The differentiation of preadipocytes and gene expression related to adipogenesis in ducks (Anas platyrhynchos)

Meat quality is closely related to adipose tissues in ducks, and adipogenesis is controlled by a complex network of transcription factors tightly acting at different stages of differentiation especially in ducks. The aim of this study was to establish the preadipocyte in vitro culture system and understand the biological characteristics of expansion of duck adipocyte tissue at the cellular and molecular level. We isolated pre-adipocytes from the subcutaneous fat of three breeds of duck and differentiated them into mature adipocytes using a mixture of insulin, rosiglitazone, dexamethasone, 3-isobutyl-1-methylxanthine, and oleic acid over 0,2, 4, 6, and 8 days. Successful differentiation was confirmed from the development of lipid droplets and their response to Oil Red O, and increasing numbers of lipid droplets were stained red over time. The expression of key marker genes, including peroxisome proliferator activated receptor γ (PPARγ), CCAAT/enhancer binding protein-α (C/EBPα), adipocyte fatty acid binding protein 4 (FABP4), and fatty acid synthetase (FAS), gradually increased during pre-adipocyte differentiation. Furthermore, it was verified by interference experiments that the knockdown of PPARγ directly reduced lipid production. Meanwhile we analyzed the role of unsaturated fatty acids in the production of poultry fat using different concentrations of oleic acid and found that lipid droplet deposition was highest when the concentration of oleic acid was 300 μM. We also compared the level of differentiated pre-adipocytes that were isolated from Jianchang ducks (fatty-meat duck), Cherry Valley ducks (lean-meat duck) and White-crested ducks (egg-producing duck). The proliferation and differentiation rate of pre-adipocytes derived from Jianchang ducks was higher than that of White-crested ducks. These results provide the foundation for further research into waterfowl adipogenesis.

Meat quality is closely related to adipose tissues in ducks, and adipogenesis is controlled by a complex network of transcription factors tightly acting at different stages of differentiation especially in ducks. The aim of this study was to establish the preadipocyte in vitro culture system and understand the biological characteristics of expansion of duck adipocyte tissue at the cellular and molecular level. We isolated pre-adipocytes from the subcutaneous fat of three breeds of duck and differentiated them into mature adipocytes using a mixture of insulin, rosiglitazone, dexamethasone, 3-isobutyl-1-methylxanthine, and oleic acid over 0,2, 4, 6, and 8 days. Successful differentiation was confirmed from the development of lipid droplets and their response to Oil Red O, and increasing numbers of lipid droplets were stained red over time. The expression of key marker genes, including peroxisome proliferator activated receptor γ (PPARγ), CCAAT/enhancer binding protein-α (C/EBPα), adipocyte fatty acid binding protein 4 (FABP4), and fatty acid synthetase (FAS), gradually increased during pre-adipocyte differentiation. Furthermore, it was verified by interference experiments that the knockdown of PPARγ directly reduced lipid production. Meanwhile we analyzed the role of unsaturated fatty acids in the production of poultry fat using different concentrations of oleic acid and found that lipid droplet deposition was highest when the concentration of oleic acid was 300 μM. We also compared the level of differentiated preadipocytes that were isolated from Jianchang ducks (fatty-meat duck), Cherry Valley ducks (lean-meat duck) and White-crested ducks (egg-producing duck). The proliferation and differentiation rate of pre-adipocytes derived from Jianchang ducks was higher than that of White-crested ducks. These results provide the foundation for further research into waterfowl adipogenesis. PLOS

