Reduced prostaglandin I₂ signaling in Arid5b^−/− primary skeletal muscle cells attenuates myogenesis

Abstract

The AT-rich interaction domain (ARID) family of proteins regulates gene expression, development, and differentiation. Although Arid5b has important functions in adipogenesis and chondrogenesis, the role of Arid5b in skeletal muscle myogenesis has not been investigated. Therefore, we isolated primary skeletal muscle cells from Arid5b^+/+ and Arid5b^−/− mice and characterized differentiation in these cells. We found that Arid5b^−/− primary skeletal muscle cells showed differentiation defects and impaired sarcomeric assembly. Microarray analysis revealed down-regulation of the prostanoid biosynthesis pathway in Arid5b^−/− myoblasts, including the genes encoding cyclooxygenase (COX)-1 (Ptgs1) and prostaglandin (PG)I synthase (Ptgis). Down-regulation of COX-1 and PGI synthase was confirmed by real-time PCR and Western blot analyses. Correspondingly, the production of PGI₂, as measured by ELISA, was reduced in Arid5b^−/− cells relative to Arid5b^+/+ cells. Boyden chamber assays showed that migration was increased but chemotaxis was impaired in Arid5b^−/− cells. Myoblast fusion was also inhibited in Arid5b^−/− cells compared with Arid5b^+/+ cells. Treatment with the PGI₂ analog iloprost rescued the defects in myotube formation, migration, and fusion. These results demonstrate that Arid5b has a novel and essential role in skeletal muscle differentiation by regulating PGI₂ production.—Murray, J., Whitson, R. H., Itakura, K. Reduced prostaglandin I₂ signaling in Arid5b^−/− primary skeletal muscle cells attenuates myogenesis.

Members of the AT-rich interaction domain (ARID) family are involved in transcriptional regulation, development, differentiation, and cell growth (1, 2). ARID proteins are characterized by the ARID domain, which is a helix-turn-helix DNA binding domain. Using binding site selection and binding interference assays, Arid5b was shown to specifically bind to an AT-rich core sequence in DNA (3). Arid5b knockout mice show increased neonatal mortality rates along with significant reductions in neonatal weight gain and adipose tissue weights (4). Adipogenesis is inhibited in Arid5b^−/− mouse embryonic fibroblasts and by short interfering RNA–mediated knockdown of Arid5b in 3T3-L1 cells (5). Arid5b also promotes chondrogenesis by functioning as a transcriptional coregulator of Sox9 (6). Adipocytes, chondrocytes, and myocytes, in addition to other cell types, are derived from mesenchymal stem cells (7). The involvement of Arid5b in adipogenesis and chondrogenesis suggests that it may also play role in myogenesis.

In skeletal muscle the satellite cells function as stem cells to repair damaged tissue and maintain the pool of stem cells (8). In adult muscle satellite cells are mitotically quiescent until activated by stimuli such as injury or exercise (9). The satellite cells then undergo several rounds of cell division to generate a population of myoblasts. A fraction of the proliferating myoblasts then returns to quiescence, thereby replenishing the population of stem cells, whereas the majority irreversibly exit the cell cycle and fuse together to form multinucleated myotubes (MTs) (10). Myogenesis is a complex, multistep process regulated by many different signaling pathways. Understanding these pathways will provide greater insight into muscle growth and repair in response to injury, disease, and aging.

Prostaglandins (PGs) are synthesized from arachidonic acid that is released from cell membrane phospholipids by phospholipase A2 (11). The enzymes cyclooxygenase (COX)-1 and COX-2 convert arachidonic acid into PGH₂, a PG precursor. PGH₂ is then converted into the bioactive PGs PGD₂, PGE₂, PGF_2α, and PGI₂ by terminal PG synthases (12). PGs bind to specific PG receptors, which are GPCRs, and activate signal transduction pathways by regulating cAMP and Ca²⁺ levels (11, 13).

PGs have been shown to play an important role in skeletal muscle myogenesis (13). PGI₂ reduces myoblast migration and thereby promotes cell–cell contact and myoblast fusion (14). Both PGE₂ and PGF_2α promote myoblast proliferation and secondary fusion of myoblasts with nascent MTs (15–18). Additionally, PGF_2α stimulates muscle protein synthesis (19, 20) and increases myoblast survival by up-regulating the inhibitor of caspase baculovirus inhibitor of apoptosis repeat ubiquitin-conjugating enzyme (21). Loss of PG signaling has detrimental effects on myogenesis. Inhibition of COX activity by nonsteroidal anti-inflammatory drugs has been shown to reduce the proliferation of satellite cells and to inhibit muscle regeneration and growth (22–24). Taken together, these studies demonstrate that PGs have an essential role in myogenesis.

