TY - JOUR
AU - Marzi, Sarah J
AB - Abstract BACKGROUND Studies in non-human mammals suggest that environmental factors can influence spermatozoal DNA methylation, and some research suggests that spermatozoal DNA methylation is also implicated in conditions such as subfertility and imprinting disorders in the offspring. Together with an increased availability of cost-effective methods of interrogating DNA methylation, this premise has led to an increasing number of studies investigating the DNA methylation landscape of human spermatozoa. However, how the human spermatozoal DNA methylome is influenced by environmental factors is still unclear, as is the role of human spermatozoal DNA methylation in subfertility and in influencing offspring health. OBJECTIVE AND RATIONALE The aim of this systematic review was to critically appraise the quality of the current body of literature on DNA methylation in human spermatozoa, summarize current knowledge and generate recommendations for future research. SEARCH METHODS A comprehensive literature search of the PubMed, Web of Science and Cochrane Library databases was conducted using the search terms ‘semen’ OR ‘sperm’ AND ‘DNA methylation’. Publications from 1 January 2003 to 2 March 2020 that studied human sperm and were written in English were included. Studies that used sperm DNA methylation to develop methodologies or forensically identify semen were excluded, as were reviews, commentaries, meta-analyses or editorial texts. The Grading of Recommendations, Assessment, Development and Evaluations (GRADE) criteria were used to objectively evaluate quality of evidence in each included publication. OUTCOMES The search identified 446 records, of which 135 were included in the systematic review. These 135 studies were divided into three groups according to area of research; 56 studies investigated the influence of spermatozoal DNA methylation on male fertility and abnormal semen parameters, 20 studies investigated spermatozoal DNA methylation in pregnancy outcomes including offspring health and 59 studies assessed the influence of environmental factors on spermatozoal DNA methylation. Findings from studies that scored as ‘high’ and ‘moderate’ quality of evidence according to GRADE criteria were summarized. We found that male subfertility and abnormal semen parameters, in particular oligozoospermia, appear to be associated with abnormal spermatozoal DNA methylation of imprinted regions. However, no specific DNA methylation signature of either subfertility or abnormal semen parameters has been convincingly replicated in genome-scale, unbiased analyses. Furthermore, although findings require independent replication, current evidence suggests that the spermatozoal DNA methylome is influenced by cigarette smoking, advanced age and environmental pollutants. Importantly however, from a clinical point of view, there is no convincing evidence that changes in spermatozoal DNA methylation influence pregnancy outcomes or offspring health. WIDER IMPLICATIONS Although it appears that the human sperm DNA methylome can be influenced by certain environmental and physiological traits, no findings have been robustly replicated between studies. We have generated a set of recommendations that would enhance the reliability and robustness of findings of future analyses of the human sperm methylome. Such studies will likely require multicentre collaborations to reach appropriate sample sizes, and should incorporate phenotype data in more complex statistical models. DNA methylation, methylome, human sperm, human spermatozoa, fertility, reproductive health, epigenetics, transgenerational inheritance, epigenetic association studies Introduction Adequate functioning of germ cells is imperative for fertility, embryogenesis and offspring health. Infertility, commonly defined as a failure to achieve a successful pregnancy after 12 months or more of regular unprotected intercourse, affects approximately one in seven couples in the UK, and is attributed to male factor infertility in around half of cases (American Society for Reproductive Medicine, 2013; Fields et al., 2013). Epidemiologically, male factor infertility has been linked to a range of lifestyle and environmental factors, including cigarette smoking, obesity and toxicants such as organophosphates (Fields et al., 2013; Mima et al., 2018). Observational studies also suggest that environmental and acquired paternal traits, such as advanced age and smoking, can influence the health of his offspring, presumably via alterations to spermatozoa (Oldereid et al., 2018). Investigation into spermatozoal genomic imprinting was further prompted by epidemiological reports of an association between ART and an increased risk of imprinting disorders in the offspring (Gosden et al., 2003), as it was hypothesized that spermatozoa from subfertile males who conceived with the aid of ART could harbour a greater frequency of abnormally imprinted genes. Together, these observations have generated considerable interest in the epigenetic landscape of human spermatozoa. Epigenetics refers to mitotically or meiotically heritable changes in gene expression that do not involve changes in the underlying DNA sequence (Dupont et al., 2009). The term is generally taken to include three distinct but interrelated mechanisms that alter the accessibility of chromatin towards the transcriptional machinery or regulate gene expression on a post-transcriptional level; DNA modifications, in humans most prevalently via methylation of CpG dinucleotides (forming 5-mC), post-translational modifications of histones (including methylation, acetylation, phosphorylation and sumoylation), and non-coding RNAs (Daxinger and Whitelaw, 2012; Desai et al., 2015) (Fig. 1). Of these, DNA methylation is the most studied epigenetic mechanism to date, owing to its relative stability and to the availability of cost-effective tools for analysing DNA methylation at single base resolution. We have therefore made DNA methylation specifically the focus of this review. Figure 1. Open in new tabDownload slide Overview of epigenetic mechanisms. The epigenetic machinery includes three distinct but interrelated mechanisms; DNA modifications, post-translational modifications of histones and non-coding RNA. The most prevalent DNA modification, DNA methylation, involves the attachment of a methyl group onto cytosine residues, forming 5-mC. lncRNA, long non-coding RNA. Figure 1. Open in new tabDownload slide Overview of epigenetic mechanisms. The epigenetic machinery includes three distinct but interrelated mechanisms; DNA modifications, post-translational modifications of histones and non-coding RNA. The most prevalent DNA modification, DNA methylation, involves the attachment of a methyl group onto cytosine residues, forming 5-mC. lncRNA, long non-coding RNA. In non-human mammals, the sperm methylome has been shown to be responsive to environmental and physiological changes, including dietary alterations, toxicants and even psychological stress (Dias and Ressier, 2014; Radford et al., 2014; Barbosa et al., 2016; Youngson et al., 2016; Sakai et al., 2018; Watkins et al., 2018). In addition, some studies suggest that acquired traits in male mice can induce epigenetic changes in sperm that influence offspring physiology (Wei et al., 2014; Huypens et al., 2016). Compared to readily available somatic tissues, such as peripheral blood, our knowledge of the human sperm methylome is considerably less detailed. Nevertheless, the last two decades have seen an increasing number of studies investigating epigenetic signatures of human germ cells, fuelled in part by an increasing interest in the potential for acquired phenotypes to influence offspring physiology. The aim of this systematic review was to critically appraise the current body of research, summarize credible findings and generate recommendations for future research in the context of investigating DNA methylation in human spermatozoa. In doing so, we interrogated the available literature to answer three specific questions: What is the strength of evidence for an association between spermatozoal DNA methylation and male subfertility, including abnormal semen parameters? What is the strength of evidence for an association between spermatozoal DNA methylation and pregnancy outcomes and offspring health? What is the strength of evidence for environmental influences on spermatozoal DNA methylation? Methods Protocol and registration The methods employed for the systematic review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Liberati et al., 2009). The protocol for the review was registered on the international prospective register for systematic reviews, PROSPERO, before data extraction was completed (CRD42019146399; https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=146399). Search strategy An exhaustive literature search was performed on 2 March 2020 using the PubMed electronic database and the following search criteria: ‘semen’ [Mesh] OR ‘sperm’ [Mesh] AND ‘DNA methylation’ [Mesh] The same search criteria were thereafter applied to search the Web of Science and Cochrane databases to identify further relevant articles. Only publications that pertained to studies in humans and were written in English were included. Reference lists of identified articles and review articles were also searched for additional references. The search and filtering process were performed independently by two authors to avoid selection bias. Selection criteria Articles were selected for inclusion in the systematic review if they fulfilled the following criteria: Studies investigating DNA methylation in human spermatozoa. Any age group. Investigations of any type of analysis method of DNA methylation. Publications in English only. Published from 2003 onwards. A cut-off date of 2003 was applied as this was deemed a realistic date for studies to begin using next-generation sequencing and microarray-based technologies, and so to minimize penalization of earlier studies due to an absence of the technology. Thus, articles published between 1 January 2003 and 2 March 2020 were included. Articles identified through the search, but which did not meet the inclusion criteria detailed above, were recorded along with their reason for being excluded (Supplementary Data S1). A flow diagram of the selection process, including the number of articles filtered at each step of the process is shown in Fig. 2. A description of semen sample parameters referred to throughout this systematic review is provided in Supplementary Data S2. Figure 2. Open in new tabDownload slide Flow chart of selection of records for the systematic review. Assembled according to PRISMA guidelines (Liberati et al., 2009). Figure 2. Open in new tabDownload slide Flow chart of selection of records for the systematic review. Assembled according to PRISMA guidelines (Liberati et al., 2009). Data extraction The following information was extracted from every identified study: Author. Year of publication. Rationale for study. Method of assaying DNA methylation. Area of research (fertility or semen parameters, pregnancy outcomes and offspring health, toxicants, metabolic interventions (including diet and lifestyle), studies of tissue specificity (including cancer, ageing and methodological studies comparing tissues) and other (including studies of genetic and psychiatric conditions, and methodological experiments)). Method of semen processing, i.e. the method of removing somatic cells. A description of different methods of semen processing relevant to this systematic review is provided in Supplementary Data S2. Global, targeted or genome-wide analysis of DNA methylation. Number of study participants. Age of study participants. Matching of cases and controls (where relevant). Main results of study. Statistical analyses, e.g. correction for multiple comparison. Covariates. Exclusion criteria applied in the study. Information was recorded by the first primary author and independently verified by the second primary author. Any disagreements were resolved in discussion with the senior author. Complete records can be found in Supplementary Data S3. Quality of evidence assessment The Grading of Recommendations, Assessment, Development and Evaluations (GRADE) criteria were used to objectively evaluate the quality of evidence in every study included in the systematic review (Ryan and Hill, 2016). The criteria were adapted for relevance to the subject of the review, as summarized in Supplementary Table SI (Supplementary Data S2). No randomized controlled trials were identified in the search; therefore, quality of evidence ratings were assigned based on the scoring of the other criteria. Studies were each scored based on risk of bias (application of appropriate statistics, including correction for multiple tests, and consideration of covariates), inconsistency, indirectness, imprecision, publication bias and other upgrading factors, such as large effect sizes, large sample sizes or use of a replication cohort. Publications with GRADE ratings of ≥0 were assigned a quality score of ‘high’, −1 were assigned ‘moderate’, −2 were assigned ‘low’ and ≤ −3 were assigned ‘very low’. Only findings of studies scoring ‘high’ or ‘moderate’ for quality of evidence were summarized. Synthesis of results The quality of evidence was graded for each study and is presented in Tables I–III. Results were summarized according to the research question they addressed. Recommendations for future DNA methylation studies in human sperm were given based on the perceived shortcomings of previous work and the gaps in the existing literature. All figures were generated in Adobe Illustrator® (Adobe, Inc., San Jose, CA, USA) and Microsoft Excel (Microsoft Corporation, Redmond, WA, USA) specifically for this publication. Table I Studies assessing the effect of DNA methylation in human sperm on fertility and semen parameters. Author, year . Imprinting studied? . Method of assaying DNA methylation . Initial processing . N (cases) . Age in years (SD); description of cases . N (controls) . Age in years; description of controls . Main results . Study quality . Laqqan et al. (2017b) No 450K array; targeted bisulphite sequencing Gradient (50%) centrifugation. Medium ND 70 35.89 (±6.03); 15 for screening and 55 for validation 71 33.98 (±5.53); 15 for screening and 56 for validation Significant assoc. between subfertility and altered DNAm at CpGs related to ALS2CR12, ALDH3B2, PRICKLE2 and PTGIR High Laqqan et al. (2017c) No 450K array; targeted bisulphite sequencing Somatic cell lysis and PureSperm gradient (45% and 90%) 101 36.4 (±3.24); 20 + 81 (screening + validation) OZS 66 36.4 (±3.24); 20 + 44 (screening + validation) proven fertility Assoc. between OZS and DNAm of UBE2G2 and cg04807108 High Tian et al. (2015) Yes Methylation-specific PCR Samples washed in PBS and sperm wash buffer 118 31.4 (±5.1); males under evaluation for subfertility None N/A LINE-1 DNAm negatively correlated with sperm motility High Wu et al. (2010) No Targeted bisulphite Sanger sequencing Swim-up 94 29.04 (±4.52); idiopathic subfertility 54 29.52 (±3.72); proven fertility Significantly ↑ DNAm of the MTHFR promoter in subfertile males cf. fertile controls High Poplinski et al. (2010) Yes Targeted bisulphite Sanger sequencing Swim-up 148 35.5 (32–41); subfertile 33 37.0 (31.5–40); NZS; infertility attributed to female side MEST ↓ DNAm significantly assoc. with OZS in subfertile males High Kabartan et al. (2019) No Methylation-specific PCR Somatic cell lysis 73 31.95 (±5.5); OZS, infertile 20 27.2 (±2.6); NZS, proven fertility No assoc. between DNA fragmentation and DNAm of BRCA1 or BRCA2 in OZS males Moderate Tang et al. (2018) Yes Bis-PCR Percoll gradient (40% and 80%) 135 29.07 (±4.25); subfertile 59 29.08 (±3.84); proven fertility Aberrant DNAm in imprinted genes of infertile men Moderate Laqqan et al. (2018) No 450K array followed by Bis-PCR-Seq Somatic cell lysis 50 38.7 (±6.9); subfertile males 28 38.5 (±5.6); proven fertility Assoc. between subfertility and DNAm of PRRC2A, ANXA2, MAPK8IP3 and GAA Moderate Camprubi et al. (2017) No 450K array Somatic cell lysis 19 25.95 (±1.41); proven fertility N/A N/A Promoter CpGs relatively ↓ DNAm in spermatozoa. Identified 94 genes that appear resistant to demethylation Moderate Dong et al. (2017) Yes Bisulphite pyrosequencing Percoll gradient (45% and 90%) 155 31.13–32.17; 48 OZS, 52 AsZS, 55 TZS 50 32.22 (±3.59); NZS In smokers, subfertility was assoc. with ↓ DNAm of H19 in OZS and ↑ DNAm of SNRPN-ICR in AsZS and TZS Moderate Laqqan et al. (2017a) No 450K array; targeted bisulphite sequencing Somatic cell lysis 72 37.4 (±6.1); 15 + 57 (screening and validation) from subfertile men (unable to conceive for at least 10 years) 36 38.5 (±5.2); 15 + 21 (screening and validation) men with proved fertility Assoc. between subfertility and DNAm in KCNJ5, MLPH and SMC1B Moderate Camprubi et al. (2016) No 450K array Somatic cell lysis 42 38.36 (±5.31); subfertile males 19 25.95 (±4.80); proven fertility Assoc. between age and DNAm of RPS6KA2; OZS and DNAm of APCS; chromosome abnormalities and DNAm of JAM3/NCAPD3 and between fecundity and DNAm of ANK2 Moderate Jenkins et al. (2016a) No 450K array ISolate gradient 94 32.0–36.32; classified according to semen parameters N/A N/A No DNAm alterations associated with TZS; abnormal semen had increased DNAm variability Moderate Jenkins et al. (2015) No 450K array ISolate gradient (35% and 90%) 20 35.69 (±2.0); subfertile N/A N/A Higher DNAm variability in low-quality sperm Moderate Botezatu et al. (2014) No Methylation-specific PCR ND 27 26–41; OAT 11 24–37; NZS ↓ DNAm of MTHFR and SNRPN in samples with low sperm motility. No significant trend for other semen parameters Moderate Schutte et al. (2013) No 27K array followed by targeted bisulphite pyrosequencing Swim-up procedure and PureSperm gradient (40% and 80%) 38 27–54; males under evaluation for subfertility (longitudinal) 4 Age ND; NZS 27K array data showed clustering by sperm count; no candidate genes showed significant difference Moderate Klaver et al. (2013) Yes Bisulphite pyrosequencing Swim-up 212 Age ND; males under evaluation for subfertility 41 34.5 (±7.2); 31 NZS samples. Longitudinal MEST meth assoc. with OZS. DNAm levels at selected genes in NZS males were stable for up to 951 days Moderate Barzideh et al. (2013) No HPLC for DNAm; flow cytometry Percoll gradient 16 20–23; NZS N/A N/A DNAm ↓ in high-density (high quality) portion of spermatozoa than in low-density portion Moderate Pacheco et al. (2011) No 27K array Percoll gradient (50%) 21 Age ND; males under evaluation for subfertility None N/A Assoc. between 9189 CpGs and low sperm motility Moderate Marques et al. (2008) Yes Bisulphite sequencing (Sanger) SupraSperm gradient (55%, 80% and 90%) and swim-up 20 Age ND; 6 OZS; 14 AsZS; all abnormal morphology ND Age ND; NZS H19 and MEST abnormally imprinted in severely OZS group Moderate Houshdaran et al. (2007) No MethyLight; Illumina bead array of 1421 CpG sites ISolate gradient 65 22–49; partners of females undergoing fertility investigation None N/A Assoc. between sperm parameters and DNAm of HRAS, NTF3, MT1A, PAX8, DIRAS3, PLAGL1, SFN, SAT2CHRM1 and MEST Moderate Marques et al. (2004) Yes Bisulphite sequencing Gradient centrifugation and swim-up 96 Age ND; OZS 27 Age ND; NZS H19 abnormally imprinted in OZS group Moderate Cheng et al. (2017) No Agilent custom 1M Promoter-CpG island microarray; targeted bisulphite pyrosequencing Enzymatic digestion and countercurrent centrifugal elutriation 17 33.2 (±0.5); 1 sample from patient with moderate HS; candidate genes in 16 patients 9 ND; 9 NZS males with obstructive AZS, 1 selected for whole genome methylation profiling HS assoc. with DNAm of BOLL, DDX4, HORMAD1 and MAEL Low Du et al. (2016) No Liquid hybridization (promoter capture) capture-based bisulphite sequencing Percoll gradient (47.5%, 57%, 76% and 95%) 7 31.7 (±4.0); AsZS 8 29.8 (± 3.2); NZS No significant assoc. between DNAm or DNAm variability and AsZS Low Jenkins et al. (2016b) No 450K array Somatic cell lysis 29 27.55 (±0.71); subfertile 27 29.74 (±0.71); proven fertility DNAm of HSPA1L and HSPA1B significantly assoc. with pregnancy rates Low Louie et al. (2016) Yes Methylation-specific PCR Swim-up for high concentration samples. Handpicked spermatozoa for low concentration samples 44 32.9–35.7; OAT ND 34.1 (±2.4); NZS No significant association between DNAm in selected imprinted genes, severe OAT and the MTHFR C677T SNP Low Cassuto et al. (2016) No 5-mC immuno Sperm separation medium gradient (45% and 90%) 10 37–50; 448 S6 spermatozoa compared to 428 S0 spermatozoa N/A N/A ↓ DNAm in morphologically higher scoring spermatozoa compared to morphologically lower scoring spermatozoa Low Laurentino et al. (2015) Yes Bisulphite pyroseq. and deep targeted bis-seq Swim-up 26 34.0–39.33; males with abnormal sperm parameters 19 33.68 (±1.58); proven fertility Significantly ↑ variation in the DNAm values of the maternally methylated gene KCNQ1OT1 in samples with abnormal sperm parameters Low Zhou et al. (2015) No Methylation-specific PCR and bisulphite sequencing PCR Percoll gradient (50%) 48 30.13 (±5.8); AsZS 42 29.0 (±4.8); NZS No DNAm of the CRISP2 promoter; expression likely regulated by miR-27b Low Urdinguio et al. (2015) No 450K array; targeted bisulphite pyrosequencing; immunostaining for DNAm PureSperm gradient (65% and 90%) 7 30–55; 29 subfertile NZS (7 included in array analysis) 5 22–49; 17 NZS fertile males (5 included in array analysis) A number of CpG sites significantly differentially methylated between subfertile cases and fertile controls Low Tavalaee et al. (2015) No 5-mC immuno Samples washed in PBS 23 31.3 (±4.3); grades II and III varicocelectomy (longitudinal) None N/A Varicocele surgery associated with improved sperm parameters, but no significant change in DNAm Low Montjean et al. (2015) Yes Global DNAm assayed by immunostaining for DNAm, targeted bisulphite sequencing Percoll gradient 30 38.3 (±6); OAT 62 38.5 (±5.3); NZS Significant association between H19 DNAm and sperm parameters Low Rotondo et al. (2013) Yes Methylation-specific PCR and combined bisulphite restriction analysis PureSperm gradient (40% and 80%) 10 36.1 (±1.2); MTHFR promoter hypermethylation; 5 with normal and 5 with abnormal semen parameters 10 Age ND; normal MTHFR promoter meth; 5 with normal and 5 with abnormal semen parameters For NZS samples, H19 meth was assoc. with MTHFR DNAm Low Sato et al. (2011) Yes COBRA and methylation-specific PCR with Luminex technology Swim-up 128 Age ND; OZS 209 Age ND; NZS Correlation between DNAm analysis methods. Higher levels of imprinting errors in OZS Low Boissonnas et al. (2010) Yes Targeted pyrosequencing Percoll gradient (45%, 60% and 90%) 41 36.6 (±5.7); divided into TZS, OZS and AsZS groups 17 Age ND; NZS ↓ DNAm of various CpG positions in the H19–IGF2 DMR in cases cf. controls Low Hammoud et al. (2010) Yes Targeted Sanger sequencing ND 20 Age ND; 10 with abnormal P1/P2 ratio and 10 with OZS ND Age ND; proven fertility ↑ DNAm of LIT1, SNRPN, MEST, ZAC, PEG3 in OZS and abnormal P1/P2 ratio samples cf. NZS samples Low Marques et al. (2010) Yes Bisulphite sequencing Testicular biopsies washed in sperm preparation medium and somatic cells removed with erythrocyte lysing buffer 24 22–44; AZS (sperm obtained from testicular biopsies) ND Age ND; NZS H19 DNAm errors identified in AZS samples obtained via testicular biopsy Low Navarro-Costa et al. (2010) No Bisulphite sequencing PureSperm gradient (40% and 80%) 5 39.4 (±7.2); OZS 5 39.2 (±7.3); NZS ↑ DNAm of the DAZL (but not DAZ) promoter more prevalent in OZS samples compared to NZS controls Low Aoki et al. (2006) No Quantitative immune-fluorescence microscopy ND 195 Age ND; subfertile (3 groups based on P1/P2 ratios) None N/A No significant differences in DNAm between P1/P2 groups Low Sujit et al. (2018) No 450K array; targeted bisulphite sequencing Somatic cell lysis for cases and swim-up procedure for controls 38 Age ND; subfertile; OZS 26 Age ND; NZS 1436 probes had ↑ DNAm and 244 had ↓ DNAm in subfertile OZS cases relative to fertile NZS controls Very low Gunes et al. (2018) No Methylation-specific PCR Somatic cell lysis 10 Age ND; OAT 29 Age ND; NZS In OAT men, MLH1 was ↑ DNAm. Seminal ROS levels were assoc. with MLH1 DNAm Very low Olszewska et al. (2017) No TLC and immuno-fluorescence Samples washed in F10 medium 9 30.7; subfertile carriers of chromosomal abnormalities 14 28.5; healthy volunteers Chromosomal abnormalities not assoc. with DNAm Very low Marques et al. (2017) Yes Methylation-specific PCR Micromanipulation from testicular biopsy 15 Age ND; AsZS ND Age ND; NZS males unable to produce semen due to spinal cord injuries No significant association between H19 and MEST DNAm and AsZS Very low Nasri et al. (2017) Yes Combined bisulphite restriction analysis PureSperm gradient (40% and 80%) 23 Age ND; samples with abnormal sperm parameters 11 Age ND; NZS No significant difference between H19 DNAm and sperm parameters Very low Xu et al. (2016a) No Methylation-specific PCR ND 27 28; AsZS 25 28; NZS AsZS significantly assoc. with abnormal DNAm of the VDAC2 promoter Very low Xu et al. (2016b) Yes Targeted DNAm quantification using the MassARRAY EpiTYPER Percoll gradient (50%) 46 31.95 (±3.1); AZS 49 32.16 (±3.26); NZS ↓ DNAm in MEST, GNAS, FAM50B, H19, LINE‐1 and P16 in AsZS males Very low Uppangala et al. (2016) No 5-mC immuno Swim-up 19 Age ND; proven fertility. 1, 3, 5 and 7 days of abstinence N/A N/A No significant association between DNAm and length of ejaculatory abstinence Very low Bahreinian et al. (2015) No Flow cytometric immunodetection for DNAm Samples washed in PBS 44 31.1 (±0.6); patients with grades II and III varicocele 15 37.8 (±2.0); proven fertility Varicocele associated with ↓ DNAm and poorer sperm parameters Very low Kuhtz et al. (2014) Yes Bisulphite sequencing Swim-up and PureSperm gradient (40% and 80%) 27 Age ND; OAT 27 Age ND; NZS No significant differences in the occurrence of abnormal methylation imprints between sperm with and without morphological abnormalities Very low Richardson et al. (2014) No Bisulphite pyrosequencing Swim-up 95 32–39; abnormal semen parameters 45 34–40; NZS The RHOX homeobox gene cluster is regulated by DNAm and RHOX gene cluster ↑ DNAm is significantly assoc. with semen abnormalities Very low Camprubi et al. (2013) Yes Bisulphite pyrosequencing Swim-up 6 Age ND; subfertile males showing hypomethylation of the H19-ICR locus None N/A No significant association between CTCFL mutations and H19-ICR sperm DNAm Very low Li et al. (2013) Yes Methylation-specific PCR Gradient (45%, 60% and 90%) gradient centrifugation. Both Percoll and PureSperm mentioned 40 Age ND; 20 OZS and 20 AsZS men 20 Age ND; NZS, proven fertility No significant association between H19 and DAZL DNAm and AsZS Very low Minor et al. (2011) Yes Bisulphite sequencing Swim-up 18 37.8 (±6.9); AZS (sperm obtained from testicular biopsies) 9 34.1 (±2.4); proven fertility (from ejaculate) ↓ DNAm of H19 in AZS males Very low Nanassy and Carrell (2011a) No Bisulphite pyrosequencing ND 92 28–33; 60 with abnormal P1/P2 ratio, 32 with OZS 40 32–33; NZS, proven fertility, normal P1/P2 ratio ↑ Freq. CREM promoter methylation in males with abnormal P1/P2 ratio and in OZS cf. fertile, NZS controls Very low Nanassy and Carrell (2011b) No Bisulphite sequencing ND 10 Age ND; abnormal P1/P2 ratio 10 Age ND; NZS No significant DNAm differences in the selected genes between samples with an abnormal P1/P2 ratio and NZS controls Very low Li et al. (2006) No Bisulphite sequencing Samples washed in PBS 4 Age ND; proven fertility N/A N/A DAZ1 promoter methylated in leukocytes but unmethylated in sperm Very low Author, year . Imprinting studied? . Method of assaying DNA methylation . Initial processing . N (cases) . Age in years (SD); description of cases . N (controls) . Age in years; description of controls . Main results . Study quality . Laqqan et al. (2017b) No 450K array; targeted bisulphite sequencing Gradient (50%) centrifugation. Medium ND 70 35.89 (±6.03); 15 for screening and 55 for validation 71 33.98 (±5.53); 15 for screening and 56 for validation Significant assoc. between subfertility and altered DNAm at CpGs related to ALS2CR12, ALDH3B2, PRICKLE2 and PTGIR High Laqqan et al. (2017c) No 450K array; targeted bisulphite sequencing Somatic cell lysis and PureSperm gradient (45% and 90%) 101 36.4 (±3.24); 20 + 81 (screening + validation) OZS 66 36.4 (±3.24); 20 + 44 (screening + validation) proven fertility Assoc. between OZS and DNAm of UBE2G2 and cg04807108 High Tian et al. (2015) Yes Methylation-specific PCR Samples washed in PBS and sperm wash buffer 118 31.4 (±5.1); males under evaluation for subfertility None N/A LINE-1 DNAm negatively correlated with sperm motility High Wu et al. (2010) No Targeted bisulphite Sanger sequencing Swim-up 94 29.04 (±4.52); idiopathic subfertility 54 29.52 (±3.72); proven fertility Significantly ↑ DNAm of the MTHFR promoter in subfertile males cf. fertile controls High Poplinski et al. (2010) Yes Targeted bisulphite Sanger sequencing Swim-up 148 35.5 (32–41); subfertile 33 37.0 (31.5–40); NZS; infertility attributed to female side MEST ↓ DNAm significantly assoc. with OZS in subfertile males High Kabartan et al. (2019) No Methylation-specific PCR Somatic cell lysis 73 31.95 (±5.5); OZS, infertile 20 27.2 (±2.6); NZS, proven fertility No assoc. between DNA fragmentation and DNAm of BRCA1 or BRCA2 in OZS males Moderate Tang et al. (2018) Yes Bis-PCR Percoll gradient (40% and 80%) 135 29.07 (±4.25); subfertile 59 29.08 (±3.84); proven fertility Aberrant DNAm in imprinted genes of infertile men Moderate Laqqan et al. (2018) No 450K array followed by Bis-PCR-Seq Somatic cell lysis 50 38.7 (±6.9); subfertile males 28 38.5 (±5.6); proven fertility Assoc. between subfertility and DNAm of PRRC2A, ANXA2, MAPK8IP3 and GAA Moderate Camprubi et al. (2017) No 450K array Somatic cell lysis 19 25.95 (±1.41); proven fertility N/A N/A Promoter CpGs relatively ↓ DNAm in spermatozoa. Identified 94 genes that appear resistant to demethylation Moderate Dong et al. (2017) Yes Bisulphite pyrosequencing Percoll gradient (45% and 90%) 155 31.13–32.17; 48 OZS, 52 AsZS, 55 TZS 50 32.22 (±3.59); NZS In smokers, subfertility was assoc. with ↓ DNAm of H19 in OZS and ↑ DNAm of SNRPN-ICR in AsZS and TZS Moderate Laqqan et al. (2017a) No 450K array; targeted bisulphite sequencing Somatic cell lysis 72 37.4 (±6.1); 15 + 57 (screening and validation) from subfertile men (unable to conceive for at least 10 years) 36 38.5 (±5.2); 15 + 21 (screening and validation) men with proved fertility Assoc. between subfertility and DNAm in KCNJ5, MLPH and SMC1B Moderate Camprubi et al. (2016) No 450K array Somatic cell lysis 42 38.36 (±5.31); subfertile males 19 25.95 (±4.80); proven fertility Assoc. between age and DNAm of RPS6KA2; OZS and DNAm of APCS; chromosome abnormalities and DNAm of JAM3/NCAPD3 and between fecundity and DNAm of ANK2 Moderate Jenkins et al. (2016a) No 450K array ISolate gradient 94 32.0–36.32; classified according to semen parameters N/A N/A No DNAm alterations associated with TZS; abnormal semen had increased DNAm variability Moderate Jenkins et al. (2015) No 450K array ISolate gradient (35% and 90%) 20 35.69 (±2.0); subfertile N/A N/A Higher DNAm variability in low-quality sperm Moderate Botezatu et al. (2014) No Methylation-specific PCR ND 27 26–41; OAT 11 24–37; NZS ↓ DNAm of MTHFR and SNRPN in samples with low sperm motility. No significant trend for other semen parameters Moderate Schutte et al. (2013) No 27K array followed by targeted bisulphite pyrosequencing Swim-up procedure and PureSperm gradient (40% and 80%) 38 27–54; males under evaluation for subfertility (longitudinal) 4 Age ND; NZS 27K array data showed clustering by sperm count; no candidate genes showed significant