Sperm DNA 5-methyl cytosine and RNA N⁶-methyladenosine methylation are differently affected during periods of body weight losses and body weight gain of young and mature breeding bulls

Abstract

Aiming to characterize the effects of nutritional status on epigenetic markers, such as DNA 5-methyl cytosine (mC) methylation and RNA N6-methyladenosine (m⁶A) methylation, of bovine sperm, 12 Angus × Hereford crossbred breeding bulls were submitted to nutritional changes for a period of 180 d: no change in body weight (BW) (phase 1 = 12 d), BW loss (phase 2 = 78 d), and BW gain (phase 3 = 90 d) in a repeated measures design. Animals were fed Beardless wheat (Triticum aestivum) hay and mineral mix. Statistical analyses were performed using SAS 9.4 (SAS Inst., Cary, NC). Higher levels of RNA m⁶A (P = 0.004) and DNA methylation (P = 0.007) of spermatic cells were observed at phase 2 compared with phase 1. In phase 3, sperm RNA m⁶A methylation levels continued to be higher (P = 0.004), whereas the DNA of sperm cells was similar (P = 0.426) compared with phase 1. Growing bulls had a tendency (P = 0.109) of higher RNA m⁶A methylation levels than mature bulls. Phase 2 altered scrotal circumference (P < 0.001), sperm volume (P = 0.007), sperm total motility (P = 0.004), sperm progressive motility (P = 0.004), total sperm count (P = 0.049), normal sperm (P < 0.001), abnormal sperm (P < 0.001), primary sperm defects (P = 0.039), and secondary sperm defects (P < 0.001). In phase 3, bulls had scrotal circumference, sperm volume, sperm motility, sperm progressive motility, total sperm count, normal and abnormal spermatozoa, and primary and secondary spermatozoa defects similar to phase 1 (P > 0.05). Serum concentrations of insulin-like growth factor-1 and leptin decreased during phase 2 (P = 0.010), while no differences (P > 0.05) were detected between phases 3 and 1; growing bulls tended (P = 0.102) to present higher leptin levels than mature bulls. Specific for mature bulls, DNA methylation was positively correlated with leptin concentration (0.569, P = 0.021), whereas for young bulls, DNA methylation was positively correlated with abnormal spermatozoa (0.824, P = 0.006), primary spermatozoa defect (0.711, P = 0.032), and secondary spermatozoa defect (0.661, P = 0.052) and negatively correlated with normal spermatozoa (−0.824, P = 0.006), total sperm count (−0.702, P = 0.035), and sperm concentration (−0.846, P = 0.004). There was no significant correlation (P > 0.05) between RNA m⁶A and hormones and semen traits. In conclusion, the nutritional status of breeding bulls alters epigenetic markers, such as DNA methylation and RNA m⁶A methylation, in sperm, and the impact of change seems to be age dependent. These markers may serve as biomarkers of sperm quality and fertility of bulls in the future. Detrimental effects on sperm production and seminal quality are observed at periods and places when and where environmental and nutritional limitations are a year-round reality and may carry hidden players that may influence a lifetime of underperformance.

Introduction

Bull fertility is critically important for cattle production systems, impacting the reproductive performance of herds through both natural breeding and artificial insemination (AI). In fact, as one bull has the potential to breed thousands of females by AI, the impact of bull fertility on the cattle industry is tremendous (Thundathil et al., 2016). Still, bulls’ subfertility rates are estimated to reach up to 25% in the United States (Kennedy et al., 2002). Therefore, factors influencing the fertility of bulls, including nutrition, health, and quality and production traits of the semen along with methods to assess fertility should not be overlooked (DeJarnette et al., 2004; Kenny and Byrne, 2018; Butler et al., 2020).

Breeding soundness exam (BSE) is currently used to evaluate bull fertility. This exam includes a general physical examination, inspection of genital organs, and assessment of sperm production and quality (Kastelic and Thundathil, 2008). Nevertheless, changes to fertility in response to environmental and nutritional factors may go undetected by this exam. Interestingly, transcriptome, metabolomic, and proteomic studies have revealed potential biomarkers in sperm and seminal plasma of farm animals, such as bulls and stallions (Novak et al., 2010; Das et al., 2013; Menezes et al., 2019). Therefore, advances in the field of molecular biology bring new possibilities to unveil the impact of the environment on semen quality, allowing the identification of potential biomarkers of fertility.

Among sperm biomarkers for fertility, epigenetic markers deserve attention for their potential to change in response to nutrition and environmental factors. Epigenetics refers to heritable changes in gene expression without changes to the DNA sequence itself (Egger et al., 2004). Well-established epigenetic mechanisms include factors such as DNA methylation, histone post-translational modifications, and noncoding RNA (Tammen et al., 2013). Different patterns of sperm DNA methylation, one of the most widely studied epigenetic mechanisms, have been associated with sperm quality and male fertility (Schagdarsurengin and Steger, 2016; Kropp et al., 2017; Khezri et al., 2020; Marcho et al., 2020) and have been reported to change in response to environmental challenges, such as exposure to toxins (Acharya et al., 2020). Nutritional factors also have been reported to affect sperm DNA methylation in animal studies: in mice, studies have shown that different dietary protein or folate levels alter sperm DNA methylation (Lambrot et al., 2013; Watkins et al., 2018); in sheep, dietary supplementation of rumen-protected methionine altered sperm DNA methylation of rams (Gross et al., 2020); and in cattle, different planes of nutrition during the prepubertal period affect sperm DNA methylation of pubertal bulls (Perrier et al., 2020). Furthermore, it is now well accepted that changes to the epigenome of sperm, including DNA methylation, in response to the environment can have an impact on the offspring health (Pembrey et al., 2014; Schagdarsurengin and Steger, 2016). Therefore, studies addressing the consequences of physiological and metabolic stressors on bull sperm DNA methylation are important for their impact on the cattle industry.

