Individual capacity for DNA repair and maintenance of genomic integrity: a fertile ground for studies in the field of assisted reproduction

ABSTRACT

Many factors may affect the chances for successful pregnancy, especially at a later age. Fertility evaluations including genetic analysis are recommended to couples that have not achieved pregnancy within 6–12 months of unprotected intercourse. This review discusses some of the common polymorphisms in genes coding for proteins functioning in DNA damage identification and repair and maintenance of genomic integrity that may affect the chances of success in natural conception as well as in assisted reproduction (AR). Common polymorphisms in genes coding for proteins functioning in DNA damage identification and repair and maintenance of genomic integrity may affect the chances of success in assisted reproduction as well as in natural conception. The effects of carriership of different alleles of key genes of DNA repair may have differential effects in men and women and at different ages, suggesting complex interactions with the mechanisms controlling cell and tissue aging and programmed cell death. Future studies in the field are needed in order to elucidate the genotype–phenotype relationships and to translate the knowledge about individual repair capacity and maintenance of genomic integrity to potential clinical applications.

Abbreviations: aCGH: microarray-based comparative genomic hybridization; AR: assisted reproduction; ATM: ataxia-telangiectasia mutated; ATP: adenosine triphosphate; BER: base excision repair; BFE: basic fertility evaluation; DMSO: dimethyl sulfoxide; FSH: follicle-stimulating hormone; GNRHR: gonadotropin-releasing hormone receptor; HMG: high-mobility group; ICSI: intracytoplasmic sperm injection; IUI: intrauterine insemination; IVF: in vitro fertilization; LH: luteinizing hormone; LIF: leukaemia inhibitory factor; MTR: methionine synthase; MTRR: methionine synthase reductase; NGS: next-generation sequencing; NER: nucleotide excision repair; NHEJ: non-homologous end joining; PAH: polycyclic aromatic hydrocarbons; PCOS: polycystic ovarian syndrome; R: restriction point; ROS: reactive oxygen species; VEGF: vascular endothelial growth factor

KEYWORDS:

• DNA repair
• individual repair capacity
• assisted reproduction
• fertility

Infertility, subfertility and delayed parenthood: definitions and trends

Infertility is usually defined as failure to achieve clinical pregnancy within 12 months of unprotected intercourse in the fertile phase of the menstrual cycle.[1] Subfertility is a more diffuse term used to describe reduced fertility with prolonged time to desired conception.[1] The pregnancy rates following unprotected intercourse are differentially distributed in time, with 80% of the pregnancies occurring within the first six months of unprotected intercourse.[1] Another 10% of couples actively trying to conceive are likely to achieve conception within 12 months of trying. Approximately half of the remaining 10% would achieve unassisted conception within the following 36 months. The remaining 5% of couples that have not conceived within 48 months of unprotected intercourse have very low future chances for unassisted normal pregnancy. The proportion of male-to-female infertility in couples with reproductive issues is roughly 1:1. Female infertility is the major issue in about half of the infertile couples.[2] According to the same study, male infertility alone accounts for 20%–30% of infertile couples and may contribute to infertility in another 20%–30%.

The proportion of couples who postpone parenthood until later in life (in their thirties and forties) tends to rise steadily on a global scale, especially in developed countries. For many of them (70%–80%), normal conception would be achieved within the timeframe typical of younger couples (within 6–12 months). In 15%–20% of those who have not conceived within 6–12 months, fertility evaluations may reveal no apparent issues and they may be diagnosed with unexplained infertility. As the underlying cause for infertility is unknown, fertility treatments do not significantly increase the chances of couples with unexplained infertility to conceive. Such couples are usually advised to wait and continue with fertility-focused intercourse (unless the female partner is over 40), as the pregnancy rates tend to reach 85–90% within 48 months even in women in the age group of 35–40. The remaining 10%–15% of infertile couples (and, potentially, those with unexplained infertility where the female partner is 40 or older) may benefit from assisted reproduction (AR).

The ability to conceive is only one factor in the capacity to produce a live-born child. The rates of miscarriage grow steeply in fertile women after the age of 30, reaching about 45% in women over 40.[3] Chromosomal anomalies are found in not less than 60% of karyotyped abortuses of women at any age, although the percentage increases with age.[4] The role of paternal age in the constitution of the risk of chromosomal abnormalities in the developing embryo is, so far, controversial.[5]

At present, it is difficult to precisely assess the chances of a couple with unexplained infertility conceiving naturally (even for younger couples) and/or the chances of benefitting from AR using the couple's own gametes. Whether the couple would eventually conceive after extended waiting, or whether that particular cycle of AR would be successful is still a matter of chance. Some couples conceive more easily than others, even when at the same age, with similar fertility issues and after same AR treatments.

Basic and extended infertility assessments: pros and cons

Female fertility is generally more prone to effects of physiological aging. It generally begins to decline after 32 years of age, but the effect of advanced age becomes significant after 35–37. Thus, the timing of recommended fertility evaluation in couples where the female partner is less than 35 years of age is after 12 months of fertility-focused intercourse and 6 months (or sooner) in couples in which the female partner is over 35 years of age.[6] Male infertility is also age-dependent, although to a much lesser degree. Advanced paternal age may influence semen volume, sperm count, motility and morphology, and may be associated with increased prevalence of genomic defects in spermatozoa, all of which may increase the risk of subfertility and infertility in men older than 55.[7] For couples who remain infertile after 6–12 months of trying to conceive naturally, basic fertility evaluation (BFE) is usually recommended in order to identify the main issue/s and, potentially, recommend strategies to overcome infertility. BFE usually includes assessment of the medical and surgical history and the lifestyle of both couple members; semen analysis for the male partner; gynaecological and obstetric history, pelvic examination, thyroid hormones, assessment of ovulation (by serial ultrasound and hormone profiles [luteinizing hormone (LH), follicle-stimulating hormone (FSH), progesterone] and a hysterosalpingogram to evaluate tubal patency – or the female partner. In women with unexplained infertility and normal hysterosalpingographic findings, diagnostic laparoscopy may be indicated.[8] Testing for infectious agents commonly associated with infertility may be part of BFE. It includes analysis for infection with Chlamydia trachomatis and Toxoplasma gondii for both the male and female partner and infection with mycoplasms (specifically, Mycoplasma genitalium, Mycoplasma hominis and Ureaplasma urealyticum) – usually, for the female partner, although men may be reservoirs for infection.[9] Some authors believe that at least part of the cases of male infertility may be attributed to infection with mycoplasms and, therefore, advocate testing for infection in both partners.[10]