In this report we show that Arid5b is an important regulator of myogenesis. Primary skeletal muscle satellite cells isolated from Arid5b^−/− skeletal muscle showed differentiation defects and immature sarcomere formation compared with Arid5b^+/+ cells. In Arid5b^−/− cells, microarray analysis revealed a down-regulation of genes in the PGI biosynthesis pathway. We found that expression of COX-1, COX-2, and PGI synthase (PTGIS) was reduced in Arid5b^−/− cells relative to Arid5b^+/+ cells. PGI₂ produced by the Arid5b^−/− cells was decreased, leading to increased migration and to inhibition of myoblast fusion in these cells. Treatment of the Arid5b^−/− cells with the synthetic PGI analog iloprost rescued MT formation and reversed the altered migration and fusion of these cells. These studies reveal a novel role for Arid5b in promoting myogenesis by regulation of the PGI biosynthesis pathway.

MATERIALS AND METHODS

Cell culture

Generation of the whole-body Arid5b^−/− mice has been previously described (4). All animal experiments were approved by the City of Hope Institutional Animal Care and Use Committee under protocol 02001. Primary muscle satellite cells were isolated from Arid5b^+/+ and Arid5b^−/− mice at 13–25 d of age. Hindlimb muscles were minced in 1x PBS containing 100 U/ml penicillin (Corning, Corning, NY, USA), 100 µg/ml streptomycin (Corning), and 0.1% Fungizone (Thermo Fisher Scientific, Waltham, MA, USA). Muscle fragments were then digested at 37°C for 15 min with calcium- and bicarbonate-free HBSS with HEPES buffer (25) containing collagenase A (Sigma-Aldrich, St. Louis, MO, USA), 20 µg/ml gentamycin (Thermo Fisher Scientific), 2 mM l-glutamine (Irvine Scientific, Santa Ana, CA, USA), 100 U/ml penicillin, 100 µg/ml streptomycin, and 0.5 µg/ml Fungizone. Cells were then triturated and incubated for 5 min at 37°C, and liberated myocytes were collected and pelleted. The trituration and incubation steps were repeated 3 times. Cell pellets were resuspended in Ham’s F10 (Thermo Fisher Scientific) containing 20% FBS (Atlanta Biologicals, Lawrenceville, GA, USA), 2.5 ng/ml basic fibroblast growth factor (bFGF; Promega, Madison, WI, USA), 100 U/ml penicillin, and 100 µg/ml streptomycin. Cells were passed through a 70-µm filter and plated on collagen-coated dishes.

The following day, satellite cells were purified by flow cytometry using a procedure modified from Sacco et al. (26). Briefly, cells were counted using a hemocytometer and trypan blue exclusion. After pelleting, the cells were resuspended in BSA Stain Buffer (BD Biosciences, San Jose, CA, USA). Biotin-labeled antibodies against CD45, CD11b, CD31, and Ly-6A/E (BD Biosciences) were then added to the cells. After incubation with Streptavidin MicroBeads (Miltenyi Biotech, San Diego, CA, USA), cells were passed through a MACS LD Column (Miltenyi Biotech) to remove biotin-labeled nonmuscle cells. The integrin α7 antibody (MBL International, Woburn, MA, USA) was fluorescently labeled using the Zenon R-Phycoerythrin (PE) Mouse IgG₁ Labeling Kit (Thermo Fisher Scientific). Cells were incubated with fluorescently labeled antibodies against CD34 eFluor 660 (Thermo Fisher Scientific) and integrin α7-PE, and flow cytometry was performed on the Aria SORP (Becton Dickenson, Franklin Lakes, NJ, USA). Experiments were performed in triplicate from 2 independent cell isolations. Growth media contained 40% DMEM (Corning), 40% Ham’s F10, 20% FBS, 100 U/ml penicillin, 100 µg/ml streptomycin, and 2.5 ng/ml bFGF. Differentiation media (DM) consisted of DMEM with 5% horse serum (Atlanta Biologicals), 100 U/ml penicillin, and 100 µg/ml streptomycin. Iloprost (Cayman Chemical, Ann Arbor, MI, USA) was dried under a stream of nitrogen gas and resuspended in DMSO. Where indicated, iloprost was added to the media at a final concentration of 0.1 µM. Myotube length (n ≥ 100 MTs per genotype and treatment) was measured using Image-Pro Premier software (Media Cybernetics, Rockville, MD, USA).

Immunofluorescence and fusion analysis

Cells were plated in 4-well, 35-mm, glass-bottom dishes (Greiner Bio-One, Monroe, NC, USA). At day 5, MTs were fixed in 4% formaldehyde for 15 min and then blocked in 0.15% Triton and 3% normal goat serum (Jackson ImmunoResearch Laboratories, West Grove, PA, USA) in PBS for 30 min. Cells were incubated with the following primary antibodies at 4°C overnight: α-actinin (Sigma-Aldrich), troponin I (Santa Cruz Biotechnology, Dallas, TX, USA), or myosin heavy chain (MF20), developed by Dr. Donald Fischman, and obtained from the Developmental Studies Hybridoma Bank, created by the National Institutes of Health, Eunice Kennedy Shriver National Institute of Child Health and Human Development (Bethesda, MD, USA), and maintained at The University of Iowa. An Alexa Fluor 488–conjugated goat anti-mouse antibody (Thermo Fisher Scientific) was used as a secondary antibody, and nuclei were stained with Hoechst 33258 (Sigma-Aldrich). Myotubes were analyzed using the Olympus IX81 inverted microscope. For fusion analysis, at least 500 nuclei for each genotype (Arid5b^+/+ and Arid5b^−/−) were counted in MF20-stained MTs in day 1 MTs (D1MTs). Fusion index was calculated as the percentage of nuclei present in MTs relative to the total number of nuclei.

