γH2AX/phospho-ATM (S1981) (pATM) double immunostaining, Ki-67 marker was not used. Irradiated cells were incubated with EdU (10 µM) 20 min before fixation. Cell fixation, permeabilization, washing, blocking, detection of EdU with Alexa Fluor 647 picolyl azide, and other protocol details were the same as in Section 2.6. After EdU detection, cells were stained with primary rabbit anti-γH2AX antibody (1:100) and mouse monoclonal anti-phospho-ATM (S1981) antibody (1:100) and 40 min with secondary antibodies: Alexa Fluor 488-conjugated goat anti-mouse IgG and Alexa Fluor 568-conjugated goat anti-rabbit IgG (1:400). DNA was counterstained with DAPI in PBS, and samples were mounted in the Citifluor antifade solution.

Quantification of γH2AX and pATM Foci
Quantification of γH2AX foci after IR was performed only in G0 and G1 cells in images of HDF cells after Ki-67 immunostaining and EdU detection. Ki-67 labels cells in G1, S, G2, and M phases, and only resting (G0) cells are Ki-67 negative. Ki-67 positive EdU-incorporated S-phase cells were excluded from the analysis. G2 cells were excluded due to the increased size of the nucleus and the specific pattern of Ki-67 nuclear distribution (1 or 2 large and bright nucleoli). For γH2AX/pATM double immunostaining, EdUincorporated S-phase cells were excluded from analysis, and G2 cells were excluded due to the increased size of the nucleus. For quantification of γH2AX and pATM foci number and the total projected area of foci per nucleus (referred to as "the total area of foci per nucleus" in the text), the IPLab v3.65 program (Scanalytics, Inc., Vernon, WI, USA) was used. Green channel images were converted to a grey scale, and the "Segmentation function" was applied to discriminate the foci from the background noise. The same thresholds of segmentation were applied in each series of experiments. The total area of foci per nucleus was measured in pixels.

Colocalization Analysis of γH2AX and pATM Foci
The images of central confocal sections of the nuclei captured at zoom 9 were taken for colocalization analysis of γH2AX and pATM foci. The analysis of colocalization was performed using the ImageJ 1.43 program (NIH, Bethesda, MD, USA). Image segmentation based on the "Difference of Gaussians" approach was performed using the GDSC Image J plugin in green (pATM) and red (γH2AX) channels separately for the elimination of background noise. Colocalization of segmented images of γH2AX and pATM foci was quantified by Pearson's correlation using the "Manders' coefficients" plugin as described in [52]. Pearson's correlation coefficient (PCC) is a commonly used parameter for measuring the extent of foci overlap in image pairs. PCC is not sensitive to differences in pixel intensities in two different color channels and can vary in the range from −1 to +1. The more PCC, the higher extent of foci overlap (colocalization) [53]. PCC was obtained for 25 cells in each series of experiments.

Statistical Analysis
Statistical analysis was performed using SigmaPlot 12.5 (Systat Software Inc., San Jose, CA, USA). Differences between groups were analyzed using one-(or two-)way ANOVA with Tukey post hoc test. p-values < 0.05 were considered to be significant.

