Combined Analysis of the Transcriptome, Proteome and Metabolome in Human Cryopreserved Sperm

1. Ethics statement and sample collection

This study was conducted with the approval of the Human Subjects Ethics Committee of National Research Institute for Family Planning (2018015). All sperm donors provided written consent before participating in the study, agreeing to the use of their anonymized information for future studies.

All sperm samples were obtained from the Human Sperm Bank at the National Research Institute for Family Planning. The healthy volunteers met the following inclusion criteria: age between 25 and 35 years, abstinence for 3 to 5 days, semen volume of 4 to 6 mL, sperm concentration of 60–80*10⁶/mL, sperm vitality >40%, and normal sperm morphology rate >4%, with no white cells or bacteria in the sperm. Volunteers with sexually transmitted diseases, cardiovascular disease, endocrine diseases, or other conditions that could clearly affect sperm quality were excluded from the study. Written consent was obtained from all participants.

2. Study design and sample collection

Each omics analysis in this study employed a paired design to investigate the difference in sperm before and after cryopreservation. The associated information of the donors for the transcriptomic, proteomic and metabolomic analyses are listed in Supplement Table 1.

For transcriptomic analysis, 15 fresh human semen samples were collected. They were randomly and equally divided into three groups of 5 samples. Then, 1 mL of semen from each sample was obtained for cryopreservation, and another 1 mL was taken and not subjected to any treatment (considered to be the normal control). All 5 samples in each group were mixed after cryopreservation when they had completely liquefied, and then the transcriptomics was performed.

For proteomics, 9 fresh human semen samples were collected. As previously described [5], they were randomly and equally divided into three groups of 3 samples. Then, 1 mL of semen from each sample was obtained for cryopreservation, and another 1 mL was taken and not subjected to treatment (considered to be the normal control). All 3 samples in each group were mixed after cryopreservation when they had completely liquefied, and then the proteomics was performed.

For metabolomic analysis, 5 fresh human semen samples were obtained. One milliliter of semen from each sample was obtained for cryopreservation and another 1 mL was taken and not subjected to any treatment (considered to be the normal control). Metabolomics was then performed on the 5 cryopreserved samples versus 5 normal controls.

3. Semen cryopreservation protocols

The sperm freezing process was conducted completely in accordance with the routine operations performed in the sperm bank [5, 6]. The cryoprotectant used was a mixture of glycerol, egg yolk, and citrate. The process involved thawing the cryoprotectant to room temperature and then adding it dropwise to the semen in a 2:1 ratio. The mixture was incubated at 30 to 35 ℃ for 5 minutes before being subjected to slow freezing using a programmable freezer. The samples in tubes were cooled at 1.5 ℃ per minute from 20 to -6 ℃, at 6 °C per minute to -100 ℃, and at -100 ℃ for 30 minutes. Then, the tubes containing samples were transferred to liquid nitrogen and was stored in liquid nitrogen for two days before being thawed for analysis.

4. Sperm motility parameter analysis and mitochondrial ultrastructure

After the semen was fully liquefied, a Makler sperm counting chamber was used to test the sperm concentration and activity. Sperm motility parameter analysis was performed by computer-aided sperm analysis (CASA, SuiPLus SAA-II).

The sperm samples were first fixed in a solution containing glutaraldehyde and paraformaldehyde and then postfixed in osmium tetroxide. The fixed samples were then dehydrated in a series of graded ethanol solutions and embedded in epoxy resin. The embedded samples were cut into 70 nm thick sections that were placed on a copper mesh with a thin film attached. The grids were then stained with uranyl acetate and lead citrate and viewed under a transmission electron microscope (TEM-1400 Plus) [10].

5. Comparative transcriptomics and bioinformatic analysis

The samples were grouped as described in the “Study design and sample collection” section. RNA sequencing was conducted by OE Biotech Company following the method described in a previous study [11]. We obtained 33.34 Gb clean data from the RNA sequencing results of 6 samples. The Q30 was 74.10% to 91.68%. The average guanine cytosine content was 51.58%. The original RNA sequencing data were imported into GeneSpring software (version 13.1; Agilent) and were standardized using the quantile method, which included the original signal value, standardized signal value, detection status, and detailed annotation information. All protein-coding genes/transcripts were used as the background list, and the list of differential protein-coding genes/transcripts was used as the candidate list screened from the background list. The differentially expressed genes were identified using paired t-test (p<0.05 and fold-change >2 or fold-change <0.5). The p-value representing whether the gene ontology (GO) functional set was significantly enriched in the list of differential protein-coding genes/transcripts was calculated using the hypergeometric distribution test, and then the p-value was corrected by the Benjamini & Hochberg multiple testing correction to obtain the false discovery rate. Enrichment analysis was performed using gene set enrichment analysis (GSEA) software (UC San Diego and Broad Institute). The cell component, biological process, and molecular function enrichment analyses were performed using the Fisher algorithm for the differentially expressed genes between samples. A database (http://geneontology.org/) was consulted for comparison analysis and selection for subsequent studies.

6. Metabolomic experiment and bioinformatic analysis

The samples were grouped as described in the “Study design and sample collection” section. Approximately 40×10⁶ fresh or cryopreserved sperm were collected for metabolomic analysis. Metabolomics was performed using a TSQ Quantiva (Thermo). A paired t-test was used as an analytical approach, and the p-value was set to 0.05. All significant differences were taken as input. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis was performed using MetaboAnalystR (https://www.metaboanalyst.ca/).