Real-time quantitative PCR

RNA was isolated using the miRNeasy kit (Qiagen, Germantown, MD, USA), and DNase was treated using the RNase-free DNase set (Qiagen). RNA was reverse transcribed using the iScript cDNA Synthesis Kit (Bio-Rad, Hercules, CA, USA). PCR was performed in 15-µl reactions containing 1X SYBR Green reagent (Bio-Rad) and 0.1 µM gene-specific primers on the CFX96 real-time PCR system (Bio-Rad). Experimental transcript levels were normalized to Rpl19 expression analyzed in separate reactions. Primer sequences are listed in Table 1. Primers marked with an asterisk were obtained from the Harvard PrimerBank (http://pga.mgh.harvard.edu/primerbank/index.html).

TABLE 1.

Sequences of primers used for real-time PCR

Gene	Gene symbol	Primer sequence, 5′−3′
		Forward	Reverse
Arid5b	Arid5b	AGAAAAACGCCCATCGAGC	CTCCCAGGATTACCACCTAAC
Cox1	Ptgs1* (144227245c1)	ATGAGTCGAAGGAGTCTCTCG	GCACGGATAGTAACAACAGGGA
Cox2	Ptgs2	AACCGTGGGGAATGTATGAG	GGTGGGCTTCAGCAGTAATT
Prostaglandin I receptor	Ptgir	GCACGAGAGGATGAAGTTTAC	AGGATGGGGTTGAAGGCGTT
Prostaglandin D synthase	Ptgds	CTCACCTCTACCTTCCTCAG	TACTCGTCATAGTTGGCCTC
Prostaglandin E synthase 2, microsomal	Ptges2	CCGTGAGAAGGACTGAGATC	AAGTGATGACCTCTTCCAGG
Prostaglandin E synthase 3, cytosolic	Ptges3* (146134999c1)	TGTTTGCGAAAAGGAGAATCCG	ACCCATGTGATCCATCATCTCA
Prostaglandin F synthase	Pgfs	CTGGATGGTGGTTACTTCTC	CTGGTAGGATGCTTAAGCTG
Prostaglandin I synthase	Ptgis* (141802458c1)	GCCAGCTTCCTTACCAGGATG	GAGAACAGTGACGTATCTGCC
Ribosomal protein L19	Rpl19	AGCCTGTGACTGTCCATTCC	GCAGTACCCTTCCTCTTCC

Primers marked with an asterisk were obtained from the Harvard PrimerBank. PrimerBank ID numbers are included in parentheses.

Western blot analysis

Total protein was collected from cells using RIPA buffer (G-Biosciences, St. Louis, MO, USA) with the addition of Halt Protease and Phosphatase Inhibitor (Thermo Fisher Scientific). Protein concentrations were determined using the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific). Proteins were resolved by SDS-PAGE and transferred to nitrocellulose membranes (Bio-Rad). Membranes were blocked in 5% nonfat milk, and primary antibody incubations were performed overnight at 4°C. The following primary antibodies were used: COX-1 (Cell Signaling, Danvers, MA, USA), COX-2 (Cell Signaling), PTGIR (LifeSpan BioSciences, Seattle, WA, USA), PTGIS (Abnova, Taipei City, Taiwan), and EFTUD2 (Proteintech Group, Rosemont, IL, USA). Horseradish peroxidase–labeled secondary antibodies were detected with Amersham ECL Prime Western Blotting Detection Reagent (GE Healthcare, Pittsburgh, PA, USA) on X-ray film (Bioland Scientific, Paramount, CA, USA) or with the Bio-Rad ChemiDoc MP Imaging System.

Microarray analysis

Total RNA was isolated from Arid5b^+/+ and Arid5b^−/− myoblasts in triplicate using the miRNeasy kit (Qiagen) and DNase treated using the RNase-free DNase set (Qiagen). RNA quality was assessed using the 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). Sense-strand cDNA was generated using the Ambion WT Expression Kit (Thermo Fisher Scientific) and was hybridized to Affymetrix GeneChip Mouse Gene 2.0 ST Arrays (Thermo Fisher Scientific). Results were analyzed using Genomics Suite software (Partek, St. Louis, MO, USA) and Ingenuity Pathways Analysis software (Qiagen). Gene expression changes with a fold difference >1.5 and P < 0.05 were considered significant.

