The Big Blue λSelect-cII selection system has been employed along with whole-exome sequencing to examine the susceptibility of the male germ line to mutation in two challenging situations (i) exposure to a chemotherapeutic regime including bleomycin, etoposide and cis-platinum (BEP) and (ii) the ageing process. A 3-week exposure to BEP induced complete azoospermia associated with a loss of developing germ cells and extensive vacuolization of Sertoli cell cytoplasm. Following cessation of treatment, spermatozoa first appeared in the caput epididymis after 6 weeks and by 12 weeks motile spermatozoa could be recovered from the cauda, although the count (P < 0.001) and motility (P < 0.01) of these cells were significantly reduced and superoxide generation was significantly elevated (P < 0.001). Despite this increase in free radical generation, no evidence of chromatin instability was detected in these spermatozoa. Furthermore, embryos obtained from females mated at this 12-week time point showed no evidence of an increased mutational load. Similarly, progressive ageing of Big Blue mice had no impact on the quality of the spermatozoa, fertility or mutation frequency in the offspring despite a significant increase in the mutational load carried by somatic tissues such as the liver (P < 0.05). We conclude that the male germ line is highly resistant to mutation in keeping with the disposable soma hypothesis, which posits that genetic integrity in the germ cells will be maintained at the expense of the soma, in light of the former’s sentinel position in safeguarding the stability of the genome.
The ‘disposable soma’ hypothesis recognizes that investment in the maintenance and repair of somatic cells has to be balanced by a commitment to the germ line and reproduction (Kirkwood 1977). Thus, species that invest heavily in reproduction early in life, experience a compensatory increase in somatic senescence that directly influences longevity. Because the germ line is essentially immortal, it is critical that DNA surveillance and repair are optimized in such cells in order to keep spontaneous mutation levels to an absolute minimum and mitigate the risk of extinction (Kirkwood & Austad 2000). Both male and female germ cells are therefore thought to be endowed with highly specialized, sophisticated mechanisms to maintain a low mutation rate, protecting the germ line at the expense of the soma. Notwithstanding such measures, the male germ line is more vulnerable to spontaneous mutations than the female, for the simple reason that male germ cells are constantly replicating, while the female germ line spends most of its post-natal existence in a state of meiotic repose (Crow 2000). Furthermore, the male gamete has to undertake a potentially hazardous journey between the site of insemination and the oocyte, during which the chromatin may be subjected to significant oxidative stress (Huang et al. 2015).
In this study, we sought to interrogate the disposable soma hypothesis by examining two paradigms in which the proposed separation of germ line and somatic cell susceptibility to mutation might be tested: ageing and exposure to chemotherapeutic reagents. These studies have been undertaken employing whole exome sequencing methodologies as well as a transgenic mouse (Big Blue), which can be used to provide objective data on mutation frequencies in multiple tissues (Nohmi et al. 2000). This animal model harbours multiple tandem copies of the λLIZ shuttle vector randomly incorporated into its genome, and in combination with the λSelect-cII system provides a quantitative measure of the mutational load carried by cells and tissues.
An analysis of the extent to which the genetic integrity of the germ line is maintained following exposure to chemotherapeutic reagents was stimulated by the rising incidence of testicular cancer, which is currently the dominant cancer of young men aged 20–34 years of age (Ziglioli et al. 2011). These tumours may be seminomas or non-seminomas and exhibit a peak incidence in the fourth decade for the former and a decade earlier for the latter (Skakkebaek 1972, Reuter 2005, Looijenga et al. 2011). For reasons that are still unclear, the incidence of testicular germ cell tumours, while low, has been increasing in recent decades, possibly as a result of foetal exposure to environmental factors such as phthalate esters polychlorinated biphenyls or polyvinyl chlorides during pregnancy (Meeks et al. 2012).
The treatment options for testicular cancer may involve combinations of surgery, radiation and chemotherapy depending on the location and type of cancer and its degree of progression (Shelley et al. 2002, Mortensen et al. 2011). Such treatments are very successful, generating extremely high 5-year survival rates that approach 100% for both types of germ cell tumour (van Basten et al. 1997, Stang et al. 2013). The available clinical and animal data suggest that the major form of chemotherapy used in the treatment of testicular cancer, comprising a mixture of bleomycin, etoposide and cisplatin (the BEP regime), very effectively suppresses spermatogenesis without destroying the spermatogonial stem cell population (Sweeney 2001, Marcon et al. 2008). As a result, spermatogenesis is known to return following BEP treatment, as is fertility (Marcon et al. 2008). However, questions remain concerning the absolute normality of the spermatozoa generated following exposure to BEP as well as the health trajectory of any progeny (Maselli et al. 2012). Thus, in rats, the return of fertility after BEP treatment is associated with an increase in preimplantation embryonic loss, which the authors attribute to an effect on the spermatogonia (Marcon et al. 2008). By contrast, clinical studies have generally failed to find evidence for an increased risk of congenital malformations in the offspring of males who have previously been exposed to alkylating reagents in the course of their cancer treatment (Chow et al. 2009, Signorello et al. 2012). Against this background, a slight increase in the relative risk of congenital birth abnormalities (RR = 1.17) was detected by Ståhl et al. (2011) in the offspring of men with a history of cancer. Given such uncertainty in the literature, Choy and Brannigan (2013) have called for additional research that will allow for the development of clinical guidelines to assist cancer patients, considering post-treatment conception, in their decision-making processes.
