Abstract
Sperm transcriptomics provide insights into subtle differences in sperm fertilization competence. For predicting the success of complex traits like male fertility, identification of hub genes involved in various sperm functions are essential. The bulls from the transcriptome profiled samples (n = 21), were grouped into good and poor progressive motility (PM), acrosome integrity (AI), functional membrane integrity (FMI) and fertility rate (FR) groups. The up-regulated genes identified in each group were 87, 470, 1715 and 36, respectively. Gene networks were constructed using up- and down-regulated genes from each group. The top clusters from the upregulated gene networks of the PM, AI, FMI and FR groups were involved in tyrosine kinase (FDR = 1.61E−11), apoptosis (FDR = 1.65E−8), translation (FDR = 2.2E−16) and ribosomal pathway (FDR = 1.98E−21), respectively. From the clusters, the hub genes were identified and validated in a fresh set of semen samples (n = 12) using RT-qPCR. Importantly, the genes (fold change) RPL36AL (14.99) in AI, EIF5A (54.32) in FMI, and RPLP0 (8.55) and RPS28 (13.42) in FR were significantly (p < 0.05) up-regulated. The study suggests that the expression levels of MAPK3 (PM), RPL36AL + RPS27A or RPL36AL + EXT2 (AI), RPL36AL or RPS27A (FMI) and RPS18 + RPS28 (FR) are potential markers for diagnosing the semen quality and fertility status of bulls which can be used for the breeding program.
Similar content being viewed by others
Introduction
In the dairy industry, bulls are selected for artificial insemination (AI) programs based on breeding soundness evaluations and conventional tests that measure sperm functions1. Conventional tests to establish fertility differences among bulls are inadequate because the results are highly variable and do not correlate with bull fertility status2. Measures for the selection of high-fertile and high-quality semen producing bulls are of utmost importance as they bring immense economic benefits to frozen semen stations and farmers. Hence, next-generation omics technologies are in high demand to identify subtle variations in molecular signatures that influence sperm quality, which ultimately determines the fertility status of bulls. Recently, sperm transcriptomic and proteomic profiling have identified the factors and pathways associated with semen quality and fertility rate (FR) in human3 and bovine1,4. Such detailed information on Murrah buffalo bulls is not currently available.
Importantly, fertilization is a multifaceted phenomenon involving numerous associated sub-processes or factors. Sperm progressive forward motility is an indispensable trait for travelling across various barriers at different segments of the cervix, uterus and oviduct. In addition, the functional membrane integrity of sperm is an essential attribute for combating the biochemical micro-milieu of the female reproductive tract. Acrosomal membrane intactness is a prerequisite for sperm capacitation in the uterus and acrosomal reaction in the oviduct. All these sperm attributes are essential components of effective fertilization and required to identify high-fertile bulls for breeding program.
Distinctly, any perturbation in these biological processes is the consequence of a disturbance in a group of genes and not from a single gene alone5. Although omics technologies generate big data, the process of extracting the desired information from them becomes challenging. Moreover, advancements in transcriptome data analysis have led to the establishment of “gene regulatory networks” and which comprise coordinately expressed genes. Hub genes in the network tend to cluster densely and represent the likely control points of the study condition6,7. Studies have established that hub genes can serve as diagnostic, predictive, or prognostic biomarkers in colon adenocarcinoma8. In bovines, hub genes involved in mastitis development9, intramuscular fat content of Nellore cattle10 and in vivo pre-implantation development11 are identified. However, such information for male fertility prediction is not available for any species.
Since effective fertilization involves diversified sperm features, the identification of hub genes for fertility prediction may aid in accurately diagnosing the fertility status of bulls. Furthermore, the identified hub genes involved in sperm function and fertility may provide new insights for developing guidelines for bull fertility assessment.
In the present study, sperm transcriptome data available in the laboratory from Murrah buffalo bulls were used retrospectively to identify the differentially expressed genes in the progressive motility (PM), acrosomal integrity (AI), functional membrane integrity (FMI) and FR groups. These differentially expressed genes can be used to identify hub genes that regulate sperm functions and FR. The present study aimed to (1) identify the hub genes involved in the regulation of sperm functions and fertility and (2) evaluate the fertility prediction ability of the identified hub genes using receiver operating characteristic (ROC) curve analysis.
Results
Grouping of bulls based on sperm functions and fertility rate
Bulls (n = 12) were classified into two groups (good and poor semen quality producers) based on the group averages for sperm functions and FR. The sperm functions and fertility rates were significantly (p < 0.05) differing between the respective good and poor groups of PM (62.31 ± 4.22 vs 32.44 ± 4.72), AI (92.86 ± 0.88 vs 83.02 ± 1.81), FMI (28.66 ± 2.00 vs 13.81 ± 1.34) and FR (52.67 ± 0.72 vs 41.16 ± 2.05) (Fig. 1a–d).
Differentially expressed genes from the transcriptome library
Differentially expressed genes with a predefined threshold of log2FC greater than 1.0 and p < 0.05 in the PM, AI, FMI and FR groups were identified (Fig. 2a–d). There were 87, 470, 1715 and 36 genes up-regulated (Table 1) and 348, 274, 455 and 81 genes down-regulated in PM, AI, FMI, and FR groups, respectively (Table 2).
