FHL3 Differentially Regulates the Expression of Myhc Isoforms Through Interactions with Myod and Pcreb

Cellular Signalling 28 (2016) 60–73

Contents lists available at ScienceDirect

Cellular Signalling

journal homepage: www.elsevier.com/locate/cellsig

FHL3 differentially regulates the expression of MyHC isoforms through interactions with MyoD and pCREB

Yunxia Zhang, Wentao Li, Mingfei Zhu, Yuan Li, Zaiyan Xu ⁎,BoZuo⁎

Key Laboratory of Swine Genetics and Breeding, Ministry of Agriculture and Key Lab of Agricultural Animal Genetics and Breeding, Ministry of Education, College of Animal Science and Veterinary Medicine, Huazhong Agricultural University, Wuhan, 430070, PR China article info abstract

Article history: In skeletal muscle, muscle fiber types are defined by four adult myosin heavy chain (MyHC) isoforms. Four Received 10 June 2015 and a half LIM domain protein 3 (FHL3) regulates myoblasts differentiation and gene expression by acting Received in revised form 9 October 2015 as a transcriptional co-activator or co-repressor. However, how FHL3 regulates MyHC expression is currently Accepted 19 October 2015 not clear. In this study, we found that FHL3 down-regulated the expression of MyHC 1/slow and up-regulated Available online 22 October 2015 the expression of MyHC 2a and MyHC 2b, whereas no significant effect was found on MyHC 2x expression. Keywords: MyoD and phosphorylated cAMP response element binding protein (pCREB) played important roles in Muscle fiber type the regulation of MyHC 1/slow and MyHC 2a expression by FHL3, respectively. FHL3 could interact with FHL3 MyoD, CREB and pCREB in vivo. pCREB had stronger interaction with the cyclic AMP-responsive elements MyoD (CRE) of the MyHC 2a promoter compared with CREB, and FHL3 significantly affected the binding capacity MyHC of pCREB to CRE. We established a model in which FHL3 promotes the expression of MyHC 2a through CREB- CREB mediated transcription and inhibits the expression of MyHC 1/slow by inhibiting MyoD transcription activity during myogenesis. Our data support the notion that FHL3 plays important roles in the regulation of muscle fiber type composition. © 2015 Elsevier Inc. All rights reserved.

1. Introduction larger in diameter. The high proportion of MyHC 2b contributes to an increase in muscle mass [5]. However, the overall proportion of MyHC 2b Skeletal muscle comprises different types of muscle fibers. Tradition- has been shown to correlate with the occurrence of pig PSE (pale, soft ally, fiber types are classified as slow-oxidative (1 type), fast oxidative- and exudative) meat, which could lead to a fast pH decline after glycolytic (2a type) and fast-glycolytic (2b type) fibers, according to slaughter [6,7]. In human beings, fast-dominant skeletal muscle their contraction and metabolic characteristics [1]. Fiber types are con- induced by denervation easily leads to skeletal muscle atrophy com- vertible in adult skeletal muscle in response to exercise training [2]. pared with type 1 fibers [8]. Oxidative enzymatic activities in human Four postnatal fiber types exist in skeletal muscle, each defined by the skeletal muscle are correlated with fiber type composition [9].The presence of dominant MyHC isoforms (MyHC 1/slow, MyHC 2a, MyHC fastest, most powerful muscle fiber type (type 2b fibers) tends to 2b, MyHC 2x) [3]. In animal production, meat quantity and quality are be lost in the elderly population [10,11], suggesting that the compo- significantly affected by the composition of the four MyHC isoforms [4]. sition of skeletal muscle is associated with aging-related loss of mus- Skeletal muscle with a higher composition of type 2b fibers tends to be cle function and muscle disease. To date, several signal transduction pathways, including myogenic regulatory transcription factors (MRFs), insulin-like growth factors (IGFs), calcineurin-nuclear factor of acti- Abbreviations: MyHC, myosin heavy chain; FHL, four and half LIM domain protein; vated T cells (CaN-NFAT), CCAAT enhancer binding protein δ (C/EBP-δ) CRE, cyclic AMP-responsive element; CREB, cAMP-response-element-binding protein; and peroxisome proliferator-activated receptor-γ coactivator (PGC)- PKA, cAMP-activited protein kinase A; MyoG, myogenin; MyoD, myogenic differentiation 1α,havebeenreportedinregulatingfiber type-specific gene expressions 1; SDS, Sodium Dodecyl Sulfate; cDNA, complementary DNA; siRNA, small hairpin RNA; – PCR, polymerase chain reaction; MRFs, myogenic regulatory transcription factors; IGFs, [12 15]. insulin-like growth factors; CaN-NFAT, calcineurin-nuclear factor of activated T cells; The LIM domain has one or more cysteine-rich zinc fingers and C/EBP-δ, CCAAT enhancer binding protein δ; PGC-1α, peroxisome proliferator-activated regulates gene transcription [16]. The LIM domain is a zinc-binding receptor-γ coactivator-1α; CDS, coding sequence; bp, base pair; PVD, polyvinylidene fluo- structural motif, of which the consensus amino acid sequence is ride; EMSA, Electrophoretic mobility shift assay; ChIP, chromatin immunoprecipitation. CX2CX16–23HX2CX2CX2CX16–21CX2-3 (C/H/D) [17]. The LIM domain ⁎ Corresponding authors. E-mail addresses: [email protected] (Z. Xu), [email protected] lacks DNA-binding ability and is involved in gene expression and cell (B. Zuo). differentiation through protein–protein interactions [16]. The four and

http://dx.doi.org/10.1016/j.cellsig.2015.10.008 0898-6568/© 2015 Elsevier Inc. All rights reserved. Y. Zhang et al. / Cellular Signalling 28 (2016) 60–73 61 a half LIM domain (FHL) proteins, including FHL1, FHL2, FHL3, FHL4 listed in Table 1. The mutants of CREB binding sites were generated and ACT, are a family of LIM-only proteins with four-and-a-half LIM using the overlapping extension PCR and mutagenic primers. domains and can act as transcriptional coactivators [18]. Fhl3 is highly expressed in skeletal muscle [18-20] and interacts with other pro- 2.2. Cell culture and differentiation of C2C12 myoblasts teins, including FHL2, CREB, Sox15, MZF-1, BKLF, Ang and MT-1X, as a transcriptional co-activator or co-repressor [18,21–26].FHL3has Mouse C2C12 (ATCC) myoblasts were cultured in 10% (v/v)fetal been reported to play important roles in cell growth, development, bovine serum (Gibco, Australia) in DMEM (high-glucose Dulbecco's tumorigenesis and cancer [27–30].Inskeletalmuscle,FHL3 modified Eagle's medium) (Hyclone, USA) under humidified air con- localizes to the nucleus and focal adhesions and is a significant regu- taining 5% CO2 at 37 °C, and were differentiated at confluence in lator of actin cytoskeletal dynamics in skeletal muscle [31].FHL3 DMEM with 2% horse serum (Gibco, USA). contributes to the regulation of MyoD dependent transcription of muscle specific genes during myogenesis [32]. In pig production, a 2.3. Transfection of plasmid DNA or siRNA oligonucleotides polymorphism within the coding region of pig Fhl3 is also significantly associated with meat mass and quality traits [19,20].Based Transfections were performed using plasmid (4 μg) or siRNA on these results, we speculat that FHL3 may regulate the expression (100 pmol) by Lipofectamine 2000 (9 μl) (Invitrogen, USA) after of MyHC isoforms of different muscle fiber types, thereby regulating seeded C2C12 myoblasts had been allowed to settle for 12–18 h, the fiber type composition of skeletal muscle. However, no studies according to the manufacturer's instruction. Fhl3 small hairpin RNA regarding the regulation of MyHC gene expression by FHL3 have (siRNA) oligonucleotides were synthesized according to Mm_FHL3_2_HP been reported thus far. In this study, we present a novel molecular siRNA (SI01002890, Qiagen, Germany) and transfected into C2C12 mechanism for the regulation of MyHC expression by FHL3, which myoblasts as previously described [32]. Creb siRNA oligonucleotides may contribute to meat quality control and may benefitmuscledisease (S: GGACCUUUACUGCCACAAATT; A: UUUGUGGCAGUAAAGGUCCTT) therapy in the future. and MyoD siRNA oligonucleotides (S: CCCCAAUGCGAUUUAUCATT; A: UGAUAAAUCGCAUUGGGGTT) were designed and synthesized by 2. Material and methods GenePharma (China, Shanghai).

