ORIGINAL RESEARCH published: 28 May 2021 doi: 10.3389/fcell.2021.669974 Metabolite Profiling of Pig Seminal Plasma Identifies Potential Biomarkers for Sperm Resilience to Liquid Preservation Yentel Mateo-Otero 1,2 , Pol Fernández-López 3 , Jordi Ribas-Maynou 1,2 , Jordi Roca 4 , Jordi Miró 5 , Marc Yeste 1,2* † and Isabel Barranco 1,2,6* † 1 Biotechnology of Animal and Human Reproduction (TechnoSperm), Institute of Food and Agricultural Technology, University Edited by: of Girona, Girona, Spain, 2 Unit of Cell Biology, Department of Biology, Faculty of Sciences, University of Girona, Girona, Ana Josefa Soler, Spain, 3 Centre d’Estudis Avançats de Blanes (CEAB), Spanish Research Council (CSIC), Girona, Spain, 4 Department University of Castilla-La Mancha, of Animal Medicine and Surgery, Faculty of Veterinary Medicine, University of Murcia, Murcia, Spain, 5 Equine Reproduction Spain Service, Department of Animal Medicine and Surgery, Faculty of Veterinary Medicine, Autonomous University of Barcelona, Barcelona, Spain, 6 Department of Veterinary Medical Sciences, Via Tolara di Sopra, Bologna, Italy Reviewed by: Felipe Martínez-Pastor, Universidad de León, Spain Metabolomic approaches allow the study of downstream gene expression events since Hideko Sone, metabolites are considered as the products of cell signaling pathways. For this reason, Yokohama College of Pharmacy, Japan many studies in humans have already been conducted to determine the influence of *Correspondence: the metabolites present in seminal plasma (SP) on sperm physiology, and to identify Marc Yeste putative biomarkers. However, in livestock species, these relationships are yet to be marc.yeste@udg.edu Isabel Barranco uncovered. Thus, the present study aimed to explore: (i) if concentrations of metabolites isabel.barranco@udg.edu in pig SP are related to sperm quality and functionality, and (ii) if they could predict the † These authors have contributed sperm resilience to liquid storage at 17◦ C. To this end, 28 ejaculates were individually equally to this work and share senior authorship collected and split into three aliquots: one was used for SP analysis through nuclear magnetic resonance (NMR) spectroscopy; another served for the evaluation of sperm Specialty section: concentration and morphology; and the last one was utilized to determine sperm This article was submitted to Cell Growth and Division, functionality parameters using computer-assisted sperm analysis (CASA) and flow a section of the journal cytometry after 0 h and 72 h of liquid-storage at 17◦ C. NMR analysis allowed the Frontiers in Cell and Developmental identification and quantification of 23 metabolites present in pig SP which, except for Biology fumarate, were not observed to follow a breed-dependent behavior. Moreover, specific Received: 19 February 2021 Accepted: 28 April 2021 relationships between metabolites and sperm variables were identified: (i) glutamate, Published: 28 May 2021 methanol, trimethylamine N-oxide, carnitine, and isoleucine were seen to be related to Citation: some sperm quality and functionality parameters evaluated immediately after semen Mateo-Otero Y, Fernández-López P, Ribas-Maynou J, collection; (ii) leucine, hypotaurine, carnitine and isoleucine were found to be associated Roca J, Miró J, Yeste M and to the sperm ability to withstand liquid storage; and (iii) Bayesian multiple regression Barranco I (2021) Metabolite Profiling models allowed the identification of metabolite patterns for specific sperm parameters of Pig Seminal Plasma Identifies Potential Biomarkers for Sperm at both 0 h and 72 h. The identification of these relationships opens up the possibility of Resilience to Liquid Preservation. further investigating these metabolites as potential sperm functional biomarkers. Front. Cell Dev. Biol. 9:669974. doi: 10.3389/fcell.2021.669974 Keywords: metabolite, metabolomic, seminal plasma, liquid storage, sperm quality, biomarkers, pig Frontiers in Cell and Developmental Biology | www.frontiersin.org 1 May 2021 | Volume 9 | Article 669974 Mateo-Otero et al. SP-Metabolites as Sperm Quality Biomarkers INTRODUCTION In a previous study, we described, for the first time, the pig SP metabolomic profile and we suggested that Artificial insemination (AI) remains the most valuable and certain SP-metabolites could play a key role in reproductive efficient reproductive biotechnology used for pig breeding, processes (Mateo-Otero et al., 2020). As far as we know, there gathering more than 90% of cases in countries with intensive are no data about the relationship between SP-metabolites pork production (Roca et al., 2016; Yeste, 2017; Waberski et al., and the sperm ability to withstand liquid preservation in 2019). In spite of the advances performed in cryopreservation any species. Therefore, an exploratory approach was carried procedures, liquid storage at 17 ◦ C continues to be the most out to evaluate the putative relationship between sperm useful and preferable biotechnology for preserving pig seminal quality/functionality parameters of liquid-stored semen samples AI-doses until insemination is carried out (Kumaresan et al., and SP-metabolites identified using nuclear magnetic resonance 2009; Waberski et al., 2019). During liquid-storage of seminal (NMR) spectroscopy. Specifically, the present study aimed AI-doses, sperm undergo injuries, leading to a decline in sperm to explore: (i) whether the concentration/presence of SP- quality, functionality and fertilizing ability within the first 72 h metabolites in ejaculates collected from AI-boars is related with of storage (Waberski et al., 2011; Yeste, 2017). One of the sperm quality/functionality parameters, and (ii) if such SP- main concerns is related to differences in the sperm ability to metabolites could predict the resilience of seminal AI-doses to withstand liquid-storage between pig seminal AI-doses (Parrilla liquid preservation at 17◦ C. In this regard, these SP-metabolites et al., 2012; Zakošek Pipan et al., 2014; Roca et al., 2015). may be of practical relevance for a better assessment of pig This situation evidences that current sperm assessments used semen quality, contributing to the successful improvement of AI- for ejaculate selection (which include sperm quantity, motility, efficiency. and morphology) are not accurate enough to predict sperm fertilizing capacity after liquid preservation (Roca et al., 2016; Waberski et al., 2019). This limitation undermines the ability MATERIALS AND METHODS to face new challenges in swine industry, as implementing new AI-procedures (such as a reduction in sperm numbers/AI- Reagents dose) may lead to a decrease on in vivo fertility outcomes Except as otherwise indicated, reagents were purchased from (Roca et al., 2016). In the last years, many efforts have been Merck KGaA (Darmstadt, Germany) and fluorochromes for aimed at identifying biomarkers of sperm resilience to liquid sperm analysis from Thermo Fisher Scientific (Waltham, MA, preservation in seminal plasma (SP) (Barranco et al., 2015a,b, USA). The commercial extender used to dilute semen samples 2016, 2019, 2020; Recuero et al., 2019; Zakošek Pipan et al., was Biosem + (Magapor, Zaragoza, Spain). 2014). The heterogeneous and complex composition of SP, together with the fact that this fluid is able to interact with Experimental Design sperm, make it a valuable source of potential biomarkers Twenty-eight entire ejaculates were individually collected from (Rodríguez-Martínez et al., 2011). 