Animals and housing
The experimental protocol was performed in accordance with guidelines established by the Animal Management Rules of the Ministry of Health of the People’s Republic of China. In addition, our study protocol was approved by the Animal Care and Use Committee of Sichuan Agricultural University (No. 20190318). The experiment was conducted at a commercial pig farm (Guangyuan, Sichuan province, China).
A total of 120 multiparous Yorkshire × Landrace sows (3–5 of parity) were randomly and equally divided into control (CON) and daidzein (DAI) groups. After confirming that all sows were in the oestrous stage, the sows were artificially inseminated twice with unfrozen semen within two days. The sows were housed in individual gestation stalls (2.20 × 0.65 m) from day 1 of mating to the day 34 of gestation with free access to drinking water. The ambient temperature was maintained at between 20 and 25 °C. During the feeding period, special attention was paid to the ventilation and tidiness of the accommodations.
Animal Treatment and Feeding
The experiment began on day 1 of gestation and ended on day 34 of gestation. During this period, the sows of control group were fed with a basal diet, and the DAI group was fed with a basal diet plus an extra 0.02% daidzein (Sichuan Jun Zheng Bio-Feed Co., Ltd., Chengdu, China). The basal diet was formulated on the basis of the nutrition needs of pregnant sows according to the National Research Council (NRC, 2012), and their compositions are shown in Supplementary Table 1. Sows were given 2.3 kg diets per day and fed twice per day (08:00 and 15:00 hours) throughout the experiment.
Sample and data collection
A random subset of sows (n = 6 per treatment) with close average body weight were selected at day 35 of gestation. The sows were then killed to obtain uteri, and the uteri were opened longitudinally along the anti-mesometrial side to obtain embryos from the attachment sites. A volume of 5 mL amniotic fluid from each embryo was harvested with sterile syringes. After all amniotic fluid samples were centrifuged at 2000 × g for 10 min at 4 °C to remove meconium, the samples were stored at − 80 °C until metabolomics and biochemical assays. Meanwhile, number of viable or mummy fetuses, average weight of viable fetuses, and size (crown-to-rump length) of viable fetuses were recorded as previously described .
Amniotic fluid reproductive hormones
Hormones including estrogen (E), progesterone (P), leptin (LEP) and insulin-like growth factor-1 (IGF-1) in the amniotic fluid were analyzed using commercially available porcine ELISA kits (Beijing Donggeboye Biological Technology Co., Ltd., Beijing, China), and the detailed operations were as per the kits’ instructions. The sensitivity and intra- and inter-assay coefficients of variation for E, P, LEP, and IGF-1 were listed in Supplementary Table 2.
Amniotic fluid immunoglobulin levels
The amniotic fluid samples from sows were assessed for immunoglobulin levels of immunoglobulin G (IgG), immunoglobulin A (IgA) and immunoglobulin M (IgM) with commercial ELISA kits (Minneapolis, MN, USA) according to the manufacturer’s instructions. A SpectraMax M2 spectrophotometer (Molecular Devices, CA, USA) was used to measure standards and samples at optical density values of 700 nm (IgG) or 340 nm (IgA, IgM). The concentration of immunoglobulin was calculated using standard curve and was expressed as µg per millilitre of amniotic fluid.
Quantification of cytokines in amniotic fluid
Concentrations of interferon γ (IFN-γ), tumour necrosis factor-α (TNF-α), interleukin-1 (IL-1), interleukin-6 (IL-6), and interleukin-10 (IL-10) were analyzed using porcine commercial ELISA kits purchased from R&D system. Assays were performed on the Immulon 2 HB 96-well plates which were coated with corresponding anti-porcine cytokines. Samples were added to the wells in a volume of 50 µl plus 50 µl of PBS–1% BSA and incubated for 2 h at room temperature (RT). The reaction was amplified with biotinylated monoclonal antibodies to porcine IFN-γ (1 µg/ml), TNF-α (250 µg/ml), IL-1 (0.1 µg/ml), IL-6 (0.2 µg/ml), and IL-10 (100 µg/ml). Plates were incubated for 1 h at RT. Detection was carried out with peroxidase-conjugated streptavidin (1:5000; Jackson Laboratories) following 60 min of incubation at RT, and the reaction was visualized with PNPP (Sigma-Aldrich). The assays were analyzed colorimetrically using a SpectraMax M2 spectrophotometer (Molecular Devices, CA, USA).
Antioxidant parameters of sows’ amniotic fluid
All antioxidant-related kits from Nanjing Jiancheng Bioengineering Institute were used to determine the antioxiant parameters according to the manufacturers’ instructions. Amniotic fluid antioxidant status was measured using a SpectraMax M2 spectrophotometer (Molecular Devices, CA, USA). Malondialdehyde (MDA) was measured using an established thiobarbituric acid (TBARS) method . MDA, a thiobarbituric acid reactive substance (TBARS), reacts with thiobarbituric acid (TBA) to form a 1: 2 MDA-TBA adduct that is absorbed at 548 nm. The concentration of MDA was calculated using standard curve and was expressed as nmol per millilitre of amniotic fluid.
