Bax deficiency extends the survival of Ku70 knockout mice that develop lung and heart diseases

Ku70 (Lupus Ku autoantigen p70) is essential in nonhomologous end joining DNA double-strand break repair, and ku70−/− mice age prematurely because of increased genomic instability and DNA damage responses. Previously, we found that Ku70 also inhibits Bax, a key mediator of apoptosis. We hypothesized that Bax-mediated apoptosis would be enhanced in the absence of Ku70 and contribute to premature death observed in ku70−/− mice. Here, we show that ku70−/− bax+/− and ku70−/− bax−/− mice have better survival, especially in females, than ku70−/− mice, even though Bax deficiency did not decrease the incidence of lymphoma observed in a Ku70-null background. Moreover, we found that ku70−/− mice develop lung diseases, like emphysema and pulmonary arterial (PA) occlusion, by 3 months of age. These lung abnormalities can trigger secondary health problems such as heart failure that may account for the poor survival of ku70−/− mice. Importantly, Bax deficiency appeared to delay the development of emphysema. This study suggests that enhanced Bax activity exacerbates the negative impact of Ku70 deletion. Furthermore, the underlying mechanisms of emphysema and pulmonary hypertension due to PA occlusion are not well understood, and therefore ku70−/− and Bax-deficient ku70−/− mice may be useful models to study these diseases.

Ku70 (Lupus Ku autoantigen p70) is a subunit of the Ku protein complex that plays an essential role in the nonhomologous end joining (NHEJ) pathway for DNA double-strand break (DSB) repair (reviewed in Downs and Jackson 1 ). Ku70-null mice have an increased sensitivity to radiation and are defective in lymphocyte differentiation because of a lack of NHEJ-dependent DSB repair. [2][3][4][5] Recent studies have shown that Ku70-null mice exhibit characteristics of premature aging, 6 some of which may arise as a result of the cellular responses to increased DNA damage caused by defective DNA repair. Furthermore, increased neuronal cell death observed during the development of ku70 −/− mice 7 suggests that Ku70dependent DNA repair may be essential for cell survival during normal brain development. There is increasing evidence that Ku70 has multiple biological activities independent of its role in the nucleus as a subunit of the Ku protein complex. 1 Ku70 is known to have anti-apoptotic activity by suppressing the intrinsic cell death pathway 8 mediated by the pro-apoptotic Bcl-2 family of proteins, such as Bax (Bcl-2-associated protein X). 9,10 Ku70-deficient cells have increased sensitivities to apoptotic stresses that are not limited to DNA-damaging stresses, 8,11,12 consistent with its anti-apoptotic activity.
Our previous studies suggest that Ku70 can interact with Bax to inhibit the conformational change required for Baxinduced apoptosis. 8 To date, several studies have shown that regulating the interaction between Ku70 and Bax in the cytosol can influence the cellular sensitivity to various types of stresses. [12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28] Recent studies have also shown that Ku70 can suppress apoptosis by maintaining anti-apoptotic Mcl-1 protein levels through deubiquitinization, 29 suppressing Apaf-1 (apoptosis protease activation factor-1) transcription 30 and inhibiting p18-Cyclin E-induced cell death produced by the caspase-dependent cleavage of cyclin E. 17 To determine the physiological significance of Ku70dependent inhibition of Bax-mediated apoptosis, ku70 −/− bax +/ − and ku70 −/− bax −/− mice were generated and their phenotypes were compared with ku70 −/− bax +/+ mice. In the absence of Ku70, we hypothesized that Bax-mediated apoptosis would be enhanced. We found that Bax deficiency decreased the mortality of ku70 −/− mice and improved the median and maximum age of survival. This study also provides evidence that apoptosis has a significant role in age-associated degenerative diseases in defective DNA repair mutant mice.