28 healthy and fertile AI-boars (one ejaculate per boar) from Over the last decade, -omics technologies (including different breeds (Pietrain, Duroc, Landrace and Tempo). The proteomics, transcriptomics, metabolomics. . .) have emerged entire ejaculates were split into three aliquots. The first one was as powerful tools to uncover the molecular factors and used for SP-harvesting, on which the metabolomic analysis was regulatory networks involved in reproductive events, thus performed. The second one was utilized for sperm concentration providing a worthwhile and broad strategy for the study of and morphology assessments. The third aliquot was extended these multifactorial processes (Egea et al., 2014). Metabolomics alike an AI-dose (30 × 106 sperm/mL in Biosem +) and stored up employs spectral and analytical approaches to provide a large- to 72 h at 17◦ C (FOC 120E Cooled Incubator; VELP Scientifica, scale identification of metabolites, which are the end-products of Usmate, Italy). Sperm quality and functionality parameters were downstream events of gene expression (Goodacre et al., 2004). assessed in this latter aliquot at different time-points (0 and Hence, this technique leads to a more realistic approach of 72 h), in terms of sperm motility, viability, acrosome damage, cellular physiological activities compared with other -omics, intracellular hydrogen peroxide (H2 O2 ) generation by viable such as transcriptomics or proteomics (Singh and Sinclair, 2007). sperm, and membrane lipid disorder in viable sperm. In the last few years, the study of metabolites as markers of patho- and physiological reproductive processes, in both males Boars and Ejaculates and females, has gained much relevance (Egea et al., 2014; Uyar Ejaculates used in the present study were supplied by a Spanish and Seli, 2014; Minai-Tehrani et al., 2016; Nobakht and Gh, AI-Centre (AIM Ibérica, Topigs Norsvin Spain; Calasparra, 2018; Darbandi et al., 2019; Asampille et al., 2020). Focusing Murcia, Spain). This AI-Centre complies with Spanish on SP, several studies performed in humans reported that (ES300130640127, August 2006) and European (ES13RS04P, July certain SP-metabolites are involved on sperm physiology and 2012) guidelines for animal health and welfare, and pig ejaculate fertility disorders (Bonechi et al., 2015; Engel et al., 2019; Wang collection and commercialization of AI-doses. As semen samples et al., 2019; Mumcu et al., 2020; Xu et al., 2020). In livestock were provided by that AI-Centre and authors did not manipulate species, only few reports conducted in cattle pointed out specific any animal, no ethical permission was needed. SP-metabolites as in vivo fertility biomarkers (Kumar et al., 2015; All entire ejaculates were collected through a semi-automatic Cazaux Velho et al., 2018). collection system (Collectist , IMV Technologies, L’Aigle, R Frontiers in Cell and Developmental Biology | www.frontiersin.org 2 May 2021 | Volume 9 | Article 669974 Mateo-Otero et al. SP-Metabolites as Sperm Quality Biomarkers France) from sexually mature (1 to 3 years old), fertile boars under an Olympus BX41 microscope (Olympus, Tokyo, Japan) that were subjected to routine collection (twice per week) for with a negative phase-contrast field (Olympus 10 × 0.30 PLAN elaborating commercial AI-doses. These AI-boars were housed in objective; Olympus). Percentages of sperm that showed an individual pens with controlled environment temperature (15– average path velocity ≥ 20 µm/s (total motile sperm) and that 25◦ C) and exposed to a total of 16 h of light (natural plus exhibited a rapid and progressive movement with a straight-line artificial). They were fed with commercial feedstuff consistent velocity ≥ 40 µm/s (progressive motile sperm) were recorded. with the nutritional necessities of AI-boars, and were provided Sperm viability, acrosome damage, intracellular H2 O2 with water ad libitum. All ejaculates collected for this study production and membrane lipid disorder were assessed with fulfilled the sperm quality thresholds for commercial AI- a flow cytometer (BD FACS Canto II; Becton Dickinson & doses (namely a minimum sperm concentration of 200 × 106 Company, Franklin Lakes, NJ, USA). The optical configuration sperm/mL and more than 70% of motile sperm and 75% of sperm of this device included three detector arrays: one octagon, which with normal morphology). detects light from 488 nm laser (blue), and two trigons, which detect light from 633 nm (red) and 405 nm (violet) lasers. Data Seminal Plasma Processing and Storage were collected and compensated using BD FACSDiva Software For SP-harvesting, ejaculates were subjected to double- (Becton Dickinson & Company). Events were triggered by centrifugation right after collection (1,500 × g for 10 min at room forward scatter (FSC) and side scatter (SSC), and non-sperm temperature Rotofix 32A, Hettich Centrifuge United Kingdom, particles were gated out based on Hoechst 33342 (H-42) Newport Pagnell, Buckinghamshire, England, United Kingdom). fluorescence, which was detected using 450/50 nm band-pass Subsequently, SP samples (second supernatants) were (BD) filter. For each semen sample and sperm parameter, microscopically analyzed (Eclipse E400; Nikon, Tokyo, Japan) three technical replicates (with 10,000 [H-42]-positive events to warrant sperm absence. Finally, SP-samples were stored in each one) were evaluated. All fluorescence parameters were cryotubes at −80◦ C (Ultra Low Freezer; Haier Inc., Qingdao, log-transformed, whereas FSC and SSC signals were linearly China) until metabolomic analysis was carried out, when they processed. Unstained and single-stained samples were used for were thawed on ice. setting electronic volume (EV) gain and compensations. For all samples, laser voltage and flow rate were constant. Sperm viability and acrosome damage were assessed by triple- Assessment of Sperm Quality and staining with H-42, propidium iodide (PI) and fluorescein- Functionality Parameters conjugated peanut agglutinin (PNA-FITC). Fluorescence of PI Eight sperm quantity and functionality parameters were assessed and PNA-FITC was detected using 670 nm long-pass (LP) in each semen sample, specifically: (i) sperm concentration, (ii) filter and 530/30 nm BP filter, respectively. With this purpose, total sperm count, (iii) sperm morphology, (iv) sperm motility, 100 µL of each semen sample (30 × 106 sperm/mL in (v) sperm plasma membrane integrity (viability), (vi) acrosome Biosem +) were incubated with 3 µL H-42 (0.05 mg/mL damage in viable sperm, (vii) intracellular H2 O2 generation in phosphate-buffered saline, PBS), 2 µL PI (0.5 mg/mL in by viable sperm, and (viii) plasma membrane lipid disorder PBS) and 2 µL PNA-FITC (100 µg/mL in PBS) for 10 min of viable sperm. at 37◦ C (Sanyo MIR-153 incubator, Gemini BV, Apeldoorn, Sperm concentration was assessed by a high-precision Netherlands). After this period, 400 µL PBS were added to each automated cell counter (NucleoCounter NC-100TM , R sample. Percentages of viable spermatozoa (H-42 + /PI-) with ChemoMetec, Allerod, Denmark). Then, the total sperm an intact (PNA-FITC-) and non-intact (PNA-FITC +) acrosome number of each ejaculate was calculated by multiplying membrane were recorded. sperm concentration per ejaculate volume (assessed in a To assess the intracellular H2 O2 production by viable sperm, graduated tube). a triple-staining with H-42, PI, and 5- and 6-chloromethyl- For sperm morphology assessment, semen samples were fixed 20 ,70 -dichlorodihydrofluorescein diacetate acetyl ester (CM- with 0.12% formaldehyde saline solution (Panreac, Barcelona, H2 DCFDA) was performed. Fluorescence of CM-H2 DCFDA was Spain) and examined under a phase-contrast microscope at detected using 530/30 nm BP filter following excitation at 488 nm. 1,000 × magnification (Nikon Labophot, Nikon, Tokio, Japan). Briefly, 50 µL of each semen sample (30 × 106 sperm/mL in A total of 200 sperm per sample were counted and classified as: Biosem +) were incubated with 1.5 µL H-42 (0.05 mg/mL in (i) morphologically normal, (ii) with abnormal head, (iii) with PBS), 1 µL PI (0.5 mg/mL in PBS), and 1 µL CM-H2 DCFDA acrosome defects, (iv) with proximal cytoplasmic droplets, (v) (1 mM in dymetilsulfoxide, DMSO) in 950 µL of PBS at 37◦ C with distal cytoplasmic droplets, (vi) with folded tails, and (vii) for 30 min. A semen sample incubated with all fluorochromes with coiled tails. plus 1 µL of tert-butyl hydroperoxide solution (70% in distilled Sperm motility was evaluated using a computer-assisted sperm water) was used as a control. The percentage of viable sperm analyzer (CASA, ISASV1 , Proiser R + D S.L., Paterna, Spain). R (H-42 + /PI-) that exhibited high intracellular H2 O2 generation With this purpose, 5 µL of each semen sample (30 × 106 (20 ,70 -dichlorofluorescein, DCF +) was recorded. sperm/mL in Biosem +) were pipetted onto a Makler chamber To evaluate membrane lipid disorder of viable sperm, a triple- (Sefi Medical Instruments, Haifa, Israel), previously warmed to staining with H-42, Yo-Pro-1, and Merocyanine 540 (M-540) was 38◦ C. In each semen sample, a total of ten different fields, carried out. Fluorescence of M-540 and Yo-Pro-1 was detected gathering more than 600 sperm, were acquired and examined using 585/42 nm and 530/30 nm BP filters, respectively. Fifty µL Frontiers in Cell and Developmental Biology | www.frontiersin.org 3 May 2021 | Volume 9 | Article 669974 Mateo-Otero et al. SP-Metabolites as Sperm Quality Biomarkers of each semen sample (30 × 106 sperm/mL in Biosem +) were Statistical Analysis incubated with 2.5 µL H-42 and 10 µL Yo-Pro-1 (2.5 µM in All analysis were performed using R software (version 4.0.2)1 . DMSO) in 950 µL of PBS at 37◦ C for 8 min. After this period, For all analysis, the level of significance was set at P ≤ 0.05. 