Introduction
Anatine adipose tissue is directly associated with the production and meat quality characteristics. It is composed of a large number of clustered adipocytes. And the adipocytes aggregated into clusters are divided into small lobes by thin layers of loose connective tissue. All known adipocytes derive from pluripotent mesenchymal stem cells (MSC) in a process called adipogenesis [1][2][3]. The proliferation and differentiation of pre-adipocytes leads to the formation of adipose tissue. [4].
Previous studies had shown that the differentiation and adipogenesis in preadipocytes was a very complex process involving a large amount of gene expression.Understanding the molecular mechanism of adipogenesis can help to improve the quality of meat, increase the utilization rate of feed, and contribute to the feed conversion ratio (FCR). Transcriptional factors including CCAAT/enhancer-binding protein (C/EBP) family including C/EBPα, C/EBPβ, and C/EBPδ [5,6], and peroxisome proliferator-activated receptor (PPAR) [7,8], as well as other genes, such as fatty acid-binding protein (FABP4) [9], lipoprotein-lipase (LPL), and fatty acid synthase (FAS) [10] both played important roles in the formation of fat. DLK1 (delta drosophila homolog-like 1 or delta-like 1 homologue) first discovered in the neuroblastoma could inhibit the differentiation of pre-adipocytes and was known as pre-adipocyte factor-1 (pref-1) [11].
Unlike mammals, adipose tissue in poultry cannot synthesize fatty acids, so that oleic acid is added to the culture medium of dexamethasone, insulin, and 3-isobutyl-1-methylxanthine (IBMX) as an exogenous fatty acid to induce pre-adipocyte differentiation. Oleic acid is both an induction factor and a triglyceride synthesis material [12], and can significantly stimulate the expression of several marker genes related to adipocyte differentiation [13]. Ding et al. [14] showed that a 12 h treatment of oleic acid significantly up-regulated the expression of FAS preadipocytes in duck.
Adipose tissue is a complex, necessary, highly active metabolic and endocrine organ [15] that could store energy as lipids [16]. Adipose tissue could serve as a crucial integrator of energy balance and glucose homeostasis in humans [17]. However, the unlimited expansion of adipose tissue resulted in obesity, which significantly contributes towards type 2 diabetes, hyperlipidemia, hyperglycemia, insulin resistance, chronic inflammation, and atherosclerosis and so on [18,19]. Therefore, the study of adipose tissue is of great importance for health.
In this study, we demonstrate changes in the expression profiles of adipogenesis-related genes during adipocyte differentiation in three breeds of duck. And we also investigated the role of oleic acid, genetic background, and PPARγ in adipogenesis in ducks. These results provide a foundation for further research into the mechanisms of duck adipogenesis.

Primary culture of pre-adipocytes from subcutaneous adipose tissue
Pre-adipocytes from Cherry Valley duck (lean-meat duck), White-crested duck (egg-laying duck), and Jianchang duck (fatty-meat duck) were isolated as described below [3]. Cherry Valley duck and Jianchang duck were provided by Jiangsu Yike Food Co., Ltd, and White-crested duck was provided by Waterfowl Institute of Zhenjiang City. Briefly, we separated subcutaneous adipose tissue from 20-day-old embryos using sterile scissors and tweezers and placed it into vial bottles. Adipose tissues were minced into 1-mm 3 sections using scissors and digested with 2 mg/mL type I collagenase in PBS at 37˚C in a shaking water bath for 90 min. The resulting mixture was filtered through 0.300 mm, 0.0750 mm, and 0.0374 mm sieve screens to remove undigested tissue and large cell aggregates after blending with 10% FBS in DMEM. The filtered suspensions were then centrifuged at 1000 × g for 10 min at room temperature to separate floating adipocytes from the pre-adipocytes. The precipitate was resuspended using Red Blood Cell Lysis Buffer for 15 min at room temperature. The pre-adipocytes were seeded into culture flasks and cultured in a humidified atmosphere of 5% CO 2 and 95% air at 37˚C after being centrifuged at 1000 × g for 10 min. The characteristics of pre-adipocytes were observed using a microscope (Olympus, Tokyo, Japan).

Cell culture and differentiation
Pre-adipocyte were digested using 0.25% Trypsin-EDTA for 1 min after the pre-adipocytes typically formed a large number of colonies and reached 100% confluence at approximately 24 h. Then the pre-adipocyte pellet was washed with PBS, centrifuged at 1000 × g for 5 min, and resuspended with 10% FBS in DMEM, and seeded into 6-well plates and 96-well plates at a density of 1 × 10 4 cells/cm 2 with the basic medium. The basic medium was changed to the differentiation medium after achieving 90% cell confluence with 0.5 mmol/L IBMX, 1 μm/L DEX, 10 μg/μL insulin, 1 μM Rosiglitazone, 300 μM oleic acid added to the medium to induce pre-adipocyte differentiation. The medium was replaced every second day. The differentiation medium was replaced with the maintenance medium of 10 μg/μL insulin only after 4 days and cells were incubated for a further 4 days.

Cell proliferation assays
Pre-adipocytes and differentiated cell proliferation was measured with the Cell Counting Kit-8 (CCK-8, Dojindo Corp, Japan) according to the instructions. Approximately 2,000 cells were transferred into a 96-well plate per well, to which 10 μL CCK-8 reagent was added to 90 μL of the culture medium. The cells were subsequently incubated for 4 h at 37˚C when the medium became yellow at 1, 2, 3, 4, 5, and 6 days culturing. Absorbance was measured at 450 nm using a microplate reader to establish the growth curve of cells (n = 5).