Bax deficiency in Ku70-null mice leads to a slight increase in body weight and brain size. Similar to previous reports, 6 ku70 −/− mice displayed growth retardation. However, the average body weight was slightly increased in ku70 −/− bax +/ − and ku70 −/− bax −/− mice (Figures 2a-c). To determine whether Bax deficiency affected specific organs to increase the overall body weight, we measured the weights of the kidneys and brain (Figures 2d and e). We found that the average weights of the kidneys in ku70 −/− bax +/ − and ku70 −/− bax −/− mice were not significantly increased compared with ku70 −/− mice ( Figure 2d). However, the average brain weight was significantly increased in ku70 −/− bax −/− (Po0.0001) mice, and to a lesser extent in ku70 −/− bax +/ − (Po0.05) mice (Figure 2e), even though ku70 −/− bax +/ − and ku70 −/− bax −/− mice have similar extended survival. This suggests that changes in brain size may not necessarily alter lifespan. Baxnull mice are known to have increased brain weight (Figure 2e) because of the suppression of neuronal cell death during development 32 , whereas ku70 −/− mice exhibit increased neuronal cell death. 7 The effects of Bax deletion on the brain weight in Ku70-null mice in this study are consistent with these previous reports.
Previous studies have reported that ku70 −/− mice have decreased lymphocyte counts because of deficient lymphocyte differentiation caused by the absence of NHEJdependent T-cell and B-cell receptor maturation. [2][3][4][5] Our blood cell count measurements were consistent with these findings (Figures 2f-j). The number of neutrophils was increased in ku70 −/− mice (Figure 2g), most likely to compensate for the decrease in lymphocytes (Figure 2h). Bax deficiency did not significantly affect the overall number of white and red blood cells and platelets in Ku70-null mice (Figures 2f, i, and j).
Ku70 −/− mice also developed rectal prolapse (12.5%) and abnormal teeth (19.7%) ( Table 1). In the ku70 −/− mice with abnormal teeth, we observed malocclusion (misaligned teeth) in mice that died at a young age (n = 3) (Table 1 and Supplementary Figure S2). Malocclusion at a young age, especially shortly after weaning, can impair the ability to eat, making these mice more prone to starvation and malnutrition (Supplementary Figure S2, see ku70 −/− mice (52 days)). In addition, we observed abnormally long teeth in adult ku70 −/− mice (n = 8) that likely occurred as a result of a decrease in gnawing behavior due to other underlying illness. For example, one of the ku70 −/− mice (130 days old) (Supplementary Figure S2) was found to have severe lung necrosis and abnormal cardiac hypertrophy at the  (Table 1), and this difference may contribute to their improved survival. When we analyzed survival, excluding malocclusion, the median lifespan of Bax-deficient Ku70-null mice remained greater than ku70 −/− mice (Supplementary Figure S3). Therefore, it is likely that other health effects besides the decreased teeth impairment from Bax deficiency are contributing to enhanced survival.
Approximately half of all of the dead mutant mice were not analyzed for the cause of death because of bodily damage caused by other mice or excessive post-mortem necrosis. These mice are listed under 'unknown cause of death' in Table 1. We suspect that some of these mice developed lymphoma or other diseases, including defects in the heart and lungs, which is discussed below. Therefore, the actual occurrence rate of these diseases may be higher than the rate that was confirmed by analyses of fresh samples. In addition,   infectious diseases may also account for the death in the unknown group as seen in Ku80 (Lupus Ku autoantigen p80) knockout mice. 35 Aging-related changes detected by macroscopic analysis. Ku70-deficient mice develop age-associated changes, such as kyphosis and alopecia, earlier than normal. 6 Interestingly, Bax deficiency delayed the development of some of these age-associated changes in Ku70-null mice (Supplementary Table S2). Furthermore, ku70 −/− mice older than 6 months old did not have visually detectable subcutaneous fat, consistent with the disappearance of subcutaneous fat reported in prematurely aging mice with mutant A-type lamins. 36,37 Bax-deficient Ku70-null mice of the same age maintained this fat layer (Supplementary Table S2). Altogether, these results suggest that the absence of Baxmediated apoptosis can slow down the progression of organismal aging and loss of subcutaneous fat in Ku70null mice.  