A total of 12 female dwarf type Xinghua chickens (the SLD chicken, with GHR gene exon 5 a T/C mutation) were randomly distributed to the HFD group, which was fed a diet consisting of 40% carbohydrate, 25% fat, and 20% protein, and the normal diet group, which was fed a diet consisting of 41% carbohydrate, 5% fat, and 22% protein (n = 6 in each group). Those 8-week-old SLD chickens with similar body weight were fed for 2 weeks and weighed before slaughter. In addition, another 12 purebred female Xinghua chickens (normal chicken) were treated in the same way as the SLD chicken. Determination on abdominal fat weight and abdominal fat rate was carried out as described previously (Chen et al., 2019). The tissues of liver and abdominal fat were immediately frozen in liquid nitrogen and stored at −80°C until use. All animal work was approved and performed in accordance with the regulations and guidelines of the Institutional Animal Care and Use Committee of the South China Agricultural University (approval number: SCAU#0017; 21 November 2017).

Abdominal fat was fixed with 4% paraformaldehyde and embedded in paraffin, and 12-μm-thick serial sections were made. Then, sections were performed H&E staining according to standard protocols. Microscopic observation and photograph were taken with Leica DM2500 microscope (Leica, Wetzlar, Germany). At least five visual fields were randomly selected for each section, and at least 20 cells were selected for each visual field. The area and diameter of adipocyte were calculated using Nikon Eclipse Ti microscope and NIS-Elements BR software (Nikon, Tokyo, Japan).

Suppressor of cytokine signaling 2 (SOCS2) and cytokine-inducible SH2-containing protein (CISH) coding sequences were amplified from chicken liver cDNA by PCR using specific primers (Supplementary File 1). The full-length coding regions of chicken SOCS2 and CISH were constructed into pcDNA3.1 vector (Invitrogen). Specific small interfering RNAs of SOCS2 and CISH were obtained from RiboBio, and non-specific siRNA was used as the control.

Liver-related cell line leghorn strain M chicken hepatoma (LMH) cell and immortalized chicken preadipocytes (ICP) were cultured in high-glucose Dulbecco’s modified Eagle’s medium (Gibco, Carlsbad, CA, United States) with 10% fetal bovine serum and 0.2% penicillin/streptomycin at the condition of 37°C with 5% CO₂. ICP cells were induced to differentiation by adding 160 μM sodium oleate (Sigma Life Science, St. Louis, MO, United States). Plasmid DNA and siRNA transfection were carried out using the transfection reagent Lipofectamine 3000 (Invitrogen Corporation, Carlsbad, CA, United States) following the manufacturer’s protocol.

RNA from tissues or cells was isolated using RNAiso reagent (Takara, Otsu, Japan). Reverse-transcription reactions were using PrimeScript RT reagent Kit with gDNA Eraser (Takara). The relative mRNA expression levels of genes were determined by using real-time quantitative PCR (qPCR) with SYBR Green. The specific primers are shown in Supplementary File 1. The 2^–ΔΔCt method was used to calculate gene expression with β-actin (for liver tissues and LMH cells) or GAPDH (for abdominal fats and ICP cells).

Three livers and abdominal fats from each of the four treatment groups were selected for RNA sequencing. RNA quantity and quality were evaluated on an Agilent 2100 Bioanalyzer (Agilent Technologies, Waldbronn, Germany), and RNA integrity was further examined using agarose gel electrophoresis. High-throughput RNA-seq was performed on the BGISEQ-500 platform (BGI, Wuhan, China). Significance was accepted at adjusted | log2FC| ≥ 0.5, P ≤ 0.001. All the sequence data have been deposited in NCBI’s Gene Expression Omnibus (GEO¹) and are accessible through GEO series accession numbers GSE129840 and GSE128340.

After 48 h of transfection, Oil Red O staining was carried out in ICP cells. Briefly, cells were washed with PBS for 5 min and then fixed with 4% paraformaldehyde for 10 min. The cells were stained with Oil Red O working solution (Solarbio, Beijing, China). The cells were observed and photographed using a microscope (Leica). The Oil Red O dyes were isolated using isopropanol solution containing 4% Non-idet P-40. Concentration was quantified by Model 680 Microplate Reader (Bio-rad, CA, United States) at 510 nm.

The Venn diagram was calculated and drawn by a web-based software². Gene ontology (GO) analysis of the enriched genes was performed using the web-based Metascape (a gene annotation and analysis resource³ (Zhou et al., 2019). Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis of differentially expressed genes (DEGs) was carried out for functional analyses⁴ (Kanehisa, 2000; Kanehisa et al., 2017).

