The present study identified genome-wide testis and epididymis-specific promoter methylation information in cfsDNA, aiming to develop noninvasive epigenetic biomarkers for male infertility. For this purpose we compared promoter methylation profiles in the cfsDNA of Nor and PV. Overall, our results demonstrate that there are a number of promoters of testis and epididymis-specific hypo- and hypermethylated genes in cfsDNA, thus providing a foundation for developing noninvasive biomarkers for studying the epigenetic mechanism and clinical diagnosis of male infertility, of which most conditions are impaired testicular spermatogenesis and epididymal sperm maturation.
In our study, the semen samples of successful vasectomy are indispensable, because these samples do not contain DNA originating from testis and epididymis owing to the ligation of vas deferens, which makes it possible to screen the testis or epididymis-specific methylated promoters. Although several previous studies reported that vasectomy could change the mRNA and protein expression of epididymis and impair the spermatogenesis of testis and induce aberrant DNA methylation at imprinted genes in testicular sperm [24–27], there are no reports in the literature describing the change of DNA methylation status in human seminal vesicles, bulbourethral glands and the prostate after vasectomy. Therefore, comparative analysis of promoter methylation information in the cfsDNA of Nor and PV can identify variations in methylation profiles at promoter regions and thus identifies promoters that are likely testis and epididymis tissue specific and potential biomarkers for male fertility.
Another rationale for identification testis and epididymis-specific methylated promoters is the high contribution of releasing or secretion of these organs to cfsDNA concentration. We previously found that concentration of cfsDNA is much higher than other body fluids, and proposed the high contribution of testicular apoptosis to cfsDNA . In the present study, averaged cfsDNA concentration in Nor was about quadruple of PV, indicating that cfsDNA is mostly originated from testis and epididymis. Considering spontaneous apoptosis happens in 75% of germ cell, which would develop to mature spermatozoa in mammalian, and the renewal of epididymal epithelium is slow [28, 29], it is more likely that the apoptosis of germ cell in testis is the primary origin of cfsDNA from testis and epididymis.
Moreover, to ensure the reliability of testis and epididymis-specific methylated promoters identified by promoter methylation profile, we excluded some factors influencing the methylation profile of cfsDNA and kept the control subjects and vasectomized donors as matched as possible. Some previous studies reported that aberrant DNA methylation occurred in distinct diseases . In addition, environmental chemicals [16, 17], age  and lifestyle factors  can also affect the DNA methylation of human tissues. Therefore, to meet inclusion criteria, both Nor and PV groups were screened for potential confounding conditions such as inflammation of reproduction organs, genetic disease, potential occupational exposure, obesity and smoking and excess alcohol consumption.
By the microarray of promoter methylation, we identified the promoters of 367 testis and epididymis-specific hypomethylated genes and 134 testis and epididymis-specific hypermethylated genes. Comparing our result with other studies about DNA methylation information of human testis, we can determine some testis-specific methylated genes. Currently, only one study has globally analyzed the human testis-specific promoter methylation to identify testis-specific hypomethylated genes . Totally 12 genes with promoter hypomethylation in this report are found in our results. Among them, MORC1 is related to spermatogenesis ; ELF5 is involved in cell proliferation ; PRAME is associated with the regulation of apoptosis ; SPACA3 is correlated to the sperm-egg recognition . But some testis-specific hypomethylated genes are inconsistent with our result. It may be mainly due to the differences in the comparison and source of DNA. The previous report  compared DNAs between testis and other tissues, whereas we compared cfsDNA between vasectomized men and healthy donors. In addition, the testicular DNA originates from spermatogenic cells, Sertoli cell and Leydig cell. However, testis and epididymis-specific cfsDNA should be mainly released by the apoptotic germ cells including spermatogonia, spermatocytes and spermatids , and epididymal epithelium. Some testis and epididymis-specific hypomethylated genes relevant to the spermatogenesis in our result, such as DNMT3L, HSF1, MSH4, THEG, SOHLH1 and CIB1, cannot be found in the methylation profile of testis tissue owing to the interference of Sertoli cell and Leydig cell . Some other reports studied promoter methylation of individual genes in testis [19, 20]. Our result confirmed the testis-specific hypomethylated genes including MAGEA1 and ANKRD30A.
The result of single-gene MeDIP-real time quantitative PCR and MethyLight analysis confirmed the reliability of our microarray data. Among the 19 promoters which were successfully determined, sixteen are consistent with the result of promoter methylation microarray. The other 3 promoters tested as false positives on the microarray, which may be due to the possibly inefficient or non-specific detection of individual probes on the microarray and normalization artifacts during data processing. The failure of measurement of MCM10 may be due to the inappropriate primers and probe. Several previous studies concurred with our result of single-gene MeDIP and MethyLight analysis: the hypomethylated promoters of CLPB, PRAME, DAZ1 MORC1, and NCK2 were confirmed in human testis ; the hypomethylated promoters of HOXA5, GML, PEG10 and SNURF were identified in human sperm [15, 39]. Methylation level of H19 in single-gene MeDIP analysis is lower than human spermatocytes and spermatid, but it is similar with spermatogonia and somatic cell , which suggests that most of H19 in cfsDNA originates from the spermatogonia and epithelial cell of male reproduction organs.
To further understand the potential function of testis and epididymis-specific hypo- and hypermethylated genes, and apply these potential epigenetic markers, we carried out Gene Ontology analysis with the testis and epididymis-specific hypomethylated genes and hypermethylated genes, respectively. Out of the biological processes of testis and epididymis-specific hypomethylated genes, it is obviously that sexual reproduction is consistent with the function of testis because the process of spermatogenesis takes place in testis and involves the development of mitotically growing spermatogonia into meiotic spermatocytes that give rise to haploid spermatids, and subsequently differentiate into mature sperm . Other biological processes, such as ion transport, organic acid transport and negative regulation of transport, participate in the reabsorption and excretion of epididymis , the biological process of defense response may protect the sperm from outside bacterial invasion in epididymis . On the other hand, the biological processes of testis and epididymis-specific hypermethylated genes involve in regulation of cellular protein metabolic process, regulation of ubiquitin-protein ligase activity during mitotic cell cycle and regulation of epidermal growth factor receptor activity, which may be the apoptotic inducement of spermatogenic cell because of either low expression or silence of these hypermethylated genes.