Semen quality and seminal plasma metabolites in male rabbits (Oryctolagus cuniculus) under heat stress

Experimental animals and design

This study was conducted at the Dongfang Rabbit Farm in Pizhou, Jiangsu Province, China (latitude 34.441208°N, longitude 117.986541°E, and altitude 28.8 m). Eight-month-old New Zealand White male rabbits (n = 20) were included in the study. All experimental male rabbits were raised in the same rabbit house by the same breeder in accordance with the conventional feeding and management mode of large-scale rabbit farms to maintain the same environmental conditions. Each male rabbit was raised in a single cage measuring 50 cm × 60 cm × 40 cm, and a nipple drinker was provided for clean water. The rabbits were fed in the morning and afternoon, every day. The rabbits were fed with commercial pellets (crude protein 16%, crude fiber 18%, crude ash 12%, calcium 1%, phosphorus 0.4%, lysine 0.6%, and H₂O 14%), containing corn, wheat bran, soybean meal, alfalfa hay powder, L-lysine, sodium chloride, calcium hydrogen phosphate, vitamin A, and vitamin D1. The animal experiments were conducted in strict accordance with protocols approved by the Animal Care Advisory Committee of the Anhui Academy of Agricultural Sciences (AAAS2022-17) under the “Guidelines for Experimental Animals” of the Ministry of Science and Technology (Beijing, China). No death or disease in the rabbits were observed during the experimental period.

The male rabbits were acclimated to the environment for 7 days prior to the start of the experiment. The rabbits were used for semen collection prior to artificial insemination. Semen was collected using an artificial vagina, a device which includes a cylindrical shell with three openings, an elastic sheath as inner tube, two rubber bands, and an attached semen collecting cup. Male rabbits were given regular contact with female rabbits to improve their sexual desire. During semen collection, a rabbit skin was fixed on the right arm of the ejaculator, the right hand held the false vagina, the gas nozzle was downward, and the finger was stuck. At the beginning of the training period, the assistant held an estrous rabbit on the rabbit skin-covered ejaculation arm. When the male rabbit climbed on the back of the female rabbit, the assistant slowly removed the female rabbit. The male rabbit fell on the rabbit skin, and the penis was inserted into the lubricated false vagina. Ejaculation was complete when the back of male rabbit curled up, the ‘cooing’ sound was made, and the male rabbit slipped to one side. The sperm extractor immediately erected the false vagina upward, deflated the sperm collection bottle, and plugged the disinfected cork. Sperm collection was performed thrice for each rabbit for subsequent experiments, with 2 to 3 day interval between each sperm collection. The semen collection dates were May 1 to 20 and August 1 to 20, 2021. The THI value was calculated; May was defined as NH for the non-heat stress group, and August as the H for the heat stress group. The semen collection occurred from 8:30 a.m. to 9:30 a.m., with 20 male rabbits participating in each session. The first sperm collection was used to detect the quality of the semen. The two remaining semen collections were centrifuged, and the supernatant and the bottom sperm were sucked out, respectively, and transferred to the enzyme-free cryo storage tubes. The samples were frozen in liquid nitrogen, and then transferred to –80 °C for storage.

After the last sperm collection, the humane endpoint for study was the death of 20 rabbits with a final live weight of approximately 4 kg each. The rabbits were stunned and exsanguinated via their carotid arteries and jugular veins. The testicular tissue of male rabbits was collected, washed with 0.9% saline, fixed with 4% paraformaldehyde, and used to produce tissue sections.

Determination of temperature and humidity index (THI)

Throughout the duration of the experiment, the ambient temperature and relative humidity were recorded thrice a day with a hygrograph. The hygrograph was located in the rabbit housing to ensure that it was not exposed to light and rain. The temperature and humidity at 8:00, 14:00, and 20:00 were measured each day. The daily ambient temperature and relative humidity were used to calculate the average daily temperature and humidity.

The temperature and humidity index (THI) was calculated using the following equation:

THI = db °C − [(0.31 − 0.31 RH) × (db °C − 14.4)]

db °C = dry bulb temperature and RH = relative humidity percentage/100. The obtained values were next classified as follows: <27.8 is no heat stress, 27.8–28.9 is moderate heat stress, 28.9–30.0 is severe heat stress, and 30.0 and above are very severe heat stress (Marai, Ayyat & Abd El-Monem, 2001).

