Deficiency of slow skeletal muscle troponin T causes atrophy of type I slow fibres and decreases tolerance to fatigue

Abstract

The total loss of slow skeletal muscle troponin T (ssTnT encoded by TNNT1 gene) due to a nonsense mutation in codon Glu¹⁸⁰ causes a lethal form of recessively inherited nemaline myopathy (Amish nemaline myopathy, ANM). To investigate the pathogenesis and muscle pathophysiology of ANM, we studied the phenotypes of partial and total loss of ssTnT in Tnnt1 gene targeted mice. An insertion of neomycin resistance cassette in intron 10 of Tnnt1 gene caused an approximately 60% decrease in ssTnT protein expression whereas cre-loxP-mediated deletion of exons 11–13 resulted in total loss of ssTnT, as seen in ANM muscles. In diaphragm and soleus muscles of the knockdown and knockout mouse models, we demonstrated that ssTnT deficiency resulted in significantly decreased levels of other slow fibre-specific myofilament proteins whereas fast fibre-specific myofilament proteins were increased correspondingly. Immunohistochemical studies revealed that ssTnT deficiency produced significantly smaller type I slow fibres and compensatory growth of type II fast fibres. Along with the slow fibre atrophy and the changes in myofilament protein isoform contents, ssTnT deficiency significantly reduced the tolerance to fatigue in soleus muscle. ssTnT-deficient soleus muscle also contains significant numbers of small-sized central nuclei type I fibres, indicating active regeneration. The data provide strong support for the essential role of ssTnT in skeletal muscle function and the causal effect of its loss in the pathology of ANM. This observation further supports the hypothesis that the function of slow fibres can be restored in ANM patients if a therapeutic supplement of ssTnT is achieved.

Introduction

Striated muscle contraction is regulated by Ca²⁺ via the troponin complex that consists of three protein subunits: troponin C (TnC), the Ca²⁺ receptor; troponin I (TnI), an inhibitor of actin-activated myosin ATPase; and troponin T (TnT), which anchors the troponin complex on the tropomyosin–actin thin filament (Perry, 1998; Chalovich, 2002; Jin et al. 2008; Wei & Jin, 2011). TnT plays a pivotal role in regulating the conformational changes in the thin myofilament regulatory system during the excitation–contraction–relaxation cycles of cardiac and skeletal muscles (Jin et al. 2008; Wei & Jin, 2011).

Vertebrate TnT has a size of ∼35 kDa and contains a viable N-terminal region and conserved middle and C-terminal regions (Jin et al. 2008; Wei & Jin, 2011). Encoded by three homologous genes, TnT is present in three muscle fibre type-specific isoforms: the slow skeletal muscle TnT (TNNT1), fast skeletal muscle TnT (TNNT3) and cardiac TnT (TNNT2) (Wei & Jin, 2011). Via alternative splicing, each TnT gene encodes multiple protein variants to add more functional modifications.

Among the three TnT isoform genes, TNNT1 is the newest member evolved. The mammalian slow skeletal muscle TnT gene consists of 14 exons and is expressed exclusively in type I slow muscle fibres (Jin et al. 2000; Chong & Jin, 2009). Alternative splicing of exon 5 of Tnnt1 transcripts generates two protein variants differing in the N-terminal region (Larsson et al. 2008; Wei & Jin, 2011).

A nonsense mutation in codon Glu¹⁸⁰ in exon 11 of TNNT1 gene is found in the Old Order Amish population to cause a progressive and postnatally lethal nemaline myopathy (ANM) with recessive inheritability (Johnston et al. 2000). The affected children have skeletal muscle weakness and die from the failure of respiratory muscles usually in their second year of life.

The nonsense mutation in Glu¹⁸⁰ result in a deletion of the C-terminal 83 amino acids of slow skeletal TnT (ssTnT), losing several binding sites for TnI, TnC and tropomyosin (Jin et al. 2003). The C-terminal truncated ssTnT protein was not detectable in the muscle samples from ANM patients, consistent with the recessive inheritance of ANM. Thus the truncated ssTnT, which lacks the T2 region tropomyosin-binding site, is not able to incorporate into myofilaments and is degraded in the muscle cells (Jin et al. 2003). The loss of ssTnT protein in slow muscle fibres is thus the molecular basis for the pathogenesis and pathophysiology of ANM (Jin et al. 2003; Wang et al. 2005).

Based on the integrative link between the cardiac troponin I (cTnI) gene and slow skeletal TnT gene, we have shown that the expression of Tnnt1 gene was decreased by ∼40% at the protein level in cardiac TnI gene-deleted mice (Feng et al. 2009). The decrease in ssTnT caused apparent atrophy of type I slow fibres and a slow-to-fast functional switch of the diaphragm muscle with compromised fatigue tolerance (Feng et al. 2009).

In the present study, we generated mouse lines with knockdown (ssTnT-KD) and knockout (ssTnT-KO) of ssTnT gene expression. ssTnT deficiency decreased the levels of other slow fibre-specific myofilament proteins in diaphragm and soleus muscles whereas fast fibre-specific myofilament proteins were increased. ssTnT deficiency also produced significantly smaller type I slow fibres and compensatory growth of type II fast fibres together with a shifted force–frequency relationship towards a fast muscle type and significantly reduced tolerance to fatigue in soleus muscle. ssTnT-deficient soleus muscle also contained a significant number of central nuclei type I fibres, indicating adaptive regeneration. The data provide strong support for the essential role of ssTnT in skeletal muscle function and the causal effect of its loss on the pathology of ANM.

Methods

ssTnT-KD and ssTnT-KO mice

All animal procedures were approved by the Institutional Animal Care and Use Committees and were conducted in accordance with the Guiding Principles in the Care and Use of Animals, as approved by the Council of the American Physiological Society.

Genomic DNA segments containing the slow skeletal muscle TnT gene were cloned from a 129SvJ strain mouse genomic DNA library as described previously (Huang et al. 1999). Restriction enzyme mapping, Southern blot analysis, and partial sequencing were carried out to verify the cloned mouse genomic DNA sequence in comparison with the Tnnt1 gene in the database (Gene ID: MGI:1333868).

