A systematic review of the validated monogenic causes of human male infertility: 2020 update and a discussion of emerging gene-disease relationships

Search strategy and study selection

A literature search was performed as described in Oud et al. (2019) to identify articles reporting on monogenic causes of male infertility or male reproductive system anomalies entered into MEDLINE-Pubmed before the 1^st of July 2020. Assessment of whether the articles met the inclusion or exclusion criteria (detailed in Supplementary Table I) was performed by two independent reviewers (BJH and MSO). The present study and the corresponding search protocol were registered with the PROSPERO registry (http://www.crd.york.ac.uk/PROSPERO) as PROSPERO 2021: CRD42021229164.

Data extraction and assessment

The clinical validity of each identified GDR was scored using a system published by Smith et al. (2017). Scoring was performed by two randomly assigned reviewers (BJH, AR-E, MJW, AS-H, MJX, LN, CF and MSO) using a standardised assessment template to: (i) extract gene names, inheritance patterns, patient phenotypes, method of discovery (sequencing method), (ii) annotate variants, and (iii) assess both functional evidence and clinical data, including the outcome of assisted reproductive technologies and recorded co-morbidities. Expression of genes across human organs was assessed by consulting data available on the Human Protein Atlas website, NCBI’s RNA-seq dataset (Fagerberg et al. 2014), GTEx, and an unpublished human testis single cell RNA-seq library tool (‘https://conradlab.shinyapps.io/HISTA; Mahyari et al. under review). To avoid bias in gene-disease evaluation and any conflicts of interest, reviewers were not allowed to score any GDRs they had published. Scoring was separated into five categories: no evidence (<3 points), limited (3-8 points), moderate (9-12 points), strong (13-15 points) or definitive (>15 points), with of a maximum of 17 points. After independent scoring, the individual scores (for each GDR) for both reviewers were compared and any inconsistencies in scoring (>1 point difference or a difference in the final classification) were settled by the two assigned reviewers. Where this was not possible, the scoring was discussed with all non-conflicted reviewers. GDRs with a moderate or higher classification were deemed as confidently linked to human male infertility and combined scoring sheets are available at http://www.imigc.org/. All results in the study are collated in Supplementary Table A PRISMA checklist is available for this study as shown in Supplementary Table III.

Summary of included studies and design

We performed a literature search using terms related to “male infertility” in combination with keywords related to the word “genetics” in MEDLINE-PubMed and used the same inclusion and exclusion criteria as described previously (Oud et al. 2019). A total of 26,250 articles were identified, of which 2,765 new articles were identified since the previous search in 2018 and not covered by the previous review (Figure 1). The total number of publications on genetic causes of male infertility has increased by 13% from 849 over the period 2010-2014 to 963 articles in the period 2015-2019 (Figure 2A). Further, while the final search performed for this review was completed in mid-2020, the total number of 2020 studies was estimated to overtake the 2019 number by 20%. The absolute and relative contribution of publications on proposed monogenic causes of phenotypes related to male infertility is also growing (46% in 2017 vs. 52% in 2020). The relative contribution of publications on association studies has declined by 42% (27% in 2017 vs. 19% in 2020), Figure 2B). The shift away from Sanger sequencing to NGS methods is continuing to take place, and 71% of all sequencing studies used NGS in 2020 (Figure 2C), which constitutes an increase of 90% since 2017. While WGS has started to emerge in the field (recent examples: Bedoni et al. 2016; Wang et al. 2017; Yuan et al. 2020; Arafat et al. 2020), WES is currently still the predominant method over panel sequencing and whole genome sequencing (75% vs. 19% and 6%, respectively). Nevertheless, there are clinical presentations where Sanger/panel sequencing is still being utilised at a high frequency, e.g. 63% of studies on 46,XX and 46,XY DSDs in the period of 2019-2020.

• Download figure
• Open in new tab

Figure 1. PRISMA Flowchart of search and assessment process.

• Download figure
• Open in new tab

Figure 2. Breakdown of genetic testing approach and number of genes associated with male infertility phenotypes with limited or moderate and higher evidence classification, over time.

A. The fraction of studies assessed in this paper that used particular sequencing technologies to elucidate causes of male infertility from 1958-2020, including karyotyping, copy number variation assessment, azoospermia factor region deletion assessment, association studies and studies investigating monogenic causes. B. The total number of studies assessed in this paper investigating male infertility, from 1958-2020, based on sequencing approach C. The fraction of studies assessed in this paper using next generating sequencing technology compared to Sanger sequencing, from 2010-2020. D. The cumulative number of genes and their strength of evidence as linked to male infertility phenotypes based the scoring criteria used in this paper, from 1989-2020.