Modulation of NAD Biosynthesis in Human Dermal Fibroblasts
To assess how changes in intracellular NAD levels can affect the efficiency of DNA DSB repair, we optimized the conditions for stimulation and suppression of NAD synthesis in human dermal fibroblasts (HDF). Cells cultured in the standard medium can synthesize NAD from the pyridine base Nam, a form of vitamin B3, via the salvage pathway. Nam is converted by the Nam phosphoribosyltransferase (NAMPT) to the Nam mononucleotide (NMN), which, in turn, is adenylated by NMN adenylyltransferases (NMNAT) to form NAD ( Figure S1) [20,54]. For the suppression of NAD synthesis from Nam, we used FK866, a highly specific noncompetitive inhibitor of NAMPT [55]. The cells were cultured in the growth medium supplemented with 10% FBS in the presence of FK866 at a concentration of 2 µM during various periods of time. We also assessed cell viability by their metabolic activity, which we estimated by MTT assay, and cell proliferation, which was assessed by flow cytometry as described in Materials and Methods. One day after the addition of FK866 to HDF cells, the NAD level was about 45% of that evaluated in control cells ( Figure 1A). At the same time, both metabolic activity ( Figure 1B) and cell proliferation ( Figure 1C) remained at the control level. During subsequent incubation with FK866, the concentration of intracellular NAD continued to fall, which was accompanied by suppression of both metabolic activity and cell proliferation ( Figure 1A-C). After 4 days of incubation with FK866, the NAD level was nearly undetectable ( Figure 1A), while the cells were still viable ( Figure 1B,C). Seven days of treatment with NAMPT inhibitor resulted in cell death ( Figure 1B,C).
It is important to note, that, besides Nam, another potential NAD precursor, tryptophan (at a concentration of 80 µM), is present in the growth medium. NAD can be synthesized from this precursor via the de novo pathway [20,54] (Figure S1). Our data indicate that the de novo NAD synthesis in HDF is inactive or insufficiently active to maintain the physiological level of NAD when the synthesis of this dinucleotide from Nam is inhibited.
OR PEER REVIEW 7 of 17  . Statistical analysis of differences the groups was carried out by one-way ANOVA with post hoc comparisons using Tuke indicates statistically significant difference at p < 0.001, * indicates statistically significant d at p < 0.05, and ns indicates that statistical difference is not significant. CTRL marks untre trol cells.
It is important to note, that, besides Nam, another potential NAD precurs tophan (at a concentration of 80 μM), is present in the growth medium. NAD synthesized from this precursor via the de novo pathway [20,54] (Figure S1). O indicate that the de novo NAD synthesis in HDF is inactive or insufficiently maintain the physiological level of NAD when the synthesis of this dinucleot Nam is inhibited.
Furthermore, in order to increase the concentration of intracellular NAD, NR, the nucleoside form of vitamin B3. NR stimulates an alternative NAD bios pathway via its phosphorylation by NR kinases (NRK) to yield NMN [56]  (C) Cell proliferation was estimated by flow cytometry. Data are presented as mean ± standard deviation (SD) (n = 3). Statistical analysis of differences between the groups was carried out by one-way ANOVA with post hoc comparisons using Tukey test. *** indicates statistically significant difference at p < 0.001, * indicates statistically significant difference at p < 0.05, and ns indicates that statistical difference is not significant. CTRL marks untreated control cells.
Furthermore, in order to increase the concentration of intracellular NAD, we used NR, the nucleoside form of vitamin B3. NR stimulates an alternative NAD biosynthesis pathway via its phosphorylation by NR kinases (NRK) to yield NMN [56] (Figure S1), and is currently one of the most widely used NAD boosting agents in various cellular and animal experimental models [57,58]. HDF were cultured in the presence of NR (at a concentration of 150 µM) for 24 h, then the concentration of NAD in the cell extract was determined. Stimulation of NAD synthesis by NR resulted in a significant increase (more than 40%) in NAD level compared to control cells ( Figure 1D). At the same time, the presence of NR in the growth medium had no effect on the metabolic activity of irradiated cells ( Figure 1E).