2.1. Plasmid constructs 2.4. Quantitative real-time PCR

The coding sequence (CDS) of the Fhl3 gene (870 bp) was obtained Total RNA from C2C12 myoblasts was extracted using TRIzol reagent by polymerase chain reaction (PCR) using cDNA (complementary (Invitrogen, USA). The concentration and quality of RNA were assessed DNA) of C2C12 myoblasts as a template. The 5′ and 3′ ends of the with a NanoDrop 2000 (Thermo, USA) and agarose gel electrophoresis. primers contain BamH IorXba Ienzymesites(Table 1). The ampli- One microgram of total RNA was used for reverse transcription with the fied CDS was digested with BamH IandXba I and was then ligated PrimeScript RTreagent kit with gDNA Eraser (Takara, Japan). The quan- into pcDNA3.1 using T4 DNA Ligase (Takara, Japan) to generate titative real-time PCR reaction was performed in a LightCycler 480 II pcDNA3.1-FHL3. (Roche, Switzerland) system using the THUNDERBIRD™ probe qPCR The CDS of Creb (1026 bp) was obtained by PCR using cDNA of Mix or SYBR®Green Real-time PCR Master Mix (Toyobo, Japan). The C2C12 myoblasts as a template. The 5′ and 3′ ends of the primers sequence of TaqMan probes and primers can be found in Table 2 and ΔΔ contain Kpn IorXho I enzyme sites (Table 1). The amplified CDS was were determined according to the literature [33].TheCt(2- Ct) digested with Kpn IandXho I and was then ligated into pcDNA3.1 method was used to analyze the relative gene expression data according using T4 DNA Ligase to generate pcDNA3.1-CREB. to the literature [34]. Five deletion fragments of the mouse MyHC 2a promoter spanning the sequence from position −2043 bp to +206 bp (2249 bp), 2.5. Western blotting −1097 bp to +206 bp (1303 bp), −909 bp to +206 bp (1115 bp), − 309 bp to +206 bp (515 bp) -109 bp to +206 bp (315 bp) (rela- Cells were lysed in RIPA buffer according to the manufacturer's tive to translation start site) were obtained by PCR (Fig. 4A). The PCR instruction (Beyotime, Jiangsu, China). Protein lysates were heated at products were cloned into the pGL3-basic vector. All primers are 95 °C for 5 min in 5× sodium dodecyl sulfate (SDS) sample buffer and were separated by 10% SDS-PAGE (30 μg each lane); then, the gel was transferred to polyvinylidene fluoride (PVDF) membranes (Millipore, USA) using a Mini Trans-Blot Cell system (Bio-Rad, USA). The mem- Table 1 brane was blocked with 5% non-fat milk for 1.5 h. The primary antibod- Primers used in the expression vector construction. ies were incubated overnight at 4 °C. The membranes were washed and Gene Primer sequence(5′-3′) Product incubated with secondary antibodies for 1 h at room temperature. The length membranes were visualized by ECL (Bio-RAD, USA). Primary antibodies fi Fhl3 S:CGGGATCCGTCATGAGCGAGGCATTTGA 870 bp speci c for FHL3 (Santa Cruz, CA, USA, sc-166,917; 1:200 dilution), A: GCTCTAGATCAGGGGCCTGCTTGGCTG myogenin (Santa Cruz, CA, USA, sc-12,732; 1:200 dilution), MyHC 2a S:GGGGTACCTTTAAAAAGCTCCAAGGTA 2249 bp (Santa Cruz, CA, USA, sc-53,095; 1:1000 dilution), CREB (Santa Cruz, pMyHC S:GGGGTACCAGGTGACAAACAGAAACACA 1303 bp CA, USA, sc-240; 1:200 dilution), β-actin (Boster, China, BM0627; 2a-delection S:GGGGTACCAAACATTTTCCCAAATATCA 1115 bp 1:1000 dilution), MyHC 2b (Hybridoma Bank, University of Iowa, S:GGGGTACCTCAGTGGTTAAAAGTAATTG 515 bp S:GGGGTACCTACAAGTCAAATCGTGATAA 315 bp 10F5; 1:500 dilution), MyHC 1/slow (Abcam, USA, ab11083; 1:6000 di- A:GGCTCGAGACCGCTCCTGCTTCTGTTTT lution), and MyHC 2x (Abcam, USA, ab127539; 1:500 dilution), p300 CREB S:GGGTACCGTCATGACCATGGAATCTGG 1026 bp (Santa Cruz, CA, USA, sc-585; 1:200 dilution), phospho-CREB at ser133 A:GCTCGAGTTAATCTGATTTGTGGCA rabbit monoclonal (Cell Signaling Technology, USA, 87G3), along with MyHC 2a-ChIP S: CTCCTAATGTTGCTACCCTG 89 bp A: AAGTACACTTATGAAGTAGC goat anti-mouse IgG-HRP (Santa Cruz, CA, USA, sc-2005; 1:3000 dilu- Mut-CRE S:TCCTCATGTTGTGGGTACCTCATGACAGAA tion), goat anti-rabbit IgG-HRP (Santa Cruz, CA, USA, sc-2004; 1:3000 A:TTCTGTCATGAGGTACCCACAACATGAGGA dilution), and anti-mouse IgM-HRP (Santa Cruz, CA, USA, sc-2064; S, sense; A, antisense; six red base pairs, enzyme digested sites; two red base pairs, mutant 1:3000 dilution) secondary antibody were used to detect protein sites; underlined, translation start site. expression. 30 μg of lysate was analyzed by Western blotting. For 62 Y. Zhang et al. / Cellular Signalling 28 (2016) 60–73

Table 2 Primers and TaqMan probes used in the Quantitative real-time PCR.

Gene. Primer sequence(5′-3′) Ann. temp(AT) Product length

Fhl3 S: CCATGAGCGAGGCATTTGAC 60 °C 85 bp A: CAGTAGGGGCCACTGTCTGTC P:AATGCAACGAGTCCTTGTACGGCCGCA MyHC 1/slow S: GCCTGGGCTTACCTCTCTATCAC 60 °C 116 bp A: CTTCTCAGACTTCCGCAGGAA P: CGTTTGAGAATCCAAGGCTCA MyHC 2a S: CAGCTGCACCTTCTCGTTTG 60 °C 81 bp A: CCCGAAAACGGCCATCT P: TGAGTTCAGCAGTCATGAG MyHC 2a S: ATGAGCTCCGACGCCGAG 62 °C 506 bp A: TCTGTTAGCATGAACTGGTAGGCG MyHC 2x S: GGACCCACGGTCGAAGTTG 60 °C 70 bp A: GGCTGCGGGCTATTGGTT P:CTAAAGGCAGGCTCTCTCACTGGGCTG MyHC 2b S: CAATCAGGAACCTTCGGAACAC 60 °C 80 bp A: GTCCTGGCCTCTGAGAGCAT P:TGCTGAAGGACACACAGCTGCACCT S: CGTGAAAAGATGACCCAGATCA 60 °C 72 bp β-actin A: CACAGCCTGGATGGCTACGT P: TTGAGACCTTCAACACCCCAGCCATG β-actin S:GCCTCACTGTCCACCTTCCA 59 °C 184 bp A:AGCCATGCCAATGTTGTCTCTT S: CCATCCAGTACATTGAGCGCCTACA 60 °C 241 bp MyoG A: ACGATGGACGTAAGGGAGTGCAGAT

S, sense; A, antisense; P, TaqMan probe. All probes were 5′-labeled with FAM and 3′-labeled with TAMRA.

Table 3 The EMSA reaction system.