difference Moderate Klaver et al. (2013) Yes Bisulphite pyrosequencing Swim-up 212 Age ND; males under evaluation for subfertility 41 34.5 (±7.2); 31 NZS samples. Longitudinal MEST meth assoc. with OZS. DNAm levels at selected genes in NZS males were stable for up to 951 days Moderate Barzideh et al. (2013) No HPLC for DNAm; flow cytometry Percoll gradient 16 20–23; NZS N/A N/A DNAm ↓ in high-density (high quality) portion of spermatozoa than in low-density portion Moderate Pacheco et al. (2011) No 27K array Percoll gradient (50%) 21 Age ND; males under evaluation for subfertility None N/A Assoc. between 9189 CpGs and low sperm motility Moderate Marques et al. (2008) Yes Bisulphite sequencing (Sanger) SupraSperm gradient (55%, 80% and 90%) and swim-up 20 Age ND; 6 OZS; 14 AsZS; all abnormal morphology ND Age ND; NZS H19 and MEST abnormally imprinted in severely OZS group Moderate Houshdaran et al. (2007) No MethyLight; Illumina bead array of 1421 CpG sites ISolate gradient 65 22–49; partners of females undergoing fertility investigation None N/A Assoc. between sperm parameters and DNAm of HRAS, NTF3, MT1A, PAX8, DIRAS3, PLAGL1, SFN, SAT2CHRM1 and MEST Moderate Marques et al. (2004) Yes Bisulphite sequencing Gradient centrifugation and swim-up 96 Age ND; OZS 27 Age ND; NZS H19 abnormally imprinted in OZS group Moderate Cheng et al. (2017) No Agilent custom 1M Promoter-CpG island microarray; targeted bisulphite pyrosequencing Enzymatic digestion and countercurrent centrifugal elutriation 17 33.2 (±0.5); 1 sample from patient with moderate HS; candidate genes in 16 patients 9 ND; 9 NZS males with obstructive AZS, 1 selected for whole genome methylation profiling HS assoc. with DNAm of BOLL, DDX4, HORMAD1 and MAEL Low Du et al. (2016) No Liquid hybridization (promoter capture) capture-based bisulphite sequencing Percoll gradient (47.5%, 57%, 76% and 95%) 7 31.7 (±4.0); AsZS 8 29.8 (± 3.2); NZS No significant assoc. between DNAm or DNAm variability and AsZS Low Jenkins et al. (2016b) No 450K array Somatic cell lysis 29 27.55 (±0.71); subfertile 27 29.74 (±0.71); proven fertility DNAm of HSPA1L and HSPA1B significantly assoc. with pregnancy rates Low Louie et al. (2016) Yes Methylation-specific PCR Swim-up for high concentration samples. Handpicked spermatozoa for low concentration samples 44 32.9–35.7; OAT ND 34.1 (±2.4); NZS No significant association between DNAm in selected imprinted genes, severe OAT and the MTHFR C677T SNP Low Cassuto et al. (2016) No 5-mC immuno Sperm separation medium gradient (45% and 90%) 10 37–50; 448 S6 spermatozoa compared to 428 S0 spermatozoa N/A N/A ↓ DNAm in morphologically higher scoring spermatozoa compared to morphologically lower scoring spermatozoa Low Laurentino et al. (2015) Yes Bisulphite pyroseq. and deep targeted bis-seq Swim-up 26 34.0–39.33; males with abnormal sperm parameters 19 33.68 (±1.58); proven fertility Significantly ↑ variation in the DNAm values of the maternally methylated gene KCNQ1OT1 in samples with abnormal sperm parameters Low Zhou et al. (2015) No Methylation-specific PCR and bisulphite sequencing PCR Percoll gradient (50%) 48 30.13 (±5.8); AsZS 42 29.0 (±4.8); NZS No DNAm of the CRISP2 promoter; expression likely regulated by miR-27b Low Urdinguio et al. (2015) No 450K array; targeted bisulphite pyrosequencing; immunostaining for DNAm PureSperm gradient (65% and 90%) 7 30–55; 29 subfertile NZS (7 included in array analysis) 5 22–49; 17 NZS fertile males (5 included in array analysis) A number of CpG sites significantly differentially methylated between subfertile cases and fertile controls Low Tavalaee et al. (2015) No 5-mC immuno Samples washed in PBS 23 31.3 (±4.3); grades II and III varicocelectomy (longitudinal) None N/A Varicocele surgery associated with improved sperm parameters, but no significant change in DNAm Low Montjean et al. (2015) Yes Global DNAm assayed by immunostaining for DNAm, targeted bisulphite sequencing Percoll gradient 30 38.3 (±6); OAT 62 38.5 (±5.3); NZS Significant association between H19 DNAm and sperm parameters Low Rotondo et al. (2013) Yes Methylation-specific PCR and combined bisulphite restriction analysis PureSperm gradient (40% and 80%) 10 36.1 (±1.2); MTHFR promoter hypermethylation; 5 with normal and 5 with abnormal semen parameters 10 Age ND; normal MTHFR promoter meth; 5 with normal and 5 with abnormal semen parameters For NZS samples, H19 meth was assoc. with MTHFR DNAm Low Sato et al. (2011) Yes COBRA and methylation-specific PCR with Luminex technology Swim-up 128 Age ND; OZS 209 Age ND; NZS Correlation between DNAm analysis methods. Higher levels of imprinting errors in OZS Low Boissonnas et al. (2010) Yes Targeted pyrosequencing Percoll gradient (45%, 60% and 90%) 41 36.6 (±5.7); divided into TZS, OZS and AsZS groups 17 Age ND; NZS ↓ DNAm of various CpG positions in the H19–IGF2 DMR in cases cf. controls Low Hammoud et al. (2010) Yes Targeted Sanger sequencing ND 20 Age ND; 10 with abnormal P1/P2 ratio and 10 with OZS ND Age ND; proven fertility ↑ DNAm of LIT1, SNRPN, MEST, ZAC, PEG3 in OZS and abnormal P1/P2 ratio samples cf. NZS samples Low Marques et al. (2010) Yes Bisulphite sequencing Testicular biopsies washed in sperm preparation medium and somatic cells removed with erythrocyte lysing buffer 24 22–44; AZS (sperm obtained from testicular biopsies) ND Age ND; NZS H19 DNAm errors identified in AZS samples obtained via testicular biopsy Low Navarro-Costa et al. (2010) No Bisulphite sequencing PureSperm gradient (40% and 80%) 5 39.4 (±7.2); OZS 5 39.2 (±7.3); NZS ↑ DNAm of the DAZL (but not DAZ) promoter more prevalent in OZS samples compared to NZS controls Low Aoki et al. (2006) No Quantitative immune-fluorescence microscopy ND 195 Age ND; subfertile (3 groups based on P1/P2 ratios) None N/A No significant differences in DNAm between P1/P2 groups Low Sujit et al. (2018) No 450K array; targeted bisulphite sequencing Somatic cell lysis for cases and swim-up procedure for controls 38 Age ND; subfertile; OZS 26 Age ND; NZS 1436 probes had ↑ DNAm and 244 had ↓ DNAm in subfertile OZS cases relative to fertile NZS controls Very low Gunes et al. (2018) No Methylation-specific PCR Somatic cell lysis 10 Age ND; OAT 29 Age ND; NZS In OAT men, MLH1 was ↑ DNAm. Seminal ROS levels were assoc. with MLH1 DNAm Very low Olszewska et al. (2017) No TLC and immuno-fluorescence Samples washed in F10 medium 9 30.7; subfertile carriers of chromosomal abnormalities 14 28.5; healthy volunteers Chromosomal abnormalities not assoc. with DNAm Very low Marques et al. (2017) Yes Methylation-specific PCR Micromanipulation from testicular biopsy 15 Age ND; AsZS ND Age ND; NZS males unable to produce semen due to spinal cord injuries No significant association between H19 and MEST DNAm and AsZS Very low Nasri et al. (2017) Yes Combined bisulphite restriction analysis PureSperm gradient (40% and 80%) 23 Age ND; samples with abnormal sperm parameters 11 Age ND; NZS No significant difference between H19 DNAm and sperm parameters Very low Xu et al. (2016a) No Methylation-specific PCR ND 27 28; AsZS 25 28; NZS AsZS significantly assoc. with abnormal DNAm of the VDAC2 promoter Very low Xu et al. (2016b) Yes Targeted DNAm quantification using the MassARRAY EpiTYPER Percoll gradient (50%) 46 31.95 (±3.1); AZS 49 32.16 (±3.26); NZS ↓ DNAm in MEST, GNAS, FAM50B, H19, LINE‐1 and P16 in AsZS males Very low Uppangala et al. (2016) No 5-mC immuno Swim-up 19 Age ND; proven fertility. 1, 3, 5 and 7 days of abstinence N/A N/A No significant association between DNAm and length of ejaculatory abstinence Very low Bahreinian et al. (2015) No Flow cytometric immunodetection for DNAm Samples washed in PBS 44 31.1 (±0.6); patients with grades II and III varicocele 15 37.8 (±2.0); proven fertility Varicocele associated with ↓ DNAm and poorer sperm parameters Very low Kuhtz et al. (2014) Yes Bisulphite sequencing Swim-up and PureSperm gradient (40% and 80%) 27 Age ND; OAT 27 Age ND; NZS No significant differences in the occurrence of abnormal methylation imprints between sperm with and without morphological abnormalities Very low Richardson et al. (2014) No Bisulphite pyrosequencing Swim-up 95 32–39; abnormal semen parameters 45 34–40; NZS The RHOX homeobox gene cluster is regulated by DNAm and RHOX gene cluster ↑ DNAm is significantly assoc. with semen abnormalities Very low Camprubi et al. (2013) Yes Bisulphite pyrosequencing Swim-up 6 Age ND; subfertile males showing hypomethylation of the H19-ICR locus None N/A No significant association between CTCFL mutations and H19-ICR sperm DNAm Very low Li et al. (2013) Yes Methylation-specific PCR Gradient (45%, 60% and 90%) gradient centrifugation. Both Percoll and PureSperm mentioned 40 Age ND; 20 OZS and 20 AsZS men 20 Age ND; NZS, proven fertility No significant association between H19 and DAZL DNAm and AsZS Very low Minor et al. (2011) Yes Bisulphite sequencing Swim-up 18 37.8 (±6.9); AZS (sperm obtained from testicular biopsies) 9 34.1 (±2.4); proven fertility (from ejaculate) ↓ DNAm of H19 in AZS males Very low Nanassy and Carrell (2011a) No Bisulphite pyrosequencing ND 92 28–33; 60 with abnormal P1/P2 ratio, 32 with OZS 40 32–33; NZS, proven fertility, normal P1/P2 ratio ↑ Freq. CREM promoter methylation in males with abnormal P1/P2 ratio and in OZS cf. fertile, NZS controls Very low Nanassy and Carrell (2011b) No Bisulphite sequencing ND 10 Age ND; abnormal P1/P2 ratio 10 Age ND; NZS No significant DNAm differences in the selected genes between samples with an abnormal P1/P2 ratio and NZS controls Very low Li et al. (2006) No Bisulphite sequencing Samples washed in PBS 4 Age ND; proven fertility N/A N/A DAZ1 promoter methylated in leukocytes but unmethylated in sperm Very low assoc., association; AsZS, asthenozoospermia; Bis-PCR, bisulphite-specific PCR; DMR, differentially methylated region; DNAm, DNA methylation levels; HPLC, high-performance liquid chromatography; HS, hypospermatogenesis; ‘5-mC immuno’, immunostaining for 5-methylCytosine; ND, not disclosed; NZS, normozoospermic; OAT, oligoasthenoteratozoospermia; OZS, oligozoospermia; P1/P2, protamine 1/protamine 2; ROS, reactive oxygen species; TLC, thin-layer chromatography; TZS, teratozoospermia; ↑, increased or hypermethylation; ↓, decreased or hypomethylation. ALS2CR12, amyotrophic lateral sclerosis 2 (juvenile) chromosome region, candidate 12; ALDH3B2, aldehyde dehydrogenase family 3 member B2; PRICKLE2, prickle planar cell polarity protein 2; PTGIR, prostaglandin I2 receptor; UBE2G2, ubiquitin-conjugating enzyme E2 G2; LINE-1, long interspersed nuclear element 1; MTHFR, methylenetetrahydrofolate reductase; MEST, mesoderm-specific transcript; BRCA, breast cancer; PRRC2A, proline-rich coiled-coil 2A; ANXA2, annexin A2; MAPK8IP3, mitogen-activated protein kinase 8 interacting protein 3; GAA, glucosidase alpha, acid; H19, imprinted maternally expressed transcript; SNRPN, small nuclear ribonucleoprotein polypeptide N; KCNJ5, potassium voltage-gated channel subfamily J member 5; MLPH, melanophilin; SMC1B, structural maintenance of chromosomes 1B; RPS6KA2, ribosomal protein S6 kinase A2; APCS, amyloid P component, serum; JAM3, junctional adhesion molecule 3; NCAPD3, non-SMC condensin II complex subunit D3; ANK2, ankyrin 2; HRAS, HRas proto-oncogene, GTPase; NTF3, neurotrophin 3; MT1A, metallothionein 1A; PAX8, paired box 8; DIRAS3, DIRAS family GTPase 3; PLAGL1, PLAG1 like zinc finger 1; SFN, stratifin; SAT2CHRM1, spermidine/spermine N1-acetyltransferase family member 2; BOLL, boule homolog, RNA-binding protein (member of the DAZ gene family); DDX4, DEAD-box helicase 4; HORMAD1, HORMA domain containing 1; MAEL, maelstrom spermatogenic transposon silencer; HSPA1L, heat shock protein family A (Hsp70) member 1 like; KCNQ1OT1, KCNQ1 opposite strand/antisense transcript 1; CRISP2, cysteine rich secretory protein 2; IGF2, insulin like growth factor 2; PEG3, paternally expressed 3; LIT1 (more commonly known as KCNQ1OT1), KCNQ1 opposite strand/antisense transcript 1; ZAC (more commonly known as PLAGL1), PLAG1 like zinc finger 1; DAZL, deleted in azoospermia like (member of the DAZ gene family); DAZ, deleted in azoospermia (group of 4 genes on chromosome Y: DAZ1-DAZ4, members of the DAZ gene family); MLH1, mutL homolog 1; VDAC2, voltage dependent anion channel 2; GNAS, GNAS complex locus; FAM50B, family with sequence similarity 50 member B; P16 (more commonly known as CDKN2A), cyclin dependent kinase inhibitor 2A; RHOX, reproductive homeobox on the X chromosome; CTCFL, CCCTC-binding factor like; CREM, CAMP responsive element modulator; DAZ1, deleted in azoospermia 1 (member of the DAZ gene family). Open in new tab Table I Studies assessing the effect of DNA methylation in human sperm on fertility and semen parameters. Author, year . Imprinting studied? . Method of assaying DNA methylation . Initial processing . N (cases) . Age in years (SD); description of cases . N (controls) . Age in years; description of controls . Main results . Study quality . Laqqan et al. (2017b) No 450K array; targeted bisulphite sequencing Gradient (50%) centrifugation. Medium ND 70 35.89 (±6.03); 15 for screening and 55 for validation 71 33.98 (±5.53); 15 for screening and 56 for validation Significant assoc. between subfertility and altered DNAm at CpGs related to ALS2CR12, ALDH3B2, PRICKLE2 and PTGIR High Laqqan et al. (2017c) No 450K array; targeted bisulphite sequencing Somatic cell lysis and PureSperm gradient (45% and 90%) 101 36.4 (±3.24); 20 + 81 (screening + validation) OZS 66 36.4 (±3.24); 20 + 44 (screening + validation) proven fertility Assoc. between OZS and DNAm of UBE2G2 and cg04807108 High Tian et al. (2015) Yes Methylation-specific PCR Samples washed in PBS and sperm wash buffer 118 31.4 (±5.1); males under evaluation for subfertility None N/A LINE-1 DNAm negatively correlated with sperm motility High Wu et al. (2010) No Targeted bisulphite Sanger sequencing Swim-up 94 29.04 (±4.52); idiopathic subfertility 54 29.52 (±3.72); proven fertility Significantly ↑ DNAm of the MTHFR promoter in subfertile males cf. fertile controls High Poplinski et al. (2010) Yes Targeted bisulphite Sanger sequencing Swim-up 148 35.5 (32–41); subfertile 33 37.0 (31.5–40); NZS; infertility attributed to female side MEST ↓ DNAm significantly assoc. with OZS in subfertile males High Kabartan et al. (2019) No Methylation-specific PCR Somatic cell lysis 73 31.95 (±5.5); OZS, infertile 20 27.2 (±2.6); NZS, proven fertility No assoc. between DNA fragmentation and DNAm of BRCA1 or BRCA2 in OZS males Moderate Tang et al. (2018) Yes Bis-PCR Percoll gradient (40% and 80%) 135 29.07 (±4.25); subfertile 59 29.08 (±3.84); proven fertility Aberrant DNAm in imprinted genes of infertile men Moderate Laqqan et al. (2018) No 450K array followed by Bis-PCR-Seq Somatic cell lysis 50 38.7 (±6.9); subfertile males 28 38.5 (±5.6); proven fertility Assoc. between subfertility and DNAm of PRRC2A, ANXA2, MAPK8IP3 and GAA Moderate Camprubi et al. (2017) No 450K array Somatic cell lysis 19 25.95 (±1.41); proven fertility N/A N/A Promoter CpGs relatively ↓ DNAm in spermatozoa. Identified 94 genes that appear resistant to demethylation Moderate Dong et al. (2017) Yes Bisulphite pyrosequencing Percoll gradient (45% and 90%) 155 31.13–32.17; 48 OZS, 52 AsZS, 55 TZS 50 32.22 (±3.59); NZS In smokers, subfertility was assoc. with ↓ DNAm of H19 in OZS and ↑ DNAm of SNRPN-ICR in AsZS and TZS Moderate Laqqan et al. (2017a) No 450K array; targeted bisulphite sequencing Somatic cell lysis 72 37.4 (±6.1); 15 + 57 (screening and validation) from subfertile men (unable to conceive for at least 10 years) 36 38.5 (±5.2); 15 + 21 (screening and validation) men with proved fertility Assoc. between subfertility and DNAm in KCNJ5, MLPH and SMC1B Moderate Camprubi et al. (2016) No 450K array Somatic cell lysis 42 38.36 (±5.31); subfertile males 19 25.95 (±4.80); proven fertility Assoc. between age and DNAm of RPS6KA2; OZS and DNAm of APCS; chromosome abnormalities and DNAm of JAM3/NCAPD3 and between fecundity and DNAm of ANK2 Moderate Jenkins et al. (2016a) No 450K array ISolate gradient 94 32.0–36.32; classified according to semen parameters N/A N/A No DNAm alterations associated with TZS; abnormal semen had increased DNAm variability Moderate Jenkins et al. (2015) No 450K array ISolate gradient (35% and 90%) 20 35.69 (±2.0); subfertile N/A N/A Higher DNAm variability in low-quality sperm Moderate Botezatu et al. (2014) No Methylation-specific PCR ND 27 26–41; OAT 11 24–37; NZS ↓ DNAm of MTHFR and SNRPN in samples with low sperm motility. No significant trend for other semen parameters Moderate Schutte et al. (2013) No 27K array followed by targeted bisulphite pyrosequencing Swim-up procedure and PureSperm gradient (40% and 80%) 38 27–54; males under evaluation for subfertility (longitudinal) 4 Age ND; NZS 27K array data showed clustering by sperm count; no candidate genes showed significant difference Moderate Klaver et al. (2013) Yes Bisulphite pyrosequencing Swim-up 212 Age ND; males under evaluation for subfertility 41 34.5 (±7.2); 31 NZS samples. Longitudinal MEST meth assoc. with OZS. DNAm levels at selected genes in NZS males were stable for up to 951 days Moderate Barzideh et al. (2013) No HPLC for DNAm; flow cytometry Percoll gradient 16 20–23; NZS N/A N/A DNAm ↓ in high-density (high quality) portion of spermatozoa than in low-density portion Moderate Pacheco et al. (2011) No 27K array Percoll gradient (50%) 21 Age ND; males under evaluation for subfertility None N/A Assoc. between 9189 CpGs and low sperm motility Moderate Marques et al. (2008) Yes Bisulphite sequencing (Sanger) SupraSperm gradient (55%, 80% and 90%) and swim-up 20 Age ND; 6 OZS; 14 AsZS; all abnormal morphology ND Age ND; NZS H19 and MEST abnormally imprinted in severely OZS group Moderate Houshdaran et al. (2007) No MethyLight; Illumina bead array of 1421 CpG sites ISolate gradient 65 22–49; partners of females undergoing fertility investigation None N/A Assoc. between sperm parameters and DNAm of HRAS, NTF3, MT1A, PAX8, DIRAS3, PLAGL1, SFN, SAT2CHRM1 and MEST Moderate Marques et al. (2004) Yes Bisulphite sequencing Gradient centrifugation and swim-up 96 Age ND; OZS 27 Age ND; NZS H19 abnormally imprinted in OZS group Moderate Cheng et al. (2017) No Agilent custom 1M Promoter-CpG island microarray; targeted bisulphite pyrosequencing Enzymatic digestion and countercurrent centrifugal elutriation 17 33.2 (±0.5); 1 sample from patient with moderate HS; candidate genes in 16 patients 9 ND; 9 NZS males with obstructive AZS, 1 selected for whole genome methylation profiling HS assoc. with DNAm of BOLL, DDX4, HORMAD1 and MAEL Low Du et al. (2016) No Liquid hybridization (promoter capture) capture-based bisulphite sequencing Percoll gradient (47.5%, 57%, 76% and 95%) 7 31.7 (±4.0); AsZS 8 29.8 (± 3.2); NZS No significant assoc. between DNAm or DNAm variability and AsZS Low Jenkins et al. (2016b) No 450K array Somatic cell lysis 29 27.55 (±0.71); subfertile 27 29.74 (±0.71); proven fertility DNAm of HSPA1L and HSPA1B significantly assoc. with pregnancy rates Low Louie et al. (2016) Yes Methylation-specific PCR Swim-up for high concentration samples. Handpicked spermatozoa for low concentration samples 44 32.9–35.7; OAT ND 34.1 (±2.4); NZS No significant association between DNAm in selected imprinted genes, severe OAT and the MTHFR C677T SNP Low Cassuto et al. (2016) No 5-mC immuno Sperm separation medium gradient (45% and 90%) 10 37–50; 448 S6 spermatozoa compared to 428 S0 spermatozoa N/A N/A ↓ DNAm in morphologically higher scoring spermatozoa compared to morphologically lower scoring spermatozoa Low Laurentino et al. (2015) Yes Bisulphite pyroseq. and deep targeted bis-seq Swim-up 26 34.0–39.33; males with abnormal sperm parameters 19 33.68 (±1.58); proven fertility Significantly ↑ variation in the DNAm values of the maternally methylated gene KCNQ1OT1 in samples with abnormal sperm parameters Low Zhou et al. (2015) No Methylation-specific PCR and bisulphite sequencing PCR Percoll gradient (50%) 48 30.13 (±5.8); AsZS 42 29.0 (±4.8); NZS No DNAm of the CRISP2 promoter; expression likely regulated by miR-27b Low Urdinguio et al. (2015) No 450K array; targeted bisulphite pyrosequencing; immunostaining for DNAm PureSperm gradient (65% and 90%) 7 30–55; 29 subfertile NZS (7 included in array analysis) 5 22–49; 17 NZS fertile males (5 included in array analysis) A number of CpG sites significantly differentially methylated between subfertile cases and fertile controls Low Tavalaee et al. (2015) No 5-mC immuno Samples washed in PBS 23 31.3 (±4.3); grades II and III varicocelectomy (longitudinal) None N/A Varicocele surgery associated with improved sperm parameters, but no significant change in DNAm Low Montjean et al. (2015) Yes Global DNAm assayed by immunostaining for DNAm, targeted bisulphite sequencing Percoll gradient 30 38.3 (±6); OAT 62 38.5 (±5.3); NZS Significant association between H19 DNAm and sperm parameters Low Rotondo et al. (2013) Yes Methylation-specific PCR and combined bisulphite restriction analysis PureSperm gradient (40% and 80%) 10 36.1 (±1.2); MTHFR promoter hypermethylation; 5 with normal and 5 with abnormal semen parameters 10 Age ND; normal MTHFR promoter meth; 5 with normal and 5 with abnormal semen parameters For NZS samples, H19 meth was assoc. with MTHFR DNAm Low Sato et al. (2011) Yes COBRA and methylation-specific PCR with Luminex technology Swim-up 128 Age ND; OZS 209 Age ND; NZS Correlation between DNAm analysis methods. Higher levels of imprinting errors in OZS Low Boissonnas et al. (2010) Yes Targeted pyrosequencing Percoll gradient (45%, 60% and 90%) 41 36.6 (±5.7); divided into TZS, OZS and AsZS groups 17 Age ND; NZS ↓ DNAm of various CpG positions in the H19–IGF2 DMR in cases cf. controls Low Hammoud et al. (2010) Yes Targeted Sanger sequencing ND 20 Age ND; 10 with abnormal P1/P2 ratio and 10 with OZS ND Age ND; proven fertility ↑ DNAm of LIT1, SNRPN, MEST, ZAC, PEG3 in OZS and abnormal P1/P2 ratio samples cf. NZS samples Low Marques et al. (2010) Yes Bisulphite sequencing Testicular biopsies washed in sperm preparation medium and somatic cells removed with erythrocyte lysing buffer 24 22–44; AZS (sperm obtained from testicular biopsies) ND Age ND; NZS H19 DNAm errors identified in AZS samples obtained via testicular biopsy Low Navarro-Costa et al. (2010) No Bisulphite sequencing PureSperm gradient (40% and 80%) 5 39.4 (±7.2); OZS 5 39.2 (±7.3); NZS ↑ DNAm of the DAZL (but not DAZ) promoter more prevalent in OZS samples compared to NZS controls Low Aoki et al. (2006) No Quantitative immune-fluorescence microscopy ND 195 Age ND; subfertile (3 groups based on P1/P2 ratios) None N/A No significant differences in DNAm between P1/P2 groups Low Sujit et al. (2018) No 450K array; targeted bisulphite sequencing Somatic cell lysis for cases and swim-up procedure for controls 38 Age ND; subfertile; OZS 26 Age ND; NZS 1436 probes had ↑ DNAm and 244 had ↓ DNAm in subfertile OZS cases relative to fertile NZS controls Very low Gunes et al. (2018) No Methylation-specific PCR Somatic cell lysis 10 Age ND; OAT 29 Age ND; NZS In OAT men, MLH1 was ↑ DNAm. Seminal ROS levels were assoc. with MLH1 DNAm Very low Olszewska et al. (2017) No TLC and immuno-fluorescence Samples washed in F10 medium 9 30.7; subfertile carriers of chromosomal abnormalities 14 28.5; healthy volunteers Chromosomal abnormalities not assoc. with DNAm Very low Marques et al. (2017) Yes Methylation-specific PCR Micromanipulation from testicular biopsy 15 Age ND; AsZS ND Age ND; NZS males unable to produce semen due to spinal cord injuries No significant association between H19 and MEST DNAm and AsZS Very low Nasri et al. (2017) Yes Combined bisulphite restriction analysis PureSperm gradient (40% and 80%) 23 Age ND; samples with abnormal sperm parameters 11 Age ND; NZS No significant difference between H19 DNAm and sperm parameters Very low Xu et al. (2016a) No Methylation-specific PCR ND 27 28; AsZS 25 28; NZS AsZS significantly assoc. with abnormal DNAm of the VDAC2 promoter Very low Xu et al. (2016b) Yes Targeted DNAm quantification using the MassARRAY EpiTYPER Percoll gradient (50%) 46 31.95 (±3.1); AZS 49 32.16 (±3.26); NZS ↓ DNAm in MEST, GNAS, FAM50B, H19, LINE‐1 and P16 in AsZS males Very low Uppangala et al. (2016) No 5-mC immuno Swim-up 19 Age ND; proven fertility. 1, 3, 5 and 7 days of abstinence N/A N/A No significant association between DNAm and length of ejaculatory abstinence Very low Bahreinian et al. (2015) No Flow cytometric immunodetection for DNAm Samples washed in PBS 44 31.1 (±0.6); patients with grades II and III varicocele 15 37.8 (±2.0); proven fertility Varicocele associated with ↓ DNAm and poorer sperm parameters Very low Kuhtz et al. (2014) Yes Bisulphite sequencing Swim-up and PureSperm gradient (40% and 80%) 27 Age ND; OAT 27 Age ND; NZS No significant differences in the occurrence of abnormal methylation imprints between sperm with and without morphological abnormalities Very low Richardson et al. (2014) No Bisulphite pyrosequencing Swim-up 95 32–39; abnormal semen parameters 45 34–40; NZS The RHOX homeobox gene cluster is regulated by DNAm and RHOX gene cluster ↑ DNAm is significantly assoc. with semen abnormalities Very low Camprubi et al. (2013) Yes Bisulphite pyrosequencing Swim-up 6 Age ND; subfertile males showing hypomethylation of the H19-ICR locus None N/A No significant association between CTCFL mutations and H19-ICR sperm DNAm Very low Li et al. (2013) Yes Methylation-specific PCR Gradient (45%, 60% and 90%) gradient centrifugation. Both Percoll and PureSperm mentioned 40 Age ND; 20 OZS and 20 AsZS men 20 Age ND; NZS, proven fertility No significant association between H19 and DAZL DNAm and AsZS Very low Minor et al. (2011) Yes Bisulphite sequencing Swim-up 18 37.8 (±6.9); AZS (sperm obtained from testicular biopsies) 9 34.1 (±2.4); proven fertility (from ejaculate) ↓ DNAm of H19 in AZS males Very low Nanassy and Carrell (2011a) No Bisulphite pyrosequencing ND 92 28–33; 60 with abnormal P1/P2 ratio, 32 with OZS 40 32–33; NZS, proven fertility, normal P1/P2 ratio ↑ Freq. CREM promoter methylation in males with abnormal P1/P2 ratio and in OZS cf. fertile, NZS controls Very low Nanassy and Carrell (2011b) No Bisulphite sequencing ND 10 Age ND; abnormal P1/P2 ratio 10 Age ND; NZS No significant DNAm differences in the selected genes between samples with an abnormal P1/P2 ratio and NZS controls Very low Li et al. (2006) No Bisulphite sequencing Samples washed in PBS 4 Age ND; proven fertility N/A N/A DAZ1 promoter methylated in leukocytes but unmethylated in sperm Very low Author, year . Imprinting studied? . Method of assaying DNA methylation . Initial processing . N (cases) . Age in years (SD); description of cases . N (controls) . Age in years; description of controls . Main results . Study quality . Laqqan et al. (2017b) No 450K array; targeted bisulphite sequencing Gradient (50%) centrifugation. Medium ND 70 35.89 (±6.03); 15 for screening and 55 for validation 71 33.98 (±5.53); 15 for screening and 56 for validation Significant assoc. between subfertility and altered DNAm at CpGs related to ALS2CR12, ALDH3B2, PRICKLE2 and PTGIR High Laqqan et al. (2017c) No 450K array; targeted bisulphite sequencing Somatic cell lysis and PureSperm gradient (45% and 90%) 101 36.4 (±3.24); 20 + 81 (screening + validation) OZS 66 36.4 (±3.24); 20 + 44 (screening + validation) proven fertility Assoc. between OZS and DNAm of UBE2G2 and cg04807108 High Tian et al. (2015) Yes Methylation-specific PCR Samples washed in PBS and sperm wash buffer 118 31.4 (±5.1); males under evaluation for subfertility None N/A LINE-1 DNAm negatively correlated with sperm motility High Wu et al. (2010) No Targeted bisulphite Sanger sequencing Swim-up 94 29.04 (±4.52); idiopathic subfertility 54 29.52 (±3.72); proven fertility Significantly ↑ DNAm of the MTHFR promoter in subfertile males cf. fertile controls High Poplinski et al. (2010) Yes Targeted bisulphite Sanger sequencing Swim-up 148 35.5 (32–41); subfertile 33 37.0 (31.5–40); NZS; infertility attributed to female side MEST ↓ DNAm significantly assoc. with OZS in subfertile males High Kabartan et al. (2019) No Methylation-specific PCR Somatic cell lysis 73 31.95 (±5.5); OZS, infertile 20 27.2 (±2.6); NZS, proven fertility No assoc. between DNA fragmentation and DNAm of BRCA1 or BRCA2 in OZS males Moderate Tang et al. (2018) Yes Bis-PCR Percoll gradient (40% and 80%) 135 29.07 (±4.25); subfertile 59 29.08 (±3.84); proven fertility Aberrant DNAm in imprinted genes of infertile men Moderate Laqqan et al. (2018) No 450K array followed by Bis-PCR-Seq Somatic cell lysis 50 38.7 (±6.9); subfertile males 28 38.5 (±5.6); proven fertility Assoc. between subfertility and DNAm of PRRC2A, ANXA2, MAPK8IP3 and GAA Moderate Camprubi et al. (2017) No 450K array Somatic cell lysis 19 25.95 (±1.41); proven fertility N/A N/A Promoter CpGs relatively ↓ DNAm in spermatozoa. Identified 94 genes that appear resistant to demethylation Moderate Dong et al. (2017) Yes Bisulphite pyrosequencing Percoll gradient (45% and 90%) 155 31.13–32.17; 48 OZS, 52 AsZS, 55 TZS 50 32.22 (±3.59); NZS In smokers, subfertility was assoc. with ↓ DNAm of H19 in OZS and ↑ DNAm of SNRPN-ICR in AsZS and TZS Moderate Laqqan et al. (2017a) No 450K array; targeted bisulphite sequencing Somatic cell lysis 72 37.4 (±6.1); 15 + 57 (screening and validation) from subfertile men (unable to conceive for at least 10 years) 36 38.5 (±5.2); 15 + 21 (screening and validation) men with proved fertility Assoc. between subfertility and DNAm in KCNJ5, MLPH and SMC1B Moderate Camprubi et al. (2016) No 450K array Somatic cell lysis 42 38.36 (±5.31); subfertile males 19 25.95 (±4.80); proven fertility Assoc. between age and DNAm of RPS6KA2; OZS and DNAm of APCS; chromosome abnormalities and DNAm of JAM3/NCAPD3 and between fecundity and DNAm of ANK2 Moderate Jenkins et al. (2016a) No 450K array ISolate gradient 94 32.0–36.32; classified according to semen parameters N/A N/A No DNAm alterations associated with TZS; abnormal semen had increased DNAm variability Moderate Jenkins et al. (2015) No 450K array ISolate gradient (35% and 90%) 20 35.69 (±2.0); subfertile N/A N/A Higher DNAm variability in low-quality sperm Moderate Botezatu et al. (2014) No Methylation-specific PCR ND 27 26–41; OAT 11 24–37; NZS ↓ DNAm of MTHFR and SNRPN in samples with low sperm motility. No significant trend for other semen parameters Moderate Schutte et al. (2013) No 27K array followed by targeted bisulphite pyrosequencing Swim-up procedure and PureSperm gradient (40% and 80%) 38 27–54; males under evaluation for subfertility (longitudinal) 4 Age ND; NZS 27K array data showed clustering by sperm count; no candidate genes showed significant difference Moderate Klaver et al. (2013) Yes Bisulphite pyrosequencing Swim-up 212 Age ND; males under evaluation for subfertility 41 34.5 (±7.2); 31 NZS samples. Longitudinal MEST meth assoc. with OZS. DNAm levels at selected genes in NZS males were stable for up to 951 days Moderate Barzideh et al. (2013) No HPLC for DNAm; flow cytometry Percoll gradient 16 20–23; NZS N/A N/A DNAm ↓ in high-density (high quality) portion of spermatozoa than in low-density portion Moderate Pacheco et al. (2011) No 27K array Percoll gradient (50%) 21 Age ND; males under evaluation for subfertility None N/A Assoc. between 9189 CpGs and low sperm motility Moderate Marques et al. (2008) Yes Bisulphite sequencing (Sanger) SupraSperm gradient (55%, 80% and 90%) and swim-up 20 Age ND; 6 OZS; 14 AsZS; all abnormal morphology ND Age ND; NZS H19 and MEST abnormally imprinted in severely OZS group Moderate Houshdaran et al. (2007) No MethyLight; Illumina bead array of 1421 CpG sites ISolate gradient 65 22–49; partners of females undergoing fertility investigation None N/A Assoc. between sperm parameters and DNAm of HRAS, NTF3, MT1A, PAX8, DIRAS3, PLAGL1, SFN, SAT2CHRM1 and MEST Moderate Marques et al. (2004) Yes