www.aging‐us.com AGING 2019, Vol. 11, No. 5 Research Paper Delayed γH2AX foci disappearance in mammary epithelial cells from aged women reveals an age‐associated DNA repair defect Teresa Anglada1, Joan Repullés1,2, Anna Espinal3, Mark A LaBarge4,5, Martha R Stampfer5, Anna Genescà1,*, Marta Martín1,* 1 Department of Cell Biology, Physiology and Immunology, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain 2 Microscopy Platform, Biomedical Research Institute Sant Pau (IIB‐Sant Pau), 08041 Barcelona, Spain 3 Servei d’Estadística Aplicada, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain 4 Department of Population Sciences, and Center for Cancer and Aging, Beckman Research Institute at City of Hope, Duarte, CA 91010, USA 5 Biological Systems and Engineering Division, Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA * Equal contribution Correspondence to: Marta Martín, Anna Genescà; email: Marta.Martin@uab.cat, Anna.Genesca@uab.cat Keywords: aging, double‐strand break repair, γH2AX, human mammary epithelial cells, DNA damage, genome integrity Received: September 16, 2018 Accepted: March 3, 2019 Published: March 14, 2019 Copyright: Anglada et al. This is an open‐access article distributed under the terms of the Creative Commons Attribution License (CC BY 3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. ABSTRACT Aging is a degenerative process in which genome instability plays a crucial role. To gain insight into the link between organismal aging and DNA repair capacity, we analyzed DNA double‐strand break (DSB) resolution efficiency in human mammary epithelial cells from 12 healthy donors of young and old ages. The frequency of DSBs was measured by quantifying the number of γH2AX foci before and after 1Gy of γ‐rays and it was higher in cells from aged donors (ADs) at all times analyzed. At 24 hours after irradiation, ADs retained a significantly higher frequency of residual DSBs than young donors (YDs), which had already reached values close to basal levels. The kinetics of DSB induction and disappearance showed that cells from ADs and YDs repair DSBs with similar speed, although analysis of early times after irradiation indicate that a repair defect may lie within the firing of the DNA repair machinery in AD cells. Indeed, using a mathematical model we calculated a constant factor of delay affecting aged human epithelial cells repair kinetics. This defect manifests with the accumulation of DSBs that might eventually undergo illegitimate repair, thus posing a relevant threat to the maintenance of genome integrity in older individuals. INTRODUCTION Of the many lesions that DNA can suffer, the DNA double strand break (DSB) poses a considerable threat The aging process is related to a loss of function and an because joining of illegitimate ends can occur. One of increased probability of developing several diseases, the earliest events in DSB signaling is the phospho- such as cancer. Cellular changes associated with aging rylation of the histone H2AX on serine 139, γH2AX are an accentuated inflammatory response, alterations in [2]. Phosphorylation of H2AX spreads over megabases the redox cellular equilibrium, telomere attrition, and surrounding the break site, resulting in a platform that changes in nuclear structure, and all of them imply enables the recruitment of effector proteins at the relevant threats to maintenance of genomic integrity [1]. damaged DNA [3]. The modification of H2AX can be www.aging‐us.com 1510 AGING identified as discrete foci forming at DSB sites and underlying age-related DSB accumulation remain scoring of γH2AX foci is a widely used tool to estimate unclear. the number of DSBs induced after exposure to damaging agents [4]. γH2AX foci disappearance over To gain insight into the consequences of organismal time is a good approach to assess kinetics of DSB repair aging on DNA damage repair capacity, we have because once DNA has been repaired, H2AX phospho- measured DSB induction and resolution in finite rylation disappears and foci are no longer detectable [5]. lifespan non-transformed (pre-stasis) human mammary The DSB repair kinetics follow a biphasic pattern: most epithelial cells (HMECs) from 12 female donors of of the DSBs are repaired by the fast component of young (≤ 27) and old (≥ 60) ages. Our work shows that repair within the first two hours after induction, while cells from aged women have a higher basal level of the remaining DSBs can be repaired by the slow DSBs and display a sharp decline of DSB repair component of repair, which acts with slower kinetics efficiency that leads to the accumulation of these lesions and might require several hours –or even days– to after exposure to low doses of IR. Both, observed data complete repair [6–8]. and mathematical modelling of DSB repair kinetics indicate that old donors display a delayed firing of the Studies with models of in vitro aging have provided DNA damage response that contributes to the accumu- evidence of a higher frequency of unrepaired DSBs with lation of damage with age. time in culture. For example, replicative senescent cells accumulate more γH2AX than dividing cells, suggesting RESULTS a reduced repair ability or accumulation of DNA damage associated with replicative halt [9]. Also, non- Defining the criteria for analyzing DNA double senescent late population doubling (PD) cells during in strand breaks in pre-stasis HMECs vitro culture present with more unrepaired DSBs and more γH2AX signaling than earlier PD cells [10,11]. A HMECs were obtained from reduction mammoplasty similar tendency is observed with organismal aging, as tissue of 12 donors, which were classified according to cells from aged human donors present with an increased age into young donors (YDs, ≤ 27, age in parentheses): frequency of chromosomal reorganizations and γH2AX YD48R(16), YD240L(19), YD168R(19), YD184(21), foci with increasing age [11–14]. Although the increas- YD59L(23) and YD123(27) and aged donors (ADs, ≥ ed frequency of DSBs with age is clear, the mechanisms 60, age in parentheses): AD153L(60), AD112R(61), underlying it are yet unknown. AD122L(66), AD29(68), AD429ER(72), AD353P(72). Cells were cultured as pre-stasis strains in M87A The presence of a greater number of lesions in the DNA medium as described by Garbe and colleagues [21], to of aged cells could be due to a progressive accumula- support their long-term growth (Figure 1A). Despite tion of lesions over time, to difficult-to-repair DSBs using a low-stress medium, there was an accumulation marked by persistent γH2AX foci or to a limited of senescent cells with time in culture (Figure 1A and capacity of aged cells to repair new DSBs [15–17]. The 1B). In order to avoid interference from replicative- general notion of declined DSB repair efficiency with senescence associated DNA damage when assessing age is supported by some studies. Accumulation of age-dependent differences in the formation and residual γH2AX foci after ex vivo ionizing irradiation resolution of DSBs, early PDs were chosen (PD < 20 (IR) exposure of fibroblasts and hematopoietic stem which correspond to passages 4th to 6th) in which the cells of healthy donors suggests that older individuals frequency of senescent cells was ≤ 10%. have a reduced DSB repair capacity [14,18]. Similarly, Garm and colleagues [19] used comet assays and flow As previously reported [21,22] we found age-related cytometry techniques to measure DSBs in peripheral differences in the fractions of myoepithelial blood mononuclear cells from twins who ranged from (CD10+/CD227-) and luminal (CD10-/CD227+) cells in 40 to 77 years of age, and observed a tendency towards HMEC culture. Flow cytometry analysis of CD10 and diminished DSB repair with increasing age. In contrast, CD227 cell-lineage specific markers confirmed an age- human dermal fibroblasts from aged donors showed a dependent decrease in the myoepithelial fraction heterogeneous capacity for DSB repair after analyzing accompanied by an increase of the luminal fraction γH2AX fluorescence intensity [12], and even an in- (CD10+/CD227- in YD240L: 56.05%; AD112R: creased DSB repair rate with age in lymphocytes from 37.29%; CD10-/CD227+ in YD240L: 12.67%; AD112R: 94 donors exposed to IR [20]. Therefore, although the 20.06%) (Figure 1C). In order to rule out radiation- collected evidence suggests that the frequency of DNA- sensitivity differences between the two breast cell types, DSBs increases with age in multiple mammalian cells from young and aged donors were exposed to 1Gy tissues, the DSB repair capacity of cells from aged of γ-rays and labelled with γH2AX and claudin-4 (Cl4), individuals is still controversial and the mechanisms a cytoplasmic membrane protein mostly expressed by www.aging‐us.com 1511 AGING luminal cells (Figure 1D). As shown in Figure 1E, there of foci scored are dependent on the phase of the cell were no differences in the frequency of γH2AX foci cycle analyzed (Supplementary Figure 1). To mitigate between Cl4+ and Cl4- cells 2h after irradiation in any variability due to cell cycle, γH2AX foci counting was of the donors analyzed (Cl4-: 31.89 and Cl4+: 33.42 in restricted to cells in G1 phase, which were identified by the YD184; Cl4-: 27.68 and Cl4+: 26.41 in the AD112R; pericentrin labelling, a centrosomal protein that Mann-Whitney test, p-value > 0.05). These results duplicates along with DNA, allowing clear distinction indicate that radiation-induction of DSBs is similar in of cell cycle phase for each individual cell analyzed myoepithelial and luminal HMECs, ruling out the need [24]. γH2AX foci were scored before and after exposure to distinctively identify them when analyzing age- of HMECs to IR (1h, 2h and 24h pIR). In order to detect dependent differences in DNA repair. differences in γH2AX foci numbers between age groups (young donors vs old donors), a generalized linear Mammary epithelial cells from aged donors show an model with repeated measures for each donor was increased basal frequency of DSBs established (see Materials andMethods section). We first estab-lished the basal frequency of DSBs in cells γH2AX foci are accepted as surrogate markers of DSBs from young and old donors, in non-irradiated samples. [23], but the pattern of γH2AX staining and the number Using the generalized linear model, the estimated mean Figure 1. Pre‐stasis HMEC characterization and culture. (A) Representative growth curves of HMECs from YD184(21) and AD112R(61) in M87A medium with supplements. Dots correspond to correlative cell passages from passage 2. The dotted thin line indicates the early passages used for the experiments. Percentages of SA‐β‐Gal positive cells are indicated within the grey box (N > 500 cells). (B) The frequency of SA‐β‐Gal positive cells increases with time in culture. (C) Diagrams of flow cytometry analysis of CD10 (PE, phycoerytrin) and CD227 (FITC, fluorescein isothiocyanate) in YD240L(19) and AD112R(61) (N > 10000 cells). (D) Images of the immuno‐fluorescent staining of claudin‐4 (expressed by luminal cells, FITC, green), γH2AX (Cy3, red) and DAPI (blue) at 2h after 1Gy of γ‐rays exposure. Claudin‐4 positive (arrows) and negative (arrowheads) cells are shown. (E) Scatter dot plot and average number (red line) of γH2AX foci/cell in claudin‐4 positive and negative cells (N > 100 cells/donor). No statistical differences were observed (Mann‐Whitney test, p‐value > 0.05). www.aging‐us.com 1512 AGING number of γH2AX foci before irradiation was 0.96 Alignment of donors at this time according to statistics (CI95% = [0.70, 1.30]) in cells from YDs and 1.94 (CI95% (Kruskal-Wallis test and Dunn’s multiple comparisons = [1.43, 2.63]) in cells from ADs, twice the level scored correction), rendered clear differences between young in YDs (Figure 2A). Statistically significant differences and old donors and most of them continued to maintain between basal γH2AX foci frequencies in YDs and ADs an age-related position (Figure 2C 1h after irradiation). were detected (p-value = 0.0013; t = -3.22). In addition Again, data from YDs showed little variance, revealing to a lower basal frequency of DSBs, in most young similar DSB repair efficiencies while ADs presented donors (5 out of 6) 60%-75% of cells were devoid of with more γH2AX foci and higher inter-cellular any γH2AX foci, whereas in most old donors (5 out of variability. Overall, 1h after irradiation cells from ADs 6) less than 45% of cells were devoid of foci (Figure accumulated higher levels of unrepaired DSBs, 2B). In addition, in most YDs less than 10% of cells suggesting that these cells elicit a less efficient response carried more than 3 γH2AX foci per cell, whereas ~20% from the fast component of DSBs repair. of cells from ADs had more than 3 foci (X2 test, p-value < 0.0001). Despite the existence of inter-individual When γH2AX foci were scored two hours after IR differences between donors of similar ages, these exposure the estimated mean number of γH2AX foci per analyses demonstrate that both, the average basal cell had already decreased in all donors and it was frequency of DSBs and the fraction of cells carrying similar for YDs (11.58 γH2AX foci/cell, CI95% = [8.54, DSBs are higher in HMECs from aged donors as 15.70]) and for ADs (14.08 γH2AX foci/cell, CI95% = compared to young donors. [10.38, 19.08]) (Figures 2A and 2C 2h after