Figure S4). Because we were not able to detect an abnormally large number of dying pulmonary cells in the other two-thirds of the ku70 −/− mice, we speculate that apoptotic cells are hard to detect in vivo, unlike in cell culture experiments, since dead cells can be cleared by macrophages and neighboring cells, especially if the ku70 −/− mice were in the stages of developing lung diseases. Pulmonary arterial occlusions can lead to pulmonary arterial hypertension (PAH). Blood vessel occlusions were detected by staining the pulmonary arteries with an endothelial cell marker, von Willebrand factor (vWF) ( Figure 5). 42 Occluded pulmonary arteries were observed in ku70 −/− mice and may be because of abnormal endothelial cell overgrowth (Figure 5b). Plexiform-like lesions surrounding the occluded blood vessels were also observed in ku70 −/− mice (Figure 5b, circled). Almost half of the ku70 −/− mice analyzed (42.9%) developed pulmonary vessel occlusions. However, Bax deficiency did not dramatically attenuate the incidence of these abnormalities (Figure 5c), suggesting that the extended survival of Bax-deficient Ku70-null mice was not mainly because of the prevention of PAH. Interestingly, a previous microarray gene expression study of lung tissue from patients with idiopathic PAH showed decreased Ku70 expression in 5 out of 6 patients. 43 Our findings of pulmonary arterial occlusion in ku70 −/− mice could provide clues to help understand the mechanism underlying PAH.
Interstitial lung disease (ILD) was also observed in our mutant mice (Supplementary Figure S7). ILD is defined by an invasion of interstitial cells into the alveolar space and can result from different mechanisms, such as fibrosis, inflammation, or abnormal growth of lung epithelial cells. 44 ILD was found more frequently in Bax-deficient Ku70-null mice than in ku70 −/− mice (Supplementary Figure S7D). However, ILD in these mice was not a result of fibrosis, excessive collagen deposition, or excessive smooth muscle growth, based on Masson's trichrome staining (Supplementary Figure S7A). In addition, ILD was not caused by inflammation or cancer as there was no accumulation of CD45-positive cells (a marker for leukocytes) or Ki67-positive cells, typical of dividing cancer cells (Supplementary Figure S7A-B). Type 1 and 2 alveolar epithelial cells were found in the regions of ILD (Supplementary Figure S7C), suggesting that this ILD, which destroyed alveolar structure, was caused by the abnormal growth and distribution of these cell types. ILD also progressed in an age-dependent manner as severe ILD was found more often in mice 46 months of age.  (Figure 6b-d). The myocardial performance index (MPI), a measure of systolic and diastolic function, 45 was determined for both the RV and left ventricle (LV), and values near 0.32 are considered normal. The MPI values for ku70 −/− mice were normal for the LV but was abnormal for the RV (Figure 6b). Longitudinal strain at the basal and mid-regions of the RV were measured to assess contractility. 34 Ku70-null mice showed significantly lower absolute RV strain, and thus less contractility, compared with wild-type mice (Figure 6c). Furthermore, there was a greater pressure load in the pulmonary artery in ku70 −/− mice than in the wild-type mice, based on the shorter Doppler-determined acceleration time of blood flow through the pulmonary artery (Figure 6d). Our preliminary experiment showed that ku70 −/− bax −/− mice had higher RV MPI than wild-type mice (in this preliminary experiment, all mice were analyzed on the same day under similar conditions, RV MPIs were 0.401 ± 0.018 (wild type), 0.50 ± 0.12 (ku70 −/− ), and 0.51 ± 0.05 (ku70 −/− bax −/− )). This observation suggests that Bax deficiency was not able to normalize heart function in ku70 −/− mice. However, this preliminary measurement was performed only once because of the difficulties with obtaining enough ku70 −/− and ku70 −/− bax −/− mice for simultaneous comparative measurements. Therefore, further detailed analysis using ku70 −/− bax −/− mice will be necessary to understand the mechanism of how Bax deletion can delay the onset of fatal heart failure in ku70 −/− mice.  Figure S6), suggesting that the hearts of Bax-deficient Ku70-null mice may be able to tolerate stresses caused by the absence of Ku70 (e.g., DNA damage). We performed TUNEL staining, but we were not able to detect dying cells in the hearts collected from actively moving mice (Supplementary Figure S5). We speculate that, unlike in in vitro settings, the rapid clearance of apoptotic cells in vivo makes it difficult to detect the actual accumulation of dead cells.