NC reaction Sample reaction Cold competitor Mutant cold competitor Super-shift Super-shift (CREB antibody) (pCREB antibody)

Nuclease Free water Up to 10 μlUpto10μlUpto10μlUpto10μlUpto10μlUpto10μl 5× binding buffer 2 μl2μl2μl2μl2μl2μl Extract 0 μl2μl2μl2μl2μl2μl Probe 1 μl1μl1μl1μl1μl1μl Competitor 0 μl0μl1μl0μl0μl0μl Mutant competitor 0 μl0μl0μl1μl0μl0μl Antibody 0 μl0μl0μl0μl10μg10μg Total 10 μl10μl10μl10μl10μl10μl quantitative Western blotting analysis, films were scanned and the treated with 0.25% Triton X-100 at room temperature for 10 min and band signal intensities determined using ImageJ software. The densi- washed twice. They were then incubated in blocking solution (5% tometry values were normalized by the corresponding β-actin densi- bovine serum albumin in PBS) to block nonspecific binding at room tometry values obtained from the same sample. temperature for 2 h. The samples were incubated with mouse monoclonal anti-MyHC 2a (Santa Cruz, CA, USA, sc-53,095; 1:50 dilution), anti- 2.6. Immunofluorescence MyHC 2b (Hybridoma Bank, University of Iowa, 10F5; 1:50 dilution) or anti-MyHC 1/slow (Abcam, USA, ab11083; 1:1000 dilution) and rabbit After transfection with plasmid (1 μg) or siRNA (25 pmol) by polyclonal anti-FHL3 (Santa Cruz, CA, USA, sc-28,692; 1:50 dilution) Lipofectamine 2000 (2 μl) (Invitrogen, USA) and differentiation for antibody at 4 °C overnight. The next day, the cells were washed three 3 days, C2C12 myoblasts cultured in 24-well plate were rinsed with times, and incubated with anti-rabbit-FITC (Beyotime, Jiangsu, China, phosphate-buffer saline (PBS) two times and fixed with 4% paraformal- A0562; 1:500 dilution) at room temperature for 1 h in the dark room. dehyde for 20 min. Next, the cells were washed twice with PBS, and The anti-rabbit-FITC antibody was removed, and then, samples were

Fig. 1. Fhl3 overexpression alters the expression levels of MyHC isoforms during C2C12 differentiation. (A and B) The mRNA and protein expression levels of Fhl3 and four MyHC isoforms after Fhl3 overexpression. C2C12 myoblasts were transfected with pcDNA3.1-FHL3 vector or empty vector when reaching confluence and then differentiated at D0, D2, D4 and D6 in DMEM with 2% horse serum. (A) Quantitative RT-PCR results of Fhl3, MyHC isoform and MyoG mRNA levels after Fhl3 overexpression at D0, D2, D4 and D6. The data are presented as mean ± S.D. (n = 3). *P b 0.05, **P b 0.01. N.S., no significance between the two groups (the same below). (B) Western blotting results of Fhl3 and MyHC isoform protein levels after FHL3 overexpression at D0, D2, D4 and D6. (C) Relative protein expression levels represented by ratio of detected protein to β-actin protein expression level after Fhl3 overexpression at D0, D2, D4 and D6. The quantifications of Western blotting data are presented as the mean ± S.D. (n = 3) (the same below). (D) Immunofluorescence results of Fhl3 and MyHC isoform protein expression levels after Fhl3 overexpression. C2C12 myoblasts overexpressing pcDNA3.1-FHL3 or vector pcDNA3.1 were differentiated for 3 days, and stained with anti-FHL3 antibodies (green), anti-MyHC 2a antibodies (red), anti-MyHC 2b antibodies (red), anti-MyHC 1/slow antibodies (red), and DAPI (blue) and then imaged by fluorescence microscopy. Bars, 200 μm. (E) The protein expression level of MyHC 2a, MyHC 2b and MyHC 1/slow observed in D was determined by ratio of the number of nuclei within MyHC isoforms positive myotubes to the total number of nuclei. Values are shown as means ±SD of three independent experiments. Y. Zhang et al. / Cellular Signalling 28 (2016) 60–73 63 washed three times and incubated with anti-mouse-Alexa Fluor 555 anti- nuclei was stained with 4′, 6-diamidino-2-phenylindole (DAPI) in the body (Beyotime, Jiangsu, China, A0460; 1:500 dilution). After the second- dark room. After washing three times, images were captured with ary antibody was removed, the cells were washed three times and the cell an Olympus IX51-A21PH fluorescence microscope (Olympus, Japan). 64 Y. Zhang et al. / Cellular Signalling 28 (2016) 60–73

Fig. 1. Fhl3 overexpression alters the expression levels of MyHC isoforms during C2C12 differentiation. (A and B) The mRNA and protein expression levels of Fhl3 and four MyHC isoforms after Fhl3 overexpression. C2C12 myoblasts were transfected with pcDNA3.1-FHL3 vector or empty vector when reaching confluence and then differentiated at D0, D2, D4 and D6 in DMEM with 2% horse serum. (A) Quantitative RT-PCR results of Fhl3, MyHC isoform and MyoG mRNA levels after Fhl3 overexpression at D0, D2, D4 and D6. The data are presented as mean ± S.D. (n = 3). *P b 0.05, **P b 0.01. N.S., no significance between the two groups (the same below). (B) Western blotting results of Fhl3 and MyHC isoform protein levels after FHL3 overexpression at D0, D2, D4 and D6. (C) Relative protein expression levels represented by ratio of detected protein to β-actin protein expression level after Fhl3 overexpression at D0, D2, D4 and D6. The quantifications of Western blotting data are presented as the mean ± S.D. (n = 3) (the same below). (D) Immunofluorescence results of Fhl3 and MyHC isoform protein expression levels after Fhl3 overexpression. C2C12 myoblasts overexpressing pcDNA3.1-FHL3 or vector pcDNA3.1 were differentiated for 3 days, and stained with anti-FHL3 antibodies (green), anti-MyHC 2a antibodies (red), anti-MyHC 2b antibodies (red), anti-MyHC 1/slow antibodies (red), and DAPI (blue) and then imaged by fluorescence microscopy. Bars, 200 μm. (E) The protein expression level of MyHC 2a, MyHC 2b and MyHC 1/slow observed in D was determined by ratio of the number of nuclei within MyHC isoforms positive myotubes to the total number of nuclei. Values are shown as means ±SD of three independent experiments.

2.7. Dual luciferase assay 2.9. Electrophoretic mobility shift assay (EMSA)

After transfection with MyHC 2a promoter constructs (0.8 μg) C2C12 myoblasts cultured in 15-cm dishes and differentiated for and pcDNA3.1-FHL3 vector (0.8 μg) by Lipofectamine 2000 (2 μl) 3 days were extracted using a Nuclear Extraction Kit (Active Motif, CA, (Invitrogen, USA), C2C12 myoblasts were differentiated for 2 days, USA) for EMSA. Double-stranded oligonucleotides (Sangon, Shanghai, cultured in a 24-well plate, washed with PBS, lysed in 100 μloflysis China) corresponding to the CRE within the MyHC 2a promoter buffer, and then, the cells were assayed for promoter activity using a were synthesized and annealed into double strands. The DNA dual luciferase reporter assay system (Promega, USA). The enzymatic binding activity of CREB protein was detected by a chemilumines- activity of luciferase was measured using a PerkinElmer 2030 Multilabel cent EMSA Kit (Beyotime, Jiangsu, China). The reactions were ana- Reader (PerkinElmer). To normalize the transfection efficiency, the cells lyzed by electrophoresis in 6.0% polyacrylamide gels at 100 V for were transfected with 0.04 μg of the Renilla luciferase reporter plasmid 1 h and were then transferred to a nylon membrane. The dried (pRL-TK, Promega, USA). nylon was visualized using ECL (Bio-Rad, USA). The antibodies were specific for CREB rabbit monoclonal (Cell Signaling Technology, USA, 48H2), phospho-CREB at ser133 rabbit monoclonal (Cell Signal- 2.8. Immunoprecipitation ing Technology, USA, 87G3). The DNA-binding reaction system and double-stranded oligonucleotides are listed in Table 3 and Fig. 5D, C2C12 myoblasts were seeded in 10-cm dishes and differentiated respectively. for 3 days. Cells were harvested and lysed in 1 ml lysis buffer (Sangon, Shanghai, China) and protease inhibitor (Sangon, Shanghai, 2.10. Chromatin immunoprecipitation (ChIP) assay China). The lysate was centrifuged to remove insoluble components and incubated with either anti-FHL3 monoclonal antibody (Santa C2C12 myoblasts were seeded in 10-cm dishes. ChIP assays were Cruz, CA, USA sc-166,917), phospho-CREB at ser133 rabbit monoclonal performed following the transfection of the pcDNA3.1-FHL3 plasmid (Cell Signaling Technology, USA, 87G3), anti-MyoD monoclonal anti- into C2C12 myoblasts and then differentiation for 3 days using the body (Santa Cruz, CA, USA sc-166917X) or IgG antibody (Beyotime, ChIP Assay Kit (Beyotime, Jiangsu, China). Briefly, after crosslinking Jiangsu, China) overnight at 4 °C in the presence of Protein A + G the chromatin with 1% formaldehyde at 37 °C for 20min and neutral- Agarose beads (Beyotime, Jiangsu, China) after removing 25 μl izing with glycine for 5 min at room temperature, C2C12 myoblasts lysates for the input control. The beads were washed four times were washed with cold PBS, scraped and collected. Nuclear lysates using lysis buffer. The proteins were analyzed by Western blotting were sonicated 15 times for 10 s with 10 s intervals on ice water as described above. using a Sonics VCX 130 (Sonics, USA). The chromatin complex was