The SOD activity was measured by using the procedure reported by Thomas . The SOD activity was assayed by reacting with 2-(4-iodophenyl)3-(4-nitrophenol)-5-phenyltetrazolium chloride to generate red formazan which could be spectrophotometrically determined at 550 nm. The content of SOD activity was determined as described by the manufacture’ instructions, which was expressed as U per millilitre of amniotic fluid.
Catalase (CAT) activity was measured spectrophotometrically at 620 nm using a previous described method . The method is based on the fact that the dichromate in acetic acid is reduced to chromic acetate when heated in the presence of hydrogen peroxide with the formation of perchloric acid as an unstable intermediate. CAT activity was expressed as U per millilitre of amniotic fluid, where one unit is defined as the amount that decreases 1 mmol/L H2O2 within 1 min per milliliter of amniotic fluid.
Glutathione peroxidase (GSH-Px) activity was measured using a colorimetric method described by Rotruck et al. . The enzymatic reaction was terminated by the addition of 5-50-dithiobis-(2-nitrobenzoic acid) (DTNB) (80 mg in 1% sodium citrate), which generated a light-yellow composite that could be measured at 412 nm. The GSH concentration in the experimental samples was extrapolated from the standard curve. In addition, GSH concentration was expressed as mg per millilitre.
The Total antioxidant capacity (T-AOC) activity was measured in accordance with the method of Prieto et al. . The reaction was assayed by the reduction of Fe3+- tripyridyltriazine to Fe2+- tripyridyltriazine and could be measured at 405 nm. T-AOC was expressed as U per milliliter, where one U represents the 0.01 increase in the absorbance value in 1 minute per millilitre.
Sample preparation and 1H-NMR measurement
For NMR analysis, amniotic fluid samples were left to thaw, and aliquots of 200 µL were mixed with 400 µL phosphate buffer (0.045 M NaH2PO4/K2HPO4, pH 7.4, 100% D2O). After centrifugation (12 000 × g, 10 min), an aliquot of 550 µL of each sample supernatant was subsequently transferred to 5-mm 1H-NMR tubes (Norell, Landisville, NJ, USA). The pure metabolite molecules used for referencing were all obtained from Sigma-Aldrich (St. Louis, MO, USA).
The NMR spectra of amniotic fluid samples were recorded at 298K using an Agilent DD2 600 MHz NMR spectrometer (Agilent Technologies, Inc., CA, USA). A standard 1H-NMR spectrum with water suppression using a standard NOESY pulse sequence (recycle delay − G1 − 90°−t1 − 90°−tm−G2 − 90°−acquisition). For each sample, parameters were set as follows: 128 scans were collected with a relaxation delay of 3 s; acquisition time of 1.71 s; TD of 32 k; and SW of 16 ppm. All NMR spectra were processed with a line broadening of 1 Hz and corrected for phased and baseline distortions using Topspin 3.0 (Bruker Biospin). The chemical shifts in amniotic fluid spectra were referenced to the anomeric proton signal of α-glucose at δ 5.23. Finally, the spectra-consistent 32,000 data points were normalized using the Probabilistic Quotient Normalization (PQN) method. The metabolite assignments were usually obtained by considering chemical shifts, coupling constants and relative intensities.
1 H-NMR spectroscopic analysis
For statistical analysis, reduction of 1H-NMR spectrum (9.0–0.5 ppm) was performed using MestReNova (Mestrelab Research). All NMR spectra were reduced to 1700 integral segments with an equal width of 0.004 ppm. Before data analysis, each amniotic fluid sample 1H-NMR spectrum was further removed the water (5.5–4.4 ppm) and urea (6.1–5.5 ppm) regions to eliminate the effects of variation in the suppression of the water and urea signals.
The SIMCA-P+ software (V14.0, Umetrics AB, Umea, Sweden) was facilitated for multivariate statistical analysis. For primary visualization, distribution and clustering, the principal component analysis (PCA) model was applied to reveal differences with biological significance. Following this, the partial least squares discriminant analysis (PLS-DA) and orthogonal partial least-squares discriminant analysis (OPLS-DA) were performed using a unit variance-scaled approach at a confidence level of 95%. To minimize the impact of the high variability in the metabolite measurements, unit variance scaling was utilized. R2X and R2Y, the fraction of variation that the model explains in the independent variables (X) and dependent variables (Y) and the predictive accuracy of the model (Q2Y), were estimated by the PLS-DA cross validation. The quality of the calculated model was accessed using a 200-iteration permutation test.
Potential differential metabolite selection was based on loading plot and variable importance in the projection (VIP), where only values with VIP > 1 and Bonferroni-corrected (Correlation Coefficient) P < 0.05 were considered statistically significant. The heat map and hierarchical cluster analysis (HCA) of differential metabolites were performed using the MeV software package (version 4.9.0). Finally, for identification of the most altered metabolic pathways, a set of significantly altered metabolites was used as the input for KEGG Pathway Analysis (http://www.kegg.com/).
Statistical analyses including all amniotic fluid antioxidant and immune parameters, as well as foetal data were performed using the Student’s t-test of SAS 9.0 (SAS Institute, Cary, NC, USA). Each sow was considered as a statistical unit. Data was presented as means ± standard deviations (SD). P-values less than 0.05 were considered statistically significant.