Bisulphite sequencing Gradient centrifugation and swim-up 96 Age ND; OZS 27 Age ND; NZS H19 abnormally imprinted in OZS group Moderate Cheng et al. (2017) No Agilent custom 1M Promoter-CpG island microarray; targeted bisulphite pyrosequencing Enzymatic digestion and countercurrent centrifugal elutriation 17 33.2 (±0.5); 1 sample from patient with moderate HS; candidate genes in 16 patients 9 ND; 9 NZS males with obstructive AZS, 1 selected for whole genome methylation profiling HS assoc. with DNAm of BOLL, DDX4, HORMAD1 and MAEL Low Du et al. (2016) No Liquid hybridization (promoter capture) capture-based bisulphite sequencing Percoll gradient (47.5%, 57%, 76% and 95%) 7 31.7 (±4.0); AsZS 8 29.8 (± 3.2); NZS No significant assoc. between DNAm or DNAm variability and AsZS Low Jenkins et al. (2016b) No 450K array Somatic cell lysis 29 27.55 (±0.71); subfertile 27 29.74 (±0.71); proven fertility DNAm of HSPA1L and HSPA1B significantly assoc. with pregnancy rates Low Louie et al. (2016) Yes Methylation-specific PCR Swim-up for high concentration samples. Handpicked spermatozoa for low concentration samples 44 32.9–35.7; OAT ND 34.1 (±2.4); NZS No significant association between DNAm in selected imprinted genes, severe OAT and the MTHFR C677T SNP Low Cassuto et al. (2016) No 5-mC immuno Sperm separation medium gradient (45% and 90%) 10 37–50; 448 S6 spermatozoa compared to 428 S0 spermatozoa N/A N/A ↓ DNAm in morphologically higher scoring spermatozoa compared to morphologically lower scoring spermatozoa Low Laurentino et al. (2015) Yes Bisulphite pyroseq. and deep targeted bis-seq Swim-up 26 34.0–39.33; males with abnormal sperm parameters 19 33.68 (±1.58); proven fertility Significantly ↑ variation in the DNAm values of the maternally methylated gene KCNQ1OT1 in samples with abnormal sperm parameters Low Zhou et al. (2015) No Methylation-specific PCR and bisulphite sequencing PCR Percoll gradient (50%) 48 30.13 (±5.8); AsZS 42 29.0 (±4.8); NZS No DNAm of the CRISP2 promoter; expression likely regulated by miR-27b Low Urdinguio et al. (2015) No 450K array; targeted bisulphite pyrosequencing; immunostaining for DNAm PureSperm gradient (65% and 90%) 7 30–55; 29 subfertile NZS (7 included in array analysis) 5 22–49; 17 NZS fertile males (5 included in array analysis) A number of CpG sites significantly differentially methylated between subfertile cases and fertile controls Low Tavalaee et al. (2015) No 5-mC immuno Samples washed in PBS 23 31.3 (±4.3); grades II and III varicocelectomy (longitudinal) None N/A Varicocele surgery associated with improved sperm parameters, but no significant change in DNAm Low Montjean et al. (2015) Yes Global DNAm assayed by immunostaining for DNAm, targeted bisulphite sequencing Percoll gradient 30 38.3 (±6); OAT 62 38.5 (±5.3); NZS Significant association between H19 DNAm and sperm parameters Low Rotondo et al. (2013) Yes Methylation-specific PCR and combined bisulphite restriction analysis PureSperm gradient (40% and 80%) 10 36.1 (±1.2); MTHFR promoter hypermethylation; 5 with normal and 5 with abnormal semen parameters 10 Age ND; normal MTHFR promoter meth; 5 with normal and 5 with abnormal semen parameters For NZS samples, H19 meth was assoc. with MTHFR DNAm Low Sato et al. (2011) Yes COBRA and methylation-specific PCR with Luminex technology Swim-up 128 Age ND; OZS 209 Age ND; NZS Correlation between DNAm analysis methods. Higher levels of imprinting errors in OZS Low Boissonnas et al. (2010) Yes Targeted pyrosequencing Percoll gradient (45%, 60% and 90%) 41 36.6 (±5.7); divided into TZS, OZS and AsZS groups 17 Age ND; NZS ↓ DNAm of various CpG positions in the H19–IGF2 DMR in cases cf. controls Low Hammoud et al. (2010) Yes Targeted Sanger sequencing ND 20 Age ND; 10 with abnormal P1/P2 ratio and 10 with OZS ND Age ND; proven fertility ↑ DNAm of LIT1, SNRPN, MEST, ZAC, PEG3 in OZS and abnormal P1/P2 ratio samples cf. NZS samples Low Marques et al. (2010) Yes Bisulphite sequencing Testicular biopsies washed in sperm preparation medium and somatic cells removed with erythrocyte lysing buffer 24 22–44; AZS (sperm obtained from testicular biopsies) ND Age ND; NZS H19 DNAm errors identified in AZS samples obtained via testicular biopsy Low Navarro-Costa et al. (2010) No Bisulphite sequencing PureSperm gradient (40% and 80%) 5 39.4 (±7.2); OZS 5 39.2 (±7.3); NZS ↑ DNAm of the DAZL (but not DAZ) promoter more prevalent in OZS samples compared to NZS controls Low Aoki et al. (2006) No Quantitative immune-fluorescence microscopy ND 195 Age ND; subfertile (3 groups based on P1/P2 ratios) None N/A No significant differences in DNAm between P1/P2 groups Low Sujit et al. (2018) No 450K array; targeted bisulphite sequencing Somatic cell lysis for cases and swim-up procedure for controls 38 Age ND; subfertile; OZS 26 Age ND; NZS 1436 probes had ↑ DNAm and 244 had ↓ DNAm in subfertile OZS cases relative to fertile NZS controls Very low Gunes et al. (2018) No Methylation-specific PCR Somatic cell lysis 10 Age ND; OAT 29 Age ND; NZS In OAT men, MLH1 was ↑ DNAm. Seminal ROS levels were assoc. with MLH1 DNAm Very low Olszewska et al. (2017) No TLC and immuno-fluorescence Samples washed in F10 medium 9 30.7; subfertile carriers of chromosomal abnormalities 14 28.5; healthy volunteers Chromosomal abnormalities not assoc. with DNAm Very low Marques et al. (2017) Yes Methylation-specific PCR Micromanipulation from testicular biopsy 15 Age ND; AsZS ND Age ND; NZS males unable to produce semen due to spinal cord injuries No significant association between H19 and MEST DNAm and AsZS Very low Nasri et al. (2017) Yes Combined bisulphite restriction analysis PureSperm gradient (40% and 80%) 23 Age ND; samples with abnormal sperm parameters 11 Age ND; NZS No significant difference between H19 DNAm and sperm parameters Very low Xu et al. (2016a) No Methylation-specific PCR ND 27 28; AsZS 25 28; NZS AsZS significantly assoc. with abnormal DNAm of the VDAC2 promoter Very low Xu et al. (2016b) Yes Targeted DNAm quantification using the MassARRAY EpiTYPER Percoll gradient (50%) 46 31.95 (±3.1); AZS 49 32.16 (±3.26); NZS ↓ DNAm in MEST, GNAS, FAM50B, H19, LINE‐1 and P16 in AsZS males Very low Uppangala et al. (2016) No 5-mC immuno Swim-up 19 Age ND; proven fertility. 1, 3, 5 and 7 days of abstinence N/A N/A No significant association between DNAm and length of ejaculatory abstinence Very low Bahreinian et al. (2015) No Flow cytometric immunodetection for DNAm Samples washed in PBS 44 31.1 (±0.6); patients with grades II and III varicocele 15 37.8 (±2.0); proven fertility Varicocele associated with ↓ DNAm and poorer sperm parameters Very low Kuhtz et al. (2014) Yes Bisulphite sequencing Swim-up and PureSperm gradient (40% and 80%) 27 Age ND; OAT 27 Age ND; NZS No significant differences in the occurrence of abnormal methylation imprints between sperm with and without morphological abnormalities Very low Richardson et al. (2014) No Bisulphite pyrosequencing Swim-up 95 32–39; abnormal semen parameters 45 34–40; NZS The RHOX homeobox gene cluster is regulated by DNAm and RHOX gene cluster ↑ DNAm is significantly assoc. with semen abnormalities Very low Camprubi et al. (2013) Yes Bisulphite pyrosequencing Swim-up 6 Age ND; subfertile males showing hypomethylation of the H19-ICR locus None N/A No significant association between CTCFL mutations and H19-ICR sperm DNAm Very low Li et al. (2013) Yes Methylation-specific PCR Gradient (45%, 60% and 90%) gradient centrifugation. Both Percoll and PureSperm mentioned 40 Age ND; 20 OZS and 20 AsZS men 20 Age ND; NZS, proven fertility No significant association between H19 and DAZL DNAm and AsZS Very low Minor et al. (2011) Yes Bisulphite sequencing Swim-up 18 37.8 (±6.9); AZS (sperm obtained from testicular biopsies) 9 34.1 (±2.4); proven fertility (from ejaculate) ↓ DNAm of H19 in AZS males Very low Nanassy and Carrell (2011a) No Bisulphite pyrosequencing ND 92 28–33; 60 with abnormal P1/P2 ratio, 32 with OZS 40 32–33; NZS, proven fertility, normal P1/P2 ratio ↑ Freq. CREM promoter methylation in males with abnormal P1/P2 ratio and in OZS cf. fertile, NZS controls Very low Nanassy and Carrell (2011b) No Bisulphite sequencing ND 10 Age ND; abnormal P1/P2 ratio 10 Age ND; NZS No significant DNAm differences in the selected genes between samples with an abnormal P1/P2 ratio and NZS controls Very low Li et al. (2006) No Bisulphite sequencing Samples washed in PBS 4 Age ND; proven fertility N/A N/A DAZ1 promoter methylated in leukocytes but unmethylated in sperm Very low assoc., association; AsZS, asthenozoospermia; Bis-PCR, bisulphite-specific PCR; DMR, differentially methylated region; DNAm, DNA methylation levels; HPLC, high-performance liquid chromatography; HS, hypospermatogenesis; ‘5-mC immuno’, immunostaining for 5-methylCytosine; ND, not disclosed; NZS, normozoospermic; OAT, oligoasthenoteratozoospermia; OZS, oligozoospermia; P1/P2, protamine 1/protamine 2; ROS, reactive oxygen species; TLC, thin-layer chromatography; TZS, teratozoospermia; ↑, increased or hypermethylation; ↓, decreased or hypomethylation. ALS2CR12, amyotrophic lateral sclerosis 2 (juvenile) chromosome region, candidate 12; ALDH3B2, aldehyde dehydrogenase family 3 member B2; PRICKLE2, prickle planar cell polarity protein 2; PTGIR, prostaglandin I2 receptor; UBE2G2, ubiquitin-conjugating enzyme E2 G2; LINE-1, long interspersed nuclear element 1; MTHFR, methylenetetrahydrofolate reductase; MEST, mesoderm-specific transcript; BRCA, breast cancer; PRRC2A, proline-rich coiled-coil 2A; ANXA2, annexin A2; MAPK8IP3, mitogen-activated protein kinase 8 interacting protein 3; GAA, glucosidase alpha, acid; H19, imprinted maternally expressed transcript; SNRPN, small nuclear ribonucleoprotein polypeptide N; KCNJ5, potassium voltage-gated channel subfamily J member 5; MLPH, melanophilin; SMC1B, structural maintenance of chromosomes 1B; RPS6KA2, ribosomal protein S6 kinase A2; APCS, amyloid P component, serum; JAM3, junctional adhesion molecule 3; NCAPD3, non-SMC condensin II complex subunit D3; ANK2, ankyrin 2; HRAS, HRas proto-oncogene, GTPase; NTF3, neurotrophin 3; MT1A, metallothionein 1A; PAX8, paired box 8; DIRAS3, DIRAS family GTPase 3; PLAGL1, PLAG1 like zinc finger 1; SFN, stratifin; SAT2CHRM1, spermidine/spermine N1-acetyltransferase family member 2; BOLL, boule homolog, RNA-binding protein (member of the DAZ gene family); DDX4, DEAD-box helicase 4; HORMAD1, HORMA domain containing 1; MAEL, maelstrom spermatogenic transposon silencer; HSPA1L, heat shock protein family A (Hsp70) member 1 like; KCNQ1OT1, KCNQ1 opposite strand/antisense transcript 1; CRISP2, cysteine rich secretory protein 2; IGF2, insulin like growth factor 2; PEG3, paternally expressed 3; LIT1 (more commonly known as KCNQ1OT1), KCNQ1 opposite strand/antisense transcript 1; ZAC (more commonly known as PLAGL1), PLAG1 like zinc finger 1; DAZL, deleted in azoospermia like (member of the DAZ gene family); DAZ, deleted in azoospermia (group of 4 genes on chromosome Y: DAZ1-DAZ4, members of the DAZ gene family); MLH1, mutL homolog 1; VDAC2, voltage dependent anion channel 2; GNAS, GNAS complex locus; FAM50B, family with sequence similarity 50 member B; P16 (more commonly known as CDKN2A), cyclin dependent kinase inhibitor 2A; RHOX, reproductive homeobox on the X chromosome; CTCFL, CCCTC-binding factor like; CREM, CAMP responsive element modulator; DAZ1, deleted in azoospermia 1 (member of the DAZ gene family). Open in new tab Table II Studies assessing the effect of DNA methylation in human sperm on pregnancy outcomes and offspring health. Author, year . Outcome studied . Method of assaying DNA methylation . Initial processing . N (cases) . Age in years (SD); description of cases . N (controls) . Age in years; description of controls . Main results . Study quality . Atsem et al. (2016) Offspring cord blood DNAm Bisulphite pyrosequencing Swim-up and PureSperm gradient (40% and 80%) 350 Age ND; 162 samples that had led to the birth of a child (cohort 1), 188 samples without pregnancy (replication cohort) None N/A DNAm of FOXK1 assoc. with paternal age in sperm and cord blood of offspring High Camprubi et al. (2012) ART success Bisulphite pyrosequencing Swim-up 107 36 (±5.5); subfertile 30 26 (±6.15); proven fertility No significant association between rates of imprinting errors and outcome of ART High El Hajj et al. (2011) ART success Bisulphite pyrosequencing PureSperm gradient (40% and 80%) 106 Age ND; subfertile 28 ND Significant association between DNAm of Alu and outcome of ART High Kobayashi et al. (2017) Pregnancy outcomes Combined bisulphite restriction analysis Swim-up 70 36.5 (±6.1); 40 males with moderate OZS and 30 males with severe OZS 151 35.4 (±5.4); NZS OZS assoc. with smoking and high consumption of carbonated drinks. Adverse pregnancy outcomes assoc. with sperm DNAm errors Moderate Rotondo et al. (2012) Spontaneous pregnancy loss Bisulphite-PCR sequencing PureSperm gradient (40% and 80%) 20 35 (±5.4); couples with recurrent spontaneous pregnancy loss 167 35.3 (±4.4); 20 males with proven fertility, 147 couples not affected by recurrent spontaneous pregnancy loss MTHFR ↑ DNAm more common in semen from couples affected by recurrent spontaneous pregnancy loss Moderate Tavalaee et al. (2009) ART success 5-mC immuno PureSperm gradient (40% and 80%) 61 36.2 (±5.2); undergoing ART None N/A DNAm negatively correlated with DNA fragmentation. No assoc. with ART success Moderate Denomme et al. (2017) Embryo quality 450K array on a subset of sperm samples (12 out of 40). Beta value distribution examined for histone-retained regions PureSperm gradient (45% and 90%), swim-up and somatic cell lysis 20 40.3 (±1.3); NZS (20 that led to ‘good’ embryogenesis) 20 42.0 (±1.1); NZS (20 that led to ‘poor’ embryogenesis) Significant assoc. between DNAm in 1634 CpG sites and embryo quality Low Feinberg et al. (2015) Autism in offspring CHARM 3.0 array; 30 samples also run on a 450K array No sperm selection method employed 44 27–51.2; fathers in a cohort enriched for autistic children None N/A 193 DMRs in paternal sperm associated with performance on the Autism Observational Scale for Infants at 12 months of age in offspring Low Aston et al. (2012) Abnormal embryogenesis 27K array followed by targeted bisulphite pyrosequencing ND 28 31.0 (±1.9); 15 males with abnormal chromatin packaging + 13 males with poor embryo outcomes 15 37.1(±2.0); NZS No significant CpG methylation differences identified Low Ankolkar et al. (2012) Spontaneous pregnancy loss Bisulphite Sanger sequencing HiSep solution gradient 26 35.4; couples with recurrent spontaneous miscarriage 26 31.3; proven fertility H19 ↓ DNAm in sperm from recurrent spontaneous miscarriage group Low Kobayashi et al. (2009) Conceptus DNAm Combined bisulphite PCR restriction analysis and bisulphite sequencing ND 78 Age ND; samples paired with aborted conceptuses conceived via ART 38 Age ND; samples paired with aborted conceptuses not conceived via ART Imprinting errors ↑ in OZS samples. Imprinting errors in sperm also evident in the resulting conceptuses in 7 out of 17 cases Low Benchaib et al. (2005) ART success 5-mC immuno PureSperm gradient (50%, 70% and 90%) 63 35.4 (±5.5); undergoing ART None N/A ↑ DNAm assoc. with ↑ rates of pregnancy and embryo development, not with fertilization rates Low Spinelli et al. (2019) No Bisulphite pyrosequencing ND 4 ND N/A N/A IDO1 ↑ in sperm cf. placenta; no DNAm change in oocytes Very low Aston et al. (2015) Embryo quality and ART success 450K array PureSperm gradient (45% and 90%); some analyses performed on whole ejaculates 127 Age ND; 54 with successful pregnancy outcomes and 72 with poor embryogenesis 54 ND; NZS, proven fertility DNAm may be predictive of embryo quality, but not of IVF outcome Very low Montjean et al. (2013) ART success Bisulphite targeted Sanger sequencing Percoll gradient (45% and 90%) 175 Age ND; OZS 119 Age ND; NZS Abnormal DNAm of selected genes more prevalent in OZS. No association between DNAm and outcome of ART Very low Al-Khtib et al. (2012) Oocyte DNAm Bisulphite pyrosequencing PureSperm gradient (50%, 70% and 90%) 5 Age ND; proven fertility None N/A ↑ DNAm of OCT4 and NANOG in sperm but ↓ DNAm in oocytes Very low Ibala-Romdhane et al. (2011) Oocyte DNAm and embryo quality Targeted bisulphite Sanger sequencing Gradient centrifugation (medium not specified) 11 Age ND; abnormal semen parameters (OZS, AsZS, TZS) ND Age ND; NZS, proven fertility Embryos with developmental failure more likely to have abnormal imprinting at H19 but all corresponding sperm had normal H19 imprinting Very low Geuns et al. (2007) Oocyte, embryo and offspring DNAm Methylation-specific PCR Percoll gradient (45% and 90%) ND ND None N/A Intergenic DMR of DLK1–GTL2 ↑ DNAm in sperm cf. somatic tissues (e.g. oocytes and preimplantation embryos) Very low Kobayashi et al. (2007) ART success Combined bisulphite PCR restriction analysis and bisulphite sequencing Swim-up method 18 27–50; OZS 79 Age ND; NZS DNAm errors ↑ in OZS. No significant association between imprinting errors and ART outcome Very low Benchaib et al. (2003) Pregnancy outcomes 5-mC immuno PureSperm gradient (50%, 70% and 90%) 23 35 (median); undergoing ART None N/A DNAm ↓ in TZS than non-TZS. No assoc. with pregnancy rate. Very Low Author, year . Outcome studied . Method of assaying DNA methylation . Initial processing . N (cases) . Age in years (SD); description of cases . N (controls) . Age in years; description of controls . Main results . Study quality . Atsem et al. (2016) Offspring cord blood DNAm Bisulphite pyrosequencing Swim-up and PureSperm gradient (40% and 80%) 350 Age ND; 162 samples that had led to the birth of a child (cohort 1), 188 samples without pregnancy (replication cohort) None N/A DNAm of FOXK1 assoc. with paternal age in sperm and cord blood of offspring High Camprubi et al. (2012) ART success Bisulphite pyrosequencing Swim-up 107 36 (±5.5); subfertile 30 26 (±6.15); proven fertility No significant association between rates of imprinting errors and outcome of ART High El Hajj et al. (2011) ART success Bisulphite pyrosequencing PureSperm gradient (40% and 80%) 106 Age ND; subfertile 28 ND Significant association between DNAm of Alu and outcome of ART High Kobayashi et al. (2017) Pregnancy outcomes Combined bisulphite restriction analysis Swim-up 70 36.5 (±6.1); 40 males with moderate OZS and 30 males with severe OZS 151 35.4 (±5.4); NZS OZS assoc. with smoking and high consumption of carbonated drinks. Adverse pregnancy outcomes assoc. with sperm DNAm errors Moderate Rotondo et al. (2012) Spontaneous pregnancy loss Bisulphite-PCR sequencing PureSperm gradient (40% and 80%) 20 35 (±5.4); couples with recurrent spontaneous pregnancy loss 167 35.3 (±4.4); 20 males with proven fertility, 147 couples not affected by recurrent spontaneous pregnancy loss MTHFR ↑ DNAm more common in semen from couples affected by recurrent spontaneous pregnancy loss Moderate Tavalaee et al. (2009) ART success 5-mC immuno PureSperm gradient (40% and 80%) 61 36.2 (±5.2); undergoing ART None N/A DNAm negatively correlated with DNA fragmentation. No assoc. with ART success Moderate Denomme et al. (2017) Embryo quality 450K array on a subset of sperm samples (12 out of 40). Beta value distribution examined for histone-retained regions PureSperm gradient (45% and 90%), swim-up and somatic cell lysis 20 40.3 (±1.3); NZS (20 that led to ‘good’ embryogenesis) 20 42.0 (±1.1); NZS (20 that led to ‘poor’ embryogenesis) Significant assoc. between DNAm in 1634 CpG sites and embryo quality Low Feinberg et al. (2015) Autism in offspring CHARM 3.0 array; 30 samples also run on a 450K array No sperm selection method employed 44 27–51.2; fathers in a cohort enriched for autistic children None N/A 193 DMRs in paternal sperm associated with performance on the Autism Observational Scale for Infants at 12 months of age in offspring Low Aston et al. (2012) Abnormal embryogenesis 27K array followed by targeted bisulphite pyrosequencing ND 28 31.0 (±1.9); 15 males with abnormal chromatin packaging + 13 males with poor embryo outcomes 15 37.1(±2.0); NZS No significant CpG methylation differences identified Low Ankolkar et al. (2012) Spontaneous pregnancy loss Bisulphite Sanger sequencing HiSep solution gradient 26 35.4; couples with recurrent spontaneous miscarriage 26 31.3; proven fertility H19 ↓ DNAm in sperm from recurrent spontaneous miscarriage group Low Kobayashi et al. (2009) Conceptus DNAm Combined bisulphite PCR restriction analysis and bisulphite sequencing ND 78 Age ND; samples paired with aborted conceptuses conceived via ART 38 Age ND; samples paired with aborted conceptuses not conceived via ART Imprinting errors ↑ in OZS samples. Imprinting errors in sperm also evident in the resulting conceptuses in 7 out of 17 cases Low Benchaib et al. (2005) ART success 5-mC immuno PureSperm gradient (50%, 70% and 90%) 63 35.4 (±5.5); undergoing ART None N/A ↑ DNAm assoc. with ↑ rates of pregnancy and embryo development, not with fertilization rates Low Spinelli et al. (2019) No Bisulphite pyrosequencing ND 4 ND N/A N/A IDO1 ↑ in sperm cf. placenta; no DNAm change in oocytes Very low Aston et al. (2015) Embryo quality and ART success 450K array PureSperm gradient (45% and 90%); some analyses performed on whole ejaculates 127 Age ND; 54 with successful pregnancy outcomes and 72 with poor embryogenesis 54 ND; NZS, proven fertility DNAm may be predictive of embryo quality, but not of IVF outcome Very low Montjean et al. (2013) ART success Bisulphite targeted Sanger sequencing Percoll gradient (45% and 90%) 175 Age ND; OZS 119 Age ND; NZS Abnormal DNAm of selected genes more prevalent in OZS. No association between DNAm and outcome of ART Very low Al-Khtib et al. (2012) Oocyte DNAm Bisulphite pyrosequencing PureSperm gradient (50%, 70% and 90%) 5 Age ND; proven fertility None N/A ↑ DNAm of OCT4 and NANOG in sperm but ↓ DNAm in oocytes Very low Ibala-Romdhane et al. (2011) Oocyte DNAm and embryo quality Targeted bisulphite Sanger sequencing Gradient centrifugation (medium not specified) 11 Age ND; abnormal semen parameters (OZS, AsZS, TZS) ND Age ND; NZS, proven fertility Embryos with developmental failure more likely to have abnormal imprinting at H19 but all corresponding sperm had normal H19 imprinting Very low Geuns et al. (2007) Oocyte, embryo and offspring DNAm Methylation-specific PCR Percoll gradient (45% and 90%) ND ND None N/A Intergenic DMR of DLK1–GTL2 ↑ DNAm in sperm cf. somatic tissues (e.g. oocytes and preimplantation embryos) Very low Kobayashi et al. (2007) ART success Combined bisulphite PCR restriction analysis and bisulphite sequencing Swim-up method 18 27–50; OZS 79 Age ND; NZS DNAm errors ↑ in OZS. No significant association between imprinting errors and ART outcome Very low Benchaib et al. (2003) Pregnancy outcomes 5-mC immuno PureSperm gradient (50%, 70% and 90%) 23 35 (median); undergoing ART None N/A DNAm ↓ in TZS than non-TZS. No assoc. with pregnancy rate. Very Low FOXK1, forkhead box K1; IDO1, indoleamine 2,3-dioxygenase 1; OCT4, octamer-binding transcription factor 4; NANOG, Nanog homeobox; DLK1, delta like non-canonical notch ligand 1; GTL2 (more commonly known as MEG3), maternally expressed 3. Open in new tab Table II Studies assessing the effect of DNA methylation in human sperm on pregnancy outcomes and offspring health. Author, year . Outcome studied . Method of assaying DNA methylation . Initial processing . N (cases) . Age in years (SD); description of cases . N (controls) . Age in years; description of controls . Main results . Study quality . Atsem et al. (2016) Offspring cord blood DNAm Bisulphite pyrosequencing Swim-up and PureSperm gradient (40% and 80%) 350 Age ND; 162 samples that had led to the birth of a child (cohort 1), 188 samples without pregnancy (replication cohort) None N/A DNAm of FOXK1 assoc. with paternal age in sperm and cord blood of offspring High Camprubi et al. (2012) ART success Bisulphite pyrosequencing Swim-up 107 36 (±5.5); subfertile 30 26 (±6.15); proven fertility No significant association between rates of imprinting errors and outcome of ART High El Hajj et al. (2011) ART success Bisulphite pyrosequencing PureSperm gradient (40% and 80%) 106 Age ND; subfertile 28 ND Significant association between DNAm of Alu and outcome of ART High Kobayashi et al. (2017) Pregnancy outcomes Combined bisulphite restriction analysis Swim-up 70 36.5 (±6.1); 40 males with moderate OZS and 30 males with severe OZS 151 35.4 (±5.4); NZS OZS assoc. with smoking and high consumption of carbonated drinks. Adverse pregnancy outcomes assoc. with sperm DNAm errors Moderate Rotondo et al. (2012) Spontaneous pregnancy loss Bisulphite-PCR sequencing PureSperm gradient (40% and 80%) 20 35 (±5.4); couples with recurrent spontaneous pregnancy loss 167 35.3 (±4.4); 20 males with proven fertility, 147 couples not affected by recurrent spontaneous pregnancy loss MTHFR ↑ DNAm more common in semen from couples affected by recurrent spontaneous pregnancy loss Moderate Tavalaee et al. (2009) ART success 5-mC immuno PureSperm gradient (40% and 80%) 61 36.2 (±5.2); undergoing ART None N/A DNAm negatively correlated with DNA fragmentation. No assoc. with ART success Moderate Denomme et al. (2017) Embryo quality 450K array on a subset of sperm samples (12 out of 40). Beta value distribution examined for histone-retained regions PureSperm gradient (45% and 90%), swim-up and somatic cell lysis 20 40.3 (±1.3); NZS (20 that led to ‘good’ embryogenesis) 20 42.0 (±1.1); NZS (20 that led to ‘poor’ embryogenesis) Significant assoc. between DNAm in 1634 CpG sites and embryo quality Low Feinberg et al. (2015) Autism in offspring CHARM 3.0 array; 30 samples also run on a 450K array No sperm selection method employed 44 27–51.2; fathers in a cohort enriched for autistic children None N/A 193 DMRs in paternal sperm associated with performance on the Autism Observational Scale for Infants at 12 months of age in offspring Low Aston et al. (2012) Abnormal embryogenesis 27K array followed by targeted bisulphite pyrosequencing ND 28 31.0 (±1.9); 15 males with abnormal chromatin packaging + 13 males with poor embryo outcomes 15 37.1(±2.0); NZS No significant CpG methylation differences identified Low Ankolkar et al. (2012) Spontaneous pregnancy loss Bisulphite Sanger sequencing HiSep solution gradient 26 35.4; couples with recurrent spontaneous miscarriage 26 31.3; proven fertility H19 ↓ DNAm in sperm from recurrent spontaneous miscarriage group Low Kobayashi et al. (2009) Conceptus DNAm Combined bisulphite PCR restriction analysis and bisulphite sequencing ND 78 Age ND; samples paired with aborted conceptuses conceived via ART 38 Age ND; samples paired with aborted conceptuses not conceived via ART Imprinting errors ↑ in OZS samples. Imprinting errors in sperm also evident in the resulting conceptuses in 7 out of 17 cases Low Benchaib et al. (2005) ART success 5-mC immuno PureSperm gradient (50%, 70% and 90%) 63 35.4 (±5.5); undergoing ART None N/A ↑ DNAm assoc. with ↑ rates of pregnancy and embryo development, not with fertilization rates Low Spinelli et al. (2019) No Bisulphite pyrosequencing ND 4 ND N/A N/A IDO1 ↑ in sperm cf. placenta; no DNAm change in oocytes Very low Aston et al. (2015) Embryo quality and ART success 450K array PureSperm gradient (45% and 90%); some analyses performed on whole ejaculates 127 Age ND; 54 with successful pregnancy outcomes and 72 with poor embryogenesis 54 ND; NZS, proven fertility DNAm may be predictive of embryo quality, but not of IVF outcome Very low Montjean et al. (2013) ART success Bisulphite targeted Sanger sequencing Percoll gradient (45% and 90%) 175 Age ND; OZS 119 Age ND; NZS Abnormal DNAm of selected genes more prevalent in OZS. No association between DNAm and outcome of ART Very low Al-Khtib et al. (2012) Oocyte DNAm Bisulphite pyrosequencing PureSperm gradient (50%, 70% and 90%) 5 Age ND; proven fertility None N/A ↑ DNAm of OCT4 and NANOG in sperm but ↓ DNAm in oocytes Very low Ibala-Romdhane et al. (2011) Oocyte DNAm and embryo quality Targeted bisulphite Sanger sequencing Gradient centrifugation (medium not specified) 11 Age ND; abnormal semen parameters (OZS, AsZS, TZS) ND Age ND; NZS, proven fertility Embryos with developmental failure more likely to have abnormal imprinting at H19 but all corresponding sperm had normal H19 imprinting Very low Geuns et al. (2007) Oocyte, embryo and offspring DNAm Methylation-specific PCR Percoll gradient (45% and 90%) ND ND None N/A Intergenic DMR of DLK1–GTL2 ↑ DNAm in sperm cf. somatic tissues (e.g. oocytes and preimplantation embryos) Very low Kobayashi et al. (2007) ART success Combined bisulphite PCR restriction analysis and bisulphite sequencing Swim-up method 18 27–50; OZS 79 Age ND; NZS DNAm errors ↑ in OZS. No significant association between imprinting errors and ART outcome Very low Benchaib et al. (2003) Pregnancy outcomes 5-mC immuno PureSperm gradient (50%, 70% and 90%) 23 35 (median); undergoing ART None N/A DNAm ↓ in TZS than non-TZS. No assoc. with pregnancy rate. Very Low Author, year . Outcome studied . Method of assaying DNA methylation . Initial processing . N (cases) . Age in years (SD); description of cases . N (controls) . Age in years; description of controls . Main results . Study quality . Atsem et al. (2016) Offspring cord blood DNAm Bisulphite pyrosequencing Swim-up and PureSperm gradient (40% and 80%) 350 Age ND; 162 samples that had led to the birth of a child (cohort 1), 188 samples without pregnancy (replication cohort) None N/A DNAm of FOXK1 assoc. with paternal age in sperm and cord blood of offspring High Camprubi et al. (2012) ART success Bisulphite pyrosequencing Swim-up 107 36 (±5.5); subfertile 30 26 (±6.15); proven fertility No significant association between rates of imprinting errors and outcome of ART High El Hajj et al. (2011) ART success Bisulphite pyrosequencing PureSperm gradient (40% and 80%) 106 Age ND; subfertile 28 ND Significant association between DNAm of Alu and outcome of ART High Kobayashi et al. (2017) Pregnancy outcomes Combined bisulphite restriction analysis Swim-up 70 36.5 (±6.1); 40 males with moderate OZS and 30 males with severe OZS 151 35.4 (±5.4); NZS OZS assoc. with smoking and high consumption of carbonated drinks. Adverse pregnancy outcomes assoc. with sperm DNAm errors Moderate Rotondo et al. (2012) Spontaneous pregnancy loss Bisulphite-PCR sequencing PureSperm gradient (40% and 80%) 20 35 (±5.4); couples with recurrent spontaneous pregnancy loss 167 35.3 (±4.4); 20 males with proven fertility, 147 couples not affected by recurrent spontaneous pregnancy loss MTHFR ↑ DNAm more common in semen from couples affected by recurrent spontaneous pregnancy loss Moderate Tavalaee et al. (2009) ART success 5-mC immuno PureSperm gradient (40% and 80%) 61 36.2 (±5.2); undergoing ART None N/A DNAm negatively correlated with DNA fragmentation. No assoc. with ART success Moderate Denomme et al. (2017) Embryo quality 450K array on a subset of sperm samples (12 out of 40). Beta value distribution examined for histone-retained regions PureSperm gradient (45% and 90%), swim-up and somatic cell lysis 20 40.3 (±1.3); NZS (20 that led to ‘good’ embryogenesis) 20 42.0 (±1.1); NZS (20 that led to ‘poor’ embryogenesis) Significant assoc. between DNAm in 1634 CpG sites and embryo quality Low Feinberg et al. (2015) Autism in offspring CHARM 3.0 array; 30 samples also run on a 450K array No sperm selection method employed 44 27–51.2; fathers in a cohort enriched for autistic children None N/A 193 DMRs in paternal sperm associated with performance on the Autism Observational Scale for Infants at 12 months of age in offspring Low Aston et al. (2012) Abnormal embryogenesis 27K array followed by targeted bisulphite pyrosequencing ND 28 31.0 (±1.9); 15 males with abnormal chromatin packaging + 13 males with poor embryo outcomes 15 37.1(±2.0); NZS No significant CpG methylation differences identified Low Ankolkar