Lower Expression of SLC27A1 Enhances Intramuscular Fat Deposition in Chicken via Down-Regulated Fatty Acid Oxidation Mediated by CPT1A

Intramuscular fat (IMF) is recognized as the predominant factor affecting meat quality due to its positive correlation with tenderness, juiciness, and flavor. Chicken IMF deposition depends on the balance among lipid synthesis, transport, uptake, and subsequent metabolism, involving a lot of genes and pathways, however, its precise molecular mechanisms remain poorly understood. In the present study, the breast muscle tissue of female Wenchang chickens (WC) (higher IMF content, 1.24 in D120 and 1.62 in D180) and female White Recessive Rock chickens (WRR; lower IMF content, 0.53 in D120 and 0.90 in D180) were subjected to RNA-sequencing (RNA-seq) analysis. Results showed that many genes related to lipid catabolism, such as SLC27A1, LPL, ABCA1, and CPT1A were down-regulated in WC chickens, and these genes were involved in the PPAR signaling pathway and formed an IPA® network related to lipid metabolism. Furthermore, SLC27A1 was more down-regulated in WRR.D180.B than in WRR.D120.B. Decreased cellular triglyceride (TG) and up-regulated CPT1A were observed in the SLC27A1 overexpression QM-7 cells, and increased cellular triglyceride (TG) and down-regulated CPT1A were observed in the SLC27A1 knockdown QM-7 cells. These results suggest that lower lipid catabolism exists in WC chickens but not in WRR chickens, and lower expression of SLC27A1 facilitate IMF deposition in chicken via down-regulated fatty acid oxidation mediated by CPT1A. These findings indicate that reduced lipid catabolism, rather than increased lipid anabolism, contributes to chicken IMF deposition.


INTRODUCTION
Meat products are important components of human food. In the last several decades, great progress has been made in meat quantity by genetic selection for growth rate and meat yield, however, higher growth rate also induced larger fiber diameters, higher proportion glycolytic fibers, and lower intramuscular fat, which seriously deteriorated the quality of meat (Dransfield and Sosnicki, 1999;Du et al., 2010;Petracci and Cavani, 2012). It is an ongoing challenge to improve meat quality meanwhile maintain growth rate.
Different from mammals, in which de novo synthesis of fatty acids mainly occurs in adipocytes, the chicken synthesize its fatty acids predominantly in the liver (Leveille, 1969;Vernon et al., 1999), similar to fish (Rollin et al., 2003), and then exported to other tissues including both muscle and adipose tissue by the peripheral vascular system. Therefore, chicken IMF accumulation is dependent on the transport and uptake of blood lipids as well as lipogenesis subsequently in muscle rather than de novo fatty acids synthesis (Griffin et al., 1987). Previous studies have identified about 20 quantitative trait loci (QTL) related to chicken IMF, which are mainly located on chromosomes 1, 2, 5, 23 (Jennen et al., 2005;D'Andre et al., 2010;Ye et al., 2010;Jia et al., 2012;Nassar et al., 2013;Sun et al., 2013;Zhang T. et al., 2015). Otherwise, a large number of genes including GPAT1, ACC, CD36, AGPAT1, and DGAT2 (Jeong et al., 2012), FABP (Ye et al., 2010;Serao et al., 2011), LPL (Zhang X. D. et al., 2015), DGAT1 (Li et al., 2013) were recognized as candidate genes for IMF, but their molecular mechanisms affecting IMF are still unclear. Obviously, the mechanism underlying chicken IMF deposition is very complicated, involving a lot of genes and metabolic pathways (Figure 1).
Although, previous studies have analyzed the transcriptome of chicken breast muscle (Cui et al., 2012) and liver (Bourneuf et al., 2006) utilizing microarrays, and identified some potential candidate genes and pathways that influence chicken breast muscle IMF deposition, no further validation has been performed. Therefore, the precise molecular mechanisms underlying chicken IMF deposition have not been fully elucidated so far. In the present study, female breast muscle of Wenchang chicken (WC, Chinese native breed, high IMF content) and Recessive White Rock chicken (WRR, imported and fast-growing breed, low IMF content) were subjected to RNA-sequencing (RNA-seq) analyses. Results showed that chicken IMF deposition is attributable to some key genes and pathways related to lipid catabolism rather than lipid anabolism, and the higher IMF deposition might resulted from lower lipid oxidation.