A mouse Tnnt1 targeted conditional mutagenesis construct was generated using the pPNT4 vector (Conrad et al. 2003), provided by Dr Marcus Conrad, Institute of Clinical Molecular Biology and Tumor Genetics GSF-Research Centre for Environment and Health, Germany). In the construct, two loxP sequences were inserted in intron 10 and intron 13 in tandem orientation to allow Cre recombinase-catalysed deletion of exons 11–13. This deletion not only removes the portion of the coding sequence but also results in a reading frame shift in the downstream mRNA with a stop codon to produce a truncated ssTnT protein mimicking that in Amish nemaline myopathy. This Cre-mediated deletion of exons 11–13 can be induced in whole animals as well as in tissues or in cultured myocytes.

A neo^R cassette constructed in the pPNT4 vector adjacent to the downstream loxP sequence was inserted into intron 10 of the Tnnt1 targeting construct for neomycin selection of transfected embryonic stem (ES) cells. The neo^R cassette is flanked by two frt sequences and can be deleted via FLP1-catalysed recombination (Meyers et al. 1998). This mechanism allows removal of the neo^R cassette after establishing the targeted loxP mutagenesis to avoid the effect of neo^R insertion on Tnnt1 expression. The induction of neo^R removal can be done in whole animals, tissues or cells.

Transfection of mouse HM-1 ES cells (x,y) with the Tnnt1 gene targeting DNA construct using electroporation was carried out at the Northwestern University Feinberg School of Medicine Gene Targeting and Transgenic Core Facility. Colonies of transfected ES cells were selected by the acquisition of neomycin resistance. Genomic DNA from the drug resistant ES cell colonies was extracted and screened with Southern blots using cloned 5′ and 3′ flanking genomic DNA probes (Huang et al. 2008) for the homologous recombination-generated change in BamHI restriction pattern.

A Tnnt1 gene targeted ES cell clone (3F3) was used to produce chimeric mice. The blastocyst injection and embryo re-implantation were carried out at the Northwestern University Feinberg School of Medicine Gene Targeting and Transgenic Core Facility. The 129SvJ mouse originated (albino) ES cells are injected into C57BL/6 (black) mouse blastocysts to produce chimeras. High chimerism males were selected to mate with C57BL/6 females for germ line transmission of the ES cell genotype. Positive offspring were first selected by the light brown coat colour in contrast to the pure C57BL/6 black littermates. The presence of targeted Tnnt1 allele in the pups was then genotyped by polymerase chain reaction (PCR) on genomic DNA extracted from tail biopsies. Two pairs of PCR primers were designed to identify the presence of the downstream loxP and the neo^R cassette, respectively. Mice bearing the targeted Tnnt1 allele were selected to breed with C57BL/6 mice for over nine generations to obtain uniform genetic background.

Removal of the neo^R cassette inserted in intron 10 was achieved by crossing the Tnnt1-flox-neo line with a Gt(ROSA)26Sor-FLP transgenic mouse line (Jackson Lab) that expresses FLP1 recombinase in most tissue types, including the developing germ line. Disruption of the Tnnt1 gene through deletion of exons 11–13 was obtained by crossing the Tnnt1-flox mouse line with the Zp3-cre mouse line (Jackson Lab) which expresses Cre recombinase in the female germ line.

SDS-PAGE and Western blotting

The expression of myofilament proteins in mouse skeletal muscles was examined using SDS-polyacrylamide gel electrophoresis (PAGE) and Western blotting. After killing by intraperitoneal injection of pentobarbital, 100 mg/kg body weight, muscle samples were immediately isolated and snap frozen on dry ice. Soleus and diaphragm muscles were homogenized in SDS-gel sample buffer containing 150 mm DTT, 2% SDS, and 50 mm Tris-HCl, pH 8.8. The tissue homogenates were heated at 80°C for 5 min, centrifuged in a microcentrifuge at top speed for 5 min to remove insoluble materials, and used for SDS-PAGE and Western blot analysis or stored at –80°C for later use.

As described previously (Feng et al. 2009), the muscle protein extracts were resolved on 14% Laemmli gels with an acrylamide to bis-acrylamide ratio of 180:1. The resulting gels were stained with Coomassie Blue R250 or electrically blotted to nitrocellulose membrane using a semidry transfer device (Bio-Rad, Hercules, CA, USA). The membranes were blocked with 1% bovine serum albumin (BSA) in Tris-buffered saline (TBS) and probed at 4°C overnight with anti-ssTnT monoclonal antibody (mAb) CT3, anti-fast skeletal muscle TnT mAb T12, and anti-TnI mAb TnI-1, diluted in TBS containing 0.1% BSA. After washes with TBS containing 0.5% Triton X-100 and 0.05% SDS to remove excessive primary antibody, membranes were subsequently probed with alkaline phosphatase conjugated anti-mouse IgG second antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) at room temperature for 1 h. After washes, the protein bands recognized by the mAbs were visualized via 5-bromo-4-chloro-3-indolylphosphate–nitro blue tetrazolium substrate reaction. The SDS-gel and Western blots were scanned at 600 dpi and two-dimensional densitometry was performed using ImageJ software to determine the relative levels of TnT and TnI.

Glycerol SDS-PAGE

Glycerol SDS-PAGE was carried out as previously described (Feng et al. 2009). Briefly, 8% SDS polyacrylamide gels with an acrylamide:bisacrylamide ratio of 50:1 were prepared in 1.5 m Tris buffer (pH 8.8) containing 30% glycerol. Myosin heavy chain (MHC) isoforms were resolved by running at 100 V at 4°C for 24 h. The resulting gels were stained with Coomassie Blue R250 and scanned for densitometry analysis.

Immunohistochemistry

Immediately after killing by intraperitoneal injection of pentobarbital, 100 mg/kg body weight, soleus and diaphragm muscles were rapidly removed and embedded in OCT compound, frozen at their in vivo length by exposure to liquid nitrogen-chilled 2-methylbutane (–159°C) for 30 s, and stored at –20°C overnight before cryosectioning. Cryosections (5 μm) were cut using a Leica CM 1950 cryostat.

