FHL1 induces formation of hypertrophic myotubes, dependent upon calcineurin signaling. C2C12 myoblasts were transiently transfected with HA-βgal control (A, B, and E) or HA-FHL1 (C, D, and F), and left untreated (A–D) or treated with the calcineurin inhibitor CsA (E and F) before and during induction of differentiation in low serum media for 96 h. Transfected cells were detected using an HA antibody (green) and costained with propidium iodide (red) to identify nuclei. Cells were visualized using laser scanning confocal microscopy. Arrows indicate hypertrophic myotubes. Bars, 100 μm.

Generation of FHL1 transgenic mice. (A) The linearized transgene contains the 2.2-kb HSA promoter, fused in frame to the human FHL1 coding sequence with an N-terminal HA tag and a C-terminal bovine growth hormone polyA tail (BGHpA). FHL1-RBM mutants C132F and H123Y are indicated. (B) Skeletal muscle lysates were immunoblotted for HA-FHL1 or β-tubulin in wild-type (WT) or FHL1 transgenic (TG) muscle (GAS, gastrocnemius; EDL, extensor digitorum longus; SOL, soleus plantaris; PLA, plantaris; ECU, extensor carpi ulnaris; FDP, flexor digitorum profundus). A lysate prepared from C2C12 cells transfected with HA-FHL1 was used as a positive control (+). (C) The relative expression of HA-FHL1 was determined by densitometry of HA-immunoreactive polypeptides, standardized relative to β-tubulin loading. HA-FHL1 expression is represented as a fold difference, relative to GAS, arbitrarily defined as one. Graph depicts mean ± SEM (GAS, EDL, SOL, and PLA: n = 3 mice; ECU and FDP: ***, P < 0.005; n = 1 mouse). (D) Tissue lysates from WT or TG mice were immunoblotted for HA-FHL1 using anti-HA antibodies or β-tubulin. Lysates prepared from C2C12 cells transfected with HA-FHL1 were used as a positive control (+). (E) Longitudinal gastrocnemius muscle sections from WT or FHL-TG mice were stained with anti-FHL1 (green) or anti-HA (red) antibodies and imaged by confocal microscopy, then images were further resolved by deconvolution. Bars, 5 μm.

FHL1 forms a complex with NFATc1. (A) GST-FHL1, GST-GATA-2, or GST and His-NFATc1 or His were coexpressed as indicated in E. coli. GST proteins and associated proteins were bound to glutathione-Sepharose. Western blots are representative of three experiments. Arrows indicate GST- and His-tagged proteins. The His tag alone did not copurify with GST-FHL1 (not depicted). (B) GST-FHL1 or GST coupled to glutathione-Sepharose was incubated with skeletal muscle lysates from wild-type FVB/n mice and immunoblotted for endogenous NFATc1, GST, or GST-FHL1 as indicated. Results shown are representative of four similar experiments. (C) Coimmunoprecipitation of HA-FHL1 and myc-NFATc1 in C2C12 cells. C2C12 cells were transfected with the indicated constructs and immunoprecipitated with a myc antibody, then immunoblotted with an HA antibody (top). Immunoprecipitation of myc-NFATc1 was confirmed by immunoblotting with myc antibody (middle). Lysates were immunoblotted with an HA antibody to confirm equal expression of FHL1 constructs (bottom). Immunoblots are representative of three experiments. (D) Densitometric quantification of HA-FHL1 wild-type or C132F/H123Y mutants, relative to immunoprecipitated myc-NFAT. The bar graph represents the mean ± SEM (n = 3 experiments; **, P < 0.01).

