Increased Smad signaling and reduced MRF expression in skeletal muscle from obese subjects


  • Disclosure: The authors declare no conflict of interest.

  • Funding agencies: This work was supported by the Deakin University Molecular and Medical Research Strategic Research Centre.



The molecular mechanisms underpinning the loss of skeletal muscle mass and strength associated with insulin resistance remain to be extensively investigated. There is mounting recognition that certain ligands of the transforming growth factor (TGF)-β family are upregulated in insulin resistant states, including obesity. This study analyses the expression of potent ligands of this family, TGF-β1 and myostatin (MSTN) and downstream components of the canonical TGF-β family signaling pathway (Smads) in skeletal muscle from lean and insulin resistant obese subjects.

Design and Methods:

Biopsies taken from the rectus abdominis muscle of lean (n = 13) and obese subjects (n = 20) were analyzed for the expression of TGF-β1 and MSTN as well as TGF-β signaling components, Smad2, 3, and 4, and transcription of the muscle regulatory factors (MRFs), MyoD and myogenin.


Increases in Smad2 and Smad3 phosphorylation, Smad4 and total Smad3 were observed to be coincident with altered transcription of MyoD and myogenin. TGF-β1 and MSTN protein levels were not significantly altered.


Thus, increased Smad signaling is likely to account for, at least, a proportion of obesity and insulin resistance-related muscle atrophy through reduced MRF, particularly MyoD, transcription. The major regulatory ligand may not be MSTN and further members of the TGF-β1 superfamily should be considered.


In addition to the metabolic alterations evident in the skeletal muscle of individuals with insulin resistance, there are further and less well understood changes that result in an accelerated rate of muscle atrophy (1) and reduced muscle strength (2). Myostatin (MSTN), a member of the transforming growth factor-β (TGF-β) family, is powerful negative regulator of skeletal muscle across the lifespan. The inactivation of MSTN results in a hypermuscular phenotype (3), whilst over-expression induces atrophy and is implicated in wasting in a variety of disease states, including HIV (4). MSTN is upregulated in myotubes isolated from insulin resistant obese women (5) and the skeletal muscle of obese mice (6). Beyond the potential for the increased expression of MSTN to influence muscle mass regulation, MSTN deletion has been shown to improve insulin sensitivity and suppress adipogenesis in obese and diabetic rodents (7, 8).

TGF-β family ligands, including MSTN, bind to trans-membrane receptors phosphorylating and activating Smad2 and Smad3, enabling complex formation with Smad4. This complex translocates to the nucleus to regulate transcription of target genes including MyoD and myogenin of the muscle regulatory family (MRF) transcription factors (9). Whilst this canonical Smad pathway is a key regulator of TGF-β family signaling, studies are lacking on this potentially important regulation within skeletal muscle of individuals with insulin resistance. Therefore, the aim of this study was to examine the expression of TGF-β1, MSTN, related Smad signaling proteins, and transcription of target genes, MyoD and myogenin, in skeletal muscle obtained from lean and insulin resistant obese subjects.

Methods and Procedures

Written informed consent was obtained from 13 lean (BMI <25 kg/m2) and 20 morbidly obese (BMI >40 kg/m2) age-matched (lean 43.8 ± 2.3, obese 37.4 ± 2.1 years old) individuals prior to undergoing routine abdominal surgery. All protocols were approved by the Human Ethics Research committees of Deakin University, the Avenue Hospital, and Cabrini Hospital, Melbourne. No subjects were taking any medications known to affect insulin sensitivity. All subjects underwent 12–18 h fasting, with general anesthesia induced and maintained using short-acting propofol and fentanyl and rocuronium mixtures, respectively. Biopsies were harvested from the rectus abdominis muscle at the commencement of surgery.

