Growth hormone receptor controls adipogenic differentiation of chicken bone marrow mesenchymal stem cells by affecting mitochondrial function

Growth hormone receptor (GHR), a member of the type I cytokine receptor family, activates several downstream signaling pathways after binding to growth hormone (GH) to regulate cell growth and development. However, in poultry, the relations among the GHR gene, mitochondrial function, and adipocyte differentiation remain unclear. In this study, we used sex-linked dwarf (SLD) chickens and normal chickens as research objects and overexpression or knockdown of GHR in bone mesenchymal stem cells (BMSCs) at the stage of adipogenic differentiation, to investigate the effects of the GHR gene on mitochondrial biogenesis, mitochondrial function and adipogenic differentiation of BMSCs.

In SLD chickens, liver mitochondrial function declines compared with that in normal chickens [22]. Therefore, SLD and normal chickens were compared in this study, and overexpression and knockdown of GHR in BMSCs were used to determine the effects of GHR on mitochondrial function and adipogenic differentiation of chicken BMSCs.

Ethics statement
All animal experiments were performed according to the protocols approved by the South China Agriculture University Institutional Animal Care and Use Committee (approval number SCAU#0017). All animal procedures followed the regulations and guidelines established by this committee and minimized the suffering of animals. The 21-d-old yellow-feather chickens included 15 SLD chickens and 15 normal chickens. To explore molecular mechanisms of GHR in vivo and determine the cause of fatty deposits in SLD chickens, mitochondrial function, mitochondrial biogenesis, and adipogenic differentiation in chicken BMSCs were examined in those chickens. The 3-d-old normal chickens were only used to isolate BMSCs, and the BMSCs were used to study GHR effects on mitochondria and adipogenic differentiation in vitro.

Para n sections and hematoxylin and eosin staining
The epiphyses of thighbone from the 21-d-old SLD and normal chickens were xed with 10% neutral formalin for 5 d and then immersed in a hydrochloric acid/formic acid working solution to complete decalci cation. After decalci cation, samples were dehydrated in alcohol and transformed into a transparent state using xylene. After completing the transparency step, samples were soaked in wax and embedded in para n. A para n sectioning machine cut 7 to 10-μm-thick sections, which were stained with hematoxylin and eosin.

Frozen sections and oil red staining
Epiphyseal parts of femurs of the 21-d-old SLD and normal chickens were cut off and soaked in 4% paraformaldehyde for 48 h and then switched to decalci cation solution for 30 d, with the solution changed every two days. After decalci cation, tissues were placed in a 15% sucrose solution in a refrigerator at 4 °C to dehydrate and sink and then were transferred to a 30% sucrose solution at 4 °C to dehydrate and sink. Dehydrated tissue was placed cut side up on a sample tray and surrounded by drops of OCT embedding agent (Servicebio, Wuhan, China). The tray was placed on the quick-freeze table of a frozen sectioning machine, and the samples were sectioned after the OCT whitened and hardened. Section thickness was 8 to 10 μm. After sectioning, tissue sections were xed on a slide and stained with oil red and hematoxylin.

Detection of triglyceride
Triglyceride was measured using a Triglyceride Assay Kit (Nanjing Jiancheng, Nanjing, China) according to the manufacturer's protocol. Triglyceride was measured at 510 nm, and absorbance was determined using a Fluorescence/Multi-Detection Microplate Reader (Bio-Tek, Winooski, USA) according to the manufacturer's protocol. Data were normalized to the control group and expressed as a percentage of the control.
Reverse-transcription quantitative PCR RNA was extracted from tissues or cells using RNAiso reagent (Takara, Shiga, Japan) according to the manufacturer's protocol. Concentration of RNA samples and optical density (OD) value of 260/280 were detected using a Nanodrop 2000c spectrophotometer (Thermo, Waltham, USA). Samples were stored at −80 °C for later use. For reverse-transcription quantitative PCR (RT-qPCR), cDNA was synthesized using MonScript™RTIII All-in-One Mix with dsDNase (Monad Co., Ltd., Guangzhou, China). ChamQ Universal SYBR qPCR Master Mix (Vazyme, Guangzhou, China) was used in RT-qPCRs run on a Bio-Rad CFX96 Real-Time Detection instrument (Bio-Rad, Hercules, USA) according to the manufacturer's protocol. The reaction procedure included initial denaturation at 95 °C for 3 min, followed by denaturation at 95 °C for 10 s and annealing at 60 °C for 30 s, for a total of 40 cycles. At the end of the cycle, the dissolution curve was analyzed, and the detection temperature was 65 °C to 95 °C. Relative gene expression was measured using RT-qPCR twice for each reaction, and β-actin was used as the control. The primers used in RT-qPCR are listed in Table 1.

