DNA damage in testicular germ cells and spermatozoa. When and how is it induced? How should we measure it? What does it mean?

This review surveys the causes and consequences of DNA damage in the male germ line from spermatogonial stem cells to fully differentiated spermatozoa. Within the stem cell population, DNA integrity is well maintained as a result of excellent DNA surveillance and repair; however, a progressive increase in background mutation rates does occur with paternal age possibly as a result of aberrant DNA repair as well as replication error. Once a germ cell has committed to spermatogenesis, it responds to genetic damage via a range of DNA repair pathways or, if this process fails, by the induction of apoptosis. When fully‐differentiated spermatozoa are stressed, they also activate a truncated intrinsic apoptotic pathway which results in the activation of nucleases in the mitochondria and cytoplasm; however, the physical architecture of these cells prevents these enzymes from translocating to the nucleus to induce DNA fragmentation. Conversely, hydrogen peroxide released from the sperm midpiece during apoptosis is able to penetrate the nucleus and induce DNA damage. The base excision repair pathway responds to such damage by cleaving oxidized bases from the DNA, leaving abasic sites that are alkali‐labile and readily detected with the comet assay. As levels of oxidative stress increase and these cells enter the perimortem, topoisomerase integrated into the sperm chromatin becomes activated by SUMOylation. Such activation may initially facilitate DNA repair by reannealing double strand breaks but ultimately prepares the DNA for destruction by nucleases released from the male reproductive tract. The abasic sites and oxidized base lesions found in live spermatozoa are mutagenic and may increase the mutational load carried by the offspring, particularly in the context of assisted conception. A variety of strategies are described for managing patients expressing high levels of DNA damage in their spermatozoa, to reduce the risks such lesions might pose to offspring health.


INTRODUCTION
Spermatozoa are highly specialized, terminally differentiated cells that are designed to achieve fertilization of the oocyte and initiation of a program of embryonic development that will culminate in the birth a healthy individual.The successful initiation of embryogenesis depends, in turn, on the ability of the spermatozoon to deliver to the egg both an appropriately programmed paternal genome and a range of epigenetic factors that will shape the developmental process in response to a variety of metabolic and environmental factors.[3] Paternally-mediated disruptions to embryonic development occur when the integrity of the paternal genome is compromised [4][5][6] or when the spermatozoon's repertoire of developmental epigenetic regulators is damaged. 7,8The evident importance of paternally mediated genetic/epigenetic changes to the long-term health trajectory of the offspring, raises important questions about how such damage arises, how it can be detected and the relative significance of such factors in the etiology of human disease.This review sets out to discuss such factors in order to shed light on when DNA damage occurs, how it occurs, what it means and how it can be clinically managed.The available information is surveyed with reference to three discrete stages of germ cell development-spermatogonial stem cells, differentiating germ cells and fully formed spermatozoa.

DNA damage in the spermatogonial stem cell population
In general, the male genome is particularly resistant to any form of disruption, as befits a cell charged with the responsibility of delivering its precious genetic cargo to the oocyte in an unmodified state.Thus within the spermatogonial stem cell population, DNA proof reading and repair are excellent, giving the male germ line one of the lowest spontaneous mutation levels in the body. 9As an exemplar of this resistance to damage, it is possible to treat male mice with the kinds of chemotherapeutic regimes used in clinical practice to treat cancer (bleomycin, etoposide and cisplatin) to the point that they are rendered completely azoospermic, and yet still the male germ cells will recover and generate normal healthy offspring, with no evidence of an enhanced mutational load. 10The resistance of the male germ line to mutation has been carefully investigated using transgenic mouse lines with recoverable mutation reporter genes, such as the Big Blue mouse.