irradiation, Table 1). The decline in γH2AX foci during this second Descriptive statistics was computed for each donor hour was higher in cells from ADs than in cells from (Table 1 and Supplementary Figure 2) and statistical YDs, suggesting that the initial impairment in DSBs differences regarding the mean number of γH2AX foci repair shown by ADs 1h after irradiation is eventually per cell of each donor were calculated (Kruskal-Wallis alleviated. test with a Dunn’s multiple comparisons correction). When donors were lined up based on statistical In order to evaluate the efficiency in the slow differences among them, most of YDs and ADs aligned component of DNA repair, we finally analyzed the according to an age-dependent order (Figure 2C Non- frequency of γH2AX foci 24 hours after IR exposure. irradiated). This analysis allowed us to detect that Both YDs and ADs have repaired most of the radiation unirradiated cells from YD123(27) and AD153L(60) induced DSBs, but while most of the YDs had reached a did not behave as the rest of the donors of their age frequency of residual DSBs close to the basal levels, group (Table 1, Figure 2C Non-irradiated), thus un- only two aged donors had reached their basal levels of masking the existence of inter-individual differences DSBs (Table 1). Thus, cells from ADs displayed a among donors. Besides these particular exceptions, the rest of YDs had a similar and low frequency of basal higher estimated mean number of γH2AX foci/cell than DSBs/cell (Table 1 and Figure 2C Non-irradiated) and, cells from YDs (YDs: 1.24 γH2AX foci/cell, CI95% = consequently, they statistically grouped together (a, b [0.91, 1.68]; ADs: 2.85 γH2AX foci/cell, CI95% = [2.10, and c) and were significantly different from most of 3.86]; p-value = 0.0001; t = -3.79) (Table 1, Figure 2A). ADs (d, e), which carry more basal DSBs/cell and Indeed, when donors were individually compared display a greater data dispersion. (Kruskal-Wallis test and Dunn’s multiple comparisons correction) the differences between YDs and ADs Aged donors accumulate higher levels of DSBs after allowed a clear age-related alignment (Figure 2C 24h irradiation after irradiation). Not only YD cells present with less γH2AX foci/cell, but also the frequency of cells devoid To study the efficiency of DSB repair with age, of γH2AX foci at 24h is 50%, close to their frequency exponentially growing cell cultures from all donors before irradiation (70%) (Figure 2D). In contrast, in were exposed to 1Gy of γ-rays. One hour after IR ADs the frequency of cells without γH2AX foci at 24h exposure, the estimated mean number of γH2AX/cell after irradiation is far from their basal frequency (15% was 15.70 (CI95% = [11.57, 21.29]) in YDs versus 22.27 vs 40%) (Figure 2E). Among cells with γH2AX foci, (CI95% = [16.42, 30.19]) in cells from ADs (Figure 2A). most of the YDs’ cells scored only 1 or 2 γH2AX foci As shown in Table 1, at this time point the mean per cell at 24h pIR, whereas ADs still accumulated 3 or number of γH2AX foci per cell strongly correlated with more γH2AX foci per cell (Figures 2D and 2E). Thus, at the age of the donors, ranging from 10.63 γH2AX foci 24 hours after irradiation more cells from ADs per cell in the youngest donor (YD48R(16)) to 32.7 accumulate DSBs, and also the frequency of DSBs per γH2AX foci per cell in the oldest donor (AD353P(72)). cell is higher than in YDs. www.aging‐us.com 1513 AGING Figure 2. Frequencies of γH2AX foci in HMECs from young and aged donors. (A) Estimated mean number of γH2AX foci/cell and confidence intervals for young and aged donors. Asterisks indicate significant differences between YDs and ADs (generalized linear model, p‐value < 0.01). The number of cells analyzed for each donor is stated in Table 1. (B) Frequency of cells with a defined number of γH2AX foci in non‐irradiated samples from the 12 donors. The number of cells analyzed for each donor is stated in Table 1. (C) Box plots of the frequency of γH2AX foci in cells from YDs and ADs in non‐irradiated samples and at 1h, 2h or 24h after exposure to 1 Gy of γ‐rays. Each donor is colored with blue or red depending on the group of age (blue for YDs and red for ADs). In each group, colors become darker with increasing age of the donor. Boxes include data from the upper to the lower quartile. The median is represented with a black line and whiskers compile 10 to 90% of the scored values. The number of cells analyzed for each donor is stated in Table 1. Statistical differences between donors are indicated following a letter code: donors signaled with the same letter do not show statistical differences and therefore different letters indicate statistically significant differences between donors (Kruskal‐Wallis test with Dunn’s multiple comparisons correction, p‐value < 0.05). (D, E) Distribution of cells according to the number of γH2AX foci/cell individually scored in YD48R(16) (D) and in AD122L(66) (E). Bars indicate the percentage of cells without foci (black bar) or with ≥1 γH2AX foci (colored bars) 24h after irradiation. The continuous line depicts this percentage before irradiation. The number of cells analyzed for each donor is stated in Table 1. (F) Hierarchical clustering of the 12 donors according to the standardized mean number of γH2AX foci scored in non‐irradiated samples and at 1, 2 and 24h after IR. The number of cells analyzed for each donor is stated in Table 1. www.aging‐us.com 1514 AGING Table 1. Descriptive analysis of the number of γH2AX foci per cell. Non-irradiated 1h post-irradiation 2h post-irradiation 24h post-irradiation Donor Mean SD N Mean SD N Mean SD N Mean SD N Young donors (YDs) 0.90 2.52 6152 16.16 7.85 2777 11.43 8.22 4046 1.41 2.33 5633 YD48R(16) 0.91 2.17 957 10.63 5.94 464 7.64 7.04 923 1.21 2.04 1000 YD240L(19) 0.71 2.64 1975 15.42 7.32 389 13.25 7.62 685 0.87 1.87 1000 YD168R(19) 0.78 2.60 809 17.13 7.34 357 8.05 6.38 740 0.58 1.40 879 YD184(21) 0.79 2.13 1000 15.04 6.89 470 12.87 8.02 752 1.05 1.72 1000 YD59L(23) 1.05 2.21 1000 19.75 7.22 609 16.17 7.15 511 2.87 2.75 956 YD123(27) 1.93 3.64 411 17.9 8.60 488 14.30 9.91 435 1.98 3.10 798 Aged donors (ADs) 2.03 3.85 4702 21.17 11.85 3616 13.79 9.52 3800 2.77 4.62 4730 AD153L(60) 0.78 2.30 998 17.36 9.00 895 13.72 8.61 733 2.08 2.85 996 AD112R(61) 2.20 3.57 770 16.31 9.75 780 14.18 8.34 671 2.33 5.18 996 AD122L(66) 2.68 4.91 984 21.45 11.15 588 9.33 6.71 822 5.38 6.85 484 AD29(68) 1.48 2.26 707 22.38 8.04 656 13.97 7.13 799 2.03 2.31 1000 AD429ER(72) 3.43 4.57 483 27.72 13.31 361 19.14 11.82 252 3.10 3.83 353 AD353P(72) 2.30 4.29 760 32.7 16.31 336 16.16 13.33 615 3.29 5.58 901 Finally, and in order to determine if the γH2AX foci Delayed