Genome-Wide Methylation Analyses of Human Sperm Unravel Novel Differentially Methylated Regions in Asthenozoospermia

Abstract

Aims & objectives: To investigate DNA methylation patterns in asthenozoospermic and normozoospermic sperm and to explore the potential roles of differential methylations in the etiology of the disease. Materials & methods: The authors performed whole-genome bisulfite sequencing analysis between normozoospermic controls and asthenozoospermic individuals. Results: The authors identified 238 significant differentially methylated regions. These differentially methylated regions were annotated to 114 protein-coding genes, with many genes showing associations with spermatogenesis, sperm motility etc. Conclusion: There are plenty of genomic regions exhibiting altered DNA methylation in asthenozoospermia, a number of which are located within or adjacent to sperm-related genes, suggesting novel methylation markers of asthenozoospermia and potential epigenetic regulation mechanisms through DNA methylation in the disease.

Keywords::

• asthenozoospermia
• DNA methylation
• epigenetics
• male infertility
• sperm
• WGBS

Infertility or subfertility has aroused wide social concerns in recent decades, and around half of the subfertile couples are affected by male factor infertility [1]. Around 80% of male infertility cases are related to asthenozoospermia with variable degrees of severity [2]. Asthenozoospermia is absent or reduced sperm motility in the fresh ejaculate, which specifically means that the proportion of progressively motile spermatozoa is below the lower limit (32%) [3]. Asthenozoospermia itself is a complex and multifactorial disease that may result from genetic, epigenetic or environmental factors. Multiple genetic factors contributing to asthenozoospermia have been identified and summarized [4–6]. However, a considerable proportion of the cases were still diagnosed with unexplained causes. Apart from genetic factors, epigenetic mechanisms, including DNA methylation, may represent an attractive explanation for the etiology of asthenozoospermia.

A widely accepted definition of epigenetics is the heritable changes of gene expression without a change of DNA sequence, of which DNA methylation is the most prominent and extensively studied epigenetic mechanism [7,8]. Methylation of the fifth atomic position of cytosines on genomic DNA is the form of DNA methylation that has been most commonly studied so far. In humans or other mammals, methylation of the fifth atomic position of cytosines primarily occurs at CpG dinucleotides, although methylation of non-CpG cytosines (CHG or CHH when H stands for A, G or T) also exists in some cases [9,10]. CpGs are functionally most important due to their symmetrical structures, which makes them peculiar epigenetic marks that can be passed on to daughter cells after DNA replication [11–14]. Through interplay with other epigenetic mechanisms, including histone modifications and transcription factors, DNA methylation pattern acts as another layer of regulation of gene expression and securer of genome stability [15–18]. In principle, methylated CpGs at regulatory genomic regions are negatively associated with gene expression, although increasing evidence in recent studies is also suggesting a positive relationship in some situations depending on their relative locations on the genome [19–22]. Importantly, DNA methylation has been reported to play vital roles in crucial development and normal physiology, including spermatogenesis and embryogenesis [23–25].

The impacts of DNA methylation on sperm development can be manifested in different aspects. Two major DNA methylation reprogramming events take place during the human developmental process, with the first wave occurring in primordial germ cells during early embryogenesis and the second in the preimplantation embryo following fertilization. Genomewide DNA methylation patterns are erased and later on re-established in both periods, which are windows of susceptibility that may introduce epigenetic errors and thus affect fertility [26]. During the long and continuous process of human sperm production, impairment of spermatogenesis or abnormal semen parameters have been reported to be associated with toxins, smoking, diet or other environmental factors [27–31]. It has been reported that some influences may last for generations, probably associated with altered DNA methylation marks in spermatozoa [32–34]. In the past 20 years, growing evidence has supported the idea that DNA methylation may also contribute to male infertility, which profits from the wide application of microarray-based technologies and next-generation sequencing. Aberrant DNA methylation has been reported to be associated with male subfertility and abnormal semen parameters, and plenty of differential methylation sites or regions have been identified in several disease states, including teratozoospermia and oligozoospermia [35–44]. For instance, several candidate gene studies of DNA methylation have found an association of aberrant methylation of H19 with male infertility, including teratozoospermia and oligozoospermia [45,46]. A methylation study on the spermatogenesis-associated genes (SPATA4, SPATA5 and SPATA6) showed that hypermethylation of these genes correlated with oligozoospermic infertility [47]. Another study demonstrated increased defective methylation of DAZL promoter CpG islands in oligoasthenoteratozoospermia patients [48]. However, only very few genes have been investigated multiple times and there has been little overlap of results from different genomewide DNA methylation studies. Moreover, many studies have used testis or blood to investigate fertility issues and only a limited number of studies have directly performed methylation profiling of human sperm samples. In studies that investigated reduced fertility, different types of sperm abnormalities may not have been distinguished, which may have led to failure in capturing subtle but functional changes in DNA methylation patterns of a certain type of abnormality. Besides, most of the previous studies used array-based or targeted-sequencing methods and thus were restricted to only a part of the human genome.