et al. (2012) Spontaneous pregnancy loss Bisulphite Sanger sequencing HiSep solution gradient 26 35.4; couples with recurrent spontaneous miscarriage 26 31.3; proven fertility H19 ↓ DNAm in sperm from recurrent spontaneous miscarriage group Low Kobayashi et al. (2009) Conceptus DNAm Combined bisulphite PCR restriction analysis and bisulphite sequencing ND 78 Age ND; samples paired with aborted conceptuses conceived via ART 38 Age ND; samples paired with aborted conceptuses not conceived via ART Imprinting errors ↑ in OZS samples. Imprinting errors in sperm also evident in the resulting conceptuses in 7 out of 17 cases Low Benchaib et al. (2005) ART success 5-mC immuno PureSperm gradient (50%, 70% and 90%) 63 35.4 (±5.5); undergoing ART None N/A ↑ DNAm assoc. with ↑ rates of pregnancy and embryo development, not with fertilization rates Low Spinelli et al. (2019) No Bisulphite pyrosequencing ND 4 ND N/A N/A IDO1 ↑ in sperm cf. placenta; no DNAm change in oocytes Very low Aston et al. (2015) Embryo quality and ART success 450K array PureSperm gradient (45% and 90%); some analyses performed on whole ejaculates 127 Age ND; 54 with successful pregnancy outcomes and 72 with poor embryogenesis 54 ND; NZS, proven fertility DNAm may be predictive of embryo quality, but not of IVF outcome Very low Montjean et al. (2013) ART success Bisulphite targeted Sanger sequencing Percoll gradient (45% and 90%) 175 Age ND; OZS 119 Age ND; NZS Abnormal DNAm of selected genes more prevalent in OZS. No association between DNAm and outcome of ART Very low Al-Khtib et al. (2012) Oocyte DNAm Bisulphite pyrosequencing PureSperm gradient (50%, 70% and 90%) 5 Age ND; proven fertility None N/A ↑ DNAm of OCT4 and NANOG in sperm but ↓ DNAm in oocytes Very low Ibala-Romdhane et al. (2011) Oocyte DNAm and embryo quality Targeted bisulphite Sanger sequencing Gradient centrifugation (medium not specified) 11 Age ND; abnormal semen parameters (OZS, AsZS, TZS) ND Age ND; NZS, proven fertility Embryos with developmental failure more likely to have abnormal imprinting at H19 but all corresponding sperm had normal H19 imprinting Very low Geuns et al. (2007) Oocyte, embryo and offspring DNAm Methylation-specific PCR Percoll gradient (45% and 90%) ND ND None N/A Intergenic DMR of DLK1–GTL2 ↑ DNAm in sperm cf. somatic tissues (e.g. oocytes and preimplantation embryos) Very low Kobayashi et al. (2007) ART success Combined bisulphite PCR restriction analysis and bisulphite sequencing Swim-up method 18 27–50; OZS 79 Age ND; NZS DNAm errors ↑ in OZS. No significant association between imprinting errors and ART outcome Very low Benchaib et al. (2003) Pregnancy outcomes 5-mC immuno PureSperm gradient (50%, 70% and 90%) 23 35 (median); undergoing ART None N/A DNAm ↓ in TZS than non-TZS. No assoc. with pregnancy rate. Very Low FOXK1, forkhead box K1; IDO1, indoleamine 2,3-dioxygenase 1; OCT4, octamer-binding transcription factor 4; NANOG, Nanog homeobox; DLK1, delta like non-canonical notch ligand 1; GTL2 (more commonly known as MEG3), maternally expressed 3. Open in new tab Table III Studies assessing the effect of environmental exposures on DNA methylation in human sperm. Author, year . Type of exposure . Method of assaying DNA methylation . Initial processing . N (cases) . Age in years (SD); description of cases . N (controls) . Age in years (SD); description of controls . Main results . Study quality . Tian et al. 2019) Toxicant High-melting resolution PCR ND 86 31.6; males under evaluation for subfertility N/A N/A High urinary levels of selected phthalate metabolites assoc. with ↓ LINE-1 promoter DNAm and low sperm parameters High Lu et al. (2018) Toxicant Bisulphite pyrosequencing Samples centrifuged at 200g for 15 min. Sperm selection ND 243 22–59; exposed to varying levels of mercury N/A N/A Negative correlation between sperm DNAm levels of H19 and urinary mercury concentrations High Laqqan et al. (2017c) Toxicant 450K array; targeted bisulphite sequencing Sperm separation medium gradient (45% and 90%) and somatic cell lysis buffer 36 40.39 (±7.32); smokers (15 for screening, 36 for validation) 42 38.38 (±8.27); non-smokers (15 for screening, 42 for validation) Smoking assoc. with DNAm of MAPK8IP3 and TKR High Darbandi et al. (2019a) Toxicant Methylation-specific PCR Centrifuged at 300g for 5 min. Sperm selection method ND 43 33–35; NZS males with high ROS levels in seminal plasma 108 33–35; NZS males with low to moderate ROS levels in seminal plasma High ROS levels assoc. with DNAm of H19 and IGF2, and with lower semen parameters Moderate Ma et al. (2019) Toxicant Bisulphite pyrosequencing Samples centrifuged for 15 min. Sperm selection ND 219 24–53; males under evaluation for subfertility N/A N/A DNAm of H19 assoc. with urinary levels of 1-OHPH and 1-OHP (PAHs) Moderate Al Khaled et al. (2018) Toxicant 450K array; validation with bisulphite sequencing PureSperm gradient (45% and 90%) 92 25–50; fertile smokers. 14 for discovery cohort and 78 for validation 14 25–50; fertile non-smokers Smoking assoc. with DNAm in PGAM5, PTPRN2 and TYRO3 Moderate Murphy et al. (2018) Toxicant RRBS ND 12 21.8 (±3.8); cannabis users 12 25.8 (±6.7); non-cannabis users Cannabis use associated with DNAm at 3979 CpG sites Moderate Pilsner et al. (2018) Toxicant Whole-genome bisulphite sequencing (WGBS) ISolate gradient (50% and 90%) 4 18–19; high serum dioxin concentration 4 18–19; low serum dioxin concentration Serum dioxin concentration assoc. with DNAm at 52 DMRs Moderate Wu et al. (2017) Toxicant 450K array Gradient (90%) centrifugation. Medium ND 48 18–55; males under evaluation for subfertility None N/A 131 DMRs were associated with at least one urinary phthalate metabolite Moderate Soubry et al. (2017) Toxicant Bisulphite pyrosequencing ISolate gradient (50% and 90%) 67 18–35; from TIEGER cohort N/A N/A Mono-isopropylphenyl diphenyl phosphate assoc. with MEG3, NDN, SNRPN DNAm. Tris(1,3-dichloro-2-propyl) phosphate exposure assoc. with DNAm at MEG3 and H19 Moderate Jenkins et al. (2017) Toxicant 450K array Somatic cell lysis 78 32.4 (±0.9); smokers 78 31.2 (±0.6); non-smokers Smoking associated with DNAm in 141 CpG sites Moderate Consales et al. (2016) Toxicant Flow cytometric immunodetection of DNAm; PCR pyrosequencing Samples washed in PBS 269 29.22 (±3.2); 75 samples from Greenland, 97 from Poland and 97 from Ukraine N/A N/A Flow cytometry suggested assoc. between persistent organic pollutant (PCBs or DDT) exposure and ↓ DNAm Moderate Leter et al. (2014) Toxicant Bis-PCR and flow cytometric immunodetection of DNAm ND 262 29.2 (± 0.3); 71 from Greenland, 96 from Poland and 95 from Ukraine N/A N/A No consistent associations between internal PFASs concentrations and any of the methylation biomarkers Moderate Hamad et al. (2018) Toxicant MethylFlash (ELISA-based) PureSperm gradient (45% and 90%), centrifugation, somatic cell lysis 55 42.66 (±8.38); smokers (>20 cigarettes/day) 54 40.07 (±6.49); non-smokers ↑ Global DNAm in smokers; smokers had ↓ sperm count Low Darbandi et al. (2019b) Toxicant Methylation-specific PCR ND 151 33.09 (±8.74); divided by seminal ROS level N/A N/A No assoc. between seminal ROS levels and DNAm of KEAP1 Low Al Khaled et al. (2018) Toxicant 450K array; validation with bisulphite sequencing PureSperm gradient (45% and 90%) 19 Age ND; smokers 20 Age ND, non-smokers No differences in DNAm between smokers and non-smokers Low Kim et al. (2015) Toxicant 5-mC immuno Swim-up 18 32.0 (±5.5); smokers 32 Age ND; non-smokers DNAm ↓ in swim-up fraction of spermatozoa; DNAm ↑ in non-smokers Low Miao et al. (2014) Toxicant Methylation-specific PCR ND 77 22–50; exposed to BPA 72 22–50; not exposed to BPA Sperm LINE-1 methylation level ↓ in BPA exposed cases cf. unexposed controls Low Chorbov et al. (2011) Toxicant Targeted bisulphite pyrosequencing ND 13 42.5; males with opioid addiction 21 29.3; healthy volunteers Significant correlation between OPRM1 methylation and opioid dependence in blood but not in sperm Very low Denham et al. (2015) Metabolic 450K array and DNAm ELISA assay kit PureSperm gradient (40% and 80%) 12 24.4 (±5.19); randomly assigned to exercise intervention 12 22.45 (±4.75); randomly assigned to not exercise during a 3-month period Global and genome-wide alterations in DNAm following 3 months of exercise High Potabattula et al. (2019) Metabolic Bisulphite pyrosequencing Swim-up then PureSperm gradient (40% and 80%) 294 38.95 (±5.9); males undergoing IVF N/A N/A No association between DNAm of imprinted genes in sperm and BMI Moderate Salas-Huetos et al. (2018) Metabolic MethylFlash (ELISA-based) Somatic cell lysis 49 24 (±4.6); healthy; consumed 60 g nuts/day 49 25 (±4.7); healthy; consumed no nuts No assoc. between nut consumption and global DNAm Moderate Ingerslev et al. (2018) Metabolic RRBS Swim-up 12 18–28; endurance athletes providing 3 samples each None N/A 330 DMRs after training and 303 DMRs after the detraining period Moderate Soubry et al. (2016) Metabolic Bisulphite pyrosequencing ISolate gradient (50% and 90%) 23 18–35; males with an overweight/obese BMI 44 18–35; lean males Relative ↓ DNAm of MEG3, NDN, SNRPN and SGCE/PEG10 DMRs in sperm from obese/overweight males Moderate Consales et al. (2014) Metabolic Bis-PCR and flow cytometric immunodetection of DNAm ND 316 29.3 (±0.3); 116 from Greenland, 100 from Poland and 100 from Ukraine N/A N/A Geographical location (Greenland versus European towns) identified as the main determinant of DNAm in repetitive sequences Moderate Tunc and Tremellen (2009) Metabolic 5-mC immuno Gradient centrifugation 45 Age ND; subfertile (longitudinal) 12 Age ND; proven fertility 3-month antioxidant supplementation assoc. with ↓ seminal ROS levels and ↓ sperm DNA fragmentation, and ↑ sperm DNAm Moderate Chan et al. (2017) Metabolic 450K array; MassARRAY EpiTYPER and other techniques Gradient centrifugation at 9400g for 20 min 18 26–36; 10 samples from men taking 400 μg folate/day for 90 days. Additional cohort of 8 exposed to long-term folate fortification 9 33 (±2); males taking placebo rather than folate No significant assoc. found between DNAm and folate supplementation Low Donkin et al. (2016) Metabolic RRBS Swim-up 16 24–40; 10 obese and 6 obese males undergoing bariatric surgery 13 30–39; lean males No DNAm difference met genome-wide significance Low Aarabi et al. (2015) Metabolic RRBS and targeted bisulphite pyrosequencing ND 30 37.9 (±1.3); subfertile NZS None N/A No significant difference in DNAm of imprinted loci after folate supplementation Low Stoger (2006) Metabolic Hairpin-bisulphite PCR ND 1 ND N/A N/A LEP CpG island ↓ meth in human and mouse sperm cf. somatic tissues Very low Jenkins et al. (2018) Ageing 450K array data from previous studies ND 329 23–56; pooled data from previous studies. Mix of subfertile patients, sperm donors and healthy volunteers 10 ND; 60 samples from 10 donors (validation cohort) Model capable of predicting age with an R2 of 0.89, a mean absolute error of 2.04 years, and a mean absolute percent error of 6.28% High Ageing 450K array and bisulphite pyrosequencing of LINE (surrogate for global levels of DNAm) No sperm selection method employed 17 23–56; proven fertility. Two samples taken 9–19 years apart N/A N/A Identified 147 regions where DNAm was significantly assoc. with age Moderate Rakyan et al. (2008) Tissue specificity MeDIP-Seq ND 4 Age ND; NZS N/A N/A Generated a reference DNA methylome for human spermatozoa Moderate Bruno et al. (2018) Cancer Bisulphite pyrosequencing PureSperm gradient (45% and 90%) 31 32.2 (±6.5); seminoma patients 61 37.1 (±5.7); healthy No difference in DNAm of imprinted genes between seminoma cases and healthy controls Low Shnorhavorian et al. (2017) Cancer MeDIP-Seq. validation using minimal read depth bisulphite sequencing ND 9 19.12–29.86; males with previous cisplatin-based treatment for osteosarcoma 9 27.5–44.4; healthy volunteers A signature of significant DMRs identified in chemotherapy-exposed sperm Low Fukuda et al. (2017) Tissue specificity WGBS ND 4 25–30; publicly available methylomes N/A N/A Sperm methylomes contained more ↓ DNAm domains than the somatic methylomes. Hypometh domains human-specific cf. chimpanzees Low Zeng et al. (2014) Tissue specificity WGBS Silica-based gradient centrifugation 2 Age ND; from previously published study; anonymous donors N/A N/A CpG islands relatively ↓ DNAm in sperm compared to somatic tissues Low Molaro et al. (2011) Tissue specificity WGBS ND 2 Age ND; healthy males N/A N/A Features that determine DNAm patterns differ between male germ cells and somatic cells; diverged between humans and chimpanzees Low Buckley et al. (2016) Tissue specificity RRBS and DNase I hypersensitive site profiles ND ND Age ND; publicly available methylomes N/A N/A Testis and sperm ↓ DNAm of BHMG1 and RSPH6A assoc. with testis-specific expression Very low Zhang et al. (2016) Tissue specificity Methylation-specific PCR Percoll gradient (40% and 80%) 15 23–34; OAT 10 23–39; NZS Promoter DNAm correlates with tissue-specific expression of BOULE and DAZL Very low Zhang et al. (2015) Cancer Bisulphite pyrosequencing ND 43 Age ND; adenocarcinoma of prostate 40 Age ND; benign prostatic hypertrophy No significant difference in the detection of RARb2 promoter DNAm when comparing bis-seq in ejaculates to bis-seq of prostatic tissue Very low Guardiola et al. (2014) Tissue specificity 450K array ND 1 ND N/A N/A APOA1/C3/A4/A5 cluster ↓ DNAm in liver, ↑ DNAm in sperm and other tissues Very low Kim et al. (2013) Tissue specificity 27K array ND 21 Age ND; publicly available methylomes N/A N/A ↓ DNAm of CTA regulatory regions in cancer tissue and sperm compared to healthy somatic tissue Very low Berthaut et al. (2013) Cancer Bisulphite pyrosequencing Centrifugation at 700g for 10 min 1 27; male undergoing temozolomide treatment. Longitudinal ND Age ND; proven fertility Temozolomide treatment assoc. with lower sperm count (although remained normal) and H19 ↓ DNAm Very low Jenkins et al. (2013) Ageing 5-mC immuno; 450K array of paired n = 2 Somatic cell lysis 67 15 for DNAm and ageing study (2 and 3 longitudinal samples); 52 for blood–sperm comparison N/A N/A ↑ DNAm levels with age Very low Nettersheim et al. (2011) Cancer Bisulphite Sanger sequencing PureSperm gradient (40% and 80%) ND ND N/A N/A NANOG promoter ↓ DNAm in spermatogonia and in several germ cell tumours, but ↑ DNAm in adult sperm Very low Fan and Zhang (2009) Tissue specificity Bisulphite sequencing of PCR amplicons (the Human Epigenome Project) ND 1 24; from the Human Epigenome Project N/A N/A CpG island methylation patterns more similar between somatic tissues than between somatic tissues and sperm Very low Grunau et al. (2005) Tissue specificity Methylation-specific PCR ND 4 ND N/A N/A ↓ DNAm of BAGE in spermatozoa and malignant tissues cf. normal, somatic tissues Very low Roberts et al. (2018) Psychiatry 450K array; pyrosequencing Somatic cell lysis. Confirmed purity of sperm samples 22 23–38; exposed to childhood abuse (5 moderate and 17 high) 12 23–29; not exposed to childhood abuse 12 DNA regions (64 probes) differentially methylated by childhood abuse identified. Also identified probes predictive of childhood abuse High Klaver et al. (2012) Methodology Bisulphite pyrosequencing Swim-up 10 Age ND; NZS N/A N/A No significant differences in DNAm in selected genes between cryopreserved and non-cryopreserved samples High Kaminsky et al. (2012) Neurological disease Bisulphite pyrosequencing ND 43 40.2 (±11.6); 29 males with bipolar disorder, 14 males with schizophrenia 30 37.7 (±10.3); healthy volunteers ↓ DNAm of one HCG9 CpG in sperm from individuals with bipolar disorder/schizophrenia High Flanagan et al. (2006) Methodology Methylation-specific PCR; microarray-based genome-wide CpG profiling ND (from cryostorage) 46 Cohort no. 1 22–35; cohort no. 2 24–56; healthy volunteers N/A N/A Significant variation between DNAm levels of selected genes both within and between samples High Ni et al. (2019) Genetics MSRE‐qPCR Somatic cell lysis 31 36.77 (±1.15); subfertile; DFI >25% 35 33.65 (±1.44); subfertile; DFI < 25% Negative assoc. between IGF2 DNAm and DFI Low Dere et al. (2016) Methodology 450K array Somatic cell lysis 12 34 (±7); longitudinal N/A N/A Intra‐individual DNAm between successive samples correlated with one another more strongly than inter-individual DNAm Low Yu et al. (2015) Methodology MeDIP-chip assay and immunostaining for DNAm v/v medium gradient (40% and 80%) 54 32.4 (±5.2); AsZS or OZS 39 33.6 (±4.8); NZS DNAm ↓ in motile sperm selected by density centrifugation compared to the less motile sperm Low Li et al. (2012) Genetics WGBS ND 4 Age ND; healthy volunteers (2 from previously published research) N/A N/A Assoc. between regions of ↓ DNAm and genomic instability Low Zeschnigk et al. (2009) Methodology Restriction digestion of DNA followed by massive parallel bisulphite sequencing of CG-rich DNA fragments ND 4 Age ND; pooled sperm samples N/A N/A Identified 824 and 482 fully methylated autosomal CGIs in blood and sperm DNA respectively Low Wu et al. (2016) Methodology MethyLight ND 68 22–47; AZS 24 22–47; NZS DNAm of the five selected testis-specific promoters was correlated between testicular DNA and paired cell-free seminal DNA Very low Hammoud et al. (2009) Genetics MeDIP followed by promoter arrays. Validation by bisulphite-sequencing of identified sites Gradient centrifugation followed by somatic cell lysis 4 ND N/A N/A Overall ↓ DNAm of developmental promoters compared to fibroblasts Very low Author, year . Type of exposure . Method of assaying DNA methylation . Initial processing . N (cases) . Age in years (SD); description of cases . N (controls) . Age in years (SD); description of controls . Main results . Study quality . Tian et al. 2019) Toxicant High-melting resolution PCR ND 86 31.6; males under evaluation for subfertility N/A N/A High urinary levels of selected phthalate metabolites assoc. with ↓ LINE-1 promoter DNAm and low sperm parameters High Lu et al. (2018) Toxicant Bisulphite pyrosequencing Samples centrifuged at 200g for 15 min. Sperm selection ND 243 22–59; exposed to varying levels of mercury N/A N/A Negative correlation between sperm DNAm levels of H19 and urinary mercury concentrations High Laqqan et al. (2017c) Toxicant 450K array; targeted bisulphite sequencing Sperm separation medium gradient (45% and 90%) and somatic cell lysis buffer 36 40.39 (±7.32); smokers (15 for screening, 36 for validation) 42 38.38 (±8.27); non-smokers (15 for screening, 42 for validation) Smoking assoc. with DNAm of MAPK8IP3 and TKR High Darbandi et al. (2019a) Toxicant Methylation-specific PCR Centrifuged at 300g for 5 min. Sperm selection method ND 43 33–35; NZS males with high ROS levels in seminal plasma 108 33–35; NZS males with low to moderate ROS levels in seminal plasma High ROS levels assoc. with DNAm of H19 and IGF2, and with lower semen parameters Moderate Ma et al. (2019) Toxicant Bisulphite pyrosequencing Samples centrifuged for 15 min. Sperm selection ND 219 24–53; males under evaluation for subfertility N/A N/A DNAm of H19 assoc. with urinary levels of 1-OHPH and 1-OHP (PAHs) Moderate Al Khaled et al. (2018) Toxicant 450K array; validation with bisulphite sequencing PureSperm gradient (45% and 90%) 92 25–50; fertile smokers. 14 for discovery cohort and 78 for validation 14 25–50; fertile non-smokers Smoking assoc. with DNAm in PGAM5, PTPRN2 and TYRO3 Moderate Murphy et al. (2018) Toxicant RRBS ND 12 21.8 (±3.8); cannabis users 12 25.8 (±6.7); non-cannabis users Cannabis use associated with DNAm at 3979 CpG sites Moderate Pilsner et al. (2018) Toxicant Whole-genome bisulphite sequencing (WGBS) ISolate gradient (50% and 90%) 4 18–19; high serum dioxin concentration 4 18–19; low serum dioxin concentration Serum dioxin concentration assoc. with DNAm at 52 DMRs Moderate Wu et al. (2017) Toxicant 450K array Gradient (90%) centrifugation. Medium ND 48 18–55; males under evaluation for subfertility None N/A 131 DMRs were associated with at least one urinary phthalate metabolite Moderate Soubry et al. (2017) Toxicant Bisulphite pyrosequencing ISolate gradient (50% and 90%) 67 18–35; from TIEGER cohort N/A N/A Mono-isopropylphenyl diphenyl phosphate assoc. with MEG3, NDN, SNRPN DNAm. Tris(1,3-dichloro-2-propyl) phosphate exposure assoc. with DNAm at MEG3 and H19 Moderate Jenkins et al. (2017) Toxicant 450K array Somatic cell lysis 78 32.4 (±0.9); smokers 78 31.2 (±0.6); non-smokers Smoking associated with DNAm in 141 CpG sites Moderate Consales et al. (2016) Toxicant Flow cytometric immunodetection of DNAm; PCR pyrosequencing Samples washed in PBS 269 29.22 (±3.2); 75 samples from Greenland, 97 from Poland and 97 from Ukraine N/A N/A Flow cytometry suggested assoc. between persistent organic pollutant (PCBs or DDT) exposure and ↓ DNAm Moderate Leter et al. (2014) Toxicant Bis-PCR and flow cytometric immunodetection of DNAm ND 262 29.2 (± 0.3); 71 from Greenland, 96 from Poland and 95 from Ukraine N/A N/A No consistent associations between internal PFASs concentrations and any of the methylation biomarkers Moderate Hamad et al. (2018) Toxicant MethylFlash (ELISA-based) PureSperm gradient (45% and 90%), centrifugation, somatic cell lysis 55 42.66 (±8.38); smokers (>20 cigarettes/day) 54 40.07 (±6.49); non-smokers ↑ Global DNAm in smokers; smokers had ↓ sperm count Low Darbandi et al. (2019b) Toxicant Methylation-specific PCR ND 151 33.09 (±8.74); divided by seminal ROS level N/A N/A No assoc. between seminal ROS levels and DNAm of KEAP1 Low Al Khaled et al. (2018) Toxicant 450K array; validation with bisulphite sequencing PureSperm gradient (45% and 90%) 19 Age ND; smokers 20 Age ND, non-smokers No differences in DNAm between smokers and non-smokers Low Kim et al. (2015) Toxicant 5-mC immuno Swim-up 18 32.0 (±5.5); smokers 32 Age ND; non-smokers DNAm ↓ in swim-up fraction of spermatozoa; DNAm ↑ in non-smokers Low Miao et al. (2014) Toxicant Methylation-specific PCR ND 77 22–50; exposed to BPA 72 22–50; not exposed to BPA Sperm LINE-1 methylation level ↓ in BPA exposed cases cf. unexposed controls Low Chorbov et al. (2011) Toxicant Targeted bisulphite pyrosequencing ND 13 42.5; males with opioid addiction 21 29.3; healthy volunteers Significant correlation between OPRM1 methylation and opioid dependence in blood but not in sperm Very low Denham et al. (2015) Metabolic 450K array and DNAm ELISA assay kit PureSperm gradient (40% and 80%) 12 24.4 (±5.19); randomly assigned to exercise intervention 12 22.45 (±4.75); randomly assigned to not exercise during a 3-month period Global and genome-wide alterations in DNAm following 3 months of exercise High Potabattula et al. (2019) Metabolic Bisulphite pyrosequencing Swim-up then PureSperm gradient (40% and 80%) 294 38.95 (±5.9); males undergoing IVF N/A N/A No association between DNAm of imprinted genes in sperm and BMI Moderate Salas-Huetos et al. (2018) Metabolic MethylFlash (ELISA-based) Somatic cell lysis 49 24 (±4.6); healthy; consumed 60 g nuts/day 49 25 (±4.7); healthy; consumed no nuts No assoc. between nut consumption and global DNAm Moderate Ingerslev et al. (2018) Metabolic RRBS Swim-up 12 18–28; endurance athletes providing 3 samples each None N/A 330 DMRs after training and 303 DMRs after the detraining period Moderate Soubry et al. (2016) Metabolic Bisulphite pyrosequencing ISolate gradient (50% and 90%) 23 18–35; males with an overweight/obese BMI 44 18–35; lean males Relative ↓ DNAm of MEG3, NDN, SNRPN and SGCE/PEG10 DMRs in sperm from obese/overweight males Moderate Consales et al. (2014) Metabolic Bis-PCR and flow cytometric immunodetection of DNAm ND 316 29.3 (±0.3); 116 from Greenland, 100 from Poland and 100 from Ukraine N/A N/A Geographical location (Greenland versus European towns) identified as the main determinant of DNAm in repetitive sequences Moderate Tunc and Tremellen (2009) Metabolic 5-mC immuno Gradient centrifugation 45 Age ND; subfertile (longitudinal) 12 Age ND; proven fertility 3-month antioxidant supplementation assoc. with ↓ seminal ROS levels and ↓ sperm DNA fragmentation, and ↑ sperm DNAm Moderate Chan et al. (2017) Metabolic 450K array; MassARRAY EpiTYPER and other techniques Gradient centrifugation at 9400g for 20 min 18 26–36; 10 samples from men taking 400 μg folate/day for 90 days. Additional cohort of 8 exposed to long-term folate fortification 9 33 (±2); males taking placebo rather than folate No significant assoc. found between DNAm and folate supplementation Low Donkin et al. (2016) Metabolic RRBS Swim-up 16 24–40; 10 obese and 6 obese males undergoing bariatric surgery 13 30–39; lean males No DNAm difference met genome-wide significance Low Aarabi et al. (2015) Metabolic RRBS and targeted bisulphite pyrosequencing ND 30 37.9 (±1.3); subfertile NZS None N/A No significant difference in DNAm of imprinted loci after folate supplementation Low Stoger (2006) Metabolic Hairpin-bisulphite PCR ND 1 ND N/A N/A LEP CpG island ↓ meth in human and mouse sperm cf. somatic tissues Very low Jenkins et al. (2018) Ageing 450K array data from previous studies ND 329 23–56; pooled data from previous studies. Mix of subfertile patients, sperm donors and healthy volunteers 10 ND; 60 samples from 10 donors (validation cohort) Model capable of predicting age with an R2 of 0.89, a mean absolute error of 2.04 years, and a mean absolute percent error of 6.28% High Ageing 450K array and bisulphite pyrosequencing of LINE (surrogate for global levels of DNAm) No sperm selection method employed 17 23–56; proven fertility. Two samples taken 9–19 years apart N/A N/A Identified 147 regions where DNAm was significantly assoc. with age Moderate Rakyan et al. (2008) Tissue specificity MeDIP-Seq ND 4 Age ND; NZS N/A N/A Generated a reference DNA methylome for human spermatozoa Moderate Bruno et al. (2018) Cancer Bisulphite pyrosequencing PureSperm gradient (45% and 90%) 31 32.2 (±6.5); seminoma patients 61 37.1 (±5.7); healthy No difference in DNAm of imprinted genes between seminoma cases and healthy controls Low Shnorhavorian et al. (2017) Cancer MeDIP-Seq. validation using minimal read depth bisulphite sequencing ND 9 19.12–29.86; males with previous cisplatin-based treatment for osteosarcoma 9 27.5–44.4; healthy volunteers A signature of significant DMRs identified in chemotherapy-exposed sperm Low Fukuda et al. (2017) Tissue specificity WGBS ND 4 25–30; publicly available methylomes N/A N/A Sperm methylomes contained more ↓ DNAm domains than the somatic methylomes. Hypometh domains human-specific cf. chimpanzees Low Zeng et al. (2014) Tissue specificity WGBS Silica-based gradient centrifugation 2 Age ND; from previously published study; anonymous donors N/A N/A CpG islands relatively ↓ DNAm in sperm compared to somatic tissues Low Molaro et al. (2011) Tissue specificity WGBS ND 2 Age ND; healthy males N/A N/A Features that determine DNAm patterns differ between male germ cells and somatic cells; diverged between humans and chimpanzees Low Buckley et al. (2016) Tissue specificity RRBS and DNase I hypersensitive site profiles ND ND Age ND; publicly available methylomes N/A N/A Testis and sperm ↓ DNAm of BHMG1 and RSPH6A assoc. with testis-specific expression Very low Zhang et al. (2016) Tissue specificity Methylation-specific PCR Percoll gradient (40% and 80%) 15 23–34; OAT 10 23–39; NZS Promoter DNAm correlates with tissue-specific expression of BOULE and DAZL Very low Zhang et al. (2015) Cancer Bisulphite pyrosequencing ND 43 Age ND; adenocarcinoma of prostate 40 Age ND; benign prostatic hypertrophy No significant difference in the detection of RARb2 promoter DNAm when comparing bis-seq in ejaculates to bis-seq of prostatic tissue Very low Guardiola et al. (2014) Tissue specificity 450K array ND 1 ND N/A N/A APOA1/C3/A4/A5 cluster ↓ DNAm in liver, ↑ DNAm in sperm and other tissues Very low Kim et al. (2013) Tissue specificity 27K array ND 21 Age ND; publicly available methylomes N/A N/A ↓ DNAm of CTA regulatory regions in cancer tissue and sperm compared to healthy somatic tissue Very low Berthaut et al. (2013) Cancer Bisulphite pyrosequencing Centrifugation at 700g for 10 min 1 27; male undergoing temozolomide treatment. Longitudinal ND Age ND; proven fertility Temozolomide treatment assoc. with lower sperm count (although remained normal) and H19 ↓ DNAm Very low Jenkins et al. (2013) Ageing 5-mC immuno; 450K array of paired n = 2 Somatic cell lysis 67 15 for DNAm and ageing study (2 and 3 longitudinal samples); 52 for blood–sperm comparison N/A N/A ↑ DNAm levels with age Very low Nettersheim et al. (2011) Cancer Bisulphite Sanger sequencing PureSperm gradient (40% and 80%) ND ND N/A N/A NANOG promoter ↓ DNAm in spermatogonia and in several germ cell tumours, but ↑ DNAm in adult sperm Very low Fan and Zhang (2009) Tissue specificity Bisulphite sequencing of PCR amplicons (the Human Epigenome Project) ND 1 24; from the Human Epigenome Project N/A N/A CpG island methylation patterns more similar between somatic tissues than between somatic tissues and sperm Very low Grunau et al. (2005) Tissue specificity Methylation-specific PCR ND 4 ND N/A N/A ↓ DNAm of BAGE in spermatozoa and malignant tissues cf. normal, somatic tissues Very low Roberts et al. (2018) Psychiatry 450K array; pyrosequencing Somatic cell lysis. Confirmed purity of sperm samples 22 23–38; exposed to childhood abuse (5 moderate and 17 high) 12 23–29; not exposed to childhood abuse 12 DNA regions (64 probes) differentially methylated by childhood abuse identified. Also identified probes predictive of childhood abuse High Klaver et al. (2012) Methodology Bisulphite pyrosequencing Swim-up 10 Age ND; NZS N/A N/A No significant differences in DNAm in selected genes between cryopreserved and non-cryopreserved samples High Kaminsky et al. (2012) Neurological disease Bisulphite pyrosequencing ND 43 40.2 (±11.6); 29 males with bipolar disorder, 14 males with schizophrenia 30 37.7 (±10.3); healthy volunteers ↓ DNAm of one HCG9 CpG in sperm from individuals with bipolar disorder/schizophrenia High Flanagan et al. (2006) Methodology Methylation-specific PCR; microarray-based genome-wide CpG profiling ND (from cryostorage) 46 Cohort no. 1 22–35; cohort no. 2 24–56; healthy volunteers N/A N/A Significant variation between DNAm levels of selected genes both within and between samples High Ni et al. (2019) Genetics MSRE‐qPCR Somatic cell lysis 31 36.77 (±1.15); subfertile; DFI >25% 35 33.65 (±1.44); subfertile; DFI < 25% Negative assoc. between IGF2 DNAm and DFI Low Dere et al. (2016) Methodology 450K array Somatic cell lysis 12 34 (±7); longitudinal N/A N/A Intra‐individual DNAm between successive samples correlated with one another more strongly than inter-individual DNAm Low Yu et al. (2015) Methodology MeDIP-chip assay and immunostaining for DNAm v/v medium gradient (40% and 80%) 54 32.4 (±5.2); AsZS or OZS 39 33.6 (±4.8); NZS DNAm ↓ in motile sperm selected by density centrifugation compared to the less motile sperm Low Li et al. (2012) Genetics WGBS ND 4 Age ND; healthy volunteers (2 from previously published research) N/A N/A Assoc. between regions of ↓ DNAm and genomic instability Low Zeschnigk et al. (2009) Methodology Restriction digestion of DNA followed by massive parallel bisulphite sequencing of CG-rich DNA fragments ND 4 Age ND; pooled sperm samples N/A N/A Identified 824 and 482 fully methylated autosomal CGIs in blood and sperm DNA respectively Low Wu et al. (2016) Methodology MethyLight ND 68 22–47; AZS 24 22–47; NZS DNAm of the five selected testis-specific promoters was correlated between testicular DNA and paired cell-free seminal DNA Very low Hammoud et al. (2009) Genetics MeDIP followed by promoter arrays. Validation by bisulphite-sequencing of identified sites Gradient centrifugation followed by somatic cell lysis 4 ND N/A N/A Overall ↓ DNAm of developmental promoters compared to fibroblasts Very low CGIs, CpG islands; DDT, dichlorodiphenyltrichloroethane; DFI, DNA fragmentation index; PCBs, polychlorinated biphenyls; PFASs, per- and polyfluoroalkyl substances; RRBS, reduced representation bisulphite sequencing. TKR, tyrosine kinase receptor (gene family); PGAM5, PGAM family member 5, mitochondrial serine/threonine protein phosphatase; PTPRN2, protein tyrosine phosphatase receptor type N2; TYRO3, TYRO3 protein tyrosine kinase (member of the TKR gene family); MEG3, maternally expressed 3; NDN, necdin, MAGE family member; KEAP1, Kelch-like ECH-associated protein 1; OPRM1, opioid receptor mu 1; SGCE, sarcoglycan epsilon; PEG10, paternally expressed gene 10; LEP, leptin; BHMG1, basic helix-loop-helix and HMG-box containing 1; RSPH6A, radial spoke head 6 homolog A; BOULE (more commonly known as BOLL), boule homolog, RNA binding protein (member of the DAZ gene family); APOA1, apolipoprotein A1 (C3, A4 and A5 are other members of the apolipoprotein gene family); CTA (more commonly known as CERNA3), competing endogenous lncRNA 3 for miR-645; BAGE, B melanoma antigen; HCG9, HLA complex group 9. Open in new tab Table III Studies assessing the effect of environmental exposures on DNA methylation in human sperm. Author, year . Type of exposure . Method of assaying DNA methylation . Initial processing . N (cases) . Age in years (SD); description of cases . N (controls) . Age in years (SD); description of controls . Main results . Study quality . Tian et al. 2019) Toxicant High-melting resolution PCR ND 86 31.6; males under evaluation for subfertility N/A N/A High urinary levels of selected phthalate metabolites assoc. with ↓ LINE-1 promoter DNAm and low sperm parameters High Lu et al. (2018) Toxicant Bisulphite pyrosequencing Samples centrifuged at 200g for 15 min. Sperm selection ND 243 22–59; exposed to varying levels of mercury N/A N/A Negative correlation between sperm DNAm levels of H19 and urinary mercury concentrations High Laqqan et al. (2017c) Toxicant 450K array; targeted bisulphite sequencing Sperm separation medium gradient (45% and 90%) and somatic cell lysis buffer 36 40.39 (±7.32); smokers (15 for screening, 36 for validation) 42 38.38 (±8.27); non-smokers (15 for screening, 42 for validation) Smoking assoc. with DNAm of MAPK8IP3 and TKR High Darbandi et al. (2019a) Toxicant Methylation-specific PCR Centrifuged at 300g for 5 min. Sperm selection method ND 43 33–35; NZS males with high ROS levels in seminal plasma 108 33–35; NZS males with low to moderate ROS levels in seminal plasma High ROS levels assoc. with DNAm of H19 and IGF2, and with lower semen parameters Moderate Ma et al. (2019) Toxicant Bisulphite pyrosequencing Samples centrifuged for 15 min. Sperm selection ND 219 24–53; males under evaluation for subfertility N/A N/A DNAm of H19 assoc. with urinary levels of 1-OHPH and 1-OHP (PAHs) Moderate Al Khaled et al. (2018) Toxicant 450K array; validation with bisulphite sequencing PureSperm gradient (45% and 90%) 92 25–50; fertile smokers. 