These studies have confirmed that the male germ line has a significantly lower spontaneous mutation rate than somatic tissues 9 and, furthermore, this rate increases much more slowly with age in the paternal germ line than in the soma, as exemplified by tissues such as the liver. 11ese findings have recently been confirmed in human tissues where, in keeping with the disposable soma hypothesis (which posits that the germ line will be protected at the expense of the soma, which is allowed to age in a trade-off between reproductive efficiency and longevity) de novo mutation rates were found to be much lower in testicular germ cells than in any of the somatic tissues investigated. 10,12A surveillance and repair are therefore extremely efficient in the spermatogonial stem cell population and de novo mutations in this cell type are relatively rare.The only real exception to this rule is the mutations that accumulate in the male germ line as a consequence of age.Paternal age is regarded a powerful driver of single-nucleotide and insertion-deletion germline mutations.The mechanism underlying the paternal age effect is traditionally thought to involve replication error within the spermatogonial stem cell population.The number of rounds of replication that the male germ line undergoes, increases as a linear function of paternal age.If we assume an average of 30 divisions before puberty and one stem cell division every 16 days (23 every year) thereafter, then the total number of chromosome replications has been calculated to increase from 35 at 15 years to 840 at the age of 50. 13With every round of replication there is a risk that an error in the fidelity of DNA duplication might occur, generating a progressive increase in mutational load. 12,14There can be no doubt that this background rate of spontaneous germ line mutations increases progressively with parental age with a vast majority (∼75%) coming from the father. 15,16However there is some doubt that this is entirely due to replication error.Such reservations stem largely from the observation by Gao et al. 16 that the profound difference in spontaneous mutations rates between the male and female germ line is evident from the earliest stages of reproductive life before gender-specific differences in germ line replication rate have had a chance to take effect.An alternative hypothesis is that the male germ line is more vulnerable to DNA damage than its female counterpart and that the progressive increase in de novo mutations with paternal age reflects a gradual deterioration in the quality of DNA repair.Certainly defects in DNA repair genes can lead to a marked increase in the incidence of spontaneous mutations observed in the male germ line. 15Furthermore, a history of paternal chemotherapy using clastogenic drugs is also known to increase the levels of de novo mutation observed in the human paternal germ line 15 (in contradistinction to the mouse data cited above 10 ) even though there is no evidence of a significant increase in birth defect rates following such treatment. 17Such results suggest that DNA damage and aberrant repair in the spermatogonial stem cell population is a factor in the etiology of spontaneous genetic mutations in the germ line.
Even though the incidence of such de novo mutations increases with paternal age, they are extremely rare and seldomly associated with the appearance of pathology in the offspring.
The one situation is which this assertion is untrue is in the case of certain dominant genetic diseases such as achondroplasia and Apert syndrome.The complex array of developmental defects associated with these conditions is known to be induced by mutations in fibroblast growth factor receptors (FGFR)-FGFR2 in the case of Apert syndrome and FGFR3 in the case of achondroplasia.The odds that such specific mutational changes might randomly occur in a genome containing over 3 billion base pairs are extremely small and bear no relation to the observed incidence of conditions such as achondroplasia, which can be as high as 35 per 100,000 in North Africa and the Middle East and around 4.6 per 100,000 worldwide. 18e unusually high prevalence of dominant genetic diseases such as Apert syndrome has been explained by the "selfish selection hypothesis," which posits that the FGFR mutations responsible for these conditions confers upon the mutated germ cells a selective advantage that allows their numbers to increase dramatically, facilitating the clonal expansion of mutant cells at locations along the seminiferous tubule. 19,20us, the seminiferous tubules of a 70+ year-old male will typically reveal foci of mutant germ cells, each one of which is capable of differentiating into a functional spermatozoon that will result in the appearance of dominant genetic disease in the progeny. 19This concept may also apply to other dominant genetic diseases impacted by paternal age, including multiple endocrine neoplasia type 2 B, where the causative mutation occurs in the RET proto-oncogene, 21 Noonan syndrome, driven by a mutation in the PTPN11 gene (encoding protein tyrosine phosphatase non-receptor type 11), 22 and Costello syndrome which is associated with a mutation in the HRAS (Harvey rat sarcoma viral oncogene homolog proto-oncogene, GTPase) gene. 23her rare conditions (so called RASopathies) are also associated with mutations in genes encoding components of the RAS/RAF/MAPK pathway including LEOPARD Syndrome, Hereditary Gingival Fibromatosis, Neurofibromatosis Type 1, and Autoimmune lymphoproliferative Syndrome, among others. 24,25Thus a range of spontaneous mutations can occur in pathways that confer upon their descendants a selective advantage over their non-mutant counterparts, permitting their clonal expansion.Thus the "selfish selection hypothesis," presents us with an extremely powerful mechanism by which rare spontaneous mutations in the male germ line can become responsible for a range of well established, autosomal, dominant conditions in our species.Such special cases apart, the spermatogonial stem cell population should generally be regarded as very resistant to the induction of genetic damage as a result of excellent DNA surveillance and repair.