Ethics Statement
All animal procedures were authorized by the Animal Care Committee of South China Agricultural University (Guangzhou, China). Animals involved in the present study were sacrificed as necessary to ameliorate their suffering.

Animals and Sample Collection
In the present study, WC chickens were collected from Longquan Wenchang Chicken Industrial co., LTD (Wenchang, Hainan, China), and WRR chickens were collected from Guangdong Wen's Food Group, LTD (Xinxing, Guangdong, China). Birds were raised up to 120 d of age (D120) or 180 d of age (D180), six female birds of similar weight from each breed per age were sacrificed for muscle sample collection. The center of right breast muscle (B) was excised, divided into three parts for RNA-seq, qRT-PCR, and Oil Red O staining respectively, snap-frozen in liquid nitrogen immediately and stored long-term at −80 • C. The whole left breast muscle was excised for IMF content measurement and stored at −30 • C.

Oil Red O Staining and IMF Content Measurement
Frozen tissues were sectioned using a Leica Kryostat (Leica CM3050S, Leica instrument GmbH, Germany), fixed 30 min with 10% paraformaldehyde, and incubated 15 min at room temperature with Oil Red O, then visualized with light microscopy.
IMF contents were measured in a Soxhlet apparatus (Soxtherm, German), according to the description elsewhere (Zerehdaran et al., 2004;Cui et al., 2012). The data are represented as percentages of the wet weight of the muscle.
After the total RNA extraction and DNase I treatment, magnetic beads with Oligo (dT) were used for mRNA enrichment. The enriched mRNA was fragmented, and used as templates for synthesizing cDNA. Short cDNA fragments were purified, end-repaired, tailed with single nucleotide adenine, and then connected with adapters. With agarose gel electrophoresis, suitable fragments were amplified with PCR. Subsequently the sample library was assessed with an Agilent 2100 Bioanaylzer and ABI StepOnePlus Real-Time PCR System for quantity and quality. Finally, the library was sequenced using Illumina HiSeq TM 2000 (BGI, China).
Primary sequencing data that produced by Illumina HiSeq TM 2000, called raw reads, were subjected to quality control (QC) to determine whether a resequencing step was needed. After QC, raw reads were filtered, and clean reads were aligned to reference sequences with SOAPaligner/SOAP2 . Then the alignment data were utilized to calculate distribution and coverage of reads on reference genes. Next, we proceeded with gene expression analysis, which included gene expression level and differential expression gene (DEG). Furthermore, we performed Gene Ontology (GO) enrichment analysis (Ashburner et al., 2000;Young et al., 2010), KEGG (Kyoto Encyclopedia of Genes and Genomes, http://www. genome.jp/kegg/) Pathway enrichment analysis (Altermann and Klaenhammer, 2005;Kanehisa et al., 2008) and Ingenuity R Pathway Analysis (IPA R , http://www.ingenuity.com/) of DEGs.

Quantitative RT-PCR Analysis
The primers for qRT-PCR were designed by Primer premier 5 software. The forward (F) and reverse (R) primer of each gene were derived from different exons, and the size of each PCR product was about 150 ∼ 250 bp. qRT-PCR was carried out in a Bio-rad CFX96 Real-Time Detection system (Biorad, Hercules, CA, USA) employing KAPA SYBR FAST q-PCR Kit (KAPA Biosystems, Wobrun, MA, USA) according to the manufacturer's instruction. The 2 −( Ct) formula was used to quantify the relative gene expression with GAPDH as a reference gene (Vandesompele et al., 2002).
SiRNA oligonucleotides and negative control (NC) for gene interference were purchased from GenePharma (GenePharma, Suzhou, China).

Cell Culture
The QM-7 myoblasts were incubated in a humidified atmosphere containing 5% CO 2 in M199 medium containing 10% fetal bovine serum (FBS), 10% tryptose phosphate broth, 0.1% penicillin, and 0.1% streptomycin (growth medium). Cells were incubated in this growth medium for 48 h, the density reached to 80 ∼ 90%. Then the cells were induced to differentiate by lowering FBS concentration to 1% (differentiation medium) for another 48 h, reaching 80 ∼ 90% preconfluent. Then the preconfluent myoblasts were used for subsequent transfection experiments.