In this study, the authors performed the gold-standard whole-genome bisulfite sequencing (WGBS) on normozoospermic and asthenozoospermic sperm samples to specifically investigate whether DNA methylation is involved in the etiology of asthenozoospermia and to build an atlas of asthenozoospermia that provides a valuable reference and resources for future studies.

Materials & methods

Ethics approval & sample acquisition

This research was approved by all respective institutional research and ethics boards. The participants in this study were selected from volunteers or patients seeking fertility treatment. Informed written consent was obtained from each participant. Participant information was de-identified before analysis. Semen samples were collected after at least 3 days of abstinence. After liquefaction for 20–30 min at 37°C, semen analyses of all samples were performed according to the standard WHO recommendations [3]. To reduce the contamination of DNA from nonsperm cells during the sample collection, the authors excluded the semen samples for which the nonsperm cell concentration exceeded 1%. The authors further microscopically examined all semen samples. Consequently, nonsperm cells were far fewer than the spermatozoa, and the spermatozoa/nonsperm cell ratios of our samples ranged from about 150 to 1000. Additionally, a recent study showed that seminal cfDNA (e.g., from broken cells) has little contaminating impact on sperm DNA methylation analysis [49]. Thus, the methylation marks that were extracted from the seminal samples should represent the true epigenetic feature of spermatozoa. In total, 12 semen samples were collected and were divided into two groups: normozoospermic control samples (n = 5) and asthenozoospermic samples (n = 7). The semen parameters of the different study groups are summarized in Table 1.

Table 1. Semen parameters of different study groups.

Parameters	Normozoospermia (n = 5)		Asthenozoospermia (n = 7)
	Mean (± standard deviation)	Range	Mean (± standard deviation)	Range
Age (years)	34.00 (± 6.04)	27–42	35.43 (± 7.04)	30–50
Semen volume (ml)	3.16 (± 1.25)	1.8–5.0	3.71 (± 1.35)	2.4–5.3
Sperm concentration (×10^6/ml)	83.50 (± 31.04)	34.7–115.1	73.96 (± 40.30)	33.2–150.0
Motility (% progressive)	71.20 (± 10.55)	61–85	10.39 (± 6.92)	5–23
Percentage of normal sperm form (%)	9.48 (± 3.52)	5.9–13.8	5.04 (± 2.32)	2.8–8.7
Vitality (%)	91.00 (± 3.24)	88–95	53.57 (± 24.37)	15–84

Sperm DNA isolation

For bisulfite sequencing, the total genomic DNA of all samples was isolated. Precipitated clean sperm samples were lysed with a lysis buffer containing a final concentration of 100 mM Tris·Cl (pH 8.0), 10 mM EDTA, 500 mM NaCl, 1% SDS and 2% ß-mercaptoethanol and were incubated together with proteinase K at 55°C for 2 h. DNA was then isolated using the QIAamp^® DNA Mini Kit (Qiagen, MD, USA) according to the manufacturer’s protocol. Inspection of DNA degradation or contamination was performed by running agarose gels. The purity of DNA was checked using the NanoPhotometer^® spectrophotometer (Implen, CA, USA), and concentration was measured by using Qubit^® DNA Assay Kit in Qubit^® 2.0 fluorometer (Life Technologies, CA, USA).

Library construction & WGBS

A total amount of 100 ng genomic DNA spiked with 0.5 ng lambda DNA were fragmented by sonication to 200–300 bp with the Covaris S220 (Covaris LLC, MA, USA). The DNA fragments then underwent bisulfite treatment with the commercial EZ DNA Methylation-Gold™ Kit (Zymo Research, CA, USA) before library construction. The concentration and quality of the library were assessed on the Agilent Bioanalyzer 2100 system. Pair-end sequencing of samples was performed on the Illumina NovaSeq platform (Illumina, CA, USA) and 150 bp reads were generated. Image analysis and base calling were performed with the Illumina CASAVA pipeline.