DNA damage during spermatogenesis
Similarly, once these stem cells enter the spermatogenic pathway and commence their differentiation into spermatozoa, DNA integrity is generally well maintained within the germ cell population.One of mechanisms by which genetic integrity is preserved during spermatogenesis may involve a process known as "transcriptional scanning."The testes are characterized by the highest level of gene transcription of any organ in the body. 26This high rate of gene transcription provides the testes with the opportunity to survey the genome for DNA damage and, if any is detected, activate transcription-coupled DNA repair. 27e induction of DNA damage during the early stages of spermatogenesis thus leads to the activation of this, and other, DNA repair mechanisms (including nucleotide excision repair, base excision repair, mismatch repair, homologous recombination and non-homologous end joining) that attempt to correct the damage.If this repair process is inadequate or inefficient, then apoptosis is induced and the DNAdamaged germ cell is eliminated from the germinal epithelium.This chain of cause-and-effect is classically induced by testicular heating.
If the testes are subjected to heat stress then DNA repair is impaired leading to the appearance of double strand breaks in differentiating germ cells at meiotic prophase 1.Such damage to secondary spermatocytes can lead to asynapsis between homologous chromosomes which are eliminated by apoptosis at the meiotic checkpoint. 28Age is also a major factor in the induction of DNA damage in the male germ line partially because of a progressive impairment of the nonhomologous end-joining and homologous recombination repair pathways with increasing age.Interestingly, in the same cells, the base excision repair pathway is actually upregulated with paternal age suggesting that the latter is associated with increased oxidative stress. 29Other factors that can induce DNA damage in the male germ line and in so doing, enhance the rate of apoptosis in differentiating testicular germ cell population, include common environmental toxicants such as bisphenol A, 30 as well as clinical treatments including chemo-and radiotherapy. 31,32, while multiple mechanisms exist for preserving DNA integrity during spermatogenesis, occasionally conditions may arise whereby these protective strategies are overwhelmed and the DNA becomes damaged.When this occurs, DNA fragmentation can only be induced via 2 mechanisms-enzymatic digestion of the DNA by nucleases or free radical attack.During spermatogenesis both of these mechanisms appear to be operative.There are some situations, such as exposure to toxicants or excessive heat, that are associated with a clear increase in oxidative stress within the testes, increasing the formation of lipid aldehydes and oxidative DNA damage that then precipitate an apoptotic state. 32,33Equally there are other conditions, such as diabetes, that induce testicular germ cell apoptosis by lowering phosphoinositide 3 kinase expression that then leads to the generation of excess ROS. 34,35Oxidative stress and apoptosis are therefore closely intertwined in the induction of testicular DNA damage; apoptosis inducing a significant increase in the intracellular generation of ROS while oxidative stress can also activate the intrinsic apoptotic cascade.As a result, both oxidative DNA damage and apoptosis-related, nuclease-induced DNA fragmentation are observed when the testes is under stress, whether this is chemically, 36-38 psychologically, 39 or surgically 40 induced or the result of exposures to excessive ionizing radiation 41 or heat. 42[52] The ability of commercial antioxidant formulations to resolve the testicular pathology associated with oxidative stress, generated either through the local application of heat or genetic inactivation of glutathione peroxidase 5 (one of the major antioxidant enzymes in the epididymis) has also been clearly demonstrated in animal models. 53tecting DNA damage in testicular germ cells is traditionally achieved via the histological examination of testicular biopsy material, which is a significant intervention.However, the discovery of M540bodies by Marchiani et al. 54 may offer an alternative approach to recovering such diagnostic information.M540-bodies are thought to be the remnants of deceased germ cells that have undergone apoptosis at some stage of spermatogenesis or epididymal maturation.These round bodies are detected in human semen samples and express multiple markers of apoptosis and contain fragmented DNA. 54Interestingly, a higher content of M540 bodies was found in oligoasthenoteratozoospermic and asthenoteratozoospermic samples suggesting that the sperm DNA damage associated with these conditions may be associated with a high incidence of apoptosis earlier in germ cell development. 