All experimental procedures were conducted in accordance with institutional guidelines for the care and use of laboratory animals in China [16]. This study is a companion study, and the details of animals, experimental design and detailed procedures have been presented previously [14]. Briefly, Eighteen Mongolian ewes carrying singletons were allocated to three groups at d 90 of gestation according to the metabolizable energy (ME): restricted group 1 (RG1, 0.18 MJ ME/body weight [BW]^0.75/d, n = 6), restricted group 2 (RG2, 0.33 MJ ME/BW^0.75/d, n = 6) and control group (CG, ad libitum, 0.67 MJ ME/BW^0.75/d, n = 6). The ewes in their second or third parity were mated with rams at a synchronized estrus, after treatment for 12 days with CIDRs (EAZI-BREED CIDR, each containing 0.3 g progesterone in inert silicone elastomer, Pfizer Australia Pty Ltd., New Zealand) and an injection of pregnant mare serum gonadotropin (400 IU). Pregnancies were confirmed by ultrasound scanning at approximately d 50 of gestation (Medison-SA-600; Medison CO. LTD, Seoul, Korea), and had similar live weights (mean 52.82±2.67 kg) at d 90 of gestation. All animals were housed in individual pens, and chopped hay (about 10 cm, mainly Leymus chinensis) was supplied during late pregnancy. Since the fetus is considered to achieve 80% to 85% of its final birth weight during the last two months of gestation [6,17], maternal undernutrition was imposed from d 90 to d 140 of pregnancy. At the beginning of restriction, the ME and chemical composition in the hay were measured (Table 1), and then the daily intake of the hay offered in the RG1 and RG2 groups was calculated by the ewe body weight, nutrition value of hay, and the designed energy plane in the restricted groups. Restricted ewes were fed at 08:30 and 16:00 h each day, and the amount of feed offered was constant throughout the restriction period (Table 2). The ewes in the CG group were offered feed at 08:30, 11:00 and 16:00 h daily (the feed refusals were approximately 10% of the total amount offered). The animals had free access to water and mineral mixture blocks (containing per kilogram: Ca, 15 g; P, 11.5 g; Mg as MgSO₄·H₂O, 1 g; Fe as FeSO₄·7H₂O, 500 mg; Cu as CuSO₄·5H₂O, 250 mg; Zn as ZnSO₄, 175 mg; Mn as MnSO₄, 100 mg; Co as CoCl₂·6H₂O, 20 mg; I as KI, 40 mg; Se as Na₂SeO₃·5H₂O, 1.5 mg; Yuantongweiye Co., Ltd., Inner Mongolian, China). All feed refusals were collected daily before feeding at 08:30, weighed and sub-sampled for chemical analysis.

Detailed slaughter procedures were described previously [14,18,19]. Briefly, all fetuses were removed at 140 d of gestation, and fetal BW was recorded. Some of the thymic tissue was snap-frozen in liquid nitrogen and stored at −80°C. After rinsing with phosphate-buffered saline (PBS, pH 7.4), portions of the fetal thymus were immediately placed in paraformaldehyde fixative solution (0.1 mol/L, pH 7.4).

After fixation for at least 2 days, the tissues were dehydrated and paraffin-embedded, sectioned at 4 to 6 μm and stained with immunohistochemistry commercial kits of proliferating cell nuclear antigen (PCNA, KGA320; KeyGEN Biotech, Nanjing, China) for microscopic examination. The paraffin-embedded tissue sections were dewaxed, rehydrated, and treated with antigen retrieval for 20 min in a microwave-oven. Sections were immersed in 3% hydrogen peroxide solution for 10 min to quench endogenous peroxidase activity. Non-specific binding was prevented by incubation with 10% normal goat serum for 10 min. The sections were incubated with mouse monoclonal anti-PCNA antibody for 1 h at 37°C and secondary antibody for 10 min at 37°C, then were visualized with diaminobenzidine (DAB) solution, lightly counterstained with hematoxylin. Each specimen was viewed under a standard microscope, and the total number of PCNA positive cells was counted with Image-Pro Express 6 software, in 6 random high-power fields (magnification×400), respectively. Data were expressed as the number of PCNA positive cells per field.

Portions of the freshly collected fetal thymus were washed with PBS (pH 7.4) and minced, then passed through a stainless-steel mesh (300 mesh). The thymic cell suspensions were centrifuged at 2,000 rpm per min for 10 min. Suspend approx. 1×10⁶ cells in 0.5 mL of PBS, and fix cells by adding 4.5 mL of pre-chilled cold 70% ethanol stored at 4°C for at least 2 h. Centrifuge 1 mL ethanol-suspended cells for 5 min at 1,200 rpm per min and decant ethanol thoroughly. After washing with PBS, the cell was suspended in 1 mL of propidium iodide staining solution (containing 0.5% propidium iodide, 0.25% TritonX-100 and 10 mg/mL Rnase; 4ABIO, Beijing, China) and incubated at room temperature for 30 min [20]. After washing and resuspending, the cells were analyzed on a flow cytometer (FACSCalibur; Becton Dickinson, Franklin Lakes, NJ, USA) and the G1-, S-, G2- and M-phase could be recognized in a proliferating cell population. Data were analyzed with CellQuest software (Becton Dickinson, USA). The proliferation index was calculated as the percentages of S, G2 and M phase cells occupying the different phases of the cell cycle.

