In China, porcine reproductive and respiratory syndrome (PRRS) and porcine circovirus (PVC) diseases are the most important factors affecting litter size. Our experiment was conducted to guide the pig breeding industry and so we chose sows without these two diseases. In addition, the sows we selected were of the same age, body condition and were fed at the same nutritional level. Most importantly, they had no history of reproductive diseases. We collected the litter size records of the Yorkshire ✕ Danish Landrace binary hybrid sows from the Dahua breeding farm (Guangxi Hanshiwei Food Co., Ltd., from 2016 to 2018, which was consisted of 8,557 litters), and used SPSS 25.0 to perform a significance test on these data. After normal distribution conversion and inspection, it is found that the total litter size (12.9 ± 2.17) approximately obeyed the normal distribution. The critical value of 15% right tail probability was 14.7 heads per litter, and the critical value of 15% left tail probability was 9.3 heads per litter.

The animal welfare protocol and research plan were approved by the Animal Care Committee of Fuyang Normal University. The ovaries used in our study were obtained from sows (Yorkshire ✕ Danish Landrace binary hybrid pigs) 4 days after their fourth litter delivery. These were collected from a commercial pig farm (Nanning, Guangxi, China), which was negative for PRRSV and PCV. For RNA-seq, six ovaries were removed from each group with small (8.48 ± 0.53/litter; three samples) and large (16.19 ± 0.43/litter; three samples) litter sizes. The samples were snap-frozen in liquid nitrogen.

Total RNA was extracted from ovary tissue samples by using Trizol reagent (TaKaRa, Dalian, China). Gel electrophoresis, Agilent 2,100 and NanoDrop protocols were used to detect the integrity and concentration of RNA samples. Before constructing the RNA-seq libraries, the epicenter Ribo-ZeroTM Kit (Illumina, San Diego, CA, United States) was used to remove rRNA. Briefly, total RNA was purified by polyacryl-amide gel electrophoresis (PAGE) to enrich the sRNAs with lengths of 15–35 nt, then the sRNAs were ligated with adapters and amplified by RT-PCR. The amplification products were then separated by PAGE, and the transcriptome sequencing was performed on the HiSeq 2000 platform (Illumina, California, United States).

The process of filtering clean reads from the original data is as follows: 1) discard low-quality reads; 2) adjust the adaptor sequence; 3) remove sequences smaller than 13 bp. Then, the clean data was mapped to the pig reference genome (Sscrofa 11.1). The clean reads was searched for miRbase (www.mirbase.org) through BLAST. The unmapped reads was annotated and classified through Rfam (version 14.0; http://rfam.sanger.ac.uk) and BLAST in Genbank (www.ncbi.nlm.nih.gov) database. Finally, miRNA candidate genes were predicted by MIREAP (http://sourceforge.net/projects/mireap/) using the remaining unannotated siRNA sequences. Sequences located in the pig genome and folded into a typical hairpin structure with neighboring sequences were considered as potential new miRNAs. The new miRNAs were compared with the mature miRNAs of other mammals by BLAST.

To find the differentially expressed miRNAs between SLS and LLS samples, DESeq2 and Cuffdiff methods were performed using R statistical software’ Bioconductor. To find genes that were differentially expressed between SLS and LLS pigs, the identified miRNAs with a fold change of 2 or greater and P-adjusted < 0.05 were considered to be differentially expressed.

Gene ontology (GO) analysis (http://www.geneontology.org) was used to explore potential functions of miRNA. This was divided into three different aspects: biological process, cell composition and molecular function of the miRNA, which not only reduced the complexity but also highlighted the molecular function. Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis (http://www.kegg.jp) was used to detect the participation of host genes in biological pathways.

Based on the binding sites of miRNAs detected in mRNA sequences, the RNAhybrid (https://bibiserv.cebitec.uni-bielefeld.de/rnahybrid/) and MiRanda (http://cbio.mskcc.org/microrna_data/miRanda-aug2010.tar.gz) software were used to predict the interactions of miRNAs and mRNAs.

cDNA Synthesis, and Real-time qPCR.

Total RNA was reversely transcribed using a Hiscript III Reverse Transcriptase Kit (Vazyme, Nanjing, China). Real-time quantitative PCR (qPCR) was performed in triplicate using qPCR SYBR Green Master Mix (Vazyme, Nanjing, China) on a Bio-Rad CF96 system (Bio-Rad, United States), and U6 and GAPDH were used for normalization of the data.