Oil Red O staining
The lipid droplets in pre-adipocytes were stained using an Oil Red O staining kit (Solarbio, China) according to the manufacturer instructions. Briefly, cells were washed using PBS and fixed using 4% formaldehyde for 30 min. Then, the fixed cells were washed using 60% isopropanol for 30 sec and stained with Oil Red O solution (solutions A and B). After 20 min, the cells were rinsed using 60% isopropanol to remove the residual Oil Red O and repeatedly washed using distilled water. Stained cells were observed using a microscope.

Triglyceride content assay
Triglyceride (TG) content was analyzed using a triacylglycerol detection kit (Jiancheng, China) according to the manufacturer protocols. Briefly, the differentiated pre-adipocytes were washed using PBS and sonicated to ensure they were ruptured. A BCA protein assay kit (Beyotime, Beijing) was used to measure the protein concentration of ruptured cells. The TG content was expressed in mmol/g protein.

Cloning the full-length coding sequence of PPARγ
Total RNA was extracted from Cherry Valley ducks using the RNAsimple Total RNA kit (Tiangen, Beijing, China) according to the manufacturer instructions. First-strand cDNA was synthesized using 1 μg RNA and a PrimeScript II 1st Strand cDNA synthesis Kit (TaKaRa, Dalian, China) according to the manufacturer protocols. Primers were designed using the BGI duck PPARγ sequence (accession: NW_004676594, Table 1). The amplified products were verified by 1.5% agarose gel electrophoresis and purified with a gel extraction kit (Omega, USA). Then, the purified products were ligated into the pMD19-T Vector (TaKaRa, Dalian, China), transferred into E. coli DH5α cells, and sequenced by Sangon (Shanghai, China).

SiRNA knockdown and transfection
Briefly, pre-adipocytes were trypsinized and resuspended in the basal culture medium, then seeded into 24-well plates and transfected with siRNA (GenePharma, Shanghai, China) to target duck PPARγ using Lipofectamine 2000 RNAi MAX (Life Technologies) with a serum-free medium at a final concentration of 50 nM. The interference sequence is designed based on the cloned PPARγ full-length sequence. A non-sense siRNA (NC) was used as a negative control for non-sequence specific effects. Two days after reaching confluency, cells were induced using the complete induction medium for another 2 days. The sequences of PPARγ siRNA were shown in Table 1. Silencing of PPARγ was monitored using qRT-PCR.

Real-time PCR analysis
Total RNA was extracted from cells using TRIzol reagent and the complementary DNA was synthesized from 1 μg of total RNA using reverse transcriptase (Tiangen, China). Quantitative real-time polymerase chain reaction (qRT-PCR) was performed using a Thermo  Table 1.

Statistical analysis
Data were presented as the mean ± standard deviation (SD). Between-group differences were determined using Student's t test. P < 0.05 was considered statistically significant. All statistical analyses were carried out using SPSS (version 17). All experiments were performed in triplicate.

Identification of pre-adipocytes and dynamic observations during differentiation
Morphological characteristics of adipogenesis in cultured duck pre-adipocytes during differentiation are shown in Fig 1. Pre-adipocytes obtained after collagenase digestion that were adhered to the bottom of the flasks reached 100% confluency after 24 h in the basal culture medium (Fig 1A). Morphology was irregular, like a spindle and triangle. The Oil Red O stain showed no TG formation at this time. After 2 days in the complete induction medium, there were a few small lipid droplets like bright spots in the cells (Fig 1B), and the Oil Red O staining showed some formation of TG. At 4 to 6 days after induction, the number of lipid droplets increased and their volume increased from small to large, some lipid droplets were fused, and there were lipid droplets and exfoliated cells floating in the culture medium (Fig 1C and 1D). At this time, most of the cells were stained red. At 8 days after induction, pre-adipocytes had developed into oval or round mature adipocytes. Lipid droplets became larger and filled the cells, and the nucleus was located at one side instead of in the center (Fig 1E). Almost all the cells were stained red. Actually, the lipid droplets formed a large accumulation of lipid droplet in cells. However, adipocytes, especially those differentiated from primary pre-adipocytes, detached and floated away from the bottom of the flask easily at this stage. Pre-adipocytes showed an "S-type" growth curve in line with primary cell growth laws ( Fig  1F). Cells entered the logarithmic phase during days 2 to 3. Thereafter, pre-adipocytes were in a slow, proliferative state (plateau phase) during which their activities remained stable. Preadipocytes essentially stopped proliferating when they started to differentiate into adipocytes. During the differentiation process, the TG content steadily increased (Fig 1G). The content of TG was significantly higher at 10 days after induction compared with that in non-induced preadipocytes.