Extraction of chicken bone mesenchymal stem cells and cell culture
Bone mesenchymal stem cells were extracted using the appropriate separation kits (TBD science, Tianjin, China) following the manufacturer's protocol.
Bone mesenchymal stem cells from 21-d-old SLD and normal chickens were extracted by cell separation kits and cultured in vitro to the appropriate density (the rst generation). Then, assays were conducted on mitochondrial function and related gene expression and protein levels.
Bone mesenchymal stem cells from 3-d-old normal chickens were extracted by cell separation kits and were cultured in vitro and passaged to the third generation. Overexpression and knockdown of GHR were to explore the effects of GHR on mitochondrial and adipogenic differentiation in chicken BMSCs.
Plasmid construction, small interfering RNA, and transfection Third generation BMSCs were plated onto 6-well plates, and transfection began when the density reached approximately 80%. After 6 h of transfection, the DMEM/F12 medium was changed to adipogenic induction medium to induce adipogenic differentiation of BMSCs.
Guangzhou RiboBio (Guangzhou, China) synthesized small interfering RNAs (siRNA) used for GHR knockdown. In preliminary experiments, four siRNAs were designed to interfere with GHR, and the si-GHR with the highest interference e ciency was used. The siRNA sequence is provided in Table 2. The si-GHR sequence was transfected in BMSCs to a nal concentration of 100 nM using Lipofectamine 3000 reagent (Invitrogen, USA) according to the manufacturer's protocol. Cells were analyzed at 72 h after transfection.

Detection of reactive oxygen species
Production of ROS in mitochondria was measured using an ROS assay kit (Beyotime, Shanghai, China) according to the manufacturer's protocol. Dichloro uorescein (DCF) uorescence was determined using a Fluorescence/Multi-Detection Microplate Reader (Bio-Tek). Data were normalized to the control group and are expressed as a percentage of the control.
Detection of ATP content ATP levels were measured using an ATP assay kit (Beyotime) according to the manufacturer's protocol. A Fluorescence/Multi-Detection Microplate Reader (BioTek) was used to determine ATP levels. Data were normalized to the control group and are expressed as a percentage of the control.

Detection of mitochondrial membrane potential
Mitochondrial membrane potential (ΔΨm) was measured using a JC-1 kit (Beyotime) according to the manufacturer's protocol. Mitochondria were xed with JC-1, and after cells were incubated with JC-1 for 20 min at 37 °C, uorescence was determined using a Fluorescence/Multi-Detection Microplate Reader (Bio-Tek). Rotenone, 10 μmol/L, was used as a standard inhibitor of ΔΨm. Data (the ratio of aggregated and monomeric JC-1) were normalized to the control group and are expressed as a percentage of the control.

Detection of enzymatic activity of mitochondrial oxidative phosphorylation complexes
Commercial assay kits (Solarbio, Beijing, China) were used to measure enzyme activity of mitochondrial oxidative phosphorylation (OXPHOS) complexes in BMSCs according to the manufacturer's protocol. Complex I enzyme activity was determined by the change in absorbance of NADH at 340 nm. Complex II enzyme activity was determined by the change in absorbance of 2,6-dichlorophenol indophenol at 600 nm. Enzyme activity of complex III and complex IV was determined by the change in absorbance of reduced cytochrome c at 550 nm. Absorbance was determined using a Fluorescence/Multi-Detection Microplate Reader (Bio-Tek). Data were normalized to the control group and are expressed as a percentage of the control.