14 for discovery cohort and 78 for validation 14 25–50; fertile non-smokers Smoking assoc. with DNAm in PGAM5, PTPRN2 and TYRO3 Moderate Murphy et al. (2018) Toxicant RRBS ND 12 21.8 (±3.8); cannabis users 12 25.8 (±6.7); non-cannabis users Cannabis use associated with DNAm at 3979 CpG sites Moderate Pilsner et al. (2018) Toxicant Whole-genome bisulphite sequencing (WGBS) ISolate gradient (50% and 90%) 4 18–19; high serum dioxin concentration 4 18–19; low serum dioxin concentration Serum dioxin concentration assoc. with DNAm at 52 DMRs Moderate Wu et al. (2017) Toxicant 450K array Gradient (90%) centrifugation. Medium ND 48 18–55; males under evaluation for subfertility None N/A 131 DMRs were associated with at least one urinary phthalate metabolite Moderate Soubry et al. (2017) Toxicant Bisulphite pyrosequencing ISolate gradient (50% and 90%) 67 18–35; from TIEGER cohort N/A N/A Mono-isopropylphenyl diphenyl phosphate assoc. with MEG3, NDN, SNRPN DNAm. Tris(1,3-dichloro-2-propyl) phosphate exposure assoc. with DNAm at MEG3 and H19 Moderate Jenkins et al. (2017) Toxicant 450K array Somatic cell lysis 78 32.4 (±0.9); smokers 78 31.2 (±0.6); non-smokers Smoking associated with DNAm in 141 CpG sites Moderate Consales et al. (2016) Toxicant Flow cytometric immunodetection of DNAm; PCR pyrosequencing Samples washed in PBS 269 29.22 (±3.2); 75 samples from Greenland, 97 from Poland and 97 from Ukraine N/A N/A Flow cytometry suggested assoc. between persistent organic pollutant (PCBs or DDT) exposure and ↓ DNAm Moderate Leter et al. (2014) Toxicant Bis-PCR and flow cytometric immunodetection of DNAm ND 262 29.2 (± 0.3); 71 from Greenland, 96 from Poland and 95 from Ukraine N/A N/A No consistent associations between internal PFASs concentrations and any of the methylation biomarkers Moderate Hamad et al. (2018) Toxicant MethylFlash (ELISA-based) PureSperm gradient (45% and 90%), centrifugation, somatic cell lysis 55 42.66 (±8.38); smokers (>20 cigarettes/day) 54 40.07 (±6.49); non-smokers ↑ Global DNAm in smokers; smokers had ↓ sperm count Low Darbandi et al. (2019b) Toxicant Methylation-specific PCR ND 151 33.09 (±8.74); divided by seminal ROS level N/A N/A No assoc. between seminal ROS levels and DNAm of KEAP1 Low Al Khaled et al. (2018) Toxicant 450K array; validation with bisulphite sequencing PureSperm gradient (45% and 90%) 19 Age ND; smokers 20 Age ND, non-smokers No differences in DNAm between smokers and non-smokers Low Kim et al. (2015) Toxicant 5-mC immuno Swim-up 18 32.0 (±5.5); smokers 32 Age ND; non-smokers DNAm ↓ in swim-up fraction of spermatozoa; DNAm ↑ in non-smokers Low Miao et al. (2014) Toxicant Methylation-specific PCR ND 77 22–50; exposed to BPA 72 22–50; not exposed to BPA Sperm LINE-1 methylation level ↓ in BPA exposed cases cf. unexposed controls Low Chorbov et al. (2011) Toxicant Targeted bisulphite pyrosequencing ND 13 42.5; males with opioid addiction 21 29.3; healthy volunteers Significant correlation between OPRM1 methylation and opioid dependence in blood but not in sperm Very low Denham et al. (2015) Metabolic 450K array and DNAm ELISA assay kit PureSperm gradient (40% and 80%) 12 24.4 (±5.19); randomly assigned to exercise intervention 12 22.45 (±4.75); randomly assigned to not exercise during a 3-month period Global and genome-wide alterations in DNAm following 3 months of exercise High Potabattula et al. (2019) Metabolic Bisulphite pyrosequencing Swim-up then PureSperm gradient (40% and 80%) 294 38.95 (±5.9); males undergoing IVF N/A N/A No association between DNAm of imprinted genes in sperm and BMI Moderate Salas-Huetos et al. (2018) Metabolic MethylFlash (ELISA-based) Somatic cell lysis 49 24 (±4.6); healthy; consumed 60 g nuts/day 49 25 (±4.7); healthy; consumed no nuts No assoc. between nut consumption and global DNAm Moderate Ingerslev et al. (2018) Metabolic RRBS Swim-up 12 18–28; endurance athletes providing 3 samples each None N/A 330 DMRs after training and 303 DMRs after the detraining period Moderate Soubry et al. (2016) Metabolic Bisulphite pyrosequencing ISolate gradient (50% and 90%) 23 18–35; males with an overweight/obese BMI 44 18–35; lean males Relative ↓ DNAm of MEG3, NDN, SNRPN and SGCE/PEG10 DMRs in sperm from obese/overweight males Moderate Consales et al. (2014) Metabolic Bis-PCR and flow cytometric immunodetection of DNAm ND 316 29.3 (±0.3); 116 from Greenland, 100 from Poland and 100 from Ukraine N/A N/A Geographical location (Greenland versus European towns) identified as the main determinant of DNAm in repetitive sequences Moderate Tunc and Tremellen (2009) Metabolic 5-mC immuno Gradient centrifugation 45 Age ND; subfertile (longitudinal) 12 Age ND; proven fertility 3-month antioxidant supplementation assoc. with ↓ seminal ROS levels and ↓ sperm DNA fragmentation, and ↑ sperm DNAm Moderate Chan et al. (2017) Metabolic 450K array; MassARRAY EpiTYPER and other techniques Gradient centrifugation at 9400g for 20 min 18 26–36; 10 samples from men taking 400 μg folate/day for 90 days. Additional cohort of 8 exposed to long-term folate fortification 9 33 (±2); males taking placebo rather than folate No significant assoc. found between DNAm and folate supplementation Low Donkin et al. (2016) Metabolic RRBS Swim-up 16 24–40; 10 obese and 6 obese males undergoing bariatric surgery 13 30–39; lean males No DNAm difference met genome-wide significance Low Aarabi et al. (2015) Metabolic RRBS and targeted bisulphite pyrosequencing ND 30 37.9 (±1.3); subfertile NZS None N/A No significant difference in DNAm of imprinted loci after folate supplementation Low Stoger (2006) Metabolic Hairpin-bisulphite PCR ND 1 ND N/A N/A LEP CpG island ↓ meth in human and mouse sperm cf. somatic tissues Very low Jenkins et al. (2018) Ageing 450K array data from previous studies ND 329 23–56; pooled data from previous studies. Mix of subfertile patients, sperm donors and healthy volunteers 10 ND; 60 samples from 10 donors (validation cohort) Model capable of predicting age with an R2 of 0.89, a mean absolute error of 2.04 years, and a mean absolute percent error of 6.28% High Ageing 450K array and bisulphite pyrosequencing of LINE (surrogate for global levels of DNAm) No sperm selection method employed 17 23–56; proven fertility. Two samples taken 9–19 years apart N/A N/A Identified 147 regions where DNAm was significantly assoc. with age Moderate Rakyan et al. (2008) Tissue specificity MeDIP-Seq ND 4 Age ND; NZS N/A N/A Generated a reference DNA methylome for human spermatozoa Moderate Bruno et al. (2018) Cancer Bisulphite pyrosequencing PureSperm gradient (45% and 90%) 31 32.2 (±6.5); seminoma patients 61 37.1 (±5.7); healthy No difference in DNAm of imprinted genes between seminoma cases and healthy controls Low Shnorhavorian et al. (2017) Cancer MeDIP-Seq. validation using minimal read depth bisulphite sequencing ND 9 19.12–29.86; males with previous cisplatin-based treatment for osteosarcoma 9 27.5–44.4; healthy volunteers A signature of significant DMRs identified in chemotherapy-exposed sperm Low Fukuda et al. (2017) Tissue specificity WGBS ND 4 25–30; publicly available methylomes N/A N/A Sperm methylomes contained more ↓ DNAm domains than the somatic methylomes. Hypometh domains human-specific cf. chimpanzees Low Zeng et al. (2014) Tissue specificity WGBS Silica-based gradient centrifugation 2 Age ND; from previously published study; anonymous donors N/A N/A CpG islands relatively ↓ DNAm in sperm compared to somatic tissues Low Molaro et al. (2011) Tissue specificity WGBS ND 2 Age ND; healthy males N/A N/A Features that determine DNAm patterns differ between male germ cells and somatic cells; diverged between humans and chimpanzees Low Buckley et al. (2016) Tissue specificity RRBS and DNase I hypersensitive site profiles ND ND Age ND; publicly available methylomes N/A N/A Testis and sperm ↓ DNAm of BHMG1 and RSPH6A assoc. with testis-specific expression Very low Zhang et al. (2016) Tissue specificity Methylation-specific PCR Percoll gradient (40% and 80%) 15 23–34; OAT 10 23–39; NZS Promoter DNAm correlates with tissue-specific expression of BOULE and DAZL Very low Zhang et al. (2015) Cancer Bisulphite pyrosequencing ND 43 Age ND; adenocarcinoma of prostate 40 Age ND; benign prostatic hypertrophy No significant difference in the detection of RARb2 promoter DNAm when comparing bis-seq in ejaculates to bis-seq of prostatic tissue Very low Guardiola et al. (2014) Tissue specificity 450K array ND 1 ND N/A N/A APOA1/C3/A4/A5 cluster ↓ DNAm in liver, ↑ DNAm in sperm and other tissues Very low Kim et al. (2013) Tissue specificity 27K array ND 21 Age ND; publicly available methylomes N/A N/A ↓ DNAm of CTA regulatory regions in cancer tissue and sperm compared to healthy somatic tissue Very low Berthaut et al. (2013) Cancer Bisulphite pyrosequencing Centrifugation at 700g for 10 min 1 27; male undergoing temozolomide treatment. Longitudinal ND Age ND; proven fertility Temozolomide treatment assoc. with lower sperm count (although remained normal) and H19 ↓ DNAm Very low Jenkins et al. (2013) Ageing 5-mC immuno; 450K array of paired n = 2 Somatic cell lysis 67 15 for DNAm and ageing study (2 and 3 longitudinal samples); 52 for blood–sperm comparison N/A N/A ↑ DNAm levels with age Very low Nettersheim et al. (2011) Cancer Bisulphite Sanger sequencing PureSperm gradient (40% and 80%) ND ND N/A N/A NANOG promoter ↓ DNAm in spermatogonia and in several germ cell tumours, but ↑ DNAm in adult sperm Very low Fan and Zhang (2009) Tissue specificity Bisulphite sequencing of PCR amplicons (the Human Epigenome Project) ND 1 24; from the Human Epigenome Project N/A N/A CpG island methylation patterns more similar between somatic tissues than between somatic tissues and sperm Very low Grunau et al. (2005) Tissue specificity Methylation-specific PCR ND 4 ND N/A N/A ↓ DNAm of BAGE in spermatozoa and malignant tissues cf. normal, somatic tissues Very low Roberts et al. (2018) Psychiatry 450K array; pyrosequencing Somatic cell lysis. Confirmed purity of sperm samples 22 23–38; exposed to childhood abuse (5 moderate and 17 high) 12 23–29; not exposed to childhood abuse 12 DNA regions (64 probes) differentially methylated by childhood abuse identified. Also identified probes predictive of childhood abuse High Klaver et al. (2012) Methodology Bisulphite pyrosequencing Swim-up 10 Age ND; NZS N/A N/A No significant differences in DNAm in selected genes between cryopreserved and non-cryopreserved samples High Kaminsky et al. (2012) Neurological disease Bisulphite pyrosequencing ND 43 40.2 (±11.6); 29 males with bipolar disorder, 14 males with schizophrenia 30 37.7 (±10.3); healthy volunteers ↓ DNAm of one HCG9 CpG in sperm from individuals with bipolar disorder/schizophrenia High Flanagan et al. (2006) Methodology Methylation-specific PCR; microarray-based genome-wide CpG profiling ND (from cryostorage) 46 Cohort no. 1 22–35; cohort no. 2 24–56; healthy volunteers N/A N/A Significant variation between DNAm levels of selected genes both within and between samples High Ni et al. (2019) Genetics MSRE‐qPCR Somatic cell lysis 31 36.77 (±1.15); subfertile; DFI >25% 35 33.65 (±1.44); subfertile; DFI < 25% Negative assoc. between IGF2 DNAm and DFI Low Dere et al. (2016) Methodology 450K array Somatic cell lysis 12 34 (±7); longitudinal N/A N/A Intra‐individual DNAm between successive samples correlated with one another more strongly than inter-individual DNAm Low Yu et al. (2015) Methodology MeDIP-chip assay and immunostaining for DNAm v/v medium gradient (40% and 80%) 54 32.4 (±5.2); AsZS or OZS 39 33.6 (±4.8); NZS DNAm ↓ in motile sperm selected by density centrifugation compared to the less motile sperm Low Li et al. (2012) Genetics WGBS ND 4 Age ND; healthy volunteers (2 from previously published research) N/A N/A Assoc. between regions of ↓ DNAm and genomic instability Low Zeschnigk et al. (2009) Methodology Restriction digestion of DNA followed by massive parallel bisulphite sequencing of CG-rich DNA fragments ND 4 Age ND; pooled sperm samples N/A N/A Identified 824 and 482 fully methylated autosomal CGIs in blood and sperm DNA respectively Low Wu et al. (2016) Methodology MethyLight ND 68 22–47; AZS 24 22–47; NZS DNAm of the five selected testis-specific promoters was correlated between testicular DNA and paired cell-free seminal DNA Very low Hammoud et al. (2009) Genetics MeDIP followed by promoter arrays. Validation by bisulphite-sequencing of identified sites Gradient centrifugation followed by somatic cell lysis 4 ND N/A N/A Overall ↓ DNAm of developmental promoters compared to fibroblasts Very low Author, year . Type of exposure . Method of assaying DNA methylation . Initial processing . N (cases) . Age in years (SD); description of cases . N (controls) . Age in years (SD); description of controls . Main results . Study quality . Tian et al. 2019) Toxicant High-melting resolution PCR ND 86 31.6; males under evaluation for subfertility N/A N/A High urinary levels of selected phthalate metabolites assoc. with ↓ LINE-1 promoter DNAm and low sperm parameters High Lu et al. (2018) Toxicant Bisulphite pyrosequencing Samples centrifuged at 200g for 15 min. Sperm selection ND 243 22–59; exposed to varying levels of mercury N/A N/A Negative correlation between sperm DNAm levels of H19 and urinary mercury concentrations High Laqqan et al. (2017c) Toxicant 450K array; targeted bisulphite sequencing Sperm separation medium gradient (45% and 90%) and somatic cell lysis buffer 36 40.39 (±7.32); smokers (15 for screening, 36 for validation) 42 38.38 (±8.27); non-smokers (15 for screening, 42 for validation) Smoking assoc. with DNAm of MAPK8IP3 and TKR High Darbandi et al. (2019a) Toxicant Methylation-specific PCR Centrifuged at 300g for 5 min. Sperm selection method ND 43 33–35; NZS males with high ROS levels in seminal plasma 108 33–35; NZS males with low to moderate ROS levels in seminal plasma High ROS levels assoc. with DNAm of H19 and IGF2, and with lower semen parameters Moderate Ma et al. (2019) Toxicant Bisulphite pyrosequencing Samples centrifuged for 15 min. Sperm selection ND 219 24–53; males under evaluation for subfertility N/A N/A DNAm of H19 assoc. with urinary levels of 1-OHPH and 1-OHP (PAHs) Moderate Al Khaled et al. (2018) Toxicant 450K array; validation with bisulphite sequencing PureSperm gradient (45% and 90%) 92 25–50; fertile smokers. 14 for discovery cohort and 78 for validation 14 25–50; fertile non-smokers Smoking assoc. with DNAm in PGAM5, PTPRN2 and TYRO3 Moderate Murphy et al. (2018) Toxicant RRBS ND 12 21.8 (±3.8); cannabis users 12 25.8 (±6.7); non-cannabis users Cannabis use associated with DNAm at 3979 CpG sites Moderate Pilsner et al. (2018) Toxicant Whole-genome bisulphite sequencing (WGBS) ISolate gradient (50% and 90%) 4 18–19; high serum dioxin concentration 4 18–19; low serum dioxin concentration Serum dioxin concentration assoc. with DNAm at 52 DMRs Moderate Wu et al. (2017) Toxicant 450K array Gradient (90%) centrifugation. Medium ND 48 18–55; males under evaluation for subfertility None N/A 131 DMRs were associated with at least one urinary phthalate metabolite Moderate Soubry et al. (2017) Toxicant Bisulphite pyrosequencing ISolate gradient (50% and 90%) 67 18–35; from TIEGER cohort N/A N/A Mono-isopropylphenyl diphenyl phosphate assoc. with MEG3, NDN, SNRPN DNAm. Tris(1,3-dichloro-2-propyl) phosphate exposure assoc. with DNAm at MEG3 and H19 Moderate Jenkins et al. (2017) Toxicant 450K array Somatic cell lysis 78 32.4 (±0.9); smokers 78 31.2 (±0.6); non-smokers Smoking associated with DNAm in 141 CpG sites Moderate Consales et al. (2016) Toxicant Flow cytometric immunodetection of DNAm; PCR pyrosequencing Samples washed in PBS 269 29.22 (±3.2); 75 samples from Greenland, 97 from Poland and 97 from Ukraine N/A N/A Flow cytometry suggested assoc. between persistent organic pollutant (PCBs or DDT) exposure and ↓ DNAm Moderate Leter et al. (2014) Toxicant Bis-PCR and flow cytometric immunodetection of DNAm ND 262 29.2 (± 0.3); 71 from Greenland, 96 from Poland and 95 from Ukraine N/A N/A No consistent associations between internal PFASs concentrations and any of the methylation biomarkers Moderate Hamad et al. (2018) Toxicant MethylFlash (ELISA-based) PureSperm gradient (45% and 90%), centrifugation, somatic cell lysis 55 42.66 (±8.38); smokers (>20 cigarettes/day) 54 40.07 (±6.49); non-smokers ↑ Global DNAm in smokers; smokers had ↓ sperm count Low Darbandi et al. (2019b) Toxicant Methylation-specific PCR ND 151 33.09 (±8.74); divided by seminal ROS level N/A N/A No assoc. between seminal ROS levels and DNAm of KEAP1 Low Al Khaled et al. (2018) Toxicant 450K array; validation with bisulphite sequencing PureSperm gradient (45% and 90%) 19 Age ND; smokers 20 Age ND, non-smokers No differences in DNAm between smokers and non-smokers Low Kim et al. (2015) Toxicant 5-mC immuno Swim-up 18 32.0 (±5.5); smokers 32 Age ND; non-smokers DNAm ↓ in swim-up fraction of spermatozoa; DNAm ↑ in non-smokers Low Miao et al. (2014) Toxicant Methylation-specific PCR ND 77 22–50; exposed to BPA 72 22–50; not exposed to BPA Sperm LINE-1 methylation level ↓ in BPA exposed cases cf. unexposed controls Low Chorbov et al. (2011) Toxicant Targeted bisulphite pyrosequencing ND 13 42.5; males with opioid addiction 21 29.3; healthy volunteers Significant correlation between OPRM1 methylation and opioid dependence in blood but not in sperm Very low Denham et al. (2015) Metabolic 450K array and DNAm ELISA assay kit PureSperm gradient (40% and 80%) 12 24.4 (±5.19); randomly assigned to exercise intervention 12 22.45 (±4.75); randomly assigned to not exercise during a 3-month period Global and genome-wide alterations in DNAm following 3 months of exercise High Potabattula et al. (2019) Metabolic Bisulphite pyrosequencing Swim-up then PureSperm gradient (40% and 80%) 294 38.95 (±5.9); males undergoing IVF N/A N/A No association between DNAm of imprinted genes in sperm and BMI Moderate Salas-Huetos et al. (2018) Metabolic MethylFlash (ELISA-based) Somatic cell lysis 49 24 (±4.6); healthy; consumed 60 g nuts/day 49 25 (±4.7); healthy; consumed no nuts No assoc. between nut consumption and global DNAm Moderate Ingerslev et al. (2018) Metabolic RRBS Swim-up 12 18–28; endurance athletes providing 3 samples each None N/A 330 DMRs after training and 303 DMRs after the detraining period Moderate Soubry et al. (2016) Metabolic Bisulphite pyrosequencing ISolate gradient (50% and 90%) 23 18–35; males with an overweight/obese BMI 44 18–35; lean males Relative ↓ DNAm of MEG3, NDN, SNRPN and SGCE/PEG10 DMRs in sperm from obese/overweight males Moderate Consales et al. (2014) Metabolic Bis-PCR and flow cytometric immunodetection of DNAm ND 316 29.3 (±0.3); 116 from Greenland, 100 from Poland and 100 from Ukraine N/A N/A Geographical location (Greenland versus European towns) identified as the main determinant of DNAm in repetitive sequences Moderate Tunc and Tremellen (2009) Metabolic 5-mC immuno Gradient centrifugation 45 Age ND; subfertile (longitudinal) 12 Age ND; proven fertility 3-month antioxidant supplementation assoc. with ↓ seminal ROS levels and ↓ sperm DNA fragmentation, and ↑ sperm DNAm Moderate Chan et al. (2017) Metabolic 450K array; MassARRAY EpiTYPER and other techniques Gradient centrifugation at 9400g for 20 min 18 26–36; 10 samples from men taking 400 μg folate/day for 90 days. Additional cohort of 8 exposed to long-term folate fortification 9 33 (±2); males taking placebo rather than folate No significant assoc. found between DNAm and folate supplementation Low Donkin et al. (2016) Metabolic RRBS Swim-up 16 24–40; 10 obese and 6 obese males undergoing bariatric surgery 13 30–39; lean males No DNAm difference met genome-wide significance Low Aarabi et al. (2015) Metabolic RRBS and targeted bisulphite pyrosequencing ND 30 37.9 (±1.3); subfertile NZS None N/A No significant difference in DNAm of imprinted loci after folate supplementation Low Stoger (2006) Metabolic Hairpin-bisulphite PCR ND 1 ND N/A N/A LEP CpG island ↓ meth in human and mouse sperm cf. somatic tissues Very low Jenkins et al. (2018) Ageing 450K array data from previous studies ND 329 23–56; pooled data from previous studies. Mix of subfertile patients, sperm donors and healthy volunteers 10 ND; 60 samples from 10 donors (validation cohort) Model capable of predicting age with an R2 of 0.89, a mean absolute error of 2.04 years, and a mean absolute percent error of 6.28% High Ageing 450K array and bisulphite pyrosequencing of LINE (surrogate for global levels of DNAm) No sperm selection method employed 17 23–56; proven fertility. Two samples taken 9–19 years apart N/A N/A Identified 147 regions where DNAm was significantly assoc. with age Moderate Rakyan et al. (2008) Tissue specificity MeDIP-Seq ND 4 Age ND; NZS N/A N/A Generated a reference DNA methylome for human spermatozoa Moderate Bruno et al. (2018) Cancer Bisulphite pyrosequencing PureSperm gradient (45% and 90%) 31 32.2 (±6.5); seminoma patients 61 37.1 (±5.7); healthy No difference in DNAm of imprinted genes between seminoma cases and healthy controls Low Shnorhavorian et al. (2017) Cancer MeDIP-Seq. validation using minimal read depth bisulphite sequencing ND 9 19.12–29.86; males with previous cisplatin-based treatment for osteosarcoma 9 27.5–44.4; healthy volunteers A signature of significant DMRs identified in chemotherapy-exposed sperm Low Fukuda et al. (2017) Tissue specificity WGBS ND 4 25–30; publicly available methylomes N/A N/A Sperm methylomes contained more ↓ DNAm domains than the somatic methylomes. Hypometh domains human-specific cf. chimpanzees Low Zeng et al. (2014) Tissue specificity WGBS Silica-based gradient centrifugation 2 Age ND; from previously published study; anonymous donors N/A N/A CpG islands relatively ↓ DNAm in sperm compared to somatic tissues Low Molaro et al. (2011) Tissue specificity WGBS ND 2 Age ND; healthy males N/A N/A Features that determine DNAm patterns differ between male germ cells and somatic cells; diverged between humans and chimpanzees Low Buckley et al. (2016) Tissue specificity RRBS and DNase I hypersensitive site profiles ND ND Age ND; publicly available methylomes N/A N/A Testis and sperm ↓ DNAm of BHMG1 and RSPH6A assoc. with testis-specific expression Very low Zhang et al. (2016) Tissue specificity Methylation-specific PCR Percoll gradient (40% and 80%) 15 23–34; OAT 10 23–39; NZS Promoter DNAm correlates with tissue-specific expression of BOULE and DAZL Very low Zhang et al. (2015) Cancer Bisulphite pyrosequencing ND 43 Age ND; adenocarcinoma of prostate 40 Age ND; benign prostatic hypertrophy No significant difference in the detection of RARb2 promoter DNAm when comparing bis-seq in ejaculates to bis-seq of prostatic tissue Very low Guardiola et al. (2014) Tissue specificity 450K array ND 1 ND N/A N/A APOA1/C3/A4/A5 cluster ↓ DNAm in liver, ↑ DNAm in sperm and other tissues Very low Kim et al. (2013) Tissue specificity 27K array ND 21 Age ND; publicly available methylomes N/A N/A ↓ DNAm of CTA regulatory regions in cancer tissue and sperm compared to healthy somatic tissue Very low Berthaut et al. (2013) Cancer Bisulphite pyrosequencing Centrifugation at 700g for 10 min 1 27; male undergoing temozolomide treatment. Longitudinal ND Age ND; proven fertility Temozolomide treatment assoc. with lower sperm count (although remained normal) and H19 ↓ DNAm Very low Jenkins et al. (2013) Ageing 5-mC immuno; 450K array of paired n = 2 Somatic cell lysis 67 15 for DNAm and ageing study (2 and 3 longitudinal samples); 52 for blood–sperm comparison N/A N/A ↑ DNAm levels with age Very low Nettersheim et al. (2011) Cancer Bisulphite Sanger sequencing PureSperm gradient (40% and 80%) ND ND N/A N/A NANOG promoter ↓ DNAm in spermatogonia and in several germ cell tumours, but ↑ DNAm in adult sperm Very low Fan and Zhang (2009) Tissue specificity Bisulphite sequencing of PCR amplicons (the Human Epigenome Project) ND 1 24; from the Human Epigenome Project N/A N/A CpG island methylation patterns more similar between somatic tissues than between somatic tissues and sperm Very low Grunau et al. (2005) Tissue specificity Methylation-specific PCR ND 4 ND N/A N/A ↓ DNAm of BAGE in spermatozoa and malignant tissues cf. normal, somatic tissues Very low Roberts et al. (2018) Psychiatry 450K array; pyrosequencing Somatic cell lysis. Confirmed purity of sperm samples 22 23–38; exposed to childhood abuse (5 moderate and 17 high) 12 23–29; not exposed to childhood abuse 12 DNA regions (64 probes) differentially methylated by childhood abuse identified. Also identified probes predictive of childhood abuse High Klaver et al. (2012) Methodology Bisulphite pyrosequencing Swim-up 10 Age ND; NZS N/A N/A No significant differences in DNAm in selected genes between cryopreserved and non-cryopreserved samples High Kaminsky et al. (2012) Neurological disease Bisulphite pyrosequencing ND 43 40.2 (±11.6); 29 males with bipolar disorder, 14 males with schizophrenia 30 37.7 (±10.3); healthy volunteers ↓ DNAm of one HCG9 CpG in sperm from individuals with bipolar disorder/schizophrenia High Flanagan et al. (2006) Methodology Methylation-specific PCR; microarray-based genome-wide CpG profiling ND (from cryostorage) 46 Cohort no. 1 22–35; cohort no. 2 24–56; healthy volunteers N/A N/A Significant variation between DNAm levels of selected genes both within and between samples High Ni et al. (2019) Genetics MSRE‐qPCR Somatic cell lysis 31 36.77 (±1.15); subfertile; DFI >25% 35 33.65 (±1.44); subfertile; DFI < 25% Negative assoc. between IGF2 DNAm and DFI Low Dere et al. (2016) Methodology 450K array Somatic cell lysis 12 34 (±7); longitudinal N/A N/A Intra‐individual DNAm between successive samples correlated with one another more strongly than inter-individual DNAm Low Yu et al. (2015) Methodology MeDIP-chip assay and immunostaining for DNAm v/v medium gradient (40% and 80%) 54 32.4 (±5.2); AsZS or OZS 39 33.6 (±4.8); NZS DNAm ↓ in motile sperm selected by density centrifugation compared to the less motile sperm Low Li et al. (2012) Genetics WGBS ND 4 Age ND; healthy volunteers (2 from previously published research) N/A N/A Assoc. between regions of ↓ DNAm and genomic instability Low Zeschnigk et al. (2009) Methodology Restriction digestion of DNA followed by massive parallel bisulphite sequencing of CG-rich DNA fragments ND 4 Age ND; pooled sperm samples N/A N/A Identified 824 and 482 fully methylated autosomal CGIs in blood and sperm DNA respectively Low Wu et al. (2016) Methodology MethyLight ND 68 22–47; AZS 24 22–47; NZS DNAm of the five selected testis-specific promoters was correlated between testicular DNA and paired cell-free seminal DNA Very low Hammoud et al. (2009) Genetics MeDIP followed by promoter arrays. Validation by bisulphite-sequencing of identified sites Gradient centrifugation followed by somatic cell lysis 4 ND N/A N/A Overall ↓ DNAm of developmental promoters compared to fibroblasts Very low CGIs, CpG islands; DDT, dichlorodiphenyltrichloroethane; DFI, DNA fragmentation index; PCBs, polychlorinated biphenyls; PFASs, per- and polyfluoroalkyl substances; RRBS, reduced representation bisulphite sequencing. TKR, tyrosine kinase receptor (gene family); PGAM5, PGAM family member 5, mitochondrial serine/threonine protein phosphatase; PTPRN2, protein tyrosine phosphatase receptor type N2; TYRO3, TYRO3 protein tyrosine kinase (member of the TKR gene family); MEG3, maternally expressed 3; NDN, necdin, MAGE family member; KEAP1, Kelch-like ECH-associated protein 1; OPRM1, opioid receptor mu 1; SGCE, sarcoglycan epsilon; PEG10, paternally expressed gene 10; LEP, leptin; BHMG1, basic helix-loop-helix and HMG-box containing 1; RSPH6A, radial spoke head 6 homolog A; BOULE (more commonly known as BOLL), boule homolog, RNA binding protein (member of the DAZ gene family); APOA1, apolipoprotein A1 (C3, A4 and A5 are other members of the apolipoprotein gene family); CTA (more commonly known as CERNA3), competing endogenous lncRNA 3 for miR-645; BAGE, B melanoma antigen; HCG9, HLA complex group 9. Open in new tab Results Study selection A total of 446 articles were identified by the search terms and through manual filtering of reference lists (Fig. 2). Ten of these were duplicate records and were therefore removed. Reviews or records that otherwise did not describe original research (n = 117) were excluded. A further 184 articles were excluded because they did not investigate humans (n = 45), did not specifically investigate DNA methylation (n = 35), investigated another tissue than sperm (n = 75) or related to forensic identification of semen (n = 29). This resulted in 135 studies that met the predetermined inclusion criteria and were included in the systematic review. To our knowledge, this represents all published studies investigating DNA methylation in human sperm that were available up to 2 March 2020. Evaluation of studies based on GRADE criteria To objectively evaluate the current body of research on humans spermatozoal DNA methylation and identify current knowledge gaps in the field, we adapted the GRADE criteria as described by