Gene-set enrichment analysis of the up-regulated genes in sperm functions and fertility rate
Gene-set enrichment analysis of the up-regulated genes revealed that biological processes such as response to oxygen-containing compounds (Normalized Expression Score: NES = 1.79), lipid metabolic processes (NES = 2.62) and defense responses (NES = 1.67), glucose metabolic processes (NES = 1.90) and fertilization (NES = 1.68) were enriched in the up-regulated genes of good PM, AI, and FMI, respectively (Table 3; Fig. 3a–c). In the FR group, there was no significantly enriched biological process.
The cellular components microtubule organizing center (NES = 1.60) in PM, endosome (NES = 1.42), plasma membrane (NES = 1.3) and acrosomal vesicle (NES = 1.90) in AI were enriched in the up-regulated genes. Interestingly, the up-regulated genes of both FMI and FR were enriched in the ribonucleoprotein complex (FMI: NES = 1.43 and FR: NES = 0.51).
Molecular functions, such as cytoskeletal protein binding (NES = 1.47) and zinc ion binding (NES = 1.70), were enriched in PM and AI, respectively. In contrast, protein serine kinase activity (NES = 1.76), catalytic activity against RNA (NES = 1.63) and phosphatase binding (NES = 1.65) were enriched in FMI group.
In the down-regulated genes, response to lipid (NES = 1.58), cell cycle (NES = 1.62) and membrane organization (NES = 1.62) were the enriched biological processes in PM, AI, and FMI, respectively (Fig. 3d–f) and they were localized at cell surface (NES = 2.08) and catalytic complex (NES = 2.17) in the AI and FMI groups, respectively. Likewise, enzyme regulator activity (NES = 1.73) and zinc ion binding (NES = 1.73) were the enriched molecular functions in AI and FMI, respectively. There was no significant enrichment in the FR group.
Gene interaction network of the sperm functions and fertility rate
The network of the up-regulated genes consisted of 90 and 121, 248 and 399, 1096 and 4779, and 44 and 91 nodes and edges corresponding to the PM, AI, FMI, and FR groups, respectively. Subsequently, merging of the four networks revealed 1736 nodes and 8787 edges. The intersection of all four networks resulted in single node EXT2 (Supplementary Fig. S1). This gene was considered along with hub genes for RT-qPCR validation and fertility prediction model analysis.
Similarly, the networks of down-regulated genes had 354 and 992, 293 and 605, 590 and 1662, and 90 and 166 nodes and edges in the PM, AI, FMI, and FR groups, respectively (Supplementary Fig. S2). However, there was no intersecting down-regulated gene for all the four groups.
Identification of the hub genes in sperm functions and fertility rate
The clusters or interconnected sub-networks obtained from each group (Table 4) indicated that the number of clusters and the cluster score were dependent on the number of up-regulated genes used for the analysis.
In the up-regulated gene network, the top cluster in the PM had 13 nodes and 58 edges, with a cluster score of 9.67. They were involved in the enzyme-linked receptor protein signaling process (FDR = 5.04E−10) and tyrosine kinase pathways (FDR = 1.61E−11). In the AI group, the top cluster had 17 nodes and 107 edges with a cluster score of 13.375. These nodes were involved in intracellular signal transduction (FDR = 1.2E−11) and apoptosis pathways (FDR = 1.65E−8). Furthermore, in the FMI, 29 nodes and 235 edges with a cluster score of 16.786 were observed. These genes were involved in protein localization to the membrane (FDR = 1.11E−13) and translation pathways (FDR = 2.2E−16). Finally, in FR, 13 nodes and 73 edges were observed with a score of 12.167. They were associated with translation (FDR = 5.68E−17) and ribosomal pathway (FDR = 1.98E−21) processes.
Likewise, in the down-regulated genes, the top cluster of PM group had 22 nodes and 103 edges with a cluster score of 9.81 and the genes were involved in amide biosynthetic process (FDR = 2.07E−7) and translation (FDR = 2.68E−7) processes. In the AI group, the top cluster had 21 nodes with 178 edges with a cluster score of 17.8 and they correspond to positive regulation of cell population proliferation (FDR = 1.03E−8) and positive regulation of gene expression (FDR = 1.03E−8) processes. Similarly, the top cluster of FMI group had 14 nodes and 89 clusters with the cluster score of 19.586 and associated with regulation of cell population proliferation (FDR = 2.98E−15) and cell surface receptor signaling pathway (FDR = 2.21E−12). In the FR group, the top cluster had 14 nodes, 89 edges and cluster score of 13.692 with the functional enrichment of translation (FDR = 7.72E−18) and SRP-dependent co-translational protein targeting to membrane (FDR = 4.42E−17).
From these clusters, the top ten up-regulated hub genes were identified in each group. Ribosomal protein gene families (both ribosomal protein large (RPL) and ribosomal protein small (RPS) subunits) were common to all four groups (Fig. 4a–d). These results indicate a probable role of sperm-retained ribosomal transcripts in fertility regulation. Similarly, the down-regulated hub genes of the PM and FR group had ribosomal protein gene families with the enrichment of translation process (Fig. 4e,h). Whereas the AI and FMI group had hub genes enriched in the regulation of response to stimulus, and regulation of proteolysis processes, respectively (Fig. 4f,g).