Mito-tracker green staining and Hoechst 33342 staining
Mito-tracker green staining and Hoechst 33342 staining were used to label mitochondria and nuclei in BMSCs, respectively. At 72 h after transfection, cells were washed twice with phosphate buffered saline (PBS) and incubated with Mito-tracker green (Beyotime) for 30 min. Cells were then suspended in PBS, and 10 µL of Hoechst 33342 dye was added (Beyotime). After washing twice with PBS, a uorescence microscope (Nikon TE2000-U, Tokyo, Japan) was used to capture ve randomly selected elds that were analyzed with NIS-Elements software.

Oil red O staining and quanti cation
Bone mesenchymal stem cells were seeded into 6-well culture plates. After transfection and differentiation for 5 d, differentiated BMSCs were washed with PBS and then xed with 4% formaldehyde for 30 min. Differentiated BMSCs were dyed with oil red O working solution (BBI, Shanghai, China) for 60 min at room temperature and then washed three times with PBS, according to the manufacturer's speci cation. After washing, a uorescence inverted light microscope (Leica DMi8, Wetzlar, Germany) was used to capture images. At the end, stain in cells was extracted by isopropanol and absorbance was measured at 510 nm with a Fluorescence/Multi-Detection Microplate Reader (Bio-Tek).

Western blot analysis
Radio-immune precipitation assay buffer (Beyotime) with phenylmethane sulfonyl uoride protease inhibitor (Beyotime) was used to lyse tissue and cellular proteins. The homogenate was centrifuged at 13,000 ×g for 10 min at 4 °C. The supernatant was collected, and protein concentration was determined immediately using a bicinchoninic acid assay protein quanti cation kit (Beyotime). Proteins were separated in 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis, transferred onto a polyvinylidene di uoride membrane, and then probed with antibodies following standard procedures.

Statistical analyses
All experiments were performed at least three times. Data are presented as the mean ± standard error of the mean (SEM). Statistical analyses were performed using Student's t-test, with statistical signi cance indicated as *P < 0.05, **P < 0.01, and ***P < 0.001.

Results
Fat deposition and differentiation of bone mesenchymal stem cells in bone marrow tissue of sex-linked dwarf and normal chickens Hematoxylin and eosin staining of bone marrow tissues from 21-d-old SLD and normal chickens showed the percentage of fat in bone marrow tissues of SLD chickens was signi cantly higher than that of normal chickens (Fig. 1a-c). Oil red O staining on bone marrow tissues of SLD and normal chickens showed the percentage of lipid droplets in bone marrow tissues of SLD chickens was higher than that of normal chickens (Fig. 1d-f). Furthermore, triglyceride content in the bone marrow tissue of SLD chickens was highly signi cantly higher than that of normal chickens (Fig. 1g). To investigate whether the cause of this phenotype in SLD chickens was due to differentiation of BMSCs, expression of marker genes of adipogenic differentiation in BMSCs of SLD and normal chickens was examined. Expression of the marker genes PPARγ, C/EBPα, and C/EBPβ increased signi cantly in SLD chickens (Fig. 1h). The results were similar in protein level assays (Fig. 1i). Thus, fat deposition in the bone marrow tissue of SLD chickens was much more severe than that of normal chickens. This result might be due to differentiation of BMSCs in SLD chickens.