26 µL of M540 (0.1 mM in DMSO) were added to each sample Statistical analysis of NMR data was performed in three steps. and incubated at 37◦ C for further 2 min. The percentage of viable First, the breed effect on SP-metabolite concentration/presence spermatozoa (H-42 + /Yo-Pro-1-) that exhibited high plasma was assessed performing a Wilcoxon rank sum test, which avoids membrane disorder (M-540 +) was recorded. the assumptions of normality and homoscedasticity as not all SP- metabolites fulfilled these requisites. The same test was also used Ejaculate Resilience to Liquid to check whether age distribution between Duroc and Pietrain Preservation breeds differed. The ability of each sample to withstand liquid preservation was Second, principal component analysis (PCA) was carried evaluated as the loss of sperm quality after 72 h. With this out to evaluate whether data could be categorized to few purpose, the following formula was used: [(q1-q0)/q0] × 100, variables/components, as this avoided using all sperm parameters where q1 is the value of each sperm parameter after 72 h and made classifying ejaculates into two groups of high and low of liquid storage at 17◦ C, and q0 is the value of each sperm quality easier. As this could not be achieved, Pearson correlations parameter evaluated immediately after ejaculate collection (0 h). were employed to quantify the relationship between single SP- This formula was applied to the following sperm parameters: metabolites and sperm parameters. (i) percentages of viable sperm, (ii) percentages of viable sperm Third, multiple linear regression analyses were performed with a non-intact acrosome, (iii) percentages of viable sperm with using a Bayesian framework with the functionality of the R high intracellular H2 O2 levels, (iv) percentages of viable sperm package ‘rstanarm’ (R package version 2.21.1; Goodrich et al., with high plasma membrane disorder, and (v) percentages of 2020). All models were performed with a prior non-informative total motile sperm. Regarding progressive motility, proportions distribution, with the highest sensitivity and 4,000 iterations of of progressively motile sperm at 0 h or 72 h were calculated sampling. The remaining sperm parameters were left by default. relative to total motility as follows: (progressive motility)/(total The Bayesian approach was chosen over the classical frequentist motility) × 100. framework due to the nature of the dataset. Because linear regression models tend to be overfitted in datasets where the 1H number of columns (variables or predictors) is similar to the RMN Analysis number of rows (samples), Bayesian models allow quantifying A 500-µL aliquot of thawed SP-samples was vortexed, and the uncertainty of such models. In other words, it is possible to centrifuged at 14,000 g and 4 ◦ C for 90 min (Rotofix 32A, have a good sense of how much sensitive or precise these models Hettich Centrifuge UK; Newport Pagnell Buckinghamshire, are. This was performed for the two time-points (i.e., 0 h and United Kingdom) using 0.5 mL-Amicon Ultra Centrifugal R 72 h) with the particularity that, for sperm resilience to liquid Filters to remove proteins and remaining cell debris. These filters storage, logistic regression was used for modeling percentages were previously washed with PBS and prepared by centrifuging of viable sperm with high intracellular H2 O2 production and three times at 14,000 g and 4 ◦ C for 20 min. Then, the eluted high plasma membrane destabilization. For these latter analyses, fraction was mixed with 100 µL PBS containing 10% D2 O with medians were used to discriminate the two quality groups: equal 0.33% of DSS (Merck KgaA, Darmstadt, Germany; pH 7.4), and or below the median was considered as 0, and above the median transferred into a 5-mm Wilmad NMR tube (Merck KgaA) to R was considered as 1. Finally, Pearson correlations were also generate 1 H NMR spectra. calculated to determine inter-metabolite relationships. NMR spectra were obtained using a Bruker 600-MHz AVANCE III NMR spectrometer (Bruker Biospin, Rheinstetten, Germany) operating at a 1 H frequency of 600.13 MHz and a RESULTS temperature of 300 ◦ K with an equilibration period of 10 min. The 1 D-1 H-nuclear Overhauser effect spectroscopy (1 D-NOESY) Metabolite Profile of Pig SP and pulse sequence from the Bruker library was used (noesygppr1d). Relevant parameters were: mixing time: 100 ms (d8), recovery Differences Between Breeds A total of 23 metabolites were identified and quantified in all SP- delay: 2 s (d1), 90◦ pulse: 10.39 µs (p1), spectral width: samples analyzed, observing that SP-metabolite concentrations 7211.539 Hz, spectral size: 32 k, number of scans: 128, and varied between ejaculates (Table 1). The 1 H-NMR (noesygppr1d) acquisition time: 2.27 s. profile of pig SP measured at 600 MHz is shown in Figure 1. The identified metabolites were categorized in the following Data Processing and Analysis chemical classes: (i) amino acids (n = 7; alanine, glutamate, Spectra were processed and analyzed with Chenomx 8.0 profiler. isoleucine, leucine, phenylalanine, tyrosine, and valine); (ii) The software provides tools for automatic phase, baseline saccharides (n = 1; glucose); (iii) salts (n = 6; acetate, citrate, correction and reference calibration as well as metabolites formate, fumarate, lactate, and malonate); (iv) alcohols (n = 2; libraries for profiling. Determination of concentration ethanol and methanol); and (v) other organic compounds for each metabolite was calculated in relation to DSS concentration (0.216 mM). 1 https://www.r-project.org/ Frontiers in Cell and Developmental Biology | www.frontiersin.org 4 May 2021 | Volume 9 | Article 669974 Mateo-Otero et al. SP-Metabolites as Sperm Quality Biomarkers (n = 5; carnitine, creatine, creatine-phosphate, hypotaurine, was used to assess if one or few principal components myo-inositol, sn-glycero-3-phosphocholine and trimethylamine (PC) explained enough variability, so as they could be used N-oxide). to categorize the sample by high and low sperm quality After identifying SP-metabolites, the breed effect on their (Figures 3A,B). The results revealed that no PC exceeded the concentration was evaluated (Supplementary Table 1) for Duroc percentage of the variance explained (Figure 3A). Moreover, (n = 11) and Pietrain (n = 14) breeds. Samples from Tempo and eight PC were required to achieve a good proportion of the Landrace boars could not be included in the analysis because variance explained (>90%; Figure 3B). Because no PC was of the insufficient number of ejaculate samples (for Tempo, representative enough for semen quality classification, all sperm n = 2, and for Landrace, n = 1). Only the concentration of parameters were individually analyzed. fumarate exhibited significant differences between breeds, the Therefore, correlations between the concentration of concentration of this metabolite being higher in Pietrain than in SP-metabolites and specific sperm quality and functionality Duroc breeds (0.0034 mM vs. 0.0018 mM, P < 0.01; Figure 2). parameters were assessed (Figure 3C). Most of the correlations These differences in the concentration of fumarate could be were observed to be from weak (0.00 < R < 0.39) to moderate assumed to be related to the boar breed due to the fact that all (0.40 < R < 0.69) (Schober et al., 2018). The statistically animals were housed in the same building. In fact, in order to significant correlations (P < 0.050) with the highest R scores evaluate a possible age effect, the Wilcoxon rank sum test was were as follows: i) the percentage of viable sperm was positively run and evidenced no differences in age distribution between correlated with glutamate concentration in SP (R = 0.46, Pietrain and Duroc breeds (P > 0.05). Therefore, considering that P = 0.014); ii) the percentage of viable sperm with high fumarate could be influenced by the boar breed, this metabolite intracellular H2 O2 levels was positively correlated with methanol was excluded from further analyses in order to specifically concentration in SP (R = 0.42, P = 0.027); iii) the percentage of evaluate the effect of SP-metabolites on sperm physiology. viable sperm with high membrane destabilization was negatively correlated with trimethylamine N-oxide concentration in SP Relationship of SP-Metabolites With (R = −0.38, P = 0.045); and iv) the percentage of progressive motile sperm was negatively correlated with both carnitine Sperm Quality and Functionality (R = −0.39, P = 0.043) and isoleucine (R = -0.39, P = 0.041) Parameters Assessed Immediately After concentrations in SP. Semen Collection After the evaluation of individual correlations between SP- Sperm quality and functionality was first evaluated immediately metabolites and sperm quality and functionality parameters, after semen collection (Supplementary Table 2). PCA analysis Bayesian multiple linear regression analysis was carried out. This multivariate analysis allows both the development of a potentially predictive model for each parameter and the quantification TABLE 1 | Concentration of metabolites (mM) in pig seminal plasma (n = 28). of the relative contribution of SP-metabolites to the model. Additionally, the use of a Bayesian framework allows the Metabolite Median Quartile 25 Quartile 75 quantification of the uncertainty associated to each coefficient, Acetate 0.1312 0.1142 0.1431 which provides the whole distribution of the coefficients and Alanine 0.1257 0.0999 0.1778 not a single point estimate (i.e., the mean). A model was Carnitine 0.7736 0.5943 1.1612 generated for every sperm quality and functionality parameter Citrate 6.6307 4.6555 8.5421 (Supplementary Figure 1A-M). In general, due to high Creatine 0.3750 0.2487 0.4934 uncertainties associated to coefficients, relationships between Creatine phosphate 0.0538 0.0433 0.0792 metabolites and sperm quality and functionality indicators could Ethanol 0.2020 0.1701 0.2203 not be reported, except for three variables: i) percentages Formate 0.1939 0.1790 0.2192 of viable sperm with high intracellular H2 O2 , which were Fumarate 0.0022 0.0017 0.0032 positively influenced by tyrosine and carnitine, and negatively by Glucose 0.1480 0.0824 0.2476 valine and hypotaurine; ii) percentages of sperm with proximal Glutamate 0.4745 0.3590 0.6526 droplets, which were positively influenced by creatine, formate, Hypotaurine 1.9639 1.5797 2.5621 myo-inositol and tyrosine, and negatively by valine; and iii) Isoleucine 2.7658 2.1357 3.6582 percentages of sperm with folded tails, which were positively Lactate 0.0477 0.0312 0.0616 influenced by carnitine and methanol, and negatively by creatine Leucine 1.8198 1.3941 2.2803 and hypotaurine (Figure 3D). Malonate 0.0633 0.0439 0.0730 Methanol 0.1106 0.0824 0.1635 Relationship Between SP-Metabolites Myo-Inositol 0.0557 0.0211 0.1208 Phenylalanine 4.2850 2.8488 5.3341 and Sperm Resilience to Liquid Storage sn-Glycero-3-phosphocholine 24.3080 19.1166 30.3360 at 17◦ C for 72 h Trimethylamine N-oxide 0.2868 0.1614 0.4411 Sperm quality and functionality were also assessed after 72 h Tyrosine 0.0237 0.0176 0.0302 of liquid storage (Supplementary Table 3). In order to evaluate Valine 0.1039 0.0795 0.1376 whether any of the identified SP-metabolites could contribute Frontiers in Cell and Developmental Biology | www.frontiersin.org 5 May 2021 | Volume 9 | Article 669974 Mateo-Otero et al. SP-Metabolites as Sperm Quality Biomarkers FIGURE 1 | 1H-NMR (noesygppr1d) profile (600MHz) from 0 to 8 ppm of pig seminal plasma. (1) DSS, (2) leucine, (3) isoleucine, (4) valine, (5) ethanol, (6) lactate, (7) alanine, (8) acetate, (9) glutamate, (10) carnitine, (11) citrate, (12) hypotaurine, (13) creatine, (14) creatine phosphate, (15) malonate, (16) sn-glycero-3-phosphocholine, (17) TMAO, (18) myo-inositol, (19) methanol, (20) glucose, (21) fumarate, (22) tyrosine, (23) phenylalanine, and (24) formate. Frontiers in Cell and Developmental Biology | www.frontiersin.org 6 May 2021 | Volume 9 | Article 669974 Mateo-Otero et al. SP-Metabolites as Sperm Quality Biomarkers FIGURE 2 | Concentration of metabolites in seminal plasma from ejaculates (one per boar) of different pig breeds (Pietrain, n = 14 and Duroc, n = 11). Boxes enclose the 25th and 75th percentiles, the whiskers extend to the 5th and 95th percentiles, and the line indicates the median. Each dot represents the metabolite concentration from each ejaculate. Differences between breeds (P < 0.05) are represented as ∗ . to explain the sperm ability to withstand liquid storage at regression analysis was run. Similar to previous findings, most 17◦ C, correlations between concentrations of SP-metabolites and of the models did not show significant contributions of SP- sperm quality and functionality parameters assessed after 72 h metabolites to sperm quality and functionality parameters after at 17◦ C were calculated (Figure 4A). As observed before, most liquid storage at 17◦ C for 72 h (Supplementary Figures 2A- correlations were found to be weak or moderate. The most G). Nevertheless, two parameters, the percentage of viable relevant and significant (P < 0.050) correlations found in terms sperm with high intracellular H2 O2 and the percentage of of higher R scores were: i) the percentage of viable sperm viable sperm with high plasma membrane destabilization, with non-intact acrosome, which was positively correlated with exhibited two groups with different sperm resilience to liquid leucine (R = 0.40, P = 0.038); ii) the percentage of total motile storage (Figure 4B). For this reason, logistic regression models sperm, which was negatively correlated with leucine (R = −0.40, were used for these two sperm parameters (Figure 4C). P = 0.039); and iii) the percentage of progressive motile sperm, Concentrations of glucose and acetate in SP showed an which was negatively correlated with carnitine (R = −0.41, opposite influence depending on the sperm parameter assessed. P = 0.031), hypotaurine (R = −0.45, P = 0.018), isoleucine Thus, whereas acetate and glucose showed a positive and (R = −0.47, P = 0.014) and leucine (R = −0.43, P = 0.024). negative effect on viable sperm with high intracellular Once the correlations between individual sperm parameters H2 O2 , respectively, the opposite pattern was observed in and SP-metabolites were analyzed, a Bayesian multiple linear viable sperm with high membrane destabilization (i.e., Frontiers in Cell and Developmental Biology | www.frontiersin.org 7 May 2021 | Volume 9 | Article 669974 Mateo-Otero et al. SP-Metabolites as Sperm Quality Biomarkers FIGURE 3 | (A) Variance explained by each principal component. (B) Cumulative variance explained by principal components. (C) Correlations between metabolites identified in pig seminal plasma and different sperm functionality parameters analyzed immediately after semen collection. The color saturation of red to blue represents the correlation coefficients (R) between metabolites, from 1 to -1, respectively. Significant correlations (P < 0.05) are marked with *. (D) Bayesian multiple linear regression models for percentage of viable sperm with high intracellular reactive oxygen species, ROS (H2 O2 ), percentages of sperm with proximal droplets and percentages of sperm with folded tails. Thin lines represent the 90% credible intervals, and the blue box represents the 50% credible intervals. Dots represent the median estimated for each metabolite in the model. glucose and acetate exhibiting a positive and a negative were found to be positive except for that observed between effect, respectively). creatine and creatine phosphate (R = −0.38, P < 0.01). The most relevant positive correlations (R > 0.8 and P < 0.01; Correlations Between SP-Metabolites Schober et al., 2018) were: alanine vs. malonate (R = 0.80), Correlations between SP-metabolites were also analyzed carnitine vs. hypotaurine (R = 0.90), carnitine vs. isoleucine and observed to vary with different magnitude (Figure 5). (R = 0.89), citrate vs. sn-glycero-3-phosphocholine (R = 0.88), Interestingly, all significant correlations between SP-metabolites citrate vs. tyrosine (R = 0.82), glutamate vs. trimethylamine Frontiers in Cell and Developmental Biology | www.frontiersin.org 8 May 2021 | Volume 9 | Article 669974 Mateo-Otero et al. SP-Metabolites as Sperm Quality Biomarkers FIGURE 4 | (A) Correlations between metabolites identified in pig seminal plasma and different sperm quality and functionality parameters evaluating the quality loss after 72 h of liquid storage at 17 ◦ C. The color saturation of red to blue represents the correlation coefficients (R) between metabolites, from 1 to −1, respectively. Significant correlations (P < 0.05) are marked with *. (B) Plot showing the relative differences of viable sperm with high membrane destabilization and with high intracellular reactive oxygen species, ROS (H2 O2 ) for each SP sample. Two different sperm resilience to liquid storage patterns can be observed (marked with a discontinuous line). (C) Bayesian logistic regression models for the percentage viable sperm with high membrane destabilization and with high intracellular H2 O2 . Thin lines represent the 90% credible intervals, and the blue box represents the 50% credible intervals. Dots represent the median estimated for each metabolite in the model. N-oxide (R = 0.83), hypotaurine vs. isoleucine (R = 0.92), tyrosine (R = 0.86), lactate vs. valine (R = 0.84), sn-glycero- hypotaurine vs. methanol (R = 0.83), isoleucine vs. methanol 3-phosphocholine vs. tyrosine (R = 0.80), and tyrosine vs. (R = 0.80), lactate vs. malonate (R = 0.92), lactate vs. valine (R = 0.91). Frontiers in Cell and Developmental Biology | www.frontiersin.org 9 May 2021 | Volume 9 | Article 669974 Mateo-Otero et al. SP-Metabolites as Sperm Quality Biomarkers FIGURE 5 | Correlation plot of the twenty-three metabolites identified in boar seminal plasma. The color saturation from red to blue represents