Validation of the expression levels of hub genes and overlapping gene
The genes (fold change) MAPK3 (17.10) and RPS27A (1.80) in PM; MCL1 (2.38), SLC9A1 (2.73), RPS27A (5.22), RPL36AL (14.99) and EXT2 (2.89) in AI; RPL36AL (2.81), EIF5A (54.32) and RPS27A (4.78) in FMI; and RPS18 (1.70), RPLP0 (8.55), RPS28 (13.42) and EXT2 (3.50) in FR were up-regulated. Among these genes, RPL36AL in AI, EIF5A in FMI and RPLP0 and RPS28 in the FR group were significantly (p < 0.05) differentially expressed (Fig. 5a–d). Genes influencing PM (ACTB), AI (MAPK3) and (JAK3) as well as the FMI gene (RSRC1) were down-regulated in their corresponding good-quality semen-producing bulls. The overlapping gene EXT2 was up-regulated in the AI and FR groups and hence, they were included in those groups for assessing sperm functions and FR predictability. The representative proteins expression levels of RPS18, RPLP0 and EXT2 showed that only EXT2 significantly higher in low fertility group as compared to high fertility group (Supplementary Fig. S3).
Influence of hub genes on sperm functions and fertility rate
The RT-qPCR expression levels of the hub genes were strongly correlated with sperm functions and fertility rate (Table 5). Importantly, RPLP0 (r = 0.58) and EIF5A (r = 0.62) were significantly (p < 0.05) correlated with FR and FMI, respectively. Interestingly, the gene RPL36AL (r = 0.73) was strongly correlated with AI rather than FMI.
Predictive ability of the genes
ROC analysis of the expression levels of the single genes MAPK3 (PM) (Fig. 6a), EXT2 and RPS27A (AI) and RPS28 (FR) showed a sensitivity of 66.67%, specificity of 83.33% and likelihood ratio of 4. Similarly, the expression levels of RPS27A and RPL36AL (FMI) (Fig. 6c) and RPS28 (FR) had a maximum sensitivity of 83.33%, specificity of 83.33% and likelihood ratio of 5. These findings suggest that these genes individually influence sperm functions and the FR.
Multiple regression analysis for the prediction of FR revealed that the combination model RPS18 + RPS28 had a maximum sensitivity of 100%, specificity of 83.33% and likelihood ratio of 6 (Fig 6d). The prediction models RPS18 + RPS28 (FR), RPL36AL + EXT2 (AI) and RPL36AL + RPS27A (AI) (Fig. 6b) had a sensitivity of 83.33%, specificity of 83.33% and likelihood ratio of 5.0. All the four models were significant (p < 0.05) in the prediction of sperm functions and FR.
Discussion
In the present study, hub genes influencing sperm functions and fertility rate were identified and validated. In PM sperm, the cellular response to the oxygen-containing compound process was the top enriched process in the up-regulated genes. Reactive oxygen species (ROS) are inherently generated by sperm during metabolic activities and capacitation process12. Excess ROS production decreases sperm motility by affecting the contractile apparatus of the flagellum13 and the defense response of the sperm against ROS in terms of the total antioxidant capacity is essential for sustaining total as well as progressive motility14. Furthermore, tyrosine kinase activity was enriched in the identified hub genes of PM sperm. Tyrosine kinases have been proved to be crucial for the sperm motility and hyperactivation15. In particular, MAPK3, known as ERK1, stimulates PM and hyperactivated motility in ejaculated sperm16. Additionally, AKAP4 protein is a substrate for ERK1 and phosphorylation of AKAP4 is crucial for the PM of sperm17. These results indicate an important role of the hub gene MAPK3 in regulating PM. ROC analysis of the present study revealed that the expression level of single gene MAPK3 alone had the maximum accuracy (66.67%) in predicting the semen samples with high percentages of PM sperm. These findings corroborate the important role of the identified hub gene, MAPK3, in regulating sperm PM. Response to lipids is an enriched process associated with downregulated genes. Lipid composition of sperm changes as it exits the male reproductive tract and when it enters the female reproductive tract. The change in composition is inevitable for a sperm to perform its functions like membrane integrity, capacitation, and acrosome reaction18. Thus, the high PM sperm is responding less to the lipid changes signifying the adaptation of high PM sperm to the changing lipids.
The enrichment of lipid metabolic processes in semen samples with a high percentage of intact acrosome suggests that lipid metabolism is crucial for maintaining AI. Recent literature suggests that the success of the male reproductive process, sperm motility, capacitation, acrosomal reaction and fusion of sperm and egg depends on the homeostasis of sperm lipids18. Sperm also utilize long-chain fatty acids for energy production19. Interestingly, knockout studies of lipid metabolism genes such as Tysnd120 and Fads21 resulted in altered plasmalogen and unsaturated fatty acid levels thereby leading to incomplete acrosome formation and failure of acrosome formation, respectively. In the good AI group, the hub genes were involved in apoptosis. MCL1, a pro-survival factor, is required for the development and homeostasis of any tissue, and MCL1 knockout mice are sterile with no mature sperm in the epididymis. The presence of MCL1 in the good AI group suggests that the gene may improve acrosome integrity by regulating anti-apoptotic process22. The phosphorylation of kinases such as JAK3 and MAPK3 is required for capacitation and acrosome reaction23. Although there is no direct evidence of the functional roles of up-regulated genes (RPS27A, RPL36AL and EXT2) in regulating AI in the present study, we speculate that these genes have a potential role in maintaining AI. Importantly, the multivariate models EXT2 + RPL36AL or RPL36AL + RPS27A had the maximum accuracy (83.3%) in predicting semen samples with good AI. Since the sperm are matured cells, down-regulation of the cell cycle genes in the high AI group indicates that these genes might have been translated into protein during spermatogenesis or these transcripts might not have been utilized by the low AI group and thus resulted in the poor acrosome quality. However, the exact function of these genes in buffalo sperm must be elucidated.
