1. Introduction
DNA double-strand breaks (DSB) involve special challenges for the cells. DNA broken ends may rejoin with unexpected partners, producing chromosomal rearrangements. Besides, persisting DSB may lead to chromo- some fragments loss during anaphase originating micronuclei. Unrepaired or mis-repaired DSB could lead to genomic or chromosomal instability, senescence, malignant transformation or cell death [1] [2] . In eukaryotes, a hierarchical and timely regulated cascade of events called DNA damage response (DDR) has evolved which in- cludes the recruitment of signaling, checkpoint control, DNA repair and chromatin remodeling protein com- plexes. DDR allows cell cycle delay, DNA repair or induce cell dead by recruiting proteins of the apoptotic cas- cade [3] . An early DDR event engages the phosphorylation of the histone variant H2AX involving the oxygen in the g-position of serine 139 in the C-terminal consensus serine-glutamine (SQ) motif (H2AX139 phosphorylation) giving rise to the named gH2AX [4] .
1.1. H2AX139 Phosphorylation Related to DNA Damage Response
In response to DSB, Ataxia Telangiectasia Mutated (ATM) and other phosphoinositide-3-Kinase-related protein Kinases (PIKK) family, such as DNA-PK, phosphorylate H2AX139 [4] . gH2AX acts as a recruiting site for the mediator of DNA damage checkpoint protein 1 (MDC1), a DDR-adaptor and signaling amplifier protein com- plex which in turn promotes the recruitment of MRN (via NBS1) and ATM, to again phosphorylate H2AX and MDC1. As a result, MDC1 creates a positive feedback loop for expanding H2AX phosphorylation bi-direction- ally up to 2 Mbp away [5] , amplifying DNA damage signaling [6] . The H2AX139 phosphorylation recruits checkpoint and repair proteins at DSB sites but also ubiquitin ligases (RNF8/UBC13 and RNF168/UBC13), chromatin remodeling complexes (i.e. 53BP1), histone acetyltransferases and cohesins that may prevent the dis- sociation of broken ends at DSB sites [7] [8] .
Employing anti-gH2AX immunolabelling, chromatin regions harboring phosphorylated H2AX139 can be de- tected in interphase nuclei as individual foci. gH2AX foci increase from 1 - 2 min up to 30 min after ionizing radiation, followed by a slow decline [5] . Foci are completely formed 10 min post-exposure and contain DSB repair factors indicating active DNA repair [9] . The bulk of foci disappear at 8 h post-irradiation, indicating DSB restoring [5] . The disappearance of damage induced-gH2AX foci could result either from gH2AX removal through histone exchange [10] or direct serine 139 dephosphorylation by protein phosphatases (including PP2A, PP4 and WIP1) [11] [12] .
Interestingly, a number of gH2AX foci can be detected 48 h to 7 days after >1 Gy irradiation being attributed to incomplete repair of complex DSB, persistent chromatin alterations or apoptosis [13] [14] . Co-immunolabel- ling of gH2AX and DDR proteins has revealed that gH2AX foci temporarily change in composition due to re- cruitment or dissociation of proteins throughout the cell cycle. For instance, NBS1, MRE11 and 53BP1 disso- ciate from gH2AX foci at the G2/M transition to re-associate at early G1, whilst MDC1 remains throughout the cell cycle [15] [16] .
Checkpoint pathways activated during the G1/S and G2/M cell cycle transitions prevent progression of cells bearing DNA damage to mitosis [17] . In cases of checkpoint failure, the antephase checkpoint, operating subse- quently (between late G2 and mid prophase), can delay mitosis or reverse mitotic progression as well [18] [19] . Once these cell cycle checkpoints are passed, mammalian cells go through mitosis, even in the presence of DNA damage. A still poorly understood gH2AX immunostaining has been observed in mitotic chromosomes, proba- bly related to these checkpoint pathways [16] [20] - [22] , serving to facilitate DDR activation and accelerate DNA damage repair in the novel cell generation [22] [23] .
1.2. H2AX139 Phosphorylation Associated to Apoptotic Cascades
Apoptotic programmed cell death represents a multistep pathway involved in cell death during development, senescence or DNA and cellular stresses. gH2AX can be required for DNA fragmentation during apoptosis, be- ing phosphorylated concomitantly with the initial formation of high molecular weight DNA fragments, although before internucleosomal DNA fragment production [24] [25] . H2AX may function in the crosstalk between cell survival and cell death in response to DNA damage depending on the balance between serine 139 phosphoryla- tion and the constitutively tyrosine 142 phosphorylation by WSTF kinase. During DDR, dephosphorylation of - tyrosine 142 (by EYA1-3) is necessary for MDC1 binding. If tyrosine 142 dephosphorylation is not accom- plished, the cellular response shifts to apoptosis with recruitment of pro-apoptotic factors [26] .