55,56It is clear that the presence of such bodies provides interesting insights into the functional state of the testes 57 and while much has been written about the potential of such bodies to interfere with the measurement of DNA damage in spermatozoa, 58,59 very little attention has been given to the significance of M540 bodies in understanding the testicular origin of DNA damage in the germ line.The presence of these structures does however tell us that the testicular response to stress is to eliminate affected germ cells via an apoptotic mechanism.It therefore follows that the DNA damage we see in mature spermatozoa either represents germ cells that have somehow escaped apoptotic elimination from the germ epithelium or, more likely, cells that have become damaged during their posttesticular journey through the male reproductive tract.The finding that DNA damage is frequently higher in epididymal than testicular spermatozoa certainly implies that post testicular DNA damage can occur in spermatozoa and that the epididymis might be a particularly challenging environment for these cells, particularly if their chromatin has not been efficiently remodeled and compacted during spermiogenesis.nucleases activated during a truncated intrinsic apoptotic cascade. 62 interesting structural feature of spermatozoa is that, uniquely amongst all cell types, the sperm nucleus is housed in a different cellular compartment from the mitochondria, which are heavily involved in orchestrating the apoptotic response (Figure 1).As a result of this unusual structural arrangement, even when these cells are triggered to undergo apoptosis, nucleases released from the mitochondria or activated in the cytoplasm cannot, as in a conventional somatic cell, move into a centrally positioned nucleus to cleave the DNA (Figure 1).In spermatozoa, even though nucleases may be activated and released during the apoptotic process, they cannot reach the sperm nucleus in order to induce DNA fragmentation.Even if they did somehow manage to cross such compartmental barriers, they would find it difficult to penetrate DNA that has been compacted to the point of near-crystallization, in contrast to the dispersed chromatin typical of somatic interphase nuclei.The only elements of the apoptotic process that have the capacity to move between different cellular compartments and penetrate sperm chromatin are ROS such as hydrogen peroxide.The latter is generated in copious amounts by the sperm mitochondria during apoptosis and, being uncharged, is free to move across the membranes isolating the sperm nucleus, and attack regions of the nuclear DNA that are not fully compacted and complexed with protamines. 62,63The structural uniqueness of spermatozoa therefore explains why a majority of DNA damage in live spermatozoa appears to be initiated by an oxidative attack rather than the nuclear incursion of activated nucleases, as in most somatic cells.
Thus, once spermiogenesis is complete, and the spermatozoa have lost their capacity for DNA repair, they become very susceptible to oxidative attack that can lead to the formation of oxidative base adducts in the form of 8-hydroxy-2′-deoxyguanosine (8OHdG) and, ultimately, DNA fragmentation.Since, in the population of male patients attending infertility clinics, around 70% of the spermatozoa were found to exhibit significant levels of oxidative DNA damage 64 we can conclude that such damage is frequently the precursor of DNA fragmentation.The involvement of oxidative stress in the initiation of DNA damage, explains why 8OHdG expression is largely observed in the live sperm population, in association with peroxidation biomarkers such as malondialdehyde. 65However, once the spermatozoa have started to die and plasma membrane integrity has been breached, a new set of mechanisms take over to ensure the complete destruction of the spermatozoa's DNA.These mechanisms involve two main elements: topoisomerase and extracellular nucleases released by the male reproductive tract. 66poisomerases are nuclear enzymes involved in the induction of double strand breaks during spermatogenesis in order to relieve the torsional stresses involved in chromatin remodeling. 67These strand breaks are normally resolved before the end of spermatogenesis 68 but if this process is deficient, then the strand breaks will persist in the mature gamete and may well impact the spermatozoon's ability to establish a normal pregnancy. 69It is also possible that the persistent presence of topoisomerase (particularly topoisomerase 2A) in defective spermatozoa 70 can contribute to DNA damage as these cells enter senescence.In the mature gamete, topoisomerase appears to be activated following its association with small ubiquitin-like modifier 1 (SUMO1). 71Interestingly, SUMOylation of proteins is known to be promoted by oxidative stress signals [72][73][74] and high concentrations of hydrogen peroxide have been shown to stimulate the SUMOylation of proteins in mouse testicular germ cells.Furthermore, in this cell type SUMO has been shown to localize to the sites of DNA F I G U R E 1 Apoptosis and DNA damage in fully differentiated spermatozoa.