Fig. 2. Fhl3 knockdown alters the levels of MyHC isoforms during C2C12 differentiation. (A and B) The mRNA and protein expression levels of Fhl3 and MyHC isoforms after Fhl3 knockdown. C2C12 myoblasts were transfected with siRNA oligonucleotides (siFHL3) or a negative control (NC) and were then differentiated at D0, D2, D4 and D6 in DMEM with 2% horse serum. (A) Quantitative RT-PCR results of Fhl3, MyHC isoforms and MyoG mRNA levels after Fhl3 knockdown at D0, D2, D4 and D6. (B) Western blotting results of Fhl3, MyHC isoforms and MyoG protein levels after Fhl3 knockdown at D0, D2, D4 and D6. (C) Relative protein levels of Fhl3, MyHC 2a, MyHC 2b, MyHC 1/slow and MyHC 2x after Fhl3 knockdown at D0, D2, D4 and D6. (D) Immunoﬂuorescence results of Fhl3 and MyHC isoform protein expression level after Fhl3 knockdown. Transfected C2C12 myoblasts were differentiated for 3 days, and stained with anti-FHL3 antibodies (green), anti-MyHC 2a antibodies (red), anti-MyHC 2b antibodies (red), anti-MyHC 1/slow antibodies (red), and DAPI (blue) and were imaged by ﬂuo- rescence microscopy. Bars, 200 μm. (E) The protein expression level of MyHC 2a, MyHC 2b and MyHC 1/slow observed in D was determined by ratio of the number of nuclei within MyHC isoforms positive myotubes to the total number of nuclei. Values are shown as means ±SD of three independent experiments. Y. Zhang et al. / Cellular Signalling 28 (2016) 60–73 65 immunoprecipitated at 4 °C overnight with antibodies against CREB 87G3) and normal mouse IgG (Beyotime, Jiangsu, China). Immuno- rabbit monoclonal (Cell Signaling Technology, USA, 48H2), phospho- precipitated products were collected after incubation with Protein CREB at ser133 rabbit monoclonal (Cell Signaling Technology, USA, A + G Agarose beads. The beads were washed, and the bound 66 Y. Zhang et al. / Cellular Signalling 28 (2016) 60–73

chromatin was eluted in ChIP Elution Buffer. Then the proteins were 3.2. FHL3 represses the expression of MyHC 1/slow through functional digested with Proteinase K for 4 h at 45 °C. The DNA was purified interaction with MyoD, but promotes the MyHC 2a expression by using phenol/chloroform. PCR was performed to analyze the DNA other mechanism fragments using 2.0% Agarose gel. The primers are listed in Table 1. MyoD is a bHLH transcription factor and binds to E-box sites of promoters of muscle-specificgenes[35]. The promoters of MyHC 2a, MyHC 2.11. Statistical analysis 2b and MyHC 1/slow all contain several E-box elements analyzed using the TFsearch software (http://www.cbrc.jp/research/db/TFSEARCH. Statistical analysis was performed using one-way ANOVA. The data html). The previous study showed that FHL3 could interact with are presented as the means ± S.D., and the level for statistical signifi- MyoD and thereby inhibited MyoD transcriptional activation [32].To cance was set at p b 0.05. determine whether FHL3 regulated the expression of MyHC isoforms through MyoD, MyoD was knocked down and Fhl3 was overexpressed in C2C12 cells. The results showed that MyoD could increase the expres- 3. Results sion of MyHC 2a, MyHC 2b and MyHC 1/slow (Fig. 3A and B). MyHC 2a and MyHC 2b expression was also increased by Fhl3 overexpression 3.1. FHL3 up-regulates the expression of MyHC 2a and MyHC 2b, and after knockdown of MyoD, but there was no effect of Fhl3 overexpres- down-regulates the expression of MyHC 1/slow sion on the MyHC 1/slow expression after MyoD knockdown (Fig. 3A and B). To further confirm this, Fhl3 was knocked down and MyoD To investigate the effects of FHL3 on the expression of four was overexpressed in C2C12 cells. MyoD overexpression stimulated MyHC isoforms during C2C12 myoblast differentiation, total RNA MyHC 1/slow expression, and knockdown of Fhl3 strengthened the pro- and protein were extracted from C2C12 myoblasts that were moting function of MyoD (Fig. 3C and D), suggesting that FHL3 re- transfected with pcDNA3.1-FHL3 or empty plasmid (pcDNA3.1) presses the expression of MyHC 1/slow through inhibition of MyoD and then differentiated at day 0 (D0), day 2 (D2), day 4 (D4), or transcriptional activity. We also performed co-expression of Fhl3 and day 6 (D6) in DMEM with 2% horse serum. Real-time PCR, Western MyoD in C2C12 myoblasts. Although MyoD or Fhl3 overexpression in- blotting and immunofluorescence were performed to detect gene creased MyHC 2a and MyHC 2b gene expression, respectively, the expression of Fhl3 and MyHC isoforms. The expression level of Fhl3 MyHC 2a expression levels after co-expression of Fhl3 and MyoD did was increased with overexpression by at least 50% compared not significantly changed compared with the control (Fig. 3E and F). with the control (Fig. 1A and B). The expression of MyHC 2a and This phenomenon may be caused by the following reasons: the exoge- MyHC 2b were up-regulated, and that of MyHC 1/slow was down- nous expressed FHL3 and MyoD may interact with each other which re- regulated after overexpression of Fhl3, while no significant change sulted in no excess FHL3 and MyoD to regulate MyHC 2a expression, was found for MyHC 2x expression (Fig. 1A-E). The mRNA level of thereby the activating effects of two proteins were abolished. Based the marker gene myogenin (MyoG) decreased (Fig. 1A), which was on the above results, we concluded that FHL3 repressed the expression in agreement with the previous study [32].TheknockdownofFhl3 of MyHC 1/slow through MyoD, but regulates the expression of MyHC 2a decreased the expression of MyHC 2a and MyHC 2b but increased and MyHC 2b through other transcription factors instead of MyoD. the expression of MyHC 1/slow at both the mRNA and protein levels (Fig. 2A-E). The results of immunofluorescence confirmed those 3.3. FHL3 up-regulates MyHC 2a expression primarily through improving of real-time PCR and Western blotting. In summary, FHL3 up- CREB transcription activity regulated the expression of MyHC 2a and MyHC 2b and down- regulated the expression of MyHC 1/slow, whereas no significant To elucidate how FHL3 regulated the expression of MyHC 2a effects on MyHC 2x expression were found. and MyHC 2b, we generated a series of MyHC 2a promoter deletions Y. Zhang et al. / Cellular Signalling 28 (2016) 60–73 67

Fig. 3. The co-expression of Fhl3 and MyoD affectstheproteinexpressionlevelsofMyHC 2a, MyHC 2b and MyHC 1/slow. (A) Western blotting results of MyHC 2a, MyHC 2b and MyHC 1/slow protein expression level after Fhl3 overexpression and MyoD knockdown. C2C12 myoblasts were co-transfected with pcDNA3.1-FHL3 and siRNA oligonucleotides and (indicated at the bottom) and differentiated for 3 days. The cell lysates were subject to Western blotting analysis with anti-FHL3, anti-MyoD, anti-MyHC 2a, anti-MyHC 2b and anti-MyHC 1/slow antibodies (indicated at the left). (B) Relative protein expression levels of MyHC 2a, MyHC 2b and MyHC 1/slow after Fhl3 overexpression and MyoD knockdown. (C) Western blotting results of MyHC 1/slow after MyoD overexpression and Fhl3 knockdown. C2C12 myoblasts were co-transfected with pcDNA3.1-MyoD and siRNA oligonucleotides and (indicated at the bottom) and differentiated for 3 days. The cell lysates were subject to Western blotting analysis with anti-FHL3, anti-MyoD and anti-MyHC1/ slow antibodies (indicated at the left). (D) Relative protein expression levels of MyHC1/slow after MyoD overexpression and Fhl3 knockdown. (E) Western blotting results of MyHC 2a and MyHC 2b protein expression levels after Fhl3 and MyoD overexpression. C2C12 myoblasts were co-transfected with pcDNA3.1-FHL3 and pcDNA3.1-MyoD (indicated at the bottom) and were then differentiated for 3 days. Cell lysates were subject to Western blotting analysis with anti-FHL3, anti-MyoD, anti-MyHC 2a, and anti-MyHC 2b antibodies (indicated to the left). (F) Relative protein expression levels of MyHC 2a and MyHC 2b after Fhl3 and MyoD overexpression.