Muscle sections were blocked in phosphate-buffered saline (PBS) containing 0.05% Tween-20 (PBS-T) and 1% BSA at room temperature for 30 min. Endogenous peroxidase activity was inactivated by incubation with 1% H₂O₂ in PBS-T at room temperature for 10 min. After wash with PBS-T, the muscle sections were probed with hybridoma culture supernatants of anti-MHC I mAb FA2, mAb CT3 or SP2/0 myeloma control at 4°C overnight. After washes with PBS-T to remove excess primary antibody, muscle sections were incubated with horseradish peroxidase conjugated anti-mouse second antibody at room temperature for 1 h. After washes to remove excess second antibody, the labelling of MHC I and slow TnT was visualized via 3,3′-diaminobenzidine-H₂O₂ substrate reaction in a dark box for 30 s. The reaction was terminated by repeated washes with 20 mm Tris-HCl, pH 7.6. Nuclei were then counterstained with Haematoxylin for 5 min followed by washes with distilled water. The muscle sections were mounted in PBS containing 50% glycerol, sealed using Cytoseal, and photographed using a Zeiss Observer 125 microscope.

Contractility measurements

Immediately after killing by intraperitoneal injection of pentobarbital, 100 mg/kg body weight, soleus muscle were quickly removed from the mouse and immersed in Krebs solution (118 mm NaCl, 25 mm NaHCO₃, 4.7 mm KCl, 1.2 mm KH₂PO₄, 2.25 mm MgSO₄, 2.25 mm CaCl₂, and 11 mm d-glucose, pH 7.4 at room temperature) bubbled with 95% O₂ and 5% CO₂. Using a no. 6–0 silk suture, tendons at both ends of soleus muscle were tied to triangle-shaped steel hooks. One end of the muscle was mounted to the movable arm in a tissue bath (Radnoti, LLC, Monrovia, CA, USA), and the other to a 300B-LR force transducer (Aurora Scientific, Inc., Aurora, ON, Canada). Isometric twitch contractions were induced by field electric pulse stimulation (20 V, 0.2 ms) via a pair of platinum electrodes and a 701B stimulator (Aurora Scientific, Inc.). The contractile force of the soleus muscle was recorded through a 604C digital controller A/D interface and Aurora 600a 4CH software at 1000 Hz sampling frequency.

During the initial equilibration period, soleus muscles were adjusted to their optimum length for the production of maximal twitch force. After being stabilized for 30 min with twitch contractions, the muscle was stimulated to produce fused tetanic contractions (500 ms each minute for 10 min) using 20 V, 0.2 ms, 100 Hz field stimulation. The force–frequency relationship of the soleus muscle was examined by following the protocol described above but changing the stimulation frequency between 40 and 220 Hz.

As an intermittent fatigability test, soleus muscles were stimulated at 100 Hz as above to produce 300 ms tetanic contractions per second (30% duty cycle) for 5 min. At the end of the intermittent fatigue protocol, muscles were tetanically stimulated as described above once per minute for 20 min. All muscle tissues used in the contractility study were recovered for immunohistochemistry confirmation of ssTnT deficiency.

Data analysis

All quantitative data are presented as means ± SEM. Statistical analysis was done with Student's t test using the Microsoft Excel computer program (two-tailed assays unless noted in the figure legend).

Results

Generation of ssTnT knockdown and knockout mice

Using the gene targeting construct (Fig. 1A) derived from 129SvJ mouse genomic DNA (Huang et al. 1999) which is syngenic to the HM-1 mouse ES cells, successful targeting of the Tnnt1 gene was achieved. Targeted ES cell clone 3F3 produced chimeric mice with high frequency germ line transmission. PCR genotyping of the ES cell-originated offspring showed Mendelian segregation of the Tnnt1-flox-neo allele. PCR genotyping further demonstrated that the neo^R selection marker can be effectively removed by crossing with the Gt(ROSA)26Sor-FLP mouse line to produce the Tnnt1-flox allele (Fig. 1B). Crossing with the Zp3-cre mouse line effectively produced deletion of exon 11–13, generating the Tnnt1-exons 11–13 deletion allele (Fig. 1C).

Figure 1.

A, Tnnt1-neo-flox is the original Tnnt1 gene-targeted allele. B, the Tnnt1-flox allele is derived from the Tnnt1-neo-flox allele by FLP1-induced removal of the neo^R cassette. C, the Tnnt1-exon 11–13 deletion (KO) allele is derived by further deletion of exons 11–13 through Cre-induced recombination, truncating the ssTnT protein to mimic the Amish nemaline myopathy mutation. The positions of primers used in PCR genotyping are indicated on the maps.

Knockdown and knockout of ssTnT in Tnnt1-targeted mice

In contrast to the neonatal lethality of human ANM, mice homozygous for the Tnnt1-KD or Tnnt1-KO allele survive to adulthood and are fertile. The Western blots in Fig. 2A showed significant (∼60%) reduction of ssTnT in the soleus muscle of adult Tnnt1-KD mice and complete loss of ssTnT in soleus muscle of Tnnt1-KO mice.

Figure 2.