Cytogenetic analysis is often included in fertility evaluations, as chromosomal abnormalities may account for a significant proportion of the cases of infertility (4%–20%, according to different studies).[11–13] In males, the most common aneuploidies diagnosable by cytogenetic analysis are Klinefelter syndrome (47, XXY), the (47, XYY) syndrome and de la Chapelle syndrome (46, XX male). In females, the most common aneuploidies that may affect fertility are Turner syndrome (45, X0) and its variants and the (46, XY) female syndrome. Structural abnormalities, such as inversions, Robertsonian or reciprocal translocations, are also commonly found in infertile males as well as females. Besides its diagnostic value in the identification of the cause of infertility in a couple, cytogenetic analysis as a component of fertility evaluations allows for prognostication of the chances of having healthy offspring using AR technologies, including assessment of the need for prenatal or preimplantation diagnosis.

Advanced hormonal assessments (estradiol, inhibin B, anti-Müllerian hormone) and ultrasound assessments (antral follicle count, ovarian volume) to assess ovarian reserve [14] may be additionally recommended for couples with unexplained infertility and/or those that are candidates for AR, as the ovarian reserve may be directly associated with the chances of natural conception (albeit within an extended timeframe) as well as with the chances for response to ovarian stimulation.

Assessment of unexplained infertility, recurrent miscarriages and stillbirths may include additional genetic analysis. The basic panel includes analysis for carriership of the common mutations increasing the risk of thrombosis: usually, Factor V Leiden (FV G1691A); the G20210A mutation in the prothrombin gene (PT G20210A), the C677T and A1298C mutation in the methylenetetrahydrofolate reductase (MTHFR) gene (MTHFR C677T and A1298C) and the 4G/5G length polymorphism in the plasminogen activator inhibitor gene (PAI 4G5G). Analysis for carriership of prothrombotic mutations usually pertains to the female partner, as paternal thrombophilia is believed to be unrelated to pregnancy loss.[15] Common mutations in the genes coding for proteins of folate metabolism (the above-mentioned MTHFR C677T and 1298C mutations, but also methionine synthase (MTR) 2756A>G, methionine synthase reductase (MTRR) 66A>G and reduced folate carrier 1 (RFC-1) 80A>G) mutations may be associated with increased risk of pregnancy loss.[16] These mutations are fairly common (the prevalence of heterozygous carriership of the T677 allele is 20%–50%; of the FV Leiden mutation, about 5%; and of PT20210A, 2% in the general population.[17–19] The increase of the risk of pregnancy loss associated with carriership of any of these mutations may be significant, depending on the dose of mutant alleles in a single locus and co-carriership of different mutations.[20] The female partner may also be tested for common polymorphisms in the vascular endothelial growth factor (VEGF) gene (specifically, the 1154G/A and −2578C/A polymorphisms) because of their association with increased risk of endometriosis, idiopathic recurrent miscarriage [21,22] and recurrent implantation failure after in vitro fertilization (IVF) and embryo transfer.[23,24] All the polymorphisms listed above are quite common (although the frequencies of different alleles may vary in different populations), may have significant effect on the risk of pregnancy loss and their deleterious effects may be decreased by various treatments (use of anticoagulation in carriers of prothrombotic polymorphisms, laparoscopic removal of lesions and/or IVF/embryo transfer in those with endometriosis in order to increase the chance of conception, etc.)

There are other types of genetic analysis that may sometimes be offered to couples with infertility. Male partners may be tested for deletions in the azoospermia factor region on the Y chromosome (associated with idiopathic azoospermia or severe oligospermia),[25] mutations in the gene coding for ubiquitin-specific protease 9 (USP9Y) – also associated with azoospermia [26] and spermatogenesis-associated protein 16 – increasing the risk of globozoospermia (a rare form of teratozoospermia).[27] Both partners may be advised to be tested for mutations in the genes coding for synaptonemal complex protein 3 – associated with azoospermia [28] and with recurrent miscarriage [29]; gonadotropin-releasing hormone receptor (GNRHR) – associated with hypogonadotropic hypogonadism [30]; LH beta subunit – associated with hypogonadotropic hypogonadism in males [31] and menstrual cycle and ovulatory disorders in females; [32] FSH beta subunit – associated with hypogonadotropic hypogonadism and azoospermia in males [33] and primary amenorrhea with infertility in females [34]; and the androgen receptor gene – associated with partial androgen insensitivity syndrome (manifesting with hypospadias/hypogonadism/cryptorchidism and infertility, or simply infertility with normal external male genitalia),[35] or, in cases of carriers of extended (≥28 repeats) (CAG)_n tandem repeat in the AR gene – defective spermatogenesis resulting in azoospermia and oligozoospermia in males [36,37] and increased risk of polycystic ovarian syndrome (PCOS) in females.[38] Many of these mutations, however, are quite uncommon and some have been reported only in isolated case studies. At present, information about carriership of many of the mutations listed directly above rarely allows tailoring out of the therapeutic strategy so that it may be better suited to the needs of the particular couple. After all, the diagnosis of azoospermia confirmed by a simple semen analysis gives sufficient information about the chances of conception using the male partner's own sperm even without the complicated and expensive genetic analysis revealing the exact genetic causes for the azoospermia. Similarly, disorders of ovulation and menstrual cycle diagnosable by ‘classic’ techniques such as basal body temperature charting, serial ultrasounds and hormone profiles may provide sufficient information about the chances of conception by natural means or using AR techniques without advanced genetic analysis. In other words, for rare mutations associated with increased risk of infertility, knowledge about one's carrier status may not significantly improve the chances of achieving conception, at least not using the couple's own gametes. Indeed, it may speed up the decision of using donated gametes and embryos instead of resorting to complex, expensive and time- and energy-consuming treatments that provide very little chance of success or continuing to wait for natural conception to occur (which, in the end, may result in other types of issues, including age-related complications of pregnancy and delivery).