Cell migration assays

Migration was analyzed using the CytoSelect 24-Well Cell Migration Assay (8-µm pores; Cell Biolabs, Inc., San Diego, CA, USA). In this Boyden chamber assay, the upper and lower chambers are separated by a polycarbonate membrane. Migratory cells plated in the top chamber pass through the polycarbonate membrane to the bottom side of the membrane. Cells were plated in triplicate in the top chamber at 1.5 × 10⁵ cells/well. For basal migration studies, both the upper and lower chambers contained differentiation media (DM). To assess migration in response to a chemoattractant, the bottom chamber was loaded with DM containing 100 ng/ml bFGF. To study the effects of iloprost, cells were plated in the upper chamber in DM with 0.1 µM iloprost. After incubation at 37°C for 5 h, membranes were stained according to the manufacturer’s protocol, and brightfield pictures were taken with a ×10 objective. The dye was extracted from the membranes and quantitated in a spectrophotometer at an optical density of 560 nm.

ELISA assays

Medium was collected from Arid5b^+/+ and Arid5b^−/− cells at MB or D1MT and flash frozen until the day of the experiment. Production of PGH₂, PGI₂, and PGE₂ was assessed using ELISA the following assay kits according to the manufacturers’ protocols: mouse PGH₂ ELISA kit (BlueGene Biotech, Shanghai, China), Urinary Prostacyclin EIA kit (Enzo Life Sciences, Farmingdale, NY, USA), and PGE₂ ELISA kit (Enzo Life Sciences).

Statistical analysis

All data are presented as the means ± se. Statistical significance was determined by unpaired Student’s t test, with a value of P < 0.05 considered significant.

RESULTS

Myogenesis is delayed in Arid5b^−/− primary skeletal muscle cells

To investigate the role of Arid5b in myogenesis, we isolated primary skeletal muscle satellite cells from Arid5b^+/+ and Arid5b^−/− mice. Arid5b expression was analyzed by real-time PCR during differentiation in Arid5b^+/+ and Arid5b^−/− cells (Supplemental Fig. 1). The expression of Arid5b was detectable at all stages of differentiation and peaked in D1MTs in Arid5b^+/+ cells, whereas it was undetectable in Arid5b^−/− cells. We then assessed myotube formation by microscopy analysis. During differentiation, the Arid5b^−/− MTs were shorter in length compared with those in the Arid5b^+/+ cells (Fig. 1A, B). Myotube number did not change significantly between Arid5b^+/+ and Arid5b^−/− cells (data not shown). Immunofluorescence analysis of the structural proteins α-actinin and troponin I was carried out to assess sarcomere assembly (Fig. 1C). A clear dentate staining pattern is an indicator of sarcomeric assembly and organization as well as MT maturity. Whereas the Arid5b^+/+ D5MT showed the characteristic dentate pattern of α-actinin and troponin I staining, the Arid5b^−/− D5MT staining appeared diffuse and disorganized relative to Arid5b^+/+ MTs. These results are consistent with the delayed maturation of MTs in the Arid5b^−/− cells.

Figure 1.

Myogenesis is impaired in Arid5b^−/− cells. A) Brightfield images of Arid5b^+/+ and Arid5b^−/− cells were taken at the indicated time points (original magnification, ×10). Cells were plated in triplicate, and representative images from 5 independent trials are shown. Scale bars, 100 µm. B) Myotube length was measured in Arid5b^+/+ and Arid5b^−/− D1MT cells using ImagePro Premier software (n ≥ 100 cells). ***P < 0.001. C) Immunofluorescence analysis for α-actinin and troponin I was performed in Arid5b^+/+ and Arid5b^−/− cells in D5MT cells. Representative images are shown. Scale bars, 10 µm.

The prostaglandin I biosynthesis pathway is down-regulated in Arid5b^−/− cells

To assess global changes in gene expression, microarray analysis was carried out in Arid5b^+/+ and Arid5b^−/− MBs using Affymetrix GeneChip Mouse Gene 2.0 ST arrays (Supplemental Table 1). This analysis identified 205 up-regulated genes and 155 down-regulated genes with a fold change ≥1.5 and P < 0.05 in Arid5b^−/− cells compared with Arid5b^+/+ cells. Using Ingenuity Pathways Analysis software, we identified the top canonical pathways affected in Arid5b^−/− cells (Table 2) and the genes with the greatest fold changes (Tables 3 and 4). This analysis showed that the PG biosynthesis pathway was significantly affected in Arid5b^−/− cells (Table 2). In particular, 2 critical genes in the PG pathway were significantly down-regulated: COX-1 (Ptgs1, −2.339 fold) and PGI synthase (Ptgis, −2.800 fold) (Table 4).

TABLE 2.

