Ionizing radiation (IR) and consequent induction of DNA double-strand breaks (DSBs) causes activation of the protein ataxia telangiectasia mutated (ATM). Normally, ATM is present as inactive dimers; however, in response to DSBs, the ATM dimer partners cross-phosphorylate each other on serine 1981, and kinase active ATM monomers are subsequently released. We have studied the presence of both nonphosphorylated as well as active serine 1981 phosphorylated ATM (pS1981-ATM) in the mouse testis. In the nonirradiated testis, ATM was present in spermatogonia and spermatocytes until stage VII of the cycle of the seminiferous epithelium, whereas pS1981-ATM was found only to be present in the sex body of pachytene spermatocytes. In response to IR, ATM became activated by pS1981 cross-phosphorylation in spermatogonia and Sertoli cells. Despite the occurrence of endogenous programmed DSBs during the first meiotic prophase and the presence of ATM in both spermatogonia and spermatocytes, pS1981 phosphorylated ATM did not appear in spermatocytes after treatment with IR. These results show that spermatogonial ATM and ATM in the spermatocytes are differentially regulated. In the mitotically dividing spermatogonia, ATM is activated by cross-phosphorylation, whereas during meiosis nonphosphorylated ATM or differently phosphorylated ATM is already active. ATM has been shown to be present at the synapsed axes of the meiotic chromosomes, and in the ATM knock-out mice spermatogenesis stops at pachytene stage IV of the seminiferous epithelium, indicating that indeed nonphosphorylated ATM is functional during meiosis. Additionally, ATM is constitutively phosphorylated in the sex body where its continued presence remains an enigma.
Ataxia telangiectasia mutated (ATM) is part of a family of phosphatidylinositol 3 kinases that includes the DNA-dependent protein kinase (DNA-PK) and ataxia telangiectasia and RAD3 related (ATR) . Induction of DNA damage triggers activation of ATM, which is involved in most, if not all, cellular responses to DNA double-strand breaks (DSBs) such as apoptosis, cell cycle arrest, and DNA damage repair . In undamaged somatic cells, ATM is present as inactive dimers. However, very shortly after induction of DSBs, for instance by ionizing radiation (IR), the ATM dimer partners cross-phosphorylate each other and are subsequently released as kinase active monomers [2, 3] capable of interacting with DNA and DSBs .
Within minutes after induction of DSBs, a substantial fraction of the nuclear ATM pool relocates to the damaged DNA and phosphorylates histone H2AX [5–7]. This results in the appearance of γ-H2AX (phosphorylated H2AX) foci specifically at the sites of DSBs [8, 9]. γ-H2AX foci function as docking sites for many DNA damage repair proteins and are thus essential for proper DNA damage repair [9–11]. Subsequently, the proteins sequestered by γ-H2AX in IR-induced foci are in their turn often also phosphorylated and activated by ATM that remains present at the sites of damage [1, 12]. One of the most important substrates of ATM is the tumor suppressor p53 [13–16], which, in spermatogonia, plays a central role in DNA damage-induced spermatogonial apoptosis [17, 18].
DSBs are predominately repaired in two ways. First, a DSB can be repaired by error-prone nonhomologous end joining (NHEJ), during which two double-stranded DNA strands are just bluntly attached. Second, there are the much more accurate homologous recombinational repair pathways (HRR) that use the homologous DNA strand as a template for repair , leading to gene conversion, recombination, single-strand annealing, or break-induced replication . Whereas the role for the ATM family member DNA-PK seems restricted to NHEJ, ATM itself is essential for HRR  and activates many proteins involved in this process such as the RAD50/MRE11/NBS1 complex  and BRCA1 and 2 [22–25].
In the testis, HRR plays an especially important role as DSBs are endogenously induced during the first meiotic prophase . These meiotic DSBs are essential for the meiotic process and are repaired by meiotic recombination, which is a meiosis-specific variant of HRR . ATM localizes at the sites of meiotic DSBs, and recombination  and nonfunctional ATM results in meiotic arrest during the meiotic prophase I [29–31] due to abnormal chromosomal synapsis and subsequent chromosome fragmentation  combined with altered interactions of telomeres with the nuclear matrix and distorted meiotic telomere clustering .
We have studied expression and localization of ATM in the testis before and after treatment with IR. Furthermore, using an antibody that specifically recognizes cross-phosphorylated ATM , we were able to compare ATM activation in response to IR with the presence of endogenously active ATM during meiosis. Finally, using ATM−/− testes, we studied in more detail the exact epithelial stage at which spermatogenic apoptosis occurs during the first meiotic prophase.