In the case that the G1/S, G2/M or antephase checkpoint systems fail, cells can activate the death pathways during the mitotic process or also after mitotic exit [17] [27] . Since the same result was obtained when BLM- treated cells were exposed to the G2/M checkpoint inhibitor UCN-01 (a CHK1 inhibitor) post-mitotic death would not be the consequence of G2/M checkpoint failure [27] . This result suggests that persisting DNA dam- age during mitosis can enhance a delayed post-mitotic cell death program.
Moreover, stimuli apart from DNA damage can trigger the extrinsic apoptotic pathway with H2AX139 phos- phorylation. For instance, the tumor necrosis factor (TNF) superfamily originating the death-inducing signaling complex (DISC) at the cell surface [28] and the TNF-related apoptosis-inducing ligand TRAIL (a member of the TNF/death receptor superfamily) can induce the apoptotic pathway. Also, it was reported that TRAIL can also produce DDR with ATM, DNA-PK and H2AX139 phosphorylation after 1 h of TRAIL exposure, without induc- tion of DNA damage [29] .
1.3. Scheduled or Cell Cycle Related H2AX139 Phosphorylation
H2AX139 is phosphorylated also in undamaged cells. A foci population reminiscent of large IR-induced foci that co-localized with many DSB repair proteins and a main population of gH2AX small foci that did not associate with repair proteins were documented in untreated asynchronic cultures of mammalian cells [21] . Large gH2AX foci may represent spontaneous DSB, while small foci are DSB-independent. Signal intensity of large and small foci increased through cell progression from G1 to S and G2 phases, reaching the highest intensity during mitosis, especially during metaphase [21] .
Scheduled H2AX139 phosphorylation originates when temporary DSB generate in the V(D)J and class-switch recombination during immune system development and also at sites of DSB formation in meiosis [30] . Addi- tionally, in somatic cells lacking telomerase, where shortened telomeres behave as DSB, H2AX139 is phospho- rylated as well [31] . A cell cycle linked H2AX139 phosphorylation, particularly takes place when DNA polyme- rase or RNA polymerase II stall at spontaneous DNA base alterations (mismatches, modified bases, abasic sites) inducing SSB during replication or transcription processes, respectively. In addition, when the resealing of DNA topoisomerase I or II cleavable complexes are blocked by collision with DNA polymerase or RNA polymerase II, producing SSB or DSB, H2AX can be also phosphorylated at serine 139 [32] .
A cell cycle related H2AX139 phosphorylation, in the absence of induced DNA damage, occurring during mi- totic chromatin structural changes, was recently proposed. The status of H2AX139 phosphorylation, during pre- mature chromosome condensation (PCC) process induced by calyculin A was investigated in both human leu- kemic (HL60) and pulmonary carcinoma (A549) cells [33] . Interestingly, untreated A549 cells showed phosphorylated H2AX139 while non-exposed HL60 cells did not display gH2AX signals. The authors reported a se- quential phosphorylation, first on serine 10 of histone H3 and then on serine 139 of H2AX. The disparity in H2AX phosphorylation observed in different cell types suggests that it may be not a common attribute of mitotic chromosomes or a mandatory feature of chromatin condensation during mitosis [33] .
It has been reported that some metaphases from untreated asynchronously growing cell cultures exhibit gH2AX signals all along the chromosomes. However, the yield and distribution of gH2AX signals was not ana- lyzed so the extension of this phenomenon remains unclear. On the other hand, it is well established that clasto- genic agents such as X-rays and bleomycin (BLM) induces H2AX139 phosphorylation during interphase.
In this work, we carried out a quantitative analysis of gH2AX signals on chromosomes of cytocentrifuged control and BLM-treated metaphase spreads in order to characterize and compare gH2AX signal distribution (per metaphases and chromosomes) in the absence (basal gH2AX signals) and presence of DNA damage (BLM- induced gH2AX signals). Chromosomes were divided in three regions (proximal, medial and distal to the cen- tromere) to test a differential sensitivity to H2AX139 phosphorylation along the chromosomes. Besides, under the hypothesis that chromatin structure influences H2AX139 phosphorylation, we selected eight CHO9 chromosomes harboring homologous regions (located in normal or rearranged chromosomes) to test if homologous regions harbor similar gH2AX signal distributions. In the same sense, one of the chromatin structure signatures, the pro- files of H4 hyperacetylation was compared to gH2AX signal profiles.
