Reagents
All reagents used in the present study were purchased from Sigma (Merck, Darmstadt, Germany) unless stated otherwise.
Animals
Semen samples were provided by a local farm (Gepork S.L.; Masies de Roda, Spain), which follows the ISO certification (ISO-9001:2008) and the EU Directive 2010/63/EU for animal experiments, the Animal Welfare Law issued by the Regional Government of Catalonia, and the current regulation on Health and Biosafety issued by the Department of Agriculture, Livestock, Food and Fisheries, Regional Government of Catalonia, Spain. As ejaculates were commercially acquired from an artificial insemination centre and animals were not manipulated for the sole purpose of the present experiment, permission from an Ethics Committee was not required.
Ejaculates from healthy and sexually mature Pietrain boars (1-3 years old) were collected using the hand-gloved method. Immediately after collection, semen samples were diluted to a final concentration of 33×106 sperm/mL using a commercial extender (Vitasem LD, Magapor S.L., Zaragoza, Spain) and stored at 17 ºC for 24 h. Upon arrival at the laboratory, semen quality was assessed. Only those samples whose sperm viability was greater than 80% and whose motility was greater than 70% motile sperm were included in the study.
On the other hand, ovaries were recovered from pre-pubertal gilts sacrificed for food purposes at a local abattoir (Frigorífics Costa Brava; Riudellots de la Selva, Girona).
Experimental design
Sixteen ejaculates that met the quality standards (each came from a separate boar; i.e., 16 boars) were included in the present study and split into three aliquots. The first was used to assess sperm quality and function. The second was intended to IVF experiments. In brief, a total of 650 oocytes were matured, fertilised and both the fertilisation rate (day 2) and rates of embryo at different pre-implantation stages (day 6) were recorded. Finally, the third aliquot was stored at –80 ºC and later served to investigate sperm metabolomics through LC-MS/MS.
Sperm quality evaluation
Sperm motility
Semen samples were pre-warmed at 38 ºC for 15 min, and 3 µL was placed into a Leja20 counting chamber (Leja Products BV; Nieuw-Vennep, The Netherlands). Following this, samples were evaluated under an Olympus BX41 microscope (Olympus; Tokyo, Japan) with a negative phase contrast objective (Olympus 10× 0.30 PLAN objective, Olympus), through a computer-assisted sperm analysis (CASA) system (Integrated Sperm Analysis System, ISAS V1.0; Proiser S.L.; Valencia, Spain). Two technical replicates were evaluated per sample and at least 1,000 sperm were examined in each replicate.
Two different parameters were recorded: the percentage of motile sperm, which considered those motile sperm whose average path velocity (VAP) was ≥ 10 µm/s; and the percentage of sperm with progressive motility, which included those motile sperm that exhibited a percentage of straightness (STR) ≥ 45 %.
Sperm morphology
Sperm morphology was evaluated after dilution in 0.12 % formaldehyde saline solution (PanReac AppliChem; Darmstadt, Germany; 1:1, v:v). Samples were observed under a phase-contrast microscope at 1,000× magnification (Nikon Labophot; Nikon; Tokio, Japan), and 200 sperm cells were examined. Sperm cells were graded as morphologically normal, or with primary or secondary alterations 29. The percentage of normal sperm was calculated from those without morphological alterations.
Sperm viability assessment
Sperm viability was assessed following the protocol of Garner and Johnson50, which uses SYBR-14, that stains sperm nuclei, and propidium iodide (PI), that only stains sperm having a compromised plasma membrane integrity. Briefly, semen samples were adjusted to a final concentration of 4×106 sperm/mL in 1× phosphate buffered saline (PBS). Next, samples were incubated for 15 min at 38 ºC with SYBR-14 (final concentration: 32 nM) and PI (final concentration: 7.5 µM). Stained cells were analysed using a CytoFLEX cytometer (Beckman Coulter; Fullerton, CA, USA), where SYBR-14 fluorescence was detected by the fluorescein isothiocyanate (FITC) channel (525/40), and PI using the PC5.5 channel (690/50). Both fluorochromes were excited with a 488-nm laser and no spill compensation was applied. Two technical replicates of at least 10,000 sperm were analysed at constant flow rate, laser voltage and sperm concentration. The percentage of viable sperm corresponded to the SYBR-14+/PI- population, after subtracting the percentage of debris particles in the analysis.