7. Statistical analysis and multiomics study

For the transcriptomic and metabolomic analyses, a paired t-test was used, and the p-value was set to 0.05 to identify the differentially expressed genes or metabolic products. Then, the transcriptomic and metabolomic results, as well as proteomic results published previously [5], were combined to explore the key genes affected by cryopreservation. The key genes were further analyzed with the DAVID Bioinformatics Resources v6.8 online server (https://david.ncifcrf.gov) for enrichment analysis.

1. Motility was altered in response to cryopreservation

The curvilinear velocity, straight-line velocity, and average path velocity, as well as the percentage of motile sperm after frozen recovery, were significantly reduced (p<0.05) (Fig. 1A-1D). The ultrastructure of mitochondria was also observed before and after cryopreservation because mitochondria are the key organells that provide energy to support the motility of sperm. Mitochondrial cristae and abundant mitochondrial matrix were observed in the normal mitochondria, as indicated by the arrows (Fig. 1E). However, the mitochondria in cryopreserved sperm showed abnormal structures such as widening of mitochondrial cristae, and increased vacuolization, as indicated by the arrows (Fig. 1F).

Fig. 1
Motility and mitochondrial ultrastructure of fresh and cryopreserved sperm. Trajectory of (A) untreated sperm and (B) cryopreserved sperm. Yellow: immotile (IM); green: progressively motile (PR); red: hyperactivated sperm (C) Motile velocity of sperm. (D) Percentage of PR sperm. The ultrastructure of mitochondria in (E) untreated and (F) cryopreserved sperm is shown. The scale bars were shown in the figure. VCL: curvilinear velocity, VSL: straight-line velocity, VAP: average path velocity. ^*p<0.05.

• Click for larger image

2. Differentially transcribed genes in sperm after cryopreservation

Fifteen semen samples were collected and treated as described in the Methods (Fig. 2A). A total of 3,212 genes were differentially expressed (p<0.05) between the cryopreserved and fresh groups. Among these genes, 3,097 (96.4%) genes were downregulated during the freezing and thawing procedure (Fig. 2B). The top 20 upregulated and downregulated genes were displayed (Fig. 2C). GO enrichment analysis was performed, and the top 30 altered pathways were shown (Fig. 2D). We observed altered function of mitochondria in the molecular function category due to altered tRNA-methyltransferase activity and adenosine tri-phosphate/adenosine di-phosphate transmembrane transporter activity. In addition, GSEA indicated that oxidative phosphorylation, which occured in mitochondria, was altered after cryopreservation (Fig. 2E). Of note, the androgen response (Fig. 2F) and spermatogenesis pathways (Fig. 2G), which were associated with sperm, were also significantly altered. All of the significantly differentially expressed genes were listed in Supplement Table 2.

Fig. 2
Transcriptomic results. (A) Protocols by which the sperm were treated in transcriptomic analysis. (B) Volcano plot of the total altered genes. Upregulated genes are shown in blue, and downregulated genes are shown in red. (C) Top 20 upregulated and top 20 downregulated genes during the freezing-thawing process. (D) GO enrichment analysis of the top 30 regulated genes. GSEA indicated that the (E) oxidative phosphorylation pathway, (F) androgen response pathway and (G) spermatogenesis pathway were altered by cryopreservation. GO: gene ontology, GSEA: gene set enrichment analysis.

• Click for larger image

3. Metabolomic analysis

Five semen samples were collected and treated as described in the Methods (Fig. 3A). The quantities of 20 metabolomic products were significantly altered: imidazole-4-acetate, succinate, beta-alanine, lactose, citrate, isocitrate, L-normetanephrine, glycerate 3-phosphate, cis-aconitate, phosphoenolpyruvate, D-gluconic acid, D-glucose, O-acetylcarnitine, NAD+, 3-dehydroxycarnitine, N2-acetyl-L-lysine, creatine, O-propanoylcarnitine, urate, and ethanolamine phosphate (Table 1), and the associated enriched pathways were identified by KEGG analysis (Supplement Table 3), and the citrate cycle (TCA cycle) in mitochondria was altered most significantly (Fig. 3B, 3C).

Fig. 3
Metabolomic analysis. (A) Protocols by which the sperm were treated in metabolomic analysis. (B) Top altered pathways indicated by metabolomic analysis. (C) Network of the metabolomic results.

• Click for larger image

Table 1
Twenty significantly altered metabolomic products

• Click for larger image

4. Combined analysis of the transcriptome, proteome and metabolome

We decided to combine the results of transcriptomics and metabolomics, as well as proteomic results that had been published previously, to explore the key genes of pathways that were affected by cryopreservation. First, we screened the results in each omics group to find 346 genes that were altered simultaneously in at least two omics analyses (Fig. 3A, Supplement Table 4). These genes were subjected to KEGG enrichment analysis (Fig. 3B). Notably, the top-ranking metabolic pathways were strongly associated with mitochondria [12, 13]. Moreover, five genes were altered in all three omics analyses: COL11A1, COL18A1, LPCAT3, NME1, and NNT (Table 2).

Table 2
The information of the 5 genes that altered in all the three omics analysis

• Click for larger image