Another epigenetic modification that has been attracting the attention of scientists for its potential to regulate mRNA processing and translation is methylation of the N6 position of adenosine (m⁶A), the most known internal modification of mammalian mRNA (Desrosiers et al., 1974; Meyer et al., 2012). Mature spermatic cells are considered transcriptionally inactive, but sperm carry RNA nonetheless (Casas and Vavouri, 2014). It is now well accepted that small noncoding RNA present in sperm are delivered to the oocyte at fertilization and are important to regulate early embryonic development (Liu et al., 2012) and to affect the offspring phenotype (McPherson et al., 2015; Rodgers et al., 2015), but the majority of transcripts in sperm consists of fragments of longer transcripts, such as mRNA, and these have a potential to serve as biomarkers of fertility (Casas and Vavouri, 2014). Indeed, recent studies have shown that m⁶A RNA methylation of sperm is critical for the regulation of spermatogenesis (Lin et al., 2017; Lin and Tong, 2019). Therefore, m⁶A methylation of fragments of mRNA in the ejaculate can be seen as a remnant of the process of spermatogenesis, making it a promising biomarker of sperm quality and fertility. Thus, studies aiming to understand what factors can change this modification in sperm, including nutritional factors, are required.

Animal experimental studies and human epidemiological observations have established the importance of parental nutrition in inducing sperm epigenetic modifications with consequences to the offspring health and metabolism (Schagdarsurengin and Steger, 2016; Donkin and Barres, 2018). Nevertheless, studies focusing on sperm epigenetic markers and their correlation with sperm quality of farm animals in response to nutrition are limited. Thus, this study aimed to characterize the effects of nutritional status on epigenetic markers, such as DNA methylation and m⁶A RNA methylation, in sperm of postpubertal bulls. We hypothesized that nutritional metabolic state, as during periods of negative energy balance (NEB) and subsequent compensatory growth, affects sperm m⁶A RNA methylation and 5-methyl cytosine (mC) DNA methylation, thus highlighting the importance of proper nutritional management of breeding bulls, including but not limited to the breeding season.

Material and Methods

The animals used in this experiment were cared for according to guidelines approved by the Institutional Animal Care and Use Committee University of Nevada, Reno (protocol #00738).

Animals, treatment, and experimental area

The dataset was obtained from 12 Angus × Hereford crossbred breeding bulls (n = 6, 23 ± 0.55 mo [young bulls], 558 ± 6.1 kg; n = 6, 47 ± 1.2 mo [mature bulls], 740 ± 30.5 kg) over a period of 180 d. Animals were housed at the Main Station Research Feedlot Facility at the Nevada Agricultural Experiment Station in Reno, NV. Bulls had free access to water and trace mineral salt during the whole trial. Beardless wheat (Triticum aestivum; 945 g/kg dry matter, 579 g/kg total digestible nutrients, and 86 g/kg crude protein) hay was delivered daily.

Three dietary regimes were offered to the bulls throughout the experimental period. Bulls coming from the breeding herd pastures were fed targeting for no body weight (BW) changes for 12 d as baseline (phase 1: BW maintenance adjustment preceding a BW loss); BW loss for 78 d (phase 2: BW loss targeted for 0.6 kg/d); fed for BW gain targeted for 1 kg/d for 90 d (phase 3: compensatory growth). Bulls were kept apart from cows on pasture and no semen was collected from the previous Spring breeding season to the beginning of this trial. In order to induce a metabolic stress and a full recovery to initial state, each animal acted as its own control in a repeated measures design. Furthermore, the feeding protocol requisites entailed simulating the same conditions at the beginning (phase 1) and at the end of the experimental period (end of phase 3) per metrics of BW and body condition score (BCS).

The animals were randomly assigned to one of the two pens (15 × 28 m). Each pen contained 30 m² of shaded area and four automated scales that recorded BW changes daily (ASMS; Model WD-1000 Master, Intergado Ltd, Contagem, Minas Gerais, Brazil). Prior to the beginning of the trial, each animal was fitted with an electronic identification ear tag (FDX-ISO 11784/11785; Allflex, Joinville, Santa Catarina, Brazil). Due to the dimensions of the scale, only one bull was allowed at a time into one individual scale when accessing the water troughs. The BW data were continuously recorded, transferred, and stored in the cloud for further analysis. The automated scale monitoring system (ASMS) scales were calibrated once weekly (using appropriate block calibration weights provided by the manufacturer) to ensure data accuracy.

Environmental conditions

Environmental conditions may trigger metabolic stress mechanisms that may change the capability of animals to perform or recover from a given stressful event. The climate of the area where the study was conducted is classified according to Köppen-Geiger (Kottek et al., 2006) as BSk (cold-semi arid climate) located at global positioning system of latitude 39°32ʹ38.87″N and longitude 119°48ʹ57.76″ W. The annual average temperature, rainfall, snowfall, and frost-free period are 10.1 °C, 18.6 mm, 228.6 mm, and 8.25 mo, respectively. The temperature–humidity index (THI) adjusted for solar radiation and wind speed according to Mader et al. (2006) was used to characterize environmental conditions. Ambient temperature (°C), relative humidity (%), solar radiation (W/m²), and wind speed (mph) were recorded daily throughout the entire trial (October to March) at 1 min intervals using a HOBO data logger (HOBO H8 Pro Series, Onset Computer Corp, Bourne, MA).

Animal data assessment

The BCS was measured using the NU Beef BCS App (University of Nebraska-Lincoln, Nebraska, USA) utilizing a scale ranging from 1 to 9, as recommended by NASEM (2016). The BCS measurements were performed on 0, 45, 90, 135, and 180 d of the trial. The BW was measured daily by the ASMS. The average daily gain (ADG, kg/d) was measured daily by the ASMS.

Reproductive traits and sperm processing

On the initial day of trial, bulls were submitted to a BSE. All bulls met (Table 1) the physical exam and the requirements for the minimal scrotal circumference (≥34 cm), sperm motility (≥30%), and normal sperm morphology (≥70%) as recommended by the Society of Theriogenology (Chenoweth et al., 1993). Bulls were restrained individually in a squeeze chute and submitted to semen collection via electroejaculation (Palmer, 2005), performed on days 0, 90, and 180 of the trial. All bulls were cleaned to remove extragonadal reserves of sperm on days 45 and 135 of the trial. Scrotal circumference was measured using a scrotal tape. Immediately after each semen collection, a sample was taken for an immediate evaluation of sperm parameters (motility, progressive motility, and vigor), and the remaining ejaculate was taken to a nearby laboratory in the experimental station where it was subdivided for processing of RNA isolation, assessment of concentration and morphology, and cryopreservation.