Sperm functionality assessment
Sperm functionality was evaluated through the assessment of sperm intracellular calcium, acrosome membrane integrity and mitochondrial membrane potential using CytoFLEX cytometer.
Sperm intracellular calcium levels were assessed following Harrison, Mairet, & Miller, 1993; sperm were stained with Fluo3-AM (final concentration: 1.2 µM) and PI (final concentration: 5.6 µM) for 10 min at 38 ºC in the dark. Fluo3 was detected through the FITC channel (525/40). The mean of Fluo3 fluorescence intensity per sperm (Fluo3+/PI-) was recorded and used for statistical analysis.
Acrosome membrane integrity was evaluated following Nagy, Jansen, Toppe, & Gadella, 2003 protocol, in which sperm were stained with PNA-FITC (final concentration: 1.2 µM) for 5 min at 38 ºC in the dark, and then with PI (final concentration: 5.6 µM) for 5 min at 38 ºC in the dark. PNA-FITC was detected by the FITC channel (525/40). The percentage of viable sperm with an intact acrosome membrane (PNA-FITC-/PI-) was recorded and used for the subsequent statistical analysis
Mitochondrial membrane potential was evaluated following the protocol set by Ortega-Ferrusola et al., 2008. Sperm were incubated with JC-1 (final concentration: 750 nmol/L) for 30 min at 38 ºC in the dark. In cells with high mitochondrial membrane potential, JC-1 aggregates and emits orange fluorescence, which is collected through the PE channel. On the contrary, in cells with low mitochondrial membrane potential, JC-1 is found in its monomeric form and generates green fluorescence, which is collected through the FITC channel. The ratio between orange and green fluorescence of JC-1 was calculated to evaluate the mitochondrial membrane potential of sperm.
Oocyte maturation, in vitro fertilisation, and embryo culture
Ovaries were transported to the laboratory in 0.9 % NaCl supplemented with 70 µg/mL kanamycin at 38 ºC. Cumulus-oocyte complexes (COC) were retrieved from follicles and selected in Dulbecco’s PBS (Gibco, ThermoFisher) supplemented with 4 mg/mL BSA. Only COCs exhibiting a complete and compact cumulus mass were included in the study.
For in vitro maturation of oocytes (IVM), TCM-199 (Gibco) supplemented with 0.57 mM cysteine, 0.1 % (w:v) polyvinyl alcohol, 10 ng/mL human epidermal growth factor, 75 µg/mL penicillin-G potassium, and 50 µg/mL streptomycin sulphate was used. COCs were matured in groups of 40-50 in four-well multi-dishes (Nunc, ThermoFisher; Waltham, MS, USA) containing 500 µL of pre-equilibrated maturation medium supplemented with 10 IU/mL equine chorionic gonadotropin (eCG; Folligon; Intervet International B.V.; Boxmeer, The Netherlands) and 10 IU/mL human chorionic gonadotropin (hCG; Veterin Corion; Divasa Farmavic S.A.; Gurb, Barcelona, Spain). After 20-22 h, oocytes were transferred into 500 µL fresh, pre-equilibrated IVM medium without hormones.
Next, mature oocytes were placed in 50-µL drops of pre-equilibrated IVF medium (Tris-buffered medium 54) containing 1 mM caffeine. Semen samples were adjusted to 1,000 sperm per oocyte in IVF medium and, thereafter, oocytes and sperm were co-incubated for 5 h in the incubator; a total of 40 oocytes per ejaculate were inseminated. The potentially fertilised oocytes were subsequently washed and transferred into 500 μL NCSU23 medium 55 supplemented with 0.4 % BSA, 0.3 mM pyruvate and 4.5 mM lactate for embryo in vitro culture (IVC). After 2 days, cleaved embryos were counted to calculate fertilisation rates, and then transferred into NCSU23 medium supplemented with 0.4 % BSA and 5.5 mM glucose. At day 6 post-fertilisation, the resulting embryos were classified following Balaban et al. 56 criteria. Specifically, the percentages of morulae, early blastocysts/blastocysts, hatching/hatched blastocysts and total embryos (sum of morulae, early blastocysts/blastocysts and hatching/hatched blastocysts) were evaluated. In addition, the sum of morulae, early blastocysts/blastocysts and hatching/hatched blastocysts was also determined to calculate the percentage of embryos in advanced stages. Finally, two different ratios were calculated, i) the developmental potential at day 6, which corresponded to the percentage of morula, early blastocysts/ blastocysts plus hatched/hatching blastocysts divided by the percentage of 2-8 cell embryos; and ii) the developmental competency of fertilised embryos, calculated as the ratio between the number of embryos at day 2 and those at day 6.