Genetic analysis may be very useful in fertility evaluations when the results of the genetic tests provide information that may increase the range of potential interventions and/or may contribute to the selection of an individualized approach that may benefit the particular patient or the particular couple. Data about the role of individual capacity for management of genomic damage in human infertility has been slowly accumulating in the last few years. The genotype–phenotype correlations of the capacity to maintain genomic integrity have only begun to reveal their potential for use in biomedical applications. Several genetic polymorphisms in genes coding for proteins responsible for repair of DNA damage and maintenance of genomic integrity that may modulate the chances of successful conception and carrying pregnancy to term have already been identified (reviewed in [39,40]). These polymorphisms are generally common (some of them, such as the TP53 Pro72Arg polymorphism, may be very common). Information about their carriership may be used as a basis for making informed lifestyle changes that may increase the chances of resolution of the reproductive issue (e.g. decreasing exogenous and/or endogenous genomic damage by avoiding or reducing certain damage-increasing factors).

Oxidative stress, DNA damage and infertility: a man's work

The potential sources of damage to sperm DNA are many and varied. The early stages of sperm development involve multiple mitotic divisions (possibly asymmetric) of primary sperm cells (type A spermatogonia), producing cells that would go on developing into mature spermatozoa and cells renewing the population of type A spermatogonia.[41] Replication of DNA preceding each cell division carries inherent risk of introduction of DNA damage (strand breaks, mismatches) and DNA mutations. The genome-wide recombination accompanying the division of genetic material between haploid gametes during meiosis may also be a source of DNA damage (specifically, strand breaks). Inadequate chromatin packaging during spermatogenesis may also render sperm DNA prone to damage. Normal sperm chromatin is strongly condensed and compactly packed in a volume significantly smaller than in other cell types, ensuring the protection of its integrity.[42] The process of packaging generates high amounts of torsion strain on the molecule of DNA that is normally managed by repeated nicking and re-ligating.[43] Thus, chromatin compactization may also be a source of strand breaks. Developing sperms are capable of repairing their DNA; therefore, damage occurring in the course of spermatogenesis is usually promptly repaired. Once the sperm is mature, however, there are no chances of further pre-fertilization DNA repair. Unlike virtually any type of cell in the multicellular body (including mature oocytes), mature spermatozoa cannot repair damaged DNA, as they do not possess the necessary repair proteins or the organelles that may produce them (e.g. ribosomes, except mitochondrial ribosomes) and the genome is too tightly packed to allow for any DNA–protein interactions.

The natural process of conception involves stringent sperm selection ensured by several barriers (physical as well as chemical) that need to be overcome within the rather narrow timeframe of pre-ovulation and ovulation. Spermatozoa need to be alive and mobile in order to be able to participate in fertilization (except in some of the more complicated AR techniques such as intracytoplasmic sperm injection (ICSI), which use immobile (or immobilized) spermatozoa). Efficient energy metabolism is crucial for sperm survival, motility and capacity for fertilization. Oxidative stress resulting from normal metabolism in the mature sperm is a significant source of DNA damage. The high levels of oxidative phosphorylation needed to maintain sperm motility are unavoidably associated with increased amounts of reactive oxygen species (ROS) that may damage the proteins, lipids and DNA of the sperm. Spermatozoa are, generally, short-lived and disposable cells. Proteins and lipid oxidation damage might not have serious effects on their overall capacity for fertilization, although it may alter membrane fluidity and permeability, thereby decreasing their functional performance. The integrity of the mitochondrial DNA of the sperm is also not critically important (as long as the mitochondria continue to produce enough adenosine triphosphate (ATP)), as the paternal contribution of mitochondrial DNA is negligible (unless the levels of damage exceed the apoptotic threshold). The integrity of the nuclear chromatin of the sperm, however, may be very important for its capacity for oocyte fertilization. Indeed, the tight packaging of the chromatin of the spermatozoon may protect its integrity, but, should damage occur in the mature sperm, it is not repaired, unless the sperm cell manages to fertilize the egg. DNA damage persists in mature sperms and may increase further after ejaculation and during the long journey through the female reproductive tract.[44] In the zygote, damage to sperm DNA may be at least partially repaired, as the capacity to repair DNA damage of the oocyte is intact.[45]

Increased overall levels of oxidative damage may increase the risk of damaging sperm DNA. Oxidative stress has been implicated in male infertility since the early 1990s, when it was demonstrated that that the levels of ROS were elevated in sperm of infertile men compared to sperm of naturally fertile men.[46] The same study reported that the levels of ROS were higher in semen that was used in intrauterine insemination (IUI) that failed than in semen that produced successful pregnancy after IUI. The initial findings that the levels of DNA damage were lower in fertile men than in infertile ones were reconfirmed in later studies.[47,48] The basic types of oxidative damage are oxidation of nitrogenous bases in DNA and strand breaks. Oxidized bases may mispair, increasing the risk of occurrence of nucleotide substitutions and small deletions. The latter does not have significant immediate effects, although high mutation burden may increase the risk of pregnancy loss at post-implantation stages. Strand breaks (especially, double-strand breaks) are the most toxic type of DNA damage. The tolerance of normal cells to double-strand breaks is usually very low. Single- and double-stranded DNA breaks are the main types of genotoxic damage in ejaculated human sperm.[49] High levels of sperm DNA damage were demonstrated to correlate with reduced overall sperm count, increased percentage of dead spermatozoa, reduced sperm motility and increased rates of abnormal morphology.[50,51] DNA fragmentation negatively correlated with rates of fertilization and embryo cleavage rates.[52,53] Relatively recently, testing of sperm DNA integrity has been proposed as a potential component of extended fertility evaluations.[54] Some authors even believe that sperm DNA integrity may be a better biomarker for prediction of the chances of conception using natural or AR techniques than conventional sperm analysis.[55] The chances of conceiving naturally or using IUI were found to be lower for the partners of males with high levels of fragmentation of sperm DNA.[56] It has been proposed that ICSI may be more successful in such cases (as damage to sperm's DNA may be repaired by the oocyte after fertilization). Thus, the couple in which the main fertility issue is high levels of sperm DNA fragmentation may be advised to proceed directly to ICSI, avoiding other time-consuming and expensive treatments that may have little chance of success.[56]