List of the top canonical pathways affected in Arid5b^−/− myoblasts relative to Arid5b^+/+ myoblasts

Canonical pathway	P	Ratio
Axonal guidance	3.19E–04	15/487 (0.031)
cAMP-mediated signaling	3.9E–04	10/226 (0.044)
GPCR signaling	1.44E–03	10/276 (0.036)
Ephrin receptor signaling	1.48E–03	8/210 (0.038)
Prostanoid biosynthesis	5.56E–03	2/15 (0.133)

Ingenuity Pathways Analysis software was used to identify pathways associated with differentially expressed genes that were significantly affected in Arid5b^−/− myoblasts. For each pathway, the ratio is the number of affected genes divided by the total number of genes in the pathway.

TABLE 3.

List of up-regulated genes with the greatest fold increase in Arid5b^−/− myoblasts relative to Arid5b^+/+ myoblasts

Gene symbol	Fold change
AGTR2	+4.850
ANKRD2	+3.541
MEIS2	+3.269
TDG	+3.064
HCN1	+3.006
CXorf64	+2.933
miR-154	+2.708
SCN7A	+2.679
COL19A1	+2.651
IGF2BP3	+2.586

TABLE 4.

List of down-regulated genes with the greatest fold decrease in Arid5b^−/− myoblasts relative to Arid5b^+/+ myoblasts

Gene symbol	Fold change
Zfp932	−3.376
Ly6a	−2.996
PTGIS	−2.800
SOSTDC1	−2.664
Tcrb-J	−2.625
AQP5	−2.603
PDE7B	−2.552
SELP	−2.511
GCHFR	−2.450
PTGS1	−2.339

Based on the microarray analysis, we investigated whether expression of genes in the PG biosynthesis pathway were affected in Arid5b^−/− cells. Real-time PCR analysis revealed that the expression level of COX-1 was significantly reduced in Arid5b^−/− cells relative to Arid5b^+/+ cells in myoblasts and at all stages of differentiation (Fig. 2A, upper panel). This down-regulation of COX-1 in Arid5b^−/− cells was also observed at the protein level by Western blot analysis (Fig. 2A, lower panel). The related COX isoform, COX-2, is also involved in the generation of PGs, and therefore we assessed COX-2 expression in Arid5b^+/+ and Arid5b^−/− cells. COX-2 expression was reduced in Arid5b^−/− cells in MB and D1MT, as seen by real-time PCR analysis (Fig. 2B, upper panel). COX-2 protein expression was lower in Arid5b^−/− cells in MB and at 8 h after induction of differentiation (Fig. 2B, lower panel) and could not be detected in either Arid5b^−/− or Arid5b^+/+ cells at later time points. Because COX-1 and COX-2 convert arachidonic acid to PGH₂, PGH₂ levels were measured by ELISA (Fig. 2C). This analysis revealed that there was no change in PGH₂ levels between Arid5b^+/+ and Arid5b^−/− cells in either MB or D1MT, even though COX-1 and COX-2 expression levels were reduced.

Figure 2.

Expression of COX-1 and COX-2 is down-regulated in Arid5b^−/− cells. A) Upper panel: real-time PCR analysis for COX-1 expression was performed using mRNA from Arid5b^+/+ and Arid5b^−/− cells at the indicated time points and normalized to Rpl19 expression. Lower panel: the expression level of COX-1 protein in Arid5b^+/+ and Arid5b^−/− cells was analyzed by Western blot. EFTUD2 was included as a loading control. B) Upper panel: COX-2 expression was analyzed by real-time PCR using mRNA from Arid5b^+/+ and Arid5b^−/− cells and normalized to Rpl19 expression. Lower panel: Western blot analysis of COX-2 protein expression levels was carried out in Arid5b^+/+ and Arid5b^−/− cells during differentiation. EFTUD2 was included as a loading control. C) PGH₂ production was assessed by ELISA in Arid5b^+/+ and Arid5b^−/− cells at the MB and D1MT stages. Samples were analyzed in triplicate, and results are expressed as the means ± sem. *P < 0.05, ***P < 0.001.

The microarray analysis also indicated that expression of Ptgis was down-regulated in Arid5b^−/− cells, which we verified by real-time PCR and Western blot analysis. Ptgis expression was reduced in Arid5b^−/− cells at all stages of differentiation at both the RNA and protein levels compared with Arid5b^+/+ cells (Fig. 3A). Therefore, we assessed the levels of PGI₂ production by ELISA (Fig. 3B). This analysis revealed that there was a significant reduction in PGI₂ production in Arid5b^−/− cells at the MB (−63%) and D1MT (−59%) stages relative to Arid5b^+/+ cells, consistent with the down-regulation of Ptgis expression in Arid5b^−/− cells. To determine if reduced PGI₂ production leads to increases in the levels of other PGs, we analyzed PGE₂ levels by ELISA. PGE₂ production in Arid5b^−/− MB was reduced by 30% relative to Arid5b^+/+ MB (P = 0.03), whereas no change was observed between Arid5b^+/+ and Arid5b^−/− D1MT (data not shown). Because Ptgis expression and PGI₂ production were reduced in Arid5b^−/− cells, we investigated the expression of the PGI receptor (IP). Real-time PCR analysis indicated the IP expression was down-regulated in Arid5b^−/− cells from MB to D1MT stages (Fig. 3C, upper panel). However, no change in IP protein levels was observed between Arid5b^+/+ and Arid5b^−/− cells by Western blot analysis (Fig. 3C, lower panel). These data suggest that defects in the PGI pathway are upstream of the PGI receptor.