driving the transcription of the luciferase reporter gene (Fig. 4A). Co- nucleotides −909 bp and +206 bp (1115 bp) containing cAMP re- transfection with pcDNA3.1 or pcDNA3.1-FHL3 and luciferase reporter sponse elements (CRE) was improved by Fhl3 overexpression indicated that the transcription activity of the region between (Fig. 4A). To conﬁrm whether CREB participated in the regulation of 68 Y. Zhang et al. / Cellular Signalling 28 (2016) 60–73

MyHC 2a expression, Creb overexpression and knockdown was per- MyHC 2a compared with the control (Fig. 4F and G), whereas no sig- formed in C2C12 myoblasts. The results showed that CREB also pro- nificant difference for MyHC 1/slow expression was observed after motedtheproteinexpressionofMyHC 2a (Fig. 4BandC).Basedon overexpression or knockdown of Creb (Fig. 4D-G), suggesting that the previous study that FHL3 could strongly interact with CREB and CREB plays an important role in the regulation of MyHC 2a expres- function as a coactivator of CREB protein [18],wehypothesizethat sion by FHL3. To further determine whether FHL3 increased the pro- FHL3 may regulate MyHC 2a through CREB mediated transcription. moter transcriptional activity of MyHC 2a through CREB binding To confirm this, Fhl3 and Creb were co-expressed in C2C12 myo- sites, we analyzed the transcriptional activity of MyHC 2a promoter blasts. Overexpression of Fhl3 or Creb promoted MyHC 2a expression, with the CRE mutation after overexpression of Fhl3 and Creb.The and co-expression of two proteins further enhanced MyHC 2a expres- transcriptional activity of the MyHC 2a promoter was enhanced by sion (Fig. 4D and E). By contrast, after transfection of Creb siRNA frag- Fhl3 or Creb overexpression, whereas no significant difference was ments, overexpression of Fhl3 did not increase the expression of found for the activity of the MyHC 2a promoter with the mutation at

Fig. 4. FHL3 up-regulates MyHC 2a expression primarily through improving CREB transcription activity. (A) Promoter activities of a series of deleted constructs determined by luciferase assay after Fhl3 overexpression and differentiation for 2 days. Left panel, schematic representation of the deleted fragments linked with the luciferase gene in the pGL3 vector. The nucleotides are numbered relative to the translation start site that was assigned as +1. Schematic diagram of MyHC 2a promoter deletion containing the CRE and E-box are displayed. Right panel, the relative activities of a series of deleted constructs of the pGL3-MyHC 2a construct determined by luciferase assay. (B) Western blotting results of MyHC 2a protein expression levels after Creb overexpression. C2C12 myoblasts were transfected with pcDNA3.1-CREB and empty vector and were then differentiated at D2 and D4; relative protein levels of MyHC 2a was also represented. (C) Western blotting results of MyHC 2a protein expression levels after Creb knockdown. C2C12 myoblasts were transfected with Creb siRNA (siCREB) and NC and were then differentiated at D2 and D4; relative protein levels of MyHC 2a was also represented. (D) Western blotting results of MyHC 2a protein levels after Fhl3 and Creb co-expression. C2C12 myoblasts were co-transfected with pcDNA3.1-FHL3 and pcDNA3.1-CREB (indicated at the bottom) and were then differentiated for 3 days. Cell lysates were subject to Western blotting with anti-FHL3, anti-CREB, anti-pCREB anti-MyHC 2a, and anti-MyHC 1/slow antibodies (indicated at the left). (E) Relative protein levels of MyHC 2a and MyHC 1/slow after Fhl3 and Creb co-expression. (F) Western blotting results of MyHC 2a and MyHC 1/slow protein expression levels after co-transfection of pcDNA3.1-FHL3 and siCREB and differentiation for 3 days in C2C12 myoblasts. (G) Relative protein expression levels of MyHC 2a and MyHC 1/slow after co-transfection of pcDNA3.1-FHL3 and siCREB. (H) Promoter activities of MyHC 2a (−909/+206) or MyHC 2a with a mutation at CRE (MutCRE -909/+206) was determined by a luciferase assay after overexpression of Fhl3 or Creb and differentiation for 2 days. The data are presented as mean ± S.D. (n = 6). Y. Zhang et al. / Cellular Signalling 28 (2016) 60–73 69

Fig. 4 (continued).

CRE (Fig. 4H), indicating that FHL3 was involved in the CREB- the C2C12 cells with 10 μM, 20 μM, 30 μM, 40 μM, 50 μM H-89, mediated gene expression. andselected20μM as the optimal working concentration. CREB phosphorylation levels were decreased by H-89 but were not 3.4. FHL3 forms a complex with pCREB to regulate MyHC 2a expression changed by Fhl3 overexpression (Fig. 5A). Thus we inferred that FHL3 might interact with pCREB to regulate MyHC 2a expression. CREB is an important transcriptional factor activated by Fhl3 overexpression up-regulated the expression level of MyHC phosphorylation at ser-133 [36]. cAMP-activated protein kinase A 2a, which was abolished by the PKA inhibitor H-89 (Fig. 5A). These (PKA) can lead to activation of the transcriptional factor CREB [37]. results showed that the effect of FHL3 on MyHC 2a expression was de- To detect whether FHL3 was involved in CREB phosphorylation- pendent on CREB protein phosphorylation by PKA. To investigate mediated MyHC 2a expression, we used the PKA inhibitor H-89 whether FHL3 could form a complex with CREB to regulate to decrease the CREB phosphorylation level [38].Wetreated MyHC 2a expression, we explored the interaction of FHL3 with 70 Y. Zhang et al. / Cellular Signalling 28 (2016) 60–73 Y. Zhang et al. / Cellular Signalling 28 (2016) 60–73 71 other proteins in vivo. CREB and p300 were all co-immunoprecipitated by the antibody against FHL3 but were not co-immunoprecipitated by control IgG, pCREB also was co-immunoprecipitated by FHL3 antibody and there was weak non-speciﬁc binding with IgG (Fig. 5B), indicating that FHL3 participated in the complex with pCREB and p300. Co-immunoprecipitation experiments were also performed with anti-pCREB and anti-MyoD. The results conﬁrmed that FHL3 could interact with both pCREB and MyoD (Fig. 5C).