WGBS data analysis

FastQC v0.11.8 was used to check the sequence quality and generate quality reports. Low-quality bases and the adaptor sequences were trimmed with Trim Galore v0.6.6 [50], and the cleaned data with high-quality reads for each sample were used for the next alignment step. Cleaned reads were mapped to the human reference genome (GRCh38) using Bismark v0.23.0 with the following alignment parameters: “bismark -X 600 -1 ${READ1} -2 ${READ2}” [51]. Duplicates were removed using sambamba v0.8.0 [52]. Methylation information was extracted using bismark_methylation_extractor from Bismark v0.23.0 for downstream analysis.

Differentially methylated region calling, gene annotation & gene ontology analysis

The DMRichR v1.7.2 package with R software v4.1.1 was used to identify differentially methylated regions (DMRs) and to annotate DMR-related genes [53–55]. CpG sites with at least 5× sequencing coverage across all samples were retained for DMR calling. The minimum number of CpGs required for a DMR was set to five. Statistical algorithms from Bioconductor R package “dmrseq” were leveraged and DMRs were identified in a two-step approach [55], which accounts for both interindividual and inter-CpG variability across the entire genome. Methylation data were first smoothed and candidate DMRs with a difference were assembled; then a regional statistic was estimated for each DMR and permutation analysis of a pooled null distribution was performed to identify significant DMRs. The absolute value for the cutoff of the single CpG coefficient was set at 0.1 for candidate region discovery. After computation, p-values, methylation differences and other statistics were assigned to each DMR. DMRs that passed the p < 0.05 threshold were selected as significant DMRs. The authors performed gene ontology (GO) analysis of biological process, cellular component and molecular function using the Bioconductor R package ‘rGREAT’ [56]. Two statistical tests, the binomial test over genomic regions and the hypergeometric test over genes, were conducted in GREAT analysis. These two statistical tests would compensate for each other, thus providing an accurate picture of annotation enrichments.

Results

Global DNA methylation profiles in normozoospermic & asthenozoospermic sperm

To obtain the DNA methylation profiling of both normozoospermic and asthenozoospermic human sperm and compare their methylomes, the authors performed WGBS on five normozoospermic and seven asthenozoospermic samples. Semen parameters of the samples are summarized in Table 1. The percentages of progressively motile sperm in the asthenozoospermic group were evidently below the normal threshold (<32%). For each sample, about 330 million paired-end reads were obtained and some basic statistics of the WGBS experiment are summarized in Supplementary Table 1. The majority of cytosine methylations occurred in the CpG context, while non-CpG cytosine methylation (CHG or CHH) was observed at a very low level (less than 1%). In this case, the authors focused only on CpG cytosine methylation across the whole genome in later analysis. To further inspect the quality of the WGBS experiment process in our study, we calculated and visualized the distribution of methylation values for all CpGs of each sample, where a bimodal distribution would suggest good experimental quality. Figure 1A shows the distribution of methylation values of one normal control sample, and such distributions of all other samples were also examined and are shown in Supplementary Figure 1. Upon extraction of methylation information, only CpGs that were covered by at least five reads in all 12 samples were retained for further analysis. A total of 23,725,450 CpG sites were retained and used for global methylation analysis as well as the regional methylation analysis. Then, the authors first explored the global level of DNA methylation in the two study groups. For each sample, the average CpG methylation values were computed and plotted, as in Figure 1B. The global methylation differences between the groups were then tested through one-way analysis of variance, and no significant differences were observed in terms of global DNA methylation between the groups (p > 0.05; Figure 1C). Nevertheless, since this measure of global methylation refers to a composite average derived from all the detected CpG sites across the entire genome, potential alterations in certain areas may have been covered up. Therefore, the authors performed a further investigation to identify regional methylation changes across the whole genome.

Figure 1. Global DNA methylation profiles and asthenozoospermia-associated differentially methylated regions.