(A) In somatic cells apoptosis conventionally involves the release of nucleases from the mitochondria or their activation in the cytoplasm.These nucleases then infiltrate a centrally placed nucleus and cleave the DNA at the inter-nucleosomal linker regions.In contrast, spermatozoa have a unique architecture that prevents DNA fragmentation during apoptosis via the conventional pathways.The nucleus is in a different compartment of the cell than the mitochondria and a majority of the cytoplasm.As a result, any nucleases activated during the intrinsic apoptotic cascade would find it difficult to gain access to the nuclear compartment.Furthermore, even if they did, they would find a highly condensed chromatin, in which the DNA has reached the physical limits of compaction, quite unlike the dispersed chromatin structure of the interphase nucleus in somatic cells.The only factors generated by the sperm midpiece during apoptosis that have the capacity to penetrate the nucleus and damage the DNA are ROS such as hydrogen peroxide.(B) It is for this reason that we find an excellent correlation between mitochondrial ROS generation and 8OHdG formation in human spermatozoa that, in this case, have been stimulated to undergo apoptosis in vitro by radiofrequency electromagnetic radiation. 101(C) Experiments in which apoptosis has been artificially induced using the phosphoinositide 3-kinase inhibitor, wortmannin, confirm that while nucleases such as Apoptosis Inducing factor (AIF) or Endonuclease G (EndoG) become activated during this process, they remain resolutely locked within the sperm midpiece and are unable to penetrate the nucleus located in the sperm head 62 ; hence the importance of ROS, such as hydrogen peroxide, in the initiation of sperm DNA damage.
double-strand breaks in stressed germ cells and to associate specifically with topoisomerase. 75ven this chain of known associations it is possible to elaborate a novel hypothesis about the origins of DNA damage in spermatozoa.
According to this model (Figure 2), a major instigator of DNA damage in human spermatozoa is oxidative stress.Under normal physiological circumstances, the generation of ROS is just sufficient to activate the redox mechanisms that regulate sperm capacitation. 6If the levels of oxidative damage start to rise, one of the major antioxidant enzymes employed by spermatozoa, peroxiredoxin 6 (PRDX6) moves to the site of damage in order to suppress the oxidative attack. 76,77At the same time, the base excision repair (BER) pathway becomes activated and any oxidized bases are cleaved from the DNA by 8-oxoguanine DNA glycosylase 1 (OGG1), creating an alkali-labile abasic site, which can be readily detected with the comet assay. 78Since the factors needed to complete the next stage of this repair pathway [apurinic endonuclease 1 (APE1) and X-ray repair complementing defective repair in Chinese hamster cells 1(XRCC1)] are not present in spermatozoa, the abasic site remains temporarily unprocessed, in anticipation that the BER pathway will be completed after fertilization, by the oocyte, which contains APE1 and XRCC1 in abundance. 78At low levels of oxidative stress, topoisomerase also becomes activated by SUMOylation and may contribute to DNA repair, possibly as a result of its DNA relaxing and reannealing activities. 79 the level of oxidative stress is slightly higher, the BER capacity of the cell becomes overwhelmed and the sperm chromatin will not only Proposed stages in the induction of DNA damage in human spermatozoa.In Stage 1, the spermatozoa are functionally normal with a low level of ROS generation supporting the redox pathways that control capacitation.In Stage 2, ROS generation is increased, possibly because the products of peroxidation, particularly lipid aldehydes such as 4-hydroxynonenal, promote mitochondrial ROS generation in a self-perpetuating cycle. 62The increased ROS generation stimulates DNA oxidation generating 8OHdG, some of which is removed from the cell by OGG1 generating multiple abasic sites that are alkali-labile and, therefore, readily detected by the comet assay.At this stage, the spermatozoa still retain their capacity for fertilization, however the presence of abasic sites as well as unresolved 8OHdG residues in the fertilizing spermatozoon increases the risk of miscarriage and an increased mutational load in any foetus that survives to term.In Stage 3, a further increase in the levels of oxidative stress suffered by the spermatozoa leads to extensive DNA strand breaks, high levels of 8OHdG expression and a severe loss of fertilizing potential.At this point in the pathway, topoisomerase activity becomes activated by SUMOylation, which may further increase the levels of DNA fragmentation.In Stage 4, the spermatozoa enter the perimortem, ROS generation has ceased, plasma membrane integrity has been breached and the reannealing properties of topoisomerase is thought to prepare the spermatozoa for attack by extracellular nucleases released by the male reproductive tract.This nuclease attack subsequently ensures complete destruction of the spermatozoa's DNA. 66,83press abasic sites generated by OGG1 but also unresolved 8OHdG lesions that will have to be subsequently processed in the oocyte in the time interval between fertilization and the initiation of cleavage.