Figure 3.

The PGI pathway is down-regulated in Arid5b^−/− cells. A) Top: real-time PCR for Ptgis expression was carried out using mRNA from Arid5b^+/+ and Arid5b^−/− cells and normalized to Rpl19 expression. Bottom: Western blot analysis was performed for PTGIS in Arid5b^+/+ and Arid5b^−/− cells. EFTUD2 was included as a loading control. B) PGI₂ production was measured by ELISA assay in Arid5b^+/+ and Arid5b^−/− cells at the MB and D1 D1MT stages. Triplicate samples were analyzed. C) Expression of IP was analyzed by real-time PCR and normalized to Rpl19 expression (top). Samples were analyzed in triplicate, and results are expressed as the means ± sem. IP protein expression was analyzed by Western blot. EFTUD2 was included as a loading control (bottom). Samples were analyzed in triplicate, and results are expressed as the means ± sem. *P < 0.05, **P < 0.01, ***P < 0.001.

We then wanted to determine if expression of other PG synthases was affected in Arid5b^−/− cells. Real-time PCR analysis showed that the expression of PGE synthases (Ptges3 and Ptges2), PGF synthase (Pgfs), and PGD synthase (Ptgds) did not change between Arid5b^+/+ and Arid5b^−/− cells at any stage of differentiation (Fig. 4). Taken together, these results suggest that only the PGI biosynthesis pathway is significantly down-regulated in Arid5b^−/− cells.

Figure 4.

Expression of PGE synthase, PGF synthase, and PGD synthase is similar in Arid5b^+/+ and Arid5b^−/− cells. Real-time PCR analysis was performed for PGE synthase 3 (Ptges3) (A), PGE synthase 2 (Ptges2) (B), PGF synthase (Pgfs) (C), and PGD synthase (Ptgds) (D). Data were normalized to Rpl19 expression levels. Samples were analyzed in triplicate, and results are expressed as the means ± sem.

Arid5b^−/− cells show increased migration and decreased fusion

Because the PGI pathway has been shown to regulate myoblast migration and fusion (14), we first investigated the ability of the Arid5b^−/− cells to migrate. Using a Boyden chamber assay, we found that the migration of Arid5b^−/− cells was increased 1.5-fold compared with Arid5b^+/+ MB under basal conditions (Fig. 5). To determine if Arid5b^−/− cells can respond to a chemotactic gradient, we performed the migration assay with bFGF, a known chemoattractant for myoblasts (27), added to the bottom chamber. Under these conditions, the migration of Arid5b^+/+ cells increased 1.7-fold in response to bFGF, whereas the migration of Arid5b^−/− cells remained at basal levels (Fig. 5). These results indicate that, although basal migration is increased in Arid5b^−/− cells, their ability to migrate toward a chemotactic gradient is diminished.

Figure 5.

Basal migration is increased in Arid5b^−/− cells, whereas chemotaxis is impaired. A) Arid5b^+/+ and Arid5b^−/− myoblasts were plated in the top well of a Boyden chamber in DM containing vehicle (PBS) or 100 ng/ml bFGF. After incubation for 5 h, migratory cells on the bottom of the membrane were stained according to the manufacturer’s protocol, and representative brightfield images are shown. Scale bars, 200 µm. B) Dye was extracted from the membranes according to the manufacturer’s protocol and quantitated at an optical density of 560 nm. N.s., not significant. *P < 0.05, **P < 0.01.

To determine if cell fusion was reduced in Arid5b^−/− cells, we assessed the fusion index in Arid5b^+/+ and Arid5b^−/− cells that were differentiated for 24 h (Fig. 6A). The fusion index was decreased by 29% in Arid5b^−/− cells compared with Arid5b^+/+ cells (Fig. 6B). In addition, we analyzed the number of nuclei per MT. The proportion of Arid5b^−/− MTs with 2–4 nuclei was increased and the proportion of Arid5b^−/− MTs with 5 or more nuclei was decreased compared with Arid5b^+/+ MTs (Fig. 6C). These results demonstrate that the MTs formed in Arid5b^−/− cells contain fewer nuclei than Arid5b^+/+ cells, indicating that fusion is impaired in Arid5b^−/− cells.

Figure 6.

Fusion is impaired in Arid5b^−/− D1MT. A) Immunofluorescence analysis was performed for myosin heavy chain (MHC) (green). Nuclei were visualized with Hoechst (blue). Representative images are shown. Scale bar, 50 µm. B) Fusion index was calculated as the percentage of nuclei within MHC-positive cells compared with total nuclei at D1MT. C) The number of nuclei per myotube was assessed and is expressed as a percentage of the total myotubes. *P < 0.05, **P < 0.01, ***P < 0.001.