A, representative SDS-PAGE gel and CT3 mAb blot showed that ssTnT was significantly decreased in ssTnT-KD mouse soleus muscle and absent in ssTnT-KO soleus muscle. Fast TnT was increased in ssTnT-KD and ssTnT-KO soleus muscles as detected in the mAb T12 Western blot. The mAb TnI-1 Western blot showed that slow TnI was also decreased in ssTnT-KD and ssTnT-KO soleus muscles with a corresponding increase in fast TnI. B, densitometry quantification of the TnT Western blots showed that ssTnT decreased in ssTnT-KD soleus muscle to 40.65 ± 8.78% of the wild-type level. In ssTnT-KO soleus muscles, fast TnT increased to 158.70 ± 6.09% of the wild-type level. C, densitometry quantification showed in ssTnT-KD and ssTnT-KO mice soleus muscle, ssTnI protein level was decreased to 77.35 ± 4.96% and 54.82 ± 3.33% of the wild-type level, respectively. Only ssTnT-KO mouse soleus muscle showed significantly increase in fsTnI (140.99 ± 7.53% of the wild-type level). The ratio of slow-to-fast TnI was decreased with statistical significance in both ssTnT-KD (0.75 ± 0.06) and ssTnT-KO (0.41 ± 0.03) soleus muscles, as compared with wild type (1.06 ± 0.09). *P < 0.05 compared with wild type in two-tailed Student's t test. D, representative glycerol SDS-gel showed that MHC I decreased whereas MHC IIa and IIx increased in ssTnT-KD and ssTnT-KO soleus muscles. Densitometry quantification of the glycerol SDS-PAGE found that MHC I accounted for 53.52 ± 8.42% of total MHC in wild-type adult mouse soleus muscle and was decreased to 27.96 ± 3.10% and 16.96 ± 5.60%, respectively, in ssTnT-KD and ssTnT-KO soleus muscles. Correspondingly, MHC IIa and IIx increased in ssTnT-KD and ssTnT-KO soleus muscles to 40.79 ± 1.34% and 55.97 ± 4.63%, respectively, compare with the wild-type level of 40.54 ± 6.21%. *P < 0.05 compared with wild-type MHC I in two-tailed Student's t test; ^&P < 0.05 compared with wild-type MHC IIx in two-tailed Student's t test; ^#P < 0.05 compared with wild-type MHC IIa in two-tailed Student's t test.

Although the deletion of exons 11–13 only truncated the T2 region of the ssTnT polypeptide (Huang et al. 1999), no truncated ssTnT protein was detected using the anti-T1 region mAb CT3 (Jin et al. 2003) in Tnnt1-KO mouse muscles. This phenotype is consistent with the complete loss of ssTnT protein in human ANM muscle due to the nonsense stop codon at Glu¹⁸⁰, which also truncates the T2 region of the ssTnT polypeptide chain (Jin et al. 2003).

A side effect of the insertion of neo^R cassette in the intron 10 of Tnnt1 gene (Fig. 1A) is a significant knockdown of ssTnT expression in soleus muscle of Tnnt1-flox-neo homozygotes (Fig. 2A), likely to be due to interruption of the transcription of Tnnt1 pre-mRNA or splicing. The ssTnT knockdown mice also survive to adulthood and are fertile. Deletion of the neo^R cassette by FLP1 recombinase rescued Tnnt1 gene expression in Tnnt1-flox mice. Therefore, the Tnnt1-flox-neo allele provides a valuable experimental system of reversible ssTnT knockdown for investigating the functional effects of rescuing ssTnT expression on muscle phenotypes.

Increased fast TnT, decreased slow TnI and decreased MHC I in ssTnT-KD and ssTnT-KO soleus muscles

To investigate the effect of ssTnT deficiency on the expression of other myofilament proteins, we first examined the level of fast skeletal muscle TnT (fsTnT). Using mAb T12, Western blots and densitometry analysis showed that ssTnT deficiency was accompanied with an increase in fsTnT by ∼60% in soleus muscles of ssTnT-KO mice (Fig. 2B). fsTnT is normally expressed as multiple splicing variants and the splicing pattern had no detectable change in ssTnT-KD and ssTnT-KO soleus muscles (Fig. 2A).

Western blots using mAb TnI-1 recognizing both slow and fast isoforms of TnI showed that ssTnT deficiency caused a decrease in slow TnI and a corresponding increase in fast TnI (Fig. 2C). The ratio of slow/fast TnI was 1.06 ± 0.09 in soleus muscles from wild-type mice, consistent to its mixed fibre nature. The slow/fast TnI ratio decreased to 0.75 ± 0.06 in ssTnT-KD soleus muscle and to 0.41 ± 0.03 in ssTnT-KO soleus muscle (Fig. 2C). The observation that the complete loss of ssTnT in soleus muscle of Tnnt1-KO mice did not abolish the expression of slow TnI indicates a preserved slow fibre cellular environment.

We further examined the isoform of thick filament protein MHC in adult ssTnT-KD and ssTnT-KO soleus muscles. Glycerol SDS-PAGE showed significant decreases in slow fibre-specific MHC I in both ssTnT-KD and ssTnT-KO soleus muscles (Fig. 2D). Densitometry quantification showed that MHC I accounts for 53.52 ± 8.42% of total MHC in wild-type adult mouse soleus muscle. In soleus muscles of ssTnT-KD and ssTnT-KO mice, MHC I contents decreased to 27.96 ± 3.10% and 16.96 ± 5.60% of total MHC proteins, respectively. Accompanying the decreases in MHC I, fast fibre-specific MHC IIa and MHC IIx increased in ssTnT-deficient soleus muscles (Fig. 2D). The observation that the expression of MHC I was not abolished in ssTnT-KO soleus muscle is consistent with the preserved slow fibre environment indicated by the continuing expression of slow TnI (Fig. 2A).

ssTnT deficiency caused severe atrophy of type I slow fibres and hypertrophic growth of type II fibres fast in adult mouse soleus muscle

Type I slow fibres in cryosections of mouse soleus muscle were stained with anti-MHC I mAb FA2. FA2 negative myofibres were identified as type II fast fibres. The isolated muscles were cross-sectioned at their optimum physiological length, at which the maximum twitch contraction force was generated. We found in both ssTnT-KD and ssTnT-KO soleus muscles that ssTnT deficiency caused a ∼40% decrease in the cross-sectional area of type I slow fibres (Fig. 3).

Figure 3.

A, representative images of immunohistochemical staining of soleus muscle serial cross-sections using mAbs FA2 and CT3 to co-localize MHC I and ssTnT in fibres. B, quantification of fibre cross-sectional area showed that the size of type I fibres in ssTnT-KD and ssTnT-KO mouse soleus muscles decreased by 40.13 ± 4.21% and 38.05 ± 6.94%, respectively, compared to wild-type control. The size of Type II fibres increased in ssTnT-KD and ssTnT-KO soleus muscles by 47.73 ± 21.78% and 76.19 ± 15.06%, respectively, compared to wild-type controls. C–D, the total numbers of type I and type II fibres per soleus muscle (C) and soleus muscle weight normalized to body weight (D) of ssTnT-KD or ssTnT-KO mice were not significantly different from wild-type controls. *P < 0.05 compared with wild-type levels in two-tailed Student's t test.

