Genome-wide 5-hydroxymethylcytosine patterns in human spermatogenesis are associated with semen quality

Abstract

We performed immunofluorescent analysis of DNA hydroxymethylation and methylation in human testicular spermatogenic cells from azoospermic patients and ejaculated spermatozoa from sperm donors and patients from infertile couples. In contrast to methylation which was present throughout spermatogenesis, hydroxymethylation was either high or almost undetectable in both spermatogenic cells and ejaculated spermatozoa. On testicular cytogenetic preparations, 5-hydroxymethylcytosine was undetectable in mitotic and meiotic chromosomes, and was present exclusively in interphase spermatogonia Ad and in a minor spermatid population. The proportions of hydroxymethylated and non-hydroxymethylated diploid and haploid nuclei were similar among samples, suggesting that the observed alterations of 5-hydroxymethylcytosine patterns in differentiating spermatogenic cells are programmed. In ejaculates, a few spermatozoa had high 5-hydroxymethylcytosine level, while in the other ones hydroxymethylation was almost undetectable. The percentage of highly hydroxymethylated (5-hydroxymethylcytosine-positive) spermatozoa varied strongly among individuals. In patients from infertile couples, it was higher than in sperm donors (P<0.0001) and varied in a wider range: 0.12-21.24% versus 0.02-0.46%. The percentage of highly hydroxymethylated spermatozoa correlated strongly negatively with the indicators of good semen quality - normal morphology (r=-0.567, P<0.0001) and normal head morphology (r=-0.609, P<0.0001) - and strongly positively with the indicator of poor semen quality: sperm DNA fragmentation (r=0.46, P=0.001). Thus, the immunocytochemically detected increase of 5hmC in individual spermatozoa is associated with infertility in a couple and with deterioration of sperm parameters. We hypothesize that this increase is not programmed, but represents an induced abnormality and, therefore, it can be potentially used as a novel indicator of semen quality.

Keywords: 5-hydroxymethylcytosine; Pathology Section; human spermatogenesis; semen quality; sperm DNA fragmentation; testicular spermatogenic cells.

Figure 1. 5-hydroxymethylcytosine (5hmC) and 5-methylcytosine (5mC) patterns in human ejaculated spermatozoa

Immunostaining for 5hmC, 5mC, merge image, and phase-contrast images of spermatozoa from patients (A-C) and a sperm donor (D), having 10.04 %, 6.04 %, 21.24 %, and 0.46 % of 5hmC-positive spermatozoa in ejaculate, respectively.

Figure 2. Scatter dot plot showing the percentage of immunocytochemically detected highly hydroxymethylated (5hmC-positive) spermatozoa in ejaculates from sperm donors versus patients from infertile couples

*The frequency of 5hmC-positive spermatozoa is significantly higher in patients compared to sperm donors (P < 0.0001, the Mann-Whitney U-test). All 5hmC-positive spermatozoa had visible anti-5-methylcytosine fluorescence.

Figure 3. Quantitative analysis of anti-5-hydroxymethylcytosine (5hmC) and anti-5-methylcytosine (5mC) fluorescence intensity in representative spermatozoa from 3 individuals (A, B, C)

* Both anti-5hmC and anti-5mC fluorescence intensities differ significantly between 5hmC-positive and 5-hmC-negative spermatozoa (P < 0.0001, the Mann-Whitney U-test).

Figure 4. Correlations between the frequency of highly hydroxymethylated (5hmC-positive) spermatozoa in human ejaculate and semen parameters

Statistically significant correlations are framed (P < 0.05, the nonparametric Spearman test).

Figure 5. 5-hydroxymethylcytosine (5hmC) and 5-methylcytosine (5mC) patterns in human testicular cells: mitotic spermatogonium at metaphase stage, meiotic spermatocytes at prophase (pachytene, diplotene and diakinesis) stage and metaphase II stage after QFH/AcD banding and immunodetection of 5hmC and 5mC

In both mitotic and meiotic chromosomes from spermatogenic cells 5hmC is undetectable, whereas 5mC is present. QFH/AcD staining and photoimaging of banded chromosomes were performed before the 5hmC and 5mC immunodetection.

Figure 6. 5-hydroxymethylcytosine (5hmC) patterns in nuclei of human testicular cells

A. The haploid and diploid nuclei of testicular cells after subsequent fluorescence in situ hybridization (FISH) with DNA probes specific for centromere 3 (3p11.1-q11.1 = D3Z1, SpectrumOrange) and 10 (10p11.1-q11.1 = D10Z1, SpectrumAqua) and immunostaining for 5hmC/Alexa488 (green) with DAPI counterstaining. In the shown field of view, two out of four diploid nuclei and one out of three haploid nuclei are hydroxymethylated, whereas in the rest nuclei 5hmC is undetectable. B. The proportions of non-hydroxymethylated diploid, hydroxymethylated diploid, non-hydroxymethylated haploid and hydroxymethylated haploid nuclei among a total of 5000 nuclei in each of testicular tissue samples from 15 azoospermic patients. The proportions of nucleus types do not differ between patients with obstructive and non-obstructive azoospermia (P = 0.978, Fisher’s z-test).

Figure 7. Immunohistochemical detection of 5-methylcytosine (5mC) and 5-hydroxymethylcytosine (5hmC) patterns in a tubular cross section

5mC is visible in nuclei of all cell types within the tubules, while 5hmC is present mostly in spermatogonia Ad lying next to the basal membrane.

Figure 8. Human meiotic oocyte after QFH/AcD staining and immunodetection of 5-methylcytosine (5mC) and 5-hydroxymethylcytosine (5hmC)

In contrast to mitotic and meiotic chromosomes from spermatogenic cells, oocyte chromosomes are methylated and hydroxymethylated along arms. QFH/AcD staining and photoimaging of banded chromosomes were performed before the 5hmC and 5mC immunodetection.