Table 1.

Descriptive statistical analyses of scoring criteria for breeding soundness exam (BSE)

Item1,3	Mean	SD	Min2	Max2	CV2, %	Amp2
SC	38.4	3.35	34	43	8.7	9
SM	75.0	16.83	55	95	22.4	40
NS	80.3	8.06	70	92	10.0	22

SC, scrotal circumference; SM, sperm motility; NS, normal spermatozoa.

Min, minimum; Max, maximum; CV, coefficient of variation; Amp, amplitude.

BSE was performed as recommended by the Society of Theriogenology (Chenoweth et al., 1993). Bulls met the physical exam requirements over all internal and external reproductive tract.

For evaluation of motility, progressive motility, and vigor, a 10-µL sample of the ejaculate was placed into a microcentrifuge tube containing 10 µL of warm (37 °C) phosphate buffered saline (PBS), homogenized, and placed in a warm (37 °C) glass with a slide cover for the analysis of motility, progressive motility, and vigor. These parameters were evaluated microscopically (Leica DM 500, 40× magnification; Wetzlar, Germany). For vigor, a scale from 0 to 5 representing the intensity of movement was used.

Processing of semen for isolation of RNA from spermatic cells was performed immediately post ejaculation as follows: fresh semen samples were centrifuged at 1,000 × g for 10 min at 4 °C for separation of sperm and seminal fluid; the cell pellets received 1 mL PBS followed by centrifugation and supernatant was discarded; the cell pellets then were incubated with lysis buffer (0.1% sodium dodecyl sulfate [SDS], 0.5% Triton-X; 500 µL) for 10 min on ice and then once again centrifuged and supernatant was discarded; finally, cell pellets were placed into sterile, RNase- and DNA-free, microcentrifuge tubes containing 500 µL of Qiazol lysis reagent (Qiagen, Hilden, Germany) and samples were frozen at −80 °C until further processing for RNA extraction.

Spermatozoa concentration was evaluated up to 30 min after semen collection using a hemocytometer. Briefly, fresh semen was added to formalin in 1:200 dilutions. Then, 8 µL of this mixture was placed on the two sides of the hemocytometer to allow a double count. The sperm count was determined microscopically (SKU: T690B-PCT200INF-PL, AmScope, CA, USA) using a 40× objective lens. Five squares on the diagonal from each side of the hemocytometer were counted. The number of sperm cells was determined by the average from the two sides multiplied by a constant of 10,000,000 and the semen volume.

For morphology assessment of spermatic cells, slides were prepared within 30 min after semen collection. A semen drop was placed at the end of a glass slide followed by a drop of eosin–nigrosin stain (Hancock Stain, Animal Reproduction Systems, CA, USA). The mixture was slowly spread on the slide with the edge of another glass slide at 30 degrees and left to dry. Slides were analyzed within 1 wk microscopically (Leica DM 500 Microscope), using 100× objective lens with immersion oil (Cargille Laboratories, NJ, USA). A total of 100 cells were counted throughout the slide, and the number of normal and abnormal sperm observed was recorded. Abnormalities were designated as defects involving the head (e.g., detached head and defects in size and shape), midpiece (e.g., distal mid-piece reflex, bowed mid-piece, and proximal droplet), or tail (e.g., bent tail and coiled tail). Then, sperm abnormalities were divided into two categories called primary (underdeveloped, double forms, acrosome defects, crater-diadem defect, pear-shaped head, small and free abnormal heads, proximal droplet, double bent and coiled tail, and accessory tail) and secondary (giant and short broadheads, detached head, detached acrosome membranes, abaxial midpiece, distal droplet, simple bent tail, and terminal coiled tail) as recommended by Chenoweth et al. (1993).

For cryopreservation, semen samples were kept in a water bath at 37 °C prior to spermatozoa concentration assessment. Following concentration assessment, samples received the necessary volume of semen extender to obtain 30 straws/bull in a concentration of 25 × 10⁶ sperm cells/straw. A Tris egg yolk extender (Two-step extender, Continental Plastic Corp., WI, USA) containing Tris 12.10 g/L, citric acid 6.9 g/L, fructose 5.0 g/L, glycerol 70 ml/L, 20% egg yolk (v/v), and an antibiotic cocktail was used. Following the manufacturer’s instructions, fraction A of the extender was added at 37 °C and fraction B was added following cooling at the semen at 4 °C for 4 h. Semen straws were then filled by suction, closed with polyvinyl alcohol, placed in a floating straw rack (cat#: FSR-101-LC, ARS, CA, USA) for 10 min, and stored in liquid nitrogen for future isolation of DNA.

RNA extraction and N⁶-methyladenosine assessment

Total RNA was extracted from fresh spermatozoa samples via miRNEasy mini kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. Samples of RNA were solubilized in RNAse- and DNAse-free molecular grade water (Invitrogen, Carlsbad, CA, USA), quantified at 260 nm using a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies Inc., Wilmington, DE), and stored at −80 °C for further analyses. The RNA m⁶A levels were colorimetrically quantified in RNA of spermatozoa cells samples via the Epiquik m⁶A RNA Methylation Quantification Kit (Epigentek, Farmingdale, NY) according to the manufacturer’s instructions using a microplate reader (SpectraMax M2e; Molecular Devices, LLC, San Jose, CA) and analyzed through the SoftMax Pro software (Molecular Devices, LLC, San Jose, CA). Briefly, 100 ng of RNA samples (260/280 ratio > 2.0) was added to each well (2 replicates per sample), plus 2 replicates per negative (0 ng/uL), and plus 2 replicates per positive controls (0.01, 0.02, 0.05, 0.1, 0.2, and 0.5 ng/uL), and absorbance was read at 450 nm. The intra-assay coefficient of variation (CV) averaged 10.05%. The cutoff limit to ensure consistent results between replicate was an intra-assay CV of 12.5%.