FHL1 regulates the transcriptional activity of NFATc1 and expression of the calcineurin-responsive gene GATA-2. (A) C2C12 myoblasts were transiently transfected with HA-tagged wild-type (WT) FHL1, FHL1^C132F, or FHL1^H123Y and allowed to differentiate for 72 h, then costained with anti-HA antibody (green) and To-Pro-3 (blue) to detect nuclei. Cells were imaged using confocal microscopy. Arrows indicate hypertrophic myotubes. Bars, 100 μm. (B) Maximum myotube diameter was measured in myotubes transiently transfected with HA-tagged wild-type (WT) FHL1, FHL1^C132F, FHL1^H123Y, or HA-βgal control, then differentiated for 96 h and stained as described in Materials and methods. The bar graph represents the mean frequency of myotubes with diameters of 0–20 μm, >20–40 μm, and >40 μm, ±SEM (n > 300 myotubes from three experiments; n = 3; *, P < 0.05). (C) C2C12 myoblasts were cotransfected with various combinations of expression vectors as indicated, and with NFATc1-dependent (pGL2-IL-2) and Renilla luciferase reporters. Transfected myoblasts were maintained in growth media for 48 h and assayed for luciferase activity. Gal4DBD activity is represented by white bars and Gal4DBD-NFATc1 by gray bars. Error bars represent ± SEM (n ≥ 5 experiments; *, P < 0.05). (D) Transfected myoblasts were differentiated for 48 h and assayed for luciferase activity (48 h; n = 4 experiments). (E) Gastrocnemius (GAS) lysates from wild-type (WT) or HA-FHL1 transgenic (TG) mice were immunoblotted for GATA-2 or β-tubulin (loading control). (F) GATA-2 protein expression levels in WT (white) versus TG (gray) gastrocnemius muscle was quantified using densitometry and standardized to the loading control. The bar graph represents the mean ± SEM (n = 5 mice per genotype; *, P < 0.05). (G) Transverse gastrocnemius muscle sections were costained with anti–GATA-2 (green), anti-vinculin (red; sarcolemma) antibodies, and To-Pro-3 iodide nuclear stain (blue), then imaged by confocal microscopy. Images were taken at the same intensity. Bars, 40 μm. (H) C2C12 cells were cotransfected with various combinations of expression vectors for Gal4DBD (white) or Gal4DBD-NFATc1 (gray) and either scrambled (control) or FHL1 siRNA oligonucleotides, as indicated, with NFATc1-dependent (pGL2-IL-2) and Renilla luciferase reporters. Assays were performed under growth conditions. Error bars represent ±SEM (n = 5 experiments; *, P < 0.05). (I) Luciferase assays performed in H were repeated under differentiation (48 h) conditions (n = 3 experiments; *, P < 0.05).

FHL1 promotes enhanced myocyte fusion. (A) C2C12 myoblasts stably expressing HA-βgal or HA-FHL1 were differentiated and costained with anti-MHC (MF20, green), myogenin antibodies (blue), and propidium iodide (red). Cells were imaged using confocal microscopy. Shown are representative images after 96 h of differentiation. Myoblast differentiation was quantified from 0–96 h. Arrows indicate examples of transfected myotubes. Bar, 70 μm. (B) The percentage of total nuclei positive for myogenin. (C) The differentiation index, determined by scoring the proportion of total nuclei within MHC-positive cells (myocytes and myotubes). (D) The fusion index, the mean number of nuclei per MHC-positive cell. Black and gray lines represent HA-βgal control and HA-FHL1–expressing C2C12 cells, respectively. Data represent the mean from three independent experiments (n = 3) ± SEM in which ≥100 cells were counted for each replicate. (*, P < 0.05; **, P < 0.01). (E) Lysates (25 μg) from differentiating C2C12 cells stably expressing HA-βgal or HA-FHL1 were immunoblotted with MyoD, Myf5, myogenin, MHC, and β-tubulin antibodies (loading control). (F) MHC expression was analyzed by densitometry and quantified relative to β-tubulin. Data represent the mean from three independent experiments ± SEM (n = 3; **, P < 0.01). (G) The DNA/protein ratio was determined. Data represent the mean from four independent experiments ± SEM (n = 4; **, P < 0.01).