The source of oxidative damage may be environmental (associated with increased levels of genotoxic agents, e.g. specific treatments, tobacco smoking, strenuous physical exercise or relative physical inactivity) or endogenous (brought about by high levels of ROS due to hyperglycaemia, inflammation, atherosclerotic vascular disease, etc.). Specific occupational hazards (e.g. cadmium, lead, iron, welding fumes, radiant heat) have been associated with poorer semen quality.[57–59] Another factor that may also increase the risk of male infertility is decreased capacity for repair of genotoxic damage. Inherited defects in genes coding for proteins involved in DNA damage detection and/or repair may be associated with infertility (provided that the associated phenotype allows survival into adulthood). In individuals affected by Bloom syndrome [caused by deficiency of one of the major helicases of homologous recombination (DNA helicase RecQ protein-like-3, RECQL3)], affected men are azoospermic or severely oligospermic, whereas women have reduced fertility and premature onset of menopause, but may conceive.[60] Carriership of defects of the gene coding for the Rad23B gene (one of the two components of the protein complex that identifies damage in non-coding regions of DNA has been shown to cause sterility in male mice.[61]

Subtle deficiency in the capacity for DNA repair conferred by carriership of ‘benign’ polymorphic alleles of key genes of DNA repair may increase the risk of male infertility. Carriership of the variant alleles of the common Arg399Gln and Arg194Trp polymorphisms in the XRCC1 gene (coding for an accessory factor to the main ligase of base excision repair (BER)) has been implicated in the risk of azoospermia and severe oligospermia.[62,63] These studies also show that the effects conferred by the genetic background may be significantly modulated by environmental factors, as the effects of carriership of XRCC1 variant alleles were only significant in males with high levels of exposure to polycyclic aromatic hydrocarbons (PAH) such as benzopyrene and benzanthracene (abundant in tobacco smoke and industrial smoke). Thus, information about the carrier status for these and other DNA polymorphisms associated with increased risk of infertility may become a basis for making informed lifestyle choices (e.g. ceasing smoking, changing jobs) in order to improve the chance of achieving desired conception.

Another polymorphism that has been implicated in male infertility is the rs1800975 polymorphism (A-to-G substitution) in the 5'-untranslated region of the XPA gene.[64] The XPA protein is part of the dimer complex that binds to damaged DNA in the early stages of nucleotide excision repair (NER) and stimulates the excision of the damaged region. Carriership of the variant allele of the rs1800975 polymorphism in males is associated with over 50% increase in sperm DNA fragmentation.[64] Polymorphisms in other genes coding for proteins of repair by nucleotide excision – namely, excision repair cross-complementation group 1 (ERCC1) – may increase the risk of impaired male gametogenesis. The ERCC1 protein is a component in the endonuclease complex introducing the 5'-strand break in the repaired strand of DNA. Carriership of the A allele of the ERCC1 C8092A polymorphism is associated with almost twofold increased risk of azoospermia.[65]

Male fertility has also been demonstrated to be affected by subtly deficient mismatch repair, probably because of the role that mismatch repair proteins play in meiotic recombination. Examples of mismatch repair proteins implicated in male fertility include the human mismatch repair genes MutL of E. coli homolog 1 (MLH1), MutL of E. coli homolog 3 (MLH3) and postmeiotic segregation increased 2 (PMS2) coding for subunits of a complex with endonuclease activity functioning in mismatch repair and meiotic recombination and the gene MutS of E. coli homolog 5 (MSH5), coding for an ATP-binding protein of meiotic recombination. Specifically, several common polymorphisms in the human MLH1 (rs4647269), MLH3 (C2531T), postmeiotic segregation increased 2 (PMS2) (rs1059060) and MSH5 (C85T) genes have been associated with increased levels of sperm fragmentation and, respectively, increased risk of severe oligozoospermia.[66,67]

Functional polymorphisms in genes coding for key proteins of repair by non-homologous end joining (NHEJ), such as the LIG4 gene (coding for ligase IV, the main ligase of NHEJ) and the recombination-activating gene 1 (RAG1) gene (coding for the endonuclease subunit of the RAG complex that mediates the introduction of DNA breaks during V(D)J recombination), may be associated with increased levels of sperm DNA fragmentation. Male carriers of the T allele of the LIG4 rs1805388 polymorphism were at threefold increased risk of infertility due to high levels of sperm DNA fragmentation.[68] In the same study, homozygous male carriers of the GG variant genotype of the RAG1 rs2227973 polymorphism were at almost 50% increased risk of infertility.[68]

Defects in genes coding for proteins functioning in the assessment of genomic integrity and damage-related signalling may be implicated in meiotic defects and fragmentation of sperm DNA that may result in male infertility. The ataxia-telangiectasia mutated (ATM) protein is, among its other functions, responsible for the induction of cell cycle arrest in the presence of DNA damage (specifically, strand breaks). ATM-deficient cells accumulate significant amounts of DNA damage, increasing the risk of cancerous transformation or premature apoptosis.[69] Infertility (male as well as female) is, beside many other issues, part of the clinical presentation in individuals with ataxia-telangiectasia due to carriership of two defective gene copies of the ATM gene. Heterozygous carriers of defective ATM alleles have been found to be at increased risk of common severe diseases, specifically cancer, but also degenerative disease.[70] It is still unclear whether carriership of a single mutant ATM allele may affect fertility. Nevertheless, heterozygous carriership of ATM mutations is common (about 2%) and may be very common (over 12%) in some populations [71]; therefore, the matter may need further investigation. Recently, it was shown that carriership of the variant allele of the rs189037 polymorphism (G>A) in the promoter of the ATM gene is associated with twofold increased risk of non-obstructive azoospermia, probably because of defective damage-related signalling in the presence of DNA breaks normally occurring in male meiosis.[72,73]