firing of the DNA Damage Response (DDR) disappearance was a good marker of chronological age, with age we carried out a hierarchical clustering analysis using standardized values of γH2AX foci from the 12 donors DSB repair is not constant, as it follows biphasic in the 4 time points (non-irradiated, 1h, 2h and 24h after exponential negative kinetics. In order to determine the IR). With these data, the donors were grouped in 3 nature of the repair defect displayed by cells from older clusters (Figure 2F). A clearly separated cluster was donors, we aimed to describe the kinetics of DSB repair constituted by the oldest donors (AD122L(66), for the two age groups. We first calculated the rate of AD429ER(72) and AD353P(72)), which displayed the γH2AX foci disappearance for each time interval worst repair efficiency among all the donors. The 4 analyzed (Table 2). Because γH2AX foci assay does not youngest donors (YD48R(16), YD168R(19), YD240L allow the scoring of the DSBs induced immediately (19) and YD184(21)), which are the ones with the best after irradiation ( ), to estimate γH2AX foci dis- DSB repair performance, clustered together and sepa- appearance at the initial time interval we have used the rated from the other donors. And finally, an previously described standard estimation of 35 DSBs intermediate cluster included the remaining young induced per Gy of radiation in G1 cells [25]. According donors (YD123(27) and YD59L(23)) along with the 3 to this, during the first hour after DNA damage aged donors (AD153L(60), AD112R(61) and induction, the rate of DSB resolution was higher for AD29(68)) that frequently did not follow an age- YDs (53.83% of γH2AX foci disappeared) than for ADs dependent order in the previous statistical analyses (39.51%) (Figure 3A and Table 2), indicating a greater (Figure 2C). Hence, hierarchical clustering of donors DSB repair ability for YDs immediately after DNA according to γH2AX foci at different times after damage induction, while ADs end the 1st hour carrying irradiation reveals that DSB repair efficiency is a good higher numbers of unresolved DSBs. In contrast, the marker of age. rate of γH2AX foci disappearance between 1 and 2h www.aging‐us.com 1515 AGING after IR was higher in AD (21.09%) than in YD samples interval analyzed (2 to 24h post-irradiation) corresponds (13.51%). Two hours after irradiation ADs have to the slow component of repair, in which the rates of repaired 60.6% of radiation-induced DSBs, very close DSBs repaired were very similar for both young to the 67.3% of DSBs repaired by YDs’ cells, (28.63%) and aged donors (31.49%) (Figure 3A and suggesting that, although with some delay, cells from Table 2), suggesting that age-related differences in ADs are eventually able to launch the DDR and DNA repair efficiency lay within the initial times after efficiently resolve the accumulated DSBs. The last time DNA damage induction. Figure 3. Dynamics of γH2AX foci disappearance after irradiation. (A) Frequency of DSBs repaired within defined time intervals after exposure to 1Gy of γ‐rays for YDs and ADs. The number of DSBs induced after 1Gy exposure ( ) of G1 cells was estimated to be of 35 according to Rothkamm & Löbrich [25]. For the other time points, the numbers of DSBs are those depicted in Tables 1 and 2. (B) Kinetics of γH2AX foci disappearance for young and aged donors following the model of first order kinetic reaction stated in the methods section. The mean number of γH2AX foci scored at each time point is represented with dots (blue for YDs, red for ADs) and it is stated in Table 1. The lines represent the kinetics of DSBs repair estimated after modeling data of all γH2AX foci/cell from the 12 donors at 1, 2 and 24h after irradiation. The number of cells analyzed for each group of age is stated in Table 1. The inset in the graph shows a detail of the early times after IR exposure. The dotted lines represent an extrapolation of the DSB repair kinetics in the time interval comprised between the DSB repair initiation and 1h after IR. Arrowheads indicate the moment of repair initiation, when the extrapolation lines for YDs and ADs reach the number of γH2AX foci present immediately after IR. Table 2. Rate of DSBs repair within intervals of time. DSB induced/ Time after IR DSB repaired (%) ** remaining Time interval exposure (h) YDs ADs YDs ADs 0* 35 35 1 16.16 21.17 0* - 1h 53.83 39.51 2 11.43 13.79 1h - 2h 13.51 21.09 24 1.41 2.77 2h - 24h 28.63 31.49 * 0 = immediately after irradiation. The current estimation of 35 DSB per Gy is used [25]. ** Frequency of DSB repair has been calculated as the difference in DSBs between one time point and the time immediately before, assuming that repairing 35 DSBs is 100% of repair. www.aging‐us.com 1516 AGING In order to test the hypothesis of an age-related delay in γH2AX foci, which are markers of unresolved DSBs. DNA-DSB repair initiation, we established a first order Hence, it is tempting to speculate that a period of kinetic reaction using a Nonlinear Regression Model. latency exists before ADs are able to fire a fully The model, described in Materials and Methods section, operative DNA repair response, although once predicts the number of γH2AX foci in AD and YD cells launched, they are able to repair with a speed similar to and it is resolved as follows: that of YDs. 24.32 . * _ if YD DISCUSSION _ . * . 24.32 _ if AD To investigate the age-associated impairment of The repair of DSBs is estimated by the constant of genomic integrity, we examined the DNA-DSB repair γH2AX foci decay β1 = -0.44, CI95% = [-0.64, -0.25] for efficiency in cells from healthy individuals of different young and aged donors. In the case of ADs, the model ages. The increased basal frequency of γH2AX foci includes a constant factor of delay in DSB repair with donor’s age observed in HMECs is in agreement initiation, which is estimated to be β0 = 0.27, CI95% = with results reported in other cell types from healthy [0.08, 0.45]. Thus, the equation is different for YDs and human donors [11–14] and reveals an age-dependent for ADs, as it assumes that ADs present a delayed start accumulation of DSBs. The observed age-related of DSB repair (e-0.44*t+0.27) with respect to YDs (e-0.44*t). increase in the number of endogenous DSBs could be The basal frequency of DSBs ([n_foci]0) is also includ- attributed to a stochastic accumulation of damage with ed in the equation because it is different for YDs and time or alternatively, an alteration of the DSB repair ADs (Table 1). Finally, the model estimates that the mechanism could account for an accelerated accumu- initial number of γH2AX foci in G1 phase HMECs lation of unresolved DSBs