Ryan and Hill (2016) to specifically be applicable to the aims of this systematic review. Details of how each study was graded are available in Supplementary Data S3. Using the adapted criteria also allowed us to identify common technical and methodological flaws, which are discussed under their respective headings below. Risk of bias There is convincing evidence in both humans and animal models that age and smoking affect DNA methylation across the genome in organs other than sperm (Horvath, 2013; Joehanes et al., 2016). Furthermore, there is extensive clinical evidence that advanced paternal age and cigarette smoking are associated with adverse effects on male fertility and offspring health (Sharm et al., 2016; de Kluiver et al., 2017; Oldereid et al., 2018). Of the 135 studies identified in our literature search, 57 studies either did not report the age of participants, did not account for age in a statistical model, such as a linear regression model with covariates, or reported significant differences in the ages of cases and controls without correcting for this. Similarly, 44 studies did not consider smoking status. There were 11 studies that did not use appropriate semen processing methods or include an analysis to assess somatic cell contamination. In such studies, one cannot disregard the possibility that somatic cells, which have considerably different DNA methylation profiles to spermatozoa, could have influenced results, particularly in the context of imprinted loci (Bartolomei and Ferguson-Smith, 2011). In seven studies, differences in semen parameters between cases and controls were not corrected for, and in two studies maternal factors were not considered in pregnancy outcomes (Benchaib et al., 2003, 2005). The matching of sperm with embryos or other tissues, i.e. that spermatozoa is derived from the same person who fathers the embryo, in studies of pregnancy outcomes or in studies of tissue specificity, is important as it limits confounding of DNA methylation data due to genetic effects (Hannon et al., 2015). Five studies were marked down for not using matched tissues under such circumstances. Appropriate statistical analysis is imperative, and this includes correcting for the use of multiple statistical tests to reduce the risk of identifying false positive outcomes by chance. We found that the poor use of statistics affected the conclusions in seven studies. Finally, one study was downgraded for applying different sperm isolation methods between cases and controls, and another was downgraded for using multiple embryos from the same family in analyses and not accounting for biases caused by relatedness. Inconsistency Studies that had inconsistent sperm processing amongst samples scored −1 for inconsistency. Only one study included a replication cohort and scored +1 in this category. Indirectness Few studies were downgraded for indirectness, but in two studies it was considered that the findings may not be generalizable due to the fact that participants were recruited from fertility clinics. Imprecision There were 18 studies that used global methods of measuring DNA methylation in sperm. Global DNA methylation refers to the total levels of DNA methylation (5-mC) across the entire genome. Due to the lack of quantitation in these methods (immunostaining or flow cytometry), all global studies were downgraded for imprecision if they did not validate their results with another method. There were 20 studies with inappropriately small sample sizes for the resolution of measurement that was being made, and these were also downgraded for imprecision. In contrast, 13 studies were upgraded for including relatively large sample sizes. Publication bias There were 70 candidate gene studies, in which targeted sequencing or PCR-based methods, or occasionally array-based techniques, were used to investigate-specific genes chosen based on prior knowledge. Such studies are prone to publication bias, given that the chosen genes were selected because of published positive results, and they also prohibit the replication of genes other than those selected. Studies using candidate gene approaches were therefore downgraded. Evidence for the association between spermatozoal DNA methylation and male subfertility, including abnormal semen parameters The most common question of studies included in this review was the association between spermatozoal DNA methylation and male subfertility, including abnormal semen parameters. Fifty-six studies investigated this topic (Table I). Among these, the quality of evidence was scored as high for 6, moderate for 16, low for 17 and very low for 17 studies. There was one moderate-quality study of global spermatozoal DNA methylation in association with semen parameters: Barzideh et al. (2013) analysed the difference in global DNA methylation between the high-density (higher quality) portion of spermatozoa and the low-density (lower quality) portion of spermatozoa separated by a Percoll density gradient and found that the high-quality sperm portion was relatively hypomethylated. Among moderate- or high-quality candidate gene studies of DNA methylation and subfertility or abnormal semen parameters, three studies found that oligozoospermia was associated with spermatozoal DNA hypomethylation of H19 (Marques et al., 2004, 2008; Dong et al., 2017), and three studies found that oligozoospermia was associated with spermatozoal DNA hypermethylation of the paternally expressed transcript 1/mesoderm-specific transcript gene (PEG1/MEST) (Marques et al., 2008; Poplinski et al., 2010; Klaver et al., 2013). Other studies, in contrast, found no association between aberrant DNA methylation of PEG1/MEST and oligozoospermia (Marques et al., 2004). Kabartan et al. (2019) found no association between oligozoospermia and DNA methylation of breast cancer gene 1 and 2 (BRCA1 and BRCA2). Tang et al. (2018) found that male subfertility, whether or not semen samples were oligozoospermic, was associated with DNA methylation defects in imprinted genes, including H19. One study found hypermethylation of the methylenetetrahydrofolate reductase (MTHFR) gene promoter in males with idiopathic subfertility and oligozoospermia, compared with fertile males (Wu et al., 2010), and another group found that DNA methylation of MTHFR was correlated with lower quality sperm parameters (Tian et al., 2015). In contrast, in a third targeted analysis of MTHFR promoter DNA methylation, hypomethylation of MTHFR was shown to be associated with low motility spermatozoa (Botezatu et al., 2014). In the context of candidate gene analyses of spermatozoal DNA methylation, it is of note that a relatively small number of genes have been investigated multiple times. Thus, as visualized in Fig. 3, spermatozoal DNA methylation of H19, PEG1/MEST and the small nuclear ribonucleoprotein polypeptide N gene (SNRPN) in relation to fertility was investigated in 20, 17 and 6 studies, respectively, whereas other genes were typically investigated in fewer than five studies, generally without resulting in replication of results. Figure 3. Open in new tabDownload slide Frequency of specific candidate genes being investigated in candidate gene analyses of human sperm. A total of 75 candidate gene studies of spermatozoal DNA methylation were identified in our literature search. Approximately half of these investigated either or both of the imprinted genes H19 or PEG1/MEST, as presented in the diagram. PEG3, paternally expressed gene 3; MEG3, maternally expressed gene 3; KCNQ1, potassium voltage-gated channel subfamily Q member 1; KCNQ1OT1, KCNQ1 opposite strand/antisense transcript 1; PLAGL1, PLAG1 like zinc finger 1. Figure 3. Open in new tabDownload slide Frequency of specific candidate genes being investigated in candidate gene analyses of human sperm. A total of 75 candidate gene studies of spermatozoal DNA methylation were identified in our literature search. Approximately half of these investigated either or both of the imprinted genes H19 or PEG1/MEST, as presented in the diagram. PEG3, paternally expressed gene 3; MEG3, maternally expressed gene 3; KCNQ1, potassium voltage-gated channel subfamily Q member 1; KCNQ1OT1, KCNQ1 opposite strand/antisense transcript 1; PLAGL1, PLAG1 like zinc finger 1. Ten moderate- or high-quality studies investigated the association between genome-wide spermatozoal DNA methylation and semen parameters or male subfertility (Houshdaran et al., 2007; Pacheco et al., 2011; Schutte et al., 2013; Camprubi et al., 2016, 2017; Jenkins et al., 2016; Laqqan et al., 2017a,b,d, 2018). Notably, none of these genome-wide analyses identified any of the frequently investigated candidate genes discussed above as significantly differentially methylated in subfertility or in association with abnormal semen parameters, such as oligozoospermia. This is despite the fact that the majority of candidate genes—all genes featured in Fig. 3—have 40 or more probes annotated to them on the Illumina 450K DNA methylation array (450K array), which was used across these 10 studies. There was also little overlap of findings from different genome-wide investigations of DNA methylation of subfertility performed using comparable methodologies: Laqqan et al. (2017b) found that subfertility was associated with differential DNA methylation at CpG sites annotated to genes encoding aldehyde dehydrogenase family 3 member B2 (ALDH3B2) and prostaglandin I2 receptor (PTGIR) using the 450K array. In a subsequent investigation using a similar approach, the same group (Laqqan et al., 2017a) found that subfertility was associated with spermatozoal DNA methylation alterations in the genes encoding potassium inwardly rectifying channel subfamily J member 5 (KCNJ5), melanophilin (MLPH) and structural maintenance of chromosomes 1B (SMC1B). In a third study of spermatozoa from subfertile males by the same group, again using the 450K array, Laqqan et al. (2018) found that subfertility was associated with DNA methylation differences in the genes encoding proline-rich coiled-coil 2A (PRRCA2), annexin A2 (ANXA2) and mitogen-activated protein kinase 8 interacting protein 3 (MAPK8IP3). There was a similar lack of overlap of findings from studies of genome-wide DNA spermatozoal methylation in oligozoospermia. Using the 450K array, Laqqan et al. (2017d) found that oligozoospermia was associated with differential DNA methylation of the gene encoding ubiquitin conjugating enzyme E2 G2 (UBE2G2), whereas Camprubi et al. (2016) found a significant association between spermatozoal hypermethylation of the gene encoding amyloid P component, serum (APCS) and oligozoospermia using a similar approach. Jenkins et al. (2016) did not identify any CpG site that met the genome-wide threshold for differential methylation in teratozoospermia, but found that teratozoospermia was associated with a higher rate of DNA methylation variability. Camprubi et al. (2017) investigated genome-wide DNA methylation in spermatozoa from males of proven fertility and found that spermatozoa exhibit relative hypomethylation compared to somatic tissues. Lastly, two groups (Pacheco et al., 2011; Schutte et al., 2013) used an earlier version of the Illumina Methylation Array, the 27K DNA Methylation Array (27K array), to investigate DNA methylation in association with sperm motility and abnormal semen parameters, respectively. Pacheco et al. (2011) found that unsupervised clustering of methylation data correctly differentiated the 21 sperm samples by their motility values, and Schutte et al. (2013) found that that unsupervised clustering of methylation data correctly separated samples according to whether they had normal or low semen parameters. Conclusion There is evidence to suggest that idiopathic male subfertility and abnormal semen parameters, in particular oligozoospermia, are associated with abnormal spermatozoal DNA methylation of imprinted regions. However, no specific DNA methylation signature of either subfertility or abnormal semen parameters has been convincingly replicated in genome-scale, unbiased analyses and the clinical benefits of findings to date is therefore limited. Furthermore, whilst these studies attempt to address the important question of the potential mechanism leading to subfertility, they do not tackle the arguably more pertinent question of whether DNA methylation in sperm can affect pregnancy outcomes and offspring health. Evidence for the association between spermatozoal DNA methylation and pregnancy outcomes, including offspring health The second question targeted by studies included in this review concerned the association between spermatozoal DNA methylation and pregnancy outcomes, including offspring health. Twenty studies addressed this question (Table II). Among these, the quality of evidence was scored as high for three, moderate for three, low for six and very low for eight studies. Among moderate- or high-quality candidate gene approaches, the majority of studies focused on methylation of imprinted genes as previously discussed (Fig. 3). Despite this considerable research effort, there is limited and conflicting evidence as to whether aberrant methylation of imprinted loci does affect ART outcomes, and there is no evidence that it affects offspring health. El Hajj et al. (2011) and Camprubi et al. (2012) found no evidence that DNA methylation at seven selected imprinted regions had an association with the success of ART, however Kobayashi et al. (2009) found that spermatozoa with aberrant methylation at one or more of 22 selected imprinted loci was associated with higher rates of miscarriage and rates of live births. This contradiction may be driven by the differing number of candidate genes chosen in the respective studies. At present, there are no high-quality genome-wide studies to validate the findings. Methylation of other genomic elements have been found to be predictors of ART outcome. El Hajj et al. (2011) found that methylation of Alu elements, the most abundant type of transposable element in the human genome (Deininger, 2011), correctly predicted whether pregnancy ended in spontaneous miscarriage or a live birth in 89% of couples. Using a multiple linear regression model, El Hajj et al. found no evidence that DNA methylation at long interspersed nuclear elements 1 (LINE-1) in sperm was correlated with live birth rates, nor were the ages of the parents, the number of retrieved oocytes, the number of fertilized oocytes, embryo quality, sperm shape, sperm count or sperm motility. Rotondo et al. (2013) identified spermatozoal hypermethylation of MTHFR as predictive of recurrent spontaneous miscarriage. Lastly, among studies of spermatozoal DNA methylation and pregnancy outcomes, Tavalaee et al. (2009) found that although spermatozoal DNA fragmentation was associated with global sperm DNA methylation and fertilization rates in males undergoing ART, none of the parameters assessed, including global sperm DNA methylation, were associated with pregnancy rates. Only one high-quality study looked at the effect of sperm DNA methylation on the future health of the offspring. Atsem et al. (2016) found that paternal age at conception was correlated with methylation of the gene encoding forkhead box K1 (FOXK1) in both sperm and on the paternal allele in cord blood of offspring. However, the attempt to link the methylation difference at the gene to the observation that the risk for autism spectrum disorder (ASD) in children increases with the age of the father did not result in a significant association, only a ‘trend’ between lower FOXK1 methylation and ASD (P = 0.07) (Atsem et al., 2016). Conclusion There is limited evidence that spermatozoal DNA methylation affects pregnancy outcomes or offspring health. This is largely due to a lack of high-quality genome-wide association studies which specifically address this topic. Further research is warranted, as this could readily inform clinical practice. Evidence for the association between spermatozoal DNA methylation and environmental factors The final research question addressed by studies summarized in this review was the association between spermatozoal DNA methylation and environmental factors (Table III). The 59 studies that addressed this topic were divided into four categories; toxicants, metabolic interventions (including diet and lifestyle), studies of tissue specificity (including cancer, ageing and methodological studies comparing tissues) and other (including genetic, psychiatric and other methodological experiments). These categories are discussed separately below. Toxicants Nineteen studies investigated the effect of various toxicants on spermatozoal DNA methylation, including six studies of cigarette smoke (Kim et al., 2015; Jenkins et al., 2017; Laqqan et al., 2017c; Al Khaled et al., 2018; Alkhaled et al., 2018; Hamad et al., 2018), two studies of recreational drugs (cannabis and opioids) (Chorbov et al., 2011; Murphy et al., 2018) and four studies of organic pollutants such as mercury, polycyclic aromatic hydrocarbons (PAHs), per- and polyfluoroalkyl substances and phthalate metabolites (Leter et al., 2014; Wu et al., 2017; Lu et al., 2018; Ma et al., 2019; Tian et al., 2019). Among these, the quality of evidence was scored as high for three studies, moderate for 10, low for five and very low for one study (Table III). As smoking is associated with lower semen sample parameters (Sharm et al., 2016), studies of the effects of cigarette smoke on spermatozoal DNA methylation were downgraded if they failed to take semen parameters into account (Jenkins et al., 2017; Hamad et al., 2018). The study of the effects of cannabis on spermatozoal DNA methylation was marked down as it did not account for the higher alcohol consumption reported in the cannabis user group (Murphy et al., 2018). Laqqan et al. (2017c), Alkhaled et al. (2018) and Jenkins et al. (2017) each published a high- or moderate-quality genome-wide association study assessing DNA methylation differences between the spermatozoa from smokers and non-smokers using the 450K array. Notably, there was no overlap of findings between studies from different groups. Laqqan et al. (2017c) reported differences between smokers and non-smokers in MAPK8IP3 and TYRO3 protein tyrosine kinase (TYRO3) genes, and Alkhaled et al. (2018) reported differences in the protein tyrosine phosphatase receptor type N2 (PTPRN2) and TYRO3. Jenkins et al. (2017) found 141 significantly differentially methylated sites associated in smokers, however none of these were associated with the genes identified by Laqqan et al. (2017c) or Alkhaled et al. (2018). Jenkins et al. (2017) and Alkhaled et al. (2018) also reported significant decreases in several semen parameters in smokers, including semen volume, total sperm count and total progressively motile sperm count, but did not account for this in a multiple regression model or by other means in the DNA methylation analyses (which resulted in the evaluation of both studies as moderate rather than high quality). In a study of 12 cannabis users compared with 12 non-users, Murphy et al. (2018) found 3979 CpG sites that differed significantly in sperm using reduced representation bisulphite sequencing (RRBS), and 46 genes which had more than 10 differentially methylated sites. One of these genes, PTPRN2, was also found to be altered by smoking by Alkhaled et al. (2018). However, Murphy et al. (2018) also saw significantly lower sperm concentration in cannabis users. Several candidate gene studies of imprinted loci showed that exposure to toxicants, such as mercury (Lu et al., 2018), PAHs 1-OHP and 1-OHPH (Ma et al., 2019) and high levels of reactive oxygen species (ROS) in seminal plasma (Darbandi et al., 2019b), was associated with H19 hypomethylation. Soubry et al. (2017) found that exposure to flame retardants in men was associated with spermatozoal DNA hypermethylation of MEG3, necdin, melanoma antigen family member (NDN) and SNRPN, and a separate group found an association between ROS levels in seminal plasma and hypermethylation of the insulin-like growth factor 2 gene (IGF2) (Darbandi et al., 2019a). High ROS levels were also found to be associated with lower semen parameters (Darbandi et al., 2019a). Tian et al. (2019) found no association between high levels of urinary phthalate metabolites and DNA methylation of imprinted loci, but found an association between urinary phthalate metabolites and LINE-1 hypomethylation. Also investigating urinary phthalates, Wu et al. (2017) identified 131 differentially methylated regions (DMRs), which were associated with exposure. Similarly, Pilsner et al. (2018) identified 52 DMRs that were differentially methylated between men with high and low serum levels of the organic pollutant dioxin. Consales et al. (2016) found no difference in methylation of repetitive sequences in relation to exposure to the persistent organic pollutants polychlorinated biphenyls and dichlorodiphenyltrichloroethane, but did find that exposure was associated with global spermatozoal DNA hypomethylation as measured by flow cytometry. Finally, Leter et al. (2014) found no association between exposure to perfluoroalkyl substances and methylation of repetitive genomic sequences. Conclusion While several studies have identified specific spermatozoal DNA methylation signatures associated with toxicants, such as cigarette smoke, cannabis and pollutants, there are yet no robust findings that have been replicated across studies. Furthermore, it is frequently difficult to distinguish between reported DNA methylation differences and the effects the toxicants themselves have on semen parameters. Lastly, the use of widely varying research methodologies (e.g. global-, candidate gene- and genome-wide analyses) makes it difficult to compare the scale and severity of the DNA methylation differences observed under these exposures. Metabolic interventions. Eleven studies assessed the effect of metabolic interventions on spermatozoal DNA methylation, including exercise (Denham et al., 2015; Ingerslev et al., 2018), diet and supplements (Tunc and Tremellen, 2009; Aarabi et al., 2015; Chan et al., 2017; Salas-Huetos et al., 2018), BMI (Consales et al., 2014; Donkin et al., 2016; Soubry et al., 2016; Potabattula et al., 2019) and endocrine factors (Stoger, 2006). Among these, the quality of evidence was scored as high for one study, moderate for six, low for three and very low for one study (Table III). Two studies analysed the effects of exercise on spermatozoal DNA methylation. Both of these compared samples collected before and after a 3-month (Denham et al., 2015) or a 6-week (Ingerslev et al., 2018) exercise intervention, using the 450K array and RRBS, respectively, to analyse spermatozoal DNA methylation. Both groups found multiple differentially methylated positions when comparing DNA methylation in sperm before and after the intervention, however there was no overlap of findings between the two studies. Two studies analysed the effect of BMI on methylation of imprinted loci in targeted analyses (Soubry et al., 2016; Potabattula et al., 2019). However, whilst Soubry et al. found that high BMI was associated with MEG3, NDN, SNRPN, sarcoglycan epsilon (SGCE) and paternally expressed gene 10 (PEG10) hypomethylation and H19 hypermethylation, Potabattula et al. found no associations that met genome-wide significance. This discrepancy could in part be explained by the fact that Soubry et al. did not correct for the use of multiple statistical tests, as well as a considerably smaller study cohort (Soubry: n = 67; Potabattula: n = 294). However, Potabattula et al. were unable to correct for participant smoking states due to lack of information on their participants, which could have confounded results. Tunc and Tremellen (2009) found no association between spermatozoal DNA methylation and serum homocysteine levels, while Salas-Huetos et al. (2018) found no association between spermatozoal DNA methylation and nut consumption in the ‘FERTINUTS’ study. In a study including 269 participants, Consales et al. (2014) measured a number of different exposures, including BMI, semen parameters, sperm chromatin integrity, biomarkers of accessory gland function and the plasma concentration of reproductive hormones, but found no global differences in sperm DNA methylation. However, spermatozoal DNA methylation of LINE-1 elements separated participants by geographical location in this study. Conclusion There is limited and conflicting evidence on whether diet and lifestyle interventions are associated with altered spermatozoal DNA methylation. Larger-scale, longitudinal analyses with appropriate controls for potential confounding factors are required. Tissue specificity and ageing Eighteen studies were categorized as relating to the effect of ageing, cancer and tissue specificity of DNA methylation in spermatozoa. Among these, the quality of evidence was scored as high for one, moderate for two, low for five and very low for nine studies (Table III). Of the three studies scoring high or moderate for quality of evidence, two were from the same group. These both identified the presence of multiple age-associated DMRs in sperm (Jenkins et al., 2014), and showed that these regions could be used to predict chronological age at relatively high accuracy (mean absolute error = 2.04 years) (Jenkins et al., 2018). The third paper generated a genome-wide DNA methylation reference of 13 normal somatic tissues, as well as human sperm, and showed broad-scale tissue-specific differences in DNA methylation patterns (Rakyan et al., 2008). Conclusion Ageing appears to be strongly associated with spermatozoal DNA methylation. It is unclear, however, whether and how such ageing-related changes have lasting effects on processes such as embryogenesis or offspring health. It is also clear that the DNA methylation signature of human spermatozoa differs considerably to that of many, if not all, somatic human tissues. Other Eleven studies that assessed the effect of a variety of environmental stimuli and exposures on sperm DNA methylation did not fit into the previous three categories. These included studies of genetics, psychiatric illness and methodological studies. Among these, the quality of evidence was scored as high for four, low for five and very low for two studies (Table III). Two of the high-quality studies related to psychiatric disease. Kaminsky et al. (2012) showed hypomethylation of a CpG site in the HLA complex group 9 (HCG9) gene in patients with schizophrenia or bipolar disorder compared to healthy controls. Grouping of these two disorders was deemed justifiable due to their shared genetics and overlap in symptoms (Maggioni et al., 2017). In a genome-wide analysis of patients that suffered from childhood abuse, 12 regions of DNA in sperm were identified as being differentially methylated in response to abuse (Roberts et al., 2018). The remaining two high-quality studies were methodological in nature. Flanagan et al. (2006) aimed to evaluate intra- and inter-individual DNA methylation variation in sperm using targeted sequencing and microarray technologies. This study showed, for the first time, that there is significant variation of DNA methylation both within and between sperm samples. Finally, Klaver et al. (2012) showed that there were no significant differences in DNA methylation at a series of imprinted genes and other loci, such as Alu and LINE-1 elements, before and after cryo-perseveration. This is a critical study, for both sperm DNA methylation and fertility research as cryo-preservation of sperm is ubiquitously used in ART, and represents the only high-quality methodological study of human sperm preservation techniques that was identified by our comprehensive search. Conclusion Although two studies suggest a possible association between psychiatric disease and spermatozoal DNA methylation, there is insufficient evidence to make conclusive statements. There appears to be little detrimental effect of cryo-preservation on sperm via spermatozoal DNA methylation. Discussion This systematic review represents the largest critical evaluation of studies of DNA methylation in human spermatozoa performed to date. The comprehensive nature of the review allowed us to identify common methodological limitations as well as draw overarching conclusions regarding our current knowledge of the human sperm DNA methylome with respect to three specific questions, as well as identify existing knowledge gaps. Overall summary of findings A currently widely held supposition in the field of human sperm epigenetics suggests that aberrant DNA methylation of imprinted loci and the promoter of MTHFR is causally associated with changes to sperm function and health, which in turn lead to lower pregnancy rates and poor pregnancy outcomes (Fields et al., 2013). Generally, the assumption made in this hypothesis is that environmental exposures can cause changes to sperm DNA methylation at these key genes, that in turn affect fertility and pregnancy outcomes. In the existing literature (between 2003 and 2020, scoring high or moderate for quality of evidence), there is suggestive evidence that idiopathic male subfertility, in particular oligozoospermia, is associated with aberrant methylation of imprinted loci and other genes in sperm. However, results are often conflicting and have not been replicated in unbiased, genome-wide analyses. There is similarly conflicting evidence that spermatozoal DNA methylation, including aberrant DNA methylation of imprinted genomic regions, can affect pregnancy outcomes or offspring health. In sum, therefore, changes in spermatozoal DNA methylation should not, at present, be seen as an established mechanism of subfertility, poor pregnancy outcomes or offspring ill health. Spermatozoal DNA methylation is also of limited clinical utility as a biomarker for such outcomes (Fig. 4). Figure 4. Open in new tabDownload slide Overview of the quality of evidence for how selected acquired and environmental factors can influence the spermatozoal DNA methylome and offspring health. This systematic review found suggestive evidence that ageing and toxicants, such as cigarette smoke, are associated with changes in human spermatozoal DNA methylation. However, there is yet insufficient evidence that such changes are associated with pregnancy outcomes, including success of assisted reproductive technologies and offspring health. Figure 4. Open in new tabDownload slide Overview of the quality of evidence for how selected acquired and environmental factors can influence the spermatozoal DNA methylome and offspring health. This systematic review found suggestive evidence that ageing and toxicants, such as cigarette smoke, are associated with changes in human spermatozoal DNA methylation. However, there is yet insufficient evidence that such changes are associated with pregnancy outcomes, including success of assisted reproductive technologies and offspring health. There is suggestive evidence that ageing and some toxicants, particularly cigarette smoke, can alter DNA methylation of sperm, but there is insufficient evidence for an effect of BMI, diet, lifestyle, psychiatric disorders, drugs or medication. In some cases, this is due to a lack of high- or moderate-quality studies. In other cases, associations have not been studied sufficiently or replicated to accumulate sufficient evidence. However, while DNA methylation in sperm may not have effects on the next generation, and may not affect fertility, it could nonetheless be used as a biomarker of exposure to certain environmental stressors, for example, to predict age and to assess exposure to pollutants. Given that human sperm is relatively accessible and composed of a comparatively homogenous cell population, DNA methylation in sperm in this context could become a useful resource clinically. Possibly the most striking conclusion from this systematic