Stimulation of NAD Biosynthesis Does Not Affect γH2AX Foci Formation and Elimination in IR-Exposed HDF
Furthermore, we optimized the conditions for induction and detection of DSB in HDF after IR treatment. IR-induced DNA DSB were indirectly monitored using immunocytochemical analysis of histone variant H2AX phosphorylated at Ser139 (γH2AX), the marker of DSB [7]. The cells cultured under standard conditions were exposed to IR at a moderate dose of 1 Gy. Next, control and irradiated cells were fixed, and γH2AX immunofluorescence staining was performed (Figure 2A). γH2AX foci can be detected in the nuclei of S-phase unirradiated cells due to occasional formation of DSB during normal replication [59]. Therefore, to exclude replicating cells from the analysis, we used DNA labeling with a thymidine analogue, 5-ethynyl-2 -deoxyuridine (EdU), which was added to the growth medium 20 min before fixation ( Figure S2). G2 cells were also excluded from analysis due to the increased size of the nucleus and specific Ki-67 staining ( Figure S2). Figure 2B,C shows the results of quantitative analysis of γH2AX foci in the nuclei of control and irradiated cells in G0 and G1 phases of the cell cycle, 1 h, 3 h, and 6 h after IR exposure. About 30-35 γH2AX foci per nucleus were detected 1 h after IR ( Figure 2B). During subsequent incubation, a decrease in the number and the total area of foci per nucleus was observed, which indicates the effective DNA DSB repair in HDF exposed to IR at a dose of 1 Gy. NR at a concentration of 150 µM was added to the cell growth medium to find out how the stimulation of NAD biosynthesis could modulate the capacity of DNA DSB repair. Overall, 24 h after the addition of NR, the cells were irradiated at a dose of 1 Gy and incubated in the growth medium containing NR (at the same concentration). In NR-treated cells, both the number ( Figure 2B) and the total area of foci per nucleus ( Figure 2C) remained at the level of NR-untreated irradiated cells 1 h, 3 h, and 6 h after IR treatment. Thus, the increase in intracellular NAD level had no effect on the efficiency of formation and elimination of γH2AX foci, i.e., it did not influence DNA DSB repair capacity after IR treatment at a dose of 1 Gy.
Cells 2023, 12, x FOR PEER REVIEW 9 of 17 Figure 2. The effect of the stimulation of NAD biosynthesis on γH2AX foci formation and elimination in IR-exposed HDF. HDF were treated with ionizing radiation (IR) at a dose of 1 Gy. Cells were fixed at different time points after IR (as indicated) and stained for γH2AX. (A) γH2AX foci 1 h after IR visualized by immunofluorescence (green). Cell nuclei were counterstained with DAPI (blue). Scale bar, 30 μm. (B,C) Cells were pretreated with NR for 24 h before IR. NR was also present in the culture medium after IR. γH2AX foci were quantified 1 h, 3 h, and 6 h after IR. The number (B) and relative total area (C) of γH2AX foci per nucleus. The total area of γH2AX foci induced in cells untreated with NR, 1 h after IR, was taken as 100%. Data are presented as mean ± standard error (SE) (n = 200). Statistical analysis of differences between the groups was carried out by one-way ANOVA with post hoc comparisons using Tukey test. *** indicates statistically significant difference at p < 0.001. ns indicates that statistical difference is not significant. CTRL marks unirradiated control cells.

IR Does Not Affect NAD Levels in HDF
To elucidate how the exposure to IR can influence the cellular NAD content, we treated HDF with IR at a dose of 1 Gy or 5 Gy and measured NAD level in cell extracts at various time intervals. NAD concentration in control cells was about 8 pmol/μg and remained unchanged at 0.5, 2, 4, and 24 h after IR at both doses ( Figure 3A). Furthermore, differences in metabolic activities of NAD(P)H-dependent dehydrogenases estimated by MTT assay were indistinguishable between irradiated and control cells (Figures 3B and  S3). Thus, IR at moderate doses does not induce any depletion of NAD in HDF. The effect of the stimulation of NAD biosynthesis on γH2AX foci formation and elimination in IR-exposed HDF. HDF were treated with ionizing radiation (IR) at a dose of 1 Gy. Cells were fixed at different time points after IR (as indicated) and stained for γH2AX. (A) γH2AX foci 1 h after IR visualized by immunofluorescence (green). Cell nuclei were counterstained with DAPI (blue). Scale bar, 30 µm. (B,C) Cells were pretreated with NR for 24 h before IR. NR was also present in the culture medium after IR. γH2AX foci were quantified 1 h, 3 h, and 6 h after IR. The number (B) and relative total area (C) of γH2AX foci per nucleus. The total area of γH2AX foci induced in cells untreated with NR, 1 h after IR, was taken as 100%. Data are presented as mean ± standard error (SE) (n = 200). Statistical analysis of differences between the groups was carried out by one-way ANOVA with post hoc comparisons using Tukey test. *** indicates statistically significant difference at p < 0.001. ns indicates that statistical difference is not significant. CTRL marks unirradiated control cells.