Semen quality test of male rabbits

Determination of biochemical components in seminal plasma

The collected semen samples were placed in an icebox and thawed slowly. The seminal palsma biochemical index levels were determined using Heat shock protein 70 (HSP70), testosterone (T), alkaline phosphatase (ALP), acid phosphatase (ACP), α-glucosidase (α-Glu), and lactate dehydrogenase (LDH) ELISA Kit (Enzyme-linked, Shanghai, China), Fructose, potassium ion (K⁺), magnesium ion (Mg²⁺), sodium (Na⁺) in the seminal plasma were detected using fructose, potassium ion (K⁺) detection kit., a magnesium ion (Mg²⁺), and a sodium ion (Na⁺) detection kits (Jiancheng, Nanjing, China), respectively. All tests were conducted following the manufacturer’s recommendations.

Extraction of sample metabolites

To extract the metabolites, 100 μL of seminal plasma samples were placed in EP tubes, and 400 μL of 80% methanol aqueous solution was added. The samples were subjected to vortex oscillation, water bath treatment for 5 min at 15,000 rpm, and were then centrifuged at 4 °C for 20 min. A certain volume of supernatant was diluted with mass spectrometry water until the methanol content was 53%. It was then centrifuged at 15,000 rpm at 4 °C for 20 min by high-speed freezing centrifuge (Scilogex, Rocky Hill, CT, USA). Lastly, the supernatant was collected for analysis (Batzias, Theodosiadou & Delis, 2004; Qiao et al., 2017).

Preparation of QC sample and blank sample

The data quality control (QC) was composed of 18 experimental samples (nine for each group) of heat stress group and non-heat stress group in equal volume. The first three QCs were used to monitor the instrument state before injection using the equilibrium chromatography–mass spectrometry system (Thermo Fisher Scientific, Waltham, MA, USA). The next three QCs were scanned in segments, together with the secondary spectra obtained from the experimental samples for the qualitative analysis of metabolites. The QC in the sample test was used to evaluate the stability of the system during the experiment and the the data were analyzed quality control. QC1, QC2, and QC3 denoted repetitive operations. In a blank sample, a 53% methanol–aqueous solution replaced the experimental sample. The pre-treatment process was the same as that of the experimental samples.

Instrument parameters

Vanquish UHPLC from Thermo Fisher was used for liquid chromatography. The Hypesil Gold column was used (100 × 2.1 mm, 1.9 μm); the column temperature was 40 °C, and the flow rate was 0.2 mL/min. Under positive ion mode, mobile phase A was 0.1% formic acid, and mobile phase B was methanol. Under negative ion mode, mobile phase A was 5 mM ammonium acetate, pH 9.0, and mobile phase B was methanol. The chromatographic gradient elution procedure is shown in Table 1.

MS conditions were scanning range of m/z 100–1,500; the eSI source was set as follows: spray voltage: 3.2 kV; sheath gas flow rate was 40 arb; aux gas flow rate was 10 arb; the capillary temperature was 320 °C. Polarity was positive; negative; the MS/MS secondary scan was a data-dependent scan.

Data preprocessing and metabolite identification

The CD 3.1 database search software was used to import the offline data file and process it. The retention time, mass charge ratio, and other parameters of each metabolite were simply screened. The retention time deviation was set to 0.2 min and the mass deviation was set to 5 ppm. The peak alignment of different samples was performed for more accurate identification. Afterward, information such as mass deviation of 5 ppm, signal strength deviation of 30%, and minimum signal strength was set to extract the peaks. In addition, the area of the peak was quantified, and then the target ions were integrated. Finally, the molecular formula was predicted using molecular ion peaks and fragment ions, and compared with the mzCloud (https://www.mzcloud.org/; mzVault, and Masslist databases). The background ions were removed by blank sample, and the original quantitative results were standardized to obtain the identification and relative quantitative results of metabolites. The data were processed using the Linux operating system (CentOS version 6.6) and R software in Python.

Data analysis

The KEGG database (https://www.genome.jp/kegg/pathway.html) and the HMDB database (https://hmdb.ca/metabolites) were used to identify the metabolites. The metaX software (Wen et al., 2017) was used to convert the data, and principal component analysis (PCA) and partial least squares discriminant analysis (PLS-DA) were performed. The screening of differential metabolites was largely based on the results of PLS-DA. The variable importance in project (VIP) of the PLS-DA model was analyzed from the obtained multivariate analysis. The VIP value represents the contribution of metabolites to grouping, and the metabolites with different varieties or tissues can be preliminarily screened. In univariate analysis, the statistical significance (P-value) of metabolites between the two groups was calculated using a t-test, and the fold change (FC) value of the metabolites between the two groups was calculated. The default criteria for screening differential metabolites were VIP > 1.0, FC > 1.5 or < 0.667, and P < 0.05.

