Oleic acid in the absence of a PPARγ agonist increases adipogenic gene expression in bovine muscle satellite cells

Abstract

We hypothesized that oleic acid (OA) in the absence of a thiazolidinedione (i.e., a synthetic peroxisome proliferator-activated receptorγ [PPARγ] agonist) would increase adipogenic gene expression in bovine muscle satellite cells (BSC). The BSC were cultured in differentiation medium containing 10 µM ciglitazone (CI), 100 µM OA, or 100 µM OA plus 10 µM CI (CI-OA). Control (CON) BSC were cultured only in differentiation media (containing 2% horse serum). The presence of myogenin, desmin, and paired box 7 proteins was confirmed in the BSC by immunofluorescence staining, demonstrating that we had isolated myogenic cells. The OA BSC had lesser paired box 3 (Pax3) and myogenic differentiation 1 expression but greater Pax7 and mygogenin (MYOG) expression (P < 0.05), than the CON BSC. The CI BSC had greater Pax3, Pax7, and MYOG expression than CON BSC (P < 0.05), suggesting that CI would promote BSC myogenesis under pro-myogenic conditions (i.e., when cultured with horse serum). However, both the OA and CI treatments upregulated the expression of PPARγ, CCAAT/enhancer-binding protein alpha (C/EBPα) and C/EBPß, sterol regulatory element-binding protein 1, lipoprotein lipase, and glycerol-3-phosphate acyltransferase 3 gene expression, as well as media adiponectin concentration (P < 0.05). The CI, OA, and CI-OA treatments also increased triacylglycerol and lipid droplet accumulation, in spite of upregulation (relative to CON BSC) of adenosine monophosphate-activated protein kinase alpha-1, perilipin 2 (PLIN2), and PLIN3 in BSC and downregulation of G protein-coupled protein receptor 43, acyl-CoA synthetase long chain family member 3, and stearoyl-CoA desaturase (P < 0.05). These results indicate that OA in the absence of a synthetic PPARγ agonist can effectively increase adipogenic gene expression in BSC.

Introduction

Our laboratory previously reported that co-culture of bovine muscle satellite cells (BSC) with preadipocytes in medium containing oleic acid (OA; 18:1n-9) and the thiazolidinedione, ciglitazone (CI), increased peroxisome proliferator-activated receptor gamma (PPARγ) and CCAAT/enhancer-binding protein beta (C/EBPß) gene expression in differentiated myoblasts and G protein-coupled receptor 43 (GPR43) gene expression in adipocytes (Choi et al., 2013). Chung et al. (2016) demonstrated that treating bovine intramuscular (i.m.) adipocytes with the combination of OA plus CI upregulated PPARγ and GPR43 gene expression, which was associated with lipid filling of i.m. adipocytes. OA is the most abundant fatty acid in bovine muscle and i.m. adipose tissue (Sturdivant et al., 1992; May et al., 1993; Zembayashi et al., 1995), and increases in i.m. adipose tissue mass are accompanied by corresponding increases in OA (Smith et al., 2006; 2009). Muscle from Japanese Black cattle is exceptionally high in OA (Sturdivant et al., 1992; May et al., 1993; Chung et al., 2006) and is unique in that it contains i.m. adipocytes within muscle fasciculi (Smith et al., 2000; Gotoh et al., 2014). Based on these findings, we hypothesized that OA might promote adipogenic gene expression in BSC. To test this hypothesis, we cultured BSC in the presence of OA in the absence or presence of a thiazolidinedione (TZD; CI) to document if OA would affect expression of genes associated with fatty acid biosynthesis and triacylglycerol (TAG) turnover.

For the current study, we selected differentiation medium containing Dulbecco’s modified Eagle medium (DMEM) plus 2% horse serum based on Singh et al. (2007), who described the trans-differentiation of porcine satellite cells to adipoblasts. To address the central hypothesis of this study that OA, in the absence of a TZD, would upregulate genes associated with lipid metabolism in BSC, it was important to minimize incubation media components. We included horse serum in the media to encourage BSC proliferation and myogenic gene expression to better reflect in situ conditions.

Materials and Methods

Animal Use

All experimental procedures were approved by the Yanbian University Institutional Animal Care and Use Committee.