All data are reported as mean ± SEM. The difference between two groups was evaluated using independent sample t-test. The differences among multiple groups were evaluated using Duncan’s multiple range test. P < 0.05 was considered to be statistically significant. All experiments were carried out at least three times. All statistical analyses were performed using SPSS 18.0 for Windows (SPSS, Inc., Chicago, IL, United States).

To investigate the difference of abdominal fat deposition between the SLD and normal chickens and test the effect of HFD on these two kinds of chicken breeds, we fed the SLD chickens and normal chickens with HFD and normal diet. After 2 weeks of feeding, HFD fed SLD chicken (HD), HFD fed normal chicken (HN), normal diet fed SLD chicken (ND), and normal diet fed normal chicken (NN) were weighed and conducted carcass determination. Results showed that the body weight of the SLD chickens was significantly lower than that of normal chickens under the same diet, whereas the body weight of either chicken breeds was not significantly affected by HFD (Figure 1A). However, HFD can significantly increase the abdominal fat weight and abdominal fat rate of chickens, and the abdominal fat weight and abdominal fat rate of the SLD chicken were significantly higher than those of normal chicken under the same diet (Figures 1B,C). Next, we analyzed the adipocytes size between chickens. The HFD can significantly increase the adipocyte size of the SLD chicken and normal chicken (Figures 1D–F). Compared with normal chicken, larger adipocytes area was observed in the SLD chicken in both HFD and normal diet (Figures 1G,H). These results demonstrated that the SLD chickens have more abdominal fat deposition and larger adipocytes size than normal chickens, and that HFD can increase abdominal fat deposition and enlarge adipocytes size in chicken.

Considering that the liver is the main organ of lipid metabolism and plays an essential role in the digestion, absorption, synthesis, decomposition, and transportation of lipids, we collected the livers of HD, HN, ND, and NN chickens for RNA sequencing to find DEGs and signal pathways involved in GHR mutation-induced difference of lipid metabolism. A total of 1,040 genes were upregulated, and 610 genes were downregulated in ND chicken compared with NN chicken (Figure 2A and Supplementary File 2), whereas a total of 1,255 genes were upregulated, and 1,218 genes were downregulated in HD chicken compared with HN chicken (Figure 2B and Supplementary File 3). GO Enrichment Analysis showed that the DEGs between the SLD chicken and normal chicken were mainly related to the metabolic process and developmental process (Figures 2C,D). KEGG pathway analysis showed that the enriched pathways of DEGs between ND and NN chickens included many lipid metabolism and adipose deposition pathways, such as metabolic pathways, linoleic acid metabolism, fatty acid degradation, PPAR signaling pathway, steroid hormone biosynthesis, fatty acid biosynthesis, fatty acid metabolism, and steroid biosynthesis (Figure 2E). On the other hand, the DEGs between HD and HN chickens were also enriched in the pathways involved in lipid metabolism and adipocyte development, such as PPAR signaling pathway, fatty acid degradation, steroid hormone biosynthesis, calcium signaling pathway, and extracellular matrix (ECM)-receptor interaction (Figure 2F). Furthermore, we identified 737 genes differentially expressed between the two kinds of chicken breeds not only under the HFD but also under the normal diet (Figure 2G and Supplementary File 4). The functions of these 737 DEGs were largely related to lipid metabolism, such as lipid catabolic process and lipid modification (Figure 2H). A total of 911 genes were identified specific differentially expressed between the SLD chicken and normal chicken under normal diet, and the functions of these DEGs were mainly related to cell cycle and lipid biosynthetic process (Figure 2I). A total of 1,734 DEGs were specific differentially expressed between the SLD chicken and normal chicken under HFD, and their functions were mainly enriched in metabolic process and lipid transport (Figure 2J). Therefore, these results not only found the DEGs of the liver between the SLD chicken and normal chicken but also obtained the potential signaling pathways and cellular processes involved in the GHR mutation-induced liver lipid metabolism difference between the SLD chicken and normal chicken.