2. Methodology
2.1. Selection of CHO9 Chromosomes
The CHO9 cell line is a subclone of CHO (Chinese hamster ovary) cells [34] . It is characterized by 21 chromo- somes, containing nine original chromosomes (1, 2, 5, 7, 9, 10, two 8 and one X) plus twelve elements (Z1-Z10, Z12 and Z13) [35] originated by rearrangements occurring during the cell line transformation. To analyze gH2AX signals we selected eight chromosomes (1, Z1, 2, Z2, Z3, Z4, 5, Z6) which can be paired according to the presence of five homologous regions as a result of rearrangements, as follows: 1-Z1, 1-Z6, 5-Z6, 2-Z2; Z3-Z4. Figure 4 shows DAPI-stained images and idiograms of these chromosomes; identical color coded bars indicate the partners of each homologous region: 5-Z6 (orange), 1-Z6 (light blue), 1-Z1 (violet), 2-Z2 (green), and Z3-Z4 (red).
2.2. Cell Culture
CHO9 cells (from Natarajan AT, Leiden) were cultivated in 95 mm Petri dishes with Ham-F12 (PAA E15-890) culture medium supplemented with 10% fetal calf serum (PAA A15-151), 200 mM glutamine and antibiotics (100 U/mL penicillin and 125 mg/mL dihydrostreptomycin sulfate) (Sigma) at 37˚C in a 5% CO2 incubator.
2.3. Treatment
Exponentially growing CHO9 cells were treated with 10 µg/mL of the radiomimetic agent bleomycin (BLM) at 37˚C and 5% CO2, during 30 min. Next, the culture medium was removed and the monolayer washed with PBS. The medium was centrifuged for 5 min at 800 rpm to collect mitotic cells which were resuspended in complete medium and incorporated into the culture. Immediately, 0.16 µg/mL Colcemid (Ciba) was added for 60 min at 37˚C and 5% CO2. Considering the recovery time of 90 min after BLM-treatment, the cells recovered at meta- phase stage were in the G2 phase or in early stages of mitosis during BLM exposure.
2.4. Metaphase Spreads
After colcemid exposure, mitotic cells were recovered, centrifuged for 5 min at 1200 rpm, re-suspended in 75 mM KCl at 37˚C at a final cell concentration of 5 × 106 cells/mL. After 10 min, the cell suspension was cyto- centrifuged onto slides at 2000 rpm during 10 min, using a Cellspin I cytocentrifuge (Tharmac GmbH) to obtain metaphase chromosome spreads.
2.5. gH2AX Immunolabeling
Chromosomal preparations were fixed (2% paraformaldehyde, 5 min), permeabilized (0.5% Triton X-100, 5 min) and blocked (2% BSA, 15 min). Subsequently, slides were incubated with mouse monoclonal anti-gH2AX (Ab- cam, 1:500 in 2% BSA) for 30 min. After washing in 2% BSA, the preparations were incubated with rabbit Alexa 488-conjugated anti-mouse antibodies (Molecular Probes, 1:250 in 2% BSA) for 30 min. Finally, nuclei were counterstained (10 min) with 1.5 µg/mL DAPI (Sigma-Aldrich) and mounted in Vectashield (Vector La- boratories).
2.6. Image Acquisition and Analysis
Metaphase spread images from controls (n = 40) and BLM-exposed (n = 40) cultures were captured using an Axioplan Mot II fluorescence microscope (Zeiss) with 100x Neofluar phase contrast objective (N.A. 1.30) and FITC/DAPI filters. A scanning digital monochromatic camera (Metasystems CV-M4 + CL) and ISIS program (Metasystems) were employed for image acquisition with a fixed exposure time.
Using the Adobe Photoshop CS program, chromosome images of captured metaphases were karyotyped based on relative size, centromere position and DAPI banding patterns, to identify chromosomes 1, 2, 5, Z1, Z2, Z3, Z4 and Z6. Images of each chromosome either from control (n = 40) or BLM-treated cells (n = 40) were ar- ranged horizontally side by side to the corresponding 229 CHO9 G-band ideogram [36] and their sizes adjusted to fix the respective ideograms. Chromosomes pertaining to the same metaphase were ordered vertically. Paral- lel lines were drawn to indicate on the images of DAPI/gH2AX stained chromosomes the regions matching the G-band ideogram.