Bax deficiency did not reduce the DNA damage load in Ku70-null brains. Ku70 deficiency has been shown to decrease neuronal survival during development because of excessive neuronal apoptosis, 7 whereas Bax deficiency has the opposite effect. 32 When compared with wild-type mice, the overall brain size and weight of ku70 −/− mice were smaller whereas bax −/− brains were larger (Figure 2e), and the number of hippocampal neurons showed a similar trend (Supplementary Figure S8A). Interestingly, ku70 −/− bax +/ − brains were only slightly heavier than ku70 −/− brains, even though ku70 −/− bax +/ − mice showed extended survival similar to the ku70 −/− bax −/− mice (Figures 1 and 2e). Ku70 −/− mice showed increased DNA DSBs in the hippocampus, based on phospho-γH2A.X staining ( Supplementary  Figures S8B and C). The neurons in ku70 −/− bax +/ − and ku70 −/− bax −/− mice were also positive for phospho-γH2A.X; therefore, Bax deficiency did not help improve neuronal DNA damage repair despite its role in increasing the survival of these neurons.
Bax deficiency in Ku70-null mice did not lead to abnormalities in the kidneys and liver. Unlike the lungs, heart, and brain, the kidneys and liver exhibited no obvious histological defects in both Ku70-null mice and Baxdeficient Ku70-null mice ( Supplementary Figures S9 and S10). Based on Masson's trichrome staining, none of the mice  Figures  S10C and F) based on Ki67 staining. Furthermore, to determine whether there were differences in the occurrence of apoptosis, cleaved Caspase-3 staining was performed ( Supplementary Figures S9C and S10D). However, there are very few detectable Caspase 3-positive cells in these tissues, most likely because of the rapid clearance of apoptotic cells.

Discussion
Previous studies have shown that Ku70 has anti-apoptotic activity by suppressing the intrinsic cell death signal mediated by Bax, in addition to its role in NHEJ DNA DSB repair. 8 Our results support the hypothesis that the absence of Ku70 leads to Bax hyperactivation, giving rise to the development of degenerative diseases that culminate in an early death in ku70 −/− mice. Furthermore, the increased accumulation of DNA damage because of the absence of Ku70 can also trigger the DNA damage response to indirectly initiate apoptosis through p53-dependent Bax activation. 46 We suspect that both of these mechanisms of Bax activation are contributing to the premature death observed in ku70 −/− mice.
In other previous studies, ku70 −/− p53 −/− and ku80 −/− p53 −/− mice developed tumors at much higher frequencies than single Ku70 or Ku80 knockout mice. 47,48 These observations indicate that NHEJ deficiency can cause oncogenic mutations and that p53-dependent cellular responses, such as cell cycle arrest and apoptosis, are important to suppress tumorigenesis. As p53 is intact in Baxdeficient Ku70-null mice, we speculate that p53 is able to suppress tumorigenesis caused by deficient NHEJ through p21-mediated cell cycle arrest and cellular senescence. Previous reports also show that cell cycle arrest, including cellular senescence, plays a greater role than apoptosis to suppress tumorigenesis in NHEJ-deficient mice. 49,50 Consistent with these findings, this study shows extended survival in male and female ku70 −/− bax +/ − mice and male ku70 −/− bax −/− mice without a significant increase in the incidence of lymphoma (Supplementary Table S1). However, the higher incidence of lymphoma in female ku70 −/− bax −/− mice may be reflective of the extended lifespan rather than an actual enhancement of tumorigenesis, as improved survival rate was more prominent in female than male ku70 −/− bax −/− mice that showed no increased tumor incidence. These results further suggest that the partial suppression of Bax-mediated cell death, rather than complete suppression, may be more beneficial for the lifespan extension in Ku70-null mice without an increase in cancer risk.