Mitochondrial function in bone mesenchymal stem cells of sex-linked dwarf and normal chickens
Mitochondria can synthesize ATP through oxidative phosphorylation to provide a major source of energy for adipogenic differentiation of MSCs [23]. Therefore, we speculated that mitochondria have an important role in the production of lipids in bone marrow tissue. The mRNA levels of mtDNA-encoded OXPHOS-related and mitochondrial biogenesis-related genes in BMSCs of SLD and normal chickens were measured by RT-qPCR. Genes included ND1, ND2, CYTB, COX1, COX2, ATP6, ATP8, PGC1α, NRF1, and TFAM. Protein levels of PGC1α, NRF1, and TOMM20 in BMSCs of SLD and normal chickens were measured by western blot. The mRNA expression of genes was elevated signi cantly in BMSCs of SLD chickens (Fig. 2a, b). Protein levels of PGC1α, NRF1, and TOMM20 also increased in SLD chickens ( Fig. 2c). Furthermore, to indicate mitochondrial function, ΔΨm, ROS production, and ATP content were examined. Compared with normal chickens, ΔΨm decreased and ROS production and ATP content increased in SLD chickens ( Fig. 2d-f). Therefore, in SLD chickens, changes in mitochondrial function might affect adipogenic differentiation of BMSCs and thus adipogenesis.
Effects of overexpression and knockdown of growth hormone receptor on mitochondrial function in chicken bone mesenchymal stem cells To investigate the effect of GHR on mitochondrial function, BMSCs were isolated and cultured in vitro, and then, GHR was overexpressed or knocked down. E ciency of GHR overexpression and knockdown e ciency in chicken BMSCs differentiated for 72 h was examined by RT-qPCR. Compared with the control group, GHR expression was signi cantly up-regulated after transfection with the vector of GHR (Fig. 3a) and signi cantly down-regulated after transfection with si-GHR (Fig. 3b). Similar to in vivo experiments, mitochondrial function was examined after overexpression and knockdown of GHR in differentiated BMSCs. After overexpression of GHR in BMSCs, mRNA expression of mtDNA-encoded OXPHOS-related and mitochondrial biogenesis-related genes decreased (Fig. 3c, e). Protein levels of PGC1α, NRF1, and TOMM20 showed similar results (Fig. 3g). In addition, ΔΨm increased, and ROS production and ATP content decreased (Fig. 3i, k, m). Opposite results were observed after knockdown of GHR (Fig. 3d, f, h, j, l,  n). These results suggested that GHR repressed mitochondrial function during adipogenic differentiation in chicken BMSCs.

Effects of overexpression and knockdown of growth hormone receptor on enzymatic activity of OXPHOS complexes in chicken bone mesenchymal stem cells
To further investigate the effects of GHR on mitochondrial function in chicken BMSCs during adipogenic differentiation, enzyme activity of mitochondrial OXPHOS complexes was examined in differentiated BMSCs after overexpression and knockdown of GHR. Enzymatic activities of OXPHOS complexes I, II, III, and IV decreased signi cantly in chicken BMSCs after overexpression of GHR (Fig. 4a-d), whereas after GHR knockdown, enzymatic activities increased signi cantly (Fig. 4e-h). Overall, the results indicated that GHR repressed mitochondrial function by suppressing enzymatic activity of OXPHOS complexes in chicken BMSCs during adipogenic differentiation.

Effects of overexpression and knockdown of growth hormone receptor on mitochondrial number and quality in chicken bone mesenchymal stem cells
Mitochondrial biogenesis increases the number of mitochondria to meet intracellular energy requirements. Therefore, the effect of GHR on mitochondrial function was explored by assaying the number of mitochondria. Mito-tracker staining was used to label mitochondria, with uorescence intensity representing mitochondrial quantity. After overexpression of GHR in BMSCs, uorescence intensity weakened, and the number of mitochondria decreased (Fig. 5a, c). After knockdown of GHR, uorescence intensity strengthened, and the number of mitochondria increased (Fig. 5b, d). Thus, in addition to repressing mitochondrial function, GHR reduced mitochondrial number and quality by repressing mitochondrial biogenesis during adipogenic differentiation of chicken BMSCs.

Effects of overexpression and knockdown of growth hormone receptor on adipogenic differentiation of chicken bone mesenchymal stem cells
Finally, GHR was overexpressed and knocked down in chicken BMSCs to investigate the effects of GHR on adipogenic differentiation. Adipogenic differentiation was induced in BMSCs, and expression of associated genes was detected by RT-qPCR. The genes were PPARγ, C/EBPα, and C/EBPβ. Protein levels of PPARγ and C/EBPα in chicken BMSCs differentiated for 72 h were measured simultaneously by western blot. Expression of adipogenic differentiation-related genes was signi cantly down-regulated after overexpression of GHR (Fig. 6a), whereas after knockdown of GHR, expression was signi cantly upregulated (Fig. 6c). Protein levels of PPARγ and C/EBPα in chicken BMSCs showed similar results (Fig. 6b, d). Furthermore, the oil red O test was used to measure lipid droplet content in chicken BMSCs differentiated for 5 d after overexpression and knockdown or GHR. Overexpression of GHR depressed the lipid droplet depot in BMSCs, whereas knockdown had the opposite effect ( Fig. 6e-h). In addition, overexpression of GHR repressed triglyceride production in BMSCs (Fig. 6i), whereas knockdown of GHR (Fig. 6j) produced opposite results. Thus, GHR repressed fat deposition in chickens by inhibiting adipogenic differentiation of chicken BMSCs.