Other pathways that play a role in mammalian fertility are some p53-related signalling pathways. Polymorphisms in the TP53 gene and the gene coding for one of its key regulators – the ubiquitin ligase mouse double minute 2 homolog (MDM2) – may also be genetic factors in male infertility. The rs2287498 (Ex2+19C>T) and rs1042522 (Pro72Arg) polymorphisms in TP53 and the rs937283 polymorphism in MDM2 have been shown to be associated with male infertility.[74–76] The rs1042522 (Pro72Arg) polymorphism is unusual in more than one aspect. Not only is it very common (the prevalence of the 72Arg variant allele may reach 80% in some populations), but its two alleles, while being both considered wildtype, may strongly modulate the risk of development of late-onset multifactorial disease, the course of the disease, the potential complications and the outcomes after different treatments (specifically, genotoxic treatments).[39,77] The 72Arg allele confers increased propensity to apoptosis in cells that have sustained genomic damage, whereas the Pro allele may stimulate cell cycle arrest and repair of damage.[78] Genotypes containing at least one 72Arg allele have been repeatedly shown to increase the risk of azoospermia and oligospermia, probably by increasing the rates of apoptosis in developing spermatocytes in a p53-dependent manner.[75,76] Notably, the TP53 Pro72Arg polymorphism plays a role in female infertility as well, although the effects of carriership of the one or the other allele on fertility may be exactly opposite in males and females (see below).

Folate metabolism polymorphisms may have an additional value in fertility evaluations, apart from the increased risk of pregnancy loss due to coagulation issues. A number of studies show that carriership of the variant allele of the very common MTHFR C677T polymorphism may be a risk factor for azoospermia and severe oligospermia in Asian males.[79,80] Similar results have been obtained for the role of MTHFR C677T and the MTR A2756G polymorphisms in the Brazilian population.[81] The association was not confirmed for European males in population samples of approximately the same size.[82] It is possible that carriership of the T allele of MTHFR C677T may be a population-specific risk factor for male infertility. As a matter of fact, most of the presently available studies about the role of polymorphisms in genes coding for proteins of DNA repair and maintenance of genomic integrity are conducted in patients of Asian ethnic origin (most commonly, Chinese and South Korean). There is pressing need for more studies of the matter in European populations, as some of the polymorphisms in DNA repair genes may be quite common and their carriership may be associated with increased risk of development of various multifactorial diseases (diabetes, vascular disease, cancer) that may directly or indirectly affect fertility.

Individual repair capacity in the oocyte: a servant of two masters

Unlike the spermatozoon, which only possesses the basics to ensure its survival, the oocyte is very well equipped to accommodate for the first divisions of the zygote. It also contains a fully functional DNA repair system that, once the oocyte is fertilized, may take care of persisting unrepaired damage in both the maternal and the paternal genome. Thus, the DNA repair machinery of the oocyte is responsible for the integrity of both parental genomes after fertilization. Fragmentation in the sperm genome may be repaired, provided that the percentage of damaged DNA is not very high. The cut-off value of sperm damage beyond which the zygote would not survive may be different in different species. This was first demonstrated in rodent models, using gamma-irradiated hamster spermatozoa for IVF, allowing the resulting zygotes to develop to the blastocyst stage, and then transferring the surviving blastocysts into surrogate mothers.[83] The rates of successful fertilization and development of the blastocysts into viable foetuses predictably decreased with increasing doses of gamma irradiation (and, respectively, with increasing levels of damage). The maximum level of sperm DNA damage that resulted in successful implantation and normal subsequent development in hamster zygotes was 8%. Damage beyond this level resulted in early pregnancy loss. Rodents have a unique DNA repair profile among all mammals. Specifically, they predominantly repair transcribed DNA at the expense of untranscribed DNA (‘rodent repairadox’), whereas all other mammals, including humans, tend to repair damage in all genomic regions.[84] Selectivity of DNA repair allows rodent cells to survive significantly larger doses of genomic damage than cells from almost any other mammalian species, including primates and humans. Thus, it could be expected that the levels of sperm DNA damage tolerated by human oocytes would be even lower than 8%. Nevertheless, a later study in human patients showed that successful pregnancy after IUI was possible with sperm in which the percentage of DNA fragmentation was 12% or less.[85] It is possible that the numerical values of the results in both studies were not that accurate, so that the cut-off percentage beyond which successful pregnancy was not possible was actually about 10% for both rodents and humans. It is also possible that the selective pattern of DNA repair in rodents briefly described above may increase the risk of ‘missing’, or rather, ‘ignoring’ genomic damage, as long as it is not in the transcribed regions of the genome. Thus, after the DNA repair of both parental genomes is complete, the rodent zygote may still contain significant amounts of ‘hidden’ damage in its untranscribed regions that may result in non-implantation and/or pregnancy loss. Human fertilized oocytes would try to repair all genomic damage in the parental genomes within the constricted timeframe before DNA replication prior to the first division of the zygote. If the repair fails, the zygote is lost, but if it succeeds, the levels of residual unrepaired damage would be very low, increasing the chances of successful pregnancy. Thus, after fertilization of rodent and human oocytes with spermatozoa with the same level of DNA damage (at the upper limit of tolerable damage for human spermatozoa, 12%), one may expect that the levels of unrepaired damage would be enough to kill the rodent zygote, but the human zygote may survive.

