The transforming growth factor-β (TGF-β) superfamily is comprised of over 30 members that include several structurally dimeric cytokines such as TGF-β, activins, and bone morphogenic proteins (BMPs)/growth and differentiation factors (GDFs) which are evolutionarily highly conserved (Shi and Massagué, 2003). TGF-β transmits its signals via dual serine/threonine kinase receptors and transcription factors called Smads. Several direct and indirect links between TGF-β and mitochondrial energy metabolism and energy sensing pathways have been reported (Casalena et al., 2012). TGF-β overexpression deregulates the cellular redox state by stimulating reactive oxygen species (ROS) production through Nox4 activation and compromises the antioxidant system that protect against mitochondrial derived ROS by inhibiting the expression of MnSOD (Michaeloudes et al., 2011). TGF-β also modulates energy metabolism by controlling mitochondrial metabolism directly by reduction of complex IV and mitochondrial respiration that results in increased ROS production (Yoon et al., 2005). Extensive interaction between TGF-β and key energy sensors adenosine monophosphate-activated protein kinase (AMPK) and sirtuin family members has been demonstrated (Mishra et al., 2008; Verdin et al., 2010). TGF-β levels correlate with obesity in mice (Samad et al., 1999) and humans (Alessi et al., 2000; Fain et al., 2005). Furthermore, TGF-β/Smad3 pathway has also been demonstrated in the regulation of insulin gene transcription and β-cell function (Lin et al., 2009). Most recently, a comprehensive study highlighted the critical involvement of TGF-β/Smad3 signaling during the process of obesity and demonstrated the beneficial effects of systemic blockade of this signaling from obesity, diabetes, and hepatic steatosis (Yadav et al., 2011). Using hyperinsulinemic-euglycemic clamp experiments, this group reported that Smad3 loss protects experimental mice from diet-induced obesity and insulin resistance. This effect was associated with the phenotypic transition of white to BATs, induction of mitochondrial biogenesis, and enhanced expression of BAT-related gene signatures. Furthermore, elevated plasma TGF-β1 levels were shown to correlate positively with BMI and increased adiposity in overweight and obese subjects compared to subjects with normal BMI. Inhibition of activin receptor IIB (ActRIIB) that integrates actions of TGF-β related ligands via dimerization with Alk4/5 and signals intracellularly via Smad2/3 promotes differentiation of primary brown adipocytes in vitro and increases the amount of brown but not white fat in mice (Fournier et al., 2012). This blockade of ActRIIB enhanced mitochondrial function and uncoupled respiration resulted in increased energy expenditure under ambient or cold temperatures but not at thermoneutrality. Interestingly, molecular signatures induced by inhibition of ActRIIB overlap with PGC-1α overexpression in vivo. PGC-1α was originally described as a co-activator of PPAR-γ that modulated expression of key brown fat specific UCP-1 and thermogenesis in brown fat (Boström et al., 2012).

Myostatin (Mst), also termed as growth/differentiation factor-8 (GDF-8) is a member of the TGF-β superfamily that regulates skeletal muscle mass (McPherron et al., 1997). Mst-deficient mice are significantly larger than wild type animals and show widespread increase in skeletal muscle mass. Initial phenotypic characterization of Mst-deficient mice revealed suppression of body fat accumulation, an unexpected finding that suggested a possible role of Mst in the control of energy balance beyond its effect on skeletal muscle (McPherron and Lee, 2002). Using muscle (gastrocnemius and levator-ani) and adipose (epididymal and subcutaneous) tissues isolated from wild type and Mst-deficient male mice, Braga et al. demonstrated significant induction of BAT-related markers in both tissue types obtained from Mst-deficient mice. Protein expression levels of adipocyte specific marker adiponectin and energy sensing AMP-activated protein kinase (AMPK) was significantly upregulated in differentiating MEFs from Mst-deficient cells compared to the wild type (Braga et al., 2013). In order to clarify how Mst deficiency affects the body fat mass and energy balance, Choi et al. compared rates of oxygen consumption, body composition, and food intake in wild type, heterozygous and Mst-deficient mice and reported significantly increased energy expenditure in Mst-deficient mice (Choi et al., 2011). These studies suggested that increased fatty acid oxidation and total energy expenditure could be key reasons for reduced adiposity in Mst-deficient mice. Significantly increased muscle mass in Mst