Presently, it is unclear why Bax deficiency was able to enhance survival in Ku70-null mice more profoundly in females than males. Our results suggest that the effects caused by Bax activation in ku70 −/− mice seem to be more detrimental to females than males, based on the observation that survival is more significantly enhanced in females when Bax is deleted. The deletion of Bax has been shown to negatively impact males, causing testicular degeneration and infertility. 51 Conversely, bax −/− females maintain fertility longer in life because of increased oocyte survival, 52 and this improvement in ovarian function may contribute to the better health span seen in Bax-deficient Ku70-null females to further enhance their survival. However, altered gonadal function does not fully explain the improved survival and extended lifespan observed in Ku70-null mice because Bax haploinsufficiency does not cause the same changes in gonadal function as the complete deletion of Bax, yet both gene mutations result in similar survival rates. Bax +/ − mice have functional ovaries and testes, and their reproduction is normal. 51 Consistently, we observed successful mating and viable offspring from male and female ku70 −/− bax +/ − mice. Therefore, the influence of gonadal function is not the only reason how Bax deficiency can extend lifespan.
In comparison with bax haploinsufficiency (bax +/ − ), the complete deletion of bax (bax −/− ) might have a greater protective influence on lung alveolar cells to suppress cell death caused by the absence of Ku70, and this might be why the maximum lifespan of ku70 −/− bax −/− mice (951 days) was longer than that of ku70 −/− bax +/ − mice (803 days) (Figure 1). However, the median survival of ku70 −/− bax −/− mice (38 weeks) and ku70 −/− bax +/ − mice (37.5 weeks) did not show significant differences (Figure 1). The complete bax deletion is likely to enhance the survival of other cell types, such as the endothelial cells in the alveolar blood vessels, and this abnormality may cause a higher incidence of pulmonary blood vessel occlusion that we observed in the ku70 −/− bax −/− mice despite the increase in overall lifespan. In addition, bax −/− mice, but not bax +/ − mice, have abnormally large brains because of increased neuronal survival during brain development 32 that may have a negative impact on survival later in life. Furthermore, testicular development is suppressed in bax −/− mice, but not in bax +/ − mice, and this leads to the absence of androgen synthesis in the testes. 51 It is possible that the some of the benefits generated by Bax deletion in a Ku70-null background may be canceled by other inherent problems caused by the absence of Bax. In order to further examine the effects of Bax deletion, we are planning to conditionally knockout Bax in specific cell types, such as alveolar epithelial or endothelial cells, in a Ku70-null line.
In support of this study in which we have demonstrated a significant role of a pro-apoptosis gene in the induction of premature death in mutant mice with defects in DNA DSB repair, another report has shown that the deletion of PUMA, a BH3-only protein that activates Bax-mediated cell death, was able to prolong the survival of telomerase-defective mutant mice. 53 Previous studies have shown that ku70 −/− , ku80 −/− , and ku70 −/− ku80 −/− mice exhibit similar abnormal aging phenotypes, including shortened lifespans. 6 As Ku70 protein levels become very low in ku80 −/− cells, 6,54 ku80 −/− mice are expected to be phenotypically similar to ku70 −/− mice and have an increased DNA damage response as well as a lower threshold to initiate Bax-mediated apoptosis. The deletion of p21 was unable to extend the survival of ku80 −/− mice, despite being able to suppress cellular senescence in ku80 −/− MEFs. 55 Importantly, this study shows that Bax deficiency can extend the survival and lifespan of ku70 −/− mice. Altogether, these observations suggest that apoptosis may potentially have a greater role than cellular senescence in lifespan determination, specifically in mice with defective NHEJ DNA repair.