Bovine Satellite Cell Isolation

Muscle-derived BSC were isolated from six 18-mo-old Yanbian yellow cattle raised at Yanbian University. Using sterile techniques, approximately 800 g of the semimembranosus muscle was dissected and transported to the cell culture laboratory within 10 min. Cell isolation procedures were conducted in the sterile environment in a culture hood.

After removal of connective tissue, the muscle was passed through a sterile meat grinder. The ground muscle was incubated with 0.1% pronase (Calbiochem, La Jolla, CA) in Earl’s Balanced Salt Solution (Sigma-Aldrich, Louis, MO) for 1 h at 37 °C with frequent mixing (Chung and Johnson, 2009). After incubation, the mixture was centrifuged at 1,500 × g for 4 min, the pellets were suspended in phosphate-buffered saline (PBS) (140 mM NaCl, 1 mM KH₂PO₄, 3 mM KCl, and 8 mM Na₂HPO₄; Gibco Life Technologies, Grand Island, NY), and the suspensions were centrifuged at 500 × g for 10 min. The supernates were centrifuged at 500 × g for 10 min to pellet the mononucleated cells. The PBS wash and differential centrifugation were repeated twice. The resulting mononucleated cell preparations were suspended in 37 °C DMEM (Gibco) containing 10% fetal bovine serum (FBS; Gibco), 1× antibiotic-antimycotic (Gibco), and 10% (vol/vol) dimethyl sulfoxide (Sigma-Aldrich) and frozen. Cells were stored frozen in liquid nitrogen.

Bovine Satellite Cell Culture

Experiments were replicated in three independent incubations (passage 2 of the BSC for all experiments). The BSC were resuspended and maintained in growth media composed of DMEM, 10% FBS, and 1% penicillin/streptomycin at 37 °C under a humidified atmosphere of 95% O₂ and 5% CO₂. Upon reaching confluence, the growth medium was replaced by differentiation medium composed of DMEM, 2% horse serum, and 1% penicillin/streptomycin.

We changed the differentiation components of Singh et al. (2007), omitting ascorbic acid, biotin, acetic acid, panthothenic acid, dexamethasone, isobutylmethylxanthine, and insulin. The BSC were cultured for 96 h in differentiation medium containing only 10 µM CI, 100 µM OA, or 100 µM OA plus 10 µM CI (CI-OA). Control (CON) BSC were cultured only in our modified differentiation medium. OA (Sigma-Aldrich) was dissolved in ethanol before addition to the BSC cultures. Cell size and cell viability were calculated by Luna Automated Cell Counter (Logo Biosystems, Annandale, VA) after trypan blue staining.

The BSC initially were cultured for 6 d and we observed that most BSC were filled with lipid droplets when exposed to the OA or CI-OA treatments. After 4-d culture, culture media was changed or the BSC were washed with PBS very carefully and gently. However, most of the lipid droplets in the BSC floated to the surface of the culture wells; thus, we were unable to harvest cells that included lipid droplets to conduct the oil red O staining. Consequently, we harvested BSC at day 4 for the measurement of oil red O staining all other dependent variables.

Immunostaining

To confirm that we have isolated BSC, we cultured BSC to approximately 70% confluence with DMEM, 10% FBS, and 1% penicillin/streptomycin without OA or CI. The BSC were grown on coverslips for 3 d and fixed with 4% paraformaldehyde for 30 min and washed with PBS. The BSC were incubated with polyclonal anti-myogenic differentiation 1 (MYOD1; BS-2442R), anti-Pax7 (AB-528428), and anti-desmin Po (BS-20702R) at 1:100 dilutions for 1 h and then exposed to fluorescein isothiocyanate (FITC)-conjugated goat antimouse or antirabbit IgG antibody (Santa Cruz Biotechnology). Finally, BSC were incubated in 0.1 µg/mL 4´,6-diamidino-2-phenylindole (DAPI) solution for 1 min and observed with a fluorescence microscope (Olympus BX53F, Tokyo, Japan).