The main sources of genomic damage in different phases of the spermatozoon life cycle are discussed above. The main sources of endogenous DNA damage in female germ cells are DNA replication preceding the mitotic division at the early stages of development and, subsequently, the first meiotic division; and the genome-wide homologous recombination occurring in pachytene of prophase I of meiosis (all occurring during intrauterine life). Multiple mitotic divisions (a potential source of replication-related damage) occur at very early stages of development of female reproductive cells (between the 7th and 20th week of intrauterine life).[86] The onset of the first meiotic division is around intrauterine week 8–13 and continues until week 24–26. Oogonia that have not proceeded to meiosis are typically routed to programmed cell death [87,88], reducing the number of potential primordial follicles. The rates of oogonial atresia peak in weeks 17–20, but may continue up to week 26 of intrauterine life. Folliculogenesis begins after week 20 and is followed by a second wave of programmed cell death in weeks 21–24 (follicular atresia) that reduces the number of potential primary oocytes.[87] Thus, the first two rounds of selection of female germ cells occur relatively early in individual development, eliminating cells that have sustained damage during mitotic or meiotic divisions. In the late weeks of the third trimester, the number of primordial follicles is already established, oogenesis is arrested at the stage of diplotene of prophase I (primary oocyte) and would not resume until the onset of puberty. Mouse models show that the levels of DNA repair in the oogonia are low at the pre-diplotene stage; then increase and remain stable throughout the enforced oocyte quiescence between birth and the onset of sexual development.[89] Despite the very low metabolic rate of quiescent follicles and the absence of DNA replication, in the considerable interval of months (or, in the case of humans, years and decades) of quiescence there is risk of DNA damage, albeit at a low rate and occurring at random sites. This damage is normally promptly repaired. It is widely known that advanced maternal age is associated with increased risk of genomic abnormalities, but maternal age does not reflect significantly on all types of genomic damage. The levels of unrepaired bulky adducts in the DNA of murine oocytes remain more or less the same throughout all the reproductive life of the female, [90] probably due to preserved capacity for repair in the mature oocyte. The main type of damage in aging female germinative tissue is chromosomal nondisjunction. It is believed that the arrest of meiotic progression lasting for several decades is the main cause for nondisjunction in aged mammalian oocytes.[91] It was already mentioned that pre-diplotene oocytes had low levels of DNA repair, which is probably a physiological mechanism to ensure that double-strand breaks occurring during normal recombination would not unduly activate the repair machinery. This comes at the expense of the risk of persistence of strand breaks after recombination is complete. Presence of reactive DNA ends generated by strand breakage may increase the risk of chromosome fusion and breakage, eventually resulting in nondisjunction and loss of chromosomal fragments. The closer is the repair capacity of the developing oocyte to normal levels, the lower is the risk that pre-diplotene suppression of DNA repair would result in persistence of unrepaired double-strand breaks and, respectively, the lower is the risk of future chromosomal abnormalities, even in advanced age. Normally functioning DNA repair machinery in the oocyte would also manage random damage occurring during follicular quiescence and in the course of oocyte maturation, and is also capable of taking care of genomic damage in the male genome after fertilization. Thus, the capacity for genomic repair in the oocyte may be an important factor in fertility, although its role may vary in different stages of oocyte development, maturation and, potentially, fertilization. Optimal repair capacity in the immature oocyte during the early stages of development may decrease the risk of later occurrence of age-dependent chromosomal defects in the oocyte. In the mature oocyte, preserved DNA repair capacity may provide the dual benefits of maintenance of integrity of the oocyte genome prior to fertilization and of the male haploid genome that has managed to penetrate the oocyte after fertilization. Thus, unlike the spermatozoon, where the capacity for genomic repair is important in the early developmental stages and has little significance once the sperm is mature, the repair capacity of the oocyte is crucial both before and after fertilization and for male and female haploid genomes. Unfortunately, the importance of DNA repair in female fertility is still significantly less well studied than in male fertility.

Role of individual repair capacity in female fertility

The above-mentioned TP53 Pro72Arg polymorphism plays a role in male as well as female fertility, although the effects of carriership of the one or the other allele may be different in men and women. Carriership of variant alleles of the TP53 gene is commonly seen in women with endometriosis, a condition that may increase the risk of infertility. The allele of the 16 bp duplication polymorphism in intron 3 of the TP53 gene is associated with increased risk of endometriosis in female carriers.[92] The16 bp duplication allele is associated with lower levels of p53 mRNA and, respectively, with less efficient induction of cell cycle arrest in the presence of damage, resulting in subtly relaxed control over cell proliferation, [93] increasing the risk of ectopic cell growth. The ‘pro-repair’ 72Pro allele of TP53 is overrepresented in women with endometriosis.[92,94,95] The balance between cell proliferation and cell death in endometrial tissue in endometriosis is pathologically altered in favour of cell proliferation, as signified by the study of Johnson et al. [96] showing that the expression of pro-apoptotic proteins such as BAX is suppressed and the expression of proteins stimulating cell division such as c-MYC is upregulated in eutopic endometrial tissue from endometriosis.[97] It is possible that the pro-apoptotic propensity conferred by the 72Arg allele suppresses invasive endometrial growth, thereby decreasing the risk of endometriosis.

Increased frequency of the 72Pro allele and marked prevalence of the homozygous 72Pro/Pro genotype have been observed among women with implantation failure after IVF and embryo transfer.[23,92,97] Defective signalling via leukaemia inhibitory factor (LIF) may play a key role in p53-mediated implantation failure. LIF is a p53-regulated secreted glycoprotein with an important role in implantation of mammalian embryos.[98,99] It is expressed in the human endometrium in the secretory phase, peaking around the prospective time of implantation. Carriership of variant alleles of the human LIF gene (e.g. the rs929271 polymorphism) is significantly more common in young women with idiopathic infertility than in control age-matched women.[100] Carriership of variant alleles of the human TP53 gene has been shown to directly influence the chances of successful embryo implantation in transgenic mice expressing either the Pro or the Arg variant of human TP53 by modulating the levels of LIF.[101,102] Specifically, mice expressing the 72Arg allele of human TP53 had higher levels of uterine LIF during the implantation window and exhibited higher rates of blastocyst implantation than mice expressing the 72Pro allele.

