Six1 homeoprotein drives myofiber type IIA specialization in soleus muscle

Abstract

Background

Adult skeletal muscles are composed of slow and fast myofiber subtypes which each express selective genes required for their specific contractile and metabolic activity. Six homeoproteins are transcription factors regulating muscle cell fate through activation of myogenic regulatory factors and driving fast-type gene expression during embryogenesis.

Results

We show here that Six1 protein accumulates more robustly in the nuclei of adult fast-type muscles than in adult slow-type muscles, this specific enrichment takes place during perinatal growth. Deletion of Six1 in soleus impaired fast-type myofiber specialization during perinatal development, resulting in a slow phenotype and a complete lack of Myosin heavy chain 2A (MyHCIIA) expression. Global transcriptomic analysis of wild-type and Six1 mutant myofibers identified the gene networks controlled by Six1 in adult soleus muscle. This analysis showed that Six1 is required for the expression of numerous genes encoding fast-type sarcomeric proteins, glycolytic enzymes and controlling intracellular calcium homeostasis. Parvalbumin, a key player of calcium buffering, in particular, is a direct target of Six1 in the adult myofiber.

Conclusions

This analysis revealed that Six1 controls distinct aspects of adult muscle physiology in vivo, and acts as a main determinant of fast-fiber type acquisition and maintenance.

Background

Adult skeletal muscles are composed of slow and fast myofiber types, and each skeletal muscle is composed of a stereotyped percentage of myofibers of different subtypes. Myofiber types are characterized by the expression of slow- or fast-type sarcomeric proteins and specific Ca²⁺-handling proteins that modulate its intracellular concentration during the excitation/contraction/relaxation cycle, by their glycolytic and mitochondrial oxidative metabolic properties as well as by their myoglobin content [1–4]. More specifically, the formation of slow or fast sarcomeric units is achieved by the expression of fiber-type specific isoforms of sarcomeric genes such as myosin heavy chain (MyHCI, MyHCIIA, MyHCIIX, MyHCIIB), troponin C (Tnnc1, Tnnc2), troponin I (Tnni1, Tnni2), and troponin T (Tnnt1, Tnnt3). Their metabolic properties are determined by glycolytic enzymes (Gck, Aldoa, Pfkm, Pfkfb1, Eno3) and by mitochondrial oxidative enzymes [5]. Upon motoneuron firing, stimulation intramyofibrillar Ca²⁺ is released from the sarcoplasmic reticulum to trigger muscle contraction through its binding to troponin C, and further activation of myosin heavy chain ATPase. The cytosolic Ca²⁺ reuptake is carried out by the sarcoplasmic reticulum SERCA proteins encoded by slow-type Atp2a2 and fast-type Atp2a1 [1–4]. SERCA activity is regulated by sarcolipin (Sln), expressed in slow and fast oxidative fibers, by phospholamban (Pln) and by myoregulin (Mln) expressed in fast fibers, known repressors of SERCA activity [6, 7] through binding to SERCA proteins [8–10]. In fast-type fibers, parvalbumin (Pvalb), a Ca²⁺ buffering protein, removes Ca²⁺ efficiently from the cytosol to promote relaxation of the myofiber [11].

In adult skeletal muscle, the expression of fiber type-specific genes is coordinated by transcription factors whose activity is modulated by cascades of signaling pathways connected with the environment; mainly by Ca²⁺ flux induced by motoneuron stimulation, O₂, hormones and nutrients availability. Slow motoneuron firing leads to sustained low amplitude elevation in intramyofibrillar calcium concentrations able to activate calcineurin, and CamK while fast motoneuron firing leads to brief intramyofibrillar calcium transients of high amplitude that do not activate calcineurin [12, 13]. Activated calcineurin and CaMK increase the activity of NFAT and MEF2 transcription factors leading to slow sarcomeric gene expression [14, 15]. Muscle mitochondrial oxidation activity which is increased during exercise is under the control of PPARβ/δ and PGC1α and PGC1β, two transcriptional coactivators of PPARβ/δ and activators of oxidative metabolism [16–18]. Whereas muscle specific deletion of both PGC1α and PGC1β does not change muscle fiber type [19], muscle-specific deletion of PPARβ/δ leads to an increased number of fast fibers with reduced oxidative capacity [20], while PPARβ/δ ectopic expression in adult myofibers can change both myosin heavy chain content and oxidative metabolism [21]. Upstream regulators controlling fast myofiber phenotypes are HDACs, Sox6 and Linc-Myh known to suppress slow-type gene expression in fast myofibers [15, 22–25], MyoD [26] and Six1 [25, 27]. Hif1α regulates the expression of genes coding for enzymes of the glycolysis pathway, but its deletion does not lead to major modification of the expression of fiber-type specific sarcomeric proteins while it impairs metabolic adaptation upon exercise [28]. Little is known however concerning the importance of these fiber-type regulators in the coordinated expression of slow or fast genes during perinatal development, the period when muscle fiber specialization takes place [1, 2]. In particular, the mechanisms presiding at the expression of a single fast Myh gene in hundreds nuclei of a myofiber have not yet been elucidated [25, 29, 30].

Six homeoproteins are major myogenic transcription factors that directly bind to DNA sequences called MEF3s to control embryonic myogenesis [31–34] and genesis of fast-type myofibers [29, 35]. Forced expression of Six1 and its Eya1 cofactor in adult slow myofibers can reprogram adult slow-twitch oxidative fibers toward a fast-twitch glycolytic phenotype [31]. In adult fast-type skeletal muscles, Six1 directly regulates the expression of numerous fast-type muscle genes [25]. Furthermore, Six1 interacts with the central enhancer of the Myh fast genes locus, and controls the expression of the fast-type Myh genes (MyHCIIA, MyHCIIX, MyHCIIB) [25].

While Six1 has been detected in soleus (SOL) muscle at the mRNA and protein levels, its physiological role in slow-type muscles has not been explored [25]. Mouse C57bl6N SOL is classified as a slow-type skeletal muscle, composed of approximately 60 % of slow-type/oxydative myofibers and 40 % of fast-type/oxydative myofibers. In this study, we analyzed the phenotypic consequences of Six1 loss in SOL myofibers during mouse perinatal development and in adult. We show that Six1 governs the specification of fast MyHCIIA myofibers and is required for the maintenance of MyHCIIA expression.

Methods

Animals and ethics statement

Animal experimentation was carried out in strict accordance with the European convention STE 123 and the French national charter on the Ethics of Animal Experimentation. Protocols were approved by the Ethical Committee of Animal Experiments of the Institut Cochin, CNRS UMR 8104, INSERM U1016. Surgery was performed under ketamine/xylazine anesthesia, and all efforts were made to minimize suffering. Six1 ^flox/flox; HSA-Cre conditional knockout mice (cSix1KO) were obtained by breeding the Six1-LoxP mice and transgenic mice expressing a CRE recombinase under the control of the human skeletal actin promoter (HSA) [20]. Six1 ^flox/flox; HSA-Cre-ER ^T2 conditional inducible knockout mice (ciSix1 KO) were obtained by breeding the Six1-LoxP mice and HSA-CRE-ER ^T2 mice [36]. two-month-old ciSix1 KO males were given intraperitoneal injection of tamoxifen (1 mg per mouse per day; Sigma) for five consecutive days.