2.7. Data Processing and Statistical Analysis
Based on our experimental design, BLM-damaged post-replicating cells were recovered 90 min after treatment as well as post-replicating control cells. Therefore, H2AX 139 phosphorylation signals on each chromatid were considered as independent events. The absence of gH2AX signal in a band was scored as 0, a signal in only one chromatid band as 1 and parallel signals in both chromatids as 2. The first scores of gH2AX signals were re- ferred to the G-bands of chromosomes 1, 2, 5, Z1, Z2, Z3, Z4 and Z6. Then, each chromosome was divided into three segments related to the centromere: proximal, medial or distal (PMD) regions of equal lengths. A score of gH2AX signals was also assigned to each PMD region through the addition of individual band scores. To enable comparisons, the scores were standardized by division between the relative lengths and the number of bands of each PMD region. Analogously, standardized scores of gH2AX signal on whole 1, 2, 5, Z1, Z2, Z3, Z4 and Z6 chromosomes were obtained. Likewise, a standardized score of gH2AX signal was assigned to each homologous region harbored in the eight selected CHO9 chromosomes (1-Z1, 2-Z2, Z3-Z4, 5-Z6 and 1-Z6). In addition, a raw band score adjusted by relative band length was added, band by band, to obtain a gH2AX signal profile along each chromosome. gH2AX signal score distribution in metaphases from control and BLM-treated cells of each whole 1, 2, 5, Z1, Z2, Z3, Z4 and Z6 chromosome, PMD and homologous chromosome regions were graphically represented, indicating 25th percentile (p25), below which were 25% of gH2AX signals, 50th percen- tile (p50), the median of gH2AX signal distribution, and 75th percentile (p75) below which were 75% of gH2AX signals. Since the variables (gH2AX signal scores of bands, PMD regions and chromosomes) did not follow a normal distribution (Shapiro-Wilk test) or showed homogeneity (Levene test), a more demanding and robust significance level (α = 0.001) was chosen when BLM-treatment, chromosomes, PMD and homologous regions were compared with ANOVA and post hoc Tukey or t-test. The non-parametric sign-test for paired samples was also applied to contrast medians of gH2AX signal distribution between the chromosomes, PMD and homologous regions.
3. Results
3.1. Distribution of gH2AX Signals in CHO9 Metaphases
As illustrated in Figure 1, metaphases of both control and BLM-treated CHO9 cells showed gH2AX signals. The regions of gH2AX signals observed in metaphase chromosomes were generally large, involving several chromosome bands. Quantification of signals along 1, 2, 5, Z1, Z2, Z3, Z4 and Z6 chromosomes in control (n = 40) and BLM-treated cells (n = 40) revealed a high dispersion in gH2AX signal distribution, ranging from no signal to nearly whole chromosome immunostaining.
![]()
Figure 1. Examples of cytocentrifuged CHO9 metaphases of three control (A) and three BLM-exposed (B) cells showing gH2AX signals (green) on DAPI-counters- tained chromosomes (merged images). Bars = 5 µm.
To characterize the chromosome gH2AX signal distribution, the percentiles p50 (median), p25 and p75 were calculated (Table 1 and Figure 2). Although chromosomes of controls showed gH2AX signals, BLM-treatment significantly increased H2AX139 phosphorylation. The percentiles p25, p50 and p75 of gH2AX signal distribu- tion were higher in BLM-exposed than in control CHO9 cells (Table 1 and Figure 2). Besides, the overall gH2AX signal distribution of control and BLM-exposed cells showed significant differences (ANOVA p < 0.001).
3.2. Distribution of gH2AX Signals along Different CHO9 Chromosomes
Figure 2 illustrates the distribution of gH2AX signals in the eight CHO9 selected chromosomes. The compari- son of gH2AX signal distribution of each chromosome between controls and BLM-exposed cells showed signifi- cant differences (Tukey test, p < 0.001). The median of gH2AX signal distribution of each chromosome of con- trols was significantly lower than those of the BLM-exposed cells, highlighting that BLM-treatment increased H2AX139 phosphorylation in all analyzed chromosomes. Within the control population, there were no differenc- es between gH2AX signals of each chromosome (Tukey test, p < 0.001 and sign test, p < 0.0001). Nevertheless, in BLM-treated cells, non-parametric sign test revealed that the median of gH2AX signal in Z6 chromosome was significantly higher compared to other chromosomes (p < 0.0001). Furthermore, the medians of gH2AX signal distributions in CHO9 chromosomes showed a tendency to decrease as follows: Z6 > Z2 > 2 > 5 > 1 > Z4 > Z1 > Z3 (sign test, p < 0.0001).