While the oocyte does contain abundant APE1 and XRCC1, it contains low levels of OGG1, despite the upregulation of BER activity in the oocyte following fertilization. 80As a result of this relative deficiency in OGG1, some 8OHdG lesions will persist into S-phase of the first mitotic division and potentially induce de novo mutations in the embryo as a consequence of inadequate or erroneous DNA repair.This is the basis of the "post-meiotic oocyte collusion hypothesis" which essentially posits that an important source of de novo mutations in our species involves the defective repair of oxidative sperm DNA damage by the oocyte following fertilization. 81In this context, it may be significant that the de novo mutational load is significantly increased in children conceived by ART, with a vast majority of mutations (87.9%) originating in the father's germ line in association with a diagnosis of male infertility. 82 the oxidative stress is more severe and spermatozoa begin to suffer a loss of vitality and membrane integrity, topoisomerase, which is integrated into the sperm chromatin, becomes further activated by SUMOylation. 71The topoisomerase, possibly by re-annealing double strand breaks within the chromatin, prepares the DNA for complete destruction by extracellular nucleases released from the male reproductive tract, particularly the vas deferens. 83The SUMOylation of topoisomerase may also be critical in the induction of sperm DNA fragmentation by extracellular DNA.which does not seem to involve oxidative stress but activation in nucleases within the sperm chromatin. 84The model summarized in Figure 2 is clearly speculative in nature but is designed to provide a testable framework for future studies in this area.

How should we measure sperm DNA damage?
The various assays that have been developed for measuring DNA damage in human spermatozoa appear correlate highly with one other even though they measure divergent aspects of the DNA damage process.Thus, the sperm chromatin structure assay (SCSA) measures the presence of acid-labile sites within the DNA, the appearance of which is highly correlated (r = ∼0.7-0.8) with the presence of strand breaks as measured by the TdT-mediated-dUTP nick end labeling (TUNEL) and the sperm chromatin dispersion (SCD) test. 85Similarly the alkaline comet assay, which measures alkali labile sites in the DNA, is significantly correlated with TUNEL, SCSA and SCD assays (r = ∼0.6-0.7). 86,87Moreover, all of these assays were found capable of differentiating between fertile and infertile populations, with the comet assay providing the highest level of sensitivity (0.85) and specificity (0.92). 86ven the fundamental importance of oxidative DNA damage in the etiology of sperm DNA fragmentation it is not surprising that measurements of 8OHdG expression in human spermatozoa are also able to discriminate between semen donors and infertile patients. 880][91] The oxidative stress associated with the cryopreservation of spermatozoa, 92,93 exposure to polycyclic aromatic hydrocarbons, 94 leukocytospermia, 64 type 1 diabetes, 95 cigarette smoking, 96 the process of apoptosis, 65 fertility status 97,98 and defective sperm quality, 99 as well as direct exposure of spermatozoa to electromagnetic radiation 100,101 or environmental estrogens such as bisphenol A, 102 have all been found to increase 8OHdG levels in spermatozoa.Since oxidative damage is so commonly involved in the etiology of DNA strand breaks it is not surprising that 8OHdG expression by human spermatozoa was found to be positively correlated with DNA fragmentation as measured by TUNEL, SCSA and alkaline comet assays. 88,95,103en comparing different methods of assessing DNA damage one factor we should be aware of, is timing.If spermatozoa are exposed to an oxidative stress they become rapidly (within 2 h) 8OHdG and SCSA positive and the two measures are highly correlated.However, analysis of the same populations of spermatozoa with the TUNEL assay reveals no evidence of damage whatsoever.