with age [15–17]. In this induced by 1Gy of γ-rays is = 24.32, CI95% = [17.28, regard, our results show that HMECs from ADs pre- 31.35]. Therefore, immediately after irradiation the cells sented increased frequencies of DSBs at all times carry those radiation-induced DSBs plus their basal analyzed after IR exposure, which manifests an frequency of DSBs ([n_foci]0) and, according to the impaired ability to repair DSBs with age. In fact, the model, these γH2AX values are 25.22 for YDs and hierarchical clustering analysis performed using data 26.35 for ADs. Although our model’s estimation of from γH2AX foci scored at all time points, efficiently induced DSBs is lower than the 35 DSBs/Gy reported grouped donors by age, thus demonstrating that analysis previously [25], it is very similar to others’ estimations of γH2AX foci disappearance after IR exposure could of ~25 DSBs/Gy in G1 cells [26,27]. Discrepancies in be a potential marker for physiological aging. It is worth the number of DSBs induced can be attributed to the to point out that this analysis efficiently unmasked inter- source of radiation, the dose rate used in each experi- individual variation amongst donors with similar ages, ment or to an overestimation of the number of DSBs which grouped in an intermediate cluster, and this is detected by PGFE methodology. especially clear in aged donors. High inter-individual heterogeneity in γH2AX analysis has also been reported As shown in Figure 3B, this model renders estimated in studies measuring γH2AX fluorescent intensity in DSB repair kinetics between 1 and 24h after IR for YDs blood samples [20,28]. Along with inter-individual and ADs that fit well the data observed. Although not variability, γH2AX foci disappearance data in HMEC strictly applicable, the model has also been used to revealed significant inter-cellular variation in the fre- make an extrapolation corresponding to the kinetics of quency of γH2AX foci in aged donors, which in fact DSB repair from the time point immediately after arises as a remarkable feature of ADs. This is in irradiation to 1h pIR, shown as dotted colored lines accordance with the recently published results from (inset in Figure 3B). The dotted line in YDs reaches the Cheung and colleagues [29] as they found an age- value of 25.22 γH2AX foci at a time close to 0 (blue associated increased cell-to-cell variability and an arrowhead in Figure 3B inset), suggesting that YDs increased inter-individual heterogeneity in chromatin initiate repair immediately after irradiation and they modifications using a mass cytometry analysis. efficiently diminish the number of DSBs during this Efficient tools that allow the detection of donors in first hour. Instead, ADs maintain the number of γH2AX which physiological aging does not completely match foci they had immediately after IR for a longer time, with chronological aging might be useful to improve because when data obtained is extrapolated from 1h pIR and adapt preventive diagnostic controls among other backward (dotted red line in Figure 3B inset) the value medical procedures. of 26.35 γH2AX foci is attained at a time between 0 and 1h (red arrowhead in Figure 3B inset), suggesting that The repair kinetics we observed in HMECs from older ADs begin to resolve DSBs later than YDs. Thus, ADs individuals after irradiation is in agreement with the reach the first hour after irradiation carrying more general notion of an age-associated decline in the DNA www.aging‐us.com 1517 AGING repair capacity, which is evidenced as a decreased low and protracted doses of radiation in which the γH2AX foci disappearance after IR exposure in repair machinery is continuously challenged. Epidemio- different cell types [14,18,19]. However, an increased logical studies have demonstrated an increased excess DSB repair rate with age has also been reported in risk for some types of cancer after exposure to low and blood mononuclear cells from 94 healthy donors [20]. protracted doses of radiation with age at exposure [38]. We propose that the apparent disparity among these For example, individuals exposed to the radioactive studies could be explained with our here presented contamination of the Techa River in the Urals had an experimental and modeled data. Although ADs show a increased excess relative risk of cancer mortality with delay in the initiation of repair, once DSB repair has increasing age at first exposure [39]. Also, data from been initiated, both groups of donors display similar workers at the Oak Ridge National Laboratory exposed DSB repair kinetics. This delayed firing translates into to low radiation doses revealed an association between the accumulation of yet to be resolved DSBs in early age at exposure and cancer mortality [40]. Thus, the times after irradiation. Eventually, ADs launch the impaired ability in DSB repair makes older people repair machinery and they start to resolve these DSBs, among the adult population particularly susceptible to appearing as even more efficient than YDs, but only ionizing radiation detrimental effects. because they have repaired less DSBs immediately before. Our study provides valuable information about the relation between aging and DNA-DSB accumulation in A delay in DSB repair initiation could be explained by human mammary epithelial cells from healthy donors, initial difficulties in loading repair proteins to DSB and we expect our results will serve as a basis for sites. Primary fibroblasts showed a delayed recruitment further studies regarding impaired DSB repair mecha- of MRE11 and RAD50 proteins with increasing donors’ nisms in aged individuals. Future studies would be age [11]. Also, a delayed recruitment to DSB sites of necessary to explore the mechanisms responsible for the 53BP1 –a repair protein that is involved in the non- delay in the initiation of DSB repair with age and their homologous end joining (NHEJ) repair pathway– was implications in the global cellular context. described in in vitro aged HMECs with a time-course experiment of 53BP1 foci formation [10]. In agreement MATERIALS AND METHODS with this, an age-associated decline of the NHEJ repair efficiency was reported in mice [30], rats [31] and Cell culture human senescent cells [32]. Cell lines defective in ATM or 53BP1 were described as presenting an accumulation Finite lifespan pre-stasis HMECs were obtained from of long-lasting residual DSBs [33], suggesting that a reduction mammoplasty tissue of 11 donors: 48R (16 defective recruitment could also translate into some yo), 240L (19 yo), 168R (19 yo), 184 (21 yo), 59L (23 kind of repair defect. In line with this, we observed that yo), 123 (27 yo), 153L (60 yo), 