IR Does Not Affect NAD Levels in HDF
To elucidate how the exposure to IR can influence the cellular NAD content, we treated HDF with IR at a dose of 1 Gy or 5 Gy and measured NAD level in cell extracts at various time intervals. NAD concentration in control cells was about 8 pmol/µg and remained unchanged at 0.5, 2, 4, and 24 h after IR at both doses ( Figure 3A). Furthermore, differences in metabolic activities of NAD(P)H-dependent dehydrogenases estimated by MTT assay were indistinguishable between irradiated and control cells (Figures 3B and S3). Thus, IR at moderate doses does not induce any depletion of NAD in HDF.

Critical NAD Depletion Decreases the Rate of γH2AX Foci Elimination after IR
Next, we intended to clarify how the depletion of the NAD pool by the NAMPT inhibitor, FK866, could affect the efficiency of DNA DSB repair. HDFs were irradiated at a dose of 1 Gy and incubated in the medium containing FK866 for 1 or 4 days, then the number of γH2AX foci and the total area of foci per nucleus were estimated in treated and untreated cells as described in Section 3.2. As a result of a 1-day treatment with FK866, there was a decrease in intracellular NAD level by 50% compared to that in control cells ( Figure 1A), but the number and the total area of γH2AX foci per nucleus did not change 1 h and 3 h after IR treatment ( Figure 4A). Thus, a mild decrease in the NAD pool did not affect both the induction and the elimination of DNA DSB. After 4 days of treatment with FK866, intracellular NAD was hardly detectable in HDF ( Figure 1A). This resulted in less efficient IR-induced γH2AX foci formation-their number per nucleus 1 h after IR was 10% less ( Figure 4B, left panel)-and the total area dropped by 25% ( Figure  4B, right panel) compared to untreated cells. Surprisingly, despite the almost complete depletion of NAD, the elimination of γH2AX foci was observed at 3, 6, and 24 h after IR, but it was less effective than in FK866-untreated cells ( Figure 4B). These results indicated that under conditions of critical NAD depletion, HDF retained the ability to eliminate DNA DSB, but the DSB repair capacity was decreased in comparison to cells with normal NAD levels.

Critical NAD Depletion Decreases the Rate of γH2AX Foci Elimination after IR
Next, we intended to clarify how the depletion of the NAD pool by the NAMPT inhibitor, FK866, could affect the efficiency of DNA DSB repair. HDFs were irradiated at a dose of 1 Gy and incubated in the medium containing FK866 for 1 or 4 days, then the number of γH2AX foci and the total area of foci per nucleus were estimated in treated and untreated cells as described in Section 3.2. As a result of a 1-day treatment with FK866, there was a decrease in intracellular NAD level by 50% compared to that in control cells ( Figure 1A), but the number and the total area of γH2AX foci per nucleus did not change 1 h and 3 h after IR treatment ( Figure 4A). Thus, a mild decrease in the NAD pool did not affect both the induction and the elimination of DNA DSB. After 4 days of treatment with FK866, intracellular NAD was hardly detectable in HDF ( Figure 1A). This resulted in less efficient IR-induced γH2AX foci formation-their number per nucleus 1 h after IR was 10% less ( Figure 4B, left panel)-and the total area dropped by 25% ( Figure 4B, right panel) compared to untreated cells. Surprisingly, despite the almost complete depletion of NAD, the elimination of γH2AX foci was observed at 3, 6, and 24 h after IR, but it was less effective than in FK866-untreated cells ( Figure 4B). These results indicated that under conditions of critical NAD depletion, HDF retained the ability to eliminate DNA DSB, but the DSB repair capacity was decreased in comparison to cells with normal NAD levels.