Morphological Analysis

In order to assess the level of myotube formation, media was aspirated from BSC after 96-h differentiation and BSC were fixed with 10% neutral buffered formalin (30 mM NaH₂PO₄ ∙ H₂O, 54.6 mM NaHPO₄, 40% formalin) for 30 min and stained with hematoxylin solution (51275, Fluka Chemie GnbH, Switzerland) at room temperature for 5 min. In addition, BSC after 96-h differentiation were fixed with 10% neutral buffered formalin for 10 min and stained with an Oil Red O solution (O0625-25G, Sigma). After washing the BSC once with 60% isopropanol and twice with distilled water, cells were photographed using a Nikon E5400 digital camera mounted on an optical microscope (Nikon TS100, Nikon Corporation, Japan).

Lipid Droplet Area Quantification

After 96-h differentiation, lipid droplets of BSC were stained with Oil Red O. Image of 20 different areas of BSC were captured with a Nikon E5400 digital camera, followed by computer image analysis using ImageJ software 1.43u (http://rsbweb.nih.gov/ij) as described by Deutsch et al. (2014). Analysis was performed by threshold converting the 8-bit red-green-blue image into a binary image, which consists only of the pixels representing the lipid droplets. Thereby, the acquired images were thresholded for the color saturation of the lipid droplet signal with the threshold command option ‘‘pass.’’ After binarization, the image was subjected to watershed object separation for image processing, which is used to identify borders of adjacent lipid droplets. After separation, the binary image was manually compared with the original image for consistency and correct binary conversion. Lipid droplets not separated by watershedding were corrected using the ImageJ pencil tool. After setting the scale of the image, the amount and individual size of the lipid droplets in the image, displayed by ImageJ as surface area in square micrometers, were measured. Incomplete droplets located at the edge of the image were excluded.

Quantitative Triacylglycerol Assay

Triacylglycerol was measured using a TAG quantification assay kit (Zen Bio, Inc.). Cultured BSC were washed with buffer to remove residual medium and the cells lysed with lysis buffer. Accumulated TAG was digested with lipase for up to 3 h to release glycerol into the buffer. Aliquots were then removed and the amount of glycerol was measured using Reagent A (Zen Bio, Inc.). The assay determines the amount of glycerol liberated from TAG by spectrophotometric detection at 540 nm and is linear over a glycerol concentration range of 0 to 500 μM.

Ribonucleic Acid Extraction and Real-Time PCR

Following 96 h of BSC culture, total ribonucleic acid (RNA) was isolated (duplicate six-well culture plates for three independent experiments) with 500 µL of TRI-Reagent (Sigma-Aldrich). Genomic DNA was removed from extracted RNA with DNase (M610A, Promega, Madison, WI) according to the manufacturer’s instructions. The purity, concentration, and integrity of the total RNA from each sample were quantified using a NanoDrop spectrophotometer (Thermo Scientific, 2000C, Washington, DC) and the RNA 6000 Nano Assay (Agilent Technologies, Palo Alto, CA) assessed the RNA integrity. The purity of the RNA (A260/A280) was greater than 1.85, and the A260/A230 ratio was approximately 2.0 in all samples. All samples had RNA integrity numbers greater than 7. Complementary DNA (cDNA) was produced from 1 μg of RNA using Taq-Man Reverse Transcriptase Reagents (Applied Biosystems, Waltham, MA) according to the protocol recommended by the manufacturer.