112R (61 yo), 122L (66 the repair defect of aged donors is accompanied by an yo), 29 (68 yo), 429ER (72 yo), or peripheral non-tumor increased frequency of γH2AX foci at 24h after irra- containing mastectomy tissue of 1 donor: 353P (72 yo). diation. Although we cannot rule out that these residual Donors were classified into two groups depending on DSBs correspond to complex damage sites or to age: young donors (YDs, ≤ 27 years old) and aged heterochromatin-located DSBs that are being repaired donors (ADs, ≥ 60 years old). When referring to donors, slowly, we hypothesize that their presence is related to the group of age is followed by the specimen the delay in the firing of the DDR. Long-lasting γH2AX identification and the age of the donor in parentheses. signaling after IR exposure was proposed to be a marker Cells were cultured as pre-stasis strains using M87A of DNA damage and aging [34] and was correlated with medium with cholera toxin and oxytocin according to radiosensitivity in mammalian cell lines [35]. previously reported methods [21], with the addition of 100 U/ml penicillin and 100 μg/ml streptomycin. Cells A delay in DSB repair initiation, probably related to were incubated at 37ºC and 5% CO2 atmosphere. The difficulties in launching an effective DDR, poses a passage number indicates each time the cells have been relevant threat to genomic integrity, as the accumulation detached form the petri dish using trypsin and seeded of unresolved DSBs leads to increasing probabilities of into new vessels. The population doubling (PD) illegitimate repair [36]. Accumulation of genomic indicates each time a cell has divided and was rearrangements arising from illegitimate DSB repair calculated as described by Greenwood and colleagues was reported in different tissues from old individuals [41], using the equation: PD = [log(cells harvested/cells [37]. These abnormalities can affect multiple genes and plated)]/log2. Cells have been obtained from mammary they are a potential source of oncogenic transformation. gland surgical discarded tissues that are subcultured It can be of particular concern in individuals exposed to twice before calculation of the PD. www.aging‐us.com 1518 AGING SA-β-Gal activity detection Olympus) and a CCD camera (CV-M4+CL, JAI). The capture methodology was adapted from the Spot- SA-β-Gal activity was detected as described by Debacq- counting system (Spot AX software, Applied Imaging) Chainiaux [42]. Blue staining was detected under an as described by Hernández [24]. Images were acquired IX71 microscope equipped with DP20 camera and automatically with a 60x objective using predefined cell A software (Olympus, Hamburg, Germany). settings. Four z-stacks were acquired for γH2AX and 6 for pericentrin, with a step size of 1.55 μm between Irradiation planes. Cells with only one pericentrin signal were selected and γH2AX foci were scored using When indicated, exponentially growing HMECs were FociPicker3D algorithm for Fiji software [43]. exposed to 1 Gy of γ-rays using an IBL-437C R-137 Cs irradiator (dose rate of 5.10 Gy/min). Statistical analysis and data modelling Immunodetection Descriptive analysis and graphics were performed using Microsoft Excel (Microsoft® Excel® 2011, v14.1, Flow cytometric analysis Redmond, Washington, USA) and GraphPad Prism 6 After trypsinization, HMECs were blocked for 20 (GraphPad Software Inc., San Diego, CA, USA) with minutes in PBS-1% BSA, incubated for 30 minutes with methods indicated in the results where applicable. When anti-CD227-FITC (clone HMPV, Becton Dickinson, comparing the number of γH2AX foci/cell among Franklin Lakes, NJ, USA) and anti-CD10-PE (clone individual donors, Kruskal-Wallis test with Dunn’s HI10a, BioLegend, San Diego, CA, USA) at a final multiple comparisons correction was applied and concentration of 1:100, all on ice. Flow cytometric different letters indicate statistical differences (p-value analysis was performed using a FACSCanto (Becton < 0.05) between donors in the graphical representation. Dickinson). In order to statistically compare the two age groups at each time point analyzed, a generalized linear model Immunofluorescence with a Negative Binomial distribution response and To detect γH2AX and pericentrin, HMECs were fixed with repeated measures for each donor was established. in 4% paraformaldehyde for 15 minutes and perme- The estimated values for γH2AX foci number and the abilized in a 1xPBS-0.5% Triton- X100 for 20 minutes. corresponding confidence intervals were obtained using To detect γH2AX and claudin-4, cells were fixed with SAS software (SAS v9.4, SAS Institute Inc., Cary, NC, ice-cold methanol for 10 minutes. Cells were incubated USA). for 1 hour with blocking solution (1xPBS-0.1% Tween20-3% FBS) before applying primary antibodies For the hierarchical cluster analysis, standardized values mouse anti-γH2AX (Ser139) (clone JBW301, Millipore, of γH2AX foci from the 12 donors along the four time Madrid, Spain), rabbit anti-pericentrin (Abcam, points (non-irradiated, 1h, 2h and 24h pIR) were used. Cambridge, UK) or rabbit anti-claudin-4 (Abcam) at Standardized data was obtained by subtracting the mean 1:1000, 1:2000 and 1:250 final concentrations respec- number of each condition (donor, time and replicate) to tively. Secondary antibodies anti-mouse Cy3 (Jackson the number of γH2AX foci scored for each cell and then ImmunoResearch Inc., Cambridge, UK) and anti-rabbit dividing this value by the standard deviation of the A488 (Thermo Fisher Scientific, Waltham, MA, USA) condition. As it was defined by Everitt and colleagues were applied at a final concentration of 1:800 and 1:500 [44], a hierarchical classification consists of a series of (claudin-4) or 1:1000 (pericentrin) respectively. Nuclei partitions, which may run from a single cluster were counterstained with 4’,6-diamidino-2-phenylin- containing all individuals, to n clusters each containing dole (DAPI) at a final concentration of 0.25 μg/ml. For a single individual. In our case we wanted to determine image acquisition an Olympus BX61 epifluorescent the inter-group (young vs old) proximity, and thus the microscope equipped with a CV-M4+CL camera (JAI, Ward method [45] was applied using R software Grosswallstadt, Germany) and Cytovision software (version 3.4.4, Vienna, Austria). In this method, the (Applied Imaging, Newcastle, UK) were used. criterion for choosing the pair of clusters to merge at each step is based on the size of the error sum-of- Automated microscopy and γH2AX foci counting squares. Hierarchical clustering is represented by a two- dimensional diagram known as a dendrogram, which γH2AX foci counting was done following a semi- illustrates the fusions or divisions made at each stage of automatic approach. Images from slides with γH2AX the analysis. and pericentrin immunofluorescence were captured using an Olympus BX61 epifluorescence microscope A first order kinetic reaction was established to obtain equipped with an automatic motorized stage (BX-UCB, estimations regarding the kinetics of DSB repair in YDs www.aging‐us.com 1519 AGING and ADs. This approach was done using methodologies 3. Nakamura AJ, Rao VA, Pommier Y, Bonner WM. The for Nonlinear Regression Model using SAS software complexity of phosphorylated H2AX foci formation (SAS v9.4, SAS Institute Inc.). The established model is and DNA repair assembly at DNA double‐strand described by the following equation: breaks. Cell Cycle. 