Genomic DNA extraction and 5-mC DNA methylation

For genomic DNA extraction, frozen semen samples were thawed at 37 °C for 30 s, and spermatic cell pellets were washed three times in 10 mL PBS via centrifugation at 10,000 × g for 10 min. After washing, cell pellets were resuspended in 500 µL of PBS and transferred to a microcentrifuge tube to be stored at −80 °C for further processing and analysis. Total genomic DNA was extracted from 50.0 × 10⁶ spermatozoa cell samples using the Purelink Genomic DNA Mini Kit (Invitrogen, Carlsbad, CA). Briefly, spermatic cells were selectively bound to Purelink Genomic DNA silica-based membrane (Invitrogen, Carlsbad, CA) in the presence of chaotropic salts. The cells were digested using Purelink Genomic proteinase K at 55 °C (Invitrogen, Carlsbad, CA). Any residual RNA was removed by digestion with Purelink Genomic RNase A (Invitrogen, Carlsbad, CA) prior to binding samples to the silica membrane. Then, samples of DNA were solubilized in 35 µL of Purelink Genomic DNA Elution Buffer (Invitrogen, Carlsbad, CA), quantified at 260 nm using a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies Inc., Wilmington, DE), and stored at −20 °C for further analyses. The 5-mC DNA methylation levels were determined using the colorimetric enzyme-linked immunosorbent assay, MethylFlash 5-mC DNA methylation quantification kit (Epigentek Inc., Farmingdale, NY, USA), and a microplate reader (SpectraMax M2e; Molecular Devices, LLC, San Jose, CA) and were analyzed through the SoftMax Pro software (Molecular Devices, LLC, San Jose, CA) according to the manufacturer’s instructions. Briefly, 100 ng of DNA samples (260/280 ratio > 1.8) was added to each well (2 replicates per sample), plus 2 replicates per negative (0 ng/uL), and plus 2 replicates per positive controls (0.1, 0.2, 0.5, 1.0, 2.0, and 5.0 ng/uL), and absorbance was read at 450 nm. The intra-assay CV averaged 5.09%. The cutoff limit to ensure consistent results between replicate was an intra-assay CV of 7.5%.

Physiological traits assessment

Blood samples were collected at 0700 hours at 0, 90, and 180 d. Blood was collected via jugular venipuncture using a vacutainer Luer-Lok access device (20-gauge × 1 in.; Airtite Product Co. Inc., Virginia Beach, VA) and a 10-mL heparinized plastic blood collection tube (Vacutainer, Becton Dickinson, Franklin Lakes, NJ) to determine plasma concentrations of insulin-like growth factor-1 (IGF-1). Additional blood samples (10 mL) were collected from all bulls, via jugular venipuncture into tubes containing no additives (Vacutainer, Becton Dickinson, Franklin Lakes, NJ) to determine serum leptin concentration. All blood samples were immediately placed on ice following collection for transportation and then centrifuged at 1,200 × g for 25 min at 4 °C in the laboratory. Serum and plasma samples were stored frozen at −20 °C until further laboratory analysis.

Plasma concentrations of IGF-1 were assessed using a human-specific commercial ELISA kit (SG100; R&D Systems Inc., Minneapolis, MN) previously validated for bovine samples (Moriel et al., 2012). Briefly, 50 uL of plasma samples was added to each well (2 replicates per sample), plus 2 replicates per negative (0 ng/mL), and plus 2 replicates per positive controls (0.094, 0.188, 0.375, 0.75, 1.5, 3.0, and 6.0 ng/mL), and absorbance was read at 450 nm adjusted for wavelength at 540 nm. Intra-assay CV averaged 1.70%, and the inter-assay averaged 6.30%. A commercial bovine ELISA kit was utilized to determine the serum concentration of leptin (EKU05589; Biomatik USA, LLC, Wilmington, DE). Briefly, 100 uL of serum samples was added to each well (2 replicates per sample), plus 2 replicates per negative (0 ng/mL), and plus 2 replicates per positive controls (0.156, 0.312, 0.625, 1.25, 2.5, 5.0, and 10.0 ng/mL), and absorbance was read at 450 nm. Intra-assay CV averaged 4.70%, and the inter-assay averaged 5.65%. For IGF-1 and leptin analyses, the cutoff limit to ensure consistent results between replicates was an intra-assay CV of 10.0%.

Statistical analyses

Data were collected and analyzed following a pre–post repeated measure design (Steiger Burgos et al., 2001). The statistical model used is shown below:

$${\displaystyle \begin{array}{l}{\displaystyle Y_{ij}=\mu +T_i+A_j+{}_j+T_i\times A_j+{\epsilon }_{ij}}\end{array}}$$

Where Y_ij is the observation taken (BW, BCS, ADG, DNA 5-mC methylation, RNA m⁶A methylation, hormones, reproductive, and semen traits) on the jth experimental unit for the pre–post ith treatment, µ is the overall mean, T_i is the effect of the ith treatment (periods of BW maintenance, BW loss, and BW gain), ϸ_j is the effect of the jth experimental unit (each animal acted as its own control), A_j is the effect of the jth age (young and mature bulls), T_i× A_j is the interaction effect of the ith treatment and the jth age, and ε_ij is the unobservable random error on the jth experimental unit associated with each ith pre–post treatment.

Statistical analyses were performed using the PROC MIXED procedure of SAS (ver. 9.4, SAS Inst. Inc., Cary, NC). Pearson correlation coefficients among variables were obtained with PROC CORR. All figures were generated by GraphPad Prism (ver. 9.0, GraphPad Inc., San Diego, CA). Outliers were tested by plotting the studentized residuals, and data points were removed if the Studentized residual was outside the range of −2.5 to 2.5. Normality assumption was tested using Shapiro–Wilk’s test, and homogeneity of variance was evaluated through Levene’s test. Percentage data (sperm motility, sperm progressive motility, normal spermatozoa, abnormal spermatozoa, primary spermatozoa defect, and secondary spermatozoa defect) were converted to a proportion, an arcsine transformation was done, and transformed data were analyzed (non-transformed data were reported). Statistical significance was declared at P ≤ 0.05 and statistical tendency 0.05 < P ≤ 0.10 using Tukey’s post hoc test. Data were evaluated as repeated measures over time (Kaps and Lamberson, 2004).