Expression of genes related to adipogenesis during differentiation
The expression profiles of six genes related to adipogenesis during pre-adipocyte differentiation in ducks are shown in Fig 2. All the genes (DLK1, LPL, C/EBPα, PPAR-γ, FAS, and FABP4) were expressed in the cells during pre-adipocyte differentiation and had different expression patterns. The expression of DLK1 was the highest in pre-adipocytes and decreased gradually during the induction process. The expression of DLK1 reached its lowest point at 10 days after induction compared with pre-adipocytes (I0) significantly (P <0.01). Both C/EBPα and FAS showed an increased expression pattern during adipogenesis until 8 days after induction, and then decreased at 10 days. FABP4 increased throughout the differentiation process until day 10 and had the highest level of expression (the highest fold-change was 41.7 at I10). LPL showed an increase-decrease expression pattern during adipogenesis and increased to its highest level on day 10, which was significantly higher than that on other culture days (P <0.01). PPARγ maintained a steady growth rate, except for a clear low peak on day 8 (P <0.05).

The role of oleic acid in adipogenesis
We used induction formulas with different concentrations of oleic acid to explore its role in the adipogenesis of pre-adipocytes under 48 h induction. Morphological observations showed little or no lipid droplet formation without oleic acid ( Fig 3A). As the concentration of oleic acid increased from 100 μM to 600 μM, the amount of fat deposition also increased (Fig 3B,  3C and 3D). The fatty deposition was at its highest when the oleic acid concentration was 300 μM (Fig 3C). There was no significant increase or decrease in lipid droplets at a concentration of 600 μM oleic acid ( Fig 3D) compared with that at 300 μM. Quantitative measurement of TG content were used to evaluate the effect of different oleic acid concentrations on adipocyte adipogenesis. The results were basically consistent with morphological observations ( Fig  3E). When oleic acid concentration was 300 μM, TG content was significantly higher than without oleic acid (P <0.05).
We examined the expression of five key fat-forming genes at different concentrations of oleic acid (Fig 3F). Except for LPL, almost all genes reached their highest peak of expression at 300 μM oleic acid. Therefore, the optimal amount of oleic acid to induce adipogenesis in poultry pre-adipocytes was 300μM.

Ability to differentiate in different genetic backgrounds
Morphological observations were performed on pre-adipocytes from three duck breeds: Cherry Valley duck (CV), White-crested duck (WC), and Jianchang duck (JC) at 3 days after induction. As shown in Fig 4, CV and JC produced more lipid droplets from pre-adipocytes than WC. The TG content of adipocytes from each of the three breeds was compared at 3 days after induction (Fig 4D). The pre-adipocytes from CV were used as a control. The TG content in JC was the highest. The proliferating cells were in the logarithmic phase from 24 h to 72 h in all three varieties. We found that JC had the highest proliferative capacity, followed by CV, while WC had the lowest proliferation of pre-adipocytes.
At 3 days after induction, there was a significant difference in the expression levels of the five genes in the three ducks, with JC showing the highest expression in almost all genes, except C/EBPα.

Effect of PPARγ interference on adipogenesis
To investigate the role of PPARγ in adipogenesis, we performed knock-down experiments using siRNA specific to PPARγ (RNAi-1343, RNAi-872, and RNAi-54). Pre-adipocytes were transfected with the siRNAs and cultured for 48 h. We found that siRNA-872 significantly suppressed the expression of PPARγ mRNA ( Fig 5A) and the highest interference efficiency. We used RNAi-872 to continue the interference of PPARγ. After 48 h of interference, the changes in expression of another adipogenic differentiation transcription factor-C/EBPα and other genes related to adipocyte differentiation, including FABP4, LPL, and FAS, were detected using qRT-PCR (Fig 5B). The expression levels of C/EBPα, FABP4, and LPL were all decreased compared with the control (induction without interference), and the expression of LPL was significantly decreased (P<0.05) after the expression of PPARγ was decreased. However, the expression level of FAS did not change significantly. Quantitative measurements of TG content were used to evaluate the impact of PPARγ interference on adipocyte adipogenesis, and showed that TG levels were lower after PPARγ interference compared to the control group (I3) (Fig 5C).
Morphological observation and Oil Red O staining showed that PPARγ siRNA-872 had almost no effect on adiposity after 48 h induction (Fig 5D), perhaps as part of PPARγ was still working, even in the knock-down experiment. As the differentiation progressed, the expression of PPARγ gradually increased and played a vital role in adipogenesis.