To verify the comparison of cross-sectional area of muscle fibres, we also examined soleus muscles of wild-type and ssTnT-deficient mice cross-sectioned at their in vivo length. Consistent with the results from muscle sections obtained at their optimal length, the cross-sectional area of type I slow fibres are significantly decreased and fast type II muscle fibres increased in the ssTnT-deficient soleus muscles at their in vivo length (data not shown).

Serial cross-sections of soleus muscle were also stained with CT3 mAb (Fig. 3A) recognizing ssTnT, and aligned with the FA2 staining of MHC I fibres. The results showed that ssTnT was exclusively co-localized with MHC I within the type I slow fibres in wild-type soleus muscle and absent in type II fast fibres. The level of ssTnT in type I fibres of ssTnT-KD soleus muscle was drastically decreased and became absent in ssTnT-KO type I fibres (Fig. 3A).

The cross-sectional area of fast type II fibres from ssTnT-KD and ssTnT-KO soleus muscles increased by 47.73 ± 21.78% and 76.19 ± 15.06%, respectively, compared with wild-type controls (Fig. 3B). Along with the significantly smaller type I slow fibres, the results suggest that ssTnT deficiency caused compensatory hypertrophic growth of fast type II fibres. Nonetheless, mAb FA2 immunostaining of whole cross-sections of ssTnT deficiency soleus muscle showed no significant change in the number of slow and fast fibres (Fig. 3C). The weight of soleus muscle of ssTnT-deficient and wild-type mice normalized to body weight also did not show significant difference (Fig. 3D).

ssTnT deficiency caused a loss of type I slow fibres in adult mouse diaphragm muscle

Diaphragm muscle also contains slow twitch fibres and functions with continuing contractions to sustain respiration. To understand how deficiency of ssTnT contributes to the pathophysiology of respiratory muscles in ANM, we examined the expression of myofilament proteins and muscle fibre type composition in diaphragm muscles of adult ssTnT-KD and ssTnT-KO mice. Western blots showed that ssTnT was non-detectable in both ssTnT-KO and ssTnT-KD diaphragm muscles (Fig. 4A and B). There was near-complete loss of slow TnI with increases in fast TnT and fast TnI (Fig. 4A and C). Glycerol SDS-PAGE showed that slow fibre-specific MHC I was largely lost (Fig. 4D). mAb FA2 staining of cryosections of ssTnT-KD and ssTnT-KO diaphragm muscles showed that type I slow fibres were abolished, indicating a switch to nearly pure fast fibre muscle (Fig. 4E).

Figure 4.

A, representative SDS-PAGE gel and CT3 mAb Western blot showed significant decrease in ssTnT in diaphragm muscles of adult ssTnT-KD and ssTnT-KO mice. Fast TnT increased as shown in the mAb T12 Western blot. Slow TnI was decreased in ssTnT-KD and ssTnT-KO diaphragm muscle with a proportional increase in fast TnI. B, densitometry analysis of CT3 and T12 blots confirmed the absence of slow TnT and slight increase in fast TnT in ssTnT-deficient diaphragm muscle. C, in ssTnT-KD and ssTnT-KO diaphragm muscles, slow TnI levels were decreased to 4.19 ± 2.48% and 0.01 ± 0.01% of the wild-type level, respectively; fast TnI protein were increased to 104.14 ± 11.25% and 120.02 ± 7.53% of the wild-type control. The slow-to-fast ratio of TnI isoforms was 0.53 ± 0.03 in wild-type diaphragm muscle, and was dramatically decreased to 0.02 ± 0.01 in ssTnT-KD diaphragm and nearly non-determinable in ssTnT-KO diaphragm muscle. *P < 0.05 and ^**P < 0.01, as compared with wild type in two-tailed Student's t test. D, glycerol SDS-PAGE showed decreased MHC I in ssTnT-KD and ssTnT-KO diaphragm muscles. Densitometry quantification showed that MHC I accounts for 5.21 ± 0.49% of total MHC in wild-type adult mouse diaphragm muscle, which was diminished in both ssTnT-deficient adult mouse diaphragm muscles. Percentage of MHC IIa in ssTnT-KO diaphragm muscles was increased to 45.33 ± 2.00%, significantly higher than that of wild-type level (33.95 ± 1.83%). *P < 0.05 as compared with wild-type MHC IIa in two-tailed Student's t test; E, anti-MHC I mAb FA2 staining of diaphragm cryosections showed diminished type I slow fibre contents in both ssTnT-KD and ssTnT-KO mice.

Interestingly, MHC IIb protein levels were also significantly decreased in both ssTnT-KD and ssTnT-KO adult diaphragm muscles (Fig. 4A) but not in soleus (Fig. 2D) or tibialis anterior and extensor digitorum longus muscles (data not shown). These observations indicate that a potentially compensatory mechanism was activated to diminish the fastest MHC IIb fibres in response to the loss of type I slow fibres following ssTnT deficiency.

ssTnT deficiency alters the contractility of adult mouse soleus muscle

Contractility measurement of superfused soleus muscle demonstrated changes in twitch and tetanic contractions (Table 1). Twitch force normalized to cross-sectional area decreased in ssTnT-KO soleus muscle by ∼25%.

Table 1.