(A) Histogram of methylation values for all CpGs in one representative sample. As expected, it showed bimodal distributions, suggesting a high experimental quality. (B) Global methylation levels of all samples. The ast1-7 and nor1-5 are sample IDs of asthenozoospermia and normozoospermia, respectively. (C) No significant global DNA methylation differences were found between groups (p > 0.05). (D) Heatmap of methylation data z-scores of 238 significantly differentially methylated regions. (E) Distribution of candidate DMRs across chromosomes, where the dotted line represents a threshold of p < 0.05, and dots above the line were chosen as significant DMRs.

DMR: Differentially methylated region.

Identification of asthenozoospermia-associated DMRs

Despite the lack of global methylation difference between the two groups, the regional analysis suggested that there are plenty of DMRs across the genome. With a rigorous permutation-based algorithm from ‘dmrseq’ the authors successfully identified 238 significant DMRs between the normozoospermic and asthenozoospermic groups (Figure 1D & E). This methodology enables the inference of DMRs with good specificity and sensitivity. In this smoothing-based approach, CpGs with higher coverage are given a higher weight and are used to infer the methylation level of adjacent CpGs that have lower coverage. In addition, this approach performs statistical testing on the DMRs themselves with a region statistic for each DMR rather than testing for the differences in a single CpG, which are then assembled into DMRs. Thus, this analysis offers accurate false discovery rate control. For the samples, the absolute value for the cutoff of the single CpG coefficient that was used to discover candidate regions was set to 0.1, with a minimum of five CpGs required for a region. The maximum distance between two neighboring CpGs that were included in the same DMR was no more than 1000 bp. A heatmap of methylation data z-scores of the 238 DMRs is illustrated in Figure 1D. In the heatmap, each row represents a DMR and each column represents a sample. Overall, the authors identified 6726 candidate DMRs across the genome and then plotted the chromosome distribution of all candidate DMRs using a Manhattan plot, as shown in Figure 1E. The 238 significant DMRs above the dotted line (p < 0.05) from the Manhattan plot were chosen for further analyses. Methylation differences and genomic annotations of all candidate DMRs were also plotted and were collected in Supplementary Figure 2.

The authors then defined hypermethylation and hypomethylation as increased or decreased methylation levels in asthenozoospermic sperm compared with normozoospermic sperm, respectively. Overall, they identified 69% hypermethylated DMRs (n = 164) and 31% hypomethylated DMRs (n = 74; Figure 2A). They further grouped the DMRs into distinct levels of methylation difference to look into the degree of alteration (Figure 2B). Most of the DMRs fell into a difference of 10–20 (83.2% of the total DMRs), while around 5.5% of the DMRs showed a higher extent of more than 20 methylation difference.

Figure 2. Distribution of the differentially methylated regions and their genomic locations.

(A) Count of hypermethylated DMRs and hypomethylated DMRs. (B) Histogram showing the distribution of degrees of methylation difference. (C) Percentage of DMRs in distinct genomic contexts. (D) Relative distances of DMRs to the TSS.

DMR: Differentially methylated region; TSS: Transcription start site.

Genomic locations of asthenozoospermia-associated DMRs

Since DNA methylation plays a crucial role in the regulation of gene expression in a genomic context-dependent manner, the authors then investigated the locations of DMRs relative to distinct genomic features (promoter, 5′UTR, 3′UTR, exon etc.). As shown in Figure 2C, differential methylations were most frequent in the distal intergenic regions (44.54%), followed by introns (29.41%) and exons (13.44%). Also, about 9.24% of the DMRs were located near promoters, suggesting potential regulation of gene expression. Less frequent locations were 3′UTR (2.52%) and 5′UTR (0.84%). In addition, the distances of DMRs relative to the transcription start site are shown in Figure 2D, and a small number of DMRs were located quite near the transcription start site (<3 kb).

Annotation of DMRs & functional-based analysis of DMR-associated genes

To further explore the biological significance of these methylation changes between groups, the authors next annotated the DMRs to find out whether they were associated with functional genes. Collectively, these 238 DMRs were annotated to 187 genes, with 27 genes harboring two or more DMRs, including 114 protein-coding genes, 65 ncRNA genes and eight pseudogenes. They first took a closer look at the protein-coding genes. As methylations of CpGs in promoter regions of coding genes are commonly associated with gene transcription, the authors summarized those protein-coding genes associated with DMR(s) located in their promoter regions (Table 2). As shown in Table 2, there were significant methylation differences in promoter regions of 17 protein-coding genes, suggesting that expression of these genes in human sperm may be associated with DNA methylation. Strikingly, the majority of these genes have been shown to play important roles in the mitotic cell cycle, transcriptional regulation and sperm functions, indicating potential involvement in sperm production. In addition, some of the genes, including IK, MAP1S, XRRA1 and AKIRIN1, also showed high expression in testis, according to previous genomewide studies [57,58]. Importantly, TDRP, which encodes the testis development-related protein, has been reported to function in sperm motility and may play a role in spermatogenesis [59,60]. Since low motility is the main manifestation of asthenozoospermia, TDRP may be an interesting candidate gene that is worth further investigation.

