Discussion
Muscle wasting is one of the hallmarks of cancer cachexia (Penna, Ballaro, Beltra, De Lucia & Costelli, 2018). STAT3 signaling plays a crucial role in the regulation of skeletal muscle generation, muscle mass, and repair processes. It is generally believed that transient STAT3 activation in muscles is beneficial to muscle regeneration and hypertrophy, but prolonged STAT3 activation in muscles has been shown to be responsible for muscle wasting (Zimmers, Fishel & Bonetto, 2016). Increased STAT3 activation has been observed in skeletal muscles in diverse models of cancer cachexia, including models based on IL-6 administration, C26 colon adenocarcinoma, LLC lung carcinoma, B16 melanoma, and ApcMin/+-induced intestinal cancer (Baltgalvis, Berger, Pena, Davis, Muga & Carson, 2008; Bonetto et al., 2012b; Pretto et al., 2015; Puppa, Gao, Narsale & Carson, 2014; Silvaet al., 2015). However, the clinical association, the factors triggering abnormal STAT3 activation in cachectic muscles as well as the molecular mechanism of muscle atrophy induced by consistent STAT3 activation still need to be delineated. Here, for the first time, our data confirm the clinical association of abnormal STAT3 activation correlated with the development of cachectic muscle wasting by measuring the biopsy specimens from liver cancer patients.
Most STAT3 molecules exist in the cytoplasm in an unphosphorylated form, and their transcriptional regulatory effects rely on the phosphorylation of the tyrosine 705 residue. The upstream activators of STAT3 include IL-6 family cytokines, growth factors such as EGF, oncogenes such as the Src family kinases, and other biological factors (Yu et al., 1995; Zhong, Wen & Darnell, 1994; Song et al., 2018). Although the excessive release of IL-6 family cytokines, including IL-6, LIF, OSM, and CNTF, has been linked to the pathological development of cancer cachexia (Fearon, 2013). Multiple studies on STAT3 signaling have demonstrated that the stimulation of non-malignant cells with IL-6 family cytokines can trigger only transient STAT3 activation, which is quickly terminated by negative regulators such as PIAS, suppressor of cytokine signaling (SOCS-1, 2, and 3), and cellular phosphatases (SHP-1/2, DUSP22, PRPRD, PTPRT, and PTPN1/2) (Johnson, O’Keefe & Grandis, 2018; Ram & Waxman, 1997; Sengupta, Schmitt & Ivashkiv, 1996). Therefore, it is reasonable to speculate that apart from IL-6 family cytokines, other regulatory factors may also be involved in maintaining the persistent STAT3 activation in cachectic muscles in cancer.
HSP90 is an ATP-dependent molecular chaperone; it has been reported to be one of the most abundant molecules within cells, accounting for nearly 2% of all cellular proteins (Joshi, Wang, Araujo, Sharma, Brodsky & Chiosis, 2018). A wide variety of stressful stimuli, including heat shock, UV radiation, and micro-organismal infections, could promote intracellular HSP90 synthesis. During cancer progression, the expression level of HSP90 increases by 2- to 10-fold in tumor cells (Joshi, Wang, Araujo, Sharma, Brodsky & Chiosis, 2018). Elevated HSP90 expression has been found in various types of human tumors and is often correlated with increased tumor cell proliferation, inadequate responses to chemotherapy, and poor survival (Kazarian et al., 2017; Ono et al., 2018; Whitesell & Lindquist, 2005; Yun, Kim, Lim & Lee, 2019). Over 200 oncogenic proteins have been identified as clients of HSP90 (Trepel, Mollapour, Giaccone & Neckers, 2010). We and others have demonstrated a direct interaction between STAT3 and HSP90 in multiple cell types, including cancer cells, epithelial cells, fibroblasts, and endothelial cells (Song et al., 2017; Shah, Patel, Fried & Sehgal, 2002). Here, in this study, we found an increased interaction of HSP90 and STAT3 in skeletal muscle from cachexia patients and experimental mice model. Administration of HSP90 inhibitors, which blocked the aberrant STAT3 activation, could successfully alleviate cachexia-related muscle wasting phenotype.
Interestingly, we revealed that the direct interaction of HSP90 with STAT3 could enhance JAK2-mediated STAT3 phosphorylation, thus induce muscle atrophy. The increased HSP90-STAT3 interaction likely causes the prolonged STAT3 activation and muscle atrophy since both HSP90 inhibitor or STAT3 siRNA could successfully alleviate the pathological muscle wasting phenotype both in vivo and in vitro ; suggesting that persistent activation of STAT3 rather than other signaling pathway is the major cause of the induction of muscle wasting in cancer cachexia. A previous study by Zhang et al. reported that tumor released HSP90 contained exosome could trigger muscle wasting by activation of TLR4, to validate whether TLR4 mediates the protective effect of HSP90 inhibitors in cachectic muscle wasting, we repeated the experiment by using the LPS induced the C2C12 myotube atrophy model, however, to our surprise, we did not observe the obvious protective effects; this result suggested that the protective effect of HSP90 inhibitors in cachectic muscle wasting is not fully TLR4 dependent. TLR4 signaling might be involved in the initial stage of STAT3 activation, but the tumor induced a prolonged STAT3 activation, which is necessary for triggering the muscle wasting as demonstrated in our mechanism study by over-expression of the consistent activated form of STAT3 (STAT3-C) in C2C12 myotube, required the participation of HSP90 as a molecular chaperone by the interaction of STAT3. Therefore, it is reasonable to speculate that the enhanced HSP90-STAT3 interaction seems to be the critical event for the induction of persistent STAT3 activation and cachectic muscle atrophy.