Measurement of Cellular TG and FFA
Cells pretreated as described above were incubated further for 16 h at 37 • C in serum free M199 medium containing 0.30 mM palmitate (Sigma, USA) bound to 1% bovine serum albumin (BSA, Sigma, USA) and 0.05 mM L-carnitine (Sigma, USA). After that, the cells were washed, collected and diluted with phosphatebuffered saline (PBS, pH 7.2-7.4). Cellular TG and free fatty acid (FFA) concentrations were measured using Chicken Triglyceride (TG) and FFA ELISA Kits (Jiyinmei Biological Technology Co., LTD, Wuhan, China), according to the manufacturer's guidelines.

Statistical Analyses
Data are presented as means ± S.E.M based on at least three replicates for each treatment. A one-way ANOVA was used to determine statistical significance, where P < 0.05 was considered significant differences.

IMF Content of WC Chicken and WRR Chicken
To visualize the difference of IMF deposition between breeds and within breeds, breast muscle samples were subjected to Oil Red O staining. Between the two breeds, the density and size of lipids in WC.D120.B (Figure 2A) and WC.D180.B ( Figure 2B) were larger than those of in WRR.D120.B ( Figure 2C) and WRR.D180.B ( Figure 2D). Within breed, the density and size of lipids in D180 muscles were larger than that of in D120 muscles. To further quantify the difference of IMF deposition in chickens, we measured the IMF content of breast muscles using Soxhlet apparatus and the data were showed in Table 1. IMF content of WC muscles were significantly higher than that of WRR muscles (P < 0.01), 1.24 vs. 0.53 in D120 and 1.62 vs. 0.90 in D180, respectively. Within breed, the IMF content in D180 muscles were significantly higher than that in D120 muscles (P < 0.01).

Transcriptome Assembly and Statistics of RNA-Seq
To identify the underlying molecular mechanism responsible for chicken IMF deposition, the transcriptome of breast muscle of WC chicken and WRR chicken at different ages were analyzed with RNA-seq. Two birds per breed of each age were selected to produce cDNA libraries for RNA-seq separately. The clean reads of each sample were over 20 million, the expressed genes ranged from 16,202 to 17,838, and a little higher in WRR samples than in WC samples ( Table 2).

Quantitative Real-Time PCR Verification of DEGs
To verify the gene expression pattern of DEGs detected by RNAseq, a total of 20 genes were selected randomly to perform quantitative real-time PCR experiments and the primers were listed in Supplementary Table 2. The gene expression pattern of quantitative real-time PCR were generally accordant with that of RNA-seq, although different in fold changes (Figure 4), which indicated that our RNA-seq data were reliable.

GO Terms for DEGs
GO-term analysis was used to investigate the function of DEGs. In the present study, the enriched GO-terms (Corrected P < 0.05) in the ontology classification "biological process" were represented in Figure 5. Between breeds, the enriched biological process mainly including multicellular organismal process, response to stimulus, biological regulation, regulation of biological process, regulation of cellular process, cellular developmental process, tissue development. The enriched biological process within WC chicken mainly focused on ion transport (Figure 5), no biological process was enriched within WRR chicken.

Key DEGs Associated with Chicken IMF Deposition
Based on GO function annotation, DEGs related to lipid metabolism would be attributable to chicken IMF deposition. Total of 33 (between breeds) and seven (between ages) DEGs were associated with IMF deposition (Supplementary Table 3). There were 24 DEGs down-regulated in WC chicken, including   Sebastian et al., 2009;Guitart et al., 2014), of which SLC27A1 was also more down-regulated in WRR.D180.B than in WRR.D120.B. In other words, SLC27A1 were generally lower expressed in the higher IMF muscles.