Glucose metabolism and fertilization were up-regulated in the FMI group. Sperm is the most differentiated yet metabolically active cell type and mainly utilizes glucose, fructose as a fuel source. Glucose metabolism is essential for sperm functions such as motility and fertilization events. In diabetic human males, altered glucose metabolism affects epigenetic dysregulation, leading to detrimental effects on sperm functional attributes and male fertility24. Bull sperm depends on oxidative phosphorylation for capacitation and binding to the oocyte, as the oviduct has a low concentration of glucose25,26. It has been reported that glycolytic enzymes are in the fibrous sheath of the flagellum27. Due to the poor metabolism of glucose in sperm, homeostasis is disturbed and ultimately affects the sperm membrane and acrosomal integrity28. Protein localization to the membrane (FDR = 1.11E−13) and translation (FDR = 2.2E−16) were the enriched processes of the hub genes in the good FMI group. EIF5A gene deletion leads to alterations in cell membrane integrity by affecting the PKC/WSC cell wall integrity pathway29 in yeast. Hypusinated EIF5A transports a subset of mRNA from the nucleus to the ribosome for translation30. The presence of ribosomal transcripts (RPS36AL and RPS27A) and eukaryotic translation initiation factors (EIF5A) in the hub genes may have led to translation enrichment. The present study indicates that the expression levels of each single gene, RPS27A or RPL36AL had maximum accuracy (83.33%) for predicting FMI. The study revealed down-regulation of RNA splicing process in high FMI group. Sperm splicing events regulate sperm functions and bull fertility31. Hence, the down-regulation of splicing might be because the associated transcripts might have been translated to proteins during spermatogenesis in the high FMI group.
The number of up-regulated genes in the FR group was comparatively lower and that may be the reason for not observing any significant enrichment. Previous research in cattle from our lab revealed that among the differentially expressed genes, the genes linked with the bull fertility rate were closely associated with the genes regulating functional membrane integrity, and acrosome integrity1. Likewise, in the present study translation was observed to be the top enriched biological process of the hub genes in FR as well as in the FMI group. The hub genes of this process include RPLs and RPSs subunits, indicating ribosomal heterogeneity between good FMI and FR. Previously, differentially expressed ribosomal transcript markers have been observed in immotile sperm32, abnormal morphology33, conception34, and fecundity35. Along with that, in the present study, ribosomal subunit genes, such as RPLP0, RPS18, and RPL36AL, were upregulated in the good FR group. The present study also suggests that the combination model RPS28 + RPS18 had the highest accuracy (83.33%) in predicting FR. Earlier studies from our laboratory have also identified a strong positive association between RPLs and FR34. In boar testis, expression levels of RPL18 have been positively associated with pubertal development36. These ribosomal proteins may aid in the translation of sperm RNAs after fertilization or may be involved in the translation of sperm mitochondrial genes or may be involved in processes other than translation. However, further investigation is required to understand the role of ribosomal gene families in predicting sperm fertility.
In the present study, there was a heterogeneity in the expression of ribosomal large and small subunit protein genes of mature buffalo sperm. Ribosomal heterogeneity in sperm is not yet elucidated completely and hence further research in this area will identify the gene regulations coded by the ribosome heterogeneity37.
Clearly, the only overlapping gene, EXT2, is a glycosyltransferase up-regulated in all four groups of DESeq2 analysis and was validated in the AI and FR groups. Quantitative dot blot analysis of the representative hub genes indicates that the protein levels of EXT2 is not in trend with the RNA-seq and RT-qPCR findings. The finding suggest that the available transcripts would have been translated to protein in the sperm. This glycosyltransferase is involved in the binding of the sperm to the ZP3 protein of an egg during fertilization38 and initiates acrosomal exocytosis. The differential gene and protein expression of EXT2 in FR group implies that EXT2 may be involved in the maintenance of acrosomal integrity and acrosomal exocytosis during sperm binding to zona, thereby, regulating the fertility of a bull.
Though ROC analysis revealed that the expression level of a single gene is sufficient for predicting PM and FMI status, the two-genes combination models are required to achieve maximum accuracy in the FR and AI groups. These results indicate that a correct combination of genes is required to achieve maximum prediction accuracy39. However, detailed studies are required to elucidate the role of these novel genes in regulating sperm functions and FR.
Thus, the expression levels of MAPK3 (PM), RPL36AL (AI), EIF5A (FMI) as well as RPLP0 and RPS28 (FR) in good-quality semen indicate that these genes can be used to diagnose the semen quality and fertility status of bulls (Fig. 7). In particular, the combination model RPS18 + RPS28 can be helpful in predicting the fertility status of bulls. Importantly, they can serve as gene markers to identify the high-fertile bulls for the breeding program. These hub genes can also be of drug targets for the improvement of sperm functions and bull FR.
Materials and methods
In brief, the methodology includes the procurement of frozen semen samples, assessment of sperm functions, RNA isolation and library preparation for RT-qPCR. Bioinformatic analysis of the transcriptome data includes the identification of differentially expressed genes, enrichment analysis and identification of the hub genes. The identified hub genes are validated using RT-qPCR and the association of the genes to the sperm functions and fertility rate was determined using correlation analysis (Fig. 8). The expression levels of the genes were used to diagnose the sperm functions and fertility rate status of the bulls.