Immunohistochemistry

For Six1 immunostaining, SOL and gastrocnemius plantaris (GP) muscles were embedded in cryomatrix and quickly frozen in isopentane cooled with liquid nitrogen. Cryostat sections (10 μm) were fixed in 4 % PFA and washed in 1× PBS. The sections were treated with Antigen Unmasking Solution (H-3300, Vector Laboratories) at 95 °C for 10 min and washed in 1× PBS for three times. Sections were treated with 1 % H₂O₂ solution for 20 min. After three washes in 1× PBS, they were permeabilized with 0.1 % Triton X-100 for 20 min and left for 1 h in blocking solution (1× PBS, 1.5 % goat serum, 0.1 % Triton X-100). Rabbit polyclonal antibodies directed against Six1 (HPA001893, Sigma) (1/100 dilution), and dystrophin (NCL-DYS2, Leica Biosystems) (1/50 dilution) were applied overnight at 4 °C to the treated sections. The next day, after three washes with 1× PBS containing 0.05 % Tween-20, sections were incubated for 1 h with appropriate fluorescent secondary antibodies (Alexa Fluor 594 goat anti-mouse IgG 1/1000 dilution, Invitrogen) and biotynilated secondary antibodies (anti-rabbit IgG 1/200, Vector Laboratories). After three washes with 1× PBS containing 0.05 % Tween-20, samples were incubated in VECTASTAIN Elite ABC Reagent (Vector Laboratories) for 30 min. After three washes with 1× PBS containing 0.05 % Tween-20, samples were incubated with a tyramide solution labeled by Alexa Fluor 488 (Tyramide Signal Amplification kit, invitrogen). After three washes with 1× PBS containing 0.05 % Tween-20, samples were mounted in Vectashield mounting medium.

For determination of Myh isoform expression, SOL and GP muscles were embedded in cryomatrix and quickly frozen in isopentane cooled with liquid nitrogen. Cryostat sections (10 μm) were washed in PBS, permeabilized with 0.1 % Triton X-100 for 20 min and left for 1 h in blocking solution (1× PBS, 1.5 % goat serum, 0.1 % Triton X-100). Rabbit poly-clonal antibodies directed against Laminin (Z0097, Dako) (1/100 dilution), and mouse monoclonal antibodies against MyHCI (NOQ7.5.4D, Sigma) (1/1000 dilution), MyHCIIA (SC-71, Developmental Studies Hybridoma Bank) (1/20 dilution), fast MyHCs (My-32, Sigma) (1/50 dilution) and MyHCemb (F1.652, sc-53091 Santa Cruz Biotechnology, Inc.) (1/20 dilution) were applied overnight at 4 °C to the treated sections. The next day, after three washes with 1× PBS containing 0.05 % Tween-20, sections were incubated for 1 h with appropriate fluorescent secondary antibodies (Alexa Fluor 488 goat anti-rabbit IgG 1/1000 dilution, Alexa Fluor 594 goat anti-mouse IgG 1/1000 dilution, Invitrogen). After three washes with 1× PBS containing 0.05 % Tween-20, samples were mounted in Vectashield mounting medium. For fiber type counting, each MyHC positive fiber was counted in the entire SOL muscle sections, and the number of positive fibers was divided by the total SOL number of fibers.

SDH/GPDH staining

Fresh-frozen sections were incubated in 0.2 M phosphate buffer (pH 7.6) containing sodium succinate and nitroblue tetrazolium, NBT (N6876, Sigma Aldrich) for 30 min at 37 °C. Sections were then washed with water and mounted in glycerine gelatin medium. GPDH staining was performed by incubation of unfrozen muscle sections with α-glycerol phosphate as described [37]. For quantification of SDH and GPDH staining, the color images were converted to thresholded images at hue (121-208) and brightness (0-140) by a threshold tool of ImageJ software. The area of thresholded images was measured by ImageJ and normalized by the whole soleus muscle area.

RNA preparation

Soleus muscles were collected from cSix1 KO and control mice. Total RNAs were extracted using Trizol Reagent (Invitrogen) according to manufacturer’s instruction.

cDNA synthesis and qPCR

RNAs were treated with DNase I (Turbo DNA-free, Invitrogen) and were reverse-transcribed with Superscript III kit (Invitrogen) according to manufacturer’s instruction. Reverse transcription was performed with 1 μg of total RNA. Quantitative real time PCR (Light Cycler 480, Roche) was performed using Light Cycler 480 SYBR Green I Master Kit (Roche) according to the manufacturer’s protocols. PCR was performed for 40 cycles of 95 °C for 15 s, 60 °C for 15 s, and 72 °C for 15 s. Gene expression levels were normalized by the expression level of the housekeeping gene Actb. The sequences of the oligonucleotides used in this study are given in Additional file 1: Table S1.

ChIP experiments

GP and tibialis anterior (TA) muscles of 2 months old female mice were minced with scissors immediately after harvesting and fixed in 1 % formaldehyde for 10 min. Formaldehyde was quenched by addition of 0.125 M glycine, and muscles were washed twice in PBS. Muscles were then incubated on ice in lysis buffer (10 mM Tris-HCl pH 7.9, 85 mM KCl, 0.5 % NP40, protease inhibitors (Complete, Roche)) for 10 min and homogenized with a mortar and subsequently with a dounce homogenizer. Nuclei were obtained by centrifugation, incubated in SDS lysis buffer (50 mM Tris-HCl pH 8, 10 mM EDTA, 1 % SDS, protease inhibitors) for 10 min, and sonicated in a bioruptor apparatus (Diagenode). The debris was removed by centrifugation. Sonicated DNA was incubated with 1 μg of Six1 antibodies (HPA001893, Sigma) under rotation at 4 °C overnight. 20 μl of dynabeads protein G (Invitrogen) were added to the lysates and incubated under rotation at 4 °C for 1 h. The beads were washed with low-salt buffer (2 mM EDTA, 20 mM Tris-HCl pH 8, 150 mM NaCl, 1 % TritonX-100, 0.1 % SDS), high-salt buffer (2 mM EDTA, 20 mM Tris-HCl pH 8, 0.5 M NaCl, 1 % TritonX-100, 0.1 % SDS), LiCl buffer (1 mM EDTA, 10 mM Tris-HCl pH 8, 0.25 M LiCl, 1 % NP40, 1 % deoxycholate), and TE buffer (1 mM EDTA, 10 mM Tris-HCl pH 8). DNA was eluted with elution buffer (1 % SDS, 0.1 M NaHCO₃) containing 0.1 mg/ml proteinase K (Invitrogen) at 62 °C for 2 h, and proteinase K was inactivated by incubation at 95 °C for 10 min. DNA was finally purified with MinElute PCR purification kit (Qiagen). The amount of specific amplified DNA is normalized by input amplification. The sequences of the oligonucleotides used in this study are as follows. Actb, 5′-TGTTACCAACTGGGACGACA-3′ and 5′-ACCTGGGTCATCTTTTCACG-3′; PvalbMEF3_1, 5′-GGAGCCTTTCATGGTGTGAT-3′ and 5′-GGCGTGTGAATCACTTTCCT-3′; PvalbMEF3_2, 5′- GGATGGGGGTGAATGTGATA-3′ and 5′- CTTCCGGTGTCAGGTACTCC-3′.