Discussion
Since their discovery in 1940, research on SLD chickens has been uninterrupted. Mutation of the GHR gene in SLD chickens interferes with binding of GH to GHR [24], and therefore, SLD chickens are a speci c animal model for mutation of the GHR gene [25]. Fat deposition is more severe in SLD chickens than in normal chickens [4]. In a previous study, compared with normal chickens, red bone marrow was severely depleted and replaced by yellow bone marrow in 7-week-old SLD chickens [26]. It was hypothesized that the SLD phenotype was due to a functional deletion of the GHR gene. Therefore, in this study, the relation between GHR and adipogenic differentiation of BMSCs was explored.
In this study, fat deposition in bone marrow tissue of 21-d-old SLD chickens was greater than that in normal chickens, consistent with previous ndings [26]. In addition, triglyceride content of bone marrow tissue in SLD chickens was twice as high as that in normal chickens, consistent with fat deposition. Fat in bone marrow tissue is primarily derived from adipogenic differentiation of MSCs [27]. The balance between adipose and bone tissues in bone marrow tissue is maintained primarily by two types of MSC differentiation: adipogenic and osteogenic [28]. When adipogenic differentiation of MSCs increases, osteogenic differentiation is relatively weakened, resulting in fat deposition [29]. Therefore, it was hypothesized that the more severe fat deposition in SLD bone marrow tissue was due to increased adipogenic differentiation capacity of BMSCs in SLD chickens because of de ciency in normal GHR gene function. To test the hypothesis, BMSCs were extracted from SLD and normal chickens, and differences in expression of genes associated with adipogenic differentiation in the two groups of cells were examined. Expression of PPARγ, C/EBPα, and C/EBPβ in BMSCs was signi cantly higher in SLD chickens than in normal chickens. PPARγ, as the predominant transcription factor in adipocyte differentiation, also plays an important role in adipogenic differentiation of BMSCs. In one study, addition of the PPARγ agonist rosiglitazone activated adipogenic differentiation of mouse BMSCs [30]. During adipogenic differentiation of MSCs, C/EBPβ, PPARγ, and C/EBPα are sequentially activated [31].
An increasing number of studies show that regulation of mitochondrial dynamics and function is critical for successful differentiation of MSCs. Adipogenic differentiation of MSCs is accompanied by changes in the mitochondrial phenotype, including increased mitochondrial biogenesis and abundance of OXPHOS complexes [32]. Therefore, mitochondrial function of BMSCs from SLD and normal chickens was examined by using ATP content, ROS, and ΔΨm assays. The ATP content was higher in the BMSCs of SLD chickens, indicating that de ciency in GHR function led to an increase in mitochondrial oxidative phosphorylation capacity and therefore production of ATP. Those changes provided the necessary conditions for adipogenic differentiation of BMSCs. Reactive oxygen species are produced by the OXPHOS pathway associated with energy production in mitochondria [33]. Only unregulated levels of ROS are harmful, whereas regulated ROS production is needed for essential signaling pathways that regulate cell functions [34]. Production of ROS by mitochondrial complex III is required to activate adipogenic differentiation of MSCs, and ROS levels increase during adipogenesis induction in MSCs [19]. In this study, ROS production was greater in BMSCs of SLD chickens than in those of normal chickens.
Reactive oxygen species promote lipid accumulation in human adipose stromal cells undergoing adipogenesis [35]. Therefore, increases in ROS may be one factor that stimulates differentiation of BMSCs toward adipogenesis. Notably, ΔΨm decreased in BMSCs of SLD chickens compared with that in normal chickens. In general, the higher ΔΨm is, the greater the energy capacity of the inner mitochondrial membrane and the higher the amount of ATP synthesis [36]. However, maintaining excessively high or excessively low ΔΨm can be harmful to mitochondria and cells [37,38]. A decrease in ΔΨm is presumed to be due to the effect of "mild uncoupling of mitochondria", which ensures the supply of ATP while appropriately lowering the ΔΨm in response to damage caused by elevated ROS [39]. In addition, expression of mitochondrial genes encoding OXPHOS, including ND1, ND2, CYTB, COX1, COX2, ATP6, and ATP8, and genes related to mitochondrial biogenesis, including PGC1α, NRF1, and TFAM, were upregulated in BMSCs of SLD chickens compared with those in normal chickens. Levels of the mitochondrial membrane protein TOMM20 and expression of genes associated with mitochondrial biogenesis (PGC1α, NRF1) also increased. The results indicated that increases in mitochondrial function in BMSCs of SLD chickens were due to the absence of GHR function. Therefore, changes in mitochondria in BMSCs of SLD chickens may affect adipogenic differentiation of BMSCs and ultimately increase fat deposition in SLD chickens.