Sample details
Animal ethics
All the experiments were conducted as per the approval of the Institute Animal Ethics Committee (IAEC approval vide: NIANP/IAEC/1/2020/12). All the methods were performed in accordance with the relevant guidelines and regulations.
Sample procurement
Frozen semen samples (n = 12) of the Murrah buffalo (Bubalus bubalis) bulls were purchased from ICAR-Central Institute for Research on Buffaloes (ICAR-CIRB), Haryana, India and stored in liquid nitrogen (− 196 °C) until further analyses.
Sperm functions
Frozen semen samples were thawed at 37 °C for 30 s in a water bath and sperm functions such as progressive motility (PM), acrosomal integrity (AI), and functional membrane integrity (FMI) were evaluated for each of the ejaculates (at least 2 ejaculates per animal)40,41,42.
Progressive motility
Progressive motility and other sperm kinematics were analyzed using a computer-assisted semen analyzer (CASA, Sperm Class Analyzer, version 6.4, Microptic, Spain). The frozen semen sample was thawed at 37 °C for 30 s in a water bath and diluted in Tris buffer (1:5). The diluted sample (10 μl) was then placed on the prewarmed (37 °C) glass slide and covered with an 18 × 18 mm coverslip. The sample was then analyzed for sperm kinematics including PM, total motility, velocities, wobble, etc. using a negative phase-contrast microscope (Nikon Eclipse 50, Nikon, Japan). The images were captured at the rate of 25 frames/s from 10 homogenous fields per ejaculate using the configuration settings: cell size: range 10–70 μm2; PM: sperm with a speed of >50μm/sec and >70% straightness (STR).
Functional membrane integrity and acrosomal integrity
Both the FMI and AI were evaluated using a single test hypoosmotic swelling- Giemsa (HOS-G) test. Thawed (37 °C) semen sample (50 μl) was added to 450 μl of hypoosmotic (100 mOsm) and isoosmotic solutions (300 mOsm) maintained at 37 °C and then incubated for 30 min. After the incubation, 10 μl of the solution from each medium was smeared onto a clean glass slide, air-dried, fixed and then stained using Giemsa. Sperm were classified into four populations: Host positive and acrosome positive (HPAP), Host positive and acrosome negative (HPAN), Host negative and acrosome positive (HNAP) and Host negative and acrosome negative (HNAN). Sperm with hairpin bent tails were considered Host positive otherwise considered Host negative. Sperm with intact acrosomal membranes were considered acrosome positive and otherwise considered acrosome negative. The actual population of both functional membrane and acrosome intact sperm (FMI) was calculated by subtracting the HPAP positive population in 300 mOsm from the HPAP of the hypoosmotic (150 mOsm) solution. Sperm population of HPAP and HNAP in 300 mOsm were added together for calculating the percentage of acrosomal intact (AI) sperm. A minimum of 200 sperm were counted using 100 × objective under the phase-contrast microscope (Nikon Eclipse 80i, Nikon, Japan). An average value for the two ejaculates per bull was calculated and considered for further statistical analysis.
Based on the group average for sperm functions and FR, the bulls were divided into two groups (n = 6, good and poor). All the functions and FRs differ significantly (p < 0.05) between the groups.
Bull fertility rate
The field fertility data for each bull were obtained from the semen bank. The fertility rate for each bull was calculated based on at least 500 inseminations per bull and verified pregnancy after 60 days of insemination.
Bioinformatics analysis
Differentially expressed genes
Differentially expressed genes (p < 0.05 and log2 fold change > 1) from the PM, AI, FMI and FR groups were analyzed from the sperm RNA-seq data (n = 21, Reproductive Physiology Laboratory, ICAR-NIANP, Adugodi, Bengaluru) of buffalo bulls. The RNA-seq datasets analyzed during the current study have been submitted to the NCBI database SRA Bioproject PRJNA803987. Due to proprietary reasons the data are not publicly available.
Enrichment analysis
Enrichment analysis of both up-regulated and down-regulated genes in each group was performed the using Gene-set enrichment analysis tool (GSEA, version 4.2.3) with the GSEA Preranked module: 1000 permutations; the minimum and the maximum size to exclude gene sets were set to 2 and 500, respectively. Enriched processes and their corresponding normalized enrichment score (NES) were obtained from GSEA, whereas the redundant processes with broad meaning were eliminated.
Construction of gene network
The network corresponding to both up-regulated and down-regulated genes from each group was imported into the Cytoscape (version 3.8.2) using the network import module with the STRING database as the background data source. The network with default combined interaction scores greater than 0.4 were imported with 10 additional interactors for subsequent analysis. Also, an intersection and union of all four networks were constructed using the merge option in the Cytoscape.
Module analysis
A molecular complex detection (MCODE) tool from the Cytoscape plug-in was used to identify the densely connected clusters with the following default parameters: score cut-off = 5; K-score = 2; node score cut-off = 0.2; max depth from seed = 100. Enrichment of the top cluster was studied using the STRING tool (version 11.5) and the significant values were denoted as False discovery rate (FDR).
Analysis of hub-genes
The top 10 hub genes were identified from the top cluster of each group through the CytoHubba plugin of Cytoscape. The hub genes were identified using each of the topological analysis methods such as Betweenness, BottleNeck, Closeness, Clustering Co-efficient, Degree, Density of Maximum Neighborhood Component (DMNC), EcCentricity, Edge Percolated Component (EPC), Maximal Clique Centrality (MCC), Maximum Neighborhood Component (MNC), Radiality, and Stress. The common 10 genes from each of these topological analysis methods were shortlisted as the key hub genes in our study since the overlap of all methods gives the most probable hub genes43. From the shortlisted hub genes, the genes with the highest average FPKM from each group were chosen for RT-qPCR validation.