3.5. FHL3 affects the binding capacity of pCREB to the CRE within the MyHC 2a promoter

The binding capacity of CREB with promoter sequences was analyzed by EMSA and ChIP assay. As shown in Fig. 5D, the incubation of nuclear extracts of C2C12 myoblasts with CRE probes formed a DNA– protein complex (Lane 2). The complex became weaker with 1× cold probes in the mixture (Lane 3). However, the complex did not change in the mutation cold probe reaction (Lane 4). No super-shift band was detected after the addition of anti-CREB to the mixture (Lane 5), while the addition of anti-pCREB to the binding reaction caused a super- shift band (Lane 6). These results suggested that the CRE of the MyHC 2a promoter was capable of binding to pCREB in vitro. ChIP was also performed on C2C12 myoblasts to determine whether FHL3 regulated the Fig. 6. Model for the regulation of MyHC 2a and MyHC 1/slow expression by FHL3 in the cell binding capacity of CREB to the CRE of the MyHC 2a promoter in vivo. nucleus. (A) FHL3 forms a complex with CREB and p300 to positively regulate MyHC 2a ex- The CRE of the MyHC 2a promoter was capable of binding to both pression. MyoD also up-regulates MyHC 2a expression. (B) FHL3 physically interacts with CREB protein and pCREB in vivo, and pCREB had a stronger binding ca- MyoD to inhibit MyHC 1/slow expression. pacity compared with CREB (Fig. 5E), which was in agreement with observation points to a novel mechanism by which FHL3 differentially fi fi the EMSA result. Knockdown of Fhl3 decreased the binding capacity of regulates muscle ber speci c gene expression. α pCREB to the CRE of the MyHC 2a promoter (Fig. 5F). Several signaling pathways, including calcineurin [43],PGC-1 [15], and Ras/ERK-1/2 signaling [44], have been implicated in skeletal muscle fast-to-slow fiber-type shift and slow fiber gene expression. The MyoG 4. Discussion gene is highly expressed in slow oxidative muscles [45]. The expression of MyHC 1/slow and MyoG shows the same upward trend after Ca2+ and MyHC has been classified into 9 to 11 classes [39]. Class II is also continuous mild heat stress stimulation which induces a fast-to-slow called “conventional” MyHCs and includes two developmental isoforms fiber-type shift, but this observation was different for other myogenic (MyHC-embryonic and MyHC-perinatal) and four adult skeletal muscle factors [46]. The present results also showed a similar expression isoforms (MyHC 2a, MyHC 2b, MyHC 2x, and MyHC 1/slow) [40].No trend of both MyHC 1/slow and MyoG after Fhl3 overexpression and reports have been published regarding the effects of FHL3 on the knockdown, and synergistic effects for MyHC 1/slow expression were expression of any individual member of the MyHC family thus far. In detected after co-transfection of FHL3 siRNA and MyoD overexpression this study, we observed that FHL3 differentially regulates the vector. As MyoD also positively regulates the expression of the MyHC 1/ expression of MyHC isoforms through interactions with MyoD and slow gene by binding to E-box in the promoter [47],weconcludedthat pCREB, suggesting that FHL3 increases fast fiber specific gene expres- the regulation mechanism of MyHC 1/slow expression by FHL3 agreed sion and decreases slow fiber specific gene expression. Adult MyHC iso- with the previous results that FHL3 negatively regulates MyoG expres- forms are highly expressed in C2C12 cells differentiated for day 4 and sion by inhibiting MyoD transcriptional activity [32]. day 6, and hardly expressed in C2C12 myoblasts [41,42]. In this study, The studies on the regulation of MyHC 2a, MyHC 2x and MyHC 2b ex- the Fhl3 overexpression peak occurred at 1–2 days of transfection, and pression were primarily focused on the specific signaling pathways, then began to decline afterward. Therefore, we infer that there may be such as p38 MAPK, IGF-1 and intracellular calcium [13,48,49]. CREB is the hysteresis effects of FHL3 change at early differentiation on the a transcription factor that can be activated by growth factor and kinases MyHC proteins at later differentiation. In general, FHL3 inhibits expres- [50,51] and that can regulate gene expression important to the develop- sion of total MyHC protein and myotube formation [32]. Our study ment of cardiovascular diseases [52,53], immune responses [54], and showed that FHL3 promoted MyHC 2a and MyHC 2b expression, and memory formation [55]. CREB promotes the transcription of target inhibited MyHC 1/slow expression after C2C12 myoblasts differentia- genes via the recruitment of the p300/CBP coactivators [56], and the tion. It can be inferred that the other members of MyHC family such as binding of CREB to p300 requires the phosphorylation of CREB at Ser- MyHC-embryonic and MyHC-perinatal may be inhibited by FHL3, 133 [57]. Emerging evidence has also revealed functions of CREB in which resulted in a decreased expression in total MyHC protein. This