To characterize the degree of myogenesis of bovine satellite cells, paired box 3 and 7 (Pax3 and Pax 7), MYOD1, and myogenin (MYOG) were selected as myogenic-specific marker genes. To characterize the degree of adipogenic/lipogenic gene expression in BSC, PPARγ, CCAAT/enhancer-binding protein alpha and beta (C/EBPα and C/EBPß), and sterol regulatory element-binding protein 1 (SREBP1) genes were selected. The genes stearoyl-CoA desaturase (SCD), carnitine palmitoyltransferase 1 (CPT1), adenosine monophosphate-activated protein kinase alpha-1 (AMPKα), lipoprotein lipase (LPL), GPR43, perilipin 2 and 3 (PLIN2 and PLIN3), acyl-CoA synthetase long chain family member 3 (ACSL3), acyl-CoA acetyltransferase 2 (ACAT2), glycerol-3-phosphate acyltransferase 3 (GPAT3), and fatty acid-binding protein4 (FABP4) were selected. Each sample of first-strand cDNA was synthesized from 600 ng of total RNA using QuantiTect Reverse Transcription Kit (Qiagen, cat#205313). Gene-specific primers (Table 1) were used to amplify target genes using 10 ng of first-strand cDNA as template in a 15-µL Synergy Brands (SYBR)-green-based quantitative real time-polymerase chain reaction (RT-PCR) reaction performed under the following conditions: 95 ºC for 2 min, 45 cycles at 95 ºC for 10 s, 60 ºC for 30 s, 72 ºC for 10 s with a melting curve from 65 to 95 ºC. The gene expression levels were normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) housekeeping gene expression in each sample. Transcript data were performed in duplicate. Relative expression was quantified by using the 2^−ΔΔCt method. The cDNA from each sample was pooled, serially diluted, quantified, and used for standard curves. Titration messenger RNA primers for all genes against increasing amounts of cDNA yielded linear responses with slopes between −2.8 and −3.0. To reduce the effect of assay-to-assay variation in the PCR assay, all values were calculated relative to a calibration standard run on every real-time PCR assay.

Table 1.

Forward and reverse primers for real-time PCR

Genes	Accession no.	Primer Sequence, 5’–3’
Pax3	NSO_4856884_001	Forward: GGCTGCGTCTCTAAGATCCT
		Reverse: ATTTCCCAGCTGAACATGCC
Pax7	NSO_4856884_004	Forward: TGCCCTCAGTGAGTTCGATT
		Reverse: CGGGTTCTGACTCCACATCT
MYOD1	8404456555	Forward: CCGACGGCATGATGGACTA
		Reverse: CTCGCTGTAGTAAGTGCGGT
MYOG	8404456557	Forward: CAGTGAATGCAGCTCCCATAG
		Reverse: GCAGATGATCCCCTGGGTTG
PPARγ	NSE_1498162_016	Forward: ATCTGCTGCAAGCCTTGGA
		Reverse: TGGAGCAGCTTGGCAAAGA
C/EBPα	NSO_5140066_004	Forward: ATCGACATCAGCGCCTACAT
		Reverse: GCCCGGGTAGTCAAAGTCG
C/EBPß	840057464	Forward: CCAGAAGAAGGTGGAGCAACTG
		Reverse: TCGGGCAGCGTCTTGAAC
SREBP1	NSE_1498162_022	Forward: CACCGAGGCCAAGTTGAATAA
		Reverse: CCAGGTCCTTCAGCGATTTG
SCD	NSE_1498162_019	Forward: TGCCCACCACAAGTTTTCAG
		Reverse: GCCAACCCACGTGAGAGAAG
CPT1	84044456553	Forward: ACACATCTACCTGTCCGTGATCA
AMPKα	8404456568	Reverse: CCCCTGAGGATGCCATTCT
		Forward: CACCAAGGTGTAAGGAAAGCA
LPL	NSE_1498162_015	Reverse: ACGGGTTTACAACCTTCCATTC
		Forward: ACGATTATTGCTCAGCATGG
		Reverse: ACTTTGTACAGGCACAACCG
GPR43	NSO_5189193	Forward: GGCTTTCCCCGTGCAGTA
		Reverse: ATCAGAGCAGCGATCACTCCAT
FABP4	AJ_4160220	Forward: AAACTTAGATGAAGGTGCTCTGG
		Reverse: CATAAACTCTGGTGGCAGTGA
PLIN2	NSO_5379931	Forward: GCGTCTGCTGGCTGATTTCT
PLIN3	NM_001077046	Reverse: TGTAAGCCGAGGAGACCAGA
		Forward: GACGAGACTGAAGCCACTGC
ACSL3	XM_019977627	Reverse: GGGGCTTTCTCGACGATTCC
ACAT2	NM_001075549_1	Forward: ACCGTCTTCCATGAAGCTGA
		Reverse: GCTCAATGTCCGCCTGGTAA
		Forward: TCAATGGTGCCTTATCGACC
GPAT3	NM_001192514_3	Reverse: ACATTGCTATTCCCATCCCACC
GAPDH	NSO_4761240	Forward: TCCTTTACACCTGGCTGACC
		Reverse: CAGTCCAGTTCCGACTTGAGA
		Forward: ACTCTGGCAAAGTGGATGTTGTC
		Reverse: GCATCACCCCACTTGATGTTG