the Pearson correlation coefficients (R) between metabolites, from 1 to –1, respectively. Significant correlations (P < 0.05) are marked with *. DISCUSSION uniform detection of equal sensibility for all proton-containing molecules in samples, and requires little or no processing As far as we know, this is the first report that performed (Goodacre et al., 2004; Emwas et al., 2019). This analytical an in-depth metabolomic profile of pig SP quantifying the method has been widely used for SP-metabolite profiling of identified metabolites. Moreover, this is the first study conducted several mammalian species, including humans (Hamamah et al., in a livestock species that explored the putative influence 1993; Segalen et al., 1995; Gupta et al., 2011; Jayaraman et al., of SP-metabolites on sperm physiology and on the sperm 2014; Bonechi et al., 2015; Mumcu et al., 2020), cattle (Kumar resilience to liquid storage at 17 ◦ C. Specifically, this exploratory et al., 2015) and pigs (Kalic et al., 1997; Mateo-Otero et al., study reported that: i) a total of 23 SP-metabolites were 2020). The present work identified and quantified a total of present at measurable concentrations in pig SP; ii) except for 23 metabolites in pig SP, including amino acids, saccharides, fumarate, no differences in SP-metabolites concentration were salts, alcohols, and other compounds. Compared to our previous identified between pig breeds (Duroc and Pietrain); iii) glutamate, work (Mateo-Otero et al., 2020), the present study quantified SP- methanol, trimethylamine N-oxide, carnitine and isoleucine metabolites, as well as identified new molecules (i.e., formate, concentrations in SP were related to specific sperm quality and fumarate, hypotaurine, isoleucine, malonate, myo-inositol, and functionality parameters evaluated right after semen collection; trimethylamine N-oxide). These new findings may result from iv) leucine, hypotaurine, carnitine and isoleucine concentrations improvements in sample handling and purification, and the in SP were found to be related with the sperm ability to withstand higher frequency and resolution of the NMR spectrometer. liquid storage; and v) Bayesian multiple regression models Additionally, while SP in pigs showed a metabolite profile similar allowed the identification of putative metabolite patterns for to that reported in humans (Hamamah et al., 1993; Segalen specific sperm parameters. In addition, very strong relationships et al., 1995; Gupta et al., 2011; Jayaraman et al., 2014; Bonechi between SP-metabolites were also observed. et al., 2015; Mumcu et al., 2020), it differed from that found Metabolomic approaches allow the study of cellular metabolic in cattle (Kumar et al., 2015). This was already reported in our products; these low-molecular-weight molecules can ultimately previous work (Mateo-Otero et al., 2020) and was confirmed in unveil the phenotype of the studied system (Deepinder et al., the present study. 2007). For this reason, studying the metabolome has been Regarding the new SP-metabolites identified in the present widely proposed as a promising tool for spermatology and work, both isoleucine and malonate were also identified in the male infertility diagnosis (Aitken, 2010; Kovac et al., 2013; human SP (Gupta et al., 2011; Bonechi et al., 2015; Mumcu Minai-Tehrani et al., 2016; Bracewell-Milnes et al., 2017; Moura et al., 2020). Moreover, only some of the NMR studies carried et al., 2018). The NMR spectroscopy is one of the most used out in humans quantified the identified metabolites (Gupta et al., technologies in this field, because it provides a non-biased 2011, 2013; Bonechi et al., 2015), generally reporting higher Frontiers in Cell and Developmental Biology | www.frontiersin.org 10 May 2021 | Volume 9 | Article 669974 Mateo-Otero et al. SP-Metabolites as Sperm Quality Biomarkers SP-metabolite concentrations than those found herein for pigs membrane destabilization was detected. These results would (Hamamah et al., 1993; Segalen et al., 1995; Gupta et al., 2011; be in agreement with the study by Darbandi et al. (2019) in Jayaraman et al., 2014; Bonechi et al., 2015; Mumcu et al., 2020). men, who found that higher levels of this metabolite in human Variations in SP metabolite profile or in their concentrations SP were related with high intracellular ROS levels in sperm could contribute to reveal physiological differences between and could ultimately cause membrane lipid peroxidation and, species in terms of sperm functionality and fertility. Nevertheless, thus, membrane destabilization. Finally, our data indicated that sample handling and processing could also serve to explain these concentrations of both carnitine (P = 0.043) and isoleucine qualitative and quantitative discrepancies. (P = 0.041) in SP were negatively correlated to progressively As aforementioned, one of the most relevant results of motile sperm. Regarding carnitine, our results are in agreement the present study was that concentrations of 23 metabolites with previous studies in humans using liquid chromatography- present in pig SP were uncovered. In addition, we observed mass spectrometry (LC-MS) (Wang et al., 2019; Xu et al., that concentrations of SP-metabolites varied between ejaculates. 2020). This metabolite was found to be a potential SP- Since all boars had similar age, were housed under the same biomarker of poor semen quality (Wang et al., 2019), and conditions, and were fed with the same commercial diet, we to be positively related with sperm deformity rate (Xu et al., hypothesize that such variability could be related to pig breeds. 2020). By contrast, our findings do not agree with the previous However, when samples were divided considering breeds, no literature, since several studies reported the beneficial effect effect on SP-metabolites concentration was observed, except of supplementing diet with L-carnitine on sperm progressive for fumarate, which exhibited higher levels in Pietrain than in motility in patients with different types of infertility (Lenzi Duroc boars. This agrees with most of the studies evaluating SP et al., 2004; Garolla et al., 2005; Micic et al., 2019). On the components in samples from different pig breeds, in which no other hand, a negative effect of isoleucine on sperm has been breed effect was reported (Barranco et al., 2015b, 2016, 2020; observed in buffalos, since poor semen quality is associated Llavanera et al., 2020). However, since the sample size for each to low levels of isoleucine (Binsila et al., 2020). Our data breed was limited in the present study, further research should indicated that each relationship has a different P value; while verify this assumption. Nevertheless, considering our findings, glutamate and sperm viability were found to have a very one could assume that differences in the concentration of SP- robust correlation coefficient (P = 0.014), the others ranged metabolites found between boars could contribute to explain the between P = 0.020 and P = 0.045. In this latter case, and disparity observed in their sperm quality parameters assessed despite the fact that these correlations had a high degree of immediately after semen collection or in their sperm resilience confidence (> 95%), one could not exclude that some happened to liquid storage. by chance (given that type I error increases with the number Many efforts have been focused on exploring pig SP of correlations tested). For this reason, except for glutamate, composition with the aim to uncover SP biomarkers able to the other relationships should be considered as a preliminary predict semen quality after ejaculate collection (López Rodríguez approximation to elucidate the effect of these SP-metabolites et al., 2012; Barranco et al., 2016, 2017, 2020; Llavanera et al., on sperm physiology. Thus, further experiments to validate 2020). In the last years, the advances in metabolomic analytical the correlations observed in the present study are warranted. methods have allowed the identification of SP-metabolites as What is more, considering the aforementioned discrepancies potential sperm quality biomarkers in mammalian species, between species, the actual effect of these metabolites on pig such as humans (Bonechi et al., 2015; Engel et al., 2019; sperm needs to be addressed in future studies because species- Wang et al., 2019; Mumcu et al., 2020; Xu et al., 2020) specific regulation of SP-metabolites could be observed. Another and cattle (Moura et al., 2018). Our results suggest that possibility to explain such discrepancies could be that only metabolomic analysis of pig SP could offer a new source moderate correlations were found between pig SP-metabolites of sperm functionality biomarkers. Therefore, by using an and sperm quality parameters. Hence, specificity and sensibility