Iloprost treatment rescues impaired fusion of Arid5b^−/− cells and restores migration to the level of Arid5b^+/+ cells

To determine whether reduced PGI signaling inhibits MT formation in Arid5b^−/− cells, we treated Arid5b^+/+ and Arid5b^−/− cells with the synthetic PGI analog iloprost. Cells were treated with 0.1 µM iloprost or DMSO vehicle control for 24 or 48 h in DM. Iloprost treatment rescued MT formation in Arid5b^−/− cells so that by 24 h the iloprost-treated Arid5b^−/− MTs were comparable in length to Arid5b^+/+ MTs by microscopy analysis (Fig. 7A, B). The same results were observed after 48 h of iloprost treatment. Myotube formation in the Arid5b^+/+ cells was not significantly affected by the addition of iloprost (Fig. 7B). Fusion index analysis indicated that, although fusion was reduced in DMSO-treated Arid5b^−/− cells, iloprost treatment fully restored fusion of Arid5b^−/− cells to the level observed in Arid5b^+/+ cells (Fig. 7C). Using the Boyden chamber assay, we tested the effect of iloprost on migration in Arid5b^−/− and Arid5b^+/+ cells. Iloprost treatment reduced the migration of Arid5b^−/− cells to the level of Arid5b^+/+ cells (Fig. 8). Taken together, these results indicate that the loss of PGI signaling in Arid5b^−/− cells led to increased migration and decreased fusion, resulting in the impairment of myogenesis in Arid5b^−/− cells.

Figure 7.

Iloprost treatment rescues myotube formation in Arid5b^−/− D1MT cells. A) Brightfield images of Arid5b^+/+ and Arid5b^−/− cells treated with DMSO or 0.1 µM iloprost in DM for 24 h (D1MT) or 48 h (D2MT). Representative images are shown. Scale bars, 100 µm. B) MT length was measured in Arid5b^+/+ and Arid5b^−/− D1MT treated with DMSO or 0.1 µM iloprost using ImagePro Premier software (n ≥ 100 cells). C) Fusion index was calculated for MHC-positive Arid5b^+/+ and Arid5b^−/− cells treated with DMSO or 0.1 µM iloprost in DM for 24 h (D1MT). *P < 0.05, ***P < 0.001.

Figure 8.

Iloprost treatment attenuates migration of Arid5b^−/− cells. A) The effect of iloprost on migration of Arid5b^+/+ and Arid5b^−/− myoblasts was assessed by Boyden chamber assay. Myoblasts were seeded in DM containing vehicle (DMSO) or 0.1 µM iloprost. After incubation for 5 h, membranes were stained for migratory cells according to the manufacturer’s protocol. Representative brightfield images are shown. Scale bars, 200 µm. B) Dye was extracted from membranes and quantitated at an optical density of 560 nm. *P < 0.05, **P < 0.01.

DISCUSSION

In this report, we demonstrate that Arid5b has a novel role in regulating mouse skeletal muscle myogenesis. Differentiation was impaired in primary skeletal muscle cells isolated from Arid5b^−/− mice relative to cells from Arid5b^+/+ mice. The loss of PGI signaling in the Arid5b^−/− cells resulted in increased migration levels and reduced fusion. These studies revealed that down-regulation of PGI signaling was an important factor leading to differentiation defects in Arid5b^−/− cells because treatment with the PGI analog iloprost rescued the impaired MT formation and the defects in migration and fusion.

We report a role for Arid5b in the differentiation of primary skeletal muscle cells. Cells isolated from Arid5b^−/− mice showed impaired differentiation and immature sarcomere formation relative to Arid5b^+/+ cells (Fig. 1). Microarray data clearly indicated a defect in PG biosynthesis pathways in Arid5b^−/− cells (Tables 1–3). Interestingly, the microarray analysis also revealed that the GPCR signaling and the cAMP-mediated signaling pathways were affected in Arid5b^−/− cells (Table 1). PG receptors are GPCRs that activate cAMP production (13), and cAMP signaling is an essential component of satellite cell differentiation, migration, and fusion (28–30). Therefore, cAMP signaling downstream of the IP may also be perturbed in Arid5b^−/− cells. Ongoing studies will address the contribution of IP and cAMP signaling to the differentiation defects in Arid5b^−/− cells.