Changes in temperature, humidity, and THI index of rabbit farm in different months

The temperature of the New Zealand white rabbit housing from May to August was monitored. The results showed that the average temperature, humidity, and THI of the rabbit housing in May were 22.04 ± 2.52 °C, 61.82 ± 13.35%, and 20.94 ± 2.22, respectively, which was defined as the non-heat stress group. In August, the average temperature, average humidity, and THI of the rabbit houses were 30.15 ± 0.67 °C, 78.62 ± 6.88%, and 29.10 ± 2.52, respectively, which were defined as the heat stress group, as shown in Fig. 1 and Table 2.

Effect of heat stress on sperm motility in male rabbits

According to the state of sperm movement, the sperm was classified into fast straight forward (A), slow straight forward (B), non-forward movement (C), and non-movement (D). As shown in Table 3, compared with the non-heat stress group, the number of grade A sperm in the heat stress group was significantly decreased (P < 0.01). The number of grade B and C sperm cells were significantly increased (P < 0.01). No significant difference in grade D sperm count was observed between the two groups (P = 0.112).

Changes in semen quality of male rabbits under heat stress

Sperm motility, semen pH, sperm malformation rate, semen volume, and semen density were calculated. The results showed that sperm motility and semen pH were significantly decreased in the heat stress group (P < 0.01), the sperm malformation rate was significantly increased (P < 0.01), and the semen volume decreased significantly (P < 0.05). The sperm density was significantly decreased (P < 0.01). TSO, TMS, and TFSF were calculated to find that TSO, TMS, and TFSF in the heat stress group were significantly lower than those in the non-heat stress group (P < 0.01). The results are shown in Table 4.

Effect of heat stress on biochemical indexes of seminal plasma in male rabbits

The biochemical indexes of seminal plasma in male rabbits were analyzed; the results are shown in Table 5. Compared with the non-heat stress group, the heat stress protein 70 (HSP70) and acid phosphatase (ACP) in the heat stress group increased significantly (P < 0.01), and the testosterone (T), α-glucosidase (α-Glu), and fructose (fructose) in the seminal plasma decreased significantly (P < 0.01). Mg²⁺, Na⁺, and K⁺ in metal ions decreased significantly (P < 0.05 for Mg²⁺, P < 0.01 for Na⁺ and K⁺); no significant difference was observed in alkaline phosphatase (P = 0.148) and lactate dehydrogenase isoenzyme (P = 0.775) between the two groups.

Total ion chromatography (TIC): After adding the intensity of all ions in the mass spectrum at each time point, the spectrum was continuously depicted and which contained the ionic strength and the retention time of each component metabolite in chromatography. The total ion flow patterns of samples in the heat stress group and non-heat stress group are shown in Fig. 2. A total of 346 metabolites were identified by screening the data.

Data quality control analysis and principal component analysis

The relative quantitative values of metabolites were used to calculate Pearson’s correlation coefficients between QC samples (Rao, Sui & Zhang, 2016). The calculation results are shown in Fig. 3A. QC sample correlation was close to 1, indicating that the stability of the whole detection was good. The PCA analysis of peaks extracted from all experimental samples and QC samples was performed, and the results are shown in Fig. 3B. The QC samples were collected together, indicating that the difference in QC samples was small. This shows that the whole method had good stability and high data quality. There was a significant difference in data between the heat stress group and the non-heat stress group, indicating that the seminal plasma metabolism of male rabbits was significantly changed after heat stress.

Selection of differential metabolites

With reference to the VIP, FC, and P-value, the identified metabolites were screened. A total of 71 differential metabolites were screened (Table S1). The R program package ggplot2 software was used to draw the volcanic map of differential metabolites. The volcanic map depicted the overall distribution of differential metabolites (Fig. 4): there were 48 upregulated metabolites and 23 downregulated metabolites.

Enrichment results from metabolic pathways of differential metabolites

HMDB and KEGG were used to annotate differential metabolites (Fig. 5). As shown in Fig. 5A, HMDB annotated 35 differential metabolites, including carboxylic acids and their derivatives, aliphatic acyl, indoles, and their derivatives. KEGG showed enrichment for 51 metabolic pathways, including ketone body synthesis and degradation, serine and threonine metabolism, tryptophan metabolism, citric acid cycle, and other metabolic pathways (Fig. 5B).