deficient mice also contributes to overall metabolism through inhibition of activin signaling, activation of AMPK/PGC-1α/Fndc5 (irisin) pathway, or by increasing Akt signaling (Guo et al., 2009; Shan et al., 2013). Subsequent studies demonstrated remarkable changes in the epididymal WAT of Mst-deficient mice compared to the wild type mice. Global gene expression analysis of these mice revealed elevated expression of genes involved in fatty acid oxidation, lipid transport, mitochondrial biogenesis, and BAT-specific transcription factors including Ucp1, Cidea, and Dio2 (Zhang et al., 2012). Histological analysis of WAT explants derived from Mst-deficient mice revealed the presence of BAT-like cells characterized by their small diameter, inclusion of multilocular lipid droplets and positive staining for UCP1. Also, these Mst-deficient mice had higher body temperature (~ 1°C) compared to the wild type mice. In the same study, the authors reported an increased abundance and activity of energy sensing enzyme AMP-activated protein kinase (AMPK) in adipose, liver, and skeletal muscle tissues of Mst-deficient mice. Subsequently several lines of evidence obtained from in vitro studies confirmed the inhibitory role of Mst during brown fat differentiation and regulation of energy metabolism. Brown preadipocyte cells obtained from mouse interscapular BAT, when treated with recombinant Mst protein significantly down-regulated brown adipocyte differentiation via regulation of Smad3-mediated β-catenin stabilization (Kim et al., 2012). Using primary cultures of mouse embryonic fibroblasts (MEFs) obtained from wild type and Mst-deficient embryos differentiating under specialized adipogenic conditions that allows brown adipose-like differentiation, it was demonstrated that the Mst-deficient group had significantly increased levels of select genes and proteins that improve lipid metabolism and energy expenditure (Braga et al., 2013). Treatment of differentiating Mst-deficient MEFs with recombinant Mst protein significantly inhibited the gene expression levels of key BAT markers (Ucp1, and Prdm16), general adipocyte markers (Cebp-α, and Pparγ) also expressed in BAT, metabolic regulators (Pgc-1α/β) as well as BAT inducer (Bmp7). Activation of AMPK-PGC-1α pathway in WATs of Mst-deficient mice was further confirmed by other investigators (Shan et al., 2013). Immunohistochemical staining of WAT from Mst-deficient mice showed numerous small-sized brown adipocyte-like cells filled with multilocular lipid droplets that stained with UCP1, typical characteristics of beige cells. Consistent with the browning phenotype, WAT from Mst-deficient mice also expressed significantly higher levels of beige cell specific Tmem26 and Cd137. These authors further demonstrated that loss of Mst or inhibition of Mst signaling in preadipocytes was unable to account for the browning phenotype observed in WAT of Mst-deficient mice and suggested that muscle-derived newly discovered myokine irisin (Fndc5) was responsible for the browning in a non-autonomous manner (Shan et al., 2013). Irisin was initially discovered as a PGC-1α-dependent myokine that is responsible for brown-fat-like characteristics both in vitro and in vivo and protects mice from diet-induced obesity (Boström et al., 2012). The authors further demonstrated that the effect of Mst inhibition on Pgc-1α and Fndc5 gene expression are abolished by compound C (AMPK inhibitor) suggesting that AMPK activation is required to mediate the effect of Mst on Pgc-1α and Fndc5. Activation of AMPK signaling pathway in Mst-deficient mice has also been reported previously that resulted in increased fatty acid oxidation, augmented SIRT1 signaling, and increased sensitivity to insulin (Zhang et al., 2011). The precise mechanism by which Mst inhibition may lead to the upregulation of AMPK is not clear. It is possible that Mst inhibition may regulate mTOR and FoxO1 pathways involved in muscle protein synthesis and degradation, respectively to increase the protein level of AMPK (McFarlane et al., 2006; Rodriguez et al., 2011). Most recently, Mst has been reported to inhibit AMPK in cardiomyocytes through the activation of TAK1, which is the main signal transducer of Mst besides Smads (Biesemann et al., 2014). In conclusion, these studies provide rational justification for inhibiting Mst through a novel class of Mst antagonists for treatment of obesity and related metabolic syndromes by enhancing brown adipocyte characteristics and increased energy expenditure not only in white adipocytes but also in skeletal muscle. A brief summary of TGF-β/Mst action on browning of white adipocytes and regulation of energy metabolism is shown in Figure 1.