review is the lack of overlap of findings obtained from the same or related research questions. Despite several high- and moderate-quality studies independently demonstrating associations between spermatozoal DNA methylation and fertility, smoking status, exercise interventions and exposure to some environmental pollutants, none of these studies identified DNA methylation changes at the same genomic regions. These discrepancies undermine the overall evidence for any one of these exposures being associated with sperm DNA methylation and could be explained in several ways. First, despite sperm being a relatively homogeneous cell type, the human sperm methylome is highly heterogeneous across individuals and populations, meaning that different results could be observed between biologically different cohorts. This is supported by the finding from Consales et al. (2014), who showed that DNA methylation in a sample of 316 men clustered significantly by geographical location and no other investigated parameter was significantly associated with the clustering. This issue can only be overcome with larger sample sizes, preferably performed using multicentre collaborations, as well as appropriate strategies controlling for genetic backgrounds. From a technical perspective, there are differences between studies in the application and effectiveness of sperm processing methods used to eliminate somatic cell contamination. Harmonization of research protocols would make cross-cohort comparisons and meta-analyses more feasible, increasing the robustness of findings. We also identified wide variations in the statistical models used to analyse DNA methylation data, particularly in how confounders relevant to human studies of fertility were accounted for. Different methods of analysing DNA methylation may also yield different results for technical reasons. For example, CpG sites found to be differentially methylated in candidate gene analyses may not be present on the 450K array, which captures around 1.7% of the ∼28 million CpG sites in the human genome (Pidsley et al., 2013), and only covers a small proportion of enhancers and other regulatory regions. The more recent Infinium MethylationEPIC array has improved coverage of enhancers and regulatory regions, and will thus make future DNA methylation analyses more comprehensive (Pidsley et al., 2016), although given the unique nature of the sperm methylome a custom array may be more suitable (Chan et al., 2019). Standardization of these methods is required to allow cross-study comparisons. Finally, for individual studies with high quality of evidence, independent replication of findings is required to solidify credible evidence, and eventually inform clinical practice. For example, while Atsem et al. (2016) report an intriguing association between age-related spermatozoal DNA methylation and neurodevelopmental problems in offspring of older fathers, this study constitutes the only study on this topic and requires replication. Recommendations for future research The drawbacks and major weaknesses we identified in the current body of literature on human spermatozoal DNA methylation were common across many studies and spanned all three investigated themes. Most of them relate to general methodological considerations for epigenetic studies of human phenotypes and tissues, and include study design, experimental processes and statistical analyses, while some challenges are specific to the nature of sperm as a tissue. In light of these prevailing criticisms, we have put forward a series of recommendations for future research, which are briefly summarized in Fig. 5 and discussed below. Figure 5. Open in new tabDownload slide Recommendations for future research on spermatozoal DNA methylation. Figure 5. Open in new tabDownload slide Recommendations for future research on spermatozoal DNA methylation. Sample size To generate robust, reproducible findings, sample sizes need to be sufficiently large to meet the power needs of the study design. Performing a power calculation prior to study commencement is recommended, and methods for a range of platforms and approaches have been described previously (Tsai and Bell, 2015; Saffari et al., 2018; Graw et al., 2019; Mansell et al., 2019). As sperm is not routinely collected outside of reproductive medicine settings, it is likely that multicentre collaborations will be required to meet the need for increased sample sizes. Such collaborations should harmonize protocols for phenotyping of participants, processing samples and data analyses to limit batch effects (Leek et al., 2010; Forest et al., 2018). Sperm sample processing A range of protocols exists for selecting motile spermatozoa and cleaning semen samples from contaminating somatic cells (Supplementary Fig. S1, Supplementary Data 2). Of importance, irrespective of which protocol is employed, samples should be checked post-processing to ensure that spermatozoa used for epigenetic analyses contain a pure sperm population. Assessing somatic cell contamination visually, and ideally also using computational methods, is highly advisable. Methodology Unbiased, epigenome-wide association studies of biologically relevant tissues have yielded important insights into a range of common human conditions, including metabolic and neurodegenerative disease (De Jager et al., 2014; Wahl et al., 2017). Whole-genome bisulphite sequencing (WGBS) remains the gold standard method to profile DNA methylation across all CpG sites in the human genome. However, achieving adequate coverage at each site can be expensive. Until the price of WGBS drops substantially, RRBS, which targets CpG rich regions, and microarray-based approaches will remain more cost-effective option for large-scale, unbiased investigations. The latest edition of human DNA methylation microarray from Illumina, the MethylationEPIC array (EPIC array), covers around 850 000 CpG sites annotated to most coding genes and regulatory regions of the human genome (Illumina, 2017). In addition, results from previous generations of arrays can largely be compared across studies. For example, the EPIC array covers >90% of CpG sites present on the 450K array, and contains ∼350 000 additional CpG sites located in enhancers. Therefore, results obtained from analyses using the EPIC array can generally be checked for consistency with previous results from studies using the 450K array, provided that similar methodologies were used. Tissue specificity For cross-tissue comparisons, biological samples should be matched, i.e. taken from the same individual or from the individual’s offspring. Because genetic variation influences the DNA methylome, not using matched tissues, especially in small sample sizes, increases the likelihood of confounding and should be avoided in future research (Bell et al., 2011; McClay et al., 2015; McRae et al., 2018). Multiple testing Performing a statistical test always comes with a risk of reporting a false positive association (∼5%). This is a particularly pertinent problem when multiple genomic loci are investigated, notably so in the case of array-based studies of DNA methylation, where there can be up to 850 000 individual tests. To reduce the chance of reporting a false positive, a correction for multiple tests is crucial. Standard methods, such as Bonferroni’s correction or a false discovery rate correction, can be applied but more accurate genome- and array-wide P-value thresholds estimating the specific testing burden have been reported recently (Mansell et al., 2019). Concerningly, multiple correction is most often omitted in targeted or candidate gene studies, despite multiple CpG sites across a specific genomic region being investigated. If, for instance, a candidate gene study investigates 20 CpG sites without correcting for multiple testing, the overall probability of detecting at least one false positive association is 64%. Modelling covariates Confounding factors, where differences in characteristics between samples could explain observed differences in DNA methylation, need to be considered when conducting epigenetic association studies. On the one hand, appropriate sampling is required to avoid complete overlap between conditions and confounding factors (e.g. when all samples in the control group are older than the case group); on the other hand, these confounding factors should be controlled for in the statistical analysis. We recommend collecting extensive phenotype data and including these as covariates in the analysis as well as using batch correcting normalization and control procedures (Pidsley et al., 2013; Fortin et al., 2014; Maksimovic et al., 2015; Saffari et al., 2018). Based on the findings of this systematic review, relevant covariates include but are not limited to: age, smoking status, BMI, alcohol consumption, exposure to environmental toxicants, medications, semen parameters, disease and exercise regimes. While associations of most of these covariates with spermatozoal DNA methylation are often still speculative and require further evidence, it is crucial to err on the side of caution with regards to covariates and confounding. In practise this means high-quality studies should clearly demonstrate that any potential and plausible confounding factors do not affect the outcomes. Expanding the scope of DNA methylation studies in human sperm The majority of studies to date have focused on differential DNA methylation in relation to subfertility or abnormal semen parameters. Widening the scope in future investigations to include more research on, for example, the influence of metabolic health, toxicants and ageing on sperm epigenetics would be recommended. Such research would also motivate improved biological hypotheses about the epidemiological associations between paternal ageing, smoking and BMI on offspring health (Oldereid et al., 2018). In this regard, however, other causes than changes to spermatozoa, whether genetic or epigenetic, need also to be taken into account. For example, there has long been an assumption that increased paternal age is associated with an increased risk for neurodevelopmental conditions, such as autism and schizophrenia, in the offspring due to age-related increases in the rate of de novo mutations in spermatozoal genomes (Gratten et al., 2016). However, recent research suggests that this association is at least partly explained by the fact that psychiatric illness is associated with an increased propensity for men to father their first child at a relatively either young or old age (Gratten et al., 2016). These two explanations for the association are of course not mutually exclusive, and it is possible that the increased rate of de novo mutations also plays a part. Nevertheless, future research exploring the association between paternal age-related changes to his (epi)genome and the propensity for his offspring to develop psychiatric disease should take these behavioural and social factors into account. Future work should also investigate the possibility that the unique chromatin packaging in spermatozoa, wherein histones are sequentially replaced with protamines during sperm maturation (Wang et al., 2019), may influence the accuracy of commonly used methodologies for interrogating DNA methylation in spermatozoa. Furthermore, chromatin packaging changes rapidly after ejaculation (Björndahl and Kvist, 2010). Therefore, the time of sperm collection after ejaculation may influence the accuracy of findings and protocols will need to be harmonized to make results comparable across studies. Overall, future research should explore the complex crosstalk between different epigenetic processes, including DNA methylation, histone modifications and small RNA species, in sperm functioning and how these may be altered by environmental and physiological change. Strengths and limitations of the systematic review This systematic review is an exhaustive literature search, including a large number of publications studying DNA methylation variation in human sperm from 2003 until 2020. Although this review is not a meta-analysis, a key strength is that it critically evaluates the findings of these studies with respect to the quality of the evidence they present, enabling the reader to navigate a large and often self-contradictory field. A limitation of this systematic review is the applicability of the GRADE criteria to fit its particular purpose; it was necessary to modify the criteria to account for the lack of randomized control trials in this field. Second, potential bias is introduced in the reporting of studies by the fact that several groups publish multiple high-quality papers, potentially using the same cohorts. Finally, it would be interesting to evaluate other epigenetic modifications, such as histone modifications and small non-coding RNA, in human sperm in response to environmental stressors and their potential impact on transgenerational inheritance, but this was beyond the scope of this review. Conclusion This systematic review of studies investigating DNA methylation in human spermatozoa found that the reliability of results from studies thus far has frequently been hampered by inadequate sample sizes and methodological drawbacks. Nevertheless, it appears from higher quality studies that factors such as ageing, cigarette smoking and environmental pollutants are associated with differences in the DNA methylation landscape of spermatozoa. Such findings are not yet sufficiently robust or replicated to be translated to clinical practice or to inform whether, and how, paternal lifestyle could influence offspring health. To increase the likelihood of uncovering biologically and clinically meaningful findings in future research, we have outlined a set of recommendations that we advise should be taken into account when designing a study in this or related areas. Further characterization of the human sperm epigenome has the potential to improve our understanding of fertility, embryogenesis and the capacity of environmental factors to impact the next generation. Acknowledgements We thank Vardhman Rakyan and David Williams for their support and advice in the preparation of this manuscript. Authors’ roles F.A. designed the study and created the figures. F.A. and A.F.D. collected the records and the data, analysed the data and wrote the manuscript. F.A., A.F.D. and S.J.M. discussed the conclusions, made the recommendations and edited the manuscript. Funding No specific funding was sought for the study. During preparation of the manuscript, F.A. was supported by funding from the Rosetrees Trust (M466). A.F.D. was supported by the Centre for Genomics and Child Health at the Blizard Institute and S.J.M. is funded by the Edmond and Lily Safra Early Career Fellowship Program and the UK Dementia Research Institute, which receives its funding from UK DRI Ltd, funded by the UK Medical Research Council, Alzheimer’s Society and Alzheimer’s Research UK. Conflicts of interest The authors declare that they have no competing interests. References Aarabi M , Gabriel MCS , Chan D , Behan NA , Caron M , Pastinen T , Bourque G , MacFarlane AJ , Zini A , Trasler J. High-dose folic acid supplementation alters the human sperm methylome and is influenced by the MTHFR C677T polymorphism . Hum Mol Genet 2015 : 24 ; 6301 – 6313 . Google Scholar Crossref Search ADS PubMed WorldCat Alkhaled Y , Laqqan M , Tierling S , Lo Porto C , Amor H , Hammadeh ME. Impact of cigarette-smoking on sperm DNA methylation and its effect on sperm parameters . Andrologia 2018 ; 50 :e12950. Google Scholar OpenURL Placeholder Text WorldCat Al Khaled Y , Tierling S , Laqqan M , Lo Porto C , Hammadeh ME. Cigarette smoking induces only marginal changes in sperm DNA methylation levels of patients undergoing intracytoplasmic sperm injection treatment . Andrologia 2018 : 50 . Google Scholar OpenURL Placeholder Text WorldCat Al-Khtib M , Blachere T , Guerin JF , Lefevre A. Methylation profile of the promoters of NANOG and OCT4 in ICSI human embryos . Hum Reprod 2012 : 27 ; 2948 – 2954 . Google Scholar Crossref Search ADS PubMed WorldCat American Society for Reproductive Medicine. Definitions of infertility and recurrent pregnancy loss: a committee opinion . Fertil Steril 2013 : 99 ; 63 . Crossref Search ADS PubMed WorldCat Ankolkar M , Patil A , Warke H , Salvi V , Mokashi NK , Pathak S , Balasinor NH. Methylation analysis of idiopathic recurrent spontaneous miscarriage cases reveals aberrant imprinting at H19 ICR in normozoospermic individuals . Fertil Steril 2012 : 98 ; 1186 – 1192 . Google Scholar Crossref Search ADS PubMed WorldCat Aoki VW , Emery BR , Carrell DT. Global sperm deoxyribonucleic acid methylation is unaffected in protamine-deficient infertile males . Fertil Steril 2006 : 86 ; 1541 – 1543 . Google Scholar Crossref Search ADS PubMed WorldCat Aston KI , Punj V , Liu LH , Carrell DT. Genome-wide sperm deoxyribonucleic acid methylation is altered in some men with abnormal chromatin packaging or poor in vitro fertilization embryogenesis . Fertil Steril 2012 : 97 ; 285 – U327 . Google Scholar Crossref Search ADS PubMed WorldCat Aston KI , Uren PJ , Jenkins TG , Horsager A , Cairns BR , Smith AD , Carrell DT. Aberrant sperm DNA methylation predicts male fertility status and embryo quality . Fertil Steril 2015 : 104 ; 1388 . Google Scholar Crossref Search ADS PubMed WorldCat Atsem S , Reichenbach J , Potabattula R , Dittrich M , Nava C , Depienne C , Bohm L , Rost S , Hahn T , Schorsch M , et al. Paternal age effects on sperm FOXK1 and KCNA7 methylation and transmission into the next generation . Hum Mol Genet 2016 : 25 ; 4996 – 5005 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Bahreinian M , Tavalaee M , Abbasi H , Kiani-Esfahani A , Shiravi AH , Nasr-Esfahani MH. DNA hypomethylation predisposes sperm to DNA damage in individuals with varicocele . Syst Biol Reprod Med 2015 : 61 ; 179 – 186 . Google Scholar Crossref Search ADS PubMed WorldCat Barbosa TD , Ingerslev LR , Alm PS , Versteyhe S , Massart J , Rasmussen M , Donkin I , Sjogren R , Mudry JM , Vetterli L , et al. High-fat diet reprograms the epigenome of rat spermatozoa and transgenerationally affects metabolism of the offspring . Mol Metab 2016 : 5 ; 184 – 197 . Google Scholar Crossref Search ADS PubMed WorldCat Bartolomei MS, , Ferguson-Smith AC . Mammalian genomic imprinting . Cold Spring Harb Perspect Biol 2011 : 3 ; 17 . Google Scholar Crossref Search ADS WorldCat Barzideh J , Scott RJ , Aitken RJ. Analysis of the global methylation status of human spermatozoa and its association with the tendency of these cells to enter apoptosis . Andrologia 2013 : 45 ; 424 – 429 . Google Scholar Crossref Search ADS PubMed WorldCat Bell JT , Pai AA , Pickrell JK , Gaffney DJ , Pique-Regi R , Degner JF , Gilad Y , Pritchard JK. DNA methylation patterns associate with genetic and gene expression variation in HapMap cell lines . Genome Biol 2011 : 12 . Google Scholar OpenURL Placeholder Text WorldCat Benchaib M , Ajina M , Lornage J , Niveleau A , Durand P , Guerin JF. Quantitation by image analysis of global DNA methylation in human spermatozoa and its prognostic value in in vitro fertilization: a preliminary study . Fertil Steril 2003 : 80 ; 947 – 953 . Google Scholar Crossref Search ADS PubMed WorldCat Benchaib M , Braun V , Ressnikof D , Lornage J , Durand P , Niveleau A , Guerin JF. Influence of global sperm DNA methylation on IVF results . Hum Reprod 2005 : 20 ; 768 – 773 . Google Scholar Crossref Search ADS PubMed WorldCat Berthaut I , Montjean D , Dessolle L , Morcel K , Deluen F , Poirot C , Bashamboo A , McElreavey K , Ravel C. Effect of temozolomide on male gametes: an epigenetic risk to the offspring? J Assist Reprod Genet 2013 : 30 ; 827 – 833 . Google Scholar Crossref Search ADS PubMed WorldCat Björndahl L , Kvist U. Human sperm chromatin stabilization: a proposed model including zinc bridges . Mol Hum Reprod 2010 : 16 ; 23 – 29 . Google Scholar Crossref Search ADS PubMed WorldCat Boissonnas CC , El Abdalaoui H , Haelewyn V , Fauque P , Dupont JM , Gut I , Vaiman D , Jouannet P , Tost J , Jammes H. Specific epigenetic alterations of IGF2–H19 locus in spermatozoa from infertile men . Eur J Hum Genet 2010 : 18 ; 73 – 80 . Google Scholar Crossref Search ADS PubMed WorldCat Botezatu A , Socolov R , Socolov D , Iancu IV , Anton G. Methylation pattern of methylene tetrahydrofolate reductase and small nuclear ribonucleoprotein polypeptide N promoters in oligoasthenospermia: a case-control study . Reprod Biomed Online 2014 : 28 ; 225 – 231 . Google Scholar Crossref Search ADS PubMed WorldCat Bruno C , Blagoskonov O , Barberet J , Guilleman M , Daniel S , Tournier B , Binquet C , Choux C , Fauque P. Sperm imprinting integrity in seminoma patients? Clin Epigenet 2018 : 10 . Google Scholar OpenURL Placeholder Text WorldCat Buckley L , Lacey M , Ehrlich M. Epigenetics of the myotonic dystrophy-associated DMPK gene neighborhood . Epigenomics 2016 : 8 ; 13 – 31 . Google Scholar Crossref Search ADS PubMed WorldCat Camprubi C , Cigliano RA , Salas-Huetos A , Garrido N , Blanco J. What the human sperm methylome tells us . Epigenomics 2017 : 9 ; 1299 – 1315 . Google Scholar Crossref Search ADS PubMed WorldCat Camprubi C , Pladevall M , Grossmann M , Garrido N , Pons MC , Blanco J. Semen samples showing an increased rate of spermatozoa with imprinting errors have a negligible effect in the outcome of assisted reproduction techniques . Epigenetics 2012 : 7 ; 1115 – 1124 . Google Scholar Crossref Search ADS PubMed WorldCat Camprubi C , Pladevall M , Grossmann M , Garrido N , Pons MC , Blanco J. Lack of association of MTHFR rs1801133 polymorphism and CTCFL mutations with sperm methylation errors in infertile patients . J Assist Reprod Genet 2013 : 30 ; 1125 – 1131 . Google Scholar Crossref Search ADS PubMed WorldCat Camprubi C , Salas-Huetos A , Aiese-Cigliano R , Godo A , Pons MC , Castellano G , Grossmann M , Sanseverino W , Martin-Subero JI , Garrido N , et al. Spermatozoa from infertile patients exhibit differences of DNA methylation associated with spermatogenesis-related processes: an array-based analysis . Reprod Biomed Online 2016 : 33 ; 709 – 719 . Google Scholar Crossref Search ADS PubMed WorldCat Cassuto NG , Montjean D , Siffroi JP , Bouret D , Marzouk F , Copin H , Benkhalifa M. Different levels of DNA methylation detected in human sperms after morphological selection using high magnification microscopy . Biomed Res Int 2016 . Google Scholar OpenURL Placeholder Text WorldCat Chan D , McGraw S , Klein K , Wallock LM , Konermann C , Plass C , Chan P , Robaire B , Jacob RA , Greenwood CMT , et al. Stability of the human sperm DNA methylome to folic acid fortification and short-term supplementation . Hum Reprod 2017 : 32 ; 272 – 283 . Google Scholar Crossref Search ADS PubMed WorldCat Chan D , Shao XJ , Dumargne MC , Aarabi M , Simon MM , Kwan T , Bailey JL , Robaire B , Kimmins S , San Gabriel MC , et al. Customized methylC-capture sequencing to evaluate variation in the human sperm DNA methylome representative of altered folate metabolism . Environ Health Perspect 2019 : 127 . Google Scholar OpenURL Placeholder Text WorldCat Cheng YS , Lu CW , Lin TY , Lin PY , Lin YM. Causes and clinical features of infertile men with nonobstructive azoospermia and histopathologic diagnosis of hypospermatogenesis . Urology 2017 : 105 ; 62 – 68 . Google Scholar Crossref Search ADS PubMed WorldCat Chorbov V , Todorov A , Lynskey M , Cicero T. Elevated levels of DNA methylation at the OPRM1 promoter in blood and sperm from male opioid addicts . J Opioid Manag 2011 : 7 ; 258 –2 64 . Google Scholar Crossref Search ADS PubMed WorldCat Consales C , Leter G , Bonde JPE , Toft G , Eleuteri P , Moccia T , Budillon A , Jonsson BAG , Giwercman A , Pedersen HS , et al. Indices of methylation in sperm DNA from fertile men differ between distinct geographical regions . Hum Reprod 2014 : 29 ; 2065 – 2072 . Google Scholar Crossref Search ADS PubMed WorldCat Consales C , Toft G , Leter G , Bonde JPE , Uccelli R , Pacchierotti F , Eleuteri P , Jonsson BAG , Giwercman A , Pedersen HS , et al. Exposure to persistent organic pollutants and sperm DNA methylation changes in Arctic and European populations . Environ Mol Mutagen 2016 : 57 ; 200 – 209 . Google Scholar Crossref Search ADS PubMed WorldCat Darbandi M , Darbandi S , Agarwal A , Baskaran S , Dutta S , Sengupta P , Khorshid HRK , Esteves S , Gilany K , Hedayati M , et al. Reactive oxygen species-induced alterations in H19–IGF2 methylation patterns, seminal plasma metabolites, and semen quality . J Assist Reprod Genet 2019 : 36 ; 241 – 253 . Google Scholar Crossref Search ADS PubMed WorldCat Darbandi M , Darbandi S , Agarwal A , Baskaran S , Sengupta P , Dutta S , Mokarram P , Saliminejad K , Sadeghi MR. Oxidative stress-induced alterations in seminal plasma antioxidants: is there any association with KEAP1 gene methylation in human spermatozoa? Andrologia 2019 : 51 . Google Scholar OpenURL Placeholder Text WorldCat Daxinger L , Whitelaw E. Understanding transgenerational epigenetic inheritance via the gametes in mammals . Nat Rev Genet 2012 : 13 ; 153 – 162 . Google Scholar Crossref Search ADS PubMed WorldCat Deininger P. Alu elements: know the SINEs . Genome Biol 2011 : 12 . Google Scholar OpenURL Placeholder Text WorldCat De Jager PL , Srivastava G , Lunnon K , Burgess J , Schalkwyk LC , Yu L , Eaton ML , Keenan BT , Ernst J , McCabe C , et al. Alzheimer’s disease: early alterations in brain DNA methylation at ANK1, BIN1, RHBDF2 and other loci . Nat Neurosci 2014 : 17 ; 1156 – 1163 . Google Scholar Crossref Search ADS PubMed WorldCat de Kluiver H , Buizer-Voskamp JE , Dolan CV , Boomsma DI. Paternal age and psychiatric disorders: a review . Am J Med Genet B Neuropsychiatr Genet 2017 : 174 ; 202 – 213 . Google Scholar Crossref Search ADS PubMed WorldCat Denham J , O’Brien BJ , Harvey JT , Charchar FJ. Genome-wide sperm DNA methylation changes after 3 months of exercise training in humans . Epigenomics 2015 : 7 ; 717 – 731 . Google Scholar Crossref Search ADS PubMed WorldCat Denomme MM, , McCallie BR , Parks JC , Schoolcraft WB , Katz-Jaffe MG. Alterations in the sperm histone-retained epigenome are associated with unexplained male factor infertility and poor blastocyst development in donor oocyte IVF cycles . Hum Reprod 2017 : 32 ; 2443 – 2455 . Google Scholar Crossref Search ADS PubMed WorldCat Dere E , Huse S , Hwang K , Sigman M , Boekelheide K. Intra- and inter-individual differences in human sperm DNA methylation . Andrology 2016 : 4 ; 832 – 842 . Google Scholar Crossref Search ADS PubMed WorldCat Desai M , Jellyman JK , Ross MG. Epigenomics, gestational programming and risk of metabolic syndrome . Int J Obes (Lond) 2015 : 39 ; 633 – 641 . Google Scholar Crossref Search ADS PubMed WorldCat Dias BG , Ressier KJ. Parental olfactory experience influences behavior and neural structure in subsequent generations . Nat Neurosci 2014 : 17 ; 89 – 96 . Google Scholar Crossref Search ADS PubMed WorldCat Dong H , Wang YX , Zou ZK , Chen LM , Shen CY , Xu SQ , Zhang J , Zhao FF , Ge SQ , Gao Q , et al. Abnormal methylation of imprinted genes and cigarette smoking: assessment of their association with the risk of male infertility . Reprod Sci 2017 : 24 ; 114 – 123 . Google Scholar Crossref Search ADS PubMed WorldCat Donkin I , Versteyhe S , Ingerslev LR , Qian K , Mechta M , Nordkap L , Mortensen B , Appel EVR , Jorgensen N , Kristiansen VB , et al. Obesity and bariatric surgery drive epigenetic variation of spermatozoa in humans . Cell Metab 2016 : 23 ; 369 – 378 . Google Scholar Crossref Search ADS PubMed WorldCat Du Y , Li MY , Chen J , Duan YG , Wang XB , Qiu Y , Cai ZM , Gui YT , Jiang H. Promoter targeted bisulfite sequencing reveals DNA methylation profiles associated with low sperm motility in asthenozoospermia . Hum Reprod 2016 : 31 ; 24 – 33 . Google Scholar Crossref Search ADS PubMed WorldCat Dupont C, , Armant DR , Brenner CA. Epigenetics: definition, mechanisms and clinical perspective . Semin Reprod Med 2009 : 27 ; 351 – 357 . Google Scholar Crossref Search ADS PubMed WorldCat El Hajj N , Zechner U , Schneider E , Tresch A , Gromoll J , Hahn T , Schorsch M , Haaf T. Methylation status of imprinted genes and repetitive elements in sperm DNA from infertile males . Sex Dev 2011 : 5 ; 60 – 69 . Google Scholar Crossref Search ADS PubMed WorldCat Fan SC , Zhang XG. CpG island methylation pattern in different human tissues and its correlation with gene expression . Biochem Biophys Res Commun 2009 : 383 ; 421 – 425 . Google Scholar Crossref Search ADS PubMed WorldCat Feinberg JI , Bakulski KM , Jaffe AE , Tryggvadottir R , Brown SC , Goldman LR , Croen LA , Hertz-Picciotto I , Newschaffer CJ , Fallin MD , et al. Paternal sperm DNA methylation associated with early signs of autism risk in an autism-enriched cohort . Int J Epidemiol 2015 : 44 ; 1199 – 1210 . Google Scholar Crossref Search ADS PubMed WorldCat Fields E , Chard J , James D , Treasure T , Guideline Dev G. GUIDELINES fertility (update): summary of NICE guidance . Br Med J 2013 : 346 . Google Scholar OpenURL Placeholder Text WorldCat Flanagan JM , Popendikyte V , Pozdniakovaite N , Sobolev M , Assadzadeh A , Schumacher A , Zangeneh M , Lau L , Virtanen C , Wang SC , et al. Intra- and interindividual epigenetic variation in human germ cells . Am J Hum Genet 2006 : 79 ; 67 – 84 . Google Scholar Crossref Search ADS PubMed WorldCat Forest M , O’Donnell KJ , Voisin G , Gaudreau H , Macisaac JL, , McEwen LM, , Silveira PP , Steiner M , Kobor MS , Meaney MJ , et al. Agreement in DNA methylation levels from the Illumina 450K array across batches, tissues, and time . Epigenetics 2018 : 13 ; 19 – 32 . Google Scholar Crossref Search ADS PubMed WorldCat Fortin JP , Labbe A , Lemire M , Zanke BW , Hudson TJ , Fertig EJ , Greenwood CMT , Hansen KD. Functional normalization of 450K methylation array data improves replication in large cancer studies . Genome Biol 2014 : 15 . Google Scholar OpenURL Placeholder Text WorldCat Fukuda K , Inoguchi Y , Ichiyanagi K , Ichiyanagi T , Go Y , Nagano M , Yanagawa Y , Takaesu N , Ohkawa Y , Imai H , et al. Evolution of the sperm methylome of primates is associated with retrotransposon insertions and genome instability . Hum Mol Genet 2017 : 26 ; 3508 – 3519 . Google Scholar Crossref Search ADS PubMed WorldCat Geuns E , De Temmerman N , Hilven P , Van Steirteghem A , Liebaers I , De Rycke M. Methylation analysis of the intergenic differentially methylated region of DLK1–GTL2 in human . Eur J Hum Genet 2007 : 15 ; 352 – 361 . Google Scholar Crossref Search ADS PubMed WorldCat Gosden R , Trasler J , Lucifero D , Faddy M. Rare congenital disorders, imprinted genes, and assisted reproductive technology . Lancet 2003 : 361 ; 1975 – 1977 . Google Scholar Crossref Search ADS PubMed WorldCat Gratten J , Wray NR , Peyrot WJ , McGrath JJ , Visscher PM , Goddard ME. Risk of psychiatric illness from advanced paternal age is not predominantly from de novo mutations . Nat Genet 2016 : 48 ; 718 . Google Scholar Crossref Search ADS PubMed WorldCat Graw S , Henn R , Thompson JA , Koestler DC. pwrEWAS: a user-friendly tool for comprehensive power estimation for epigenome wide association studies (EWAS) . BMC Bioinformatics 2019 : 20 . Google Scholar OpenURL Placeholder Text WorldCat Grunau C , Sanchez C , Ehrlich M , van der Bruggen P , Hindermann W , Rodriguez C , Krieger S , Dubeau L , Fiala E , De Sario A. Frequent DNA hypomethylation of human juxtacentromeric BAGE loci in cancer . Genes Chromosomes Cancer 2005 : 43 ; 11 – 24 . Google Scholar Crossref Search ADS PubMed WorldCat Guardiola M , Oliva I , Guillaumet A , Martin-Trujillo A , Rosales R , Vallve JC , Sabench F , del Castillo D , Zaina S , Monk D , et al. Tissue-specific DNA methylation profiles regulate liver-specific expression of the APOA1/C3/A4/A5 cluster and can be manipulated with demethylating agents on intestinal cells . Atherosclerosis 