Parameters of twitch and tetanic contractions of wild-type and ssTnT-deficient adult mouse soleus muscles

		Wild type (n = 4)	ssTnT-KD (n = 3)	ssTnT-KO (n = 4)
Twitch	Force (mN mm−2)	60.49 ± 4.4	50.24 ± 7.55	44.56 ± 4.16*
	+dF/dt (mN s−1)	3.74 ± 0.41	3.91 ± 0.89	3.58 ± 0.43
	–dF/dt (mN s−1)	0.76 ± 0.09	1.04 ± 0.21*	1.24 ± 0.07*
	TPT (ms)	40.25 ± 1.49	32.67 ± 2.96*	29.75 ± 2.87*
	TP50 (ms)	12.5 ± 0.87	10.33 ± 0.67	10.5 ± 0.65
	TR75 (ms)	75 ± 13.71	54.33 ± 3.18	55.25 ± 5.5
Tetanic	Force (mN mm−2)	296.53 ± 25.51	276.45 ± 17.04	268.14 ± 13.82
	+dF/dt (mN s−1)	4.22 ± 0.28	4.23 ± 1.16	4.51 ± 0.61
	–dF/dt (mN s−1)	2.43 ± 0.34	4.56 ± 0.62*	6.12 ± 0.86*
	TR75 (ms)	161.5 ± 12.61	102.67 ± 7.36*	84.75 ± 7.65*

Force, developed force in twitch and tetanic contractions; +dF/dt, maximum contractile velocity; –dF/dt, maximum relaxation velocity; TPT, time to develop maximum force; TP50, time to develop 50% of maximum force; TR75, time to 75% relaxation. Data are presented as means ± SEM.

^*P < 0.05 compared with wild-type group in two-tailed Student's t test.

Relaxation velocity (–dF/dt) of twitch and tetanic contractions were all accelerated in both ssTnT-KD and ssTnT-KO soleus muscles. Although the contractile velocity (+dF/dt) did not show significant difference among the groups, the time for maximum force development during twitch contraction (TPT) was significantly shorter in soleus muscles of ssTnT-KD and ssTnT-KO mice than wild-type controls.

Force–frequency study showed that at low stimulation frequency (40 Hz), force production in ssTnT-deficient soleus muscles was significantly lower than that of wild-type controls (Fig. 5).

Figure 5.

Normalized to the maximum etanic force at optimal frequency of stimulation, adult ssTnT-KD and ssTnT-KO mouse soleus muscles produced significantly lower force at low frequency stimulation (40 Hz) than that of wild-type soleus muscle. At higher pulse frequencies (180, 200 and 220 Hz), twitch force of ssTnT-KO soleus muscles were significantly lower than that of wild-type control. *P < 0.05 compared with wild-type control in two-tailed Student's t test.

Although ssTnT-deficient soleus muscles did not show a significant loss of total weight (Fig. 3D), there is evidence for the effect of type I slow fibre atrophy on gross atrophy of the soleus muscle. When normalized to tibial length and excluding tendon length, the optimal length of ssTnT-deficient soleus muscles, at which maximum twitch force was achieved in ex vivo contractility experiments, was significantly shorter than that of wild-type controls (data not shown). Consistently, the in vivo length of ssTnT-deficient soleus muscles was also significantly shorter than that of wild-type controls. Therefore, despite the compensatory hypertrophy of type II fast fibres, the shorter muscle length indicated longitudinal atrophy of soleus muscle due to ssTnT deficiency and slow type I fibre atrophy. On the other hand, the insignificant decrease in total tissue mass of ssTnT-deficient soleus muscle may reflect an increase in non-muscle cell contents resulting from continuing remodelling.

ssTnT deficiency decreased fatigue tolerance of adult mouse soleus muscle with impaired recovery

Aerobic-oxidative slow fibres exhibit higher tolerance to fatigue than glycolytic fast fibres (Westerblad et al. 1991; Fitts, 1994; Allen et al. 2008). To investigate whether the ssTnT deficiency and consequential slow-to-fast switch of myofilament protein isoforms alters muscle fatigability, we examined tetanic force production of ssTnT-deficient soleus muscles using an intermittent fatigue protocol. Figure 6 shows the results from the fatigue treatment consisting of tetanic contractions with 30% duty cycles generated using 0.2 ms pulses at predetermined optimum frequency. At the end of the 5 min protocol, force production of ssTnT-KD and ssTnT-KO soleus muscles decreased to 25.57 ± 1.59% and 17.92 ± 1.11%, respectively, of their pre-fatigue levels, significantly lower than that of wild-type soleus muscle (32.53 ± 2.20% of the pre-fatigue level; Fig. 6A).

Figure 6.

A, 5 min intermittent tetanic fatigue protocol with 30% duty cycle demonstrated that the force development of ssTnT-KD and ssTnT-KO soleus muscles decreased to 25.57 ± 1.59% and 17.92 ± 1.11%, respectively, of the pre-fatigue level, significantly more severe than that of wild-type control (32.53 ± 2.20%). B, ssTnT-KD and ssTnT-KO soleus muscles showed slower and less recovery from fatigue. 1 min after the fatigue treatment, tetanic force development of wild-type soleus muscle recovered to 78.92 ± 3.42% of the pre-fatigue level whereas ssTnT-KD and ssTnT-KO soleus muscles recovered to only 67.18 ± 1.56% and 48.15 ± 3.01%, respectively, of the pre-fatigue levels. *P < 0.05 compared with wild type in two-tailed Student's t test.

ssTnT-KD and ssTnT-KO soleus muscles recovered from fatigue slower and less than wild-type controls (Fig. 6B). After 1 min of recovery from the fatigue treatment, force production of ssTnT-KD and ssTnT-KO soleus muscles returned to 67.18 ± 1.56% and 48.15 ± 3.01%, respectively, of their pre-fatigue levels, both values being significantly lower than the recovery of wild-type soleus muscle (78.92 ± 3.42% of the pre-fatigue level) (Fig. 6B). The results demonstrate that ssTnT deficiency significantly compromised fatigue tolerance of adult mouse soleus muscle and hindered post-fatigue recovery.

Effects of ssTnT deficiency on the expression of myofilament protein isoforms in soleus and diaphragm muscles during postnatal development

To assess the effect of ssTnT deficiency on muscle development, soleus and diaphragm muscles were examined on postnatal days 1, 3, 7, 14 and 21 and ssTnT-KD muscle showed significantly diminished ssTnT throughout postnatal development (Fig. 7). The slow isoforms of TnI and MHC I were at lower levels in early postnatal soleus and diaphragms muscles. Interestingly, deficiency of ssTnT led to a delayed down-regulation of cardiac TnT, especially embryonic cardiac TnT, in postnatal soleus and diaphragm muscles (Fig. 7). However, SDS-PAGE gel showed that there was no significant overall change in the protein profiles of developing ssTnT-KD soleus and diaphragm muscles (Fig. 7).