2010; 9:389–97. https://doi.org/10.4161/cc.9.2.10475 _ if YD 4. Rothkamm K, Barnard S, Moquet J, Ellender M, Rana _ _ if AD Z, Burdak‐Rothkamm S. DNA damage foci: meaning and significance. Environ Mol Mutagen. 2015; where [n_foci]t is the number of γH2AX foci at a 56:491–504. https://doi.org/10.1002/em.21944 concrete time after irradiation, is the number of 5. Chowdhury D, Keogh MC, Ishii H, Peterson CL, radiation-induced DSBs, β1 is the γH2AX foci decay Buratowski S, Lieberman J. γ‐H2AX dephosphorylation proportion, [n_foci]0 is the basal frequency of γH2AX by protein phosphatase 2A facilitates DNA double‐ foci (before irradiation) and β0 is a constant of delayed repair onset. This model assumes that (1) the same strand break repair. Mol Cell. 2005; 20:801–09. number of DSBs per unit of radiation are induced in YD https://doi.org/10.1016/j.molcel.2005.10.003 and AD cells immediately after irradiation ; (2) cells 6. Metzger L, Iliakis G. Kinetics of DNA double‐strand do not reach complete repair, but instead they reach the break repair throughout the cell cycle as assayed by basal frequency [n_foci]0 of DSBs and (3) cells from pulsed field gel electrophoresis in CHO cells. Int J ADs suffer a delay in DSBs repair initiation (β0). Radiat Biol. 1991; 59:1325–39. https://doi.org/10.1080/09553009114551201 ACKNOWLEDGEMENTS 7. DiBiase SJ, Zeng ZC, Chen R, Hyslop T, Curran WJ Jr, Iliakis G. DNA‐dependent protein kinase stimulates an We gratefully acknowledge Gemma Fuster (Hospital independently active, nonhomologous, end‐joining Clínic, Barcelona) for her help with flow cytometry. apparatus. Cancer Res. 2000; 60:1245–53. CONFLICTS OF INTEREST 8. Löbrich M, Jeggo P. A Process of Resection‐ Dependent Nonhomologous End Joining Involving the The authors declare that they have no conflicts of Goddess Artemis. Trends Biochem Sci. 2017; 42:690– interest. 701. https://doi.org/10.1016/j.tibs.2017.06.011 9. Sedelnikova OA, Horikawa I, Zimonjic DB, Popescu FUNDING NC, Bonner WM, Barrett JC. Senescing human cells and ageing mice accumulate DNA lesions with This work was supported by by Consejo de Seguridad unrepairable double‐strand breaks. Nat Cell Biol. Nuclear and Fundación Dexeus Salud de la Mujer 2016 2004; 6:168–70. https://doi.org/10.1038/ncb1095 grants. Additional support was provided by Generalitat 10. Hernández L, Terradas M, Martín M, Feijoo P, Soler D, de Catalunya (2017-SGR-503). TA has been a bene- Tusell L, Genescà A. Increased mammogram‐induced ficiary of FI-DGR 2015-2017 (Generalitat de Catalunya) and Universitat Autònoma de Barcelona DNA damage in mammary epithelial cells aged in PhD program fellowships. MS was supported by the vitro. PLoS One. 2013; 8:e63052. Office of Health and Biological Research, US. Depart- https://doi.org/10.1371/journal.pone.0063052 ment of Energy under Contract No. DE-AC02- 11. Sedelnikova OA, Horikawa I, Redon C, Nakamura A, 05CH11231. Zimonjic DB, Popescu NC, Bonner WM. Delayed kinetics of DNA double‐strand break processing in REFERENCES normal and pathological aging. Aging Cell. 2008; 7:89–100. https://doi.org/10.1111/j.1474‐ 1. López‐Otín C, Blasco MA, Partridge L, Serrano M, 9726.2007.00354.x Kroemer G. The hallmarks of aging. Cell. 2013; 12. Kalfalah F, Seggewiß S, Walter R, Tigges J, Moreno‐ 153:1194–217. Villanueva M, Bürkle A, Ohse S, Busch H, Boerries M, https://doi.org/10.1016/j.cell.2013.05.039 Hildebrandt B, Royer‐Pokora B, Boege F. Structural 2. Rogakou EP, Pilch DR, Orr AH, Ivanova VS, Bonner chromosome abnormalities, increased DNA strand WM. DNA double‐stranded breaks induce histone breaks and DNA strand break repair deficiency in H2AX phosphorylation on serine 139. J Biol Chem. dermal fibroblasts from old female human donors. 1998; 273:5858–68. Aging (Albany NY). 2015; 7:110–22. https://doi.org/10.1074/jbc.273.10.5858 https://doi.org/10.18632/aging.100723 www.aging‐us.com 1520 AGING 13. Joyce NC, Harris DL, Zhu CC. Age‐related gene 23. Löbrich M, Shibata A, Beucher A, Fisher A, Ensminger response of human corneal endothelium to oxidative M, Goodarzi AA, Barton O, Jeggo PA. gammaH2AX stress and DNA damage. Invest Ophthalmol Vis Sci. foci analysis for monitoring DNA double‐strand break 2011; 52:1641–49. https://doi.org/10.1167/iovs.10‐ repair: strengths, limitations and optimization. Cell 6492 Cycle. 2010; 9:662–69. https://doi.org/10.4161/cc.9.4.10764 14. Rübe CE, Fricke A, Widmann TA, Fürst T, Madry H, Pfreundschuh M, Rübe C. Accumulation of DNA 24. Hernández L, Terradas M, Martín M, Tusell L, Genescà damage in hematopoietic stem and progenitor cells A. Highly sensitive automated method for DNA during human aging. PLoS One. 2011; 6:e17487. damage assessment: gamma‐H2AX foci counting and https://doi.org/10.1371/journal.pone.0017487 cell cycle sorting. Int J Mol Sci. 2013; 14:15810–26. 15. Vijg J. Somatic mutations and aging: a re‐evaluation. https://doi.org/10.3390/ijms140815810 Mutat Res. 2000; 447:117–35. 25. Rothkamm K, Löbrich M. Evidence for a lack of DNA https://doi.org/10.1016/S0027‐5107(99)00202‐X double‐strand break repair in human cells exposed to 16. Garinis GA, van der Horst GT, Vijg J, Hoeijmakers JH. very low x‐ray doses. Proc Natl Acad Sci USA. 2003; DNA damage and ageing: new‐age ideas for an age‐ 100:5057–62. https://doi.org/10.1073/pnas.0830918100 old problem. Nat Cell Biol. 2008; 10:1241–47. https://doi.org/10.1038/ncb1108‐1241 26. Stenerlöw B, Karlsson KH, Cooper B, Rydberg B. 17. Gorbunova V, Seluanov A. DNA double strand break Measurement of prompt DNA double‐strand breaks repair, aging and the chromatin connection. Mutat in mammalian cells without including heat‐labile Res. 2016; 788:2–6. sites: results for cells deficient in nonhomologous end https://doi.org/10.1016/j.mrfmmm.2016.02.004 joining. Radiat Res. 2003; 159:502–10. https://doi.org/10.1667/0033‐7587(2003)159[0502: 18. Li Z, Zhang W, Chen Y, Guo W, Zhang J, Tang H, Xu Z, MOPDDS]2.0.CO;2 Zhang H, Tao Y, Wang F, Jiang Y, Sun FL, Mao Z. Impaired DNA double‐strand break repair contributes 27. Iliakis GE, Cicilioni O, Metzger L. Measurement of to the age‐associated rise of genomic instability in DNA double‐strand breaks in CHO cells at various humans. Cell Death Differ. 