Fig. 5. FHL3 promotes MyHC 2a expression through its interaction with pCREB and p300 in C2C12 myoblasts. (A) The effects of FHL3 on MyHC 2a expression protein levels through pCREB. After Fhl3 overexpression, C2C12 myoblasts were differentiated for 3 days and then treated or not with 20 μM H-89 for 1 h before harvest, and the control cells were treated with 0.1% DMSO (indicated at the bottom). Protein expression levels of pCREB and MyHC 2a were detected by Western blotting; Relative protein expression levels of pCREB and MyHC 2a were also represented (B) The immunoprecipitation results showing the interaction among FHL3, p300 and pCREB in vivo. C2C12 myoblasts were differentiated for 3 days, harvested and pull-downed by antibodies. Lane 1: the analysis of total cell lysates before immunoprecipitation to verify expression of CREB and pCREB. Lane 2: immunoprecipitation results with anti IgG monoclonal antibody as a negative control; Lane 3: immunoprecipitation results with anti-FHL3 monoclonal antibody. (C) The immunoprecipitation results showing the interaction of FHL3 with pCREB and MyoD. Lane 3: immunoprecipitation results with anti-pCREB monoclonal antibody; Lane 4: immunoprecipitation results with anti-MyoD monoclonal antibody. (D) EMSA results showing the binding of CREB and pCREB to the MyHC 2a promoter in C2C12 myoblasts differentiated for 3 days. The probes were incubated with nuclear extract in the absence or presence of 1-fold excess of various competitor probes (mutant or non-labeled probe) or antibodies (anti-CREB or pCREB). The specificDNA–protein complex and the super- shift complex (DNA–protein–antibody complex) bands are indicated by arrows. The sequences of various probes are shown under the panel. *, non-specific binding. (E) ChIP assay results showing the interaction of CREB and pCREB with the MyHC 2a promoter in vivo in C2C12 myoblasts differentiated for 3 days. Immunoprecipitated DNA was amplified by PCR for 35 cycles. Total chromatin was used as the input, and normal mouse IgG was used as the negative control (the same as below). (F) ChIP assay results showing the effectoftheFhl3 knockdown on the binding capacity of pCREB to the MyHC 2a promoter. Immunoprecipitated DNA was amplified by PCR for 30 cycles. 72 Y. Zhang et al. / Cellular Signalling 28 (2016) 60–73 myogenesis, muscle degeneration and regeneration in muscular dystro- [6] Y.C. Ryu, Y.M. Choi, B.C. Kim, Variations in metabolite contents and protein fi – denaturation of the longissimus dorsi muscle in various porcine quality classi ca- phy [58 60]. However, the role of CREB protein in the expression regu- tions and metabolic rates, Meat Sci. 71 (2005) 522–529. lation of different MyHC isoforms has not been studied. Our results [7] J.E. Hammelman, B.C. Bowker, A.L. Grant, J.C. Forrest, A.P. Schinckel, D.E. Gerrard, showed that pCREB played a more important role in the regulation of Early postmortem electrical stimulation simulates PSE pork development, Meat Sci. 63 (2003) 69–77. MyHC 2a expression than unphosphorylated CREB, and FHL3 regulated [8] B. Wei, W. Dui, D. Liu, Y. Xing, Z. Yuan, G. Ji, MST1, a key player, in enhancing fast MyHC 2a expression through its interaction with pCREB and possible skeletal muscle atrophy, BMC Biol. 11 (2013) 12. interaction with p300. This conclusion was different from the previous [9] A. Jaworowski, M.M. Porter, A.M. Holmbäck, D. Downham, J. Lexell, Enzyme report in which FHL3 did not require phosphorylation of CREB at ser- activities in the tibialis anterior muscle of young moderately active men and women: relationship with body composition, muscle cross-sectional area and 133 to stimulate transcriptional activity [18]. The present data also fibre type composition, Acta Physiol. Scand. 176 (2002) 215–225. showed that MyoD and FHL3 promoted MyHC 2a expression with differ- [10] E.R.Barton-Davis,D.I.Shoturma,A.Musaro,N.Rosenthal,H.L.Sweeney,Viral ent molecular mechanisms. FHL3 played the predominant role in the mediated expression of insulin-like growth factor I blocks the aging-related loss of skeletal muscle function, Proc. Natl. Acad. Sci. U. S. A. 95 (1998) up-regulation of MyHC 2a expression by improving pCREB transcrip- 15603–15607. tional activity rather than changing MyoD transcriptional activity. In ad- [11] G. Grimby, B. Danneskiold-Samsøe, K. Hvid, B. Saltin, Morphology and enzymatic dition, MyoD also could bind to pCREB and FHL3 [32,61], a tertiary capacity in arm and leg muscles in 78–81 year old men and women, Acta Physiol. Scand. 115 (1982) 125–134. complex of FHL3/MyoD/pCREB may exist and play the roles in the [12] M.T. Wheeler, E.C. Snyder, M.N. Patterson, S.J. Swoap, An E-box within the MHC IIB gene regulation by another mechanism. As the promoter of MyHC 2b gene is bound by MyoD and is required for gene expression in fast muscle, Am. also contains CRE, we inferred that FHL3 promoted the MyHC 2b expres- J. Physiol. 276 (1999) C1069–C1078. [13] R.A. Shanely, K.A. Zwetsloot, T.E. Childs, S.J. Lees, R.W. Tsika, F.W. Booth, IGF-I acti- sion with the same mechanism as that of MyHC 2a expression regula- vates the mouse type IIb myosin heavy chain gene, Am. J. Physiol. Cell Physiol. tion, but the molecular mechanism remains to be determined. 297 (2009) C1019–C1027. [14] S.J. Swoap, R.B. Hunter, E.J. Stevenson, H.M. Felton, N.V. Kansagra, J.M. Lang, K.A. Esser, S.C. Kandarian, The calcineurin-NFAT pathway and muscle fiber-type gene ex- 5. Conclusions pression, Am. J. Physiol. Cell Physiol. 279 (2000) C915–C924. [15] J. Lin, H. Wu, P.T. Tarr, C.Y. Zhang, Z. Wu, O. Boss, L.F. Michael, P. Puigserver, E. Isotani, E.N. Olson, B.B. Lowell, R. Bassel-Duby, B.M. Spiegelman, Transcriptional We propose a model whereby FHL3 forms a complex with CREB or co-activator PGC-1a drives the formation of slow-twitch muscle fibres, Nature 418 MyoD to regulate the expression of MyHCs in the nucleus of myoblasts (2002) 793–801. (Fig. 6). Both FHL3 and MyoD participate in the up-regulation of the [16] I. Sánchez-García, T.H. Rabbitts, The LIM domain: a new structural motif found in zinc-finger-like proteins, Trends Genet. 10 (1994) 315–320. MyHC 2a expression, and FHL3 plays the leading role in the up- [17] K.L. Schmeichel, M.C. Beckerle, Molecular dissection of a LIM domain, Mol. Biol. Cell regulation of MyHC 2a expression, primarily through its direct interac- 8(1997)219–230. tion with pCREB. p300 may also participate in the complexes by a pos- [18] G.M. Fimia, D. De Cesare, P. Sassone-Corsi, A family of LIM-only transcriptional coactivators: tissue-specific expression and selective activation of CREB and sible interaction with FHL3 (Fig. 6A), as FHL3 does not directly bind to CREM, Mol. Cell. Biol. 20 (2000) 8613–8622. p300 [29]. In contrast, FHL3 down-regulates the expression of MyHC [19] B. Zuo, Y.Z. Xiong, C.Y. Deng, Y.H. Su, J. Wang, M.G. Lei, F.E. Li, S.W. Jiang, R. Zheng, CDNA cloning, genomic structure and polymorphism of the porcine FHL3 gene, 1/slow possibly through decreasing the binding capacity of MyoD to – fi Anim. Genet. 35 (2004) 230 233. the E-box of its promoter (Fig. 6B). Regulating muscle ber composition [20] B. Zuo, Y.Z. Xiong, H. Yang, J. Wang, Full-length cDNA, expression pattern and asso- is possible by altering Fhl3 expression levels, thereby improving meat ciation analysis of the porcine FHL3 gene, Asian-Aust. J. Anim. Sci. 20 (2007) quality and providing therapy for muscle disease. 1473–1477. [21] J.M. Müller, E. Metzger, H. Greschik, A.K. Bosserhoff, L. Mercep, R. Buettner, R. Schüle, Our results elucidated a novel molecular mechanism regarding the The transcriptional coactivator FHL2 transmits Rho signals from the cell membrane regulation of MyHCs expression by FHL3. FHL3 might be required into the nucleus, EMBO J. 21 (2002) 736–748. for repression of MyHC 1/slow expression during C2C12 myoblasts dif- [22] A.P. Meeson, X. Shi, M.S. Alexander, R.S. Williams, R.E. Allen, N. Jiang, I.M. Adham, fi S.C. Goetsch, R.E. Hammer, D.J. Garry, Sox15 and Fhl3 transcriptionally coactivate ferentiation. FHL3 promoted the expression of fast ber genes (MyHC Foxk1 and regulate myogenic progenitor cells, EMBO J. 26 (2007) 1902–1912. 2a and MyHC 2b) through its interaction with pCREB but inhibited the [23] K. Takahashi, C. Matsumoto, C. Ra, FHL3 negatively regulates human high-affinity expression of slow fiber genes (MyHC 1/slow)byinhibitingMyoDtran- IgE receptor beta-chain gene expression by acting as a transcriptional co-repressor – scriptional activity. of MZF-1, Biochem. J. 386 (2005) 191 200. [24] J. Turner, H. Nicholas, D. Bishop, J.M. Matthews, M. Crossley, The LIM protein FHL3 binds basic Kruppel-like factor/Kruppel-like factor 3 and its co-repressor C- terminal-binding protein 2, J. Biol. Chem. 278 (2003) 12786–12795. Acknowledgments [25] W. Xia, W. Fu, L. Cai, H. Kong, X. Cai, J. Liu, Y. Wang, M. Zou, D. Xu, Identification and characterization of FHL3 as a novel angiogenin-binding partner, Gene 504 (2012) – This study was financially supported by the Key National High Tech- 233 237. [26] X. Cai, J. Wang, X. Huang, W. Fu, W. Xia, M. Zou, Y. Wang, J. Wang, D. Xu, Identifica- nology Development Project of China (Project No. 2011AA100301), tion and characterization of MT-1X as a novel FHL3-binding partner, PLoS One 9 the National Natural Science Foundation of P.R. China (Grants (2014), e93723. No. 31001043), the Agricultural Innovation Fund of Hubei Province [27] W. Han, Z. Xin, Z. Zhao, W. Bao, X. Lin, B. Yin, J. Zhao, J. Yuan, B. Qiang, X. Peng, RNA- binding protein PCBP2 modulates glioma growth by regulating FHL3, J. Clin. Invest. 3 (Grant No. 2007-620), the Fundamental Research Funds for the Central (2013) 03–2118. Universities (Grant No. 2014PY038). We also thank Prof. Guoquan Liu [28] C. Niu, Z. Yan, L. Cheng, J. Zhu, H. Zhang, X. Xu, J. Lin, Q. Ye, Downregulation and and Prof. Chunyan Mu for their kind assistance in manuscript writing antiproliferative role of FHL3 in breast cancer, IUBMB Life 63 (2011) 764–767. [29] M.E. Hubbi, D.M. Gilkes, J.H. Baek, G.L. Semenza, Four-and-a-half LIM domain pro- and Dr. Xinyu Wu for technical help in the experiment. teins inhibit transactivation by hypoxia-inducible factor 1, J. Biol. Chem. 287 (2012) 6139–6149. [30] G.M. Fimia, D. De Cesare, P. Sassone-Corsi, CBP-independent activation of CREM and References CREB by the LIM-only protein ACT, Nature 398 (1999) 165–169. [31] I.D. Coghill, S. Brown, D.L. Cottle, M.J. McGrath, P.A. Robinson, H.H. Nandurkar, J.M. [1] S.E. Handel, N.C. Stickland, The growth and differentiation of porcine skeletal muscle Dyson, C.A. Mitchell, FHL3 is an actin-binding protein that regulates alpha-actinin- fibre types and the influence of birth weight, J. Anat. 152 (1987) 107–119. mediated actin bundling: FHL3 localizes to actin stress fibers and enhances cell [2] D. Pette, Training effects on the contractile apparatus, Acta Physiol. Scand. 162 spreading and stress fiber disassembly, J. Biol. Chem. 278 (2003) 24139–24152. (1998) 367–376. [32] D.L. Cottle, M.J. McGrath, B.S. Cowling, I.D. Coghill, S. Brown, C.A. Mitchell, FHL3 [3] L. Lefaucheur, P. Ecolan, L. Plantard, N. Gueguen, New insights into muscle fiber binds MyoD and negatively regulates myotube formation, J. Cell Sci. 120 (2007) types in the pig, J. Histochem. Cytochem. 50 (2002) 719–730. 1423–1435. [4] H. Hu, J. Wang, R. Zhu, J. Guo, Y. Wu, Effect of myosin heavy chain composition [33] N. da Costa, J. Edgar, P.T. Ooi, Y. Su, J.D. Meissner, K.C. Chang, Calcineurin differential- of muscles on meat quality in Laiwu pigs and Duroc, Sci. China C Life Sci. 51 ly regulates fast myosin heavy chain genes in oxidative muscle fibre type conver- (2008) 127–132. sion, Cell Tissue Res. 329 (2007) 515–527. [5] Y.C. Ryu, Y.M. Choi, S.H. Lee, H.G. Shin, J.H. Choe, J.M. Kim, K.C. Hong, B.C. Kim, Com- [34] K.J. Livak, T.D. Schmittgen, Analysis of relative gene expression data using real-time paring the histochemical characteristics and meat quality traits of different pig quantitative PCR and the 2(−delta delta C(T)) method, Methods 25 (2001) breeds, Meat Sci. 80 (2008) 363–369. 402–408. Y. Zhang et al. / Cellular Signalling 28 (2016) 60–73 73