3.3. Distribution of gH2AX Signals in PMD Chromosome Regions
Proximal, medial and distal (PMD) chromosome regions of equal lengths were assigned for each chromosome in order to analyze the influence of centromere or chromosome end proximities in the distribution of gH2AX sig- nals. Figure 3 shows the distribution (p25 to p75) and the median of gH2AX signals in control and BLM-treated cells. ANOVA (p < 0.001) and Tukey (p < 0.001) tests demonstrated that BLM-treatment significantly increased gH2AX signals in each region compared to controls. In the control population, the gH2AX signal distribution
![]()
Table 1. Sample sizes and parameters of gH2AX signal distributions in control and bleomycin-treated CHO9 cell popula- tions.
BLM: bleomycin-treated CHO9 cells. n: pooled number of: metaphases; chromosomes 1, 2, 5, Z1, Z2, Z3, Z4 and Z6 (8 chromosomes × 40 meta- phases); proximal, medial and distal (PMD) chromosome regions (3 PMD regions × 8 chromosomes × 40 metaphases); chromosome bands scored in controls or BLM-exposed CHO9 cells (229 bands/metaphase × 40 metaphases). p25: 25th percentile, p50: 50th percentile (median values), and p75: 75th percentile of gH2AX chromosome signal distributions.
![]()
![]()
Figure 2. Bar-graph representing the distribution of gH2AX signals on CHO9 chromosomes of control and n = 40 BLM-treated (BLM) cells. Vertical grey bars represent the percentile range (p25 to p75) of gH2AX signal distribution on each chromosome. The median (p50) of the distribution is indicated by the horizontal black line on each bar. From left to right: Z3, Z1, Z4, 1, pooled chromosomes (T), 5, Z2, 2 and Z6.
was similar among PMD regions of the eight chromosomes (Tukey test, sign-test). In BLM-exposed cells, proximal regions revealed significant differences in gH2AX signal distribution (sign-test, p < 0.0001), exhibiting less H2AX139 phosphorylation than medial or distal regions. Moreover, a tendency to higher H2AX139 phospho- rylation in the distal chromosome regions was observed (Figure 3).
3.4. gH2AX Signal Distribution between Homologous Regions of CHO9 Chromosomes
The homologous regions of the eight rearranged CHO9 chromosomes are shown in Figure 4. Distinct color bars illustrate the location and extension of the homologous chromosome regions. Partners of each homologous re- gion are depicted with the same color bars.
![]()
![]()
Figure 3. Bar-graph representing the distribution of gH2AX signals considering pooled proximal, medial and distal (PMD) regions of CHO9 chromosomes from control and BLM-treated (BLM) cells. Vertical grey bars represent the percentile range (p25 to p75) of gH2AX signal distribution on each PMD chromosome region. The median (p50) of the distribution is indi- cated by the horizontal black line on each bar. From left to right in both controls and BLM: pooled distal, medial and proxi- mal chromosome regions.
![]()
Figure 4. Homologous regions in CHO9 rearranged chromosomes. Images of DAPI-stained (blue) 1, 2, 5, Z1, Z2, Z3, Z4 and Z6 chromosomes aligned to the respective G-banding ideo- gram are depicted. Color bars at the right side of each ideogram indicate homologous regions between 5 and Z6 (orange), 1 and Z6 (light blue), 1 and Z1 (violet), 2 and Z2 (green), and between Z3 and Z4 (red), originated through chromosome rearrangements. Z1 and the acro- centric Z6 derived from a reciprocal translocation (1:5) involving distal regions of 1p and 5q. Chromosome Z2 arose from a 2q deletion. A pericentric inversion of chromosome 3 involving entirely 3p and the proximal portion of 3q generated acrocentric Z4. A translocation (3p; 4p) produced Z3 [35] .
With the aim to test whether homologous regions behaved similarly or not with respect to gH2AX signal dis- tribution, we contrasted them between the homologous regions in each pair of chromosomes (i.e. 1-Z1) and then, we compared these distributions between all five homologous regions, both in controls and BLM-exposed cells. The median and the p25 to p75 range of gH2AX signal distributions in homologous regions (1-Z, 1-Z6, 2-Z2, Z4-Z3 and 5-Z6) of control and BLM-treated cells are shown in Figure 5.