RNA extraction, cDNA synthesis, and RT-PCR

Total cellular RNA was extracted using the Totally RNA Kit (Ambion Inc., Austin, TX). RNA quality and concentration were determined using an Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA). Equal amounts of total RNA were synthesized into cDNA using an AMV RT kit (Promega, Madison, WI). Samples were analyzed by real-time PCR using oligonucleotide primers for human TGF-β1 (forward 5′-CGCGTGCTAATGGTGGCAAA-3′; reverse 5′-ATGCTGTGTGTACTCTGCTTGAACT- 3′), MSTN (forward 5′-CCAGGAGAAGATGGGCTGAA-3′; reverse 5′-CAAGACCAAAATCCCTTCTGGAT-3′), MyoD (forward 5′-CCGCCTGAGCAAACTAAATGA-3′; reverse 5′-GCAACCGCTGGTTTGGAT-3′), and myogenin (forward 5′-GGTGCCCAGCGAATGC-3′ reverse 5′-TGATGCTGTCCACGATGGA-3′) on a 7500 Real Time PCR System (Applied Biosystems, Foster City, CA) using Power SyBR Green (Applied Biosystems, Warrington, UK). Fluorescent emission data were analyzed for the critical threshold (CT) values, with the expression of the gene of interest normalized to cyclophilin and expressed as 2−ΔCT. Cyclophilin expression was unaltered between lean and obese subjects.


Smad2, phospho-Smad2 (Ser465/467), Smad3, phospho-Smad3 (Ser423/425), Smad4, phospho-Akt (Ser473), and Akt antibodies were purchased from Cell Signaling Technology (Danvers, MA). TGF-β1 and MSTN antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA) and Chemicon (Billerica, MA), respectively. Anti-actin was purchased from Sigma (Sydney, Australia). Secondary antibodies, goat anti-mouse IgG horseradish peroxidase (HRP) (Pierce Biotechnology, Rockford, IL), and goat anti-rabbit IgG HRP (Chemicon) were used. Tissue sample were homogenized in ice-cold extraction buffer (20 mM Tris-HCl, 5 mM EDTA, 10 mM Na-pyrophosphate, 100 mM NaF, 2 mM Na3VO4, 1% Igepal, 10 μg/ml Aprotinin, 10 ug/ml Leupeptin, 3 mM Benzamindine, and 1 mM phenylmethylsulfonyl fluoride). Total protein concentrations were determined by the Bradford method. Skeletal muscle lysates were separated by sodiumdodecyl sulfate (SDS)-polyacrylamide gel electrophoresis and transferred onto polyvinylidene fluoride (PVDF) or nitrocellulose membranes. Membranes were incubated in blocking buffer (50 mmol/l Tris-HCl, pH 7.5, 0.75 mol/l NaCl, and 0.1% Tween) containing 5% bovine serum albumin (BSA) or cold fish gelatin (CFG) and incubated overnight at 4°C with primary antibody, followed by a 60-min incubation period with secondary antibody. Immunoreactive bands were detected using enhanced chemiluminescence (Western Lighting Chemiluminescent Reagent Plus, Perkin Elmer, Boston, MA). Densitometric analysis of the protein bands was performed using Kodak ID 3.5 Imaging Software (Kodak Digital Services, Rochester, NY) and expressed in arbitrary units.

Statistical analysis

Data calculations and statistical analysis were performed using SPSS 14.0. All data are presented as mean ± SEM. Comparisons between the lean and obese groups were assessed using unpaired t-tests. Differences were considered significant at P < 0.05.

Results and Discussion

Obesity is associated with endocrine abnormalities that are either precipitated by or precede the onset of peripheral insulin resistance (10). Despite comparable fasting plasma glucose (lean 5.1 ± 0.2, obese 5.2 ± 0.1 mmol/L), fasting plasma insulin was significantly elevated in obese subjects (lean 4.3 ± 0.8, obese 16.4 ± 1.3 mU/l, P < 0.01), resulting in HOMAR-IR scores indicative of insulin resistance (lean 0.9 ± 0.2, obese 3.8 ± 0.3, P < 0.001) (11). Western blot analysis of Akt phosphorylation was performed due to its involvement in insulin-stimulated glucose uptake. Data analysis confirmed a significant decrease (58%, P < 0.001) in basal phospho-Akt in the rectus abdominis muscle from obese subjects (Figure 1a), consistent with the HOMAR-IR data. Predominately composed of type II muscle fibers, rectus abdominis is a muscle group involved in stability as well as bending and rotational movement. Rectus abdominis may itself be a minor component of whole body insulin resistance, however previous reports demonstrate both impaired insulin-mediated signaling and atrophy, comparable to other peripheral tissue in both gestational diabetes and critical illness (12, 13).

Figure 1.