2016; 23:1765–77. stages of the cell cycle using pulsed field gel https://doi.org/10.1038/cdd.2016.65 electrophoresis: calibration by means of 125I decay. Int J Radiat Biol. 1991; 59:343–57. 19. Garm C, Moreno‐Villanueva M, Bürkle A, Petersen I, https://doi.org/10.1080/09553009114550321 Bohr VA, Christensen K, Stevnsner T. Age and gender 28. Ismail IH, Wadhra TI, Hammarsten O. An optimized effects on DNA strand break repair in peripheral blood mononuclear cells. Aging Cell. 2013; 12:58–66. method for detecting gamma‐H2AX in blood cells reveals a significant interindividual variation in the https://doi.org/10.1111/acel.12019 gamma‐H2AX response among humans. Nucleic Acids 20. Sharma PM, Ponnaiya B, Taveras M, Shuryak I, Turner Res. 2007; 35:e36. H, Brenner DJ. High throughput measurement of https://doi.org/10.1093/nar/gkl1169 γH2AX DSB repair kinetics in a healthy human population. PLoS One. 2015; 10:e0121083. 29. Cheung P, Vallania F, Warsinske HC, Donato M, https://doi.org/10.1371/journal.pone.0121083 Schaffert S, Chang SE, Dvorak M, Dekker CL, Davis MM, Utz PJ, Khatri P, Kuo AJ. Single‐Cell Chromatin 21. Garbe JC, Bhattacharya S, Merchant B, Bassett E, Modification Profiling Reveals Increased Epigenetic Swisshelm K, Feiler HS, Wyrobek AJ, Stampfer MR. Variations with Aging. Cell. 2018; 173:1385– Molecular distinctions between stasis and telomere 1397.e14. https://doi.org/10.1016/j.cell.2018.03.079 attrition senescence barriers shown by long‐term culture of normal human mammary epithelial cells. 30. Vaidya A, Mao Z, Tian X, Spencer B, Seluanov A, Gorbunova V. Knock‐in reporter mice demonstrate Cancer Res. 2009; 69:7557–68. https://doi.org/10.1158/0008‐5472.CAN‐09‐0270 that DNA repair by non‐homologous end joining declines with age. PLoS Genet. 2014; 10:e1004511. 22. Garbe JC, Pepin F, Pelissier FA, Sputova K, https://doi.org/10.1371/journal.pgen.1004511 Fridriksdottir AJ, Guo DE, Villadsen R, Park M, Petersen OW, Borowsky AD, Stampfer MR, Labarge 31. Vyjayanti VN, Rao KS. DNA double strand break repair MA. Accumulation of multipotent progenitors with a in brain: reduced NHEJ activity in aging rat neurons. basal differentiation bias during aging of human Neurosci Lett. 2006; 393:18–22. https://doi.org/10.1016/j.neulet.2005.09.053 mammary epithelia. Cancer Res. 2012; 72:3687–701. https://doi.org/10.1158/0008‐5472.CAN‐12‐0157 32. Seluanov A, Mittelman D, Pereira‐Smith OM, Wilson www.aging‐us.com 1521 AGING JH, Gorbunova V. DNA end joining becomes less positive results. Environ Mol Mutagen. 2004; 43:36– efficient and more error‐prone during cellular 44. https://doi.org/10.1002/em.10207 senescence. Proc Natl Acad Sci USA. 2004; 101:7624– 42. Debacq‐Chainiaux F, Erusalimsky JD, Campisi J, 29. https://doi.org/10.1073/pnas.0400726101 Toussaint O. Protocols to detect senescence‐ 33. Riballo E, Kühne M, Rief N, Doherty A, Smith GC, associated beta‐galactosidase (SA‐betagal) activity, a Recio MJ, Reis C, Dahm K, Fricke A, Krempler A, biomarker of senescent cells in culture and in vivo. Parker AR, Jackson SP, Gennery A, et al. A pathway of Nat Protoc. 2009; 4:1798–806. double‐strand break rejoining dependent upon ATM, https://doi.org/10.1038/nprot.2009.191 Artemis, and proteins locating to γ‐H2AX foci. Mol 43. Du G, Drexler GA, Friedland W, Greubel C, Hable V, Cell. 2004; 16:715–24. Krücken R, Kugler A, Tonelli L, Friedl AA, Dollinger G. https://doi.org/10.1016/j.molcel.2004.10.029 Spatial dynamics of DNA damage response protein 34. Siddiqui MS, François M, Fenech MF, Leifert WR. foci along the ion trajectory of high‐LET particles. γH2AX responses in human buccal cells exposed to Radiat Res. 2011; 176:706–15. ionizing radiation. Cytometry A. 2015; 87:296–308. https://doi.org/10.1667/RR2592.1 https://doi.org/10.1002/cyto.a.22607 44. Everitt BS, Landau S, Leese M, Stahl D. Cluster 35. Taneja N, Davis M, Choy JS, Beckett MA, Singh R, Kron Analysis. 5th ed. Quality and Quantity. 2011. 348 p. SJ, Weichselbaum RR. Histone H2AX phosphorylation 45. Ward JH Jr. Hierarchical Grouping to Optimize an as a predictor of radiosensitivity and target for Objective Function. J Am Stat Assoc. 1963; 58:236– radiotherapy. J Biol Chem. 2004; 279:2273–80. 44. https://doi.org/10.1074/jbc.M310030200 https://doi.org/10.1080/01621459.1963.10500845 36. Martín M, Terradas M, Tusell L, Genescà A. ATM and DNA‐PKcs make a complementary couple in DNA double strand break repair. Mutat Res Rev Mutat Res. 2012; 751:29–35. https://doi.org/10.1016/j.mrrev.2011.12.006 37. Suh Y, Vijg J. Maintaining genetic integrity in aging: a zero sum game. Antioxid Redox Signal. 2006; 8:559– 71. https://doi.org/10.1089/ars.2006.8.559 38. Preston DL, Kitahara CM, Freedman DM, Sigurdson AJ, Simon SL, Little MP, Cahoon EK, Rajaraman P, Miller JS, Alexander BH, Doody MM, Linet MS. Breast cancer risk and protracted low‐to‐moderate dose occupational radiation exposure in the US Radiologic Technologists Cohort, 1983‐2008. Br J Cancer. 2016; 115:1105–12. https://doi.org/10.1038/bjc.2016.292 39. Krestinina LY, Preston DL, Ostroumova EV, Degteva MO, Ron E, Vyushkova OV, Startsev NV, Kossenko MM, Akleyev AV. Protracted radiation exposure and cancer mortality in the Techa River Cohort. Radiat Res. 2005; 164:602–11. https://doi.org/10.1667/RR3452.1 40. Richardson DB, Wing S. Greater sensitivity to ionizing radiation at older age: follow‐up of workers at Oak Ridge National Laboratory through 1990. Int J Epidemiol. 1999; 28:428–36. https://doi.org/10.1093/ije/28.3.428 41. Greenwood SK, Hill RB, Sun JT, Armstrong MJ, Johnson TE, Gara JP, Galloway SM. Population doubling: a simple and more accurate estimation of cell growth suppression in the in vitro assay for chromosomal aberrations that reduces irrelevant www.aging‐us.com 1522 AGING SUPPLEMENTARY MATERIAL Supplementary Figure 1. Immunofluorescent labeling of γH2AX (Cy3, red), pericentrin (FITC, green) with DAPI (blue) counterstain at different phases of the cell cycle. Scoring of individualized γH2AX foci was restricted to G1 phase cells (red box) by selecting those cells with one pericentrin signal. From late S to mitosis cells have two pericentrin signals. Also the pan‐nuclear pattern of γH2AX foci during S phase is very characteristic. Supplementary Figure 2. Mean number of γH2AX foci per cell and their standard deviation in young and aged HMEC donors before irradiation and at 1, 2 and 24 hours after IR. www.aging‐us.com 1523 AGING