In order to find DEGs involved in abdominal fat deposition and adipogenesis, we collected abdominal fat from HD, HN, ND, and NN chickens for RNA sequencing. A total of 536 genes were upregulated, and 1,044 genes were downregulated in fat between ND chicken and NN chicken (Figure 3A and Supplementary File 5). Under an HFD, a total of 622 genes were upregulated expression, and 643 genes were downregulated expression (Figure 3B and Supplementary File 6). GO analysis of DEGs between the SLD chicken and normal chicken under a normal diet found that these DEGs were related to metabolism process, developmental process, and growth (Figure 3C). DEGs under an HFD also exhibited similar biological process enrichment (Figure 3D). KEGG analysis of DEGs between ND and NN chickens showed that the mainly enriched pathways included metabolic pathway, cell adhesion molecules (CAMs), and fat digestion and absorption (Figure 3E), whereas calcium signaling pathway, ECM-receptor interaction, type I diabetes mellitus, and CAMs were included in the enriched KEGG pathways of DEGs between HD and HN chickens (Figure 3F). Furthermore, we identified 541 genes differentially expressed between the two kinds of chicken breeds not only under the HFD but also under the normal diet (Figure 3G and Supplementary File 7). The functions of these 541 DEGs were largely related to hormone metabolism and growth, such as sterol metabolic process, regulation of IGF transport, steroid biosynthesis, triglyceride metabolic process, and regulation of growth (Figure 3H). A total of 1,037 genes were identified specific differentially expressed between the two kinds of chicken breeds under the normal diet, and the functions of these DEGs were mainly involved in the cell proliferation and growth processes (Figure 3I). A total of 763 DEGs were specific differentially expressed between the two kinds of chicken breeds under HFD, and the functions of these DEGs were involved in signal transduction, development, and lipid transport (Figure 3J). Together, these results not only found the DEGs of the abdominal fat between the SLD chicken and normal chicken but also obtained the potential signaling pathways and cellular processes involved in the GHR mutation-induced fat deposition and adipogenesis.

To further analyze the different effects of HFD on the SLD chicken and normal chicken, we screened out the common and specific DEGs between the SLD chicken and normal chicken. In the liver, we found a total of 191 common DEGs (Figure 4A and Supplementary File 8), which mean that HFD induces the differential expression of these genes not only in the SLD chickens but also in normal chickens. The functions of these common DEGs were involved in transport of small molecules, hemostasis, and cell cycle (Figure 4B). We also found that 606 DEGs were specific differentially expressed between HD and ND chickens (Figure 4A), and that the functions of these DEGs were related to metabolic processes and hepatocyte differentiation (Figure 4C). Besides, a total of 880 DEGs were specific differentially expressed between HN and NN chickens, and these DEGs were mainly enriched in cell cycle process (Figure 4D).

In abdominal fat, a total of 163 genes were not only differentially expressed between HN and NN chickens but also differentially expressed between HD and ND chickens (Figure 4E and Supplementary File 9). The functions of these common DEGs were related to IGF transport and lipid metabolism (Figure 4F). Additionally, the functions of the 600 DEGs specific differentially expressed between HN and NN fats were mainly enriched in cell cycle, such as cell division, mitotic, regulation of cell cycle process, regulation of chromosome segregation, positive regulation of cell cycle, and meiotic cell cycle process (Figure 4G). The functions of the 730 DEGs, which were specific differentially expressed between HD and ND fats, were related to multiple metabolic processes and homeostasis (Figure 4H).

Next, we compared the common and specific DEGs between different treatment groups in the liver and in the abdominal fat, respectively (Figures 4I,J). We found that a total of 66 DEGs were common differentially expressed in the liver among the four treatment groups (Figure 4I and Supplementary File 10), and that 48 DEGs were common differentially expressed in the abdominal fat among the four treatment groups (Figure 4J and Supplementary File 10). These common DEGs may play important roles in chicken abdominal fat deposition. To test the quality and accuracy of RNA-seq results, we selected several common DEGs, which are related to lipid metabolism and adipogenesis, in the liver and in the abdominal fat, respectively (Figures 4K,L). Then, we used qPCR to validate the expression of DEGs among the four treatment groups. Results showed that the qPCR data were consistent with RNA-seq data (Figures 4M,N), demonstrating that the RNA-seq results were highly reliable.