It is likely that p53-regulated pathways play important roles beyond the post-implantation stage as well. So far, this has been unequivocally confirmed in mammalian models. Isolated uterine p53 deficiency in female mice resulted in accelerated decidual senescence and significant increase in the incidence of preterm birth, despite normal ovulation, fertilization and implantation.[103] It is possible that uterine deficiency of p53 (e.g. because of lower expression of TP53 conferred by carriership of variant alleles) may affect the successful carrying of pregnancy to term in humans as well. Of course, not all results obtained in rat and mouse models pertain directly to humans, as there may be significant differences in some aspects of the biology of the two species. This may be especially valid for studies of the capacity for DNA repair and maintenance of genomic integrity (e.g. the above-mentioned ‘rodent repairadox’). There are also some peculiarities in the responses to DNA damage in mouse and human embryonic cells associated with differences in the stringency of the early checkpoints for DNA damage.[104] It has been established, nevertheless, that carriership of the 72Pro allele has an adverse effect on the capacity not only to conceive but also to sustain pregnancy in humans in an age-dependent manner. The reproductive disadvantages conferred by the carriership of the Pro72 allele were demonstrated to be significant only in women that were older than 30 years of age at the time of birth of their first child.[105] In women pregnant for the first time between 20 and 30, carriership of the Pro72 allele/s did not seem to have noticeable effects on the rates of reproductive failure. It is possible that this effect only becomes significant after 30 because it adds up to other age-dependent potential factors that may decrease the chances of establishing and carrying pregnancy to term. Older women are more likely to have health issues interfering with the capacity to conceive and sustain pregnancy and the ovarian reserve generally declines after the age of 32–35. Thus, it is possible that carriership of 72Pro augments any other factors that contribute to reproductive failure.

While the pro-apoptotic tendency conferred by the 72Arg allele may explain the increased risk of azoospermia and oligospermia in male carriers, it is truly intriguing why the 72Pro allele (believed to be the ancestral allele) may have an adverse effect on female fertility and, specifically, on the rates of early pregnancy loss. It is known that the 72Pro allele confers, among other properties, increased propensity for transcriptional activation of the genes regulated by p53.[78] Among these are genes coding for proteins of damage-associated signalling, proteins functioning in the enforcement of cell cycle arrest in the presence of damage and key proteins of DNA repair. Normally, the DNA of a dividing cell is subjected to multiple checks before and after it is replicated in order to avoid copying of damaged templates. In the presence of damage, the progression through the cell cycle is halted and repair activities are launched. Generally, the checkpoint in the transition phase between G1 and S phases has priority over all other checkpoints in the cell cycle (hence the name restriction point, or R). DNA repair and subsequent checks whether damage has been repaired may take time, as the level of unrepaired damage must be minimal (ideally zero) before the cell is allowed to re-enter the cell cycle. Very early embryos, however, are forced to operate on a very tight schedule that includes multiple divisions occurring within pre-set time limits and delays may be fatal. This may be one of the causes of the relatively relaxed stringency of R in very early human embryos and its complete lack in cells of early rodent embryos.[106] Thus, rodent embryonic cells with damaged DNA may continue with their development, with some cells dying because of accumulation of damage beyond the apoptotic threshold (and, potentially, being replaced by division of the remaining cells), and some being directed to differentiation, where R operates normally.[107] Thus, at least some embryos from the same pregnancy may have a chance of developing, albeit at the risk of premature delivery or and/or birth defects. In early human embryos, however, R operates, although not at full efficiency. Thus, cells that have sustained genomic damage may enforce cell cycle arrest and attempt to repair the damage. If the repair activities are not completed fairly quickly, however, the development of the embryo may be delayed beyond the timeframe of early development, which may prevent normal implantation (resulting in early pregnancy loss) or significantly delay it. The latter is also likely to result in early embryo loss, as the endometrium is receptive only within a narrow window of opportunity for implantation (in humans, between day 6 and 11 post-ovulation). Thus, pro-repair tendencies conferred by the 72Pro allele are likely to interfere seriously with embryo development and cause ‘missing’ of the implantation window and/or intrauterine growth restriction associated with pregnancy loss. At the same time, the pro-apoptotic tendency conferred by the 72Arg allele may result in death of embryonic cells that have sustained too much damage, but the very early embryo at pluripotent stage (up to 16–32 blastomeres) may make up for missing blastomere/s by cell division.

IVF procedures often use frozen gametes and embryos pre-treated with substances with significant genotoxic potential, such as dimethyl sulfoxide (DMSO). This may have additional adverse effects on the survival of embryos, as high-grade damage may not be repaired in time for successful implantation. Indeed, IVF using frozen embryos or embryos produced by fertilization of frozen gametes may exhibit lower rates of success.[108] Carriership of the 72Pro allele may contribute to the risk of pregnancy loss in IVF using frozen gametes or embryos because of increased propensity for imposing cell cycle arrest in the presence of damage.

Carriership of a specific genetic polymorphism may have only subtle effects on female fertility, but the effect of co-carriership of variant alleles in several loci may result in a cumulative effect. Homozygocity by the variant allele of the single-nucleotide polymorphism rs2279744 in the MDM2 gene coding for the key regulator of p53 stability has been shown to be associated with increased risk of recurrent pregnancy loss.[109] Homozygous carriership of the variant allele of the rs2279744 polymorphism in the human TP63 gene (coding for a protein closely related to p53) has been shown to act as an additional risk factor for recurrent miscarriage when co-inherited in the same genotype with the variant allele of the MDM2 rs2279744 polymorphism.[110]

Recently, it has been shown that carriership of mutations in the major tumour suppressor gene BRCA1 (but not BRCA2) are associated with accelerated ovarian aging and, respectively, with decreased ovarian reserve.[111,112] BRCA1 and BRCA2 are subunits in a large enzyme complex that is rapidly recruited to sites of DNA damage. The majority of cases of familial breast cancer are associated with carriership of mutations in either the one or the other gene, and ovarian cancer may develop in a significant proportion of carriers. It is currently believed that fertility treatments do not significantly modify the risk for carriers of BRCA1 mutations and are therefore not contraindicated.[113] As the carriership of BRCA1 mutations is common and may be very common (up to 2%) in some populations, screening for mutations in the BRCA1 gene as part of the fertility evaluation may have dual benefits for prospective carriers: on the one hand, it may serve as an early warning sign for a potential rapid decline in ovarian reserve and, on the other hand, it offers opportunities for early intervention in order to prevent development of cancer after the reproductive plans have been completed.