exploratory approach, the present study evaluated the potential of SP-metabolites should be tested and validated before their relationship of the concentration of SP-metabolites with sperm use as biomarkers. quality and functionality parameters assessed immediately after This study was also able to explore whether concentrations semen collection. While a strongly significant (P = 0.014) of metabolites in SP were able to explain variations in positive correlation between the concentration of glutamate in sperm quality/functionality parameters using Bayesian multiple SP and sperm viability was observed, this does not fully agree linear regression models. Only metabolites whose 90% credible with previous findings in which the addition of monosodium intervals did not cross 0 in a Pareto distribution were considered glutamate to fresh pig semen had no effect on the percentage as valid and, thus, as promising biomarkers. We did not find of viable sperm during in vitro capacitation (Spinaci et al., clear patterns for most of the quality parameters evaluated 2017). On the other hand, a positive relationship (P = 0.027) immediately after semen collection. However, models based on between methanol concentration in SP and viable sperm with viable sperm with high intracellular ROS, sperm with proximal high H2 O2 was found. These results do not come as a surprise droplets and sperm with folded tails showed a positive or when one considers the well-known oxidative stress-inducer negative effect of specific SP-metabolites. Most relationships role of this metabolite (Parthasarathy et al., 2006). In addition, between metabolites have not yet been described in the literature; a negative relationship (P = 0.045) between trimethylamine thus, further studies focused on determining whether these N-oxide concentration in SP and viable sperm with high relationships have a biological significance are required. Frontiers in Cell and Developmental Biology | www.frontiersin.org 11 May 2021 | Volume 9 | Article 669974 Mateo-Otero et al. SP-Metabolites as Sperm Quality Biomarkers Sperm quality of pig seminal AI-doses gradually declines agreement with the literature, as acetate production from glucose during their storage at 17◦ C (Waberski et al., 2011; Yeste, through pyruvate has been previously reported to occur in 2017), showing differences in their sperm resilience to liquid mammalian mitochondria (Liu et al., 2018). Our models showed preservation (Parrilla et al., 2012; Zakošek Pipan et al., 2014; that low levels of glucose and high levels of acetate have an impact Roca et al., 2015). While several studies have demonstrated that on the percentage of viable sperm with high levels of ROS, which some SP-proteins are able to predict pig sperm resilience to liquid could be explained by the link between acetate production and storage (Zakošek Pipan et al., 2014; Barranco et al., 2016, 2019; ROS generation (Liu et al., 2018). On the other hand, glucose Parrilla et al., 2019; Llavanera et al., 2020), no study has, to could also be involved in fatty acids and cholesterol synthesis the best of our knowledge, evaluated the potential relationship (Wolfe, 1998). As we found that high levels of glucose and low between SP-metabolite composition and the sperm ability to levels of acetate influenced the percentage of viable sperm with sustain 72 h of liquid storage. Our approach suggested that some high membrane destabilization, further studies should explore SP-metabolites were positively and negatively related with the the exact mechanism that would underlie such observation. sperm functionality decline during liquid preservation. Leucine In conclusion, the use of NMR spectroscopy analysis allowed concentration in SP was found to be positively correlated with the identification and quantification of 23 metabolites in pig the percentage of viable sperm with a non-intact acrosome SP. Our exploratory approach was able to suggest that the (P = 0.038) and negatively associated with the loss of total concentration of selected SP-metabolites was related to sperm (P = 0.039) and progressive (P = 0.024) motility after 72 h of quality and functionality parameters of liquid-stored semen liquid storage. Despite the fact that no information regarding samples and to the sperm ability to withstand liquid storage, leucine effect on sperm performance is available in mammals, which led to uncover potential biomarkers. Specifically, the the addition of this metabolite is known to have a positive concentration of glutamate in SP appeared to be related to sperm effect on the motility of zebrafish sperm (Zhang et al., 2017). viability when evaluated immediately after ejaculate collection. Similarly, our findings revealed that hypotaurine concentration Moreover, while concentrations of hypotaurine and isoleucine in in SP was negatively related with higher loss of progressive SP could be involved in the sperm resilience to liquid storage, the motility (P = 0.018) after liquid storage at 17 ◦ C. In this context, pattern observed for glucose and acetate suggests that they could it is worth mentioning that hypotaurine supplementation has also play a role. Yet, these relationships can only be considered as been reported to exert a positive impact upon sperm motility a first steppingstone for metabolite biomarker identification, as in mammalian species, such as hamsters (Boatman et al., 1990) they are pending to be validated in future studies. Nevertheless, or humans (Holmes et al., 1992; Brugnon et al., 2013). Finally, this could have a very relevant impact on AI industry because the both carnitine (P = 0.031) and isoleucine (P = 0.014) were found identification of such metabolites could contribute to improve: to be negatively related with the loss of progressive motility i) quality assessment of seminal AI-doses prior to AI; and ii) AI after liquid storage. While information regarding the effect of extenders through supplementation with specific metabolites to isoleucine on sperm motility is yet to be reported, a study improve the sperm resilience to liquid storage. Moreover, further on L-carnitine supplementation during liquid storage of pig research including more semen samples should be conducted to semen demonstrated that this metabolite could maintain sperm determine whether these SP-metabolites could also be used as motility (Yang et al., 2020). Whilst all these data suggest that in vivo fertility biomarkers. these metabolites may be promising biomarkers for motility prediction of seminal AI-doses after liquid storage, the different P values may, as pointed out before, imply different robustness DATA AVAILABILITY STATEMENT in the relationships observed. In this regard, while correlations of isoleucine and hypotaurine with loss of progressive motility over The raw data supporting the conclusions of this article will be liquid storage were high, those with P values oscillating between made available by the authors, without undue reservation. 0.020 and 0.040 would need further validation. For this reason, although our study can be considered as a first steppingstone in establishing biomarkers for the sperm resilience to liquid storage, ETHICS STATEMENT further experiments are needed to confirm the exact role of SP-metabolites during that process. Ethical review and approval for animal participants was not In a similar fashion to the case of sperm quality evaluated required as samples were provided by the AI-Centre and authors immediately after semen collection, Bayesian multiple linear did not manipulate any animals. regression models did not identify clear metabolite concentration patterns explaining sperm quality loss. However, the percentage of viable sperm with high intracellular H2 O2 and the percentage AUTHOR CONTRIBUTIONS of viable sperm with high plasma membrane destabilization showed two distinct groups of low- and high-quality loss. For YM-O, MY, and IB: conceptualization. YM-O, PF-L, and this reason, logistic regression models were considered to be IB: methodology. YM-O, PF-L, JR-M, MY, and IB: formal appropriate to further study such relationships. Interestingly, analysis and investigation. YM-O: writing – original concentrations of acetate and glucose in SP were observed to draft preparation. JR-M, JM, JR, MY, and IB: writing – have opposite effects on both parameters. These findings are in review and editing. JM, JR, and MY: funding acquisition. Frontiers in Cell and Developmental Biology | www.frontiersin.org 12 May 2021 | Volume 9 | Article 669974 Mateo-Otero et al. SP-Metabolites as Sperm Quality Biomarkers IB and MY: supervision. All authors have read and agreed to the SUPPLEMENTARY MATERIAL published version of the manuscript. The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fcell.2021. FUNDING 669974/full#supplementary-material