The second scenario where investment in the maintenance of genetic integrity in the germ line might be called into question is ageing. The latter results from the progressive accumulation of somatic damage as a consequence of limited investment in maintenance and restoration at the cellular level, particularly with respect to DNA repair and antioxidant protection pathways (Kirkwood & Austad 2000). The extent to which the male germ line is protected from the impact of ageing is controversial since there is ample evidence to suggest that ageing does induce high level of oxidative stress and DNA damage in male germ cells even though other evidence suggests that the germ line is relatively protected from the processes driving senescence (Hill et al. 2004).
In light of the above considerations, we have used the Big Blue model and whole-exome sequencing to quantify the vulnerability of the male germ line to mutate in the wake of two potentially mutagenic stimuli, chemotherapy and advancing age.
Materials and methods
Chemicals and reagents
Unless otherwise stated, all chemicals and reagents were obtained from Sigma-Aldrich while fluorescent probes were obtained from Molecular Probes (Eugene, OR, USA).
Animal treatment and embryo collection
All experimental procedures were conducted with the approval of the University of Newcastle’s Animal Care and Ethics Committee. Big Blue transgenic rodents (Agilent) homozygous for the λLIZ shuttle vector were obtained from Australian BioResources Ltd (Moss Vale, NSW, Australia). Adult males treated with chemotherapy were subjected to a BEP (Bleomycin, Etoposide, Platinum agent) regime commencing when males were between 2 and 4 months of age and extending for 3 weeks. Mice receiving BEP treatment received etoposide (5 mg/kg, Pfizer, New York, NY, USA) and cisplatin (1 mg/kg, Sigma) in 0.9% saline via IP injection on days 1, 3–5 of each week, and etoposide, cisplatin and bleomycin (828 IU/kg, Hospira, Lake Forest, IL, USA) via IP injection on day 2 of each week. Control mice received 0.9% saline via IP injection on days 1–5 of each week. Following the 3-week treatment period, males were allowed to recover for up to 24 weeks. Big Blue females were then used for natural matings at 12, 16 and 24 weeks post treatment in order to assess the long-term impact of BEP exposure on fertility in vivo. The foetuses generated from the 12-week matings were used for mutation analysis. For this purpose, viable embryos were collected at 14 days post coitum (dpc), snap-frozen in liquid nitrogen and stored at −80°C. For the 12-week time point, there were 16–31 mice per group. Testis weight, sperm motility and cell counts were conducted on 15–31 mice, while for all other assays, a minimum of 3 mice were used. For all other time points, 3–7 mice were analysed. For mating studies, 2–5 pairings were conducted for each time point.
For the ageing study, liver, testis and spermatozoa were collected from Big Blue male mice at 2–4, 9–11, 17–20 and 21–23 months of age. For the mating studies, males aged 2-4 months and 17-20 months were mated to females 1–3 (young) or 8–10 months (old) of age (corresponding to approaching middle age but still capable of generating offspring). Embryos from these matings were collected 14 dpc, snap-frozen in liquid nitrogen and stored at −80°C. For these studies, 19–22 male mice were analysed per age group. Testis weight and sperm motility were conducted on 16–22 mice per group, while all other assays were conducted on 3–10 mice per group. For mating studies, 4–10 pairings were conducted for each age group.
Collection of spermatozoa and tissue
Mice were killed via carbon dioxide asphyxiation and spermatozoa and testes were collected immediately. Caudal epididymal spermatozoa were collected by back-flushing and deposition of the perfusate under oil at 37°C. The sperm population was then diluted in 1 mL modified Biggers, Whitten and Whittingham medium (BWW) (Biggers et al. 1971) supplemented with 0.1% PVA, 0.27 mM sodium pyruvate, 44 mM sodium lactate, 5 U/mL penicillin, 5 mg/mL streptomycin and 20 mM HEPES buffer. Caput spermatozoa were obtained via gentle dissection of the tissue in 1 mL BWW followed by filtration through a 70 μm nylon mesh strainer. Testes were fixed overnight in Bouin’s solution and washed 3× in 70% ethanol before being embedded in paraffin, sectioned and stained with haemotoxylin and eosin. Testicular sections were subsequently examined to determine the percentage of tubular cross sections exhibiting evidence of impaired spermatogenesis as described by Marcon et al. (2008). One tissue section was analysed per testis, from 3 to 4 animals per time point. An average of 97 tubular cross sections were analysed per slide. Tubules were classified as exhibiting disrupted spermatogenesis when there was extensive vacuolisation of the Sertoli cells, disorganisation of the germ cells or a lack of later germ cell types (e.g. beyond spermatogonia or, sometimes, spermatocytes) present.
Computer-assisted sperm analysis
Caudal epididymal sperm motility was assessed with a Hamilton Thorne CASA System, version 12 IVOS (Hamilton Thorne Biosciences, Beverly, MA, USA). Motility was scored for at least 200 cells per sample. Motility was defined as an average path velocity greater than 5 μm/s while progressive motility was defined as an average path velocity greater than 25 μm/s and a straightness value greater than 80%. The image acquisition settings were negative phase-contrast optics; recording rate 60 frames/s; minimum contrast 80; minimum cell size 3 pixels; low size gate 1.0; high size gate 2.9; low-intensity gate 0.6; high-intensity gate 1.4; non-motile head size 6 and non-motile head intensity 160.
MitoSOX Red and dihydroethidium assays
To determine the generation of mitochondrial or overall cellular superoxide anion, MitoSOX Red (MSR) or dihydroethidium (DHE) was used respectively (Aitken et al. 2013). These probes were added to cells at a final concentration of 2 μM, accompanied by the viability stain, SYTOX Green, at a final concentration of 0.05 μM. Following 15-min incubation at 37°C, the spermatozoa were centrifuged for 3 min at 600 g before being resuspended in 300 μL BWW. Fluorescence was then measured on a FACSCalibur flow cytometer with CellQuest Pro software (Becton Dickinson, San Diego, CA, USA). Argon laser excitation at 488 nm was coupled with emission measurements using 530/30 band pass (green) and 585/42 band pass (red) filters. A total of 10,000 events were recorded following gating out of non-spermatozoa-specific events.