Results

Environmental conditions

Descriptive statistical analyses from adjusted THI are presented in Table 2. According to the Livestock Weather Safety Index (LCI, 1970), animals were in their thermal comfort zone during the entirety of the experimental period.

Table 2.

Descriptive statistical analyses for adjusted temperature–humidity index (THI) during periods of body weight (BW) maintenance, body weight loss, and compensatory growth of young (n = 6) and mature (n = 6) crossbred Angus × Hereford breeding bulls

Phase1	Mean (THI)	SD	Min2	Max2	CV2, %	Amp2	% days THI3 Normal4	% days THI Alert4	% days THI Danger4
Phase 1	56.9	3.5	51.6	63.0	6.2	11.4	100.0	0	0
Phase 2	48.9	7.5	34.8	60.1	15.4	25.2	100.0	0	0
Phase 3	49.5	4.9	36.7	58.8	9.9	22.05	100.0	0	0

Phase 1: BW maintenance adjustment preceding BW loss (12 d, October); phase 2: BW loss (78 d, mid-October to December); and phase 3: compensatory growth (90 d, January to March).

Min, minimum; Max, maximum; CV, coefficient of variation; Amp, amplitude.

THI, adjusted temperature–humidity index (unitless; Mader et al., 2006).

Thresholds cutoffs point from Livestock Weather Safety Index (LCI, 1970): THI ≤ 74 = Normal, 74 > THI ≤ 79 = Alert, 79 > THI ≤ 84 = Danger, and THI ≤ 84 = Emergency.

Performance

When bulls experienced NEB, growing bulls decreased 26.8% and 20.4% of BW and BCS, respectively, whereas mature bulls decreased 25.7% and 26.8% (Figures 1A and 2). A full recovery to the initial state of BW and BCS was observed for growing and mature bulls at the end of the compensatory growth (phase 3) as planned. The ADG experienced by growing bulls was similar (Figure 1B; P ≥ 0.14) to mature bulls over time.

Figure 1.

Body weight (kg, A) and daily gain (kg/d, B) changes of young (n = 6, filled diamond symbol) and mature (n = 6, plus symbol) crossbred Angus × Hereford breeding bulls undergoing periods of body weight maintenance (PS1), body weight loss (PS2), and compensatory growth (PS3). Light- and dark-shaded areas indicate 95% confidence interval.

Figure 2.

Body condition score (scale 1 to 9) changes of young (n = 6, filled diamond symbol) and mature (n = 6, plus symbol) crossbred Angus × Hereford breeding bulls undergoing periods of body weight maintenance (PS1), body weight loss (PS2), and compensatory growth (PS3). Error bars show the standard error of the mean.

5-Methyl cytosine DNA methylation

The abundance of DNA methylation of spermatic cells tended to be higher (P = 0.07; Figure 3A) during the phase of BW loss compared with the other phases. No differences in DNA methylation were detected between the periods of BW maintenance and compensatory growth. Effects of age (P = 0.426) or time × age (P = 0.791) were not observed for DNA methylation of spermatic cells.

Figure 3.

5-Methyl cytosine (5-mC) DNA methylation (ng, A) and N⁶-methyladenosine (m⁶A) RNA methylation (ng, B) in sperm of young (n = 6, unshaded) and mature (n = 6, shaded bar) crossbred Angus × Hereford breeding bulls undergoing periods of body weight maintenance (day = 0), body weight loss (day = 90), and compensatory growth (day = 180). Error bars show the standard error of the mean. Least square means followed by different letters are statistically different (RNA m⁶A; Tukey’s test; P < 0.05) or statistical tendency (DNA 5-mC; Tukey’s test; 0.05 < P ≤ 0.10).

RNA m⁶A methylation

The abundance of RNA m⁶A methylation in spermatic cells was greater (P = 0.004; Figure 3B) during periods of NEB and compensatory growth compared with the initial day of the trial (bulls coming from the breeding herd pastures). No differences (P > 0.05) were detected between NEB and compensatory growth periods for RNA m⁶A methylation. Young bulls (0.13 ± 0.01 ng) had a tendency (P = 0.109) of higher RNA m⁶A methylation levels than mature bulls (0.10 ± 0.009 ng). No significant (P = 0.368) time × age interaction was observed for RNA m⁶A methylation of spermatic cells.

Reproductive and sperm morphology traits

The bulls’ reproductive parameters are shown in Figures 4 and 5. Time influenced scrotal circumference (P < 0.001; Figure 4A), sperm motility (P = 0.004; Figure 4B), sperm progressive motility (P = 0.004; Figure 4C), sperm volume (P = 0.007; Figure 5A), and total sperm count (P = 0.049; Figure 5C). During the NEB, bulls decreased 8.4% of scrotal circumference, ejaculated 40.8% less sperm volume, decreased 34.0% of sperm motility, decreased 35.9% of sperm progressive motility, and decreased 50.0% of total sperm count compared with the initial day of the trial. By the end of the compensatory phase, bulls had similar (P > 0.05) scrotal circumference, sperm volume, sperm motility, sperm progressive motility, and total sperm count compared with the initial day of the trial. Young bulls had lower (P = 0.005) scrotal circumference (33.9 ± 1.34 cm) than mature bulls (40.6 ± 0.95 cm). Effect of time × age interaction was not detected for scrotal circumference (P = 0.558). There were no effects of age (P ≥ 0.168) and time × age interaction (P ≥ 0.438) in sperm volume, sperm motility, sperm progressive motility, and total sperm count. Additionally, no effects of time (P ≥ 0.459), age (P ≥ 0.598), and time × age interaction (P ≥ 0.406) were detected in sperm vigor (Figure 4D) and sperm concentration (Figure 5B).