The present study was funded by the Ministry of Science and Supplementary Figure 1 | (A-M). Complete Bayesian multiple linear Innovation, Spain (Grants: RYC-2014-15581, AGL2017-88329- regression models for all the parameters evaluated immediately after semen collection. R, and FJCI-2017-31689), the Seneca Foundation Murcia, Spain (19892/GERM-15), and the Regional Government of Catalonia Supplementary Figure 2 | (A-F) Complete Bayesian multiple linear regression (Grants: 2017-SGR-1229 and 2020-FI-B-00412). models for all the parameters evaluated immediately after semen collection. (G,H) Complete Bayesian logistic regression models for percentages of viable sperm with high membrane destabilization, and with high intracellular Ros. ACKNOWLEDGMENTS Supplementary Table 1 | Metabolite concentrations (mM) in seminal plasma of Pietrain and Duroc breeds. Significant correlations (P < 0.05) are marked with ∗ . The authors would like to thank Pau Nolis, from the Supplementary Table 2 | Sperm quality and functionality parameters of pig Nuclear Magnetic Resonance Service, Autonomous University of semen samples (n = 28) assessed immediately after semen collection. Sd: Barcelona, for the metabolomic analysis and technical support. Standard Deviation. Moreover, the authors would also wish to thank AIM Iberica Supplementary Table 3 | Sperm quality loss of pig semen samples (n = 28) after (Topigs Norsvin Iberica) for supplying the boar ejaculates. 72 h of liquid storage at 17◦ C. Sd: Standard Deviation. REFERENCES and statistical analysis. Syst. Biol. Reprod. Med. 61, 353–359. doi: 10.3109/ 19396368.2015.1054003 Aitken, R. J. (2010). Whither must spermatozoa wander? The future of laboratory Bracewell-Milnes, T., Saso, S., Abdalla, H., Nikolau, D., Norman-Taylor, J., seminology. Asian J. Androl. 12, 99–103. doi: 10.1038/aja.2008.42 Johnson, M., et al. (2017). Metabolomics as a tool to identify biomarkers to Asampille, G., Cheredath, A., Joseph, D., Adiga, S. K., and Atreya, H. S. (2020). The predict and improve outcomes in reproductive medicine: a systematic review. utility of nuclear magnetic resonance spectroscopy in assisted reproduction. Hum. Reprod. Update 23, 723–736. doi: 10.1093/humupd/dmx023 Open Biol. 10:200092. doi: 10.1098/rsob.200092 Brugnon, F., Ouchchane, L., Pons-Rejraji, H., Artonne, C., Farigoule, M., and Barranco, I., Fernandez-Fuertes, B., Padilla, L., Delgado-Bermúdez, A., Janny, L. (2013). Density gradient centrifugation prior to cryopreservation Tvarijonaviciute, A., and Yeste, M. (2020). Seminal plasma anti-müllerian and hypotaurine supplementation improve post-thaw quality of sperm from hormone: a potential Ai-boar fertility biomarker? Biology 9, 1–13. infertile men with oligoasthenoteratozoospermia. Hum. Reprod. 28, 2045–2057. doi: 10.3390/biology9040078 doi: 10.1093/humrep/det253 Barranco, I., Padilla, L., Tvarijonaviciute, A., Parrilla, I., Martínez, E. A., Rodriguez- Cazaux Velho, A. L., Menezes, E., Dinh, T., Kaya, A., Topper, E., Moura, A. A., Martinez, H., et al. (2019). Levels of activity of superoxide dismutase in seminal et al. (2018). Metabolomic markers of fertility in bull seminal plasma. PLoS One plasma do not predict fertility of pig AI-semen doses. Theriogenology 140, 13:e0195279. doi: 10.1371/journal.pone.0195279 18–24. doi: 10.1016/j.theriogenology.2019.08.010 Darbandi, M., Darbandi, S., Agarwal, A., Baskaran, S., Dutta, S., Sengupta, P., et al. Barranco, I., Perez-Patiño, C., Tvarijonaviciute, A., Parrilla, I., Vicente-Carrillo, (2019). Reactive oxygen species-induced alterations in H19-Igf2 methylation A., Alvarez-Rodriguez, M., et al. (2017). Active paraoxonase 1 is synthesised patterns, seminal plasma metabolites, and semen quality. J. Assist. Reprod. throughout the internal boar genital organs. Reproduction 154, 237–243. doi: Genet. 36, 241–253. doi: 10.1007/s10815-018-1350-y 10.1530/REP-17-0300 Deepinder, F., Chowdary, H. T., and Agarwal, A. (2007). Role of metabolomic Barranco, I., Tvarijonaviciute, A., Perez-Patino, C., Alkmin, D. V., Ceron, J. J., analysis of biomarkers in the management of male infertility. Expert Rev. Mol. Martinez, E. A., et al. (2015a). The activity of paraoxonase type 1 (PON-1) in Diagn. 7, 351–358. doi: 10.1586/14737159.7.4.351 boar seminal plasma and its relationship with sperm quality, functionality, and Egea, R. R., Puchalt, N. G., Escrivá, M. M., and Varghese, A. C. (2014). OMICS: in vivo fertility. Andrology 3, 315–320. doi: 10.1111/andr.309 CURRENT and future perspectives in reproductive medicine and technology. Barranco, I., Tvarijonaviciute, A., Perez-Patinõ, C., Parrilla, I., Ceron, J. J., J. Hum. Reprod. Sci. 7, 73–92. doi: 10.4103/0974-1208.138857 Martinez, E. A., et al. (2015b). High total antioxidant capacity of the porcine Emwas, A. H., Roy, R., McKay, R. T., Tenori, L., Saccenti, E., Nagana Gowda, G. A., seminal plasma (SP-TAC) relates to sperm survival and fertility. Sci. Rep. 5, 1–9. et al. (2019). NMR spectroscopy for metabolomics research. Metabolites 9:123. doi: 10.1038/srep18538 doi: 10.3390/metabo9070123 Barranco, S., Tvarijonaviciute, A., Perez-Patiho, C., Vicente-Carrillo, A., Parrilla, Engel, K. M., Baumann, S., Rolle-Kampczyk, U., Schiller, J., von Bergen, N., Ceron, J. J., et al. (2016). Glutathione peroxidase 5 is expressed by the entire M., and Grunewald, S. (2019). Metabolomic profiling reveals correlations pig male genital tract and once in the seminal plasma contributes to sperm between spermiogram parameters and the metabolites present in human survival and in vivo fertility. PLoS One 11:e0162958. doi: 10.1371/journal.pone. spermatozoa and seminal plasma. PLoS One 14:e0211679. doi: 10.1371/journal. 0162958 pone.0211679 Binsila, B. K., Archana, S. S., Ramya, L., Swathi, D., Selvaraju, S., Gowda, N. S., Garolla, A., Maiorino, M., Roverato, A., Roveri, A., Ursini, F., and Foresta, et al. (2020). Elucidating the processes and pathways enriched in buffalo sperm C. (2005). Oral carnitine supplementation increases sperm motility in proteome in regulating semen quality. Cell Tissue Res. 383, 881–903. doi: 10. asthenozoospermic men with normal sperm phospholipid hydroperoxide 1007/s00441-020-03303-9 glutathione peroxidase levels. Fertil. Steril. 83, 355–361. doi: 10.1016/j. Boatman, D. E., Bavister, B. D., and Cruz, E. (1990). Addition of hypotaurine fertnstert.2004.10.010 can reactivate immotile golden hamster spermatozoa. J. Androl. 11, 66–72. Goodacre, R., Vaidyanathan, S., Dunn, W. B., Harrigan, G. G., and Kell, doi: 10.1002/j.1939-4640.1990.tb01581.x D. B. (2004). Metabolomics by numbers: acquiring and understanding global Bonechi, C., Collodel, G., Donati, A., Martini, S., Moretti, E., and Rossi, C. (2015). metabolite data. Trends Biotechnol. 22, 245–252. doi: 10.1016/j.tibtech.2004.03. Discrimination of human semen specimens by NMR data, sperm parameters, 007 Frontiers in Cell and Developmental Biology | www.frontiersin.org 13 May 2021 | Volume 9 | Article 669974 Mateo-Otero et al. SP-Metabolites as Sperm Quality Biomarkers Goodrich, B., Gabry, J., Ali, I., and Brilleman, S. (2020). rstanarm: Bayesian Applied Mumcu, A., Karaer, A., Dogan, B., and Tuncay, G. (2020). Metabolomics analysis of Regression Modeling via Stan. Available online at: https://mc-stan.org/rstanarm/ seminal plasma in patients with idiopathic Oligoasthenoteratozoospermia using (accessed November 2, 2020). high-resolution NMR spectroscopy. Andrology 8, 450–456. doi: 10.1111/andr. Gupta, A., Mahdi, A. A., Ahmad, M. K., Shukla, K. K., Jaiswer, S. P., and Shankhwar, 12707 S. N. (2011). 1H NMR spectroscopic studies on human seminal plasma: a Nobakht, M., and Gh, B. F. (2018). Application of metabolomics to preeclampsia probative discriminant function analysis classification model. J. Pharm. Biomed. diagnosis. Syst. Biol. Reprod. Med. 64, 324–339. doi: 10.1080/19396368.2018. Anal. 54, 106–113. doi: 10.1016/j.jpba.2010.07.021 1482968 Gupta, A., Mahdi, A. A., Shukla, K. K., Ahmad, M. K., Bansal, N., Sankhwar, P., Parrilla, I., del Olmo, D., Sijses, L., Martinez-Alborcia, M. J., Cuello, C., Vazquez, et al. (2013). Efficacy of Withania somnifera on seminal plasma metabolites J. M., et al. (2012). Differences in the ability of spermatozoa from individual of infertile males: a proton NMR study at 800 MHz. J. Ethnopharmacol. 149, boar ejaculates to withstand different semen-processing techniques. Anim. 208–214. doi: 10.1016/j.jep.2013.06.024 Reprod. Sci. 132, 66–73. doi: 10.1016/j.anireprosci.2012.04.003 Hamamah, S., Seguin, F., Barthelemy, C., Akoka, S., Le Pape, A., Lansac, J., et al. Parrilla, I., Perez-Patiño, C., Li, J., Barranco, I., Padilla, L., Rodriguez-Martinez, H., (1993). 1H nuclear magnetic resonance studies of seminal plasma from fertile et al. (2019). Boar semen proteomics and sperm preservation. Theriogenology and infertile men. J. Reprod. Fertil. 