To measure mitochondrial membrane potential, JC-1 probe was added to cells at a final concentration of 6.25 µM. Following 15-min incubation at 37°C, the spermatozoa were centrifuged for 3 min at 600 g before being resuspended in 400 µL BWW. Just prior to analysis, propidium iodide was added to each sample at 10 µg/mL. Fluorescence was then measured on a FACSCalibur flow cytometer with CellQuest Pro software (Becton Dickinson). Argon laser excitation at 488 nm was coupled with emission measurements using >670 long pass filter (far red; propidium iodide) to exclude dead cells from analysis, followed by 530/30 band pass (green; low MMP) and 585/42 band pass (red; High MMP) filters. A total of 10,000 events were recorded following gating out of non-spermatozoa-specific events.
Snap-frozen spermatozoa samples were thawed and incubated overnight at 4°C in 25 µg/mL DNA/RNA Damage Antibody (Novus Biologicals, Littleton, CO, USA) in phosphate-buffered saline containing 0.1% Tween (PBST). Cells were washed with phosphate-buffered saline (PBS), incubated with Alexa Fluor 488 antibody (ThermoFisher) in PBST for 1 h at room temperature and then washed twice with PBS. Cells were mounted on slides and counted on a fluorescent Zeiss Axio Imager A.1 (Carl Zeiss GmbH). A total of 100 cells were counted for each sample.
Sperm chromatin structure assay
Following isolation, spermatozoa were washed with BWW, snap-frozen in liquid nitrogen and stored at −80°C. Sperm chromatin structure assay (SCSA) was performed as described by Evenson and Jost (2000) using a FACScan Flow Cytometer with CellQuest software (Becton Dickinson). Briefly, 200 μL acid detergent solution (0.08 N HCl, 0.1 5 M NaCl, 0.1% Triton X-100, pH 1.2) was added to thawed cells. Following 30 s incubation, 600 μL acridine orange staining solution was added and the sample was run through the flow cytometer for 2.5 min before acquisition. A total of 5000 events were recorded following gating out of debris. Cells in the main population and those with high DNA stainability, representing immature spermatozoa, were subtracted from the total number of cells analysed to calculate %COMP, the proportion of cells outside the main population with abnormal chromatin structure.
DNA extraction for λSelect-cII mutation detection assay and whole-exome sequencing
High-molecular-weight genomic DNA was extracted from adult testes, livers and 14 dpc whole embryos using the RecoverEase DNA Isolation Kit (Agilent) according to the manufacturer’s protocol with the following modifications. A 2 mL Kontes dounce tissue grinder (Kimble Chase, Rockwood, TN, USA) was used to homogenise each embryo in 1 mL lysis buffer. Pestle B was used for 10–12 strokes to homogenise each sample. After pouring each sample through the cell strainer, a further 7 mL of lysis buffer was used to wash the tissue grinder to bring the total sample volume to 8 mL. Following addition of digestion solution, 70 μL of warm proteinase K solution (2 mg/mL proteinase K, 2% SDS, 0.1 M EDTA) was added before each sample was incubated at 50°C for 45 min. Following this incubation, the sample was dialysed against TE Buffer (10 mM Tris–HCl pH 7.5, 1 M EDTA) for 24–72 h, as required. On completion of dialysis the fully hydrated genomic DNA was recovered and stored at 4°C.
Determination of cII mutant frequency
The λLIZ shuttle vector was packaged using Transpack Packaging Extract (Agilent) according to the manufacturer’s protocol, with several modifications. Briefly, following the final 90 min incubation, 1.1 mL SM buffer (100 mM NaCl, 8 mM MgSO4·7H2O, 50 mM Tris–HCl pH 7.5, 0.01% w/v gelatin) were added to each sample as described in the λ Select-cII Mutation Detection System for Big Blue Rodents protocol (Agilent). Titering and screening then proceeded as described in the λ Select-cII protocol using G1250 E. coli cells on TB1 agar (1.2% bacto agar, 1% bacto-tryptone, 0.5% NaCl, 0.0001% thiamine hydrochloride). Ultimately, cII− mutants were measured by growing infected G1250 bacteria on TB1 plates at 24°C for approximately 46 h. During this incubation, infections involving phage with a WT cII gene undergo lysogenisation (a process by which bacteria acquire phage sequences that become integrated into their genome) and the infected bacteria become part of the developing lawn, while phage with a mutated cII gene undergo a lytic cycle, giving rise to plaques. At 37°C, all phage undergo the lytic cycle and form plaques regardless of whether they are carrying a WT or mutant cII gene, enabling titering of the total plaque forming units (pfu) screened. The cII mutant frequency was calculated as the number of mutant plaques per total plaques screened.