[35] A.B. Lassar, R.L. Davis, W.E. Wright, T. Kadesch, C. Murre, A. Voronova, D. Baltimore, [48] J.D. Meissner, K.C. Chang, H.P. Kubis, A.R. Nebreda, G. Gros, R.J. Scheibe, The p38 H. Weintraub, Functional activity of myogenic HLH proteins requires hetero- alpha/beta mitogen-activated protein kinases mediate recruitment of CREB- oligomerization with E12/E47-like proteins in vivo, Cell 66 (1991) 305–315. binding protein to preserve fast myosin heavy chain IId/x gene activity in myotubes, [36] G.Y. Wu, K. Deisseroth, R.W. Tsien, Activity-dependent CREB phosphorylation: con- J. Biol. Chem. 282 (2007) 7265–7275. vergence of a fast, sensitive calmodulin kinase pathway and a slow, less sensitive [49] D.L. Allen, L.A. Leinwand, Intracellular calcium and myosin isoform transitions. Cal- mitogen-activated protein kinase pathway, Proc. Natl. Acad. Sci. U. S. A. 98 (2001) cineurin and calcium-calmodulin kinase pathways regulate preferential activation 2808–2813. of the IIa myosin heavy chain promoter, J. Biol. Chem. 277 (2002) 45323–45330. [37] G.Li-Sha,C.Yi-He,Z.Na-Dan,Z.Teng,L.Yue-Chun,Effectsofcarvediloltreatmenton [50] X.Y.Li,X.R.Zhan,X.M.Liu,X.C.Wang,CREBisaregulatorytargetfortheproteinkinase cardiac cAMP response element binding protein expression and phosphorylation in Akt/PKB in the differentiation of pancreatic ductal cells into islet beta-cells mediated by acute coxsackievirus B3-induced myocarditis, BMC Cardiovasc. Disord. 13 (2013) 100. hepatocyte growth factor, Biochem. Biophys. Res. Commun. 404 (2011) 711–716. [38] B.Wang,L.Zhu,S.Sui,C.Sun,H.Jiang,D.Ren,Cilostazolinducesmitochondrialfatty [51] A.J. Shaywitz, M.E. Greenberg, CREB: a stimulus-induced transcription factor activated acid beta-oxidation in C2C12 myotubes, Biochem. Biophys. Res. Commun. 447 by a diverse array of extracellular signals, Annu. Rev. Biochem. 68 (1999) 821–861. (2014) 441–445. [52] R.C. Fentzke, C.E. Korcarz, R.M. Lang, H. Lin, J.M. Leiden, Dilated cardiomyopathy in [39] A. Weiss, D. McDonough, B. Wertman, L. Acakpo-Satchivi, K. Montgomery, R. transgenic mice expressing a dominant-negative CREB transcription factor in the Kucherlapati, L. Leinwand, Organization of human and mouse skeletal myosin heart, J. Clin. Invest. 101 (1998) 2415–2426. heavy chain gene clusters is highly conserved, Proc. Natl. Acad. Sci. U. S. A. 96 [53] C.S. McPherson, T. Featherby, E. Krstew, A.J. Lawrence, Quantification of phosphory- (1999) 2958–2963. lated cAMP-response element-binding protein expression throughout the brain of [40] S. Schiaffino, C. Reggiani, Molecular diversity of myofibrillar proteins: gene regula- amphetamine-sensitized rats: activation of hypothalamic orexin A-containing neu- tion and functional significance, Physiol. Rev. 76 (1996) 371–423. rons, J. Pharmacol. Exp. Ther. 323 (2007) 805–812. [41] E. Zebedin, W. Sandtner, S. Galler, J. Szendroedi, H. Just, H. Todt, K. Hilber, Fiber type [54] A.Y. Wen, K.M. Sakamoto, L.S. Miller, The role of the transcription factor CREB in conversion alters inactivation of voltage-dependent sodium currents in murine immune function, J. Immunol. 185 (2010) 6413–6419. C2C12 skeletal muscle cells, Am. J. Physiol. Cell Physiol. 287 (2004) C270–C280. [55] J.C. Yin, J.S. Wallach, M. Del Vecchio, E.L. Wilder, H. Zhou, W.G. Quinn, T. Tully, Induc- [42] D.M. Brown, T. Parr, J.M. Brameld, Myosin heavy chain mRNA isoforms are tion of a dominant negative CREB transgene specifically blocks long-term memory expressed in two distinct cohorts during C2C12 myogenesis, J. Muscle Res. Cell in Drosophila, Cell 79 (1994) 49–58. Motil. 32 (2012) 383–390. [56] R.P. Kwok, J.R. Lundblad, J.C. Chrivia, J.P. Richards, H.P. Bächinger, R.G. Brennan, S.G. [43] U. Delling, J. Tureckova, H.W. Lim, L.J. De Windt, P. Rotwein, J.D. Molkentin, A Roberts, M.R. Green, R.H. Goodman, Nuclear protein CBP is a coactivator for the tran- calcineurin-NFATc3-dependent pathway regulates skeletal muscle differentiation scription factor CREB, Nature 370 (1994) 223–226. and slow myosin heavy-chain expression, Mol. Cell. Biol. 20 (2000) 6600–6611. [57] D. Parker, K. Ferreri, T. Nakajima, V.J. LaMorte, R. Evans, S.C. Koerber, C. Hoeger, M.R. [44] M. Murgia, A.L. Serrano, E. Calabria, G. Pallafacchina, T. Lomo, S. Schiaffino, Ras is in- Montminy, Phosphorylation of CREB at Ser-133 induces complex formation with volved in nerve-activity-dependent regulation of muscle genes, Nat. Cell Biol. 2 CREB-binding protein via a direct mechanism, Mol. Cell. Biol. 16 (1996) 694–703. (2000) 142–147. [58] A.E. Chen, D.D. Ginty, C.M. Fan, Protein kinase A signalling via CREB controls [45] S.M.Hughes,J.M.Taylor,S.J.Tapscott,C.M.Gurley,W.J.Carter,C.A.Peterson,Selective myogenesis induced by Wnt proteins, Nature 433 (2004) 317–322. accumulation of MyoD and myogenin mRNAs in fast and slow adult skeletal muscle [59] R. Berdeaux, N. Goebel, L. Banaszynski, H. Takemori, T. Wandless, G.D. Shelton, M. is controlled by innervation and hormones, Development 118 (1993) 1137–1147. Montminy, SIK1 is a class II HDAC kinase that promotes survival of skeletal [46] T. Yamaguchi, T, Suzuki, H, Arai, S, Tanabe, Y, Atomi, Continuous mild heat stress in- myocytes, Nat. Med. 13 (2007) 597–603. duces differentiation of mammalian myoblasts, shifting fiber type from fast to slow, [60] R. Stewart, L. Flechner, M. Montminy, R. Berdeaux, CREB is activated by muscle Am. J. Physiol. Cell Physiol. 298 (2010) C140–C148. injury and promotes muscle regeneration, PLoS One 6 (2011), e24714. [47] J.D. Meissner, P.K. Umeda, K.C. Chang, G. Gros, R.J. Scheibe, Activation of the b my- [61] A. Magenta, C. Cenciarelli, F. De Santa, P. Fuschi, F. Martelli, M. Caruso, A. Felsani, osin heavy chain promoter by MEF-2D, MyoD, p300, and the calcineurin/NFATc1 MyoD stimulates RB promoter activity via the CREB/p300 nuclear transduction pathway, J. Cell. Physiol. 211 (2007) 138–148. pathway, Mol. Cell. Biol. 2893-906 (2003).