Increased levels of oxidative damage play a significant role in male infertility but may have some significance in female infertility as well. Mitochondrial DNA exhibits very little natural variance when compared to nuclear DNA, mainly because it is significantly more gene-dense (hence the risk that even a small change may be deleterious) and has very low levels of recombination. There are, nevertheless, relatively benign alterations in the sequence of human mitochondrial DNA that tend to segregate in the same DNA molecule (mitochondrial haplogroups). Different mitochondrial haplogroups are associated with different ATP output and different levels of production of ROS.[114] Generally, the higher the output of ATP, the higher the levels of ROS and, respectively, the higher the risk of oxidative damage to DNA. Two phylogenetically related mitochondrial haplogroups (J and T, forming the macro-haplogroup JT) are associated with lower amounts of synthesized ATP but also with lower levels of ROS than the most common haplogroup H (which is also the haplogroup with the highest ATP and ROS output).[115] Recently, it was shown that the JT macrohaplogroup was significantly under-represented in Caucasian females with low ovarian reserve.[116] The authors of [116] estimated that the risk of premature ovarian failure may be at least three times lower in JT carriers than in carriers of any other mitochondrial haplogroup. Considering that the ‘protective’ macrohaplogroup JT accounts for only 15–20% of all Eurasian haplogroups, [117] the risk associated with carriership of mitochondrial haplogroups other than J and T may need to be taken into account when evaluating the ovarian reserve in women of European or Asian origin.

Aspects of individual repair capacity and maintenance of genomic integrity that may affect both male and female fertility

Obesity has been implicated as a significant factor in both male and female infertility.[118,119] Obesity affects female reproductive function on virtually all levels: it disturbs the hypothalamic-pituitary-ovarian axis, causes hormone dysregulation, affects oocyte quality and endometrial receptivity and increases the risk of development of various diseases and conditions that may affect adversely the chances of pregnancy and successful pregnancy outcomes. Male obesity is associated with impairment in the regulation of reproductive hormones, increased scrotal temperatures (affecting spermatogenesis) and may be associated with abnormal levels of insulin and glucose in the seminal fluid (decreasing sperm motility). Obesity is often part of the phenotype of metabolic syndrome or diabetes type 2. Polymorphisms in the human gene NEIL1, which codes for one of the major glycosylases of BER, were identified in several per cent of human patients with diabetes type 2 [120] and the potential role of allelic variants of NEIL3 in the pathogenesis of insulin resistance was shown an year later.[121] Four allelic variants of the human gene HMGA1 (coding for a major regulator of gene expression and DNA repair) were identified in individuals with severe insulin resistance.[122] HMGA1 polymorphisms have been implicated in age-dependent increase in oxidative damage by the dual mechanism of insulin resistance/hyperglycemia and suppression of DNA repair.[123–125] BER is one of the two main mechanisms for repair of oxidative damage. It is possible that carriership of alleles of genes coding for variant proteins of BER and/or regulators of gene expression such as the high-mobility group (HMG) of nonhistone proteins may affect human fertility by interference in the glucose metabolism, increasing the risk of metabolic syndrome and diabetes type 2. The later, in turn, increases oxidative stress, which contributes to the deleterious effects of hyperglycaemia on fertility by increasing the risk of azoospermia and ovarian failure, probably via the mechanisms described above.

Individual capacity for repair of DNA damage and maintenance of genomic integrity is not a simple sum of the disparate effects of different genetic polymorphisms. The phenotype is a complex product of genetic traits and environmental influences and may dynamically change under physiological conditions (e.g. as the organism ages) or in various pathological states. The overall capacity to maintain the integrity of the genome may vary in the same individual (that is, in the same genetic context) depending on the momentous state of the organism. The latter may change with age, in the course of treatment/s for infertility and in different phases of pregnancy. Thus, it may be quite important to be able to assess the individual repair capacity at baseline and monitor potential changes in time (e.g. while waiting to conceive), in the course of specific fertility treatments and in the course of pregnancy (if and when finally achieved). Conventional genetic analysis may not be capable of assessing the dynamics of the capacity to repair genomic damage. There are several currently available methods that allow monitoring of individual repair capacity at a phenotypic level.[126–129] Indeed, these methods usually require living cells from the patient (skin fibroblasts, peripheral leukocytes) taken at fixed intervals of time. Considering the invasive nature of many fertility treatments, introduction of additional manipulations may be undesirable. A combined approach of genotypic and phenotypic analysis, however, may yield better results in fertility evaluations than routine genotype–phenotype correlation analysis and expression studies. Several recent reports demonstrate the opportunities offered by high-throughput methodologies for analysis of nucleic acids, such as next-generation sequencing (NGS) and microarray-based comparative genomic hybridization (array CGH, aCGH) in research related to infertility and pre-implantation genetic diagnosis.[130–133] The number of studies in the field is still limited, but it could be expected that these methods may be implemented in the clinical practice in the near future as they allow for very high resolution (single-nucleotide level) on a genome-wide scale.

The research in the area of genotype–phenotype correlations between individual repair capacity/capacity for maintenance of genomic integrity and human fertility is still limited. Future research in the field may facilitate monitoring of the effects of targeted lifestyle changes (giving up harmful habits, modification of the pattern of physical activity, eating a healthier diet, changing jobs) and the course of therapies decreasing oxidative stress (antidiabetic treatments, antioxidant therapies, etc.) and may serve as an early warning system for any changes in the individual repair capacity status that may increase the genotoxic pressure, in order to intervene and/or change the therapeutic modalities.

Conclusions

Many factors may affect the chances of successful pregnancy, especially in later age. Inter-individual variance of the capacity to detect and repair genomic damage and the potential age-dependent effects of carriership of specific alleles and haplotypes may have significant effect on male and/or female fertility. Current data show that common polymorphisms in genes coding for proteins responsible for DNA repair and/or maintenance of genomic integrity may modulate the chances of natural and assisted conception, successful implantation and carrying pregnancy to term, especially in older couples. Further research about the role of individual repair capacity is needed in order to increase the reliability of genetic markers for assessment of risk of reproductive failure, to clarify the role of biologic age of both prospective parents and to devise strategies improving the chances of reproductive success by natural means or using AR technologies.

Disclosure statement

No potential conflict of interest was reported by the authors.