Determination of cII mutation frequency
The cII sequence from every successfully replated mutant was sequenced following PCR amplification and purification as follows: 5 μL sample were combined with 2.5 mM MgCl2, 1× reaction buffer, 0.5 mM each dNTP, 1 μM each primer (upstream: 5′-CCGCTCTTACACATTCCAGC-3′, downstream: 5′-CCTCTGCCGAAGTTGAGTAT-3′) and either 2.5 U or 1 U Taq (Taq2000 DNA Polymerase, Agilent or Taq DNA Polymerase Native, Invitrogen, respectively) in a total reaction volume of 20 μL. A 476 bp amplicon was produced in an Eppendorf Mastercycler (Eppendorf, Hamburg, Germany) with the following cycling conditions: an initial 3 min denaturation step at 95°C, 35 cycles of denaturation at 95°C for 30 s, annealing at 62°C for 1 min and extension at 72°C for 1 min, followed by a final extension at 72°C for 10 min. Amplicons were purified using the Wizard SV Gel and PCR Clean-Up System (Promega) and ethanol precipitated before being sequenced. The PCR primers detailed were also used as sequencing primers. Mutations were identified using NCBI Nucleotide BLAST, FinchTV Version 1.4 software (Geospiza, Seattle, WA, USA) and the published WT lambda sequence (Accession #J02459). When this analysis did not identify a mutation, PCR and sequencing were repeated using a second upstream primer to cover the PR promoter region (5′-GCGACAGATTCCTGGGATAA-3′); however, this strategy did not reveal any additional mutations in the promoter region. When the same mutation was found multiple times in one sample, it was assumed to be a jackpot mutation and was therefore counted as a single independent mutation. When a mutant exhibited multiple types of mutations, these mutations were scored as a single complex mutation. The mutation frequency was then calculated as the number of independent mutations per total plaques screened.
Calculation of the mutational load present in exome of embryos
The genomic DNA extracted from embryos conceived by young and old parents were prepared for whole-exome sequencing. In the young group, males and females were 2–4 and 1–3 months of age respectively, while in the aged group, the corresponding figures were 17–20 and 8–10 months of age respectively; n = 4 for all analyses. Samples containing approximately 100 µg/mL of genomic DNA were placed in a Covaris Sonicator (Covaris, Woburn, MA, USA) to generate 150–200 bp DNA fragments. Each sample was treated with SureSelectXT Target Enrichment System for Illumina Paired-End Sequencing Library (Agilent) and sequencing was performed on a Mi-Seq Sequencer (Illumina). Raw sequence reads obtained by sequencing were imported into and analysed with the NextGENe Software (SoftGenetics, State College, PA, USA). Sequence reads were mapped against the NCBI and Mouse Genome Sequence Consortium Build 37 (NCBI37/mm9) (July 2007, NCBI). Mutations detected on genes with coverage lower than 15 reads and calling scores below 12 were considered to be alignment or base-calling errors and were excluded from further analysis. Mutational load in each sequenced sample was calculated from validated mutations per nucleotide and averaged for comparison between mating groups. Mutations identified as matching common known SNPs present in the dbSNP database (Sherry et al. 2001) were excluded from the calculation of de novo mutation rate. Mutations present in at least 90% of all the samples were considered established mutations in the mouse strain and also excluded from the calculations.
All experiments were replicated at least 3× on independent samples and the results analysed by one- and two-way ANOVA using the SuperANOVA program (Abacus Concepts Inc, CA, USA) on a MacIntosh G4 Powerbook computer; post hoc comparison of group means was by Fisher’s PLSD (Protected Least Significant Difference). Mutation spectra were analysed using the Cariello et al. (1994) program. For the exome analysis, all experimental results obtained were imported into R Statistical Software (R Core Team 2004). Differences in mutation frequency were determined by performing the Mann–Whitney U test in light of the data’s non-parametric distribution. Paired comparisons of litter sizes and embryonic resorptions per successful mating of young and aged parents were also calculated using the Mann–Whitney U tests. A Kruskal–Wallis one-way ANOVA and the post hoc multiple comparison test, kruskalmc test in R, was performed to compare mutational frequencies in tissue samples.
One week after cessation of the 3-week treatment period, the histology of the testes revealed a profound disruption of spermatogenesis in the BEP-treated animals, in contrast to the vehicle-only controls (Fig. 1). In the treated animals, a majority of the tubules were devoid of differentiating germ cells and instead presented with highly vacuolated Sertoli cells and exfoliated immature germ cells in the tubule lumina. Around the periphery of the seminiferous tubules, a single layer of spermatogonia remained bound to the basement membrane and still appeared to be viable. Similarly, the interstitial tissue did not appear to have been dramatically affected by temporary exposure to BEP (Fig. 1). During the ensuing 20-week spermatogenesis gradually recovered, with fewer and fewer tubular cross sections showing evidence of severely disrupted spermatogenesis (Figs 1 and 2A). Nevertheless, even 24 weeks after the cessation of BEP treatment, occasional tubules could be found that still exhibited vacuolated Sertoli cell cytoplasm and a complete absence of germ cell development (Figs 1 and 2A). The long-lasting effects of BEP treatment were also evident from the testicular weights which, though improved, had still not returned to control levels after 24 weeks of recovery (Fig. 2B). Despite such lingering effects of BEP exposure, it should be noted that litter sizes were not significantly different between control and BEP-exposed fathers 12, 16 and 24 weeks after treatment (Fig. 2C). Similarly, the number of embryonic resorption sites was not significantly different between control and treated animals at these time points (Fig. 2D).
One week after the cessation of treatment there were very few spermatozoa in the caput epididymis, indicating that sperm production had been severely compromised by BEP exposure (P < 0.001; Fig. 3A). Spermatozoa were present in low numbers at 6 and 12 weeks and afterwards numbers increased but remained significantly (P < 0.001) below control levels as long as 24 weeks the cessation of treatment (Fig. 3A). Some residual spermatozoa were recovered from the cauda epididymis 1 week after the cessation of treatment (Fig. 3B) but by 6 weeks had declined to negligible levels reflecting the severe disruption of spermatogenesis (P < 0.001). Even by the 12th post-treatment week, the numbers of spermatozoa stored in the cauda epididymis were still low and remained significantly below control values 24 weeks after the cessation of treatment (P < 0.01; Fig. 3B). The few spermatozoa recovered 6 weeks after the cessation of treatment were all immotile (Fig. 3C and D); by the 12th week, both motility and progressive motility were clearly improving, although these quality measures were still significantly below control levels at this time point (P < 0.01 and P < 0001 respectively). However, by the 16th week, neither motility nor progressive motility was significantly different from the controls. An analysis of reactive oxygen species (ROS) generation by the spermatozoa revealed a significant increase in the generation of mitochondrial ROS as early as 1 week post treatment (P < 0.05), which then remained elevated until the 24th week (P < 0.01; Fig. 3E). Cytoplasmic superoxide generation as measured by DHE oxidation was also significantly elevated 12 weeks post treatment (P < 0.01) but then returned to control levels by the 16th week (Fig. 3F).