97, 51–55. doi: 10.1530/jrf.0.0970051 137, 23–29. doi: 10.1016/j.theriogenology.2019.05.033 Holmes, R. P., Goodman, H. O., Shihabi, Z. K., and Jarow, J. P. (1992). The Parthasarathy, N. J., Kumar, R. S., Manikandan, S., Narayanan, G. S., Kumar, R. V., taurine and hypotaurine content of human semen. J. Androl. 13, 289–292. and Devi, R. S. (2006). Effect of methanol-induced oxidative stress on the doi: 10.1002/j.1939-4640.1992.tb00317.x neuroimmune system of experimental rats. Chem. Biol. Interact. 161, 14–25. Jayaraman, V., Ghosh, S., Sengupta, A., Srivastava, S., Sonawat, H. M., and Narayan, doi: 10.1016/j.cbi.2006.02.005 P. K. (2014). Identification of biochemical differences between different forms Recuero, S., Fernandez-Fuertes, B., Bonet, S., Barranco, I., and Yeste, M. (2019). of male infertility by nuclear magnetic resonance (NMR) spectroscopy. J. Assist. Potential of seminal plasma to improve the fertility of frozen-thawed boar Reprod. Genet. 31, 1195–1204. doi: 10.1007/s10815-014-0282-4 spermatozoa. Theriogenology 137, 36–42. doi: 10.1016/j.theriogenology.2019. Kalic, M., Kamp, G., and Lauterwein, J. (1997). Nuclear magnetic resonance studies 05.035 of boar seminal plasma. Problems encountered in the identification of small Roca, J., Broekhuijse, M. L. W. J., Parrilla, I., Rodriguez-Martinez, H., Martinez, molecules: hypotaurine and carnitine. NMR Biomed. 10, 341–347. doi: 10.1002/ E. A., and Bolarin, A. (2015). Boar differences in artificial insemination (SICI)1099-1492(199710)10:7<341::AID-NBM464<3.0.CO;2-M outcomes: can they be minimized? Reprod. Domest. Anim. 50, 48–55. doi: Kovac, J. R., Pastuszak, A. W., and Lamb, D. J. (2013). The use of genomics, 10.1111/rda.12530 proteomics, and metabolomics in identifying biomarkers of male infertility. Roca, J., Parrilla, I., Bolarin, A., Martinez, E. A., and Rodriguez-Martinez, H. Fertil. Steril. 99, 998–1007. doi: 10.1016/j.fertnstert.2013.01.111 (2016). Will AI in pigs become more efficient? Theriogenology 86, 187–193. Kumar, A., Kroetsch, T., Blondin, P., and Anzar, M. (2015). Fertility-associated doi: 10.1016/j.theriogenology.2015.11.026 metabolites in bull seminal plasma and blood serum: 1H nuclear magnetic Rodríguez-Martínez, H., Kvist, U., Ernerudh, J., Sanz, L., and Calvete, J. J. (2011). resonance analysis. Mol. Reprod. Dev. 82, 123–131. doi: 10.1002/mrd.22450 Seminal plasma proteins: what role do they play? Am. J. Reprod. Immunol. 66, Kumaresan, A., Kadirvel, G., Bujarbaruah, K. M., Bardoloi, R. K., Das, A., Kumar, 11–22. doi: 10.1111/j.1600-0897.2011.01033.x S., et al. (2009). Preservation of boar semen at 18 degrees C induces lipid Schober, P., Boer, C., and Schwarte, L. A. (2018). Correlation coefficients: peroxidation and apoptosis like changes in spermatozoa. Anim. Reprod. Sci. appropriate use and interpretation. Anesth. Analg. 126, 1763–1768. doi: 10. 110, 162–171. doi: 10.1016/j.anireprosci.2008.01.006 1213/ANE.0000000000002864 Lenzi, A., Sgrò, P., Salacone, P., Paoli, D., Gilio, B., Lombardo, F., et al. (2004). Segalen, J., de Certaines, J. D., Le Calvé, M., Colleu, D., Bansard, J. Y., and Rio, A placebo-controlled double-blind randomized trial of the use of combined M. (1995). Fertilization attempts: use of automatic spectrum analysis. J. Reprod. L-carnitine and L-acetyl-carnitine treatment in men with asthenozoospermia. Fertil. 103, 181–187. Fertil. Steril. 81, 1578–1584. doi: 10.1016/j.fertnstert.2003.10.034 Singh, R., and Sinclair, K. D. (2007). Metabolomics: approaches to assessing Liu, X., Cooper, D. E., Cluntun, A. A., Warmoes, M. O., Zhao, S., Reid, M. A., oocyte and embryo quality. Theriogenology 68(Suppl. 1), S56–S62. doi: 10.1016/ et al. (2018). Acetate production from glucose and coupling to mitochondrial j.theriogenology.2007.04.007 metabolism in mammals. Cell 175, 502.e13–513.e13. doi: 10.1016/j.cell.2018.08. Spinaci, M., Bucci, D., Gadani, B., Porcu, E., Tamanini, C., and Galeati, G. (2017). 040 Pig sperm preincubation and gamete coincubation with glutamate enhance Llavanera, M., Delgado-Bermúdez, A., Mateo-Otero, Y., Padilla, L., Romeu, X., sperm-oocyte binding and in vitro fertilization. Theriogenology 95, 149–153. Roca, J., et al. (2020). Exploring seminal plasma GSTM3 as a quality and in vivo doi: 10.1016/j.theriogenology.2017.03.017 fertility biomarker in pigs—relationship with sperm morphology. Antioxidants Uyar, A., and Seli, E. (2014). Metabolomic assessment of embryo viability. Semin. 9, 1–14. doi: 10.3390/antiox9080741 Reprod. Med. 32, 141–152. doi: 10.1055/s-0033-1363556 López Rodríguez, A., Rijsselaere, T., Beek, J., Vyt, P., Soom, A., Van, et al. Waberski, D., Henning, H., and Petrunkina, A. M. (2011). Assessment of storage (2012). Boar seminal plasma components and their relation with semen effects in liquid preserved boar semen. Reprod. Domest. Anim. 46, 45–48. doi: quality. Syst. Biol. Reprod. Med. 59, 5–12. doi: 10.3109/19396368.2012.72 10.1111/j.1439-0531.2011.01836.x 5120 Waberski, D., Riesenbeck, A., Schulze, M., Weitze, K. F., and Johnson, L. (2019). Mateo-Otero, Y., Fernández-López, P., Gil-Caballero, S., Fernandez-Fuertes, B., Application of preserved boar semen for artificial insemination: past, present Bonet, S., Barranco, I., et al. (2020). 1H nuclear magnetic resonance of pig and future challenges. Theriogenology 137, 2–7. doi: 10.1016/j.theriogenology. seminal plasma reveals intra-ejaculate variation in metabolites. Biomolecules 10, 2019.05.030 1–16. doi: 10.3390/biom10060906 Wang, Y. X., Wu, Y., Chen, H. G., Duan, P., Wang, L., Shen, H. Q., et al. Micic, S., Lalic, N., Djordjevic, D., Bojanic, N., Bogavac-Stanojevic, N., Busetto, (2019). Seminal plasma metabolome in relation to semen quality and urinary G. M., et al. (2019). Double-blind, randomised, placebo-controlled trial on the phthalate metabolites among Chinese adult men. Environ. Int. 129, 354–363. effect of L-carnitine and L-acetylcarnitine on sperm parameters in men with doi: 10.1016/j.envint.2019.05.043 idiopathic oligoasthenozoospermia. Andrologia 51:e13267. doi: 10.1111/and. Wolfe, R. R. (1998). Metabolic interactions between glucose and fatty acids in 13267 humans. Am. J. Clin. Nutr. 67, 519S–526S. doi: 10.1093/ajcn/67.3.519S Minai-Tehrani, A., Jafarzadeh, N., and Gilany, K. (2016). Metabolomics: a state-of- Xu, Y., Lu, H., Wang, Y., Zhang, Z., and Wu, Q. (2020). Comprehensive metabolic the-art technology for better understanding of male infertility. Andrologia 48, profiles of seminal plasma with different forms of male infertility and their 609–616. doi: 10.1111/and.12496 correlation with sperm parameters. J. Pharm. Biomed. Anal. 177:112888. doi: Moura, A. A., Memili, E., Portela, A. M. R., Viana, A. G., Velho, A. L. C., Bezerra, 10.1016/j.jpba.2019.112888 M. J. B., et al. (2018). Seminal plasma proteins and metabolites: effects on Yang, K., Wang, N., Guo, H. T., Wang, J. R., Sun, H. H., Sun, L. Z., et al. (2020). sperm function and potential as fertility markers. Anim. Reprod. 15, 691–702. Effect of L-carnitine on sperm quality during liquid storage of boar semen. doi: 10.21451/1984-3143-AR2018-0029 Asian Austral. J. Anim. Sci. 33, 1763–1769. doi: 10.5713/ajas.19.0455 Frontiers in Cell and Developmental Biology | www.frontiersin.org 14 May 2021 | Volume 9 | Article 669974 Mateo-Otero et al. SP-Metabolites as Sperm Quality Biomarkers Yeste, M. (2017). State-of-the-art of boar sperm preservation in liquid Conflict of Interest: The authors declare that the research was conducted in the and frozen state. Anim. Reprod. 14, 69–81. doi: 10.21451/1984-3143- absence of any commercial or financial relationships that could be construed as a AR895 potential conflict of interest. Zakošek Pipan, M., Mrkun, J., Kosec, M., Nemec Svete, A., and Zrimšek, P. (2014). Superoxide dismutase: a predicting factor for boar semen characteristics Copyright © 2021 Mateo-Otero, Fernández-López, Ribas-Maynou, Roca, Miró, for short-term preservation. Biomed Res. Int. 2014:105280. doi: 10.1155/2014/ Yeste and Barranco. This is an open-access article distributed under the terms 105280 of the Creative Commons Attribution License (CC BY). The use, distribution or Zhang, J., Zhang, X., Liu, Y., Su, Z., Ullah Dawar, F., Dan, H., et al. (2017). Leucine reproduction in other forums is permitted, provided the original author(s) and the mediates autophagosome-lysosome fusion and improves sperm motility by copyright owner(s) are credited and that the original publication in this journal activating the PI3K/Akt pathway. Oncotarget 8, 111807–111818. doi: 10.18632/ is cited, in accordance with accepted academic practice. No use, distribution or oncotarget.22910 reproduction is permitted which does not comply with these terms. Frontiers in Cell and Developmental Biology | www.frontiersin.org 15 May 2021 | Volume 9 | Article 669974