Adiponectin in BSC Culture Media

Adiponectin was measured in BSC culture media. Media samples were analyzed using a bovine adiponectin ELISA kit (CSB-E14054B, CUSABIO, Houston, TX) according to the manufacturer’s protocol. Adiponectin concentration was measured at 450 nm with a Biotek Epoch Microplate Reader (Winooski, VT). The amount of adiponectin in the samples was calculated from a reference curve generated in the same assay with reference standards of known concentrations of adiponectin. Mean interassay and intraassay precision for the adiponectin assay were <8.4% and <5.7%, respectively.

Statistical Analysis

All data are presented as mean ± SEM. One-way analysis of variance with the post hoc Tukey’s multiple comparison test was used to evaluate statistical significance of differences among the treatments (Graph Pad Prism 6.0). Statistical significance was set at P < 0.05.

Results

Yanbian Yellow Cattle BSC Are Myogenic

In our study, we used MYOD1 antibodies to monitor the myogenic differentiation of BSC (Figure 1A) and desmin and Pax7 antibodies to characterize proliferation-positive BSC by immunofluorescence staining (Figure 1B and C).

Figure 1.

Immunofluorescence staining showed that the nucleus after DAPI staining was blue and MYOD1 (A), Desmin (B), and Pax7 (C) were green in the cytoplasm. The original magnification was ×40 (scale bars = 200 µm). Results are representative of three separate experiments.

Average BSC Size and Cell Viability

A total of 1,055 ± 33.82 cells for each treatment were analyzed for the measurement of average cell size. The CI-OA treatment increased average cell size (20.10 ± 0.11 µm) relative to CON (16.03 ± 0.40 µm), CI (17.63 ± 0.12 µm), and OA (16.03 ± 0.25 µm) BSC (P < 0.001) (Figure 2A). There was no difference in BSC viability among treatments (P = 0.79) (Figure 2B).

Figure 2.

Effects of CI and/or OA on average cell size (A) and cell viability (B) in bovine satellite cells after 96-h culture. CON, DMEM + 2% horse serum; CI, CON + 10 µM CI; OA, CON + 100 µM OA; CI-OA, CON + 10 µM CI +100 µM OA. Data are means of six culture dishes derived from six animals. ^a,b,cMeans within a panel with common superscripts are not different (P > 0.05).

BSC Differentiation and Lipid Accumulation

Exposure of postconfluent cells to the differentiation medium with added CI and/or OA led to a marked increase in the area of cells containing lipid droplets and formation of adipoblasts (Figure 3). Control BSC proliferated and subsequently fused to form myotubes. Upon incubation of BSC with CI and/or OA, the cells became adipocyte-like, an effect that was clearly evident after 96 h (Figure 3D).

Figure 3.

Oil red O- and hematoxylin-stained BSC cultured with OA and/or CI in DMEM containing 2% horse serum after 96-h culture. (A) CON (DMEM + 2% horse serum); (B) CI (CON + 10 µM CI); (C) OA (CON + 100 µM OA); (D) CI-OA (CON + 10 µM CI +100 µM OA; scale bars = 200 µm).

Concentration of Triacylglycerol and Lipid Droplet Area

Addition of CI and/or OA caused a profound increase in TAG and lipid droplet accumulation in BSC (both P < 0.0001; Figure 4). The concentration of cellular TAG, the main component of lipid droplets, was significantly elevated in BSC by the CI, OA, and CI-OA treatments and, correspondingly, the CI, OA, and CI-OA treatments increased the percent area of lipid droplets compared to CON (P < 0.05). The greatest concentration of TAG was observed in the CI-OA BSC (P < 0.05).

Figure 4.

Effect of CI and/or OA on triacylglycerol (A) and percent lipid droplet area (B) in BSC at 96-h culture time. CON, DMEM + 2% horse serum; CI, CON + 10 µM CI; OA, CON + 100 µM OA; CI-OA, CON + 10 µM CI + 100 µM OA. Data are means of six culture dishes derived from six animals. ^a,b,cMeans within a panel with common superscripts are not different (P > 0.05).