These results indicated that by 12 weeks after the cessation of BEP treatment, small numbers of spermatozoa were appearing in the epididymis with significantly reduced total motility levels accompanied by high levels of ROS generation. In order to determine whether the oxidative stress observed in the spermatozoa during this recovery phase influenced DNA integrity in these cells, a SCSA analysis was performed (Fig. 4A and B). According to this assay, no significant induction of DNA damage was evident at any time during the 24-week recovery period (Fig. 4B). At all time intervals tested, the %COMP values recorded in the SCSA assay were low and not significantly different from the control animals (Fig. 4B). As a positive control for this assay, we also examined caput epididymal spermatozoa, which are known to generate high SCSA values (Pérez-Cerezales et al. 2012). This is because the chromatin of caput epididymal cells is still not fully stabilized by disulphide bond formation and, as a result, exhibits an enhanced susceptibility to the acid denaturation conditions employed in the SCSA system (Bennetts & Aitken 2005). As anticipated, immature spermatozoa from the caput epididymis exhibited significantly higher SCSA signals compared with caudal cells (P < 0.001) but at no time point did caput spermatozoa from BEP–exposed animals generate higher SCSA signals than their unexposed controls (Fig. 4A).
Mutation assays following chemotherapy and ageing
Examination of testicular tissue from mice sampled 12 weeks after BEP treatment revealed a significant increase in mean mutation frequency relative to the controls (P = 0.036; Table 1) in the absence of any particular change in the spectrum of independent mutations observed (Table 2). In order to determine whether these mutations were in the somatic or germ cell component of the testes the mean mutant frequency, describing the fraction of λLIZ shuttle vectors carrying mutations in the cII gene, was examined in spermatozoa and found to be identical (3.1 ± 1.0 × 105 vs 3.1 ± 1.0 × 105 in BEP-treated and control groups respectively, with 2,640,668 plaques screened). A subsequent analysis of mutation frequencies in day 14 embryos generated from BEP-treated and control males mated to untreated females revealed no increase in mutation frequency as a consequence of chemotherapeutic exposure (2.9 ± 0.2 × 105 vs 2.7 ± 0.3 × 105; Table 3). In Table 4, the types of independent mutations recorded in these two groups of animals are presented. In both the control and BEP-exposed groups, the predominant type of mutation was a G-to-A transition followed by G-to-T transversions. Analysis of the mutation spectra between the two groups did however reveal a significant difference (P < 0.05), with the offspring of chemotherapy-treated males notably exhibiting fewer C-to-T transitions and more G-to-T transversions than the controls (Table 4).
Mutation frequencies in the cII gene in testes of BEP-treated mice after 12 weeks recovery.
|Group||n||Plaques screened||Mutant plaques||Mutant frequencya (×10−5)||Mean mutant frequency ± s.e. (×10−5)||Independent mutations||Mutation frequencyb (×10−5)||Mean mutation frequency ± s.e. (×10−5)|
|Control||4||1.4 ± 0.3||1.1 ± 0.2|
|BEP||4||6.3 ± 2.9*||4.1 ± 1.8*|
Spectra of independent mutations in the cII gene in testes of BEP-treated mice after 12 weeks recovery.
Mutation frequencies in the cII gene in day 14 whole embryos produced from pairings between BEP-treated fathers and untreated mothers.
|Group||n||Plaques screened||Mutant plaques||Mutant frequencya (×10−5)||Mean mutant frequency ± s.e. (×10−5)||Independent mutations||Mutation frequencyb (×10−5)||Mean mutation frequency ± s.e. (×10−5)|
|Control||15c||4.6 ± 0.4||2.9 ± 0.2|
|BEP||15d||4.5 ± 0.6||2.7 ± 0.3|
Spectra of independent mutations in the cII gene in day14 whole embryos produced from pairings between BEP-treated fathers and untreated mothers.
Ageing of either males (or their female partners) also had no significant impact on the mutation frequencies seen in day 14 embryos (Table 5) or the spectra of independent mutations observed (data not shown). The whole-exome analysis of embryos conceived from differentially aged parents generated data for 203,225 exonic regions of the mouse genome, collectively comprising 4.54 × 10−8 bases per sample with an average coverage of 10 per base. The absolute number of de novo mutations in the exomes of the offspring of young or aged parents was found to be not statistically different (Fig. 5A) comprising 1352.0 ± 19.8 and 1356.5 ± 8.8 mutations per individual for matings involving young and old progenitors respectively. Overall, the type of mutation found in the genome of the offspring of young and aged mice, was predominantly base substitutions with a transition/transversion ratio of 1.05.