Figure 4.

Scrotal circumference (cm, A), sperm motility (%, B), sperm progressive motility (%, C), and sperm vigor (scale 0 to 5, D) of young (n = 6, unshaded) and mature (n = 6, shaded) crossbred Angus × Hereford breeding bulls undergoing periods of body weight maintenance (day = 0), body weight loss (day = 90), and compensatory growth (day = 180). Error bars show the standard error of the mean. Least square means followed by different letters are statistically different (Tukey’s test; P < 0.05).

Figure 5.

Sperm volume (mL, A), sperm concentration (spermatozoa cell/mL, B), and total sperm count (×10⁹, C) of young (n = 6, unshaded) and mature (n = 6, shaded) crossbred Angus × Hereford breeding bulls undergoing periods of body weight maintenance (day = 0), body weight loss (day = 90), and compensatory growth (day = 180). Error bars show the standard error of the mean. Least square means followed by different letters are statistically different (Tukey’s test; P < 0.05).

In terms of spermatozoa morphology, effect of time was detected for normal spermatozoa (P < 0.001; Figure 6A), abnormal spermatozoa (P < 0.001; Figure 6B), primary spermatozoa defect (P = 0.039; Figure 6C), and secondary spermatozoa defect (P < 0.001; Figure 6D). During the NEB period, bulls decreased 47.7% of normal spermatozoa, increased 205.9% abnormal spermatozoa, increased 106.4% of primary spermatozoa defect, and increased 321.7% of secondary spermatozoa defect compared with the initial day of the trial. By the end of the compensatory growth period, bulls had similar (P > 0.05) normal and abnormal spermatozoa and primary and secondary spermatozoa defects to the initial day of the trial. In addition, there were no effects of age (P ≥ 0.342) and time × age interaction (P ≥ 0.206) in normal and abnormal spermatozoa, primary and secondary spermatozoa defects.

Figure 6.

Normal spermatozoa (%, A), abnormal spermatozoa (%, B), primary spermatozoa defect (%, C), and secondary spermatozoa defect (%, D) of young (n = 6, unshaded) and mature (n = 6, shaded) crossbred Angus × Hereford breeding bulls undergoing periods of body weight maintenance (day = 0), body weight loss (day = 90), and compensatory growth (day = 0). Error bars show the standard error of the mean. Least square means followed by different letters are statistically different (Tukey’s test; P < 0.05).

Physiological traits

The IGF-1 and leptin concentrations are shown in Figure 7. Bulls decreased (P = 0.010; Figure 7A) IGF-1 concentration during the NEB, whereas, comparing the baseline with compensatory growth, bulls had similar (P > 0.05) IGF-1 concentrations. Effects of age (P = 0.577) and time × age interaction (P = 0.868) were not detected for plasma concentrations of IGF-1.

Figure 7.

Insulin-like growth factor-1 (ng/mL, A) and leptin (ng/mL, B) concentration of young (n = 6, unshaded) and mature (n = 6, shaded) crossbred Angus × Hereford breeding bulls undergoing periods of body weight maintenance (day = 0), body weight loss (day = 90), and compensatory growth (day = 180) as function of the periods of sampling. Error bars show the standard error of the mean. Least square means followed by different letters are statistically different (Tukey’s test; P < 0.05).

Effect of time (P = 0.155) was not detected for serum concentrations of leptin (Figure 7B). A tendency was detected for the time × age interaction (P = 0.082) for serum leptin concentrations. Young bulls tended (P = 0.102) to present greater leptin concentration than mature bulls during the overall evaluated sampling days of the trial, which was more evident under the compensatory growth period.

Pearson correlation analyses

Pearson correlation among DNA methylation, RNA m⁶A methylation, hormones, and semen traits is presented in Table 3. Specific for mature bulls, DNA methylation was positively correlated with leptin concentration (0.569, P = 0.021) and tended to be negatively correlated with sperm total motility (−0.446, P = 0.083), whereas for young bulls, DNA methylation was positively correlated with abnormal spermatozoa (0.824, P = 0.006), primary spermatozoa defect (0.711, P = 0.032), and secondary spermatozoa defect (0.661, P = 0.052) and negatively correlated with normal spermatozoa (−0.824, P = 0.006), total sperm count (−0.702, P = 0.035), and sperm concentration (−0.846, P = 0.004). Furthermore, in young bulls, DNA methylation tended to be negatively correlated with IGF-1 serum levels (−0.578, P = 0.103).

Table 3.

Pearson correlation specific for age between DNA 5-mC, RNA m⁶A, hormones and semen traits during periods of body weight maintenance, body weight loss, and compensatory growth of young (n = 6) and mature (n = 6) crossbred Angus × Hereford breeding bulls