Pathways for DEGs
In vivo, lots of genes execute their biological function concertedly, and pathway enrichment analysis can reveal the main biochemical metabolism process and signal transduction pathways in which DEGs are involved. In the present study, pathways significantly enriched with DEGs were assessed by a hypergeometric test using R packages (P < 0.05, FDR adjusted). As presented in Figure 6, there were 14 (in D120) and 13 (in D180) pathways identified in DEGs between breeds, with eight being shared by two ages which included MAPK signaling, TGF-beta, cytokine-cytokine receptor interaction, focal adhesion, regulation of actin cytoskeleton, arrhythmogenic right ventricular cardiomyopathy (ARVC), cardiac muscle contraction, hypertrophic cardiomyopathy (HCM). Within breeds, between different ages, DEGs in WC chicken were enriched in MAPK signaling and regulation of actin cytoskeleton, while DEGs in WRR chicken were enriched in VEGF signaling and fat digestion and absorption (Figure 6). Similar results were obtained by previous study on Beijing-you chickens and Arbor Acres chickens using Agilent cDNA microarray (Cui et al., 2012). These results indicate that not only the pathways related to lipid metabolism (MAPK signaling; Kokta et al., 2004;Du et al., 2010), but also those involved in cell-links [cytokine-cytokine receptor interaction (Ozaki and Leonard, 2002), focal adhesion (Petit and Thiery, 2000), regulation of actin cytoskeleton (Pollard, 2003)] and cell cycle (TGF-beta; Shi and Massague, 2003) could dedicate to the IMF deposition, for their crucial role in maintaining the morphology of tissue and cell growth and proliferation. In addition, key DEGs related to lipid metabolism include CPT1A, CETP, LPL, and SLC27A1, all being involved in PPAR signaling pathway (Figure 7B), although it did not reach up to statistical significance (P > 0.05).

Gene Networks for DEGs
IPA R was performed to investigate genes affecting lipid metabolism between breeds. A total of 16 networks were identified in WRR.D120.B-VS-WC.D120.B (Supplementary Table 4) to be associated with post-translational modification, organismal injury and abnormalities, cell-to-cell signaling and interaction, energy production, cancer and lipid metabolism. The DEGs ABCA1, CPT1A, LPL, and SLC27A1, which were down-regulated in WC.D120.B and involved in PPAR signaling pathway (Figure 7B), acted as important node gene and interacted in the network of lipid metabolism, molecular transport, small molecule biochemistry ( Figure 7A). CETP was another important node gene in this network, but up-regulated in WC.D120.B. In WRR.D180.B-VS-WC.D180.B, five networks were identified (Supplementary Table 4) and they were mainly related to connective tissue development and function, cancer, organismal injury and abnormalities, nerve system development and function, dermatological diseases and conditions.

Proposed Network for SLC27A1 to Regulate Chicken IMF Deposition
Based on the above results and previous study, a network for SLC27A1 to regulate chicken IMF deposition was proposed ( Figure 7C). In this proposed network, SLC27A1 might regulate LPL (Auwerx et al., 1996), or CPT1A (Mascaro et al., 1998;Song et al., 2010), or CETP (Li and Chiang, 2009), or ABCA1 (Chinetti et al., 2001), respectively or together via PPARs, and consequently regulate the lipid metabolism of WC chickens.

PCR (
To investigate the molecular mechanism and biological pathway for SLC27A1 affecting IMF deposition, we assessed the mRNA levels of PPARs and DEGs involved in PPAR signaling pathway in pcDNA3.1(+)-SLC27A1-transfected cells compared with control pcDNA3.1(+)-EGFP-transfected cells by quantitative real-time PCR (Figure 8E). Overexpressed SLC27A1 resulted in the up-regulated for and CPT1A (P < 0.05), while no changes were observed for other genes.
SLC27A1 Knockdown Increased Cellular TG and Decreased CPT1A mRNA SLC27A1 knockdown was accomplished by RNA interference to further assess the function of SLC27A1. SiRNA SLC27A1-1220 for SLC27A1 and NC for GAPDH were transfected into QM-7 cells respectively. SLC27A1 mRNA levels decreased nearly to 50% in SLC27A1-1220-transfected cells compared with control NCtransfected cells (Figure 9A). To assess the effect of the SLC27A1 knockdown in lipid metabolism, we measured the incorporation of palmitate into TG in QM-7 cells. SLC27A1 knockdown resulted in significant increases of cellular TG concentration (4.04 ± 0.10 and 3.55 ± 0.17 mm/L, P < 0.05, Figure 9B), and significant decreases of cellular FFA concentration (394.27 ± 18.36 µm/L and 465.01 ± 18.21, P < 0.05, Figure 9C), compared with control cells. The change pattern of FFA concentration in SLC27A1 knockdown was discrepant with that in SLC27A1 overexpression, suggesting that SLC27A1 may not be primary fatty acid transporter but function on lipid metabolism in QM-7 cell.
As well, we assessed the mRNA levels of PPARs and DEGs involved in the PPAR signaling pathway in SLC27A1-1220transfected cells compared with control NC-transfected cells by quantitative real-time PCR (Figure 9E). SLC27A1 knockdown resulted in the down-regulated for CPT1A (P < 0.01), and no changes were observed for other genes.