Systematic evaluation of evidence

A total of 1,523 publications met all inclusion criteria and were used in this systematic review. Of these, 186 publications were published in the period from 2019 to mid-2020 and were thus not included in the previous clinical validity assessment. Overall, 657 individual GDRs described in these 1,523 publications were investigated, of which 136 were novel and 521 were re-evaluated with our updated assessment criteria in order to incorporate any additional supporting evidence (Supplementary Table II).

The quality and the extent of all evidence for the GDRs was assessed by two independent reviewers, using a standardised scoring method. This assessment was performed for GDRs that were new (n=136), existing with newly published evidence (n=145), or previously confidently linked (moderate, strong or definitive) to male infertility with no newly published evidence (n=19). This score was used to assign a clinical validity of each GDR as: no evidence, limited, moderate, strong, or definitive. We assessed the experimental quality, patient phenotype information, functional evidence based on gene expression and the study of in vitro and in vivo loss-of-function animal/cell models. Variants were also reclassified according to the widely accepted ACMG-AMP standards and then recorded in spreadsheets for each GDR.

Of the newly identified GDRs, 17 were classified as having moderate evidence or higher (confidently associated with the phenotype) and 43 were classified as having limited evidence. After classification of the existing GDRs that had new evidence published and scoring of GDRs that were previously confidently linked to male infertility with our updated scoring criteria, 45 increased in score since 2019, 11 decreased in score, and 9 changed from unable to be classified to no evidence (or vice versa). As detailed above, GDRs were listed as ‘unable to classify’ when the predicted inheritance pattern or genotypes were not reported. In comparison to the previous assessment, 456 scores did not change because no (or insufficient) new evidence was published during the search period (Supplementary Table II).

The novel and established GDRs combined resulted in a total of 104 genes confidently linked to 120 human male infertility or abnormal genitourinary development phenotypes with moderate (n=45), strong (n=27) or definitive (n=48) evidence (Tables I and II). An overview of the organ or cell-level role(s) of these genes is detailed in Figure 3. Another 138 GDRs were classified as ‘limited’ and are thus candidate genes whose dysfunction may result in a male infertility disorder (Supplementary Table II). We propose that latter (limited evidence) group will be of particular interest for review in the next few years.

• Download figure
• Open in new tab

Figure 3. Overview of all genes associated with male infertility phenotypes at an organ/cell level.

Definitive genes are labelled in red, strong in orange and moderate in yellow. Organs top to bottom: brain, adrenal gland (and kidney), testes and epididymides with vas deferens. Bottom left: sperm fertilising an oocyte surrounded by cumulus cells. Right: seminiferous tubule cross-section. Leydig cells (blue), Sertoli cells (purple), basement membrane (pink), spermatogonia (green), spermatocytes (brown), spermatids (pink and orange). ** denotes germ cell arrest gene for FANCM and TEX14. Genes were classified as unclear when they were not clearly linked to a specific organ.

Insights into the genetic causes of male infertility

As outlined above, in recent years, the identification of genes confidently linked to a male infertility phenotype has risen by 150% from a stable average of approximately 4 per year in the period of 2000-2016 to 10 per year in the period of 2017-2019 (Figure 2C). Additionally, the number of GDRs with limited (emerging) evidence is growing quickly, at approximately 17 new genes per year in the period 2017-2019 (Figure 2C).

The majority of confident GDRs were endocrine disorders or reproductive system syndromes (n=67) and isolated infertility phenotypes (n=36), while a minority were linked to syndromic infertility (n=17), including primary ciliary dyskinesia (Table I). The systematic analysis of patients who present with multiple morphological abnormalities of the sperm flagella (MMAF) phenotypes was also of particular note. In 2018 there were 5 genes confidently linked to human MMAF. This has now tripled to a total of 15 genes as of mid-2020. Although representing only a small fraction of patients with isolated infertility, the use of NGS has the potential to diagnose up to ∼50% of MMAF patients (Toure et al. 2021). A further 15 genes were confidently linked to the most frequently presenting clinical presentations, NOA and oligozoospermia (Tables I and II). The majority of all confident GDRs represented an autosomal recessive inheritance pattern (n=68), while autosomal dominant (n=38), X-linked (n=12) and Y-linked (n=2) inheritance patterns were also reported.