The overall gH2AX signals in homologous regions were significantly higher in BLM-treated than control cells (ANOVA, p < 0.001). Either in controls or BLM-exposed cells, the distribution of gH2AX signals between the homologous regions of each chromosome pair was similar (t-test, p > 0.98). However, gH2AX signals dis- tribution showed significant differences between homologous regions located in distinct chromosome pairs (t-test, p < 0.001; sign-test, p < 0.0001). Homologous regions in 1-Z1 and Z3-Z4 displayed lower H2AX139 phosphorylation than 5-Z6 and 1-Z6. Interestingly, chromosome 1, which harbors both 1-Z1 and 1-Z6 homo- logous regions (Figure 4 and Figure 5), exhibited higher H2AX139 phosphorylation on the latter (t-test, p < 0.001; sign-test, p < 0.0001).
3.5. Comparison between Chromosome gH2AX Signals and Hyperacetylated Regions
Figure 6(A) depicts gH2AX signal profiles of CHO9 chromosomes harboring homologous regions of both con- trols and BLM-exposed cells. Figure 6(B) compares chromosomal gH2AX signal profiles with the correspond- ing H4+ac pattern of BLM-treated cells. Both profiles are referred to G-bands ideograms. The patterns of H4+ac in CHO9 chromosomes were previously established in our laboratory [37] , using an antibody that recognizes di-, tri- and tetra-acetylated lysine 12 of histone H4 (H4K12ac).
As can be observed in Figure 6(B), similarities between gH2AX and H4+ac profiles were observed for each ana- lyzed chromosome. Additionally, a tendency of both profiles to increase towards terminal chromosome regions mainly in BLM-exposed cells was detected.
4. Discussion
In the present work, a quantitative analysis of spontaneous or BLM-induced gH2AX signals on CHO9 metaphase chromosomes was addressed.
4.1. Spontaneous gH2AX Signals Were Observed in Control Metaphases
Since CHO9 control metaphases exhibited noticeable gH2AX signals (Figure 1) one may assume the presence of H2AX139 phosphorylation independent of DNA damage. Cell cycle-dependent H2AX139 phosphorylation in nocodazole synchronized human cells confirmed through gH2AX immunostaining and western blot analysis was reported [20] . gH2AX foci were observed in a small proportion of interphase nuclei but in nearly all mitotic chromosomes, reaching the maximum level at metaphase. However, gH2AX induction in mitotic chromosomes
![]()
![]()
Figure 5. Bar-graph representing the distribution of gH2AX signals in homologous regions of CHO9 chromosomes. Vertical grey bars represent the percentile range (p25 to p75) of gH2AX signal distribution of each homologous chromosome region. The median (p50) of the distribution is indicated by the horizontal line on each bar. From left to right in both controls and BLM: 1 and Z1 (violet), Z3 and Z4 (red), 2 and Z2 (green), 5 and Z6 (orange) as well as 1 and Z6 (light blue) homologous regions.
![]()
Figure 6. (A) gH2AX and H4+ac signal profiles of CHO9 chromosome harboring homologous regions of both controls and BLM-exposed cells. Homologous regions (bars) shared by chromosomes 5 and Z6 (orange), 1 and Z6 (light blue), 1 and Z1 (violet), 2 and Z2 (green), as well as between Z3 and Z4 (red) are shown at the left side of each ideogram. gH2AX signal profiles are illustrated at the right side of each ideogram. (B) gH2AX and H4+ac signal profiles are respectively depicted at the left and right sides of ideograms corresponding to chromosomes 1, 2, Z2, Z4, 5 and Z6. gH2AX signal profiles belong to BLM-treated cells. The chromosome patterns of H4+ac were previously established in our laboratory [37] .
was not associated with CHK2 or p53 signaling protein phosphorylations, suggesting that in this context H2AX phosphorylation is independent of DNA damage response [20] .
In this respect, the role of mitotic chromosome condensation on H2AX139 phosphorylation should be consi- dered. It was reported that mitotic gH2AX induction takes place in parallel with histone H3 phosphorylation at Ser 10 by Aurora kinase [38] . Since H3 phosphorylation at Ser 10 is considered a cytogenetic mark of chromatin condensation during mitosis [39] , it was suggested that mitotic gH2AX could play a role in proper condensation of chromatin [20] . Furthermore, mitotic gH2AX could arise in response to the tensional stress generated during mitotic chromatin remodelling, turning DNA more sensitive to nucleases [40] .