EMSA

EMSA was carried out with Six1 and Six4 full-length mouse cDNA cloned into the pCR3 vector (Clontech) as previously described [38]. Recombinant mouse Six1 and Six4 proteins were obtained separately with a T7 transcription/translation kit (Promega) and mixed before contact with the DNA. Myogenin MEF3 DNA was incubated with recombinant proteins. Competition experiments were performed in the presence of a tenfold and hundredfold molar excess of unlabeled identified Pvalb MEF3_1, Pvalb MEF3_2, Myogenin MEF3, Myod1DRR MEF3, or Myogenin NFI sites. The sequences of the oligonucleotides used are as follows, the MEF3 consensus sequence is in italic; MyogF 5′-TGG GGG GGC TCA GGT TTC TGT GGC GT-3′. MyogR 5′-ACG CCA CAG AAA CCT GAG CCC CCC CA-3′. NF1F 5′-TAT CTC TGG GTT CAT GCC AGC AGG G-3′. NF1R 5′-CCC TGC TGG CAT GAA CCC AGA GAT A-3′. PvalbMEF3_1F, 5′- TGA GCA TCT GTA ACC TGA CCC TTG G -3′. PvalbMEF3_1R, 5′- CCA AGG GTC AGG TTA CAG ATG CT-3′. PvalbMEF3_2F, 5′- TGA GTA CCT GAC ACC GGA AGG GGA G-3′. PvalbMEF3_2R, 5′- CTC CCC TTC CGG TGT CAG GTA CT-3′. MyodDRRF, 5′- AGT TGG ATC CGG TTT CCA GAG GC -3′. MyodDRRR, 5′- TGA GAC AGT AAT TTT ATC CTG CT -3′.

Plasmids construction

For the construction of the pGL3-Pvalb, C57bl6N mouse DNA was first used as a template to clone the 700 bp promoter of Pvalb with forward MluI, 5′- GTAACCTGACCCTTGGAAACCAG -3′ and reverse BglII, 5′- CTTGGATGAGCAGAGGCCGGA-3′ primers. This Pvalb promoter fragment was subsequently inserted into an MluI-BglII digested pGL3 basic plasmid. For the construction of the pGL3-PvalbmutMEF3-1, PvalbmutMEF3-2 and PvalbdoublemutMEF3, the two MEF3 sites of the promoter were mutated as follows; Pvalb MEF3-1: 5′GTAACCTGA to 5′CGCGTCTGA; Pvalb MEF3-2: 5′GACACCGGA to 5′CTCGAGGGA. All plasmids sequences were confirmed by sequencing.

Electroporation

In vivo transfections were also carried out on ten-week old C57Bl6N mice. For each experimental conditions three to five TA muscles belonging to different mice were used. Under isoflurane anesthesia, legs were shaved and muscles were pre-treated by injection of a sterile 0.9 % NaCl solution containing 0.4 U of bovine hyaluronidase/μl 2 h before plasmid injection. Two micrograms of Luciferase-expressing vector and one hundred ng of pRL-TK vector (Promega) were introduced into TA muscles of ten-week-old mice by electroporation as previously described [27]. Two weeks following electroporation, electroporated muscles were frozen in liquid nitrogen before processing for Luciferase assays.

Luciferase assays

Two weeks after electroporation, TA were dissected and frozen in liquid nitrogen before processing. TA were homogenized in Passive Lysis Buffer (Dual-Luciferase Reporter Assay System, Promega) and rotated for 15 min. The homogenate was centrifuged to remove debris, and the supernatant was used for Luciferase activity measurement according to manufacturer’s instruction (Dual-Luciferase Reporter Assay System, Promega).

Microarrays

After validation of RNA quality with the Bioanalyzer 2100 (using Agilent RNA6000 nano chip kit), 50 ng of total RNA were reverse-transcribed following the Ovation PicoSL WTA System (Nugen). Briefly, the resulting double-strand cDNA was used for amplification based on SPIA technology. After purification according to Nugen protocol, 5 μg of single strand DNA was used for generation of Sens Target DNA using Ovation Exon Module kit (Nugen). 2.5 μg of Sens Target DNA were fragmented and labelled with biotin using Encore Biotin Module kit (Nugen). After control of fragmentation using Bioanalyzer 2100, the cDNA was then hybridized to GeneChip® Mouse Gene 1.0 ST (Affymetrix) at 45 °C for 17 h. After overnight hybridization, the ChIPs were washed using the fluidic station FS450 following specific protocols (Affymetrix) and scanned using the GCS3000 7G. The scanned images were then analyzed with Expression Console software (Affymetrix) to obtain raw data (cel files) and metrics for quality controls. The analysis of some of these metrics and the study of the distribution of raw data show no outlier experiment. RMA normalization was performed using R and normalized data was subjected to statistical tests.

Preparation of nuclear and cytosolic proteins

SOL and GP muscles of 2, 3, and 12 weeks old mice were frozen with liquid nitrogen and were homogenized with a mortar. Homogenates were fractionated by NE-PER kit (78833, Thermo Scientific) according to manufacturer’s instruction.

Western blot

Cell or tissue lysates of GP and SOL from cSix1KO and control mice (20–40 μg) were denatured in Laemmli buffer, separated by SDS-polyacrylamide gel electrophoresis and transferred to nitrocellulose membrane. Membranes were blocked in 50 mM Tris-HCl pH 7.6, 137 mM NaCl and 0.1 % (v/v) Tween-20 containing 10 % (w/v) skimmed milk or 5 % (w/v) BSA for 1 h at room temperature and incubated overnight at 4 °C with the indicated primary antibodies (Complex I, NADH dehydrogenase, ab14713, Abcam (Cambridge, UK); Complex II, succinate dehydrogenase, ab109865, Abcam (Cambridge, UK); cytochrome bc1 complex, ab110252, Abcam (Cambridge, UK); Complex IV, cytochrome C oxidase, ab14744, Abcam (Cambridge, UK); Complex V, ATP synthase, ab14748, Abcam (Cambridge, UK); hexokinase II, Sc-6521 (Santa Cruz Biotechnology); glycogen synthase 1, CST #3893, Cell Signaling Technology; AS160, 07-741, Millipore; GLUT4, kind donation from Geoffrey Holman, University of Bath); lamin B, sc-6216 (Santa Cruz Biotechnology); β-tubulin, 05-661 (Millipore); Six1, HPA001893 (Sigma). Detection was performed using horseradish peroxidase conjugated secondary antibodies and enhanced chemiluminescence reagent.

Statistical analysis

All graphs represent mean values ± SEM. Significant differences between mean values were evaluated using two-tailed, unpaired Student’s t test (when two groups were analyzed) or one-way ANOVA followed by Student Newman-Keuls test (for three or more groups).

Microarray data accession number

Microarray data have been deposited in the Gene Expression Omnibus as accession no. GSE50023.