RNA isolation and cDNA synthesis
Buffalo sperm RNA was isolated using the previously established protocol established in our laboratory44. Briefly, frozen semen samples were washed using 50% Bovipure (Nidacon, Sweden) solution. After washing, 40 × 106 sperm per sample were taken and lysed with a double lysis method followed by extraction using a silica membrane-based column (PureLink RNA mini kit, Invitrogen, USA). The extracted total RNA was subjected to DNase treatment (Turbo DNA-free kit, Ambion, USA) to remove the contaminating genomic DNA. The total RNA concentration was measured using a fluorometer (Qubit 4.0, Invitrogen, USA) and RNA quality was measured using a spectrophotometer (NanoDrop ND-1000, Thermo Scientific, USA). The RNA integrity was estimated using Bioanalyzer (2100 Bioanalyzer, Agilent Technologies, USA).
Total sperm RNA (100 ng) was used for library preparation and amplification using the NEBNext Ultra II Directional RNA library kit (New England Biolabs, USA). Precisely, total RNA was reverse transcribed to cDNA using random hexamers. The complementary strand was synthesized using DNA polymerase-I, subsequently amplified for 15 PCR cycles and used for gene expression studies.
Gene expression studies
The expression levels of the hub genes identified from the CytoHubba tool were quantified using the RT-qPCR (StepOnePlus, Applied Biosystems, USA). Each RT-qPCR reaction consists of an equal concentration of amplified cDNA from each bull, 1X SYBR Green Mastermix with ROX (TB Green Premix Ex Taq II, Takara Bio, Japan) and 125 nM each of forward and reverse primers (Table 6). The absence of RNA from other contaminating cells was ensured in each sperm sample with the cell-specific primers for KIT (germ cells), CDH1 (epithelial cell) and PTPRC (leukocytes) (Supplementary Fig. S4). The PCR cycle conditions were 95 °C for 30 s, 40 cycles of 95 °C for 5 s and 60 °C for 1 min followed by the default melt curve settings. The data were acquired and analyzed in the StepOne software (v2.2.2). Relative gene expression levels were calculated using the 2−ΔΔCt method45 using GAPDHS as the housekeeping gene. The PCR products were also checked using 2% agarose gel electrophoresis (Supplementary Fig. S4).
Quantitative dot blot
Quantitative dot blot was performed with minor modifications46. Differential protein estimation was done using new set of bulls (n = 13). Total sperm lysate (1 µg/bull) from high (n = 8) and low (n = 5) fertility groups were applied directly to the nitrocellulose membrane. The membrane was allowed to dry at 37 °C for 30 min and blocked with 5% bovine serum albumin in TBST (19.97 mM Tris–HCl, 136.89 mM NaCl, pH 7.6 plus 0.1% Tween-20) and incubated at 37 °C for 1 h. Then, the primary antibodies, rabbit anti-RPS18 polyclonal antibody (A11687, ABclonal, USA), rabbit anti-EXT2 polyclonal antibody (A1973, ABclonal, USA), and rabbit anti-RPLP0 polyclonal antibody (PA5-41717, ThermoFisher, USA) were added to the respective membrane and incubated at 37 °C for 120 min. Rabbit anti-GAPDHS polyclonal antibody (A10471, ABclonal, USA) was used as housekeeping protein. Then the membrane was washed thrice for 5 min each in TBST buffer. Later, the secondary antibody was added to the membrane and incubated at 37 °C for 60 min. After incubation, the membrane was washed in TBST for three times. The image was developed using ECL substrate solution (Immobilon ECL Ultra substrate, Thermofisher, USA) and the signals were captured by chemi-documentation system (iChemi XR, Syngene, UK). The relative abundance of the protein between high and low fertility group was calculated based on densitometric analysis using GENE TOOLS software (Syngene, UK).
Receiver operating characteristic (ROC) curve analysis
The fertility predictive value of the validated genes was evaluated using ROC curve analysis. First, the predictive power of individual genes was analyzed using univariate regression analysis. Subsequently, multivariate regression analysis was performed by combining the expression levels of these genes. The linear regression models were developed by employing PM, AI, FMI and FR as independent variables and ΔCt of the genes as the dependent variable. The ROC analysis was performed to assess the sensitivity (%), specificity (%), accuracy (%) and diagnostic efficiency (%) of the univariate and multivariate regression models at the chosen cut-off with a maximum likelihood ratio for classifying the bulls into their respective good or poor categories34.
Statistical analysis
The sperm functions were subjected to statistical analysis using the IBM SPSS statistics 20 and GraphPad Prism 6. All the sperm function data were normally distributed and hence the Student’s t-test was used for calculating the significance between the groups. The correlation between the gene expression levels and the functional parameters was analyzed using the Pearson correlation coefficient. The correlation (r) values of < 0.1, 0.1 to 0.3, 0.3 to 0.5 and > 0.5 were considered trivial, small to medium, medium to large and large to very large, respectively47. All the values were presented as mean ± SEM and the significance is set at p < 0.05.
Data availability
RNA-seq data are not generated in this study. The data used for the present study are not publicly available due to proprietary nature but are available from the corresponding author on reasonable request.