Recent developments in genetics of male infertility research and diagnostics

The diagnostic rate of genetic tests for all types of isolated male infertility combined currently sits between 4 and 9.2% (Tüttelmann et al. 2018; Punab et al. 2017; Olesen et al. 2017). These rates are notably behind levels seen in other heterogeneous disorders with a large genetic contribution such as developmental delay (∼30%) and cardiomyopathies (30-40%) (Rehm, 2017). The biggest difference between these diseases is the slow uptake of NGS approaches in the male infertility field in research, but even more so in diagnostics. WES and WGS are now routinely being applied in the diagnostic follow-up of patients with other genetic disorders, resulting in the availability of very large cohorts for disease gene discovery (e.g. Kaplanis et al. 2020, Bourinaris et al. 2020, Cuvertino et al. 2020). This, unfortunately, is still not the case for severe forms of male infertility in most countries. However, as highlighted in Figures 2C/D, the field of male infertility genetics is now expanding rapidly, largely due to the reduced costs and increasing accessibility of NGS, which allows for a more complete and economic testing of patients. Consequently, the number of novel candidate genes and validated disease genes is rapidly growing, which is essential for further diagnostic implementation of these approaches in male infertility.

Identifying the variant(s) causative for male infertility

Currently, approximately 33% of all identified GDRs proposed in the literature fell into the “unable to classify” category. For most, this classification is given because the inheritance pattern is unclear. The list, however, likely contains multiple genes that play an important role in the aetiology of monogenic male infertility. As with any disease, but perhaps exacerbated with the wide range of individual biological processes required for male fertility, it can be difficult to determine the role of poorly characterised genes, and whether loss of their function results in a disease state. To assist in filtering for high confidence variants, members of the GEMINI consortium have developed the ‘Population Sampling Probability’ pipeline (PSAP; Wilfert et al. 2016). This is particularly useful when it is the first time a damaging variant has been found in a novel gene, where calculations are based on inheritance model (autosomal dominant, recessive, etc.) and allele frequency. As has been highlighted by the GEMINI consortium (Nagirnaja et al. submitted), this tool is useful for discovering promising new candidate disease genes and individual pathogenic variants, but independent replication and functional validation is still critically important before genes can be used in a diagnostic setting as a ‘validated disease gene’.

Dynamics of genotype-phenotype mapping

Recent genetic male infertility studies suggest that a high percentage of diagnoses can be revealed, especially in cohorts of men with maturation arrest. A clear determination of the point of arrest (i.e., at spermatogonia or spermatocytes) can assist in refining possibly relevant genes containing variants, based on any known biological function(s). However, very little so far has been done outside of meiotic and spermatogonial arrest. Various genes have also been linked to multiple phenotypic outcomes; thus, the establishment of a clear genotype-phenotype correlation is not necessarily straightforward. One such example is the TEX14 gene, where pathogenic variation can lead to a Sertoli cell-only phenotype or spermatocyte arrest (Fakhro et al. 2018). The progressive loss of germ cells may occur as a factor of age due to a role in spermatogonial stem cell renewal, as suggested for Nxf2 knockout mice (Pan et al. 2009). Alternatively, environmental exposures may interact with the genotype to lead to an exacerbation of the phenotype. Variability in the genotype-phenotype relationship has also been identified for single gene variants in reproductive system syndromes/endocrine disorders (Domenice et al. 2016; Goncalves et al. 2017) in 12 of the 13 (92%) confident genes – ANOS1, AR, CHD7, CYP19A1, CYP21A2, FGF8, FGFR1, KISS1R, LHCGR, NR0B1, NR5A1 and WDR11. These disorders are largely complex in that they effect multiple hormonal pathways and tissues beyond the testis.

Other presentations with variability in genotype and phenotype include congenital hypogonadotropic hypogonadism/Kallmann syndrome. In addition to Mendelian causes, incomplete penetrance and oligogenic forms are thought to be causative for these disorders (Maione et al. 2018). One example is the PROK2 gene, for which autosomal dominant, recessive and oligogenicity have been proposed to be causative (Leroy et al. 2008). For the disorders/differences of sexual development (DSD) spectrum and hypogonadotropic hypogonadism (HH), research groups are now actively exploring the role of oligogenic inheritance in contributing to the disease (Cangiano et al. 2021). For DSDs, this genetic explanation has been proposed following the finding that many patients present with at least 1 variant in 2 individual known or novel DSD genes, particularly when MAMLD1 is affected (Camats et al. 2020, Li et al. 2020). Similarly, variants in FGFR1 and other HH genes have shown to cause Kallman syndrome in an oligogenic manner (reviewed in Cangiano et al. 2021).