Results

Six1 protein subcellular localization in adult myofibers

To analyze the properties of Six1 homeoprotein in adult SOL muscles, we first determined its expression pattern. We showed previously that Six1 mRNA accumulates in both fast and slow muscles and that Six1 protein is produced in both type of muscles but accumulates more robustly in the nuclei of fast myofibers as evaluated by immunohistochemistry [25, 27]. We here show that Six1 protein is produced in the GP (fast gastrocnemius and plantaris muscles) and in the SOL of 2 weeks, 3 weeks and adult mice (Fig. 1a), but that the nuclear accumulation of Six1 is drastically reduced between 3 weeks and the adult stage in the SOL (Fig. 1b), although no obvious cytoplasmic Six1 accumulation is observed in the SOL as estimated by Western blot experiments. Six1 protein was further detected in the fast GP and the slow SOL muscles by immunofluorescence using antibodies directed against Six1 at several developmental stages. At embryonic day E18.5, we observed that Six1 is present in the nuclei of myofibers of the GP and SOL expressing at that stage both MyHCI (slow/β) and fast MyHC (detected by My32 antibodies that recognize MyHCemb, MyHCneo and adult fast MyHC) (Fig. 1c). On adult sections, Six1 protein accumulation was detected preferentially in the nuclei of GP as compared with SOL nuclei (Fig. 1d), in agreement with previous results [27]. While GP muscles are mainly composed of myofibers expressing MyHCIIB or MyHCIIX, SOL muscles are composed mainly by myofibers expressing MyHCI (slow/β), MyHCIIA and by few myofibers expressing MyHCIIX. Nuclear Six1 is nevertheless detected in the SOL in the fast MyHCIIA, and in some MyHCI fibers (Fig. 1c). These results showed that in adult hindlimb muscles, Six1 proteins accumulate preferentially in the nuclei of GP fast myofibers than in the nuclei of SOL myofibers (Fig. 1c), in accordance with the low amount of Six1 proteins detected in myonuclei of SOL by Western blot analysis (Fig. 1a), and that this preferential nuclear Six1 accumulation takes place during the perinatal period.

Fig. 1

Six1 proteins are predominantly localized in nuclei of fast type skeletal muscles. a, b Western blot analysis of Six1 proteins in total lysates (a) and in nuclear fractions (b) of GP and SOL muscles of 2-week, 3-week, and 12-week-old animals. Quantification of nuclear Six1 proteins at these developmental stages is presented. Lamin B is used as a nuclear control protein, and α-tubulin as a cytoplasmic control protein. c Immunostaining of Six1 (green), Dystrophin (red), MyHCI (red) Laminin (green) and MY32 (red) in soleus and gastrocnemius (GP) muscles of embryonic E18.5 mouse fetuses. The second image is an enlargement of the yellow square drawn on the first image. White arrows indicate Six1 nuclear staining. d. Immunostaining of Six1 (green), Dystrophin (red), MyCHI (red), MyHCIIA (red) and Laminin (green) in SOL and GP of 20-week-old mice. The lower panels are enlargment of the yellow squares drawn on the top image. White arrows indicate nuclear Six1 detection. Some MyHCI and MyHCIIA positive fibers are indicated as I or IIA, respectively. SOL soleus muscle, GP gastrocnemius plantaris muscle

Six1 deficiency impairs adult muscle fast type phenotype acquisition in SOL

To characterize the role of Six1 in adult SOL, we analyzed myofiber-specific Six1 knockout (cSix1 KO) mice [25]. cSix1 KO mice were viable, and Six1 mRNA and protein were not detectable in adult GP or SOL muscles [25]. As fiber type specialization that leads to the expression of a single Myh gene in all the myonuclei of a given myofiber proceeds during the neonatal stage, we next performed immunohistochemistry to analyze the content of MyHCI, MyHCIIA, and MyHCemb of SOL in 3-week-old cSix1 mutant myofibers. In control (Ctrl) mice, 10 % of myofibers expressed MyHCemb, 50 % expressed MyHCI, and 60 % expressed MyHCIIA (Fig. 2a, b). In cSix1 KO mice, 40 % of myofibers still expressed MyHCemb, 96 % of myofibers expressed MyHCI, and 33 % of myofibers expressed MyHCIIA. This result suggested that sustained expression of Six1 in the perinatal period is important for embryonic to adult fast myofiber transition in SOL and that the decrease of nuclear Six1 accumulation may be important to allow MyHCI exclusive expression in slow-type SOL myofibers. It is known that perinatal muscle growth in mice takes place by accretion of new satellite cells (SC) [39]. As Six1 is expressed in SC [40–42], it is possible that Six1 may be transiently expressed by new accreted myonuclei until P21 [39], before its efficient deletion by the HSA-CRE recombinase only active in post mitotic myofibers. To test this hypothesis, we measured Six1 protein accumulation in the nuclei of SOL at 3 weeks of development and detected Six1 positive myonuclei in cSix1 SOL at this development stage (Fig. 2e). This showed that HSA-CRE recombinase had not yet recombined the Six1 flox allele in all P21 myonuclei, allowing Six1 to be detected in some nuclei of perinatal myofibers of mutant animals.

Fig. 2

Six1 deficiency induced a lack of MyHCIIA fibers in soleus. a Immunostaining of MyHCI (red), MyHCIIA (red), MyHCemb (red), and laminine (green) in soleus muscles of 3-week-old control and cSix1 KO mice. SOL soleus, GP gastrocnemius plantaris, F fibula. b Percentage of myofibers expressing MyHCI, MyHCIIA and MyHCemb in soleus of 3-week-old control and cSix1 KO mice, Ctrl (n = 3), cSix1 (n = 3). c Immunostaining of MyHCI (red), MyHCIIA (red), and laminine (green) in soleus of 12-week-old control and cSix1 KO mice. d Percentage of myofibers expressing MyHCI, MyHCIIA in soleus of 12-week-old control and cSix1 KO mice, Ctrl (n = 3), cSix1 (n = 3). e Immunostaining of Six1 (green) and Dystrophin (red) in soleus of 3 weeks old cSix1 KO mice and Ctrl mice. White arrows indicate nuclear Six1 staining. Nuclei are stained with Dapi (blue). *P < 0.05, **P < 0.01, ***P < 0.001

In 12-week-old adult animals, at a stage when Six1 is no longer detected in adult cSix1 myonuclei [25], 100 % of SOL myofibers in mutant mice expressed MyHCI while MyHCIIA was not detectable (Fig. 2c, d), contrary to Ctrl SOL where 45 % of myofibers expressed MyHCIIA and 55 % expressed MyHCI (Fig. 2c, d). In adult cSix1 mutant SOL, MyHCIIX and MyHCIIA mRNA became undetectable (Fig. 4b), while the amount of MyHCI mRNA increased twice (Fig. 4b).

The number of SOL myofibers present in Ctrl and cSix1 KO was comparable, excluding that absence of MyHCIIA myofibers in adult mutant animals is the consequence of their death (Additional file 2: Figure S1). We observed no significant modification of the CSA between wt and cSix1 KO adult myofibers, excluding that Six1 is a main regulator of MyHCIIA myofibers growth in SOL (Additional file 2: Figure S1).

To analyze the oxidative/glycolytic metabolism of mutant SOL muscles, SDH, and GPDH staining were performed on 12-week-old soleus muscle sections (Fig. 3a, b). A two- to threefold decrease of SDH activity was observed in cSix1 SOL as compared with that of control (Fig. 3a). Nevertheless, no significant difference of mitochondrial protein contents was observed in cSix1 SOL extracts as determined by Western blots (Fig. 3c). We also observed a robust decrease in GPDH activity in cSix1 SOL as compared with that of control, suggesting that absence of Six1 decreases glycolytic flux in adult SOL myofibers (Fig. 3b).