Growth hormone binds to GHR and activates a variety of downstream signaling pathway. In order to better understand how GHR mutation affects downstream signaling transduction, we constructed a GHR-mediated downstream genes interaction network, which includes JAK–STAT, PI3K–AKT–mTOR, FOXO signaling, GH–GHR–IGFs, and MAPK signaling pathways (Figure 5). Next, we used this network to map the DEGs between the SLD chicken and normal chicken. The DEGs between ND liver and NN liver were enriched in PI3K–Akt–mTOR signaling pathway, which regulated lipid biosynthesis, lipolysis, and cell autophagy (Figure 5A). Additionally, the DEGs between HD liver and HN liver were enriched in JAK–STAT, AKT, and FOXO signaling pathways. Notably, the function of these DEGs was mainly related to cell cycle regulation (Figure 5B). On the other hand, the DEGs between ND and NN abdominal fats were enriched in FOXO signaling pathway, and the functions of these DEGs were related to glycolysis, cell cycle, lipid biosynthesis, and lipid metabolism (Figure 5C). The DEGs between HD and HN abdominal fats were enriched in JAK–STAT signaling pathway, and the functions of these DEGs were related to lipogenesis, glycolysis, and cell cycle (Figure 5D). Notably, SOCS2 and CISH were the only two genes that were differentially expressed in all of the four treatment groups, indicating their important roles on lipid metabolism and GHR mutation-induced fat deposition.

In order to understand the roles of SOCS2 and CISH on GHR mutation-induced abdominal fat deposition, we tested their functions in lipid metabolism and adipocyte differentiation in vitro. The expression of SOCS2 and CISH was significantly downregulated in the liver and abdominal fat of dwarf chicken compared with normal chicken, no matter in an HFD or normal diet condition (Figures 6A–D). Next, we used LMH cell and ICP cell to test the function of SOCS2 and CISH in vitro. To assess the function of SOCS2 and CISH in lipid metabolism and fat deposition and find the potential downstream genes that can be affected by SOCS2 and CISH, we selected several genes that were not only involved in lipid metabolism and fat deposition but also differentially expressed between the SLD chicken and normal chicken as candidate genes to define the status of lipid metabolism and fat deposition. In LMH cell, SOCS2 overexpression significantly inhibited the expression of most of the lipid metabolism-related DEGs, whereas CISH overexpression can only inhibit the expression of CIDEC and FASN (Figure 6E). On the other hand, SOCS2 knockdown significantly promoted the expression of most of the lipid metabolism-related DEGs, whereas CISH knockdown can only promote the expression of CIDEC and FASN (Figure 6F). In ICP cell, SOCS2 overexpression significantly inhibited the expression of fat deposition-related DEGs, whereas CISH overexpression can only inhibit the expression of DGAT2 and ApoB (Figure 6G). The knockdown of SOCS2 promoted the expression of most of the genes involved in fat deposition, whereas CISH knockdown only promoted ApoA4 and ApoB expression (Figure 6H). To investigated whether SOCS2 and CISH can affect lipid droplet accumulation during ICP differentiation, we transfected the overexpression vectors or siRNA of these two genes to the ICP and then induced the cells to differentiation for 48 h. As judged by Oil Red O staining and extraction assays, SOCS2 overexpression significantly reduced lipid droplet accumulation (Figure 6I), whereas SOCS2 inhibition significantly increased lipid droplet accumulation (Figure 6J). However, the overexpression and inhibition of CISH have no significant impacts on lipid droplet accumulation (Figures 6I,J). Taken together, these data indicate that SOCS2 inhibits lipid metabolism and decreases lipid droplet accumulation.

SOCS2 and CISH were both downregulated in the liver and fat of the SLD chicken compared with normal chicken (Figures 6A–D). In order to explore the important roles of SOCS2 and CISH on GHR mutation-induced fat deposition, we used primary liver cell and preadipocyte isolated from the SLD chicken and normal chicken, respectively, to test the function of SOCS2 and CISH on lipid metabolism and lipid droplet accumulation. In accordance with the RNA-seq data, the expression of SOCS2, CISH, and DEGs involved in lipid metabolism and fat deposition was dysregulated in primary cells from the SLD chicken (Figures 7A,B). Overexpression of SOCS2 or CISH alone in the cells from the SLD chicken can partly rescue the dysregulated expression of genes involved in lipid metabolism and fat deposition, whereas co-overexpression of SOCS2 and CISH can rescue most of the dysregulated expression of these genes (Figures 7A,B). Importantly, primary preadipocyte isolated from the SLD chicken can accumulate more lipid droplet than that from the normal chicken (Figure 7C). Overexpression of SOCS2 can significantly decrease lipid droplet accumulation in preadipocyte from the SLD chicken, whereas no significant alteration was observed after transfection of CISH relative to control. However, co-transfection of SOCS2 and CISH decreased more lipid droplet accumulation than individual transfection of SOCS2, suggesting a complementary role between SOCS2 and CISH (Figure 7C). Taken together, these results indicate that co-overexpression of SOCS2 and CISH can make the lipid droplet accumulation of the SLD chicken return to the level of normal chicken, suggesting that SOCS2 and CISH are important in GHR mutation-induced fat deposition in the SLD chicken.