The isolation steps of pig ovarian GCs were performed as follows. The generic ovarian slaughterhouse tissues were digested with collagenase (Sigma, United States) in a 37°C water bath for about 1 hour. Then, the digested tissues were centrifuged at a low speed, and the supernatant was discarded, and the remaining tissue was filtered through a 200-mesh filter. The filtrate was washed three times with PBS, and finally the cells were resuspended with DMEM/F12 containing 1% penicillin/streptomycin. Then, the cells were seeded at a certain density on a cell culture plate at 37°C in an atmosphere of 5% CO₂. In order to detect the transfection efficiency of the miR-138 mimic, miR-183 agomir (RiboBio, Guangzhou, China) was transfected into ovarian GCs when they were at 50% confluence using Lipofectamine 2000 (Invitrogen, United States).

For measurement of the estradiol (E2) and progesterone (P4) levels in the culture supernatants, ELISA kits (Solarbio, Beijing, China) were used according to the manufacturer’s instructions. The inter- and intra-assay coefficients of variation were < 15 and < 10%, respectively for each kit. Each sample was measured in triplicate.

The proliferation state of ovarian GCs was investigated with Cell Counting Kit-8 (CCK-8) (Tiandz, Beijing, China) and Cell-Light™ EdU Apollo^® 567 In Vitro Imaging Kit (RiboBio, Guangzhou, China), respectively. The manufacturer’s instructions were followed for all procedures.

Proteins were extracted from cell samples by using RIPA lysate buffer containing 1% PMSF (Solarbio, Beijing, China). The BCA Kit (Solarbio, Beijing, China) was used to measure the concentration of proteins in solution. The proteins were separated by 10% SDS- polyacrylamide gel (Bio-Rad, United States) electrophoresis and transferred to 0.22 μm nitrocellulose membranes (PALL, United States). After the membranes were incubated with primary antibodies specific for anti-cyclin D (cat. no. ab16663; dilution ratio = 1/200), anti-Bcl-2 (cat. no. ab692; dilution ratio = 1/200), anti-cyclin E (cat. no. ab33911; dilution ratio = 1/1000), anti-P27 (cat. no. ab32034; dilution ratio = 1/5000), anti-GAPDH (cat. no. ab8245; dilution ratio = 1/1000), and β-actin (cat. no. ab6276; dilution ratio = 1/5000) from Abcam (Cambridge, MA, United States) and secondary antibodies (dilution ratio = 1/5000; Solarbio, Beijing, China), the membranes were exposed with ECL Plus (Solarbio, Beijing, China). The images obtained were captured with the ChemiDoc XRS + system (Bio-Rad, United States).

GCs from pig ovaries were treated with miR-183 agomir in 6-well plates. After 24 h, the cells were treated with the Cell Cycle Assay Kit (Multisciences, Hangzhou, China), washed with PBS buffer and centrifuged to remove the supernatants. Then permeabilization and DNA stain solutions were added to the re-suspended cells and mixed by vortex shaking. After incubation in the dark for 30 min, the cells were analyzed by Flow Cytometry (FACS Canto™ II, BD BioSciences, United States).

Results are shown as means ± standard error of the means (SEM) for at least three independent experiments. SPSS 25.0 software (Chicago, United States) was used to perform a significance test on these data. p-values were calculated by Student’s t-test. p < 0.05 was considered as statistically significant for RNA expression, protein expression, CCK-8 and EdU assays and measurement of E2 and P4.

In order to discover the role that miRNAs play in the determination of pig litter size, six sow ovary samples with complete fertility records were selected to perform high-throughput sequencing. After sequencing, Fast-QC (http://www.bioinformatics.babraham.ac.uk) was used to make an overall evaluation of the original data. As shown in Figure 1A, the Q30 values of the LLS pigs in the tested samples were: 92.56, 92.40 and 82.21%, and the SLS pigs were 90.21, 92.46 and 78.43% (Figure 1A). It was worth mentioning that the Q30 values of two samples were low (only about 80%), but the mapped reads and unique mapped reads were not less than other samples, indicating that the data quality of this sequencing could be used for subsequent research (Table 1). The results of 64.5 million (M) and 61.6 M total reads were obtained from the ovarian libraries of SLS and LLS pigs, respectively. After removing the low quality and adaptor sequences, a total of 44.2 and 43.8 M clean reads, respectively, were obtained (Figure 1B; Table 1). The length of miRNA-seq reads ranged from 20 to 24 nucleotides; the analysis revealed that a majority of reads were 21 and 23 nt (Supplementary Figure S1).