Figure 7.

Soleus and diaphragm muscles were collected from postnatal wild-type and ssTnT-KD mice at days 1, 3, 7, 14 and 21. Thin and thick myofilament proteins were examined by SDS-PAGE, Western blotting using mAb CT3, TnI-1, T12, and glycerol gel. ssTnT-KD soleus and diaphragm muscles showed significantly diminished ssTnT protein throughout postnatal development as compared with that of wild-type muscles. Slow isoform of TnI and MHC I were also at lower levels in postnatal soleus and diaphragms muscles. Deficiency of ssTnT was accompanied by a delayed down-regulation of cardiac TnT in both soleus and diaphragm muscles during postnatal development with a high level of the alternatively spliced embryonic isoform. eTnT, embryonic cardiac TnT; eMHC, embryonic cardiac MHC.

ssTnT deficiency is accompanied by centrally nucleated type I slow fibres in adult mouse soleus muscle

mAb FA2 staining of soleus muscle cross-sections detected a significant number of type I slow fibres characterized by centrally nucleated and small-sized fibres in both ssTnT-KD and ssTnT-KO mice, but not in wild-type, groups (Fig. 8A). The type I slow fibres with this marker of regeneration accounted for 5.25 ± 1.93% and 3.97 ± 0.89% of total slow type I myofibres in ssTnT-KD and ssTnT-KO soleus muscles, respectively (Fig. 8B).

Figure 8.

Immunohistochemical staining of cross sections of adult mouse soleus muscles using mAb FA2 to identify MHC I-positive fibres (A) demonstrated central nucleated, small-sized type I slow fibres (examples are indicated by the arrowheads) in ssTnT-KD and ssTnT-KO groups (5.25 ± 1.93% and 3.97 ± 0.89% of total slow type I fibres, respectively; B), but not in their type II fibres or type I and II fibres of the wild-type group, suggesting that ssTnT deficiency resulted in active regeneration of type I slow fibres.

Discussions

Gene targeting-generated C-terminal truncation of ssTnT reproduces the complete loss of ssTnT in ANM

The ANM nonsense mutation at codon Glu¹⁸⁰ of Tnnt1 gene produces a truncated ssTnT lacking the C-terminal tropomyosin binding site and interacting sites with TnI and TnC (Jin et al. 2003; Wang et al. 2005). We previously showed that the truncated ssTnT protein was not detectable in muscle biopsies of ANM patients, consistent with the recessive inheritance of the disease and the fact that the truncated ssTnT is unable to form troponin complex or incorporate into the myofilament, which is thus rapidly degraded in the muscle cells (Wang et al. 2005). Our present study showed that no intact or fragmented ssTnT was detectable in ssTnT-KO mice in which the coding exons 11–13 were deleted by gene targeting (Fig. 1) to produce a C-terminal truncated ssTnT similar to the ANM mutant. Therefore, this mouse model mimics the ANM mutation and provides an experimental system for the study of ANM pathogenesis and pathophysiology.

ssTnT deficiency results in severe atrophy and loss of type I slow muscle fibres and myopathic phenotypes

Significantly smaller-sized type I slow fibres were found in the soleus muscle of ssTnT-KO and ssTnT-KD mice (Fig. 3). Therefore, the decrease in ssTnT and possibly secondary decreases in other slow type myofilament proteins have a determining role in the growth and maintenance of type I slow fibres. While the number of type I fibres was not significantly decreased in ssTnT-deficient soleus muscle (Fig. 3), adult ssTnT-KD and ssTnT-KO mouse diaphragm muscles contained nearly no type I fibres (Fig. 4). Besides the fact that mouse diaphragm muscle normally has fewer slow fibres than soleus, the complete loss of type I fibres in ssTnT-KO and even in ssTnT-KD diaphragms may suggest an intriguing difference between these type I fibres, which needs to be further investigated.

The longitudinal atrophy of ssTnT-deficient soleus muscle is an interesting observation. The decreased number of sarcomeres per myofibril would certainly contribute to the decreased force production (Table 1). Mouse soleus muscle is a mixed fibre muscle that contains a significant portion of fast fibres. The mechanism by which ssTnT deficiency and subsequent reduction in the slow fibre content cause potentially compensatory growth in the diameter of fast fibres (Fig. 3), but decrease myofibril length, may represent an adaptive regulation and needs to be further investigated.

The overall phenotype of ssTnT deficiency includes myopathic features such as decreases in contractile force (Table 1) and tolerance to fatigue (Fig. 6). The data demonstrated that the deficiency in a minor isoform of TnT had major negative impacts on skeletal muscle function, clearly indicating the differentiated and critical function of the fibre type-specific TnT isoforms.

ssTnT deficiency in adult mouse soleus muscle resulted in a compensatory hypertrophic growth of fast fibres

The decrease or total loss of ssTnT in ssTnT-KD and ssTnT-KO mouse soleus muscles caused decreases in the slow isoforms of other thin and thick myofilament proteins and corresponding increases in the level of fast-type myofilament protein isoforms (Fig. 2). Myofilament protein Isoforms are expressed in fibre type-specific manner (Yu et al. 2007). Our previous studies have demonstrated that the expression of fibre type-specific troponin subunits was coordinated at the fibre level (Brotto et al. 2006). For example decreased ssTnT in soleus muscle during hindlimb unloading resulted in a slow-to-fast fibre type switch of other myofilament protein isoforms (Yu et al. 2007). Therefore, the increases in fast fibre-specific myofilament protein isoform contents correspond with the hypertrophic growth of type II fast fibres in adult soleus muscle that are experiencing slow fibre atrophy (Fig. 3).