Table 2. Summary of genes harboring differentially methylated regions in their promoter regions.

Gene symbol	Protein	Genomic location	Methylation direction^†	Methylation difference	Function summary
ACOT7	Acyl-CoA Thioesterase 7	Promoter	Hypomethylation	-15	Hydrolyze the CoA thioester of palmitoyl-CoA and other long-chain fatty acids
IK	IK Cytokine	Promoter	Hypomethylation	-13	Required for normal mitotic cell cycle progression
ZNF442	Zinc Finger Protein 442	Promoter	Hypomethylation	-19	May be involved in transcriptional regulation
MAP1S	Microtubule Associated Protein 1S	Promoter	Hypermethylation	16	Microtubule-associated protein that mediates aggregation of mitochondria, resulting in cell death and genomic destruction
TDRP	Testis Development Related Protein	Promoter	Hypomethylation	-14	Contributes to normal sperm motility
TPO	Thyroid Peroxidase	Promoter	Hypermethylation	11	Plays a central role in thyroid gland function
XRRA1	X-ray Radiation Resistance Associated 1	Promoter	Hypomethylation	-16	May be involved in the response of cells to x-ray radiation
ENO1	Enolase 1	Promoter	Hypomethylation	-14	Glycolytic enzyme that catalyzes the conversion of 2-phosphoglycerate to phosphoenolpyruvate
OR52B4	Olfactory Receptor Family 52 Subfamily B Member 4	Promoter	Hypermethylation	12	Odorant receptor
CHAMP1	Chromosome Alignment Maintaining Phosphoprotein 1	Promoter	Hypomethylation	-12	A zinc finger protein that functions as a regulator of chromosome segregation in mitosis
SMYD2	SET And MYND Domain Containing 2	Promoter	Hypomethylation	-11	Protein-lysine N-methyltransferase that methylates both histones and nonhistone proteins
TUBGCP5	Tubulin Gamma Complex Associated Protein 5	Promoter	Hypomethylation	-10	γ-Tubulin complex, necessary for microtubule nucleation at the centrosome
OR52K2	Olfactory Receptor Family 52 Subfamily K Member 2	Promoter	Hypermethylaton	12	Odorant receptor
AKIRIN1	Akirin 1	Promoter	Hypomethylation	-13	May regulate chemotaxis of both macrophages and myoblasts
PSTPIP1	Proline-Serine-Threonine Phosphatase Interacting Protein 1	Promoter	Hypermethylaton	16	A cytoskeletal protein involved in regulation of the actin cytoskeleton
RBM15B	RNA Binding Motif Protein 15B	Promoter	Hypomethylation	-13	An RNA-binding protein that regulates several different processes such as alternative splicing
RPS25	Ribosomal Protein S25	Promoter	Hypomethylation	-14	A ribosomal protein that is a component of the 40S subunit

† Methylation direction is the change of methylation in asthenozoospermia compared with normal controls.

In addition, increasing evidence is supporting the roles of DNA methylation even when they do not show in promoter regions. For instance, it has been reported that some highly expressed genes also show high DNA methylation levels in their gene bodies. In this case, attention should also be paid to those DMRs that locate in gene bodies or intergenic regions. The authors then did a thorough search of all the protein-coding genes in the literature and databases and found that a number of those genes (i.e., LRGUK, DZIP1, DNAAF5, SRPK1 and TDRP) had been directly involved in spermatogenesis or in maintaining sperm functions. For instance, DZIP1 encodes for a centrosome-related protein and its mutation is related to morphological abnormalities of the sperm flagella, which affects the motility of sperm [61]. DNAAF5 is involved in the assembly or stability of axonemal dynein arms, which is also related to the functions of the sperm flagella [62]. These genes are summarized in Table 3, and the DMRs regarding these genes are illustrated in Figure 3A–E.