2014 : 237 ; 528 – 535 . Google Scholar Crossref Search ADS PubMed WorldCat Gunes S , Agarwal A , Henkel R , Mahmutoglu AM , Sharma R , Esteves SC , Aljowair A , Emirzeoglu D , Alkhani A , Pelegrini L , et al. Association between promoter methylation of MLH1 and MSH2 and reactive oxygen species in oligozoospermic men—a pilot study . Andrologia 2018 : 50 . Google Scholar OpenURL Placeholder Text WorldCat Hamad MF , Abu Dayyih WA , Laqqan M , AlKhaled Y , Montenarh M , Hammadeh ME. The status of global DNA methylation in the spermatozoa of smokers and non-smokers . Reprod Biomed Online 2018 : 37 ; 581 – 589 . Google Scholar Crossref Search ADS PubMed WorldCat Hammoud SS , Nix DA , Zhang HY , Purwar J , Carrell DT , Cairns BR. Distinctive chromatin in human sperm packages genes for embryo development . Nature 2009 : 460 ; 473 – U478 . Google Scholar Crossref Search ADS PubMed WorldCat Hammoud SS , Purwar J , Pflueger C , Cairns BR , Carrell DT. Alterations in sperm DNA methylation patterns at imprinted loci in two classes of infertility . Fertil Steril 2010 : 94 ; 1728 – 1733 . Google Scholar Crossref Search ADS PubMed WorldCat Hannon E , Lunnon K , Schalkwyk L , Mill J. Interindividual methylomic variation across blood, cortex, and cerebellum: implications for epigenetic studies of neurological and neuropsychiatric phenotypes . Epigenetics 2015 : 10 ; 1024 – 1032 . Google Scholar Crossref Search ADS PubMed WorldCat Horvath S. DNA methylation age of human tissues and cell types . Genome Biol 2013 : 14 . Google Scholar OpenURL Placeholder Text WorldCat Houshdaran S , Cortessis VK , Siegmund K , Yang A , Laird PW , Sokol RZ. Widespread epigenetic abnormalities suggest a broad DNA methylation erasure defect in abnormal human sperm . PLoS One 2007 : 2 . Google Scholar OpenURL Placeholder Text WorldCat Huypens P , Sass S , Wu M , Dyckhoff D , Tschop M , Theis F , Marschall S , de Angelis MH , Beckers J. Epigenetic germline inheritance of diet-induced obesity and insulin resistance . Nat Genet 2016 : 48 ; 497 . Google Scholar Crossref Search ADS PubMed WorldCat Ibala-Romdhane S , Al-Khtib M , Khoueiry R , Blachere T , Guerin JF , Lefevre A. Analysis of H19 methylation in control and abnormal human embryos, sperm and oocytes . Eur J Hum Genet 2011 : 19 ; 1138 – 1143 . Google Scholar Crossref Search ADS PubMed WorldCat Illumina. Pub. No. 1070-2015-008-B. Infinium MethylationEPIC BeadChip Datasheet. San Diego, USA: Illumina, 2017 . Ingerslev LR , Donkin I , Fabre O , Versteyhe S , Mechta M , Pattamaprapanont P , Mortensen B , Krarup NT , Barres R. Endurance training remodels sperm-borne small RNA expression and methylation at neurological gene hotspots . Clin Epigenet 2018 : 10 . Google Scholar OpenURL Placeholder Text WorldCat Jenkins TG , Aston KI , Cairns BR , Carrell DT. Paternal aging and associated intraindividual alterations of global sperm 5-methylcytosine and 5-hydroxymethylcytosine levels . Fertil Steril 2013 : 100 ; 945 . Google Scholar Crossref Search ADS PubMed WorldCat Jenkins TG , Aston KI , Cairns B , Smith A , Carrell DT. Paternal germ line aging: DNA methylation age prediction from human sperm . BMC Genomics 2018 : 19 . Google Scholar OpenURL Placeholder Text WorldCat Jenkins TG , Aston KI , Hotaling JM , Shamsi MB , Simon L , Carrell DT. Teratozoospermia and asthenozoospermia are associated with specific epigenetic signatures . Andrology 2016 : 4 ; 843 – 849 . Google Scholar Crossref Search ADS PubMed WorldCat Jenkins TG , Aston KI , Meyer TD , Hotaling JM , Shamsi MB , Johnstone EB , Cox KJ , Stanford JB , Porucznik CA , Carrell DT. Decreased fecundity and sperm DNA methylation patterns . Fertil Steril 2016 : 105 ; 51 . Google Scholar Crossref Search ADS PubMed WorldCat Jenkins TG , Aston KI , Pflueger C , Cairns BR , Carrell DT. Age-associated sperm DNA methylation alterations: possible implications in offspring disease susceptibility . PLoS Genet 2014 : 10 . Google Scholar OpenURL Placeholder Text WorldCat Jenkins TG , Aston KI , Trost C , Farley J , Hotaling JM , Carrell DT. Intra-sample heterogeneity of sperm DNA methylation . Mol Hum Reprod 2015 : 21 ; 313 – 319 . Google Scholar Crossref Search ADS PubMed WorldCat Jenkins TG , James ER , Alonso DF , Hoidal JR , Murphy PJ , Hotaling JM , Cairns BR , Carrell DT , Aston KI. Cigarette smoking significantly alters sperm DNA methylation patterns . Andrology 2017 : 5 ; 1089 – 1099 . Google Scholar Crossref Search ADS PubMed WorldCat Joehanes R , Just AC , Marioni RE , Pilling LC , Reynolds LM , Mandaviya PR , Guan WH , Xu T , Elks CE , Aslibekyan S , et al. Epigenetic signatures of cigarette smoking . Circ Cardiovasc Genet 2016 : 9 ; 436 – 447 . Google Scholar Crossref Search ADS PubMed WorldCat Kabartan E , Gunes S , Arslan MA , Asci R. Investigating the relationship between BRCA1 and BRCA2 genes methylation profile and sperm DNA fragmentation in infertile men . Andrologia 2019 : 51 . Google Scholar OpenURL Placeholder Text WorldCat Kaminsky Z , Tochigi M , Jia P , Pal M , Mill J , Kwan A , Ioshikhes I , Vincent JB , Kennedy JL , Strauss J , et al. A multi-tissue analysis identifies HLA complex group 9 gene methylation differences in bipolar disorder . Mol Psychiatry 2012 : 17 ; 728 – 740 . Google Scholar Crossref Search ADS PubMed WorldCat Kim R , Kulkarni P , Hannenhalli S. Derepression of cancer/testis antigens in cancer is associated with distinct patterns of DNA hypomethylation . BMC Cancer 2013 : 13 . Google Scholar OpenURL Placeholder Text WorldCat Kim SK , Jee BC , Kim SH. Histone methylation and acetylation in ejaculated human sperm: effects of swim-up and smoking . Fertil Steril 2015 : 103 ; 1425 – 1431 . Google Scholar Crossref Search ADS PubMed WorldCat Klaver R , Bleiziffer A , Redmann K , Mallidis C , Kliesch S , Gromoll J. Routine cryopreservation of spermatozoa is safe—evidence from the DNA methylation pattern of nine spermatozoa genes . J Assist Reprod Genet 2012 : 29 ; 943 – 950 . Google Scholar Crossref Search ADS PubMed WorldCat Klaver R , Tuttelmann F , Bleiziffer A , Haaf T , Kliesch S , Gromoll J. DNA methylation in spermatozoa as a prospective marker in andrology . Andrology 2013 : 1 ; 731 – 740 . Google Scholar Crossref Search ADS PubMed WorldCat Kobayashi H , Hiura H , John RM , Sato A , Otsu E , Kobayashi N , Suzuki R , Suzuki F , Hayashi C , Utsunomiya T , et al. DNA methylation errors at imprinted loci after assisted conception originate in the parental sperm . Eur J Hum Genet 2009 : 17 ; 1582 – 1591 . Google Scholar Crossref Search ADS PubMed WorldCat Kobayashi H , Sato A , Otsu E , Hiura H , Tomatsu C , Utsunomiya T , Sasaki H , Yaegashi N , Arima T. Aberrant DNA methylation of imprinted loci in sperm from oligospermic patients . Hum Mol Genet 2007 : 16 ; 2542 – 2551 . Google Scholar Crossref Search ADS PubMed WorldCat Kobayashi N , Miyauchi N , Tatsuta N , Kitamura A , Okae H , Hiura H , Sato A , Utsunomiya T , Yaegashi N , Nakai K , et al. Factors associated with aberrant imprint methylation and oligozoospermia . Sci Rep 2017 : 7 . Google Scholar OpenURL Placeholder Text WorldCat Kuhtz J , Schneider E , El Hajj N , Zimmermann L , Fust O , Linek B , Seufert R , Hahn T , Schorsch M , Haaf T. Epigenetic heterogeneity of developmentally important genes in human sperm: implications for assisted reproduction outcome . Epigenetics 2014 : 9 ; 1648 – 1658 . Google Scholar Crossref Search ADS PubMed WorldCat Laqqan M , Solomayer EF , Hammadeh M. Aberrations in sperm DNA methylation patterns are associated with abnormalities in semen parameters of subfertile males . Reprod Biol 2017 a: 17 ; 246 – 251 . Google Scholar Crossref Search ADS PubMed WorldCat Laqqan M , Solomayer EF , Hammadeh M. Association between alterations in DNA methylation level of spermatozoa at CpGs dinucleotide and male subfertility problems . Andrologia 2018 : 50 . Google Scholar OpenURL Placeholder Text WorldCat Laqqan M , Tierling S , Alkhaled Y , Lo Porto C , Hammadeh ME. Alterations in sperm DNA methylation patterns of oligospermic males . Reprod Biol 2017 d: 17 ; 396 – 400 . Google Scholar Crossref Search ADS PubMed WorldCat Laqqan M , Tierling S , Alkhaled Y , Lo Porto C , Solomayer EF , Hammadeh M. Spermatozoa from males with reduced fecundity exhibit differential DNA methylation patterns . Andrology 2017 b: 5 ; 971 – 978 . Google Scholar Crossref Search ADS PubMed WorldCat Laqqan M , Tierling S , Alkhaled Y , Lo Porto C , Solomayer EF , Hammadeh ME. Aberrant DNA methylation patterns of human spermatozoa in current smoker males . Reprod Toxicol 2017 c: 71 ; 126 – 133 . Google Scholar Crossref Search ADS PubMed WorldCat Laurentino S , Beygo J , Nordhoff V , Kliesch S , Wistuba J , Borgmann J , Buiting K , Horsthemke B , Gromoll J. Epigenetic germline mosaicism in infertile men . Hum Mol Genet 2015 : 24 ; 1295 – 1304 . Google Scholar Crossref Search ADS PubMed WorldCat Leek JT , Scharpf RB , Bravo HC , Simcha D , Langmead B , Johnson WE , Geman D , Baggerly K , Irizarry RA. Tackling the widespread and critical impact of batch effects in high-throughput data . Nat Rev Genet 2010 : 11 ; 733 – 739 . Google Scholar Crossref Search ADS PubMed WorldCat Leter G , Consales C , Eleuteri P , Uccelli R , Specht IO , Toft G , Moccia T , Budillon A , Jonsson BAG , Lindh CH , et al. Exposure to perfluoroalkyl substances and sperm DNA global methylation in arctic and European populations . Environ Mol Mutagen 2014 : 55 ; 591 – 600 . Google Scholar Crossref Search ADS PubMed WorldCat Li B , Li JB , Xiao XF , Ma YF , Wang J , Liang XX , Zhao HX , Jiang F , Yao YQ , Wang XH. Altered DNA methylation patterns of the H19 differentially methylated region and the DAZL gene promoter are associated with defective human sperm . PLoS One 2013 : 8 . Google Scholar OpenURL Placeholder Text WorldCat Li J , Harris RA , Cheung SW , Coarfa C , Jeong M , Goodell MA , White LD , Patel A , Kang SH , Shaw C , et al. Genomic hypomethylation in the human germline associates with selective structural mutability in the human genome . PLoS Genet 2012 : 8 . Google Scholar OpenURL Placeholder Text WorldCat Li ZX , Ma X , Wang ZH. A differentially methylated region of the DAZ1 gene in spermatic and somatic cells . Asian J Androl 2006 : 8 ; 61 – 67 . Google Scholar Crossref Search ADS PubMed WorldCat Liberati A , Altman DG , Tetzlaff J , Mulrow C , Gotzsche PC , Ioannidis JPA , Clarke M , Devereaux PJ , Kleijnen J , Moher D. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration . Br Med J 2009 : 339 . Google Scholar OpenURL Placeholder Text WorldCat Louie K , Minor A , Ng R , Poon K , Chow V , Ma S. Evaluation of DNA methylation at imprinted DMRs in the spermatozoa of oligozoospermic men in association with MTHFR C677T genotype . Andrology 2016 : 4 ; 825 – 831 . Google Scholar Crossref Search ADS PubMed WorldCat Lu ZX , Ma YF , Gao LY , Li YJ , Li Q , Qiang M. Urine mercury levels correlate with DNA methylation of imprinting gene H19 in the sperm of reproductive-aged men . PLoS One 2018 : 13 . Google Scholar OpenURL Placeholder Text WorldCat Ma YF , Lu ZX , Wang L , Qiang M. Correlation of internal exposure levels of polycyclic aromatic hydrocarbons to methylation of imprinting genes of sperm . Int J Environ Res Public Health 2019 : 16 . Google Scholar OpenURL Placeholder Text WorldCat Maggioni E , Crespo-Facorro B , Nenadic I , Benedetti F , Gaser C , Sauer H , Roiz-Santianez R , Poletti S , Marinelli V , Bellani M , et al. Common and distinct structural features of schizophrenia and bipolar disorder: the European Network on Psychosis, Affective disorders and Cognitive Trajectory (ENPACT) study . PLoS One 2017 : 12 . Google Scholar OpenURL Placeholder Text WorldCat Maksimovic J , Gagnon-Bartsch JA , Speed TP , Oshlack A. Removing unwanted variation in a differential methylation analysis of Illumina HumanMethylation450 array data . Nucleic Acids Res 2015 : 43 . Google Scholar OpenURL Placeholder Text WorldCat Mansell G , Gorrie-Stone TJ , Bao YC , Kumari M , Schalkwyk LS , Mill J , Hannon E. Guidance for DNA methylation studies: statistical insights from the Illumina EPIC array . BMC Genomics 2019 : 20 . Google Scholar OpenURL Placeholder Text WorldCat Marques CJ , Carvalho F , Sousa M , Barros A. Genomic imprinting in disruptive spermatogenesis . Lancet 2004 : 363 ; 1700 – 1702 . Google Scholar Crossref Search ADS PubMed WorldCat Marques CJ , Costa P , Vaz B , Carvalho F , Fernandes S , Barros A , Sousa M. Abnormal methylation of imprinted genes in human sperm is associated with oligozoospermia . Mol Hum Reprod 2008 : 14 ; 67 – 73 . Google Scholar Crossref Search ADS PubMed WorldCat Marques CJ , Francisco T , Sousa S , Carvalho F , Barros A , Sousa M. Methylation defects of imprinted genes in human testicular spermatozoa . Fertil Steril 2010 : 94 ; 585 – 594 . Google Scholar Crossref Search ADS PubMed WorldCat Marques PI , Fernandes S , Carvalho F , Barros A , Sousa M , Marques CJ. DNA methylation imprinting errors in spermatogenic cells from maturation arrest azoospermic patients . Andrology 2017 : 5 ; 451 – 459 . Google Scholar Crossref Search ADS PubMed WorldCat McClay JL , Shabalin AA , Dozmorov MG , Adkins DE , Kumar G , Nerella S , Clark SL , Bergen SE , Hultman CM , Magnusson PKE , et al. High density methylation QTL analysis in human blood via next-generation sequencing of the methylated genomic DNA fraction . Genome Biol 2015 : 16 . Google Scholar OpenURL Placeholder Text WorldCat McRae AF , Marioni RE , Shah S , Yang J , Powell JE , Harris SE , Gibson J , Henders AK , Bowdler L , Painter JN , et al. Identification of 55,000 replicated DNA methylation QTL . Sci Rep 2018 : 8 . Google Scholar OpenURL Placeholder Text WorldCat Miao M , Zhou X , Li Y , Zhang O , Zhou Z , Li T , Yuan W , Li R , Li DK. LINE-1 hypomethylation in spermatozoa is associated with bisphenol A exposure . Andrology 2014 : 2 ; 138 – 144 . Google Scholar Crossref Search ADS PubMed WorldCat Mima M , Greenwald D , Ohlander S. Environmental toxins and male fertility . Curr Urol Rep 2018 : 19 . Google Scholar OpenURL Placeholder Text WorldCat Minor A , Chow V , Ma S. Aberrant DNA methylation at imprinted genes in testicular sperm retrieved from men with obstructive azoospermia and undergoing vasectomy reversal . Reproduction 2011 : 141 ; 749 – 757 . Google Scholar Crossref Search ADS PubMed WorldCat Molaro A , Hodges E , Fang F , Song Q , McCombie WR , Hannon GJ , Smith AD. Sperm methylation profiles reveal features of epigenetic inheritance and evolution in primates . Cell 2011 : 146 ; 1028 – 1040 . Google Scholar Crossref Search ADS WorldCat Montjean D , Ravel C , Benkhalifa M , Cohen-Bacrie P , Berthaut I , Bashamboo A , McElreavey K. Methylation changes in mature sperm deoxyribonucleic acid from oligozoospermic men: assessment of genetic variants and assisted reproductive technology outcome . Fertil Steril 2013 : 100 ; 1241 . Google Scholar Crossref Search ADS PubMed WorldCat Montjean D , Zini A , Ravel C , Belloc S , Dalleac A , Copin H , Boyer P , McElreavey K , Benkhalifa M. Sperm global DNA methylation level: association with semen parameters and genome integrity . Andrology 2015 : 3 ; 235 – 240 . Google Scholar Crossref Search ADS PubMed WorldCat Murphy SK , Itchon-Ramos N , Visco Z , Huang ZQ , Grenier C , Schrott R , Acharya K , Boudreau MH , Price TM , Raburn DJ , et al. Cannabinoid exposure and altered DNA methylation in rat and human sperm . Epigenetics 2018 : 13 ; 1208 – 1221 . Google Scholar Crossref Search ADS PubMed WorldCat Nanassy L , Carrell DT. Abnormal methylation of the promoter of CREM is broadly associated with male factor infertility and poor sperm quality but is improved in sperm selected by density gradient centrifugation . Fertil Steril 2011 a: 95 ; 2310 – 2314 . Google Scholar Crossref Search ADS PubMed WorldCat Nanassy L , Carrell DT. Analysis of the methylation pattern of six gene promoters in sperm of men with abnormal protamination . Asian J Androl 2011 b: 13 ; 342 – 346 . Google Scholar Crossref Search ADS PubMed WorldCat Nasri F , Gharesi-Fard B , Jahromi BN , Farazi-fard MA , Banaei M , Davari M , Ebrahimi S , Anvar Z. Sperm DNA methylation of H19 imprinted gene and male infertility . Andrologia 2017 : 49 . Google Scholar OpenURL Placeholder Text WorldCat Navarro-Costa P , Nogueira P , Carvalho M , Leal F , Cordeiro I , Calhaz-Jorge C , Goncalves J , Plancha CE. Incorrect DNA methylation of the DAZL promoter CpG island associates with defective human sperm(dagger ). Hum Reprod 2010 : 25 ; 2647 – 2654 . Google Scholar Crossref Search ADS PubMed WorldCat Nettersheim D , Biermann K , Gillis AJM , Steger K , Looijenga LHJ , Schorle H. NANOG promoter methylation and expression correlation during normal and malignant human germ cell development . Epigenetics 2011 : 6 ; 114 – 122 . Google Scholar Crossref Search ADS PubMed WorldCat Ni WH , Pan CS , Pan QQ , Fei QJ , Huang XF , Zhang CN. Methylation levels of IGF2 and KCNQ1 in spermatozoa from infertile men are associated with sperm DNA damage . Andrologia 2019 : 51 . Google Scholar OpenURL Placeholder Text WorldCat Oldereid NB , Wennerholm UB , Pinborg A , Loft A , Laivuori H , Petzold M , Romundstad LB , Soderstrom-Anttila V , Bergh C. The effect of paternal factors on perinatal and paediatric outcomes: a systematic review and meta-analysis . Hum Reprod Update 2018 : 24 ; 320 – 389 . Google Scholar Crossref Search ADS PubMed WorldCat Olszewska M , Barciszewska MZ , Fraczek M , Huleyuk N , Chernykh VB , Zastavna D , Barciszewski J , Kurpisz M. Global methylation status of sperm DNA in carriers of chromosome structural aberrations . Asian J Androl 2017 : 19 ; 117 – 124 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Pacheco SE , Houseman EA , Christensen BC , Marsit CJ , Kelsey KT , Sigman M , Boekelheide K. Integrative DNA methylation and gene expression analyses identify DNA packaging and epigenetic regulatory genes associated with low motility sperm . PLoS One 2011 : 6 . Google Scholar OpenURL Placeholder Text WorldCat Pidsley R , Wong CCY , Volta M , Lunnon K , Mill J , Schalkwyk LC. A data-driven approach to preprocessing Illumina 450K methylation array data . BMC Genomics 2013 : 14 . Google Scholar OpenURL Placeholder Text WorldCat Pidsley R , Zotenko E , Peters TJ , Lawrence MG , Risbridger GP , Molloy P , Van Djik S , Muhlhausler B , Stirzaker C , Clark SJ. Critical evaluation of the Illumina MethylationEPIC BeadChip microarray for whole-genome DNA methylation profiling . Genome Biol 2016 : 17 . Google Scholar OpenURL Placeholder Text WorldCat Pilsner JR , Shershebnev A , Medvedeva YA , Suvorov A , Wu HT , Goltsov A , Loukianov E , Andreeva T , Gusev F , Manakhov A , et al. Peripubertal serum dioxin concentrations and subsequent sperm methylome profiles of young Russian adults . Reprod Toxicol 2018 : 78 ; 40 – 49 . Google Scholar Crossref Search ADS PubMed WorldCat Poplinski A , Tuttelmann F , Kanber D , Horsthemke B , Gromoll J. Idiopathic male infertility is strongly associated with aberrant methylation of MEST and IGF2/H19 ICR1 . Int J Androl 2010 : 33 ; 642 – 649 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Potabattula R , Dittrich M , Schorsch M , Hahn T , Haaf T , El Hajj N. Male obesity effects on sperm and next-generation cord blood DNA methylation . PLoS One 2019 : 14 . Google Scholar OpenURL Placeholder Text WorldCat Radford EJ , Ito M , Shi H , Corish JA , Yamazawa K , Isganaitis E , Seisenberger S , Hore TA , Reik W , Erkek S , et al. In utero undernourishment perturbs the adult sperm methylome and intergenerational metabolism . Science 2014 : 345 ; 785 . Google Scholar Crossref Search ADS WorldCat Rakyan VK , Down TA , Thorne NP , Flicek P , Kulesha E , Graf S , Tomazou EM , Backdahl L , Johnson N , Herberth M , et al. An integrated resource for genome-wide identification and analysis of human tissue-specific differentially methylated regions (tDMRs) . Genome Res 2008 : 18 ; 1518 – 1529 . Google Scholar Crossref Search ADS PubMed WorldCat Richardson ME , Bleiziffer A , Tuttelmann F , Gromoll J , Wilkinson MF. Epigenetic regulation of the RHOX homeobox gene cluster and its association with human male infertility . Hum Mol Genet 2014 : 23 ; 12 – 23 . Google Scholar Crossref Search ADS PubMed WorldCat Roberts AL , Gladish N , Gatev E , Jones MJ , Chen Y , MacIsaac JL , Tworoger SS , Austin SB , Tanrikut C , Chavarro JE , et al. Exposure to childhood abuse is associated with human sperm DNA methylation . Transl Psychiatry 2018 : 8 . Google Scholar OpenURL Placeholder Text WorldCat Rotondo JC , Bosi S , Bazzan E , Di Domenico M , De Mattei M , Selvatici R , Patella A , Marci R , Tognon M , Martini F. Methylenetetrahydrofolate reductase gene promoter hypermethylation in semen samples of infertile couples correlates with recurrent spontaneous abortion . Hum Reprod 2012 : 27 ; 3632 – 3638 . Google Scholar Crossref Search ADS PubMed WorldCat Rotondo JC , Selvatici R , Di Domenico M , Marci R , Vesce F , Tognon M , Martini F. Methylation loss at H19 imprinted gene correlates with methylenetetrahydrofolate reductase gene promoter hypermethylation in semen samples from infertile males . Epigenetics 2013 : 8 ; 990 – 997 . Google Scholar Crossref Search ADS PubMed WorldCat Ryan R , Hill S. How to GRADE the quality of the evidence. In: Cochrane Consumers and Communication Group . Melbourne : La Trobe University , 2016 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Saffari A , Silver MJ , Zavattari P , Moi L , Columbano A , Meaburn EL , Dudbridge F. Estimation of a significance threshold for epigenome-wide association studies . Genet Epidemiol 2018 : 42 ; 20 – 33 . Google Scholar Crossref Search ADS PubMed WorldCat Sakai K , Ideta-Otsuka M , Saito H , Hiradate Y , Hara K , Igarashi K , Tanemura K. Effects of doxorubicin on sperm DNA methylation in mouse models of testicular toxicity . Biochem Biophys Res Commun 2018 : 498 ; 674 – 679 . Google Scholar Crossref Search ADS PubMed WorldCat Salas-Huetos A , Moraleda R , Giardina S , Anton E , Blanco J , Salas-Salvado J , Bullo M . Effect of nut consumption on semen quality and functionality in healthy men consuming a Western-style diet: a randomized controlled trial . Am J Clin Nutr 2018 : 108 ; 953 – 962 . Google Scholar Crossref Search ADS PubMed WorldCat Sato A , Hiura H , Okae H , Miyauchi N , Abe Y , Utsunomiya T , Yaegashi N , Arima T. Assessing loss of imprint methylation in sperm from subfertile men using novel methylation polymerase chain reaction Luminex analysis . Fertil Steril 2011 : 95 ; 129 – 134 . Google Scholar Crossref Search ADS PubMed WorldCat Schutte B , El Hajj N , Kuhtz J , Nanda I , Gromoll J , Hahn T , Dittrich M , Schorsch M , Muller T , Haaf T. Broad DNA methylation changes of spermatogenesis, inflammation and immune response-related genes in a subgroup of sperm samples for assisted reproduction . Andrology 2013 : 1 ; 822 – 829 . Google Scholar Crossref Search ADS PubMed WorldCat Sharm R , Harlev A , Agarwal A , Esteves SC. Cigarette smoking and semen quality: a new meta-analysis examining the effect of the 2010 World Health Organization Laboratory methods for the examination of human semen . Eur Urol 2016 : 70 ; 635 – 645 . Google Scholar Crossref Search ADS PubMed WorldCat Shnorhavorian M , Schwartz SM , Stansfeld B , Sadler-Riggleman I , Beck D , Skinner MK. Differential DNA Methylation regions in adult human sperm following adolescent chemotherapy: potential for epigenetic inheritance . PLoS One 2017 : 12 . Google Scholar OpenURL Placeholder Text WorldCat Soubry A , Guo LS , Huang ZQ , Hoyo C , Romanus S , Price T , Murphy SK. Obesity-related DNA methylation at imprinted genes in human sperm: results from the TIEGER study . Clin Epigenet 2016 : 8 . Google Scholar OpenURL Placeholder Text WorldCat Soubry A , Hoyo C , Butt CM , Fieuws S , Price TM , Murphy SK , Stapleton HM. Human exposure to flame-retardants is associated with aberrant DNA methylation at imprinted genes in sperm . Environ Epigenet 2017 : 3 . Google Scholar OpenURL Placeholder Text WorldCat Spinelli P , Latchney SE , Reed JM , Fields A , Baier BS , Lu X , McCall MN , Murphy SP , Mak W , Susiarjo M. Identification of the novel IDO1 imprinted locus and its potential epigenetic role in pregnancy loss . Hum Mol Genet 2019 : 28 ; 662 – 674 . Google Scholar Crossref Search ADS PubMed WorldCat Stoger R. In vivo methylation patterns of the leptin promoter in human and mouse . Epigenetics 2006 : 1 ; 155 – 162 . Google Scholar Crossref Search ADS PubMed WorldCat Sujit KM , Sarkar S , Singh V , Pandey R , Agrawal NK , Trivedi S , Singh K , Gupta G , Rajender S. Genome-wide differential methylation analyses identifies methylation signatures of male infertility . Hum Reprod 2018 : 33 ; 2256 – 2267 . Google Scholar Crossref Search ADS PubMed WorldCat Tang QQ , Pan F , Yang J , Fu ZQ , Lu YW , Wu X , Han XM , Chen MJ , Lu CC , Xia YK , et al. Idiopathic male infertility is strongly associated with aberrant DNA methylation of imprinted loci in sperm: a case-control study . Clin Epigenet 2018 : 10 . Google Scholar OpenURL Placeholder Text WorldCat Tavalaee M , Bahreinian M , Barekat F , Abbasi H , Nasr-Esfahani MH. Effect of varicocelectomy on sperm functional characteristics and DNA methylation . Andrologia 2015 : 47 ; 904 – 909 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Tavalaee M , Razavi S , Nasr-Esfahani MH. Influence of sperm chromatin anomalies on assisted reproductive technology outcome . Fertil Steril 2009 : 91 ; 1119 – 1126 . Google Scholar Crossref Search ADS PubMed WorldCat Tian MP , Bao HQ , Martin FL , Zhang J , Liu LP , Huang QY , Shen HQ. Association of DNA methylation and mitochondrial DNA Copy number with human semen quality (vol 91, 101, 2014) . Biol Reprod 2015 : 92 . Google Scholar OpenURL Placeholder Text WorldCat Tian MP , Liu LP , Zhang J , Huang QY , Shen HQ. Positive association of low-level environmental phthalate exposure with sperm motility was mediated by DNA methylation: a pilot study . Chemosphere 2019 : 220 ; 459 – 467 . Google Scholar Crossref Search ADS PubMed WorldCat Tsai PC , Bell JT. Power and sample size estimation for epigenome-wide association scans to detect differential DNA methylation . Int J Epidemiol 2015 : 44 ; 1429 – 1441 . Google Scholar Crossref Search ADS PubMed WorldCat Tunc O , Tremellen K. Oxidative DNA damage impairs global sperm DNA methylation in infertile men . J Assist Reprod Genet 2009 : 26 ; 537 – 544 . Google Scholar Crossref Search ADS PubMed WorldCat Uppangala S , Mathai SE , Salian SR , Kumar D , Singh VJ , D’Souza F , Kalthur G , Kamath A , Adiga SK. Sperm chromatin immaturity observed in short abstinence ejaculates affects DNA integrity and longevity in vitro . PLoS One 2016 : 11 . Google Scholar OpenURL Placeholder Text WorldCat Urdinguio RG , Bayon GF , Dmitrijeva M , Torano EG , Bravo C , Fraga MF , Bassas L , Larriba S , Fernandez AF. Aberrant DNA methylation patterns of spermatozoa in men with unexplained infertility . Hum Reprod 2015 : 30 ; 1014 – 1028 . Google Scholar Crossref Search ADS PubMed WorldCat Wahl S, , Drong A, , Lehne B, , Loh M , Scott WR , Kunze S , Tsai PC , Ried JS , Zhang WH , Yang YW , et al. Epigenome-wide association study of body mass index, and the adverse outcomes of adiposity . Nature 2017 : 541 ; 81 . Google Scholar Crossref Search ADS PubMed WorldCat Wang T , Gao H , Li W , Liu C. Essential role of histone replacement and modifications in male fertility . Front Genet 2019 : 10 . Google Scholar OpenURL Placeholder Text WorldCat Watkins AJ , Dias I , Tsuro H , Allen D , Emes RD , Moreton J , Wilson R , Ingram RJM , Sinclair KD. Paternal diet programs offspring health through sperm- and seminal plasma-specific pathways in mice . Proc Natl Acad Sci USA 2018 : 115 ; 10064 – 10069 . Google Scholar Crossref Search ADS PubMed WorldCat Wei YC , Yang CR , Wei YP , Zhao ZA , Hou Y , Schatten H , Sun QY. Paternally induced transgenerational inheritance of susceptibility to diabetes in mammals . Proc Natl Acad Sci USA 2014 : 111 ; 1873 – 1878 . Google Scholar Crossref Search ADS PubMed WorldCat Wu CL , Ding XF , Tan HP , Li HG , Xiong CL. Alterations of testis-specific promoter methylation in cell-free seminal deoxyribonucleic acid of idiopathic nonobstructive azoospermic men with different testicular phenotypes . Fertil Steril 2016 : 106 ; 1331 – 1337 . Google Scholar Crossref Search ADS PubMed WorldCat Wu HT , Estill MS , Shershebnev A , Suvorov A , Krawetz SA , Whitcomb BW , Dinnie H , Rahil T , Sites CK , Pilsner JR. Preconception urinary phthalate concentrations and sperm DNA methylation profiles among men undergoing IVF treatment: a cross-sectional study . Hum Reprod 2017 : 32 ; 2159 – 2169 . Google Scholar Crossref Search ADS PubMed WorldCat Wu W , Shen OX , Qin YF , Niu XB , Lu CC , Xia YK , Song L , Wang SL , Wang XR. Idiopathic male infertility is strongly associated with aberrant promoter methylation of methylenetetrahydrofolate reductase (MTHFR) . PLoS One 2010 : 5 . Google Scholar OpenURL Placeholder Text WorldCat Xu AM , Hua YB , Zhang JZ , Chen W , Zhao K , Xi W , Wang HN , Fang JZ , Su SF , Tang M , et al. Abnormal hypermethylation of the VDAC2 promoter is a potential cause of idiopathic asthenospermia in men . Sci Rep 2016 : 6 . Google Scholar OpenURL Placeholder Text WorldCat Xu J , Zhang A , Zhang Z , Wang P , Qian Y , He L , Shi H , Xing Q , Du J. DNA methylation levels of imprinted and nonimprinted genes DMRs associated with defective human spermatozoa . Andrologia 2016 : 48 ; 939 – 947 . Google Scholar Crossref Search ADS PubMed WorldCat Youngson NA , Lecomte V , Maloney CA , Leung P , Liu J , Hesson LB , Luciani F , Krause L , Morris MJ. Obesity-induced sperm DNA methylation changes at satellite repeats are reprogrammed in rat offspring . Asian J Androl 2016 : 18 ; 930 – 936 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Yu BL , Zhou H , Liu M , Zheng T , Jiang L , Zhao M , Xu XX , Huang ZF. Epigenetic alterations in density selected human spermatozoa for assisted reproduction . PLoS One 2015 : 10 . Google Scholar OpenURL Placeholder Text WorldCat Zeng J , Nagrajan HK , Yi SV. Fundamental diversity of human CpG islands at multiple biological levels . Epigenetics 2014 : 9 ; 483 – 491 . Google Scholar Crossref Search ADS PubMed WorldCat Zeschnigk M , Martin M , Betzl G , Kalbe A , Sirsch C , Buiting K , Gross S , Fritzilas E , Frey B , Rahmann S , et al. Massive parallel bisulfite sequencing of CG-rich DNA fragments reveals that methylation of many X-chromosomal CpG islands in female blood DNA is incomplete . Hum Mol Genet 2009 : 18 ; 1439 – 1448 . Google Scholar Crossref Search ADS PubMed WorldCat Zhang CW , Xue P , Gao LZ , Chen X , Lin KB , Yang XY , Dai YF , Xu EY. Highly conserved epigenetic regulation of BOULE and DAZL is associated with human fertility . FASEB J 2016 : 30 ; 3424 – 3440 . Google Scholar Crossref Search ADS PubMed WorldCat Zhang T , Zhang L , Yuan QG , Wang XL , Zhang Y , Wang JH , Zhang YP. The noninvasive detection of RAR beta 2 promoter methylation for the diagnosis of prostate cancer . Cell Biochem Biophys 2015 : 71 ; 925 – 930 . Google Scholar Crossref Search ADS PubMed WorldCat Zhou JH , Zhou QZ , Lyu XM , Zhu T , Chen ZJ , Chen MK , Xia H , Wang CY , Qi T , Li X , et al. The expression of cysteine-rich secretory protein 2 (CRISP2) and its specific regulator miR-27b in the spermatozoa of patients with asthenozoospermia . Biol Reprod 2015 : 92 . Google Scholar OpenURL Placeholder Text WorldCat Author notes Fredrika Åsenius and Amy F. Danson contributed equally to this study. © The Author(s) 2020. Published by Oxford University Press on behalf of European Society of Human Reproduction and Embryology. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - DNA methylation in human sperm: a systematic review
JF - Human Reproduction Update
DO - 10.1093/humupd/dmaa025
DA - 2020-11-01
UR - https://www.deepdyve.com/lp/oxford-university-press/dna-methylation-in-human-sperm-a-systematic-review-OqZz02lCCY
SP - 841
EP - 873
VL - 26
IS - 6
DP - DeepDyve
ER -