DISCUSSION
The Higher IMF Content Is an Important Factor for the WC Chicken Excellent Meat Quality IMF content is positively correlated with flavor, tenderness, and juiciness (Fernandez et al., 1999b;Chartrin et al., 2006;Gao and Zhao, 2009;Cannata et al., 2010;Hocquette et al., 2010;Madeira et al., 2013). In pigs, increases in IMF content are accompanied by the increase in sensory tenderness (P = 0.001) and sensory FIGURE 7 | Continued overexpression increases fatty acids afflux, which offers ligands (Forman et al., 1995) for binding and activating PPARs; expression of FATP genes (SLC27A1) involves both PPARα and -γ (Motojima et al., 1998); PPAR control triglyceride metabolism by transcriptional control the expression of LPL (Auwerx et al., 1996); Though binding to PPAR responsive element in CPT1A, PPAR promotes fatty acid oxidation (Mascaro et al., 1998;Song et al., 2010); PPAR activation may inhibit plasma CETP activity (Li and Chiang, 2009); PPARs and LXRs are involved in the regulation of ABCA1 expression (Chinetti et al., 2001). Genes colored in red are up-regulated expression in WC chickens, while genes colored in green are down-regulated expression in WC chickens. Color intensity correlates to the size of the fold change. juiciness scores (P = 0.017; Cannata et al., 2010). In ducks, increasing lipid levels in breast muscle promoted tenderness and flavor, with correlation coefficients of 0.43 and 0.28, respectively (Chartrin et al., 2006). IMF affects meat quality positively mainly for two reasons. First, the oxidation of IMF can dissolve the muscle fiber bundle, thus improving the tenderness and juiciness of muscles. Second, IMF contains plenty of phospholipids, the degradation of phospholipids in heating produces various kinds of volatile aromatic compounds, thus improving the flavor of muscles. The present study showed that the IMF content of WC chickens were significantly higher than that of WRR chickens at two ages (Table 1), suggesting that the higher IMF content resulted in better meat quality of WC chickens.

Lower Lipid Catabolism Exists in WC Chicken
In the present study, we compared and analyzed the breast transcriptional profiles of WC chickens and WRR chickens of The regulative mechanism of lipid metabolism for SLC27A1. The data presented are representative of three independent experiments; n = 3, mean ± S.E.M. *P < 0.05, SLC27A1-1220 vs. NC; **P < 0.01, SLC27A1-1220 vs. NC. distinct genetic background, and a lot of DEGs between breeds and within breeds affecting IMF deposition were identified. These DEGs such as ABCA1, CPT1A, LPL, and SLC27A1 were mostly down-regulated in WC chickens (Supplementary Table 3).
Moreover, ABCA1, CPT1A, LPL, and SLC27A1 were all involved in the PPAR signaling pathway (Figure 7B), a wellknown pathway affecting lipid metabolism (Kokta et al., 2004;Du et al., 2010), and formed a IPA R network related to lipid metabolism, molecular transport, small molecule biochemistry (Figure 5A), implying that these genes may be key genes affecting chicken IMF deposition and their down-regulated expression means that a lower lipid catabolism exists in WC chicken.