To further con rm the conjecture on the mechanism of fat deposition in bone marrow tissue of SLD chickens, BMSCs were extracted from chickens for cellular veri cation. Expression of mitochondrial genes encoding OXPHOS and genes associated with mitochondrial biogenesis was examined after overexpression and knockdown of GHR in BMSCs. The results were consistent with those obtained in vivo. When GHR was knocked down in chickens BMSCs, ΔΨm decreased, ROS production and ATP content increased, and protein levels of PGC1α, NRF1, and TOMM20 was enhanced. The opposite result after overexpression of GHR. Collectively, the results suggest that GHR inhibits mitochondrial function during adipogenic differentiation in chicken BMSCs. This conclusion is also supported by enzymatic activity of complexes I, II, III, and IV after overexpression and knockdown of GHR. Enzymatic activity of complexes I, II, III, and IV was enhanced after knockdown of GHR. The opposite result after overexpression of GHR. Complexes I, II, III, and IV are important components of the mitochondrial electron transport chain and are involved in the adipogenic differentiation of BMSCs through mitochondrial oxidative phosphorylation. In one study, inhibition of the mitochondrial electron transport chain suppressed adipogenic differentiation of MSCs [23]. Mitochondrial biogenesis is regulation of the number of mitochondria through mitochondrial self-renewal in response to energy demands triggered by developmental signals and environmental stressors [40]. Mitochondrial biogenesis increases during adipogenic differentiation of MSCs [23]. In immortalized human MSCs, overexpression of PGC-1α increases mitochondrial function and biogenesis and promotes adipogenic differentiation of MSCs [41]. TFAM can bind to the mitochondrial light strand promoter and functions in mitochondrial transcription regulation [41], and knockdown of TFAM in MSCs inhibits adipogenic differentiation [23]. Furthermore, Mito-tracker staining validated the effect of GHR on mitochondrial biogenesis. The number of mitochondria decreased after overexpression of GHR, indicating that GHR inhibited mitochondrial biogenesis. The opposite result was observed after knockdown of GHR.
In addition, whether GHR inhibited adipogenic differentiation of chicken BMSCs in vitro was investigated. After overexpression of GHR in chicken BMSCs, expression of differentiation-related genes, including PPARγ, C/EBPα, and C/EBPβ, was repressed and lipid droplet production and triglyceride levels decreased. Notably, PPARγ determines the direction of adipogenic differentiation of MSCs [42]. With knockdown of GHR, opposite results were obtained. Thus, GHR can inhibit adipogenic differentiation of chicken BMSCs.

Conclusions
Sex-linked dwarf chickens had severe fat deposition in bone marrow tissue than normal chickens. Increased adipogenic differentiation of BMSCs in SLD chickens was associated with increases in mitochondrial biogenesis and function and expression of genes related to differentiation. After overexpression of GHR in chicken BMSCs, mitochondrial function and adipogenic differentiation of BMSCs were repressed. The opposite results were observed after knockdown of GHR. Therefore, GHR inhibits excessive adipogenic differentiation of chicken BMSCs by repressing mitochondrial function.
TTCAGGGTCAGGATACCTCTTT Table 2 Oligonucleotide sequence in this study. BMSCs; e and f expression of genes related to mitochondrial biogenesis in BMSCs; g and h protein levels of PGC1α, NRF1, and TOMM20; i and j mitochondrial membrane potential in BMSCs; k and l reactive oxygen species (ROS) production in BMSCs; m and n the concentration of ATP in BMSCs. Data are expressed as the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001