Mutation frequencies in the cII gene in day 14 whole embryos produced from pairings between aged mice.
|Group||n||Plaques screened||Mutant plaques||Mutant frequencya (×10−5)||Mean mutant frequency ± s.e. (×10−5)||Independent mutations||Mutation frequencyb (×10−5)||Mean mutation frequency ± s.e. (×10−5)|
|YM × YF||5c||3.5 ± 1.1||1.6 ± 0.1|
|YM × OF||5d||1.9 ± 0.2||1.5 ± 0.1|
|OM × YF||5e||2.4 ± 0.3||1.7 ± 0.1|
|OM × OF||7f||3.0 ± 0.5||1.8 ± 0.2|
The mutation frequency obtained from sequencing the whole-exome of the offspring of young and aged parents was also determined to be not statistically different (Fig. 5B). The mutation frequency observed in the offspring resulting from the mating of young or old mice was 2.98 × 10−7 ± 4.36 × 10−9 and 2.99 × 10−7 ± 1.95 × 10−9 mutations per base pair (bp), respectively. The overall mutation frequency carried in the exome of embryos conceived by both groups was calculated to be 2.98 × 10−7 ± 2.40 × 10−9 mutations per bp.
These data suggested that the male germ line is protected from the impact of ageing on this strain of mouse in contrast to somatic tissues such as the liver where a significant increase (P < 0.05) in mutation frequency was observed with age (Table 6). The concept that the germ line of Big Blue mice is protected from age-dependent deterioration was also consistent with the analysis of male reproductive function during ageing in this strain (Fig. 5). Thus, age was found to have no significant impact on testes weight (Fig. 5C), sperm motility (Fig. 5D), sperm DNA damage (Fig. 5E) or 8OHdG formation (Fig. 5F). Significantly elevated levels of DNA fragmentation were observed in caput vs caudal epididymal spermatozoa (P < 0,001; Fig. 5E) reflecting the differences in DNA compaction in these two populations of spermatozoa; however, this change was not influenced by age. Interestingly, mitochondrial superoxide generation by caudal epididymal spermatozoa was found to significantly decline during epididymal maturation (P < 0.001; Fig. 5G) in concert with a significant increase in mitochondrial membrane potential (P < 0.001; Fig. 5H), suggesting that the efficiency of mitochondrial electron transport improves as a consequence of epididymal passage. Importantly however, this biomarker of epididymal function showed no deterioration as a consequence of age (Fig. 5G and H).
Mutation frequencies in the cII gene in livers from ageing mice.
|Age (months)||n||Plaques screened||Mutant plaques||Mutant frequencya (×10−5)||Mean mutant frequency ± s.e. (×10−5)||Independent mutations||Mutation frequencyb (×10−5)||Mean mutation frequency ± s.e. (×10−5)|
|2–4||5||3.3 ± 0.4||2.4 ± 0.3|
|9–11||5||6.6 ± 0.5**||4.6 ± 0.3*|
|21–23||4||6.8 ± 0.8**||4.4 ± 0.5*|
One of the key tenets of the disposable soma hypothesis is that energy will be invested in maintaining the genetic integrity of the germ line at the expense of the soma, which is more vulnerable to ageing as a consequence. We have used whole-exome sequencing and the Big Blue transgenic mouse to examine this concept in two situations: the administration of chemotherapeutic agents and ageing.
In terms of chemotherapy, this study has confirmed the devastating effect that chemotherapeutic agents such as bleomycin, etoposide and cisplatin can have on spermatogenesis (Bieber et al. 2006, Delbes et al. 2007, Maselli et al. 2012). After a 3-week course of treatment, spermatogenesis was severely impaired with significant exfoliation of developing precursor germ cells in the lumen of the seminiferous tubules and extensive vacuolization of the Sertoli cell cytoplasm. As long as 12 weeks after a course of BEP treatment, evidence of disrupted spermatogenesis was still evident, testis weights were still reduced, sperm numbers were still low in the caput and cauda epididymis and the spermatozoa were actively generating ROS. At this point in the recovery pathway however, sperm motility was beginning to reappear, the animals were fully fertile and we could find no evidence of DNA damage in the spermatozoa or enhanced mutation rates in the offspring. Litter sizes were also not statistically different between treated animals and controls and the offspring showed no gross morphological abnormalities. However, more subtle changes that might have been detected by weighing the pups or conducting detained autopsies cannot be excluded. Nevertheless, at face value, these results suggest that as the germinal epithelium recovers from BEP exposure and spermatogenesis resumes, the spermatozoa that are produced are normal and fully capable of fertilization, even if they are surrounded by dysfunctional spermatozoa generated by regions of the seminiferous tubule that are still in the process of recovery.
Studies in the rat have emphasized the importance of germ cell apoptosis in BEP-induced azoospermia and have also suggested that recovery of the spermatogenic process may be impaired following exposure, with BEP-treated animals exhibiting abnormal chromatin stability according to SCSA, TUNEL and Comet assays (Delbes et al. 2007). However, when BEP-treated animals were left for a further period of recovery, all evidence of DNA fragmentation receded (Maselli et al. 2012), although long-term changes to sperm chromatin composition was suggested by high chromomycin A3 stainability, indicative of poor protamination and a concomitant increase in the levels of histone expression (Maselli et al. 2012). Such data suggest a long-lasting impact of BEP exposure on the chromatin structure of mammalian spermatozoa. The histone enrichment observed under these circumstances may be an adaptive change designed to place an increased number of genes in a non-protaminated ‘poised’ state, ready for early, facilitated expression in the embryo (Hammoud et al. 2009). Alternatively, such poor protamination may place the germ line at risk by facilitating the onset of DNA damage. A large number of studies have demonstrated that poor chromatin compaction is positively correlated with a susceptibility to sperm DNA damage (Sakkas et al. 1998, De Iuliis et al. 2009, Nili et al. 2009, Castillo et al. 2011, Manochantr et al. 2012). Such observations have important implications for young male cancer survivors wishing to establish a family because DNA damage in spermatozoa has been associated with a wide range of diseases in the offspring including brain disorders such as autism, bipolar disease, spontaneous schizophrenia and epilepsy as well as childhood cancer (Aitken & De Iuliis 2007).