Fig. 3

Metabolic properties of cSix1 KO muscles. a SDH staining of soleus of 12-week-old cSix1 KO and wt mice, and percentage of the area of positive fibers. b GPDH staining of soleus of 12-week-old cSix1 KO and wt mice, and percentage of the area of positive fibers. c Western blot analysis of mitochondrial proteins and glucose metabolism proteins in SOL and GP of Ctrl and cSix1 KO mice. C2C12, mitochondrial fraction of C2C12

Networks of genes under the control of Six1 in adult SOL

To explore the networks of genes under the control of Six1 in adult SOL, we performed Affymetrix microarray analyses using RNA from adult Ctrl and cSix1 KO mice. Genes whose expression is the most up- or down-regulated (Additional file 3: Table S2) are shown as a heat map and as a bar graph (Fig. 4a and Additional file 4: Figure S2). The expression of several fast-type genes is down-regulated in cSix1 KO including Pvalb, Mybpc2, MyHCIIA, Myl1, Myoz1, and Mylpf (Fig. 4a). We validated the expression of those fiber type-specific genes by qPCR. Consistent with immunohistochemistry data, MyHCIIA mRNA was no longer detectable in SOL of 12-week-old cSix1 KO mice, and MyHCI mRNA level was increased by twofold as compared with control (Fig. 4b). The expression levels of slow-type genes (Tnnt1, Tnni1, Tnnc1) were also increased by twofold in cSix1 KO soleus muscles while the expression of fast-type genes Tnnt3 and Tnni2 was not detected and expression of Tnnc2 was markedly decreased (Fig. 4b). These data show that nuclear accumulation of Six1 observed in SOL MyHCIIA-myofibers is necessary to activate the expression of MyHCIIA and of other fast-type sarcomeric genes and to suppress the expression of MyHCI and other slow-type muscle genes.

Fig. 4

Affymetrix microarray analysis in soleus of cSix1 mice. a Microarray analysis of soleus of 3-month-old cSix1 KO mice (n = 3) and Ctrl mice (n = 3): a heat map of genes significantly upregulated (red) or downregulated (green) by twofold compared with Ctrl mice. b–f mRNA expression levels of sarcomeric genes (b), energy metabolism (c), calcium regulation (d), transcription factors (e), and most down-regulated or up-regulated genes (f) in soleus of 3-month-old cSix1 KO mice, Ctrl (n = 4), cSix1 (n = 3). *P < 0.05, **P < 0.01, ***P < 0.001

The second most down-regulated gene in cSix1 SOL is Pvalb (Fig. 4a), which is a calcium buffering protein allowing muscle relaxation in fast-type muscles [11, 43–45]. Furthermore, other genes related with calcium regulation were also modified in cSix1 KO mice (Additional file 4: Figure S2). We validated the down-regulation of Pvalb mRNA level in cSix1 SOL by qPCR experiments (Fig. 4c). The expression level of Sln, Atp2a2, and Ryr3 was increased in cSix1 SOL muscles while the expression of fast-type genes Myoz1 and Atp2a1 was strongly decreased (Fig. 4d). These data suggest that Six1 regulates intramyocellular calcium transients through the control of the expression of several fiber type-specific calcium-binding proteins, and suggest a feedback loop between Six1 nuclear accumulation and resting intracellular calcium concentration.

Concerning Affymetrix results, the expression of several genes encoding glycolytic enzymes (Aldoa, Ldha, Pfkfb1, PFKm, and Eno3) was also down-regulated (Fig. 4a and Additional file 4: Figure S2 and Additional file 5: Figure S3). Aldoa possesses several promoters, and its fast-type one is a known direct target of Six proteins [25, 27]. Muscle-specific Aldoa mRNA expression was strongly down-regulated in SOL of cSix1 KO mice (Fig. 4c). We also observed the down-regulation of Slc2a4 (Glut4) and Tbc1d1—involved in Glut4 vesicular traffic [46]—of glycolytic genes (Eno3, Pfkfb1), of muscle Creatine Kinase, of Ldha, and Ldhd of the NAD⁺-dependent isocitrate dehydrogenase Idh3a, of Idh1, the down-regulation of Pdp1, an activator of and the up-regulation of Pdk3 an inhibitor of pyruvate dehydrogenase activity, and validated by qPCR experiments of their expression (Fig. 4d). No major up- or down-regulation of genes related with oxidative metabolism was detected (Additional file 4: Figure S2), which is consistent with the results of the Western blot analysis showing no major modification of the quantity of the electron transport chain proteins (Fig. 3c).

To investigate potential interaction between Six1 and known fiber-type regulators that may participate downstream of Six1 to the observed phenotype, we analyzed their expression levels by Affymetrix and qPCR (Fig. 4e and Additional file 4: Figure S2). Sox6 gene, a known repressor of slow sarcomeric genes [22, 24], showed a twofold expression decrease in SOL of cSix1 KO mice (Fig. 4e). This down-regulation may be involved in the up-regulation of the expression of its known slow-type gene targets MyHCI, Tnnt1, Tnnt2, Tnni1, and Tnnc1 [22–24, 47] observed in cSix1 mutant myofibers. Expression of Esr1 gene (estrogen receptor 1), an activator of the slow/oxidative phenotype in females [48] was decreased in cSix1 KO mice. We tested the expression of Ppargc1a and Nfatc1 known transcriptional activators of slow-oxidative genes [1, 2], but found no significant modifications of their expression levels in SOL of cSix1 mice (Fig. 4e). We were also unable to detect modifications of the expression of mRNA encoding TFAM, PPARβ, Rev-erb-α, or Mef2C, known transcription factors controlling the slow/oxidative phenotype of adult myofibers (data not shown). Altogether our results indicate that in SOL of adult mice, Six1 controls the expression of genes coding for fast-type sarcomeric and calcium handling proteins, for glycolytic proteins, and for Sox6 a known repressor of slow type sarcomeric genes; absence of Six1 leading to a myofiber switch toward a slower phenotype. Interestingly, in SOL of cSix1 KO, we also observed the up-regulation of Prox1 (Fig. 4f), a known repressor of the fast genes Tnnt3, Tnni2, MyHCIIA, and Myl1, and which may also account for their down-regulation [49].

Among the genes showing the most up- or down-regulation in mutant SOL, Affymetrix analysis also revealed a strong reduction of the expression levels of Ddit4l (Redd2) an inhibitor of the mTOR pathway [50], Rspo3 (R-Spondin3) and Aldh1a1 and the increase of the expression level of Nuak1 (NUAK family, SNF1-like kinase, 1), Ptgr1 (Prostaglandin reductase1), Chnrγ and Chnrα1 (Acetylcholin receptor gamma and alpha1), and Cidea (cell death-inducing DNA fragmentation factor, alpha subunit-like effector A), an inhibitor of AMPK [51] (Fig. 4a). We also validated the expression of those genes by qPCR (Fig. 4f).

These results indicate that Six1 has a role to enhance the glycolytic pathway in myofibers of adult SOL through the transcription of genes coding for glycolytic proteins, and may modulate more generally glucose utilization in adult myofibers by modulating the expression of key modulators of glucose flux among which Krebs cycle genes (Additional file 5: Figure S3).

Skeletal muscle-specific Tamoxifen-inducible Six1 knockout mice also showed fast-to-slow fiber type transition

To examine the role of Six1 in the maintenance of myofibers phenotype in the adult SOL, we bred Six1 ^flox/flox mice with transgenic mice expressing Cre-ER^T2 recombinase under the control of the HSA promoter [36] to obtain Six1 ^flox/flox; HSA-Cre-ER ^T2 conditional inducible knockout mice (hereafter named ciSix1 KO), to induce Six1 deletion after Tamoxifen injection. Eight-week-old ciSix1 KO mice and their littermate Ctrl were injected with Tamoxifen. One month after Tamoxifen injection, Six1 mRNA level in ciSix1 KO was reduced by ~90 % compared with that in Ctrl mice (Fig. 5a). We further analyzed SOL fiber-type composition by immunohistochemistry (Fig. 5b, c). The percentage of MyHCI positive fiber in ciSix1 KO mice increased to 90 %, and the percentage of MyHCIIA positive fiber in ciSix1 KO mice decreased to 10 %. This result showed that Six1 is required for the maintenance of MyHCIIA expression in SOL at the adult stage.