Item1	5-mC	m6A	IGF-1	LEP	NS	AS	PSD	SSD	SpV	SM	SPM	TSPTZ	SV	[S]
			Young bulls2
5-mC	1.000 (1.00)	0.081 (0.837)	−0.578 (0.103)∗	−0.107 (0.783)	−0.824 (0.006)∗	0.824 (0.006)∗	0.711 (0.032)∗	0.661 (0.052)∗	−0.742 (0.021)∗	−0.275 (0.474)	−0.469 (0.202)	−0.702 (0.035)∗	−0.469 (0.202)	−0.846 (0.004)∗
m6A	—	1.000 (1.000)	0.345 (0.362)	0.520 (0.151)	0.168 (0.665)	−0.165 (0.664)	−0.264 (0.492)	0.002 (0.996)	−0.161 (0.678)	−0.100 (0.797)	−0.101 (0.799)	0.004 (0.992)	0.152 (0.697)	0.216 (0.577)
IGF-1	—	—	1.000 (1.000)	0.493 (0.178)	0.824 (0.006)∗	−0.839 (0.005)∗	−0.705 (0.034)∗	v0.667 (0.049)∗	0.351 (0.353)	0.524 (0.147)	0.528 (0.141)	0.146 (0.708)	0.493 (0.177)	−0.073 (0.851)
LEP	—	—	—	1.000 (1.000)	0.297 (0.436)	−0.297 (0.436)	−0.473 (0.198)	0.011 (0.978)	−0.050 (0.898)	0.431 (0.247)	0.430 (0.248)	0.152 (0.696)	0.369 (0.408)	0.363 (0.337)
			Mature bulls2
5-mC	1.000 (1.00)	0.011 (0.968)	−0.316 (0.233)	0.569 (0.021)∗	−0.250 (0.349)	0.225 (0.401)	0.148 (0.584)	0.071 (0.794)	−0.172 (0.523)	−0.446 (0.083)∗	−0.172 (0.524)	−0.300 (0.218)	−0.146 (0.589)	0.012 (0.963)
m6A	—	1.000 (1.000)	−0.077 (0.761)	−0.211 (0.401)	−0.024 (0.923)	0.046 (0.857)	−0.134 (0.595)	0.163 (0.517)	−0.047 (0.850)	−0.220 (0.379)	−0.229 (0.362)	−0.110 (0.665)	0.117 (0.644)	0.121 (0.633)
IGF-1	—	—	1.000 (1.000)	0.179 (0.476)	0.524 (0.025)∗	−0.498 (0.035)∗	−0.430 (0.075)∗	−0.306 (0.216)	0.542 (0.019)∗	0.414 (0.087)∗	0.402 (0.098)∗	0.504 (0.033)∗	−0.019 (0.937)	0.261 (0.294)
LEP	—	—	—	1.000 (1.000)	−0.291 (0.242)	0.258 (0.300)	0.195 (0.436)	0.125 (0.621)	−0.216 (0.388)	−0.309 (0.211)	−0.289 (0.245)	−0.320 (0.195)	−0.335 (0.174)	−0.328 (0.183)

5-mC, DNA 5-mC methylation; m⁶A, N⁶-methyladenosine methylation; IGF-1, insulin-like growth factor 1; LEP, leptin; NS, normal spermatozoa; AS, abnormal spermatozoa; PSD, primary spermatozoa defect; SSD, secondary spermatozoa defect; SpV, sperm volume; SM, sperm motility; SPM, sperm progressive motility; TSPTZ, total sperm count; SV, sperm vigor; [S], sperm concentration.

Young bulls (n = 6) and mature bulls (n = 6).

Statistical significance was declared at P ≤ 0.05 and statistical tendency 0.05 < P ≤ 0.10.

In mature bulls, IGF-1 serum levels were positively correlated with normal sperm (0.524, P = 0.025), sperm volume (0.542, P = 0.019), and total sperm count (0.504, P = 0.033) and negatively correlated with abnormal spermatozoa (−0.498, P = 0.035). There was a tendency for IGF-1 levels to be negatively correlated with sperm total motility (−0.414, P = 0.087), sperm progressive motility (−0.402, P = 0.098), and primary spermatozoa defect (−0.430, P = 0.075). In young bulls, IGF-1 was positively correlated with normal sperm (0.824, P = 0.006) and negatively correlated with abnormal sperm (−0.839, P = 0.005), primary spermatozoa defects (−0.705, P = 0.034), and secondary spermatozoa defects (−0.667, P = 0.049). RNA m⁶A methylation and leptin serum concentrations had no significant correlation (P ≥ 0.151) with any parameter measured.

Discussion

The fertility of bulls is critically important for the success of cow–calf operations and the AI industry (Thundathil et al., 2016). Investigating the impact of management practices and nutritional and environmental factors on semen quality and fertility is, therefore, crucial. However, testing male fertility via mating or AI is considered time-consuming and expensive (Larsson and Rodrigues-Martinez, 2000). Sperm quality parameters individually evaluated have limited value for predicting fertility, but the more sperm parameters can be tested, especially in combination, the better is the prediction of fertility (Farrell et al., 1998; Gillan et al., 2005; Morrell et al., 2017). Thus, unveiling biomarkers in sperm of bulls is important to improve fertility predictions. The objective of this study was to investigate if periods of nutritional and metabolic changes affect two sperm epigenetic markers, DNA methylation and RNA m⁶A methylation, and their correlations with sperm quality parameters in postpubertal bulls.

Nutrition and energy balance are known to alter sperm quality parameters of ruminants (Martin et al., 2010; Ros-Santaella et al., 2019). Grazing breeding bulls in extensive production systems, such as the ones observed in the rangelands of Western United States, often undergo NEB at least at some point of their year-round production cycle (NASEM, 2016). Indeed, in the present study, the rates of weight loss and weight gain successfully induced metabolic changes according to NASEM (2016), and a detrimental impact of NEB on reproductive and sperm quality parameters was evident. Specifically, the period of NEB decreased scrotal circumference, sperm volume, sperm motility, and normal spermatozoa count while increasing abnormal spermatozoa count and spermatozoa primary and secondary defects. Although a period of NEB did not decrease sperm concentration, it decreased the total sperm count in the ejaculate because of less ejaculate volume. These results may have important implications for the bovine AI industry and for beef operations in rangeland conditions. For example, the difference in total sperm count per ejaculate does reduce the total number of insemination doses obtained from one ejaculate, and bulls undergoing NEB in grazing conditions may produce a lower number of calves. Therefore, ranchers and the AI industry should pay attention to the nutritional management of breeding bulls, as nutritional strategy could likely result in large differences in total sperm output.