Besides, chromatin condensation along mitosis could explain the presence of large gH2AX signals observed in CHO9 metaphase chromosomes. Large gH2AX signals may arise by recruitment of single gH2AX focus dur- ing chromosome condensation. Since γH2AX signals enlarge on swollen metaphase chromosomes of HeLa and Indian muntjac cells, aggregation of individual gH2AX foci producing a large gH2AX signal may take place [16] .
Furthermore, G2/M and antephase checkpoints are not only sensitive to DNA damage, but also to a variety of insults such as hypothermia, anoxia, osmotic shock or even to non-clastogenic drugs like colcemid or nocoda- zole [41] [42] .
During the experimental procedures, CHO9 cells could be stressed by colcemid treatment, hypothermia or osmotic changes activating pathways involved in H2AX139 phosphorylation. However, since γH2AX signals were previously detected on mitotic chromosomes of asynchronously growing cell cultures not exposed to col- cemid [16] [20] - [23] , this compound may not impact on H2AX139 phosphorylation.
It must be taken into account that a proportion of CHO9 basal gH2AX signals could be related to persisting scheduled H2AX139 phosphorylation due to temporary SSB or DSB induced during replication or transcription [30] [31] [43] . Besides, SSB or DSB produced by endogenous oxidative DNA damage along DNA replication could also contribute to gH2AX signaling. Additional experimental approaches are needed to deepen our under- standing of H2AX139 phosphorylation unrelated to the DNA damage response, especially gH2AX induced dur- ing mitosis.
4.2. BLM-Exposure Increased gH2AX Signals on Metaphase Chromosomes
Although metaphase chromosomes of control cells showed gH2AX immunolabelling, BLM-treatment increased mitotic H2AX139 phosphorylation (ANOVA, p < 0.001). Occurrence of gH2AX signals during mitosis after DNA damage induction has been previously reported [16] [20] [21] leading to the assumption that persisting gH2AX signals may serve to facilitate DDR in the novel cell generation [22] [23] . Thus, the observed gH2AX signal increase in CHO9 metaphases treated with BLM could correspond to a DDR related H2AX139 phospho- rylation persisting through mitosis.
Further, when chromosomes were exposed to g-rays or treated with adriamycin during mitosis, strong gH2AX signals as well as elevated rates of segregation and cytokinesis failures were observed [44] [45] . This could mean that H2AX139 phosphorylation not only occurs in interphase nuclei, but also could arise during mitosis.
4.3. Spontaneous and BLM-Induced gH2AX Signals Varied between Metaphases
High dispersion in the overall level of gH2AX signals between CHO9 metaphases was observed ranging from complete absence to nearly whole immunostained metaphases (Figure 1). Several factors competent for H2AX139 phosphorylation could combine, inducing a particular level of gH2AX signal in a metaphase. For instance, ex- tensive metaphase immunostaining could result from BLM-exposure plus a contribution of endogenous SSB and DSB, originated during DNA metabolism, oxidative DNA damage or chromatin winding. By contrast, if few or no factors interact, gH2AX signals will be respectively low or absent.
Besides, a mitotic-specific amplification of gH2AX preceding the apoptotic inter-nucleosomal DNA frag- mentation has been detected in damaged cells starting mitosis [27] . Considering this finding, the overall chro- mosome gH2AX immunostaining observed in CHO9 metaphases (Figure 1) could correspond to an amplified pre-apoptotic H2AX139 phosphorylation process.
Our results are in agreement with the assumption that mitotic gH2AX immunostaining is not a general phe- nomenon. In this sense, variability in the extent of H2AX139 phosphorylation during mitosis among distinct cell lines (human HeLa and neuroblastoma, green monkey SV40 transformed kidney, C3H mouse embryo and In- dian muntjac skin primary cultures) has been reported previously [21] . These observations indicate that H2AX phosphorylation is not a common feature of mitotic cells and that may be not indispensable for chromatin con- densation during mitosis [33] . Further experimental approaches will elucidate the grounds of mitotic H2AX phosphorylation.