Table 3. Differentially methylated region(s)-harboring genes involved in sperm functions or spermatogenesis.

Gene symbol	Protein	Genomic location	Methylation direction^†	Methylation difference	Function summary
LRGUK	Leucine-Rich Repeat and Guanylate Kinase Domain-Containing Protein	Exon	Hypermethylation	16	Involved in multiple aspects of sperm assembly, including acrosome attachment, shaping of the sperm head and early aspects of axoneme development
DZIP1	DAZ Interacting Zinc Finger Protein 1	Intron	Hypomethylation	-11	Participates in spermatogenesis and has a role in primary cilium formation
DNAAF5	Dynein Axonemal Assembly Factor 5	Intron	Hypomethylation	-16	Essential for the preassembly or stability of axonemal dynein arms and plays a crucial role in cilium motility
SRPK1	Serine/Arginine-Rich Splicing Factor Kinase 1	Distal Intergenic	Hypomethylation	-14	Can influence chromatin reorganization in somatic and sperm cells and cell cycle progression
TDRP	Testis Development Related Protein	Promoter	Hypomethylation	-14	Contributes to normal sperm motility

† Methylation direction is the change of methylation in asthenozoospermia compared with normal controls.

Figure 3. Methylation levels of selected representative differentially methylated regions and gene ontology analysis of annotated genes.

(A–E) Plots of five selected DMRs with regard to genes functioning in sperm motility or spermatogenesis. Each dot represents the methylation level of an individual CpG in a single sample, and the size of the dot is representative of coverage. The lines represent smoothed methylation levels for each sample, either normozoospermic controls (blue) or asthenozoospermic samples (red). CpG annotations are shown below the plot. If the DMR is located in the gene body of a gene, the exon of the gene is also shown below. (F) Gene ontology analysis of biological process, cellular component and molecular function of DMR-annotated genes.

DMR: Differentially methylated region.

GO analysis of DMR-associated genes

Apart from those spermatogenesis- or motility-related genes, more than 60 of the DMR-associated genes have been reported to show altered expression in the testis of azoospermic or teratozoospermia patient samples, or to have restricted expression toward testis, suggesting that these genes may also indirectly function in male fertility [57,58]. Some of these genes are involved in transcription regulation, cell proliferation and microtubules, such as MAF and TUBA3E [63–65]. To further investigate all the DMR-associated genes, the authors then performed GO analysis of the biological process, cellular component and molecular function (Figure 3F). For biological process, the differentially methylated genes were mostly clustered to genes related to protein complex oligomerization, purine ribonucleotide metabolic process, ribonucleotide metabolic process, activation of protein kinase A activity etc. For cellular component, these genes were mostly enriched in genes involved in voltage-gated potassium channel complex. For molecular function, the top three related functions were connexin binding, calcium- and calmodulin-responsive adenylate cyclase activity and voltage-gated potassium channel activity. Notably, cAMP/protein kinase A and calcium pathways are shown to be two general sperm motility-regulating signaling pathways [66,67]. The results of the present study suggest that DNA methylation may also be involved in sperm motility through epigenetic regulation of genes in these two signaling pathways.

Discussion

Asthenozoospermia contributes to a large proportion of subfertility cases and thus affects many couples [1]. While some genetic factors have already been discovered in asthenozoospermia and its resulting male infertility, there are still many pieces missing from the puzzle regarding its etiology [2]. The emerging picture is that DNA methylation may play an important role in human developmental processes relevant to male fertility or spermatogenesis [35,36,39,41]. Several studies examined sperm DNA methylation profiles of specific genes or genomic regions from males with unexplained infertility, varicocele, oligozoospermia and many other types of abnormal sperm conditions [35,39,40,68]. However, these studies only targeted limited gene loci or obtained information from a restricted number of CpG sites on the human genome [69], and few studies have directly investigated sperm DNA methylation in asthenozoospermia. In this study, the authors performed WGBS of sperm from asthenozoospermia patients as well as normozoospermic controls to get a comprehensive map of DNA methylation patterns in both diseased and healthy states. With the single-base resolution method, the authors were able to identify asthenozoospermia-related DMRs and their associated genes. To their knowledge, this is the first study to directly compare sperm DNA methylation patterns from asthenozoospermia and normozoospermic controls and report regional specific changes at a genomewide level.