Fig. 5

Tamoxifen-induced conditional muscle specific Six1 knockout reduced fast-type gene expression in soleus. a Six1 mRNA expression levels in SOL of 3 month-old control (Ctrl, n = 3) and ciSix1 KO (n = 4) mice 1 month after tamoxifen injection. b Percentage of myofibers expressing MyHCI or MyHCIIA in SOL of 3 month-old control (Ctrl, n = 3) and ciSix1 KO (n = 4) mice 1 month after tamoxifen injection. c Immunostaining of MyHCI (red) and MyHCIIA (red) in SOL of 3 month-old control and ciSix1 KO mice 1 month after tamoxifen injection. Arrows indicate fibers expressing MyHCIIA, and an arrow head indicates a fiber expressing both MyHCI and MyHCIIA in ciSix1 KO mice. d-g mRNA expression levels of sarcomeric genes (d), energy metabolism (e), calcium regulation (f), transcription factors (g), and most down-regulated or up-regulated genes (h) in SOL of 3 month-old control (Ctrl, n = 3) and ciSix1 KO (n = 4) mice 1 month after tamoxifen injection. *P < 0.05, **P < 0.01, ***P < 0.001

To compare the phenotypic consequences of Six1 deletion in the adult myofibers and during development, mRNA expression levels of genes identified previously were estimated in SOL of ciSix1 KO by qPCR. The expression of fast-type sarcomeric genes (MyHCIIX, MyHCIIA, Tnnt3, Tnni2, Tnnc2) was also down-regulated (Fig. 5d), consistent with immunohistochemistry results. In contrast, we did not observe a significant up-regulation of slow-type sarcomeric genes (MyHCI, Tnnt1, Tnni1, Tnnc1), that may be related to the absence of Sox6 down regulation (Fig. 5d). For genes controlling glucose metabolism, we observed that the expression level of glycolytic enzymes (Aldoa, Pfkfb1, and Eno3) was down-regulated, and that the expression level of Ldha and Pdk3 was not altered (Fig. 5e). Concerning genes controlling calcium handling, mRNA level of Pvalb was decreased by 90 % compared with that of Ctrl mice, and an increase of Sln and Ryr3 mRNA levels was observed (Fig. 5f). mRNA level of fiber type regulators such as Sox6 was not significantly changed between ciSix1 KO mice and Ctrl mice (Fig. 5g), contrary to what was observed in cSix1 KO. The decreased expression level of Ddit4l, Rspo3, and Aldh1a1, and the increased expression level of Cidea were also observed in the SOL of ciSix1 KO mice (Fig. 5h). These data show that one main function of Six1 in adult SOL could be to activate the expression of a network of fast-glycolytic specific genes among which are MyHCIIA, Aldoa, and Pvalb. The modulation of slow-type genes expression was not observed in the time window of 4 weeks following Six1 deletion, suggesting that it may take more time to completely transdifferentiate fast-MyHCIIA myofibers into slow myofibers.

Pvalb is a direct target of Six1

A robust reduction of Pvalb mRNA observed in cSix1 KO mice and in ciSix1 KO mice led us to test the hypothesis that Pvalb is a direct target of Six1. We analyzed the Pvalb promoter sequence and identified MEF3 regulatory elements. Two MEF3 sites located at −725 bp (Pvalb MEF3-1) and −148 bp (Pvalb MEF3-2) from the transcription start site of Pvalb were identified (Fig. 6a). Six1 binding at these MEF3 sites was demonstrated in vivo by ChIP experiments with Six1 antibodies on adult fast GP and TA muscles (Fig. 6b) and confirmed for both of these sites by EMSA assays with recombinant Six1 and Six4 proteins (Fig. 6c). We next tested the transcriptional activating potential of these elements in vivo in transient transfection assays. A 748 bp DNA fragment of the Pvalb promoter, including the two identified MEF3 sites, was ligated to pGL3 basic plasmids to generate pGL3-Pvalb constructs. Mutations of the two MEF3 sites were introduced in the promoter sequence, giving rise to Pvalb mut-MEF3-1, Pvalb mut-MEF3-2, and Pvalbmut-MEF3-1/2. Luciferase activity was tested after electroporation of these reporter plasmids in adult TA muscles. The activity of a pGL3-Pvalb was 100-fold higher than that of the empty pGL3 vector. Luciferase activity was strongly decreased when the MEF3-2 site at -148 was mutated. No further transcriptional decrease of the Pvalb promoter was observed when both MEF3 sites were mutated (Fig. 6d). Altogether, these results demonstrate that Pvalb is a direct target of Six1, and that Six1 binding to the proximal MEF3 site present in the Pvalb promoter is essential for its transcriptional activity in vivo.

Fig. 6

Pvalb is a direct target of Six1. a Schematic representation of the Pvalb promoter. b qPCR values of ChIP experiments performed with Six1 antibodies or IgG on GP and TA chromatin, and showing Six1 binding to Pvalb MEF3_1 and Pvalb MEF3_2. c Competitive Electromobility shift assays performed with recombinant Six1 and Six4 proteins and labeled Myogenin MEF3 oligonucleotide and 10 or 100 fold molar excess of unlabelled oligonucleotides containing Myogenin MEF3 site, NF1 site, or Pvalb MEF3_2 (left panel) and 100-fold molar excess of unlabelled oligonucleotides containing Myogenin MEF3, NF1 site, Myod DRR MEF3 site, or Pvalb MEF3_1 sites (right pannel) whose sequence is presented on a. d Luciferase assays from adult TA muscles electroporated by the indicated luciferase vectors and the TK-renilla luciferase vector allowing normalization. *P < 0.05, ***P < 0.001

Discussion

In this study, we analyzed the subcellular distribution of Six1 homeoprotein during postnatal development in adult SOL muscle, and the consequences of Six1 loss during muscle fiber type specialization. We showed that Six1 proteins accumulated differentially in the myonuclei of adult fast and slow fibers. Absence of Six1 delayed the transition from embryonic MyHCemb fiber type to adult fast-type MyHCIIA fiber type, leading to a concerted down-regulation of the fast-type program and up-regulation of the slow-type program. Transcriptomic analysis of Ctrl and mutant SOL identified a network of down-regulated and up-regulated genes upon Six1 deficiency. Particularly, Six1 is required in the SOL to activate the expression of several fast sarcomeric genes, glycolysis genes, fast-type calcium-handling genes as well as Sox6, a repressor of slow genes. Conversely, up-regulated expression was observed in Six1 mutant SOL for numerous slow sarcomeric genes and for Prox1, a repressor of fast genes. We further identified Pvalb, a key player in intracellular Ca²⁺ buffering, as a direct target of Six1. Altogether our analysis revealed that Six1 is an essential genetic determinant of fast-type specialization during the post-natal period, and that Six1 is required in the adult for the maintenance of this phenotype.