An additional goal of this study was to delineate the mechanism by which the prolonged STAT3 activation induces muscle wasting. The pathogenesis of muscle wasting in cancer cachexia has been demonstrated to be the result of an imbalance between the anabolic and catabolic pathways. The STAT3 pathway has been found to be a crucial pathway involved in this process. During the catabolic response in skeletal muscles, prolonged STAT3 activation could disturb the sensitivity of the insulin/IGF-1/AKT signaling pathway by enhancing the expression of inhibitory SOCS proteins (Mashili, Chibalin, Krook & Zierath, 2013; Ueki, Kondo & Kahn, 2004). During the anabolic response, current models indicate that STAT3 could enhance the expression of myostatin via activation of CAAT/enhancer-binding protein (C/EBP) (Zhang et al., 2013). Myostatin is a TGF-beta family cytokine that can repress muscle growth and disturb myogenesis. Our study observed the elevated expression of myostatin accompanied by prolonged STAT3 activation, which fully supports our conclusion. However, the most interesting mechanism revealed by our study is that the prolonged activation of STAT3 in skeletal muscle derived from either C26 tumor-bearing cachectic mice or C2C12 myotubes treated with C26 CM was correlated with elevated expression of UPS-related muscle-specific E3 ubiquitin ligases, such as Atrogin-1 and MuRF1. In addition, the transfection-induced overexpression of constitutively activated STAT3 (STAT3-C) also triggered E3 ligases, indicating that constitutive activation of STAT3 itself might be enough to induce muscle wasting by activating E3 ligases mediated protein degradation pathways. However, the pathway linking STAT3 and the activation of the protein degradation pathway has never been evaluated.
FOXO1, as a key member of the forkhead box protein O (FOXO) transcription factor family, has been found to be widely expressed in various types of metabolism-associated organs, such as the liver and skeletal muscles. In skeletal muscles, the FOXO1 transcription factor represents a master regulator of muscle growth, regeneration, and atrophy. Increased FOXO1 expression and activation have been found in skeletal muscle in energy-deprived states such as fasting, cancer, and severe diabetes (O’Neill et al., 2019; Reed, Sandesara, Senf & Judge, 2012). Enhanced FOXO1 activation in skeletal muscle triggered the muscle atrophy by activation of the UPS and lysosomal proteolysis through the upregulation of FOXO downstream target genes, including atrogin-1 and MuRF1 (Milan et al., 2015). Genetic data demonstrated that STAT3 expression and activation are correlated with the intact function of FOXO transcription factors; however, whether FOXO1 is necessary for STAT3-induced muscle wasting or vice versa is still not known. In this study, we demonstrated for the first time that STAT3 could regulate the expression and nuclear translocation of FOXO1 in the skeletal muscle, subsequently up-regulated E3 ligases and FOXO1-dependent triggering of muscle wasting, which could partly explain the pathological mechanism by which consistent STAT3 activation in the skeletal muscle could trigger muscle atrophy. Since the bioinformatics study demonstrated that there are several STAT3 consensus motifs in FOXO1 promoter regions, our subsequent ChIP analysis indicated that STAT3 could bind to the FOXO1 promoter and directly regulate the FOXO1 gene transcription in skeletal muscles. To test the pathogenic effects of STAT3 on muscle atrophy, a constitutively active recombinant STAT3 mutant expression plasmid, STAT3-C, was adopted in our study. Our results showed that the overexpression of the constitutively activated form of STAT3 in skeletal muscle could mimic the effect of the C26 cell-conditioned medium on muscle wasting phenotype induction. Knocking down the expression of FOXO1 in STAT3-C-transfected or C26 cell-conditioned medium-treated myotubes almost completely abolished the activation of the ubiquitin-proteasome system and eliminated muscle wasting, suggesting that STAT3 activation-induced muscle atrophy is indeed FOXO1 dependent.
In summary, our study demonstrated that HSP90 inhibitors could effectively alleviate cancer cachexia-induced skeletal muscle wastingin vivo ; these effects might be associated with the abnormal activation of the JAK2/STAT3/FOXO1 and ubiquitin-proteasome pathways. The close monitoring of HSP90 levels in cancer patients and the development of strategies to block HSP90 activity could potentially prevent or reverse cachexia development.