Discussion
So far, there have been few systematic studies on the isolation, culture, and induction of duck primary fat cells. In 2012, He et al. [20] reported the functions of A-FABP and its regulatory  genes related to adipogenesis in duck primary pre-adipocytes. Subsequently, Ding et al. [3] reported a more comprehensive and systematic method for the isolation and induction of duck pre-adipocytes. In this study, we successfully isolated and obtained pre-adipocytes in 20-dayold embryos using these methods. During the induction process, we added rosiglitazone, which differed from the previous methods [21]. At the beginning of induction, the pre-adipocytes gradually shrank from a fusiform shape and became round. The lipid droplets in the cells formed and small lipid droplets gradually accumulated into large lipid droplets. The lipid droplet filled the cytoplasm and it became a mature white adipocyte. This process was similar to that in pre-adipocytes from chicken, which share similar morphologies with primary pre-adipocytes in ducks [12]. The lipid droplets eventually formed large lipid droplets in cells; however, adipocytes that were differentiated from primary pre-adipocytes easily detached from the bottom of the flask at this stage and floated as dead cells.
The expression of DLK1 in adipocytes is highly specific. We observed that it was highly expressed in pre-adipocytes, but gradually disappeared as the adipocytes differentiated, and it was almost absent in mature adipocytes. A previous study has shown that transfection of constitutive expression components of Pref-1 significantly reduced the degree of 3T3-L1 adipocyte differentiation [22]. The expression levels of FABP4, C/EBPα, FAS, and LPL increased with induction, and maintained a high level of expression compared with pre-adipocytes.
Monounsaturated fatty acids regulate the proliferation and differentiation of pre-adipocytes. Unlike in mammals, poultry adipocytes do not synthesize fatty acids, so extra fatty acids must be added to induce adipogenesis. Previously, it was reported that a concentration of 0.1 mmol L-1 oleate induced the expression of transcriptional differentiation factors, including PPARγ and SREBP-1, while high levels of oleic acid reduced their expression [23]. In this study, the TG content decreased at 600 μM oleic acid concentration, compared to 300 μM, indicating that high concentrations of oleic acid induced duck adipocyte differentiation, which was consistent with observations of mammalian pre-adipocytes. Thus, high concentrations of oleic acid may have inhibitory effects on fat production.
The number and size of adipose cells determines the amount of adipose tissue [24], so the results for adipocytes in different breeds of duck were inconsistent. Wang et al. [25] made a comparison between the fat and lean broiler lines, a total of 33 miRNAs were found to be significantly differentially expressed, and the target genes of these differentially expressed miRNAs were enriched in pathways related to adipocyte development and lipid metabolism, such as transcription regulation, response to insulin stimulation, and more interestingly, and IGF-1 signalling pathway. Different genetic backgrounds of the same species have different patterns of fat deposition [26]. In our study, we found that CV duck and JC duck had a high amount of adipose tissue, and their pre-adipocyte content and growth rate were faster than WC duck. After induction, the TG content in adipocytes from WC duck was significantly lower than that in JC duck.
PPARγ is the most adipocyte-specific and adipogenic member of the PPAR family. It is mainly expressed in adipose tissue and plays an important role in lipid metabolism and adipocyte differentiation [27]. This study explored the role of PPARγ in adipogenesis by interfering with PPARγ. After the expression of PPARγ was knocked-down, the adipogenic capacity was reduced, although not significantly. This could be explained by the decrease in TG content after gene interference. Previous interference experiments with chicken PPARs also showed that PPARγ inhibits chicken pre-adipocyte proliferation and promotes differentiation [28]. PPARγ was at the center of this regulatory network and played an essential role in adipocyte differentiation. In the absence of PPARγ, pre-adipocyte differentiation did not begin properly [29]. The interference of PPARγ regulated the expression of other fat-related genes, which significantly reduced the expression of LPL and had a slight effect on the expression of FABP4, C/ EBPα, and FAS, albeit not significantly.