Development of the fast-type IIA phenotype in SOL

Adult mouse SOL is composed of distinct types of myofibers characterized by their contractile and metabolic properties and expressing either the slow MyHCI, the fast MyHCIIA, or, more rarely, the fast MyHCIIX gene [5]. We observed that the Six1 proteins accumulate differentially in adult myonuclei, with a more pronounced enrichment in fast fibers of the GP (Fig. 1) as compared with the fibers of the SOL. However, Six1 mRNA and total protein levels were found similar between SOL and the fast GP [25]. Altogether, these observations suggest that Six1 could be sequestered outside of most MyHCI myonuclei, or actively retained in the nuclei of fast-type fibers, and we show that this control takes place during the perinatal period. Nevertheless, in the SOL Six1 proteins coordinate the expression of genes specific of the adult fast phenotype (Mybpc2, Tnni2, Tnnt3, Pvalb, Sox6), most probably in MyHCIIA fibers where these genes are coexpressed with MyHCIIA. Most of those fast-type genes are also down-regulated in fast GP or TA muscles of cSix1 mutant, as already reported [25], demonstrating that Six1 acts as a major determinant of fast type gene expression in the different fast fibers subtypes, its absence leading to a “slower phenotype”. In the GP and fast tibialis muscles, absence of Six1 leads to the down regulation of MyHCIIB and MyHCIIX and the up-regulation of MyHCI and MyHCIIA [25], while in the SOL it abrogates MyHCIIX and MyHCIIA expression. To explain this discrepancy one can suggest that the absolute quantity of Six homeoproteins (Six1, Six2, Six4, and Six5) present in the myonucleus tightly controls the expression of the fast Myh genes cluster [52]; high levels being required to activate MyHCIIB, while lower levels would favor MyHCIIA expression. In fast muscles, absence of Six1 would lead to decrease the overall Six level which would nevertheless remain sufficient to activate MyHCIIA [25], while in the SOL the threshold of Six proteins reached in absence of Six1 would not allow to maintain MyHCIIA expression. Quantification of nuclear Six2, Six4, and Six5 proteins remaining in SOL and GP muscles of Six1 mutant mice may help testing this hypothesis. Alternatively, other yet unidentified specific transcription factors may compensate Six1 loss in fast TA and GP muscles allowing MyHCIIA expression observed in Six1cKO, while their absence in SOL would preclude MyHCIIA expression. Identification of the transcription factor machinery present specifically in MyHCIIA, MyHCIIX and MyHCIIB myofibers and responsible for the expression of a single fast Myh gene at the locus remains to be established. We also observed that in cSix1 mutant SOL, SDH and GPDH activities are decreased, and that many genes coding for glycolytic proteins are downregulated. More particularly, we identified a decrease of Phosphofructo-kinase, of AldolaseA, of Glyceraldehyde-3-phosphate dehydrogenase and of Enolase 3 mRNAs, arguing for a decreased glucose use in cSix1 mutant SOL. Furthermore, we observed an increase of Pdk3 and a decrease of Pdp1 mRNAs, which should lead to a decrease of Pyruvate Dehydrogenase activity and consequently a decrease of acetyl CoA production. Last, expression of Isocitrate dehydrogenase 3, higher in MyHCIIA and MyHCIIX myofibers [5], is also decreased in mutant SOL. Altogether these results suggest a concerted control of sarcomeric genes and of genes controlling glucose metabolism by Six1 homeoproteins.

We observed that 3-week-old animals show already a lower accumulation of Six1 in myonuclei of the SOL as compared to the fast gastrocnemius. Mechanisms underlying Six1 nucleocytoplasmic shuttle and specific accumulation in perinatal and adult fast myofibers remain to be identified. At birth, SOL myofibers express the embryonic MyHCemb, neonatal and the slow MyHCI genes [53]. Fast MyHCIIA is detected at the mRNA level at post-natal stages, and its expression increases during the 3 weeks after birth at the expense of MyHCemb and MyHCneo, whose expression declines with neuromuscular junction (NMJ) maturation [2]. We observed that in cSix1 SOL the transition from MyHCemb to MyHCIIA is impaired, and that three weeks old mutant animals still express MyHCemb in the SOL, showing the requirement of Six1 for the transition from embryonic Myh to adult Myh expression and myofiber specialization.

During post-natal development, three main factors are involved in the emergence of adult myofiber specialization [2] and that may control Six1 accumulation specifically in fast myonuclei.

The first factor involves the influence of slow and fast neuromuscular junctions. Myofibers are still polyinnervated at birth in the SOL, and retraction of polyinnervation only takes place in the first weeks after birth giving rise to slow–twitch fatigue resistant and fast–twitch fatigue resistant alpha motoneurons on slow/MyHCI and fast/MyHCIIA myofibers, respectively [54, 55]. During this period, in the rodents’ SOL, the expression of embryonic and neonatal Myh is replaced by the expression of either adult MyHCI or MyHCIIA [53, 56]. Accordingly, low Six1 nuclear accumulation in MyHCI myofibers may be the consequence of slow motoneurons activity. The main second messenger of slow tonic firing in the myofiber is the Ca⁺⁺ concentration flux that modulates Calcineurin and CaMK activities [57]. Calcineurin phosphatase activity then controls the subcellular localization of NFATc transcription factors [14, 58]. Calcineurin signaling from the perinatal period on appears to be essential for fiber-type specialization, as supported by experiments in transgenic animals showing that its blockade by forced MCIP1 expression impairs slow fibers specialization [59].

In SOL of Myod1-MCIP1 transgenic mice, Myoglobin expression is not reduced, while the expression of MyHCI is downregulated from day 7 and is undetected from day 14 [59]. We did not detect in cSix1KO or ciSix1 SOL modification of the expression of Myoglobin, suggesting that in both cases mutant myofibers are not completely reprogrammed. Transient Six1 expression in perinatal myofibers of cSix1 SOL may explain why MyHCIIA is transiently expressed in the perinatal period, until the end of SC accretion that takes place in the first 3 weeks post natal [39], and of regular Six1 positive nuclei supply. Thus, the incomplete reprogramming observed in adult cSix1 mutant SOL may be the consequence of accretion of new satellite cells into the growing post-natal fast MyHCIIA myofibers that provides transiently a genetically or epigenetically programmed fast phenotype that impairs their total reprogramming.

It is known that MyHCI expression during the perinatal period as well as in the adult is nerve-dependent [60–62]. Accordingly, all cSix1 mutant SOL myofibers that express only MyHCI should be innervated by slow motoneurons. We observed a strong decrease of the expression of R-Spondin3 (Rspo3) in cSix1 mutant myofibers. R-spondins are secreted proteins known to enhance Wnt/β-catenin signaling [63] which is a major actor of neuromuscular jonctions (NMJ) [64]. Mice mutant for β-catenin in the myofiber show presynaptic differentiation defects [65]. Whether a Rspo3/β-catenin pathway controlled by Six1 may favor specific stabilization of fast–twitch fatigue resistant alpha motoneuron on future MyHCIIA myofibers is an interesting issue that remains to be tested. Related to this observation, we noted the increase of the expression of Chnrγ and Chnrα1, suggesting NMJ remodeling in cSix1 SOL.

A second important mechanism involved in fast myofiber specialization during the perinatal period concerns the influence of thyroid hormone [66, 67]. In agreement, hypothyroidic animals show delayed fast Myh transition and prolonged MyHCemb expression [68], a phenotype also found in cSix1 mutant SOL. Affymetrix transcriptomic analysis did not reveal a link between the presence of Six1 in SOL and the level of thyroid receptors (TR) or TR co-factors expression. This suggesting that the absence of fast phenotype acquisition in cSix1 mutant SOL is not due to a decreased expression of TR, although more detailed analysis of the expression of miR-133a1, a direct TR target gene in adult skeletal muscle involved in the control of TEAD1 expression [67], remains to be performed. While Six1 does not control TR expression in SOL, the possibility that nuclear Six1 accumulation in fast myofibers could be controlled by the thyroid hormone axis remains an interesting possibility.