4.4. BLM-Induced gH2AX Signals Differed between Chromosomes and PMD Regions
gH2AX signals unrelated to BLM-exposure distributed uniformly among CHO9 chromosomes and PMD regions (t-test, p > 0.001; sign-test, p > 0.0001). This finding might indicate that basal gH2AX signals occur regularly per band unit. Conversely, BLM-induced gH2AX signals varied between chromosomes as follows: Z6 > Z2 > 2 > 5 > 1 > Z4 > Z1 > Z3 (sign-test, p < 0.0001) (Figure 2). Besides, BLM-induced gH2AX signals varied also be- tween PMD regions. Chromosome segments close to centromeres showed low gH2AX signaling (sign-test, p < 0.0001) with a tendency to increase towards distal regions (Figure 3).
This uneven gH2AX signal distribution between chromosomes and chromosome regions is in agreement with the inter- and intra-chromosomal variability of the induced-damage reported by researchers of our laboratory in CHO cells [46] - [52] and others [53] - [61] both in human and in other mammalian cells.
4.5. Spontaneous and BLM-Induced gH2AX Signals Were Similar in CHO9 Homologous Chromosome Regions
Homologous regions are homologous chromosome segments among CHO9 chromosomes originated by rear- rangements occurring during the cell line evolution. An equivalent gH2AX signal distribution in control and BLM-trated cells was detected between the partners of homologous regions (t-test, p > 0.98). Since chromatin features through homologous regions may be indistinguishable, the influence of the same chromatin structure in a particular chromosome region independently of the position in the karyotype could be put forward to explain this observation.
Chromosome regions of higher condensed chromatin (owing to compaction or protein coating) may be less accessible to chemical DNA damaging agents like BLM, and also less available to kinases related to H2AX139 phosphorylation. Conversely, less condensed chromosome regions may be more sensitive to chemical agents and H2AX139 phosphorylation kinases [47] [62] [63] . Since it is expected that along homologous regions similar chromatin states are present (i e. early replicating G-light bands, late replicating G-dark bands, heterochromatic constitutive C band, interstitial telomeric sequences) [64] [65] , a comparable response to DNA damage and H2AX139 phosphorylation is envisaged.
On the other hand, homologous regions from distinct chromosome pairs displayed differential sensitivity in H2AX139 phosphorylation related to BLM. Interestingly, different gH2AX signal distributions were also ob- served between distinct homologous pairs belonging to the same chromosome, such as regions 1-Z1 and 1-Z6 of chromosome 1 (Figure 4 and Figure 6). The distal 1-Z6 region revealed more H2AX139 phosphorylation than 1-Z1. 1-Z6 homologous region is located distally on chromosome 1 and 1-Z1 expands through proximal, medial and distal segments of this chromosome. Therefore, the different sensitivity of chromatin states to damaging agents could explain the difference in gH2AX signaling between 1-Z1 and 1-Z6. In this respect, it was reported that regions with dissimilar chromatin organization may exhibit distinct sensibilities to chemical clastogens [47] [62] .
Furthermore, the overall histone acetylation status of distinct interphase chromatin compartments is retained in metaphase chromosomes both in human and CHO9 metaphase chromosomes [66] . Less acetylated chromo- some regions colocalize within G-dark and C-bands. On the contrary, hyperacetylated regions mainly corres- pond to G-light bands and gene-rich human telomeres (T-band regions) [66] - [68] . In this respect, the similarity between gH2AX signal and H4+ac profiles, increasing towards terminal CHO9 chromosome regions (Figure 6(B)) suggests that the organizational and functional state of the chromatin [37] [69] might underlie both types of epi- genetic changes. However, more experimental evidences are required to support this hypothesis.
5. Conclusion
To sum up, gH2AX signals were detected in a high percentage of control metaphases, probably involving endo- genous DNA damage or tensional stress. Nevertheless, it is clear that BLM exerted an effect on H2AX139 phos- phorylation, prevailing towards acetylated and gene-rich distal chromosome segments. Both basal and BLM- induced gH2AX signal distributions were equal between CHO9 homologous chromosome regions. The compa- rable H2AX139 phosphorylation of homologous regions suggests its dependence on chromatin structure or func- tion, being irrespective of the position in the karyotype. Co-immunostaining of gH2AX and proteins related to DDR, repair, apoptosis or chromatin remodeling as well as the induction of changes in chromatin status, could provide more information to understand the biological significance of basal mitotic H2AX139 phosphorylation.
Acknowledgements
We thank PEDECIBA Postgraduate Program and the National Agency of Investigation and Innovation (ANNI) for financial support and G. Folle for critical reading of the manuscript.