FHL1 transgenic mice exhibit skeletal muscle hypertrophy, increased strength, and fatigue resistance. (A) Transverse gastrocnemius sections were stained with a dystrophin antibody and viewed by confocal microscopy. Higher-magnification images are shown in the insets, with one fiber outlined. Bars, 50 μm. Images were analyzed for fiber area (B) and diameter (C; n = 3–4 mice; *, P < 0.05). GAS, gastrocnemius. (D) Transverse FDP sections were stained with MHC isoform-specific antibodies, and the percentage of 2A, 2X, and 2B fibers was calculated. The bar graph displays the frequency of each fiber type in wild-type (WT) versus transgenic (TG) mice (n = 3–5 mice; ***, P < 0.005). (E) Transverse FDP sections were stained for 2B MHC fibers; one fiber is outlined per image. Bars, 40 μm. (F) The diameters of FDP-2B fibers from WT versus TG mice are shown (n = 3–4 mice; *, P < 0.05). (G) The frequency of Pax-7–positive nuclei in the gastrocnemius was scored, relative to total nuclei. The bar graph represents the frequency of Pax-7–positive nuclei for WT and TG mice (n = 3–5 mice; **, P < 0.01). (H) Transverse FDP sections were hematoxylin and eosin stained. Arrows indicate centralized nuclei. Bars, 40 μm. (I) The frequency of centralized nuclei were scored, relative to total myofiber nuclei, in transverse gastrocnemius and FDP sections (n = 4–6 mice; *, P < 0.05; ***, P < 0.001). (J) A whole animal strength test measured overall limb and abdominal strength as a percentage pass rate over 15 trials for WT versus TG 12-mo-old mice (n = 4–8; **, P < 0.01). All graphs represent mean ± SEM; WT, white; TG, gray. (K) Muscle fatigability was assessed as a drop in force with repeated tetanic contractions in EDL muscles from WT (black) versus TG (red) mice, expressed as a percentage of initial contraction over time, mean ± SEM (n = 7, P < 0.02).

RBM-FHL1 mutations alter localization of NFATc1. (A) C2C12 myoblasts were cotransfected with HA-FHL1 wild-type (top), HA-FHL1^C132F (middle), or HA-FHL1^H123Y (bottom), and myc-NFATc1, then differentiated for 72 h. Myotubes were fixed and costained with HA (green) and myc (red) antibodies, costained with menadione-NBT, and imaged by fluorescence microscopy. White arrowheads indicate perinuclear aggregates colocalizing with the reducing-body marker menadione-NBT in the absence of substrate (black arrowheads). Yellow arrowheads reveal the region shown in the high-magnification images showing colocalization between FHL1, NFATc1, and menadione-NBT in the insets. (B) Transverse skeletal muscle sections were prepared from human RBM muscle biopsies, with FHL^C132F (top) or FHL1^H123Y (bottom) mutations. Sections were stained with antibodies recognizing FHL1 (green) and NFATc1 (red), then costained with DAPI (nuclei). Arrowheads identify reducing bodies. (C) Vector control, FHL1 wild-type (FHL1-WT), or indicated FHL1 RBM mutants were cotransfected with myc-NFATc1 into C2C12 myoblasts. Myoblasts were differentiated for 72 h, then ionomycin-stimulated for 30 min, and costained with anti-myc (NFATc1, glow-over images are shown) and anti-HA (to identify cotransfected cells, not depicted here). White arrows identify NFATc1 nuclear staining, yellow arrows indicate NFATc1-negative nuclei. High-magnification images are shown in the insets. (D) The percentage of transfected cells containing nuclear NFATc1 was scored after ionomycin stimulation. The bar graph equals mean ± SEM (n = 4 experiments; *, P < 0.05). Bars: (A and B) 20 μm; (C) 50 μm.

FHL1 regulates NFAT-mediated signaling in skeletal muscle. Wild-type FHL1 complexes with NFATc1 to increase its transcriptional activity, promoting skeletal muscle hypertrophy and oxidative fiber type expression. FHL1-RBM mutations sequester NFATc1 into reducing body aggregates, resulting in its reduced nuclear translocation and reduced activation of gene transcription.