If the same cells are then left for 24 h, the TUNEL results are increased and significantly correlated with the outcome of the SCSA assay, while incubation for a further 24 h reveals an even closer correlation between the TUNEL and SCSA (R 2 = 0.85) in association with a significant decline of cell viability. 78The TUNEL assay is therefore detecting a later stage in the DNA damage process than SCSA or 8OHdG and may be reflecting the double strand breaks induced by topoisomerase following the activation of this enzyme by SUMOylation. 67

What does sperm DNA damage mean?
The relationship between sperm DNA damage and clinical outcomes, particularly pregnancy, has been shrouded in controversy from the inception of this field.For example, analyses of the relationship between sperm DNA damage and clinical pregnancy rates following intrauterine insemination (IUI) therapy indicated that high levels of sperm DNA fragmentation were significantly associated with lower pregnancy rate, 104 as were high levels of oxidative sperm DNA damage. 97However, other studies have failed to find such a relationship 105  where the functional competence of the male gamete and the integrity of the sperm DNA are largely irrelevant. 106,107The quality of sperm DNA also has a variable impact on embryo quality in IVF/ICSI cycles, with only 21/62 (34%) of studies examined finding a consistent negative relationship. 106The exception appeared to be analyses conducted using the alkaline comet assay, which revealed an association between high levels of sperm DNA fragmentation and impaired embryo quality in 67% of papers published. 106The assessment of 8OHdG levels in human spermatozoa has also been found to correlate well with the quality of embryos generated in IVF cycles, 108 possibly because this base adduct is a feature of live cells that have the potential to participate in fertilization.Subsequent to embryo implantation, a much more consistent relationship has been observed between high of levels of sperm DNA damage and clinical pregnancy outcome, regardless of whether IVF or ICSI is used to achieve fertilization. 106,109,110Inconsistencies between studies in detecting this relationship between sperm DNA damage and clinical pregnancy rates may simply reflect a lack of rigour in terms of the patient selection criteria, with many studies failing to eliminate couples with female factor infertility for whom sperm DNA quality is an irrelevance. 106spite the negative relationships between high levels of sperm DNA damage and fertility described above, the Practice Committee of the American Society for Reproductive Medicine (ASRM) has concluded that "current methods for assessing sperm DNA integrity do not reliably predict treatment outcomes and cannot be recommended routinely for clinical use" (Practice Committee of the American Society for Reproductive Medicine, 2013). 111More recently the American Urological Association in association with the ASRM have commented that "Sperm DNA fragmentation analysis is not recommended in the initial evaluation of the infertile couple" because "There are no prospective studies that have directly evaluated the impact of DNA fragmentation testing on the clinical management of infertile couples (i.e., that the fertility outcomes of those who had testing are different from those who did not)." 112The point that appears to have been missed in arriving at such a conclusion is that the impact of sperm DNA damage on clinical pregnancy rates is just the tip of the iceberg.[115] Indeed, the AUA/ASRM guidelines 112 acknowledge that "it is possible that very high levels of sperm DNA fragmentation will have a more substantial adverse impact on pregnancy outcomes with IVF as well as an increased risk of miscarriages."However, it is also theoretically possible that lower levels of sperm DNA damage may still influence the health and wellbeing of the progeny even when pregnancy and parturition are apparently normal, through the induction of genetic and epigenetic mutations in the embryo.It is now well established that the paternal genome represents the source of most de novo mutations in our species and that such mutations arise as a result of defective DNA damage repair. 81To resolve this important issue, it is now critical that studies are performed exploring the association between DNA damage in spermatozoa and the mutational load carried by the offspring.