Materials and Methods
Animals, Irradiation, and Fixation
Testes of groups of four male FvB/NAU mice of at least 7 weeks of age were irradiated and fixed in 4% phosphate-buffered formaldehyde (pH 6.6–7.2) for 4 h and postfixed in a diluted Bouin solution (71% picric acid [0.9%], 24% formaldehyde [37%], 5% acetic acid) for 16 h at room temperature. One testis of each mouse was fixed at various time intervals after a dose of 4 Gy of 200 kV x-rays (Philips, Eindhoven, The Netherlands; 200 kV, 20 mA, 0.5 mm Cu filter) using the contralateral testes for protein isolation. The animals were used and maintained according to the regulations provided by the animal ethical committee of the University Medical Center Utrecht, which also approved the experiments. ATM−/− mice were generated as described .
Immunohistochemistry was performed as described previously [34–36]. Five-micrometer paraffin sections of testes taken at different intervals after irradiation were mounted together on a TESPA-coated glass slide and dried overnight at 37°C. Sections were dewaxed in xylene and hydrated in a graded series of alcohols. Subsequently, the sections were boiled three times for 10 min in 0.01 M sodium citrate using a microwave oven (H2500, Bio-Rad, Hercules, CA). After this, sections were incubated in 0.35% H2O2 in PBS for 10 min. Blocking occurred in 5% BSA (Sigma, St. Louis, MO)/5% goat or rabbit serum (Aurion, Wageningen, The Netherlands) in PBS. The slides were then incubated with goat polyclonal antibodies against total ATM (C-20, Santa Cruz Biotechnology, Inc., Santa Cruz, CA) or rabbit polyclonal antibodies against phosphorylated ATM (pS1981-ATM, Rockland, Gilbertsville, PA) diluted 1:50 in PBS including 1% BSA in a humidified chamber overnight at 4°C. Incubation with secondary biotinylated rabbit-anti-goat or goat-anti-rabbit (Santa Cruz Biotechnology) IgGs diluted 1:100 in PBS including 1% BSA was performed in a humidified chamber for 60 min at room temperature. The horseradish peroxidase avidin-biotin complex reaction was performed according to the manufacturer’s protocol (Vector Laboratories). Between each step, sections were washed in PBS. Bound antibodies were visualized using 0.3 μg/μl 3,3′-diaminobenzidine (DAB, Sigma) in PBS, to which 0.03% H2O2 was added. Sections were counterstained with Mayer hematoxylin. For negative control sections, one volume of primary antibodies was incubated with five volumes of blocking peptide (Santa Cruz Biotechnology or Rockland, respectively) for 2 h. Adjacent sections were used for a periodic acid Schiff (PAS) staining to identify the stages of the cycle of the seminiferous epithelium. Prior to mounting with Pertex (Cellpath Ltd., Hemel Hempstead, GB), sections were dehydrated in a series of graded alcohols and xylene.
Western Blot Analysis
Western blot analysis was performed as described previously [34–36]. Total protein lysates were prepared by homogenizing the testes in a polytron device (Janke & Kunkel GmbH, Staufen, Germany), after which the cells were lysed in RIPA buffer (PBS with 1% NP40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM phenylmethanesulphonylfluoride [PMSF], 10 μg/ml leupeptin, 10 μg/ml aprotinin, 2 mM Na3VO4, 1 mM Na2MoD4, and 10 mM NaF) for 30 min on ice. Lysates were sonicated on ice and cleared by centrifugation. Protein levels were measured using BCA analysis (Pierce Chemical Co., Rockford, IL), and 50 μg of protein was loaded in each lane. SDS-page was performed as described by Laemmli . Proteins were blotted onto a polyvinylidene difluoride (PVDF) membrane (MilliPore, Bedford, MA).
Western blots were blocked using Blotto-A containing 5% Protifar (Nutricia, Zoetermeer, The Netherlands) in Tris-buffered saline (10 mM Tris-HCl, pH 8.0; 150 mM NaCl), including 0.05% Tween-20 (TBT), and washed in TBT in between each step. First antibodies (see Immunohistochemistry) were diluted 1:1000 in Blotto-A. After incubation with secondary antibodies conjugated to horseradish peroxidase (DAKO A/S, Glostrup, Denmark) and diluted 1:5000 in Blotto-A, the antigens were visualized using chemiluminescence (ECL, Amersham Pharmacia Biotech Benelux, Roosendaal, The Netherlands) and exposure to x-ray film (RX, Fuji Photo Film Co., Ltd., Tokyo, Japan).