This study also shows that Ku70 is essential to maintain normal lung alveolar structure and pulmonary arteries. Furthermore, this study suggests that Bax-mediated apoptosis plays a role in the development of emphysema in ku70 −/− mice. The hearts in aged ku70 −/− mice (47 months old) also had functional and structural defects, suggesting that Ku70 is required for the maintenance of heart function in older mice. It is not well understood how the DNA damage response promotes the development of emphysema, pulmonary artery occlusion, and heart failure; therefore, further studies are clearly needed to understand the role of Ku70 in the maintenance of homeostasis in the pulmonary and cardiovascular systems.

Materials and Methods
Mouse husbandry and phenotypic observation. Genotyping for the Ku70 and Bax genes was performed as previously reported. 2,51 First, ku70 +/ − bax +/ − breeding pairs generated ku70 +/+ bax +/+ , ku70 −/− bax +/ − , and ku70 −/− bax −/− mice in order to analyze their age-associated disease phenotypes and overall lifespans. The housing and phenotypic observation of these mice were performed as previously reported in articles analyzing the abnormal aging phenotype in ku70 −/− mice. 6 All mice were observed at least 6 times a week for the entire course of their lifespans. Moribund mice showing weight loss and decreased responsiveness were continually monitored multiple times a day, and all the mice were killed when they became immobile and could no longer reach the water source. Morbidities were scored by Kaplan-Meier analysis and measured for statistical significance by the log-rank test. The killed mice were observed by necropsy, and organs were removed and fixed for histology. All mice were housed in microisolator cages in a specific pathogen-free environment. The rodent diet was irradiated, and the bedding, wire top, isolator cage, cardholders, water, and water bottles were autoclaved. All mouse procedures were done in accordance with the Guide for the Care and Use of Laboratory Animals and approved by the institutional IACUC.
Analysis of lung alveolar structure and size. Lungs were collected and fixed by intratracheal instillation of 4% paraformaldehyde to maintain the alveolar structure according to previously published methods. 52 Tissue sections (7 μm thick) of the upper left lobe were prepared for analysis. To determine the differences in alveolar size, measurements of mean linear intercept (L m ) were used. L m was calculated as previously described. 53,54 Pictures of at least 12 representative fields of lung sections were taken, avoiding large airways and blood vessels. Immunohistochemistry. Harvested lungs were perfused, and hearts were incubated in a solution of 250 mM potassium chloride before fixation. All tissues were fixed overnight in 4% neutral buffered paraformaldehyde and then stored in sterile PBS at 4°C before being processed, embedded in paraffin, and sectioned (5-7 μm). Sections were deparaffinized in xylene, rehydrated in a graded ethanol series, and incubated in a solution of 3% H 2 O 2 in methanol for 20 min in order to quench endogenous peroxidase activity. Antigen retrieval was performed according to the manufacturer's instructions for the Biocare Medical Decloaking Chamber system (Concord, CA, USA). All sections were blocked for 1 h at room temperature before an overnight incubation in the primary antibody. The Vectastain Elite ABC Kit (Vector Laboratories, Burlingame, CA, USA; PK-6101) was used according to the manufacturer's instructions. Positive staining was visualized with DAB (ImmPACT DAB, Vector Laboratories, SK-4105). All sections were counterstained with hematoxylin for 30 s and dipped in acid alcohol as needed before being dehydrated and coverslipped. The following primary antibodies were used: Generation and culture of MEFs. MEF cultures were prepared from 14-to 15-day-old embryos from intercrossed ku70 +/ − bax +/ − mice. After each embryo was removed and washed in PBS, the dark-colored visceral tissue from the abdominal region was removed. The remaining tissue was minced and digested in 0.25% trypsin for 20 min at 37°C in 5% CO 2 . The tissue was further broken down by pipetting before being plated onto a 10 cm 2 dish in DMEM supplemented with 10% FBS, 1% non-essential amino acids, 1% (10 mM) L-glutamine, 1% (5 mM) sodium pyruvate, and 1% (50 U) of penicillin and streptomycin. At 80-90% confluency, the cells were passaged and plated in a 1 : 5 dilution. The cells were deemed MEFs once they acquired fibroblast cell morphology. The genotype of each embryo was confirmed with PCR as previously reported. 2,51 Cell death assay. Wild-type and ku70 −/− MEFs were plated onto a 6-well plate at a density of 200 000 cells per well and treated with 1.0 μM Doxorubicin (Sigma-Aldrch, St. Louis, MO, USA; 44583) for 24 h. Floating dead cells were collected, and the live cells were detached using 0.05% trypsin. After pelleting the cells and resuspending the cells to get a total volume of 100 μl, the live and dead cells were counted manually based on Trypan blue dye exclusion. Apoptosis in tissue sections was detected by TUNEL staining using the in situ Apoptosis Detection Kit (4810-30-K, TAVS2 TdT-DAB; Trevigen, Gaithersburg, MD, USA).