A third factor that may participate in myofiber specialization involves intrinsic cell autonomous differences in distinct population of myogenic progenitors leading to myofibers heterogeneity. It is suspected that adult satellite cells associated with slow or fast muscles have intrinsic different genetic properties, although extrinsic factors arising in extracellular matrix or from muscle position and usage in the limb has not been completely excluded [69]. It is therefore tempting to speculate that satellite cells (SC), associated with future MyHCIIA and MyHCI myofibers in the SOL, accreted in the growing myofiber during post-natal development possess specific heritable properties. As myonuclei present in the SOL at 3 weeks show lower Six1 accumulation as compared with myonuclei of the GP, this may suggest that perinatal SC accreted in growing SOL and GP myofibers express different level of Six1, depending of their localization in the niche of MyHCI or subtypes of fast Myh myofibers. It would thus be interesting to test Six1 expression level in GP and SOL associated SC to confirm this hypothesis.

Fast MyHCIIA expression in adult soleus

In the mammalian genome, a number of genes are organized in clusters such as the β-globin [70], the Hox [71], and the fast Myh cluster [52]. Within these clusters, the precise order of the genes allows their sequential expression through shared enhancers. We showed previously that in adult fast TA and GP muscles Six1 is bound on a central enhancer located at the Myh fast locus between MyHCemb and MyHCIIA, and that it controls the expression of MyHCIIB [25]. We show here that in addition Six1 is essential for efficient fiber type shift from MyHCemb to MyHCIIA in SOL muscles during postnatal development (Fig. 2). The above mentioned fast Myh enhancer is able to activate the transcription of the MyHCIIX, MyHCIIA and MyHCIIB genes and may control higher order chromatin conformation at the locus to allow a single fast Myh gene to be expressed in all myonuclei of each fiber [25]. MyHCIIA activation during the perinatal period in SOL may also depend on this enhancer, and on its own promoter elements known also to interact with Six1 [25, 72].

It will be of major interest to unravel the mechanisms presiding the choice of the expression of a single fast Myh gene at the locus in a given myofiber during the perinatal period, its coordinated associated metabolic specialization and the involvement of Six1 in this matter. In the case of the β-globin, an LCR localized upstream of Globin genes controls the spatiotemporal and sequential expression of each gene at the locus. The transcription factor Sox6 binds to each private Globin regulatory elements and coordinates the interactions with the LCR through chromatin conformation modification [73, 74]. Interestingly, Sox6 also binds intergenic regions of Myh locus in C2 myotubes [22, 24]. An interesting hypothesis would thus be that mechanisms similar to those controlling the β-globin locus could act at the Myh locus with Six1 and Sox6 cooperating to orchestrate the spatiotemporal expression of the Myh fast genes. Sox6 mRNA is expressed in adult fast type muscles, not in slow type muscles [22], and skeletal muscle specific Sox6 knockout mice showed fiber type transition from fast to slow. cSix1 KO SOL mice have an increased expression of slow type genes that may be partly the consequence of decreased expression of Sox6 mRNA level observed. We have shown previously that Six homeoproteins also control the nuclear accumulation of Sox6 in fetal muscle [29], and we cannot exclude that in cSix1 KO this is not also the case. It would be interesting to determine whether Six1 is relocalized ouside of the nucleus in adult Sox6 mutant myofibers since all myofibers in SOL of Sox6 mutant mice express MyHCI [22, 24] and that in transgenic mice overexpressing Sox6 the expression of slow type genes like MyHCI and Tnni1 is down regulated [75]. The consequence of Sox6 down regulation observed in cSix1 SOL may also be responsible of Prox1 up regulation. Prox1 is a known repressor of fast muscle gene expression in Zebrafish lying downstream of Sox6 [76] and its deletion in mouse leads to increased expression of fast MyHCIIA, Tnni2, and Tnnt3 genes in SOL [49]. Analysis of the consequences of Prox1 knock down in cSix1 SOL would allow to test its involvement in the down regulation of MyHCIIA, Tnni2, and Tnnt3.

Six genes redundancy in adult myofibers

In our study, HSA-CREert2 recombinase induced deletion of Six1 in 2 months adult myofibers led to a switch toward a slow phenotype within 4 weeks following Six1 deletion. This inducible switch was characterized by the decreased expression of MyHCIIX, MyHCIIA, Pvalb, AldoA, and Eno3 and an increased number of MyHCI positive myofibers. In this model, the up-regulation of slow-type muscle genes 4 weeks after Six1 deletion was less obvious than observed in HSA-CRE animals, possibly because in this paradigm the expression of Sox6, which inhibits a battery of slow muscle genes, is faintly down-regulated. It is possible that during this time period, Six1 direct targets are efficiently down-regulated, while the up-regulation of slow-type genes might require other modifications such as a switch from fast to slow innervation, and/or decrease of Sox6 expression. Furthermore, the nuclear level of Six4 and Six5 homeoproteins, both expressed in adult myofibers, might be higher in SOL of ciSix1 than in cSix1. In adult muscles, the Six4-Baf60c transcription complex controls Deptor expression and transgenic animals with muscle-specific forced expression of Deptor show a switch from oxidative to glycolytic metabolism and are protected from diet-induced insulin resistance [77]. In cSix1 mutant SOL, the level of Deptor mRNA is unchanged, suggesting that the specific Six4/Deptor genetic axis, if maintained in cSix1 mutant SOL, is sufficient to maintain the level of expression of genes coding for proteins of the glycolytic pathway. cSix1 SOL presents a severe decrease of GPDH enzymatic activity, which may be correlated to a decreased expression of Gpd1 in mutant myofibers. Whether further decreased Six activity by combining Six1 and Six4 loss in adult myofibers would lead to increased defects of glycolytic flux remains to be evaluated.

Pvalb is a calcium binding protein responsible for fast calcium concentration decrease and relaxation in fast type muscle fiber after excitation. Pvalb knockout mice showed prolonged contraction-relaxation cycle [11] and increased fatigue resistance [44] associated with increased mitochondria composition without major modifications of fiber type specific sarcomeric proteins [43]. We demonstrated that Pvalb is a direct target of Six1 (Fig. 6), but did not observe up regulation of oxidative metabolism components neither at the mRNA level nor at the protein level (Figs. 3, 4, 5, and 6) in cSix1 mutant SOL, with on the contrary a decreased SDH activity. Recent study with single fiber proteomics revealed that MyHCI fibers have less mitochondrial proteins and SDH activity compared to MyHCIIA fibers [5]. The loss of MyHCIIA fiber in SOL of Six1 KO mice might explain the decreased SDH activity observed. In addition, down-regulation of Esr1 (estrogen receptor 1), which is known to activate mitochondrial oxidation activity in females [48], was also observed in SOL muscles of cSix1 KO males. These changes may participate in down-regulation of oxidative metabolism in mutant samples (Fig. 3). In that context, down regulation of Esr1 could cancel the effect of Pvalb expression decrease in regard of mitochondrial activity.

Conclusions

We present the evidence that Six1 homeoproteins are required in mouse SOL muscles for the acquisition of the fast MyHCIIA phenotype by controlling MyHCIIA expression and the expression of other fast-type muscle genes among which Sox6, and by controlling efficient glucose utilization through the control of glycolytic and of Krebs cycle genes.