NAD Depletion Suppresses the Accumulation of the Activated Form of ATM Kinase at DSB Sites and Its Colocalization with γH2AX
ATM kinase is one of the key regulators of cellular response to the induction of DNA DSB. At sites of DSB, ATM is activated by autophosphorylation at Ser1981 and phosphorylates various substrate proteins, including H2AX [6,8]. Our next objective was to test the hypothesis that decreased DNA DSB repair capacity in NAD-depleted cells might be associated with an impaired activation of ATM kinase at the sites of DNA damage. To do this, we treated cells with IR at a dose of 1 Gy after 4 days of incubation with 2 µM FK866. Cells were fixed 1 h after irradiation, and double immunostaining for γH2AX and activated form of ATM kinase (phosphorylated at Ser1981, pATM) was performed. The cells in G0 and G1 phases of the cell cycle were analyzed. Using confocal microscopy, it was shown that the irradiation of control cells induced the accumulation of pATM foci highly colocalized with γH2AX ( Figure 5A). After 4 days of treatment with FK866, we observed a considerable decrease in the number and the total area of pATM foci per nucleus ( Figure 5A,B). The degree of colocalization of pATM/γH2AX foci was visually reduced ( Figure 5A). For quantitative colocalization analysis of foci in confocal fluorescence microscopy images, Pearson's correlation coefficient (PCC) was used [53]. Consistent with confocal microscopy visual observations, high values of PCC (more than 0.7) for irradiated control cells indicated a high degree of colocalization of pATM and γH2AX foci. PCC values estimated for FK866-treated irradiated cells were significantly decreased (approx. 0.4), reflecting a much lower degree of colocalization of pATM and γH2AX in NAD-depleted cells ( Figure 5C). These results indicate that NAD depletion impairs ATM kinase activation and its colocalization with γH2AX in response to IR.

NAD Depletion Suppresses the Accumulation of the Activated Form of ATM Kinase at DSB Sites and Its Colocalization with γH2AX
ATM kinase is one of the key regulators of cellular response to the induction of DNA DSB. At sites of DSB, ATM is activated by autophosphorylation at Ser1981 and phosphorylates various substrate proteins, including H2AX [6,8]. Our next objective was to test the hypothesis that decreased DNA DSB repair capacity in NAD-depleted cells might be associated with an impaired activation of ATM kinase at the sites of DNA damage. To for irradiated control cells indicated a high degree of colocalization of pATM and γH2AX foci. PCC values estimated for FK866-treated irradiated cells were significantly decreased (approx. 0.4), reflecting a much lower degree of colocalization of pATM and γH2AX in NAD-depleted cells ( Figure 5C). These results indicate that NAD depletion impairs ATM kinase activation and its colocalization with γH2AX in response to IR. Figure 5. NAD depletion suppresses the accumulation of the activated form of ATM kinase at DSB sites and its colocalization with γH2AX. HDF were treated with FK866 for 4 days, and then irradiated at a dose of 1 Gy (IR). One hour after IR, cells were fixed and stained for γH2AX and ATM kinase phosphorylated at Ser1981 (pATM). (A) Immunofluorescence of γH2AX (red) and pATM Figure 5. NAD depletion suppresses the accumulation of the activated form of ATM kinase at DSB sites and its colocalization with γH2AX. HDF were treated with FK866 for 4 days, and then irradiated at a dose of 1 Gy (IR). One hour after IR, cells were fixed and stained for γH2AX and ATM kinase phosphorylated at Ser1981 (pATM). (A) Immunofluorescence of γH2AX (red) and pATM (green) foci. Cell nuclei were counterstained with DAPI (blue). Scale bar, 20 µm. (B) The number (left panel) and relative total area (right panel) of pATM foci per nucleus. Total area of pATM foci induced in cells untreated with FK866, 1 h after IR, was taken as 100%. (C) Quantification of colocalization of γH2AX and pATM by Pearson's correlation. Data are presented as mean ± SE (n = 200). Statistical analysis of differences between the groups was carried out by one-way ANOVA with post hoc comparisons using the Tukey test. *** indicates statistically significant difference at p < 0.001. CTRL marks unirradiated control cells.