A total of 411 miRNAs were identified, and 17 miRNAs were significantly down-regulated and 16 miRNAs were up-regulated when comparing LLS pigs to SLS pigs, respectively (Figures 2A,B, Supplementary Table S1). It can be seen that miR-183-5p is one of the most significant differential expressions. Its expression level in the LLS group was 1,512 (TPM) compared to 522 in the SLS group (TPM, Supplementary Table S2). It can also be seen that miR-183-5p is conserved in more than 35 species including humans, mice and pigs ( Figure 2C, Figure 2D).

To explore the possible biological functions of miRNAs identified in SLS and LLS porcine ovarian tissues, GO analysis was performed (Supplementary Figure S2). The results showed that these miRNAs were significantly enriched by 20 biological processes, including steroid biosynthesis process, lipid metabolism process, ovarian follicle development, granular cell differentiatio. In order to better understand the molecular regulation of DEGs, we performed KEGG pathway analysis (Supplementary Figure S3). The hippo signaling pathway was the most significant (path: 04390). From the three enrichment pathways of steroid biosynthesis (path: 00100), ovarian steroid production (path: 04913) and steroid hormone biosynthesis (path: 00140), it could be seen that the synthesis and metabolism of steroids play an important role in the ovaries of pigs.

We confirmed the differences in the abundance of certain miRNAs when comparing SLS and LLS tissue samples by using qPCR. We randomly selected 10 differentially expressed miRNAs and amplified these using qPCR primers (Figure 3). Our qPCR analysis found the presumed patterns of up- and down-regulation of these 10 miRNAs to be remarkably similar in both types of analysis (Figure 3), suggesting that miRNA-Seq data provided reliable information about the relative abundance of miRNAs. Among these miRNAs, miR-183 is conserved in 38 species and has high expression level in the ovaries of LLS sows. This suggests that it may play a potential role in regulating porcine litter size.

In order to understand the effect of miR-183 on the function of ovarian GCs, the expression changes of key functional genes in GCs after overexpression of miR-183 were examined. After transfection of miR-183 agomir (a mimic), in porcine GCs, the overexpression efficiency was measured by qPCR. The cells were collected after transfection with three different concentrations of agomir for 24 h. The results showed that the concentrations of agomir at 10, 20 and 25 nM were able to stably overexpress miR-183, while the transfection concentration of 20 nM could significantly reduce the synthesis of E2 and promote the synthesis of P4. Therefore, for the next round of functional verification, we chose a transfection concentration of 20 nM to overexpress miR-183 in GCs (Figure 4).

After treatment with 20 nM miR-183 agomir for 24 h, the GCs were collected and RNA was extracted, and measured by qPCR. It was found that the expression of hormone synthesis related genes Star, Cyp11a1 and CYP19 were decreased significantly (p < 0.05), but the expression levels of 3β-HSD were markedly increased (p < 0.0001; Figure 5).

To further clarify the effect of miR-183 on the proliferation and apoptosis of ovarian GCs, miR-183 was overexpressed in these cells. The effect of miR-183 on the proliferation ability of GCs was assessed by EdU staining. Compared with the control group, miR-183 agomir significantly increased the number of EdU GCs positive cells (Figures 6A,B). Similarly, miR-183 could significantly increase cell viability as detected with the CCK-8 kit (p < 0.05, Figure 6C). In order to detect the effect of miR-183 on apoptosis, we treated GCs with 20 nM miR-183 agomir for 24 h. After the GCs were collected, the protein expression levels of the anti-apoptotic protein, Bcl-2, was measured by using Western blot. It was found that there was a tendency to decrease the expression levels between control and treatment groups (p < 0.05, Figure 6D). Moreover, miR-183 significantly increased the number of cells in the S phase and this was shown by flow cytometry (Figure 7A). It was also found that miR-183 overexpression increased the expression of cyclins D and E, and decreased the expression of p27 at the protein level (Figure 7B).