Recommendations for genetic testing in male infertility

Although NGS methods are frequently used in research laboratories to study the genetics of male infertility, they have not yet been extensively employed in clinical genetic diagnostics for this condition. While there are no standardised international guidelines for clinical genetic testing in male infertility, a small number of guidelines have been published by professional andrology groups. Examples include guidelines for oligoasthenoteratozoospermia (OAT; Colpi et al. 2018) and Klinefelter syndrome (Zitzmann et al. 2021) by the European Academy of Andrology, and diagnostics/treatment prior to use of assisted reproductive technologies (Toth et al. 2019) by the German Association of the Scientific Medical Societies. However, sperm count cut-offs and genetic testing approaches vary widely between countries. Patients with oligozoospermia, <5 million sperm/ml (EU/US), or azoospermia are generally offered karyotyping and Y-chromosome microdeletion analysis, while CFTR testing is recommended for men with suspected congenital bilateral absence of the vas deferens (Krausz and Riera-Escamilla, 2018). These guidelines, or in some cases purely expert opinions, thus limit the genetic testing to the most common infertility presentations and exclude patients with other sperm phenotypes such as teratozoospermia/asthenoteratozoospermia (e.g. globozoospermia and MMAF) and other sperm motility disorders (asthenozoospermia/OAT). According to the average clinical presentation of German men over the last 30 years, this means that at least 40% of infertile men are not being offered any form of genetic testing (Tüttelmann et al. 2018).

In line with the definitive evidence that pathogenic variants in at least 36 genes may result in isolated infertility, we advocate for an update of genetic testing guidelines for male infertility. Ideally, the blood of infertile men should be used for WES to identify any protein-coding variants in a non-biased manner. We appreciate there will often be an issue with a lack of expertise in this technology, or access to sequencing platforms. IMIGC members (imigc.org) are situated across the US, Europe and Australia, and are interested in forming additional collaborations for clinical and research purposes. As a less impactful, backup approach, the genetic targets of panels used in targeting sequencing should at least be updated to include the genes validated in this study (those moderately linked and stronger). Additionally, a wider subset of patients/phenotypes should be sequenced. Such approaches will help to achieve more diagnoses, better personalised treatment, improved risk assessment for the transmission of infertility to offspring, and better counselling for potential health risks in the infertile man. It will also help to improve counselling of azoospermic men prior to testicular biopsy, as finding a genetic diagnosis may help in predicting the chances for successful testicular sperm extraction, which is an incredibly invasive procedure.

Wider implications of male infertility genetics

Precise phenotyping of men with pathogenic variants will assist in evaluating potential future health risks. Mounting evidence suggests that male infertility may pose as a predictive condition for life-threatening comorbidities in later life, such as various types of cancer and cardiovascular disease (Jensen et al. 2009; Eisenberg et al. 2014; Ehrlich, 2015). While the precise mechanism(s) underpinning this observation are unknown, it is notable that many genes that are highly expressed in the testis and lead to male infertility when mutated are also expressed in other organs, but at considerably lower levels. For example, multiple DNA repair genes linked to infertility and various cancers (reviewed in Nagirnaja et al. 2018), and TEX11, which is expressed in the testis and the pancreas according to RNA-seq data (GTEx Portal). As has been made clear above, this is the case for several cilia genes, where men often present with chronic bronchitis and other lung conditions (reviewed in Leigh et al. 2009; Sironen et al. 2020). Thus, additional studies where the health of infertile men is broadly investigated are urgently needed.

Although the topic could constitute an entire article itself, the rapidly increasing use of assisted reproductive technologies worldwide (de Mouzon et al. 2020) is resulting in the propagation of infertility-causing variants throughout the population. As we identify these variants in men who undergo assisted reproductive technologies, we will also be faced with delivering this information back to the patients through genetic counselling services. This is especially relevant for men with Y-linked variants, where any sons will presumably be inflicted with the same infertility phenotype as their affected fathers and thus also require ARTs if wanting their own biological child.

Strengths and limitations

We used the joint expertise of the IMIGC to systematically evaluate all available evidence (as of July 1st, 2020) for monogenic causes of isolated or syndromic male infertility, endocrine disorders or reproductive system abnormalities affecting the male sex organs. No other published reviews in this field come close to including the approximately 600 genes comprehensively evaluated here. We improved the quality of evidence scoring by including a larger number of reviewers from independent research groups with broad expertise in male infertility. As all research groups have been involved in disease gene discovery of various genes described in this study, we prevented bias as much as possible by not allowing scoring by reviewers who have worked on respective genes.