The SCSA assay is one of the most sensitive assays of DNA damage that can readily discriminate the vulnerability of poorly compacted spermatozoa in the caput epididymis from their compacted caudal epididymal counterparts and with an efficiency that is beyond the conventional TUNEL assay (Pérez-Cerezales et al. 2012). Within our data set, we found exactly the same high sensitivity to DNA damage in caput epididymal spermatozoa when the latter were exposed to the acid DNA denaturation conditions characteristic of SCSA (Fig. 4; Evenson & Jost 2000). This sensitivity is thought to be due to the limited degree of chromatin stabilization observed in immature caput epididymal spermatozoa as a consequence of the incomplete formation of inter- and intra-molecular disulphide bridges between, and within, sperm protamines (Pérez-Cerezales et al. 2012). By the time epididymal maturation has been completed and these disulphide brides have been created, the chromatin is in a highly stable, compacted state and in this mature condition is relatively resistant to DNA damage, whether this is via an oxidative attack (Sawyer et al. 2003) or acid-induced denaturation in the SCSA assay. The fact that the anticipated decline in SCSA reactivity as spermatozoa matured in the epididymis was observed with equal facility in both our control animals and our BEP-exposed males suggests that chromatin compaction was occurring normally in the latter. The spermatozoa recovered from the cauda epididymis 12–24 weeks after the cessation of BEP therapy therefore showed no signs of spontaneous DNA damage or any susceptibility to the induction of such damage under denaturing conditions. Nevertheless, more subtle changes in sperm chromatin integrity and structure that remain undetected by the SCSA assay cannot be excluded.
These results accord with the outcome of the Big Blue® mutation assay, which failed to find any evidence of an increase in the mutational load carried by the offspring of BEP-exposed fathers. They also accord with a growing volume of clinical data suggesting that there is no major increase in the incidence of congenital birth defects in the offspring of fathers who have previously been exposed to alkylating agents in the course of cancer treatment (Dohle et al. 2010). While evidence does exist in the rat model for impacts of paternal chemotherapy with reagents such as cyclophosphamide (Hales et al. 2005), these concerns were not borne out in the present animal model as far as BEP is concerned. It is possible that more stringent dosing regimens may have generated a different outcome; however, the chemotherapeutic protocol employed in this study was clearly competent to completely suppress spermatogenesis. The power of the Big Blue model is that it permits a quantitative approach to genotoxicity that can be backed up by the rigour of sequencing every single mutant plaque, in order to confirm the existence of de novo mutations. These quantitative data serve to emphasize the remarkable capacity of the spermatogonial stem cell population to not only withstand chemotherapeutic attack, but also to recover from such an insult by reinitiating spermatogenesis in a manner that may be quantitatively impaired but maintains a high level of genetic fidelity in the spermatozoa. Thus, while there was clear evidence of an increase in mutation rate within the testes following BEP exposure (Table 1), there was no evidence of an increase in the mutational load carried in the germ line, suggesting that the increase must have been due the somatic cell constituents of this organ (for example, Sertoli or Leydig cells). Of course, it is always possible that whole genome sequencing might have revealed damage outside of the exomic regions. However, it should also be acknowledged that the Big Blue cassette inserts itself into random regions of the genome and yet we still did not see a significant increase in mutation frequency following treatment with BEP.
The only subtle change observed as a consequence of BEP exposure was an increase in G-T transversions at the expense of C-T transitions (Table 4). The impact of such a change on the health trajectory of the offspring is difficult to ascertain. However, it may be significant that transversions, rather than transitions, have been found to have a greater impact on gene expression and have been associated with the early onset of cancer and other genetic diseases (Leonard et al. 2013, Lyons & Lauring 2017, Zhan et al. 2017).
Analysis of the impact of paternal ageing on the mutational load carried by the progeny generated very similar results. While mutation rates significantly increased in the liver as a consequence of ageing, there was no evidence of any corresponding change in the germ line. The offspring of old fathers exhibited exactly the same mutation rate as the offspring of young fathers and the relative age of the mother did not seem to contribute to the overall levels of mutation observed. Furthermore, we could see no change in the quality of the spermatozoa with age. As with the BEP study, we noted a change in the susceptibility of chromatin to the acid conditions associated with the SCSA assay as a consequence of epididymal maturation (Fig. 5E); however, this stabilization process was unaffected by age (Fig. 5E). Similarly, we noted that the mitochondria acquire functionality as a consequence of epididymal maturation, reducing the leakage of electrons to oxygen to generate superoxide anion and increasing the competence of these organelles to generate a potential difference across the inner mitochondrial membrane (Fig. 5G and H). This attribute of epididymal function did not change with paternal age.
Thus, for this strain of mouse, there is a clear preferential investment in the maintenance of the germ line with increasing age, as would have been predicted by the disposable soma hypothesis and adding weight to the general evolutionary significance of this concept. In contrast, humankind clearly places limits on any investment in the germ line since there is incontrovertible evidence indicating that paternal ageing is associated with a loss of human fertility, elevated rates of miscarriage and an increase in the mutational load carried by the offspring (Kong et al. 2012, Aitken 2013a, Johnson et al. 2015, Jónsson et al. 2017, Kaarouch et al. 2018). Thus, the human condition may represent a departure from the disposable soma hypothesis, with its implied high level of investment in the maintenance of the germ line. The elevated levels of infertility observed in our own species as well as the abundant evidence for paternal age impacts on a complex sweep of conditions including achondroplasia, autism and spontaneous schizophrenia, raise fundamental questions about our species’ commitment to its germ line and the validity of mouse models to study the impact of intrinsic and extrinsic factors on germ cell integrity (Aitken et al. 2013b, Yatsenko & Turek 2018). The evolutionary advantage associated with such a reckless dereliction of duty in a human context may suggest the existence of a trade-off that places increased emphasis on the role of the germ line in creating genetic diversity and our resultant ability to adapt to environmental change.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported. Robert Aitken is a member of the Editorial Board of Reproduction.