In the present study, the nutritional management of bulls successfully induced a period of NEB and a subsequent return to their initial metabolic state. The adjusted THI supported our hypothesis that animals would undergo similar metabolic challenges simultaneously. Leptin and IGF-1, important regulators of energy homeostasis (Kawai and Rosen, 2010; Park and Ahima, 2015), were used in this study as metabolic parameters. IGF-1 is involved in the main metabolic pathways related to animal growth and distribution of nutrients to different tissues (Owens et al., 1993), whereas leptin has been positively associated with dry matter intake, ADG, and body fatness measures of beef steers and heifers (Foote et al., 2016). In the present study, IGF-1 serum concentrations were lower during the period of NEB, whereas no changes were observed in circulating levels of leptin in response to diet. The return to positive energy status following NEB was characterized by a steep rise in circulating IGF-1 levels, which likely promoted an accelerated growth during the compensatory growth period shifting tissue deposition patterns toward fat deposition on depots where it could be more easily reassessed in case another NEB period was to come (Yang et al., 2019; Barboza et al., 2020). Interestingly, IGF-1 was positively correlated with ejaculate volume and several quality sperm parameters, including sperm motility and progressive motility, normal spermatozoa, and total sperm count in mature bulls, while it was negatively correlated with abnormal spermatozoa irrespective of bulls age. This is in agreement with a previous study that reported a positive correlation between serum IGF-1 levels and sperm motility and concentration in buffalo bulls (Kumar et al., 2019), which indicates that IGF-1 may be a potential biomarker for the fertility of bulls.

DNA methylation is one the most extensively studied epigenetic mechanisms and has been reported to be altered in various human tissues in response to changes in body composition (Aronica et al., 2017; Nishida et al., 2020). The impact of nutrition on this epigenetic mechanism has also been reported in studies that found that BW changes affect the methylation pattern of several genes, including genes associated with obesity (Cordero et al., 2011), metabolism (Martín-Núñez et al., 2014), and fertility (Sujit et al., 2018; Liu et al., 2019). In sperm, several animal studies have unveiled the impact of nutrition on DNA methylation (Lambrot et al., 2013; Watkins et al., 2018; Gross et al., 2020; Perrier et al., 2020), but studies investigating how differential tissue utilization and metabolic states affect the sperm epigenome and its potential reversal by nutritional management are limited. In the present study, sperm DNA methylation tended to increase in bulls undergoing NEB and was negatively correlated with normal spermatozoa, total spermatozoa count, and ejaculate volume and was positively correlated with spermatozoa defects in young, but not mature, bulls, which indicates that DNA methylation may serve as a biomarker of semen quality in young bulls. Sperm DNA methylation has been associated with sperm abnormalities and poor sperm quality in infertile men (Houshdaran et al., 2007; Lambrot et al., 2013). Reasons for why young bulls were more susceptible to changes in sperm DNA methylation in response to nutrition and what mechanisms are involved in this process remain to be unveiled by future studies. Since no techniques were employed for the separation of live and dead sperm prior to sperm processing in the present study, it is possible that changes in DNA methylation were affected by differences in sperm viability in response to diet, but this hypothesis requires further study. Furthermore, since DNA methylation can influence the transcription of small noncoding RNAs (Sato et al., 2011) and sperm microRNAs affect the offspring phenotype (McPherson et al., 2015; Rodgers et al., 2015), the possibility that a period of NEB regulates the sperm epigenome of bulls with effects on next generations impacting cattle productivity deserves further investigation. This possibility is strengthened by the recent mice studies reporting that paternal diet influences the sperm epigenome with consequences to reproductive parameters and health of the offspring (Fullston et al., 2012; Schagdarsurengin and Steger, 2016; Watkins et al., 2018).

The RNA m⁶A methylation is crucial for the regulation of various biological processes, including spermatogenesis (Zheng et al., 2013). It is becoming clear that RNA m⁶A methylation is an important regulator of gene expression, involved in the regulation of mRNA processing, including actions related to mRNA splicing, stability, translation, and nuclear export, and decay (Liu and Jia, 2014; Chandola et al., 2015; Roignant and Soller, 2017). The regulation of RNA m⁶A methylation is complex and involves m⁶A writers (methyltransferases), erasers (demethylases), and readers (m⁶A-binding proteins) (Liu and Jia, 2014; Roignant and Soller, 2017). Various m⁶A readers have been identified to date and they are important to determine the fate of m⁶A-containing mRNA, such as enhancing translation or mRNA decay (Lee et al., 2020). In the present study, sperm m⁶A RNA methylation was greater in periods of NEB and compensatory weight gain in comparison to a period of BW maintenance. It is possible that different genes were methylated or different m⁶A readers were active during steroidogenesis in periods of NEB in comparison to compensatory weight gain, but further research is required to confirm these hypotheses. In the present study, no correlation was observed between m⁶A RNA methylation and semen quality and sperm morphology, which may indicate that m⁶A RNA methylation is acting on the regulation of genes not directly related to the traits measured herein. These findings highlight the need to further investigate molecular changes affecting sperm quality in response to the environment as these can impact the efficiency of cattle production and can go undetected because of inherent limitations of measuring cellular metabolism in field conditions. Therefore, further studies are required to unveil what is the impact of changes in RNA m⁶A methylation in response to metabolic changes and whether m⁶A RNA methylation can be used as a biomarker of fertility.

In conclusion, the nutritional status of breeding bulls alters epigenetic markers, such as DNA methylation and RNA m⁶A methylation, in sperm and the impact of change seems to be age dependent. These markers may serve as biomarkers of sperm quality and fertility of bulls in the future. The period of NEB during the breeding season of cattle may affect the fertility of bulls and even have an impact on the health and performance of offspring through changes in sperm DNA methylation which may or may not be reversible. Detrimental effects on sperm production and seminal quality are observed at periods and places when and where environmental and nutritional limitations are a year-round reality and may carry hidden players that may influence a lifetime of underperformance.

Glossary

Abbreviations

5-mC

5-methyl cytosine

ADG

average daily gain

AI

artificial insemination

AS

abnormal spermatozoa

ASMS

automated scale monitoring system

BCS

body condition score

BSE

breeding soundness exam

BW

body weight

IGF-1

insulin-like growth factor-1

LEP

leptin

LRNS

Large Ruminant Nutrition System

m⁶A

N ⁶-methyladenosine methylation

NEB

negative energy balance

NS

normal spermatozoa

PSD

primary spermatozoa defect

[S]

sperm concentration

SC

scrotal circumference

SM

sperm motility

SPM

sperm progressive motility

SpV

sperm volume

SSD

secondary spermatozoa defect

SV

sperm vigor

THI

temperature-humidity index

TSPTZ

total sperm count

Conflict of interest statement

The authors declare no conflict of interest.