We sought to mimic the high serum IL-6, acute phase response and muscle wasting of patients with cancer cachexia. We chose C26 adenocarcinoma, which exhibits increased circulating levels of IL-6 that coincide with muscle wasting 
. Certain clones of the C26 that do not cause cachexia coincidently do not produce IL-6 
. Consistent with a causative role in muscle wasting in humans, circulating IL-6 has been reported to be a marker of weight loss in patients afflicted by various forms of cancer 
. Moreover, direct administration of IL-6 to mice induces systemic muscle wasting. The consistency of such muscle wasting across various routes of IL-6 administration, including by direct injection of recombinant IL-6 
, by transgenesis 
, by implantation of osmotic pump delivering recombinant IL-6 
, by injection of IL-6 expressing CHO cells into athymic nude mice 
, and by transfection of plasmid DNA encoding IL-6 
, testifies to the potency by which IL-6 causes the cachectic phenotype.
Regardless, IL-6 is likely not the only cytokine mediating muscle wasting in cancer or even in the C26 model. Inhibition of IL-6 only partially rescues muscle wasting in the C26 model, causing some to conclude that IL-6 is only one of several players involved in the C26 model and cannot by itself induce the full cachectic syndrome 
. The other cytokines we observed increased in C26 mice that might also play a role in muscle wasting include IL-6-family ligands such as LIF, as well as TNF-a, IFN-gamma, all of which can induce cachexia independently. Zhou et al. recently hypothesized IL-6 might serve only a marginal role in cachexia 
. Consistent with prior studies 
, they reported that myostatin inhibition using a soluble receptor-fusion protein developed by Lee et al. 
reduced muscle loss in the C26 model. Serum IL-6 levels in such mice were not different from control treated C26 mice. As well, the authors were unable to induce muscle loss in mice given recombinant IL-6 by osmotic pump. They conclude that muscle depletion in cancer cachexia might depend upon myostatin and related ligands rather than on pro-inflammatory cytokines alone. An alternative interpretation of those data, however, is that myostatin inhibition results in muscle hypertrophy, which balances IL-6 or other cytokine-induced muscle wasting. Secondly, it is also likely that the authors failed to achieve a sufficient dose of IL-6 to effect wasting. They used human IL-6, which has a 5- to 10-fold lower activity than murine IL-6 on murine cells 
, at levels 40-60-fold less than those reported to induce severe wasting 
. (The ED50
for recombinant human IL-6 in the T1188.8.131.52 mouse plasmacytoma assay is 0.2–0.8 ng/ml, while it is 0.02–0.06 ng/ml for recombinant murine IL-6.) Thus the preponderance of evidence indicates that IL-6 can cause muscle wasting. Whether the effect is direct or indirect is still unclear and what mechanism leads to muscle wasting is still unknown.
Here we document that the STAT3 pathway is activated in skeletal muscle in C26-bearing mice and that expression of STAT3 target genes including the acute phase response genes are activated. Among the STAT3 target genes significantly induced was SOCS3, a classical feedback inhibitor of STAT3 activation. STAT3 induces expression of SOCS3 which binds to activated JAKs and receptors to inhibit STAT3 activation in at least three ways: by preventing binding of STAT to activated receptors, by binding and inhibiting activated JAKs and by targeting JAKs and receptors for degradation 
. Thus high levels of SOCS3 protein should inhibit STAT3 activation. In contrast to the high SOCS3 mRNA levels in muscle, however, we observed little to no increase in SOCS3 protein either in muscle or in liver. This lack of SOCS3 protein explains in part how high pSTAT3 levels could persist regardless of high SOCS3 RNA levels and how sustained STAT3 activation might continue to drive muscle wasting while ostensibly activating its inhibitor.