Consistent with previous studies regarding sperm DNA methylations [35,36,39,44,70], the findings of the present study strengthen the association between DNA methylation and male fertility. Although the global levels of DNA methylation in the two study groups showed no significant difference, the authors were able to identify 238 significant DMRs from the detected methylated CpG sites, indicating the presence of differential DNA methylation in the genomic region from asthenozoospermia patients. Given that clustered CpGs are more likely functionally important in impacting the gene activity during development [71], the present novel analysis thus focused on the genomic regional analysis of showing clustered differential methylation instead of single CpG sites. The data indicated that these DMRs are located most frequently in distal intergenic regions and gene bodies (introns and exons), though nearly 9% of the DMRs were also located near promoters.

These promoter-located DMRs could be further annotated to 17 protein-coding genes, including several cell cycle- or mitosis-related genes such as IK, MAP1S, CHAMP1, TUBGCP5 and PSTPIP1, which is reasonable, considering the importance of mitosis during sperm production. Notably, the gene PSTPIP1 was also reported to overlap differentially methylated CpGs in a previous study comparing sperm methylomes of fertile men and in vitro fertilization patients [36], which coincides with the present study’s results. Further screening and validation of these genes may be promising and rewarding in the field of reproductive research. On the other hand, while these DMRs in promoter regions most likely suggest potential roles in the regulation of relevant gene expression, the DMRs in gene bodies or distal intergenic regions cannot be overlooked, given that increasing evidence is supporting the idea that DNA methylation can function distinctly in different genomic contexts [20].

The authors annotated hundreds of protein-coding genes associated with these asthenozoospermia-related DMRs. Particularly, several sperm-related genes, such as LRGUK, DZIP1, DNAAF5, SRPK1 and TDRP, have been identified [59,61,62,72,73], indicating that expression levels of these genes in sperm may be under the regulation of DNA methylation-dependent mechanisms. For example, TDRP was reported to be expressed in spermatogenic cells and its expression was lower in the testis tissues of azoospermic men [59,60]; DZIP1 had a crucial function in sperm motility and its mutation can lead to morphological abnormalities of the sperm flagella and may induce asthenozoospermia [61]; DNAAF5 also encodes for a protein involved in the function of sperm flagella and thus may affect sperm motility [62]. Apart from these genes with known direct or indirect functions in sperm, there were also a large number of DMR-associated genes showing restricted expression in the testis or altered expression in the testis of patients with abnormal sperm, suggesting that these genes may also be related to spermatozoon development and male fertility [57]. Notably, GO analysis of the DMR-related genes has shown an enrichment in the activation of PKA activity, which is consistent with the previous evidence that the cAMP/protein kinase A pathway is important in regulating sperm motility [66,67]. Taken together, the current findings provide promising targets for further investigation of asthenozoospermia and male infertility.

Most of the current body of research on human spermatozoal DNA methylations is limited, either focusing on semen parameters or simply using array-based assays [35,40]. In terms of asthenozoospermia, only one promoter-targeted sequencing study has identified a few CpGs/DMRs in promoter regions, thus providing only limited information [69]. In addition, the regional analysis of CpGs may provide more prominent functional insights into related genes, compared with a single CpG variation analysis. The present study shows that there are plenty of unique DMRs in asthenozoospermia and that DNA methylation alterations are multifacetedly associated with the abnormality, which may be primarily through the regulation of crucial genes involved in sperm production, either directly or indirectly. These sperm-related genes may be promising candidate targets for further investigation of the etiology of asthenozoospermia, and experimental validation of expressions of these genes in sperm with low motility can provide more solid evidence of these results.

Nevertheless, there are also a few limitations of this study. First, the sample size was relatively small. Practically, collecting sperm samples and extracting DNA are rather difficult to achieve in a short period of time. Above all, it usually takes a much longer time to select and collect suitable samples due to the specialty of semen samples, especially for those asthenozoospermia cases. In addition, the isolation of genomic DNA of mature sperm remains a challenge, since DNA is very tightly packed in sperm and there is no perfect experimental protocol to address this issue. Therefore, considering the accessibility of sperm samples and the loss rate of sperm sample DNA, the authors maximized the utilization of a sperm sample as much as possible to achieve the current results. Certainly, the validation of DMRs may be an indispensable step, which is supposed to take at least a few more months, given the challenges mentioned above. In fact, a large-scale multi-omics (i.e., bisulfite sequencing, RNA sequencing and assay for transposase-accessible chromatin using sequencing) cohort analysis as well as loci-specific methylation status validation experiments on those selected promising target genes will be implemented in the next step of the authors’ research plan.