Growth hormone (GH) and insulin-like growth factor 2 (IGF-2) were measured by commercial kits (NJJCBIO, Nanjing, China). According to the manufacturer’s recommendations, approximately 0.5 g of the fetal thymus was rinsed and homogenized in 0.85% chilled normal saline to obtain a 10% thymus homogenate, then 50 μL aliquots of the supernatants were added to the microtiter plates coated by sheep antibodies of GH and IGF-2 labeled with horse radish peroxidase and incubated at 37°C for 30 min to become antibody-antigen-enzyme-antibody complex. After washing with PBS twice, 50 μL staining solution of tetramethyl benzidine was added. The reaction is terminated 15min later by a sulphuric acid solution, then determined at 450 nm in Microplate Reader (ELX800; BIO-TEKINSTUMENTS, Winooski, VT, USA).

Total RNA from thymus was isolated using RNA extraction kits (TianGen, Beijing, China), and reverse-transcribed with a first-strand cDNA synthesis kit according to the manufacturer’s instructions (TransGen Biotech Co., Ltd, China). Using cDNA synthesized from 500 ng of total RNA as a template for one amplification, real-time reverse transcriptase (RT)-polymerase chain reaction (PCR) was performed using TransStart Green qPCR mix (TransGen Biotech Co., Ltd, China) in a Bio-Rad iQ 5 Real Time PCR System according to the instructions provided. The thermal cycling parameters were as follows: 95°C for 30 s for initial denaturation then 40 cycles of 95°C for 5 s, 60°C for 30 s and 72°C for 30 s. The sequences of gene-specific primers for real-time RT-PCR are listed in Table 3. All samples were measured in triplicate, and the real-time PCR assay had similar efficiency and within the range of 90% to 110%. The relative expression ratio of mRNA was calculated by R = 2−ΔΔCt [21]. Threshold cycle (Ct) was the value of PCR cycles at which the fluorescence signal of the PCR reaction reached a fixed threshold. For each sample, the Ct both for the target gene and endogenous control gene were determined to calculate ΔCt sample (Ct target gene – Ct endogenous control). In this study, the β-actin gene was used as an endogenous control. Subsequently, ΔΔCt (ΔCt sample – ΔCt CG) was determined, and the relative expression was calculated by 2−ΔΔCt. For the CG group, the ΔΔCt is equal to 0. Negative controls were performed in which cDNA was substituted by water.

A 10% thymus homogenate was used for the measurement of E-cadherin (E-cad), OB-cadherin, and vimentin (VIM) in fetal thymuses according to the procedures of commercial kits (NJJCBIO, China). In the kits, purified sheep antibodies of E-cad, OB-cadherin, and VIM labeled with horse radish peroxidase are used to coat microtiter plate wells.

All data were analyzed by using the analysis of variance procedure as implemented in SAS software [22]. Duncan’s test was used to identify significant differences between mean values. Significance was declared at p≤0.05.

The cell cycle, proliferation index and expression of PCNA in ovine fetal thymuses are summarized in Figure 1. With the reduction of maternal energy intake, the G0/G1 phase cell number in fetal thymus of the RG1 group was increased compared with the controls (p<0.05), but reduced PCNA expression and proliferation index were found in the RG1 fetal thymuses compared with the CG group (p<0.05). For the S phase and G2/M phase cell number in fetal thymuses, there were no differences between the restricted groups and the CG group (p>0.05).

GH, IGF-2, and their receptors expressions in fetal thymuses are presented in Table 4. Fetuses in the RG1 group exhibited decreased thymic GH and IGF-2 (p<0.05), and the mRNA expressions of thymic growth hormone receptor (GHR) and IGF-2R were lower than those of the CG group (p<0.05). For the RG2 group, GH and GHR mRNA expression in fetal thymuses were decreased relative to the controls (p<0.05).

The mRNA expressions of cell cycle regulators in fetal thymuses are shown in Figure 2. Reduced thymic mRNA expressions of CDK1, CDK2, CDK4, CCNA, CCNB, CCND, and CCNE in the RG1 group were found compared with the controls (p<0.05). For the RG2 group, the mRNA expressions of CDK1, CDK2, CDK4, and CCNE were lower than those of the CG group (p<0.05). Thymic mRNA expressions of E2F1, E2F2, and E2F5 were reduced in the RG1 and RG2 groups (p<0.05) and reduced thymic E2F4 expression was found in the RG1 group (p = 0.0037).

The markers expressions of epithelial-mesenchymal transition (EMT) in fetal thymuses are presented in Figure 3. The decreased expressions of E-cad in the RG1 and RG2 groups were found relative to the controls (p<0.05), and the expression of OB-cadherin in the RG1 group was higher than that of the CG group (p<0.05). For the VIM expressions in fetal thymuses, there were no differences among the RG1, RG2 and CG groups (p>0.05).