Localization of ATM in the Testis Before and After X-Irradiation
In order to investigate ATM localization during spermatogenesis, sections of mouse testes fixed before and at various time points (0–24 h) after irradiation were immunohistochemically stained using a polyclonal antibody against ATM (Fig. 1, A and B). In the normal adult testis, all germ cells showed a granular staining in the cytoplasm. Type A spermatogonia also showed a light granular staining in the nucleus, which became more pronounced and evenly spread in intermediate spermatogonia. From B spermatogonia onward to pachytene spermatocytes in stage VII, a pronounced nuclear staining was observed. No nuclear staining was observed in later spermatogenic cell types. Neither the myoid cells nor the Sertoli cells showed any staining. The Leydig cells, however, showed a heavy granular staining in the cytoplasm but no staining in the nucleus.
No changes in ATM localization were found in response to IR. Negative controls for both irradiated and nonirradiated testes were performed using an antibody-specific blocking peptide. No staining, apart from a vague cytoplasmic staining throughout the testis and signal in the residual bodies, could be observed in any of the negative controls.
To investigate whether ATM levels in the mouse testis change after irradiation, a Western blot analysis was performed on whole testis lysates of normal and irradiated testes (Fig. 1C). No change in ATM levels was found in response to irradiation.
ATM Activation in the Testis
To study when and where during spermatogenesis ATM gets activated, we used an antibody that specifically recognizes ATM that is phosphorylated at serine 1981 (pS1981-ATM) . In contrast to the staining pattern found using antibodies that recognize total ATM, phosphorylated ATM appeared only to be present in the sex body of pachytene spermatocytes (Fig. 2, A and B).
However, after irradiation active, pS1981 phosphorylated ATM was also observed in the nuclei of all spermatogonia and was very pronounced in Sertoli cells (Fig. 2, C and D). However, no additional nuclear staining beyond the already present staining of the sex body appeared in pachytene spermatocytes. This staining pattern was found at all time points after irradiation up until 24 h.
Again, the negative controls performed using an antibody-specific blocking peptide showed no staining apart from a vague cytoplasmic staining throughout the testis and occasional cytoplasmic “dots” in Leydig cells, elongated spermatids, and a stronger signal in residual bodies.
The appearance of pS1981 phosphorylated ATM in spermatogonia and Sertoli cells, as shown using immunohistochemistry, could also be visualized using Western blot analysis of total testis lysates (Fig. 2E). Phosphorylated ATM was found to increase in response to IR with a maximum between 3 and 6 h after irradiation.
Stage IV Pachytene Arrest in ATM−/− Testes
In order to investigate in more detail at which exact stage spermatogenic arrest occurs during the meiotic prophase in the absence of ATM, we studied sections of ATM−/− testes. Spermatogenesis in the ATM−/− testis did not proceed any further than pachytene spermatocytes in stage IV of the cycle of seminiferous epithelium (Fig. 3). Prior to this point, spermatogenesis appeared normal. However, during stage IV, distinguished by the division of intermediate into B spermatogonia, massive apoptosis of spermatocytes occurred, and no spermatocytes were observed in subsequent stages.
In spermatogonia, ATM is present before and after irradiation. However, using an antibody specific for serine 1981 phosphorylated ATM, we did not find activated ATM in these cells. Only after irradiation did ATM become phosphorylated on serine 1981 in spermatogonia and Sertoli cells (Fig. 4). Hence, as described for cultured somatic cells , in spermatogonia and Sertoli cells “dormant” ATM dimers most likely get activated by auto-cross-phosphorylation. Subsequently, activated ATM has been shown to phosphorylate histone H2AX [5–7], which forms the previously described γ-H2AX foci at the sites of DSBs . These foci then recruit DNA repair and damage response proteins , including p53, which subsequently also is activated by ATM [13–16]. Finally, in spermatogonia, p53 can induce apoptosis or cell cycle arrest [17, 18].
Also in Sertoli cells, irradiation induces p53 and serine 1981 ATM phosphorylation. However, these cells do not undergo apoptosis in response to irradiation and are not visibly affected by the absence of p53  or ATM. In the adult mouse, Sertoli cells normally do not divide. This could be a reason for these cells to repair damaged DNA instead of undergoing cell-cycle arrest or apoptosis. The fact that both p53 and ATM are not essential for Sertoli survival indicates that these proteins are merely involved in stress responses of these cells.