Second, considering the crosstalk between different layers of epigenetic marks, such as histone modifications, a combination of multi-omics that includes both evidence from gene expression level and other epigenetic levels may provide more detailed insights into the whole picture. Finally, the heterogeneity of individual sperm may cover up some methylation changes that the study failed to detect, since motilities of sperm even from the same individual can vary from each other, which may represent different fates during spermatogenesis. It will be a good idea to separate different fractions of sperm with varied motilities and perform genomewide methylation analysis in future studies.

Conclusion

Our study has provided a novel and detailed analysis of the DNA methylation patterns in asthenozoospermia and normal controls and has identified methylation alterations associated with genes and pathways that may contribute to the abnormal phenotypes as well as infertility. This is the first study to implement whole-genomewide profiling of DNA methylation in asthenozoospermia, as previous studies used either array-based methods or targeted-sequencing methods. Although our current results still lack evidence for direct causation of the disease, the identification of asthenozoospermic-related DMRs and their associated genes is intriguing, and these genomic regions and genes provide promising candidate sites or genes for further investigations of the etiology behind asthenozoospermia and male infertility. Our study has further consolidated DNA methylation as a crucial participant in asthenozoospermia and male infertility. We have successfully identified novel asthenozoospermic-related DMRs and their associated genes, providing promising targets for further investigation of this disease. Importantly, a larger scale, multi-omics study with integrated analysis of both methylation and gene expression data, as well as experimental validation for promising candidate genes, will be performed to investigate the etiology in detail. With the rapid progress of advanced technologies such as scRNA sequencing and single-cell bisulfite sequencing, inspiring findings in this area will surely flourish in the near future and great contributions will also be made to health and medical areas.

Summary points

• There was no significant difference in terms of global DNA methylation levels between normozoospermic and asthenozoospermic sperm.

• Regional analysis identified 238 asthenozoospermia-associated differentially methylated regions (DMRs), of which 164 DMRs were hypermethylated and 74 DMRs were hypomethylated.

• The 238 DMRs were annotated to 187 genes, which included 114 protein-coding genes, 65 ncRNA genes and eight pseudogenes.

• Gene ontology analysis showed clusters of DMR-associated genes related to biological pathways, including protein complex oligomerization, purine ribonucleotide metabolic process, ribonucleotide metabolic process and activation of protein kinase A activity.

• Seventeen DMRs were located near promoter regions of corresponding protein-coding genes, suggesting potential regulation of gene expression.

• Five DMR-related genes (LRGUK, DZIP1, DNAAF5, SRPK1 and TDRP) have been reported to be directly involved in spermatogenesis or the maintenance of sperm functions, indicating that these genes may be interesting candidates for further investigation.

Author contributions

C Fan, F Li and B Shen conceived and supervised the study. J Li performed DNA extraction, library construction and bisulfite sequencing. J Xu, T Yang and F Li performed sperm collection and semen analysis. J Li, J Chen and C Fan performed data and statistical analyses. J Li and C Fan wrote the manuscript. All authors contributed to and approved the final version of the manuscript.

Ethical conduct of research

The authors state that this work has been approved by all respective institutional research and ethics boards (WHO International Clinical Trials Registry ChiCTR2200056153). The authors state that they have obtained appropriate institutional review board approval or have followed the principles outlined in the Declaration of Helsinki for all human or animal experimental investigations. In addition, for investigations involving human subjects, informed consent has been obtained from the participants involved.

Acknowledgments

The authors thank the participants of the study who made this work possible, the medical staff of West China Second University Hospital of Sichuan University and the members of Institutes for Systems Genetics at West China Hospital, Sichuan University.

Supplementary data

To view the supplementary data that accompany this paper please visit the journal website at: www.tandfonline.com/doi/suppl/10.2217/epi-2022-0122

Financial & competing interests disclosure

This work was supported and funded by the International Cooperation Initiative grant (no. 139190042) from West China Hospital, Sichuan University, Chengdu, China. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was used in the creation of this manuscript.

Additional information

Funding

This work was supported and funded by the International Cooperation Initiative grant (no. 139190042) from West China Hospital, Sichuan University, Chengdu, China. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. No writing assistance was used in the creation of this manuscript.