Effect of obesity on TGF-β family-related signaling in human skeletal muscle. Protein content of (a) phospho-Akt relative to total Akt, (b) phospho-Smad2 relative to total Smad2, (c) phospho-Smad3, (d) total Smad3, (e) Smad4, (f) TGF-β1, and (g) MSTN normalized to actin in rectus adbominis muscle biopsies from lean and obese subjects. Representative Western blots are displayed in (j). Real time PCR was performed to assess (h) MyoD and (i) myogenin gene expression. Data are presented as mean ± SEM for 13 lean and 20 obese subjects. *P < 0.05, **P < 0.01 compared to lean values.

Smad2 phosphorylation was increased in the obese subjects compared to the lean individuals (Figure 1b). Phosphorylated and total Smad3 levels increased by 71% (P < 0.05) and 222% (P < 0.01) in obese muscle, respectively (Figure 1c and d), whilst Smad4 protein abundance doubled (Figure 1e; P < 0.05). Our reasoning for normalizing phosphorylated- and total Smad3 to actin was to present the data in a manner that clearly demonstrated increased expression of both forms, observations that would have been masked if presented simply as phosphorylated Smad3/total Smad3.

Translocation of the Smad complex to the nucleus leads to a downregulation in the transcription of MRF members, MyoD and myogenin (9). The MRFs are responsible for the transcriptional regulation of skeletal muscle specific genes required for subsequent differentiation (9). Smad3 is the effector by which TGF-β1 (14) and MSTN (9) repress skeletal muscle differentiation by inhibiting the formation of multinucleated myotubes. Reduced transcription of MyoD (57%, P < 0.01) and myogenin (46%, P < 0.05) in the muscle of obese subjects (Figure 1h and i) may contribute to, at least a proportion of functional skeletal muscle mass seen in such conditions.

Whilst increased TGF-β1 release into the circulation of insulin resistant patients has been described (15), research has primarily focused on increased MSTN abundance (5) and its corresponding metabolic effects when reduced, such as prevention of diet-induced obesity and improved insulin-sensitivity (7, 8). Despite increased activation of the canonical Smad pathway analysis of TGF-β1 and the active protein subunit (26 kDa) of MSTN showed no increase (Figure 1f and g), indicating that neither TGF-β1 or MSTN are alone responsible for such signaling.

In the present study, analysis was made of the expression of TGF-β1, MSTN, and their common key downstream effectors in rectus abdominis biopsy samples harvested from lean and obese subjects. Whilst it is unlikely that TGF-β1 or MSTN alone are implicated in the atrophic effects on myogenesis originally observed by Tajiri et al. (2010), we cannot discount that firstly the subcellular localization and association with prodomains impacts TGF-β1 and MSTN ligand activity. Secondly, it is possible that alternative TGF-β family ligands are responsible for the increase in Smad signaling with subsequent MyoD and myogenin downregulation. Although this constitutes an important limitation of this study, it may be speculated that reduced MyoD and myogenin transcription may lead to significant reductions in functioning skeletal muscle mass through heightened Smad signaling mediated by an alternative ligand.

As the major site of insulin-stimulated glucose uptake, the ramifications for a reduction in skeletal muscle regeneration are likely to compromise functional and metabolic health in insulin resistant individuals (1). Heightened Smad signaling, an important downstream signaling effector pathway of the TGF-β family ligands, previously demonstrate to increase muscle catabolism and impair anabolism was demonstrated within the rectus abdominis muscle from obese individuals. Thus this data suggests this pathway as an additional catabolic pathway that may influence the regulation of skeletal muscle mass in states of insulin resistance. It is unclear from this small cross-sectional study whether the alterations in Smad signaling are present due to the insulin resistance or arise as a consequence to lifestyle changes, including reduced physical activity. Further analysis is also required to determine whether single or multiple ligands are responsible for such signaling in obesity-induced-insulin resistance, its influence on the regenerative ability, hence mass of functioning skeletal, are likely to influence whole body insulin resistance.


The authors wish to thank the participants who were willing to undergo this additional procedure during surgery. We also acknowledge the helpful advice and technical assistance of Dr Nicole Stupka, Institute for Technology Research and Innovation, Biodeakin, and Dr Marissa Trenerry, School of Exercise and Nutrition Sciences, Deakin University.