References
Selvaraju, S. et al. Deciphering the complexity of sperm transcriptome reveals genes governing functional membrane and acrosome integrities potentially influence fertility. Cell Tissue Res. https://doi.org/10.1007/s00441-021-03443-6 (2021).
Amann, R. P., Saacke, R. G., Barbato, G. F. & Waberski, D. Measuring male-to-male differences in fertility or effects of semen treatments. Annu. Rev. Anim. Biosci. 6, 255–286 (2018).
Sendler, E. et al. Stability, delivery and functions of human sperm RNAs at fertilization. Nucleic Acids Res. 41, 4104–4117 (2013).
Somashekar, L. et al. Comparative sperm protein profiling in bulls differing in fertility and identification of phosphatidylethanolamine-binding protein 4, a potential fertility marker. Andrology 5, 1032–1051 (2017).
Mathur, P. P., Francispillai, M., Vaithinathan, S. & Agarwal, A. NF-κB in male reproduction: A boon or a bane ?. Open Reprod. Sci. J. 3, 85–91 (2011).
Doering, T. A. et al. Network analysis reveals centrally connected genes and pathways involved in CD8 + T cell exhaustion versus memory. Immunity 37, 1130–1144 (2012).
Goymer, P. Why do we need hubs?. Nat. Rev. Genet. 9, 1000140 (2008).
Xu, S. et al. Identification of hub genes for early diagnosis and predicting prognosis in colon adenocarcinoma. Biomed Res. Int. 2022, 1893351 (2022).
Bakhtiarizadeh, M. R., Mirzaei, S., Norouzi, M. & Sheybani, N. Identification of gene modules and hub genes involved in mastitis development using a systems biology approach. Front. Genet. 11, 1–16 (2020).
dos Santos Silva, D. B. et al. Prediction of hub genes associated with intramuscular fat content in Nelore cattle. BMC Genom. 20, 520 (2019).
Jiang, Z. et al. Transcriptional profiles of bovine in vivo pre-implantation development. BMC Genom. 15, 756 (2014).
Lamirande, D. Human sperm hyperactivation and capacitation as parts of an oxidative process. Free Radic. Biol. Med. 14, 157–166 (1993).
Guthrie, H. D. & Welch, G. R. Effects of reactive oxygen species on sperm function. Theriogenology 78, 1700–1708 (2012).
Kumar, P., Saini, M., Kumar, D., Bharadwaj, A. & Yadav, P. S. Estimation of endogenous levels of osteopontin, total antioxidant capacity and malondialdehyde in seminal plasma: Application for fertility assessment in buffalo (Bubalus bubalis) bulls. Reprod. Domest. Anim. https://doi.org/10.1111/rda.12882 (2016).
Bajpai, M. Effect of tyrosine kinase inhibitors on tyrosine phosphorylation and motility. Arch. Androl. 246, 229–246 (2003).
Almog, T. et al. Identification of extracellular signal-regulated kinase 1/2 and p38 MAPK as regulators of human sperm motility and acrosome reaction and as predictors of poor spermatozoan quality. J. Biol. Chem. 283, 14479–14489 (2008).
Rahamim, L., Almog, T., Yao, Z., Seger, R. & Naor, Z. A-Kinase Anchoring Protein 4 (AKAP4) is an ERK1/2 substrate and a switch molecule between cAMP/PKA and PKC/ERK1/2 in human spermatozoa. Sci. Rep. https://doi.org/10.1038/srep37922 (2016).
Shan, S., Xu, F., Hirschfeld, M. & Brenig, B. Sperm lipid markers of male fertility in mammals. Int. J. Mol. Sci. 22, 8767 (2021).
Lenzi, A., Picardo, M., Gandini, L. & Dondero, F. Lipids of the sperm plasma membrane: From polyunsaturated fatty acids considered as markers of sperm function to possible scavenger therapy. Hum. Reprod. Update 2, 246–256 (1996).
Mizuno, Y. et al. Tysnd1 Deficiency in Mice interferes with the peroxisomal localization of PTS2 enzymes, causing lipid metabolic abnormalities and male infertility. PLoS Genet. 9, 1–16 (2013).
Roqueta-rivera, M., Abbott, T. L., Sivaguru, M., Hess, R. A. & Nakamura, M. T. Deficiency in the Omega-3 fatty acid pathway results in failure of acrosome biogenesis in mice. Biol. Reprod. 732, 721–732 (2011).
Ah-cann, C. et al. Male sterility in Mcl-1-flox mice is not due to enhanced Mcl1 protein stability. Cell Death Dis. 7, e2490–e2492 (2016).
Luna, C., Mendoza, N., Casao, A., Perez-pe, R. & Jose, A. JNK and p38 MAPK pathways link capacitation with apoptosis and seminal plasma proteins protect sperm by interfering with both routes. Biol. Reprod. 96, 800–815 (2017).
Ding, G. L. et al. The effects of diabetes on male fertility and epigenetic regulation during spermatogenesis. Asian J. Androl. 17, 948–953. https://doi.org/10.4103/1008-682X.150844 (2015).
Carlson, D. et al. Oviduct secretion in the cow. J. Reprod. Fertil. 22, 549–552 (1970).
Ruiz-Pesini, E., Diez-Sanchez, C., Lopez-Perez, M. J. & Enriquez, J. A. The role of the mitochondrion in sperm function: Is there a place for oxidative phosphorylation or is this a purely glycolytic process?. Curr. Top. Dev. Biol. 77, 3–19 (2007).