Gene Expression

Myogenic

The CI and CI-OA treatments increased Pax3 expression, whereas the OA treatment decreased Pax3 expression (P < 0.001; Figure 5A). The CI, OA, and CI-OA treatments increased Pax7 expression relative to CON, with the OA treatment causing the greatest increase (Figure 5B; P < 0.001). CI and OA strongly depressed expression of MYOD1 relative to CON and CI-OA BSC (P < 0.0001; Figure 5C). The expression of MYOG was greatest in the OA-treated BSC and least in the CI-treated BSC (P < 0.0001; Figure 5D).

Figure 5.

Pax3 (A), Pax7 (B), MYOD1 (C), and MYOG (D) gene expression in BSC cultured with OA and/or CI in DMEM containing 2% horse serum at 96-h incubation time. CON, DMEM + 2% horse serum; CI, CON + 10 µM CI; OA, CON + 100 µM OA; CI-OA, CON + 10 µM CI + 100 µM OA. Data are means ± pooled SEM for three separate incubations. ^a,b,cMeans within a gene with common superscripts are not different (P > 0.05).

Adipogenic

Expression of PPARγ, C/EBPα, C/EBPß, and SREBP1 was greater in CI and OA BSC than in CON BSC (P ≤ 0.0012; Figures 6A–D). The CI-OA treatment did not increase PPARγ or C/EBPα gene expression but increased C/EBPß and SREBP1 expression relative to CON BSC (P < 0.0001).

Figure 6.

PPARγ (A), C/EBPα (B), C/EBPß (C), SREBP1 (D), SCD (E), CPT1 (F), AMPKα (G), LPL (G), GPR43 (I), and FABP4 (J) gene expression in BSC cultured with OA or/and CI in DMEM containing 2% horse serum at 96-h incubation time. CON, DMEM + 2% horse serum; CI, CON + 10 µM CI; OA, CON + 100 µM OA; CI-OA, CON + 10 µM CI + 100 µM OA. Data are means ± pooled SEM for three separate incubations. ^a,b,c,dMeans within a gene with common superscripts are not different (P > 0.05).

Associated with lipid metabolism

Relative SCD gene expression was least in OA BSC and greatest in CI (P < 0.0001; Figure 6E). Expression of CPT1 was elevated in CI BSC only, relative to CON BSC (Figure 6F). Expression of AMPKα was increased only by OA treatment (P < 0.0001) and LPL expression was greatest in CI-OA BSC (P < 0.0001; Figure 6G). Expression of GPR43 was depressed by the OA and CI-OA treatments relative to CON BSC (P = 0.0073; Figure 6I) and FABP4 expression was increased stepwise by CI, OA, and CI-OA (P < 0.0001; Figure 6J).

Expression of PLIN2 (P < 0.0001) and PLIN3 (P = 0.0003) was greater in CI, OA, and CI-OA BSC than in CON BSC (Figure 7). Relative ACSL3 (P = 0.0017) and ACAT2 (P = 0.009) gene expression decreased with CI, OA, and CI-OA treatments, and expression of GPAT3 was increased by CI, OA, and CI-OA treatments (P = 0.0043; Figure 8).

Figure 7.

PLIN2 (A) and PLIN3 (B) gene expression in BSC cultured with OA and/or CI in DMEM containing 2% horse serum at 96-h incubation time. CON, DMEM + 2% horse serum; CI, CON + 10 µM CI; OA, CON + 100 µM OA; CI-OA, CON + 10 µM CI + 100 µM OA. Data are means ± pooled SEM for three separate incubations. ^a,b,c,dMeans within a gene with common superscripts are not different (P > 0.05).

Figure 8.

ACSL3 (A), ACAT2 (B), and GPAT3 (C) gene expression in BSC cultured with OA and/or CI in DMEM (I) containing 2% horse serum at 96-h incubation time. CON, DMEM + 2% horse serum; CI, CON + 10 µM CI; OA, CON + 100 µM OA; CI-OA, CON + 10 µM CI + 100 µM OA. Data are means ± pooled SEM for three separate incubations. ^a,b,c,dMeans within a gene with common superscripts are not different (P > 0.05).