If such DNA damage occurs early in spermatogenesis, for example in the stem cell population, then it is possible for mutations to be created that are present in the spermatozoa at ejaculation.This is the case, for example, with the dominant genetic mutations occurring in the fibroblast growth factor receptors responsible for conditions such as Apert syndrome and achondroplasia. 116However, we would argue that a more common mechanism for inducing de novo mutations in the offspring is the defective repair of sperm DNA damage in the oocyte following fertilization in accordance with the post-meiotic oocyte collusion hypothesis described above. 81As a result of such mechanisms, the presence of DNA damage (particularly oxidative DNA adducts) in spermatozoa would be expected in increase the incidence de novo mutations in the newly formed embryo and, through such mechanisms, have a long-term impact on the health trajectory of the progeny.Such a mechanism has been clearly demonstrated in the Gpx5 knockout mouse where the selective induction of oxidative DNA damage in epididymal spermatozoa has been clearly linked with the occurrence of miscarriages and developmental defects in the offspring of wild-type female mice mated to GPx5 KO males over 1 year of age. 117,118Moreover, recent human data suggest a similar mechanism for the increased de novo mutation rate observed in children conceived in vitro. 82Given this chain of associations between DNA damage in mature spermatozoa and the mutational load carried by the offspring we have previously argued that such damage should be assessed clinically as a matter of good practice.since it will potentially impact the health of any progeny generated as a result of assisted conception therapy. 119

How should sperm DNA damage be managed
Monitoring DNA damage in spermatozoa is not only important because it may represent a challenge to the genetic and epigenetic integrity of the offspring but also because it should influence clinical practice.Once extensive DNA damage has been detected in spermatozoa, a number of therapeutic options are available to ameliorate the situation.For example, the recovery of testicular spermatozoa has been proposed as a strategy to combat elevated levels of DNA fragmentation in ejaculated human spermatozoa, based on the understanding that significant DNA damage may occur as spermatozoa transit the epididymis. 120Adoption of sophisticated sperm isolation procedures (swim up, electrophoretic separation, density gradient centrifugation, microfluidic systems) that limit the concentration of DNA damaged spermatozoa in the suspensions used for IVF or ICSI could also be used to mitigate the risk of generating mutations in the zygote.This strategy could even include the incorporation of antioxidants into the media used for isolation purposes including vitamin C and vitamin E. [121][122][123] When patients are about to engage in a form of treatment that is known to compromise DNA integrity in the germ line, such as radio-and chemo-therapy for cancer, consideration could also be given to cryostoring the spermatozoa before treatment is instigated.
In pursuing this approach, we should be awarethat the methods we currently use to cryopreserve human spermatozoa induce a significant increase in oxidative DNA damage and, in this sense, may be counterproductive. 92Improved cryostorage protocols are clearly needed to optimize this approach to patient management.Finally, if significant levels of oxidative DNA damage are detected in the spermatozoa then consideration may be given to a course of antioxidant therapy.In animal models there is incontrovertible evidence that if oxidative stress in spermatozoa is experimentally induced by, for example, testicular heating or deletion of the GPx5 gene, then normal reproductive function can be restored through the administration of exogenous antioxidants. 53Attempts to conduct clinical trials in this area have been an egregious waste of time because, in general, the patients were not selected on the basis of evidence that their spermatozoa were suffering from oxidative stress. 124Giving powerful antioxidants to patients who are not suffering from oxidative stress.runs the risk of creating a state of reductive stress that will further compromise, rather than cure, the genetic integrity of their spermatozoa.

CONCLUSIONS
Monitoring DNA damage in the germ line is important for four reasons.Firstly, it is a biomarker for impaired male reproductive function, reflecting the impact of age, and a variety of environmental, lifestyle and clinical factors, on the production and maturation of spermatozoa.Secondly, when the damage is severe, it may be associated with an impaired capacity for fertilization or, even if conception is achieved, provide a robust indication of pending miscarriage. 112,125Thirdly, even if the pregnancy carries to term, DNA damage in spermatozoa may act as a precursor for genetic and epigenetic mutations that will impact the entire health trajectory of the offspring. 119Fourthly, the presence of significant DNA damage in spermatozoa should alter the clinical management of the patient in terms of the isolation procedures employed to separate spermatozoa from seminal plasma, the decision as to whether ejaculated or testicular spermatozoa should be used for ICSI, whether the spermatozoa might be cryopreserved for use in a subsequent ART cycle and whether the use of antioxidant therapy is advised.
possibly because there are multiple factors outside of sperm quality that will determine the outcome of IUI therapy, not least the timing of the insemination relative to ovulation, oocyte quality and the relative hostility of the female reproductive tract.More definitive conclusions have been reached concerning the clinical significance of sperm DNA damage in the case of ICSI and IVF.Thus, a recent metaanalysis of this topic has found that relationships between sperm DNA fragmentation and fertilization are weak, particularly in the case of ICSI