Follistatin (Fst) is a secreted glycoprotein that has been recognized for many years as a high affinity binding and neutralizing protein for several members of the TGF-β superfamily including activins and Mst (Welt et al., 2002; Sidis et al., 2006). A series of genetic studies have convincingly demonstrated an essential role of Fst in skeletal muscle development (Matzuk et al., 1995; Lee et al., 2010). However, the role of Fst in other metabolic tissues is less well-known. Fst-deficient mice created by a targeted deletion of the Fst gene survived until birth but died within hours of delivery (Matzuk et al., 1995). Several defects, including reduced size of intercostal and diaphragm muscle, were observed in these mice, and these severe musculoskeletal defects were suggested as the possible cause of neonatal death for these mice. Since skeletal muscle and brown fat share Myf5⁺ as common precursor, and there is an absolute requirement of brown fat to maintain body temperature during first few hours of neonatal life, Braga et al. tested possible novel role of Fst in regulating overall thermogenic program (Braga et al., 2014). Initial gene expression analysis of a mouse tissue panel prepared from C57BL6J mice showed very high levels of Fst expression in BAT and skeletal muscle; intermediate expression in inguinal WAT and liver, and significant but lower levels in other metabolic tissues. Expression levels of Fst along with other known brown adipocyte markers were significantly induced from nearly undetectable levels after differentiation of mouse brown preadipocyte cells. Furthermore, treatment of the differentiating cells with the recombinant Fst protein further upregulated the expression of several thermogenic genes including Ucp1 and Prdm16, suggesting a possible role of Fst in the brown fat differentiation program. As direct in vivo analysis of lipid and energy metabolism as well as brown fat differentiation program in Fst-deficient mice was not possible because of perinatal death of Fst-deficient pups, Braga et al. isolated MEFs from wild type and Fst-deficient embryos after heterozygous breeding and allowed these cells to undergo specialized adipogenic differentiation conditions to assess possible defects in lipid and energy metabolism in Fst-deficient MEF cultures compared to the wild type (Braga et al., 2013). Surprisingly, a significant lack of stimulation of key markers for lipid and energy metabolism and mitochondrial biogenesis including Adn, Pgc1a, Bmp7, Cidea, Cyt C, Thrsp, Hp, Acsl1, Pparg, and Cd36 amongst others in Fst-deficient MEFs was observed in comparison to the wild type MEFs. Treatment of these differentiating Fst-deficient MEF cultures with recombinant Fst rescued the expression levels of these thermogenic genes and proteins. Interestingly, using a similar MEF-based approach Braga et al. also demonstrated reciprocal effects on Bmp7, Pgc1a, and Cidea gene expression in Mst-deficiency cultures compared to the wild type (Braga et al., 2013), suggesting possible cross-talk between Mst and Fst pathways during brown fat differentiation. Since PGC-1α is known to drive brown-fat-like development of white fat and thermogenesis (Boström et al., 2012), its reciprocal regulation by Fst and Mst is quite logical. BMP7, a key regulator of brown adipogenesis and energy expenditure (Tseng et al., 2008) is known to be induced by Fst (Amthor et al., 2002), but antagonized by Mst (Rebbapragada et al., 2003). Fst-deficient MEFs showed decreased oxygen consumption rate (OCR) compared to the wild type; but recombinant Fst rescued this respiration impairment indicating the crucial role of Fst during maintenance of cellular respiration. These findings therefore, suggest that Fst is a novel modulator of brown fat differentiation and energy metabolism, and its loss-of-function results in severe metabolic defects that may be essential for survival. Follistatin like-3 (FSTL3), another major Mst binding protein is reported to regulate fat mass and glucose homeostasis, it is not known whether it also regulates BAT mass and activity (Brown et al., 2011). FSTL3 knockout mice developed distinct phenotypes including increased pancreatic islet number and size, decreased visceral fat mass, and increased insulin sensitivity, suggesting that the role of Fsts may be more complicated (Mukherjee et al., 2007). The molecular mechanisms responsible for Fst-induced brown adipocyte characteristics have not been investigated; however, it is possible that Fst may antagonize components of the Mst/TGF-β signaling pathways during the process. It will be interesting to explore whether Fst antagonizes Mst/Smad3 signaling during adipocyte differentiation and browning of white adipocytes. Myf5-expressing precursor cells are found in skeletal muscle as well as in BAT. It is possible that Fst may induce classical BAT mass and its activity by directly targeting Myf5-expressing cells. On the other hand, Fst may also target pp38 MAPK/Cox-2 pathways in Myf5-negative cells to promote browning in WAT. Most recently, irisin, a PGC-1α and exercise inducible myokine has been reported to stimulate browning of WAT through p38 MAP Kinase and ERK signaling (Zhang et al., 2014). Similar to irisin, Fst levels also increase after exercise (Hansen et al., 2011). Furthermore, recombinant Fst or anti-Mst antibody treatment of mouse muscle cells increases irisin encoded Fndc5 gene (Shan et al., 2013). It is, therefore, possible that Fst may engage irisin/Mst dependent signaling to promote browning. Detailed in vivo studies using Fst-transgenic mice expressing Fst under the control of both white and brown adipose-specific promoters on their metabolic parameters including whole body energy expenditure, glucose tolerance, and insulin sensitivity would provide valuable information regarding potential adipocyte-specific effects of Fst. Delineating the specific tissue targets of Fst responsible for its regulation of overall energy homeostasis would provide a novel therapeutic approach for the treatment of obesity and related metabolic syndromes. Initial gene therapy preclinical studies performed on non-human primates demonstrate that sustained Fst expression caused no aberrations in the structures or functions of a variety of organs, suggesting that Fst use in clinical settings may be safe (Kota et al., 2009). A brief summary of Fst action in adipocyte differentiation and energy metabolism is shown in Figure 2.

Patent application (RS) entitled “Composition and methods for treating or preventing metabolic syndrome disorders” is in process. Authors declare no other potential conflict of interest.