Culture Media Adiponectin

Media adiponectin concentration was increased stepwise by CI, OA, and CI-OA (P < 0.0001; Figures 6J and 9).

Figure 9.

Adiponectin concentration (µg/mL) in media from BSC cultured for 96 h with OA and/or CI in DMEM containing 2% horse serum at 96-h incubation time. CON, DMEM + 2% horse serum; CI, CON + 10 µM CI; OA, CON + 100 µM OA; CI-OA, CON + 10 µM CI + 100 µM OA. Data are means ± pooled SEM for three separate incubations. ^a,b,cMeans within a panel with common superscripts are not different (P > 0.05).

Discussion

This study addressed the hypothesis that the OA would promote adipogenic gene expression in BSC in the absence of a TZD, i.e., a synthetic PPARγ agonist. The development of bovine subcutaneous (s.c.) and i.m. adipose tissues is accompanied by a corresponding increase in OA (Chung et al., 2006; Brooks et al., 2011; Smith et al., 2009), and muscle and adipose tissues from Japanese Black cattle contain unusually high quantities of OA (Sturdivant et al., 1992; May et al., 1993; Zembayashi et al., 1995; Chung et al., 2006). Additionally, there is evidence that longissimus muscle from Japanese Black cattle contains i.m. adipocytes within muscle fasciculi in addition to the extraordinarily high abundance of i.m. adipocytes surrounding muscle fasciculi (Smith et al., 2000; Gotoh et al., 2014).

There are no fibroblasts that potentially could differentiate to adipocytes in muscle fasciculi, and the source of intrafascicular adipocytes in the muscle of Japanese Black cattle has not been documented. Muscle satellite cells are considered to be stem-like cells due to their potential to proliferate (Beauchamp et al., 1999). The BSC from single myofibers are multipotential and spontaneously differentiate into myocytes, adipocytes, or osteocytes under the appropriate culture conditions (Asakura et al., 2001; Wada et al., 2002; Yeow et al., 2001; Fux et al., 2004; Kook et al., 2006). It is possible that OA produced by i.m. and/or s.c. adipocytes overlying muscle fasciculi may act as a paracrine factor, stimulating the conversion of BSC to intrafascicular adipocytes. This is difficult if not impossible to demonstrate in situ, so we utilized a BSC cell culture system in the current study. For the current study, it was essential that we demonstrate that OA, in the absence of a TZD, promotes adipogenic gene expression as BSC in situ would not be exposed to synthetic PPARγ agonists.

Previous studies have documented that fatty acids and/or PPARγ agonists promote lipid accumulation and adipogenic gene expression in BSC. Teboul et al. (1995) first reported almost complete conversion of C2C12 myoblastic cells incubated with linoleic acid (18:2n-6) plus pioglitazone to lipid-filled cells, as indicated by a cessation of myoblast fusion, rounding up lipid filling of myoblasts and a strong downregulation of MYOG and upregulation of adipocyte lipid binding protein gene expression. Yeow et al. (2001) subsequently demonstrated that rosiglitazone in the absence of supplemental fatty acids strongly upregulated PPARγ expression in C2C12 cells, and they proposed that rosiglitazone inhibited myogenesis. During embryonic and fetal development, there are common mesenchymal stem cell precursors from which skeletal muscle develops. Commitment to the myogenic lineage is dependent upon expression of regulatory factors in the mesenchymal stem cells precursors. Pax 3 and Pax7 are markers of differentiation of mesenchymal cells toward a myogenic cell lineage. Upregulation of Pax3 and Pax7 expression increases expression of myogenic regulatory factors (Cossu and Borello, 1999), which include MYOD1 and MYOG. Upregulation of MYOD1 expression promotes the conversion of precursor cells into myoblasts; MYOG is involved in the fusion of myoblasts to form myotubes (Stewart and Rittweger, 2006).

