Molecular Profiling of Spermatozoa Reveals Correlations between Morphology and Gene Expression: A Novel Biomarker Panel for Male Infertility

Abstract

Choosing spermatozoa with an optimum fertilizing potential is one of the major challenges in assisted reproductive technologies (ART). This selection is mainly based on semen parameters, but the addition of molecular approaches could allow a more functional evaluation. To this aim, we used sixteen fresh sperm samples from patients undergoing ART for male infertility and classified them in the high- and poor-quality groups, on the basis of their morphology at high magnification. Then, using a DNA sequencing method, we analyzed the spermatozoa methylome to identify genes that were differentially methylated. By Gene Ontology and protein-protein interaction network analyses, we defined candidate genes mainly implicated in cell motility, calcium reabsorption, and signaling pathways as well as transmembrane transport. RT-qPCR of high- and poor-quality sperm samples allowed showing that the expression of some genes, such as AURKA, HDAC4, CFAP46, SPATA18, CACNA1C, CACNA1H, CARHSP1, CCDC60, DNAH2, and CDC88B, have different expression levels according to sperm morphology. In conclusion, the present study shows a strong correlation between morphology and gene expression in the spermatozoa and provides a biomarker panel for sperm analysis during ART and a new tool to explore male infertility.

Figure 1

Study design. (a) Morphological criteria are used to score spermatozoa at high magnification (6100x) and to assess blastocyst quality under an inverted microscope. (A) Representative images of a spermatozoon with “score 0” (top) and the bad-quality blastocyst obtained by ICSI using this spermatozoon. (B) Representative images of a spermatozoon “score 6” and the good-quality blastocyst obtained. (b) Simplified flowchart of the strategy to identify candidate biomarkers of good-quality spermatozoa. DNA from sperm samples (n = 3 with “score 0” and n = 3 with “score 6”) is sequenced to identify an initial set of genes that are differentially methylated in spermatozoa with good (score 6) and poor (score 0) morphology. The correlation between the expression profiles of candidate genes and spermatozoon morphology is then evaluated by reverse transcription quantitative polymerase chain reaction (RT-qPCR) using RNA from an independent set of sperm samples (n = 5 with “score 0” and n = 5 with “score 6”) from different patients with infertility.

Figure 2

Analysis of the differentially methylated genes. (a) Distribution of differentially methylated genes between score 0 and score 6 sperm samples in the human chromosomes. (b) Gene Ontology (GO) classification of the differentially methylated genes in molecular function categories. (c) GO classification of the differentially methylated genes in biological process categories. The two pie charts were generated with the PANTHER tool.

Figure 3

Top-ranked functional networks of the differentially methylated genes. (a) Top network identified by ingenuity pathway analysis (IPA) of differentially methylated genes related to cell cycle, cellular assembly, and organization. (b) Top network identified by IPA of differentially methylated genes related to organ development and function. Colored nodes indicate differentially methylated genes. Noncolored nodes were proposed by IPA and suggest potential targets functionally coordinated with the differentially methylated genes. Dashed lines represent indirect relationships, and solid lines indicate direct molecular interactions. In each network, edge types are indicatives: a line without arrowhead indicates binding only; a line finishing with a vertical line indicates inhibition; a line with an arrowhead indicates ‘acts on.' (c) Protein-protein interaction network of selected differentially methylated genes. Using the OmicsNet database, six genes (AURKA, HDAC4, CARHSP1, CACNA1H, CACNA1C, and DNAH2) were used to construct a top-ranked functional protein-protein interaction network.

Figure 4

Relative expression level of 10 genes that are differentially expressed in score 0 and score 6 spermatozoa. (a) Graphical representation of the gene types. (b) Gene expression was compared between score 0 (white) and score 6 (black; reference, set to 1) sperm samples by RT-qPCR analysis. p value were calculated with the Mann–Whitney test.

Figure 5

Expression profile of candidate genes in different human tissues. Expression values (in Log2 (RPKM)) of HDAC4, CARHSP1, SPATA18, AURKA, CCDC60, DNAH2, and CFAP46 in 30 tissues from the Genotype-Tissue Expression (GTEx) consortium. For each gene, the colored circle belonging to each tissue indicates the valid RPKM value of all samples in the tissue. RPKM: reads per kilobase of transcript per million mapped reads.