Several studies have demonstrated that skeletal muscle myoblasts endogenously produce PGs and that PGs have important roles in muscle growth and differentiation. Primary mouse myoblasts were shown to produce PGI₂ (14), and additional studies showed that both human and mouse myoblasts generate PGE₂ and PGF_2α (15, 17). Treatment of the developing chick embryo with the PGI₂ synthase inhibitor tranylcypromine reduced the number of skeletal muscle fibers formed (31). The COX inhibitors aspirin and indomethacin were shown to decrease the rate of proliferation and fusion of human myoblasts (15). COX-1 expression is constitutive in most tissues, and COX-2 expression is inducible in response to inflammation, cell division, and growth factors (12, 32). However, expression of COX-1 and COX-2 was down-regulated in Arid5b^−/− cells, suggesting that both isoforms contribute to the phenotype in Arid5b^−/− cells. Our results confirmed that primary skeletal muscle cells produce PGI₂ (Fig. 3B), and this is the first report of PGH₂ production by primary skeletal muscle cells. No change in PGH₂ levels between Arid5b^+/+ and Arid5b^−/− cells was observed, even though COX-1 and COX-2 were down-regulated (Fig. 2). Although the reduction in COX-1 and COX-2 levels may decrease production of PGH₂, there is also a reduction in the conversion of PGH₂ to PGI₂, which may lead to the accumulation of PGH₂. Therefore, the net effect of lower production accompanied by lower utilization may result in PGH₂ levels in Arid5b^−/− cells that are similar to those in Arid5b^+/+ cells. Additionally, although PTGIS expression was significantly reduced throughout differentiation in Arid5b^−/− cells (Fig. 3A), we did not observe changes in the expression of other PG synthases (Fig. 4). The production of PGE₂ was mildly reduced only in MB (data not shown) compared with the strong reduction of PGI₂ production in both MB and D1MT (Fig. 3B). Therefore, our data indicate that PGI₂ is the major cause of the defects observed in differentiation, migration, and fusion in Arid5b^−/− cells.

Skeletal muscle myofibers are formed by the fusion of many myoblasts. Efficient myoblast fusion involves migration to the appropriate location, recognition of another myoblast or an existing MT, adhesion, and finally merging together into 1 cell (33). Cell–cell contact is important for myoblast fusion, and increased rates of migration impede fusion by preventing myoblast contact and adhesion (34). Loss of PGI₂ signaling in Arid5b^−/− cells impaired differentiation by increasing rates of migration and reducing myoblast fusion (Figs. 5 and 6). Importantly, the rescue of the migration and fusion defects in Arid5b^−/− cells by the PGI₂ analog iloprost (Figs. 7 and 8) indicates that the inhibition of PGI₂ signaling was sufficient to impair differentiation in these cells.

Our results are consistent with the phenotype observed in IP^−/− myoblasts, in which cell motility increased and fusion was impaired (14). Bondesen et al. (14) showed that iloprost treatment decreased rates of migration and promoted fusion in wild-type myoblasts but did not increase fusion in IP^−/− myoblasts because the receptor was ablated. We found that IP protein was expressed at the same levels in Arid5b^−/− cells as Arid5b^+/+ cells (Fig. 3). Because iloprost treatment reversed the phenotype (Figs. 7 and 8), these results indicate that IP was functional in Arid5b^−/− cells.

Myoblast migration is required during development for population of limb muscles and after injury for muscle repair, and this directed migration is regulated by chemotaxis. A previous report suggested that Arid5b may be involved in cell migration (35). In a scratch assay, mouse embryonic fibroblasts isolated from Arid5b^−/− mice showed reduced migration into the scratched area relative to Arid5b^+/+ mouse embryonic fibroblasts in the presence of platelet-derived growth factor. However, the scratch assay does not measure chemotaxis because a chemokine gradient was not generated in this assay (36) and therefore may not directly correlate with our results from Boyden chamber assays. Our Boyden chamber assays established a gradient of bFGF that stimulated migration of Arid5b^+/+ cells while the Arid5b^−/− cells remained unresponsive (Fig. 5). Therefore, Arid5b ablation appears to reduce the sensitivity of cells to a chemotactic gradient. Reduced chemotaxis in Arid5b^−/− cells suggests that in vivo these cells may not be able to migrate to sites of injury, which produce chemoattractants, and therefore muscle regeneration may be impaired. Future studies will examine the role of Arid5b in muscle repair after injury.

Taken together, our results demonstrate that Arid5b is an important regulator of skeletal muscle differentiation. We revealed a novel function for Arid5b in maintaining PGI signaling in skeletal muscle cells. By regulating myoblast migration and fusion, Arid5b may also have a role in skeletal muscle development and regeneration after injury.

Supplementary Material

This article includes supplemental data. Please visit http://www.fasebj.org to obtain this information.

Glossary

ARID

AT-rich interaction domain

bFGF

basic fibroblast growth factor

COX

cyclooxygenase

D1MT

day 1 myotube

DM

differentiation medium

IP

prostaglandin I receptor

MB

myoblast

MHC

myosin heavy chain

MT

myotube

PG

prostaglandin

PTGIS

prostaglandin I synthase

AUTHOR CONTRIBUTIONS

J. Murray performed the research, analyzed the data, and wrote the paper; J. Murray and R. H. Whitson designed the research; and K. Itakura analyzed the data and critically revised the manuscript.