Cellular senescence assay. Wild-type and ku70 −/− MEFs were plated onto a 24-well plate at a density of 10 000 cells per well and cultured overnight. On the following day, cells were washed with warm (37°C) HBSS and then fixed with 4% neutral buffered paraformaldehyde for 10 min at room temperature. The senescence detection kit (Biovision, Milpitas, CA, USA; no. K320-250) was used according to the manufacturer's instructions to observe senescence-associated β-galactosidase activity in the MEFs.
Western blotting. At 80-90% confluency, the floating dead cells were removed with HBSS. The remaining cells were scraped on ice, pelleted, and incubated in RIPA lysis buffer (150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, and 0.1% SDS in PBS) containing a protease inhibitor cocktail (Thermo Scientific, Waltham, MA, USA; no. 87786) and PMSF for 20 min at 4°C. The total cell lysate was cleared via centrifugation at a speed of 14 000 r.p.m. (18 626 r.c.f.) for 30 min at 4°C. After determining the protein concentration using a Bradford assay (Bio-Rad, Hercules, CA, USA; no. 500-0006), 5 μg of protein was prepared in Laemmli sample buffer, heat inactivated at 95°C for 5 min, separated using a 4-12% SDS-polyacrylamide gel (Life Technologies, Carlsbad, CA, USA, NuPAGE), and then transferred overnight onto a nitrocellulose membrane. The membranes were then dried at 65°C for 20 min before being blocked for 1 h at room temperature in 5% BSA, 3% milk, and 2% normal serum of the host species from which the secondary antibody was derived. The following primary antibodies were used: Ku70 N3H10 (NeoMarkers, Fremont, CA, USA; 1 : 500), Bax N20 (Santa Cruz Biotechnology, Inc., Dallas, TX, USA; sc-493, 1 : 1000), Caspase 3 (Cell Signaling; 1 : 1000), and β-actin (Sigma-Aldrich, A5441, 1 : 40 000). Horseradish peroxidase-conjugated goat anti-mouse (Life Technologies) or anti-rabbit IgG (Dako) was used as the secondary antibody. The incubation time for each primary antibody was 1 h at room temperature or overnight at 4°C, and the secondary antibodies were incubated for 1 h at room temperature. The bands were visualized using an ECL detection system (GE Healthcare, Pittsburgh, PA, USA; RPN2132).
Echocardiography. Echocardiography was performed as previously described. 45 Briefly, animals were anesthetized with 2% isoflurane supplemented with O 2 in an isoflurane induction chamber and maintained with 1.5% isoflurane by nose cone. After depilating the chest, the extremities were secured to a warming pad (Braintree Scientific, Braintree, MA, USA) with paper tape, and needle electrodes were connected to a preamplifier to simultaneously record a single lead electrocardiogram. All image acquisitions and offline measurements were conducted by a single investigator who was blinded to animal groups. Image processing and data analysis were performed on the ultrasonograph using Syngo Vector Imaging technology software (Siemens Medical Solutions, Malvern, PA, USA).
Statistical analyses. Statistical analyses were performed using GraphPad Prism for Windows, version 6.0 (GraphPad Software, San Diego, CA, USA). The log-rank test was used to compare the survival curves, and Student's t-test was used to compare the genotypes where appropriate.