SLC27A1 Negatively Regulated Lipid Accumulation of Chicken via CPT1A
Our RNA-seq data revealed that SLC27A1 was lower expressed in WC.D120.B and WRR.D180.B compared with in WRR.D120.B, while the IMF content of WC.D120.B and WRR.D180.B were higher than that of WRR.D120.B (Table 1). In other words, SLC27A1 is generally expressed lower in muscles with higher IMF contents. Our findings in cell experiments further confirmed these results. Overexpression of SCL27A1 in QM-7 cells caused the decrease in the cellular TG content (Figure 8B), while knockdown of SLC27A1 with RNAi resulted in the increase in the cellular TG content ( Figure 9B). The reduced intramuscular TG was also observed in mouse skeletal muscle overexpressed SLC27A1 (Guitart et al., 2014). Overall, SLC27A1 is negatively correlated with the lipid accumulation.
Previous studies have shown that overexpressed SLC27A1 in the skeletal muscle promoted fatty acid oxidation (Holloway et al., 2011;Guitart et al., 2014). Overexpression of SLC27A1 increased the expression of CPT1A (Figure 8E), and knockdown of SLC27A1 decreased the expression of CPT1A (Figure 9E), suggesting that SLC27A1 may regulate the expression of CPT1A in some way. PPARs are known to be pivotal transcription factors of genes involved in lipid metabolism (Lemberger et al., 1996). Their binding to PPAR responsive element localized in the 5 ′ -flanking region (Mascaro et al., 1998) and the second intron (Song et al., 2010) both increased CPT1A expression. As fatty acids and their derivatives are ligands for PPARs (Forman et al., 1995), it is possible that SLC27A1 overexpression increases fatty acids afflux and activation, which offers ligands for binding and activating PPARs, in turn activating the PPARs-dependent gene transcription, such as CPT1A, a rate-limiting enzyme of mitochondrial fatty acid oxidation (McGarry and Brown, 1997;Zammit, 2008;Akkaoui et al., 2009). In addition, cellular FFA change patterns were discrepant in SLC27A1 overexpression ( Figure 8C) and knockdown (Figure 9C), indicating that SLC27A1 plays a supplementary role in fatty acid transport, similar in rat skeletal muscle (Nickerson et al., 2009). Therefore, SLC27A1 involves in fatty acids oxidation through collaboration with CPT1A (Sebastian et al., 2009), similar to FAT/CD36 (Campbell et al., 2004;Schenk and Horowitz, 2006), but not for its transport activity across mitochondrial membrane.
Several fatty acid transporters, fatty acid translocase (FAT/CD36; Campbell et al., 2004;Schenk and Horowitz, 2006;Smith et al., 2011), plasm membrane fatty acid binding protein (FABPpm/GOT2; Clarke et al., 2004), long chain Acyl-CoA synthetase (ACSLs; Ellis et al., 2011), and fatty acid transporters (FATPs; Stahl, 2004;Kazantzis and Stahl, 2012), have been identified contributing to fatty acids uptake and oxidation. Some transporters may cooperate concertedly in fatty acids uptake, such as FABPpm and CD36 in heart and muscle (Chabowski et al., 2007), CD36 and ACSL1 in MDCK cell (Schneider et al., 2014), and ACSL1 and FATP1 in adipocytes (Richards et al., 2006). Both FATP1 and FATP4 belong to the FATP family, and their amino acid identity is 60.3% (Herrmann et al., 2001). FATP1 may involve in fatty acid uptake by hormonal regulation, while FATP4 may mediate fatty acid uptake on the basal level (Stahl et al., 2002). These fatty acid transporters are coexpressed in skeletal muscle for unknown reasons, and their functional impact with each other remains unknown. In this study, overexpression or knockdown of the SLC27A1 could not change the expression of the ACSL1, GOT2, and CD36 (Figures 8D, 9D). These findings indicate that different fatty acid transporters have distinct transcription regulatory mechanism, while SLC27A4, for its high homology with SLC27A1, was regulated similar to SLC27A1.
In conclusion, the present study showed that a higher IMF content resulted in a better meat quality of WC chickens. Lower lipid metabolism exists in WC chickens, the higher IMF deposition in WC chickens may be due to its lower fatty acid oxidation, and lower expression of SLC27A1 down-regulated the fatty acid oxidation by collaboration with CPT1A. Moreover, these findings also indicate that reduced lipid catabolism, rather than increased lipid anabolism, contributes to chicken IMF deposition.