Singh et al. (2007) reported that CI upregulated PPARγ and C/EBPα gene expression and depressed myoblast fusion in porcine BSC. Similarly, in the current study, the CI strongly downregulated MYOG expression and also downregulated MYOD1 gene expression. However, OA appeared to attenuate the effects of CI in depressing expression of these genes and, in fact, the level of MYOG gene expression was greatest when BSC were treated with OA in the absence of CI (the OA treatment). These results indicate that oleic did not cause complete conversion of BSC to adipocytes, perhaps, because the incubation media also contained horse serum, which would promote myogenic gene expression.

The current study is unique in that it demonstrated that OA, in the absence of a TZD, upregulated C/EBPα and C/EBPß and increased incubation media adiponectin. OA downregulated genes associated with lipid metabolism (SCD, GPR43, ASCSL3, and ACAT2) but also upregulated AMPKα, PLIN2, PLIN3, LPL, GPAT3, FABP4, and SREBP1. We previously demonstrated a strong downregulation of SCD gene expression in BSC by OA (Choi et al., 2015), representing feedback regulation of OA concentration in cells. The adipogenic transcriptional factors, PPARγ, C/EBPα, and C/EBPß, regulate adipocyte differentiation (Umek et al., 1991; Hu et al., 1995; Boone et al., 1999). Previous in vitro studies reported that PPARγ and C/EBPß were not expressed when myoblasts were treated under conditions that promoted myogenic differentiation (Hu et al., 1995; Kook et al., 2006; Singh et al., 2007).

In the current study, expression of PPARγ as well as C/EBPα were strongly upregulated by the CI and OA treatments but not by CI-OA. As indicated above, the CI-OA treatment also did not affect MYOD1 or MYOG expression. We cannot explain why the CI-OA treatment did not upregulate PPARγ and C/EBPß expression or downregulate MYOD1 or MYOG expression; at best, we can conclude that conditions that favor myogenesis attenuate adipogenesis in BSC. However, this is difficult to reconcile with the increases in FABP4 and GPAT3 expression and media adiponectin in response to the CI, OA, and CI-OA treatments. Adiponectin expression was thought to be specific to adipose tissue (Hu et al., 1996), but recent studies reported adiponectin in rat muscle (Capllonch-Amer et al., 2014; Jiménez et al., 2019). Consistent with this, we detected adiponectin in media from CON BSC, and media adiponectin increased stepwise in response to the CI, OA, and CI-OA treatments. These results support at least partial conversion of BSC to adipocytes in the presence of CI and/or OA.

As a cellular energy sensor, AMPK promotes adenosine triphosphate-generating pathways such as glycolysis and lipolysis (Hardie and Carling, 1997). The GPR43 receptor is a cell-surface, volatile fatty acid receptor in adipose tissue, which, when activated, decreases cyclic adenosine monophosphate concentrations and attenuates lipolysis (Brown et al., 2003; Ge et al., 2008; Hong et al., 2005). In previous studies from our laboratory, GPR43 expression in BSC was not affected by media OA (Choi et al., 2013, 2015). In the present study, OA decreased GPR43 and ACSL3 expression (the latter potentially reducing the production of acyl-CoA) and increased AMPKα, PLIN2, and PLIN3 expression, all of which should have depressed TAG accumulation. However, based on the increases in TAG accumulation and lipid droplet size elicited by OA, this clearly was not the case. The accumulation of lipid caused by OA likely was the result of upregulation of LPL, FABP4, and GPAT3 expression, which would have promoted fatty acid uptake and esterification. We conclude that the magnitude of TAG synthesis exceeded that of TAG hydrolysis, typical of adipocytes actively accumulating lipids. Conversely, the CI, OA, and CI-OA treatments depressed ACAT2 gene expression in stepwise fashion, suggestion that CI and/or OA decreased cholesterol esters deposition in the lipid-filling BSC. This is consistent with the upregulation of SREBP1 observed in the current study.

In conclusion, we have demonstrated that OA, in the absence of a TZD, promotes the conversion of BSC to lipid-filled adipocytes. The current study suggests the possibility that OA released from i.m. and/or s.c. adipocytes in situ may act as a paracrine factor, stimulating the conversion of BSC to i.m. adipocytes. This may be one of the mechanisms responsible for the presence of i.m. adipocytes with muscle fasciculi reported for highly marbled Japanese Blac cattle (Smith et al., 2000; Gotoh et al., 2014).