All procedures (oocyte IVM, IVF, and IVC) were carried out at 38.5 °C under a humidified atmosphere of 5 % CO2 in air.
Metabolomic analysis
Sperm lysis
A total of 100 million sperm were lysed in 500 µL of lysis buffer (0.1% SDS 0.1% Triton in PBS). After samples were vortexed for 45 min at 4 °C, and lysates were centrifuged at 18,000 g for 20 min at 4 °C. Supernatants were recovered and stored at -80 °C until LC-MS/MS analysis was carried out. Two technical replicates per semen sample were processed. In addition, and in order to prepare the blank, all the protocols were applied in parallel to four replicates that did not contain sperm samples.
LC-MS/MS analysis
Cell lysates were analysed by adapting a previously reported method for the quantification of polycarboxylic acids 57. The method involved a derivatisation with o-benzylhydroxylamine, a liquid-liquid extraction with ethyl acetate and LC-MS/MS detection using selected reaction monitoring mode. A LC-MS/MS system consisting of an Acquity UPLC instrument (Waters Associates, Milford, MA, USA) coupled to a triple quadrupole (TQS Micro, Waters) mass spectrometer was used for the analysis. Lactic acid, citric acid, isocitric acid, α-ketoglutarate, succinic acid, fumaric acid, malic acid, acetoacetate and α-hydroxyglutarate were determined. In addition to the concentration of each metabolite, nine ratios between metabolites with potential information about enzyme activity were calculated. MassLynx software V4.1 (Waters Associates) was used for peak integration and data management.
Statistical Analysis
Data analysis
Data preprocessing and statistical analyses were conducted using the R software version 4.2.0. First, missing values were replaced by half of the minimum value within the dataset. Shapiro-Wilk test was used to assess normality. The metabolomics dataset was log-transformed before modelling.
Sperm physiology and in vitro fertility parameters were classified into three main blocks: sperm quality, sperm function and IVF outcomes, and were analysed separately. Each group was log-transformed and scaled prior to running PCA for dimensionality reduction purposes. PCA disposes an orthogonal projection onto a lower dimensional subspace, which captures the majority of the variance of the dataset 58. Then, variables of each block were projected onto a few principal loading vectors independently, condensing most of the variability of the original data 59. Score values from the first PC of each block were utilized as a reduced-dimension feature-vector in the response block (Y-block), predicted in function of the metabolomics set (X-block) using a multivariate PLS regression. The generation of the PLS model was carried out through the root mean square error of prediction as metric in a repeated double cross-validation framework 60 including a recursive ranking based on variable importance in projection and sequential backward feature removal 61. The whole operation was repeated 20 times for improved coverage of inner and outer segments and modelling performance. The model performance was assessed by means of a permutation test of 500 iterations between permuted models, with a random assignation of the observations, and the actual model obtained. Furthermore, linear models were performed on metabolic data using the reduced-dimension feature-vectors as response. The Benjamini–Hochberg procedure was carried out on all analyses to control the false discovery rate (FDR) 62. Only P-values lower than 0.05 were considered statistically significant.
Multi-block data integration
Integration of multiple data sets measured on the same observations were carried out utilizing the N-integration with Projection to Latent Structures model 63. This model was performed to assess multi-block correlations between sperm quality, sperm function, IVF outcomes, and metabolomic blocks from the same observational units, using the mixOmics R package v 6.18.1. 64. Pair-wise similarity matrix was obtained from the two correlated latent components obtained through the projection to latent structures method. A relevance network graph was created to describe connections between the four datasets, based on the rule of similarity score ≥ 0.3 65.