This study was funded by a Discovery grant from the Australian Research Council (# DP110103951).
The authors are extremely grateful to Dr Jennette Sakoff for supplying the chemotherapeutic drugs and to the staff of the Animal Services Unit, University of Newcastle, for their assistance with animal care and maintenance. They are also grateful to Tegan Smith for helpful discussions in relation this project and the Analytical and Biomolecular Research Facility (ABRF) of the University of Newcastle for access to its flow cytometry core.
AitkenRJSmithTBLordTKuczeraLKoppersAJNaumovskiNConnaughtonHBakerMADe IuliisGN 2013 On methods for the detection of reactive oxygen species generation by human spermatozoa: analysis of the cellular responses to catechol oestrogen, lipid aldehyde, menadione and arachidonic acid. Andrology 1 192–205. (https://doi.org/10.1111/j.2047-2927.2012.00056.x)
ChowEJKamineniADalingJRFraserAWigginsCLMineauGPHamreMRSeversonRKDrews-BotschCMuellerBA 2009 Reproductive outcomes in male childhood cancer survivors: a linked cancer-birth registry analysis. Archives of Pediatrics and Adolescent Medicine 163 887–894. (https://doi.org/10.1001/archpediatrics.2009.111)
De IuliisGNThomsonK.MitchellLAFinnieJMKoppersAJHedgesANixonBAitkenRJ 2009 DNA damage in human spermatozoa is highly correlated with the efficiency of chromatin remodeling and the formation of 8-hydroxy-2′-deoxyguanosine, a marker of oxidative stress. Biology of Reproduction 81 517–524. (https://doi.org/10.1095/biolreprod.109.076836)
HillKABuettnerVLHalangodaAKunishigeMMooreSRLongmateJScaringeWASommerSS 2004 Spontaneous mutation in Big Blue® mice from fetus to old age: tissue-specific time courses of mutation frequency but similar mutation types. Environmental and Molecular Mutagenesis 43 110–120. (https://doi.org/10.1002/em.20004)
HuangVWLeeCLLeeYLLamKKKoJKYeungWSHoPCChiuPC 2015 Sperm fucosyltransferase-5 mediates spermatozoa-oviductal epithelial cell interaction to protect human spermatozoa from oxidative damage. Molecular Human Reproduction 21 516–526. (https://doi.org/10.1093/molehr/gav015)
JónssonHSulemPKehrBKristmundsdottirSZinkFHjartarsonEHardarsonMTHjorleifssonKEEggertssonHPGudjonssonSA et al. 2017 Parental influence on human germline de novo mutations in 1548 trios from Iceland. Nature 549 519–522. (https://doi.org/10.1038/nature24018)
KaarouchIBouamoudNMadkourALouanjliNSaadaniBAssouSAboulmaouahibSAmzaziSCopinHBenkhalifaMSefriouiO 2018 Paternal age: negative impact on sperm genome decays and IVF outcomes after 40 years. Molecular Reproduction and Development 85 271–280. (https://doi.org/10.1002/mrd.22963)
LooijengaLHGillisAJStoopHBiermannKOosterhuisJW 2011 Dissecting the molecular pathways of (testicular) germ cell tumour pathogenesis; from initiation to treatment-resistance. International Journal of Andrology 34 e234–e251. (https://doi.org/10.1111/j.1365-2605.2011.01157.x)
MarconLHaleBFRobaireB 2008 Reversibility of the effects of subchronic exposure to the cancer chemotherapeutics bleomycin, etoposide, and cisplatin on spermatogenesis, fertility, and progeny outcome in the male rat. Journal of Andrology 29 408–417. (https://doi.org/10.2164/jandrol.107.004218)
SignorelloLBMulvihillJJGreenDMMunroHMStovallMWeathersREMertensACWhittonJARobisonLLBoiceJDJr 2012 Congenital anomalies in the children of cancer survivors: a report from the childhood cancer survivor study. Journal of Clinical Oncology 30 239–245. (https://doi.org/10.1200/JCO.2011.37.2938)
StangAJansenLTrabertBRusnerCEberleAKatalinicAEmrichKHolleczekBBrennerH & GEKID Cancer Survival Working Group 2013 Survival after a diagnosis of testicular germ cell cancers in Germany and the United States, 2002–2006: a high resolution study by histology and age. Cancer Epidemiology 37 492–497. (https://doi.org/10.1016/j.canep.2013.03.017)
StåhlOBoydHAGiwercmanALindholmMJensenAKjærSKAndersonHCavallin-StåhlERylanderL 2011 Risk of birth abnormalities in the offspring of men with a history of cancer: a cohort study using Danish and Swedish national registries. Journal of the National Cancer Institute 103 398–406. (https://doi.org/10.1093/jnci/djq55)
ZhanHJiangJSunQKeAHuJHuZZhuKLuoCRenNFanJ et al. 2017 Whole-exome sequencing-based mutational profiling of hepatitis B virus-related early-stage hepatocellular carcinoma. Gastroenterology Research and Practice 2017 2029315. (https://doi.org/10.1155/2017/2029315)