Bunch, D., Welch, J. E., Magyar, P. L., Eddy, E. M. & Briens, D. A. O. Glyceraldehyde 3-phosphate dehydrogenase-s protein distribution during mouse spermatogenesis. Biol. Reprod. 841, 834–841 (1998).
Zhu, Z., Zhang, W., Li, R. & Zeng, W. Reducing the glucose level in pre-treatment solution improves post-thaw boar sperm quality. Front. Vet. Sci. 9, 1–10 (2022).
Chatterjee, I., Gross, S. R., Kinzy, T. G. & Chen, K. Y. Rapid depletion of mutant eukaryotic initiation factor 5A at restrictive temperature reveals connections to actin cytoskeleton and cell cycle progression. Mol. Genet. Genom. 275, 264–276. https://doi.org/10.1007/s00438-005-0086-4 (2006).
Gobert, A. P. et al. Article Hypusination orchestrates the antimicrobial response of macrophages. Cell Rep. 33, 108510 (2020).
Song, H., Wang, L., Chen, D. & Li, F. The function of Pre-mRNA alternative splicing in mammal spermatogenesis. Int. J. Biol. Sci. 16, 38–48 (2020).
Bansal, S. K., Gupta, N., Sankhwar, S. N. & Rajender, S. Differential genes expression between fertile and infertile spermatozoa revealed by transcriptome analysis. PLoS One https://doi.org/10.1371/journal.pone.0127007 (2015).
Zhang, T. et al. System analysis of teratozoospermia mRNA profile based on integrated bioinformatics tools. Mol. Med. Rep. 18, 1297–1304 (2018).
Selvaraju, S. et al. Orchestrating the expression levels of sperm mRNAs reveals CCDC174 as an important determinant of semen quality and bull fertility. Syst. Biol. Reprod. Med. 67, 1–13 (2021).
Bonache, S., Mata, A. & Larriba, S. Sperm gene expression profile is related to pregnancy rate after insemination and is predictive of low fecundity in normozoospermic men. Hum. Reprod. 27, 1556–1567 (2012).
Lervik, S. et al. Gene expression during testis development in Duroc boars. Animal 9, 1832–1842. https://doi.org/10.1017/S1751731115000907 (2015).
Genuth, N. R. & Barna, M. The discovery of ribosome heterogeneity and its implications for gene regulation and organismal life. Mol. Cell 7, 364–374 (2018).
Ensslin, M. A., Lyng, R., Raymond, A., Copland, S. & Shur, B. D. Novel gamete receptors that facilitate sperm adhesion to the egg coat. Soc. Reprod. Fertil. Suppl. 63, 367–383 (2007).
Gerszten, R. & Wang, T. J. The search for new cardiovascular biomarkers. Nature 451, 949–952. https://doi.org/10.1038/nature06802 (2008).
Swathi, D. et al. X chromosome-linked genes in the mature sperm influence semen quality and fertility of breeding bulls. Gene 839, 146727 (2022).
Selvaraju, S., Ravindra, J. P., Ghosh, J., Gupta, P. S. P. & Suresh, K. P. Evaluation of sperm functional attributes in relation to in vitro sperm-zona pellucida binding ability and cleavage rate in assessing frozen thawed buffalo (Bubalus bubalis) semen quality. Anim. Reprod. Sci. 106, 311–321 (2008).
Selvaraju, S. et al. Evaluation of maize grain and polyunsaturated fatty acid (PUFA) as energy sources for breeding rams based on hormonal, sperm functional parameters and fertility. Reprod. Fertil. Dev. 24, 669–678 (2012).
Kaur, B., Mukhlis, Y., Natesh, J., Penta, D. & Meeran, S. M. Identification of hub genes associated with EMT-induced chemoresistance in breast cancer using integrated bioinformatics analysis. Gene 809, 146016 (2022).
Parthipan, S. et al. Spermatozoa input concentrations and RNA isolation methods on RNA yield and quality in bull (Bos taurus). Anal. Biochem. 482, 32–39 (2015).
Livak, K. J. & Schmittgen, T. D. Analysis of relative gene expression data using real-time quantitative PCR and the 2−ΔΔCT method. Methods 25, 402–408 (2001).
Tian, G. et al. Quantitative dot blot analysis (QDB), a versatile high throughput immunoblot method. Oncotarget 8, 58553 (2017).
Cohen, J. A power primer. Quant. Methods Psychol. 112, 155–159 (1992).
Acknowledgements
The authors sincerely acknowledge Dr. Raghavendra Bhatta, Director, ICAR-NIANP, Bengaluru, India for his critical technical inputs and necessary facilities to carry out this work.
Funding
This research was carried out under the ICAR-National Fellow project funded by the Indian Council of Agricultural Research, Government of India. Dr. S. Selvaraju is supported by the ICAR-National Fellow project, ICAR, Ministry of Agriculture, Government of India.
Author information
Authors and Affiliations
Contributions
Designed the experiment: D.S. and S.S.; Conducted the experiment: D.S., L.R., and S.S.A.; Analyzed the data and drafted the manuscript: D.S., S.S., and B.K., B.K.B.; All authors reviewed and finalized the manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Swathi, D., Ramya, L., Archana, S.S. et al. Identification of hub genes and their expression profiling for predicting buffalo (Bubalus bubalis) semen quality and fertility. Sci Rep 13, 22126 (2023). https://doi.org/10.1038/s41598-023-48925-5
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-023-48925-5
- Springer Nature Limited