Despite the compensatory hypertrophy of type II fast fibres, ssTnT-deficient soleus muscle showed decreased contractility (Table 1). Therefore, the loss of slow fibre function could not be fully compensated by the growth fast fibres. It has been shown that ssTnT confers higher Ca²⁺ sensitivity of the myofilament (Brotto et al. 2006), which may have a unique and critical contribution to contractile function in mixed fibre muscles.

Whereas predicted secondary decreases in slow TnI and MHC I expression are seen early on in postnatal developing soleus and diaphragm muscles of ssTnT-deficient mice, together with prolonged and potentially compensatory expression of cardiac TnT, the expression of thick and thin filament proteins in ssTnT-deficient soleus and diaphragm muscles showed no drastic change during postnatal development (Fig. 7). Therefore the functional changes demonstrated in the present study are likely to be due to the primary effects of ssTnT deficiency and slow fibre reduction.

ssTnT deficiency decreased the tolerance of soleus muscle to fatigue; relevance to the pathophysiological phenotype of ANM

ANM patients have severe muscle weakness and primarily die from failure of respiration muscles that are under continuing usage and require high tolerance to fatigue. Contractility studies demonstrated that ssTnT-KO and ssTnT-KD mouse soleus muscles had significantly decreased tolerance to fatigue in a dose-dependent manner (Fig. 6A). The decreased tolerance to fatigue was accompanied by slower and less extensive post-fatigue recovery (Fig. 6B). These results strongly support the causal effect of the loss of ssTnT on the muscle weakness-and respiratory muscle failure-based lethality of ANM (Johnston et al. 1997).

In comparison to its effect in soleus muscle, ssTnT-KD caused nearly complete loss of ssTnT in diaphragm muscle together with nearly complete loss of type I slow fibres (Fig. 4). This phenotype is consistent with our previous observation that ssTnT-KD resulted from partial promoter deletion, with the most significant ssTnT deficiency in the diaphragm muscle (Feng et al. 2009). Although normal mouse diaphragm contains fewer type I slow fibres than soleus muscle (Figs 3 and 4), the higher sensitivity of diaphragm muscle to decreases in ssTnT expression suggests a higher turnover rate of troponin, possibly due to its unique pattern of continuous contractions in sustaining respiration.

This correlation is worth further investigation for its significance in the respiratory failure of ANM patients. Although ANM carriers did not report clinical symptoms (Johnston et al. 1997), the decreased tolerance of ssTnT-KD slow muscle to fatigue suggests that the haploid carriers might encounter ssTnT deficiency under muscle wasting conditions, thus experiencing transient fatigue intolerance and other slow muscle dysfunction-related symptoms, especially in the diaphragm muscle. This hypothesis will be investigated in future studies.

ssTnT-KO mouse soleus muscle maintains a slow fibre cellular environment and exhibits a sign of regeneration

Slow TnI and MHC I are decreased but remain at significant levels in ssTnT-KO soleus muscle (Fig. 2). Immunohistochemistry images showed type I slow fibres in ssTnT-KO soleus muscle with atrophy but no decrease in number (Fig. 3). These data demonstrate that slow fibres are preserved in ssTnT-deficient soleus muscle of adult mice, although not in the diaphragm (Fig. 4). The continuing expression of highly fibre type-specific myofilament proteins such as slow skeletal muscle TnI and MHC I indicates a preserved cellular environment in the ssTnT-deficient slow fibres. This is a plausible finding, suggesting that a restoration of ssTnT in slow fibres may readily rescue muscle function in ANM patients.

Our data further showed that in soleus muscles of 3-to 5-month-old ssTnT-KD and especially ssTnT-KO mice, there were significant numbers of centrally nucleated type I slow fibres (Fig. 8). No centrally nucleated type II fast fibres were observed. This phenotype suggests that type I slow fibres of ssTnT-KD and ssTnT-KO mouse soleus muscles remain capable of regeneration, despite the significant adaptive up-regulation of fast fibre type myofilament proteins and compensatory hypertrophy of type II fast fibres. This observation further supports the hypothesis that the function of slow fibres can be restored in ANM patients if a therapeutic supplement of ssTnT is achieved.

Key points

• Loss of slow skeletal muscle troponin T (ssTnT) due to a nonsense mutation in codon Glu¹⁸⁰ causes a lethal form of recessively inherited nemaline myopathy.

• We studied the phenotypes of partial and total loss of ssTnT in the diaphragm and soleus muscles of Tnnt1 gene targeted mice.

• ssTnT deficiency resulted in significant decreases in other slow fibre-specific myofilament proteins whereas fast fibre-specific myofilament proteins were increased correspondingly.

• ssTnT deficiency produced significantly smaller type I slow fibres and compensatory growth of type II fast fibres with significantly reduced contractile force and tolerance to fatigue.

• ssTnT-deficient soleus muscle exhibited a significant numbers of small-sized central nuclei type I fibres, indicating regeneration.

• The data support the critical role of ssTnT in skeletal muscle function and the causal effect of its loss on the pathogenesis of nemaline myopathy.

Glossary

ANM

Amish nemaline myopathy

ES

embryonic stem (cell)

fsTnT

fast skeletal muscle TnT

mAb

monoclonal antibody

MHC

myosin heavy chain

ssTnT

slow skeletal muscle troponin T

ssTnT-KD

knockdown ssTnT

ssTnT-KO

knockout ssTnT

TnC

troponin C

TnI

troponin I

TBS

Tris-buffered saline

Additional information

Competing interests

None declared.

Author contributions

B.W. performed the phenotype studies, analysis and interpretation of data, and drafting and revising of the article. Y.L. contributed to the construction of Tnnt1 targeted mouse line. J.-P.J. contributed to the conception and design of the study, analysis and interpretation of data, drafting and revising of the article, and final approval of the version to be published. The construction of Tnnt1 targeted mice was done at Northwestern University, and the phenotype studies were done at Wayne State University School of Medicine. All authors read and approved the final version of the manuscript. Dr Yingru Lu participated in this study when our research group was at the ENH Research Institute, Northwestern University Feinberg School of Medicine.

Funding

This work was supported by a grant from the National Institutes of Arthritis, Musculoskeletal and Skin Diseases at the National Institutes of Health (AR-048816) to J.-P.J.