Discussion
DSB repair is promoted by PARP and sirtuin family proteins that use NAD as a substrate to catalyze ADP-ribosylation or deacetylation of various targets in response to DSB induction. PARP1 is a major consumer of NAD in mammalian cells under conditions of genotoxic stress. PARP1 is activated by both single-and double-strand breaks of DNA [60] and catalyzes the cleavage of NAD to ADP-ribose and Nam, followed by the transfer of ADP-ribose units to target proteins, including itself [24]. DNA oxidative and DNA alkylating agents, which generate DNA single-strand breaks as intermediates of base excision repair (BER), can induce hyperactivation of PARP1 and massive poly-ADP-ribosylation leading to a dramatic drop of the intracellular NAD, ATP depletion and cell death [61][62][63][64][65][66]. Supplementation of cells with various forms of vitamin B3 replenishes NAD levels and recovers cell viability under genotoxic conditions [65,67], and it is suggested that elevated NAD can promote single-strand break/BER repair [68,69]. Yet, not much is known about how the induction of DSB affects the level of intracellular NAD and how changes in NAD bioavailability can influence DSB repair.
In the present study, we tested how the stimulation and inhibition of NAD biosynthesis changed DSB repair capacity in HDF exposed to moderate doses of IR. First, we demonstrated that the boosting of NAD levels by NR did not change the efficiency of DSB repair (as indicated by γH2AX foci elimination) after the exposure of cells to IR at a dose of 1 Gy. Moreover, IR, even at a dose of 5 Gy, did not lead to any changes in intracellular NAD concentration. Similar results were obtained in the previous study by Weidele et al. on peripheral blood mononuclear cells. Irradiation with X-ray at a dose of 2.5 or 5 Gy led to only a slight decrease in NAD levels, whereas the treatment with NAD precursor, nicotinic acid, had no effect on the DNA break repair as revealed by fluorimetric alkaline DNA unwinding assay. A significant (by 50%) drop in NAD was observed only after the exposure of cells to X-ray at a dose of 25 Gy. In this case, the treatment with nicotinic acid slightly increased DNA repair efficiency [70]. Interestingly, in another study, very high doses (50 Gy and above) of IR also caused NAD depletion in several mouse and human cancer cell lines. Moreover, replenishment of NAD levels by treatment of cells with 3 mM Nam improved DSB repair capacity [71]. Thus, our and other results suggest that, under conditions of moderate IR-induced genotoxic stress, the activation of PARP1 and other NAD-dependent enzymes (e.g., PARP2, PARP3, SIRT1, SIRT6, and SIRT7) involved in DNA damage response does not change the concentration of NAD, and it remains optimal for the efficient repair of DSB. The positive effect of NAD boosting on DSB repair can be achieved when IR induces a substantial decrease in intracellular NAD levels, and this requires exposure to very high doses of radiation.
Next, we tested how an imposed depletion of NAD can affect the efficiency of DSB repair in HDF after IR exposure. To do this, we suppressed NAD biosynthesis by FK866, an inhibitor of NAMPT, catalyzing the first step of NAD generation from Nam [55]. We showed that the decrease in intracellular NAD by 50% after 1 day of treatment with FK866 did not affect the DSB repair capacity. Moreover, even after 4 days of FK866 treatment, when intracellular NAD was almost undetectable, cells were still able to eliminate IR-induced DNA DSB, although DSB repair capacity was reduced compared to cells with normal NAD levels. Specifically, we observed a decrease in both number and total area of γH2AX foci per nucleus 1 h after a moderate (1 Gy) IR exposure under conditions of critical depletion of NAD. On the contrary, the kinetics of the subsequent elimination of γH2AX foci was significantly slowed down without NAD in comparison to control cells.
ATM-dependent phosphorylation of H2AX at serine 139, forming γH2AX foci at the sites of DSB, is one of the key events in the early cell response to DSB [7,8]. The NAD-dependent enzymes PARP1 and SIRT1 promote the recruitment and activation (autophosphorylation) of ATM at DSB [27,28,38]. Moreover, PARP1-and SIRT1-deficient cells display an altered activation of ATM and a reduced γH2AX foci formation in response to DSB-inducing agents [28,38]. Therefore, we hypothesized that the reduced γH2AX foci formation in response to IR, followed by the suppressed elimination of these foci in NAD-depleted cells, might be caused by the impaired activation and recruitment of ATM to DSB sites. In support of this assumption, we showed that IR-induced accumulation of the activated form of ATM at DSB sites and its colocalization with γH2AX were significantly decreased in HDF pretreated with FK866.
Since the homologous recombination is not available in G0-and G1-phase cells that were selected for the DSB repair analysis, we can assume that the maintenance of physiological concentrations of NAD is necessary for the efficient NHEJ, which is the major DSB repair pathway in G0/G1-phase. However, the very fact of DSB repair in cells lacking NAD indicates that NAD-dependent processes, such as protein deacetylation and ADPribosylation, may be important but not crucial factors controlling the NHEJ of DSB induced by moderate doses of IR.