Although nonphosphorylated ATM is present in spermatocytes until stage VII, we only find phosphorylated ATM in the sex body of these cells (Fig. 4). This is surprising since ATM plays an active role during the first meiotic prophase [28, 32, 33, 38, 39]. During the meiotic prophase, DSBs are endogenously induced, an event necessary for meiotic recombination [26, 27]. During normal homologous recombination, ATM plays an essential role [1, 19, 20, 40–50], activating many proteins like the RAD50/MRE11/NBS1 complex  and BRCA1 and 2 [22–25]. Therefore, it is not surprising that when studying surface spreads of spermatocytes, ATM also localizes at the sites of meiotic DSBs and recombination . In mice deficient for ATM, apoptosis of spermatocytes occurs at the zygotene/pachytene stage of prophase I as a result of abnormal chromosomal synapsis and subsequent chromosome fragmentation . In addition, these mice exhibit altered interactions of telomeres with the nuclear matrix combined with distorted meiotic telomere clustering . In line with these results, we found that in ATM knock-out mice spermatogenesis stops at pachytene stage IV of the cycle of the seminiferous epithelium, precisely at the stage at which a pachytene checkpoint has been described . At this epithelial stage, spermatogenesis is known to be interrupted in several transgenic mice, such as MSH5−/− mice, that have a spermatogenic phenotype similar to ATM−/− , and DNA-PKcs deficient scid mice, in which many spermatocytes fail to pass epithelial stage IV and undergo apoptosis .
Probably during the prophase of meiosis, ATM does not need to be phosphorylated in order to be active, or perhaps it is phosphorylated at a different site. In short, although not being phosphorylated at serine 1981, ATM is clearly active during the first meiotic prophase as illustrated by its localization on the meiotic chromosomes  and the phenotype of the ATM−/− mice. This is in contrast to ATM activation by pS1981 phosphorylation during DNA repair of IR-induced DSBs in irradiated spermatogonia and other mitotic cells .
The staining pattern of pS1981 phosphorylated ATM in the testis, before and after irradiation, coincides almost exactly with that of p53 (Fig. 4) [17, 35]. The main difference is that p53 is not present in the sex body of spermatocytes, which contains the X and Y chromosomes. Although they are both induced in irradiated spermatogonia, neither p53 nor pS1981 phosphorylated ATM are induced in spermatocytes in response to IR [17, 35]. Moreover, whereas in spermatogonia p53 plays a central role in the induction of IR-induced apoptosis [17, 18], the apoptotic elimination of spermatocytes with synaptic errors has been described to be p53-independent . Hence, although active during meiotic recombination [28, 32], after irradiation ATM in spermatocytes still remains nonphosphorylated on pS1981 and does not activate p53. Additionally, in response to IR, clear γ-H2AX foci appear in pachytene spermatocytes (Fig. 4) . The fact that ATM remains nonphosphorylated on pS1981 in these cells, even after irradiation, indicates that ATM phosphorylation at serine 1981 is not required for γ-H2AX foci formation in meiotic cells. Most likely ATM is required but activated differently.
The presence of pS1981 phosphorylated ATM in the sex body of spermatocytes is intriguing. This presence coincides with the presence of γ-H2AX (Fig. 4), which is not only present in the sex body [26, 35], but also essential for its formation and functionality [54, 55]. In the H2AX−/− testis, pairing of the sex chromosomes fails and they become fragmented or associated with autosomal chromosomes . Additionally, in H2AX−/− spermatocytes, meiotic sex chromosome inactivation is not initiated, and several sex body-specific proteins are not recruited . It is very likely that, as is the case for damaged DNA [5–7], ATM is also responsible for histone H2AX phosphorylation in the sex body. Furthermore, in the sex bodies, ATM could also be responsible for activation of BRCA1 [22–25], a process that is required for γ-H2AX to appear in the sex bodies . Both H2AX and BRCA1 are involved in meiotic X-chromosome inactivation [55–57] and seem to play an as yet unidentified role in keeping the X and Y chromosomes intact during the first meiotic prophase .
In conclusion, ATM is functional during the spermatogonial DNA damage response and during the first meiotic prophase [28, 32, 33, 38, 39]. Whereas in response to irradiation spermatogonial ATM is activated by cross-phosphorylation at serine 1981, during the first meiotic prophase ATM remains nonphosphorylated at this specific site. However, ATM localization on meiotic chromosomes  and stage IV arrest in ATM−/− testes shows that ATM, although not phosphorylated at pS1981, is functional during meiosis. Additionally, pS1981 phosphorylated ATM is present in the sex body of pachytene spermatocytes. In these structures it may be responsible for phosphorylation and activation of proteins including H2AX and BRCA1.