Emerging data indicate that SOCS3 is regulated transcriptionally, but also post-transcriptionally and post-translationally. TNF stabilizes SOCS3 mRNA elicited by lipopolysaccharide 
. N-terminal truncated splice variants of SOCS3 generated under stress conditions show greater stability than full-length, revealing an important role for transcriptional control of SOCS3 
. Furthermore, Jak-mediated phosphorylation of SOCS3 at two tyrosine residues in the conserved SOCS box, Tyr204 and Tyr221, fully destabilizes SOCS3 protein and activates its proteasome-mediated degradation, while, on the contrary, a phosphorylation-deficient mutant of SOCS3, Y204F/Y221F, can remain stable in the presence of activated Jak2 and receptor tyrosine kinases 
. These results indicate that JAK/STAT activation drives not only SOCS3 mRNA expression, but also its proteolytic degradation. Recently, a reduction in SOCS3 levels has been found to be associated with, and even preceding, a decrease in MyHC in an experimental model of muscle unloading 
. Taken together, these observations suggest that SOCS3 may not be present in sufficient quantities to inhibit STAT3 activation in cachectic muscle. Thus stabilization of SOCS3 protein might represent an intervention for muscle wasting with high IL-6.
In this experimental work we also confirm that at least two acute phase response proteins, fibrinogen and SAA1 are expressed in muscle, and that in the case of the former, at levels about half of that expressed in the liver. The significance of these results is at least three-fold. First, they establish skeletal muscle as an important source of acute phase protein synthesis. Second, they establish a molecular link between the observations of high IL-6, increased acute phase response proteins and muscle wasting in cancer. Third, they suggest a molecular mechanism through which STAT3 might causally influence muscle wasting by altering the profile of genes expressed and mRNAs translated in muscle.
Generally, the acute phase response is considered to be hepatic in origin, although several reports document expression of acute phase response genes in lung and mammary tissue 
. Our results indicate that skeletal muscle may be a major physiological source of acute phase response proteins, both at baseline and in pathological conditions of high IL-6, including cancer. Skeletal muscle constitutes ~40% of total body weight, while the liver is 10-fold smaller, at 4.5-5%. If the relationship of fibrinogen content we observed is representative of other acute phase response proteins, muscle might be the greater source of acute phase proteins, synthesizing about 5-times the protein produced in the liver. Thus skeletal muscle might be a key player in innate immunity.
In addition to its functional and metabolic roles, skeletal muscle is the major protein reservoir in the body. Under disease conditions, the mobilized free amino acids can also be utilized for metabolism of vital organs such as the liver, heart, brain or lung 
. Highly plastic, skeletal muscle proteolysis has been proposed to be the main source for free amino acids for the hepatic acute phase response. Prolonged synthesis of acute phase response proteins, such as fibrinogen, is metabolically expensive and might induce nutritional deprivation of skeletal muscle 
. Calculation by others suggest that catabolism of 2.6 grams of muscle protein is required to produce 1 gram of fibrinogen 
. As our experiments suggest, in addition to providing amino acids to the liver for production of acute phase response proteins, the skeletal muscle itself synthesizes acute phase response proteins. Indeed, we have shown the production and the release of fibrinogen from muscle in vitro
in C2C12 myotubes following activation of the STAT3 signaling pathway. STAT3 is the main in vivo
inducer of hepatic acute phase response expression 
, thus STAT3 is a strong candidate for mediating the muscle acute phase response.
Inflammation and correspondingly increased acute phase response protein levels are a hallmark of cancer cachexia. It has been hypothesized that hepatic synthesis of positive acute phase response proteins using amino acids liberated from skeletal muscle proteins is a major driver of skeletal muscle proteolysis, although the nature of the signal mediating both processes was not suggested 
. Our findings show that IL-6 apparently mediates both the hepatic and skeletal muscle acute phase response through STAT3 activation in cancer, thereby positioning the IL-6/STAT3 pathway as a potentially important target to reduce skeletal muscle wasting. STAT3-mediated production of acute phase response proteins could represent a major re-prioritization of protein synthesis in skeletal muscle, away from structural proteins and towards secreted proteins. Given the enhanced proteolysis of structural proteins in cachexia, freed amino acids may be synthesized into acute phase response proteins. Ultimately this would drain protein reserves in skeletal muscle and thus represents a causal mechanism for muscle wasting in cancer.
STAT3 activation has also been observed in muscle in other experimental models of cancer cachexia with high IL-6, namely ApcMin/+
. Extending these observations to other settings of muscle wasting and elevated IL-6 and related cytokines, STAT3 activation might also mediate muscle loss in obesity 
, advanced age or sarcopenia 
, inflammatory myopathies 
, burn 
, and other diseases.