STAT3 activation in skeletal muscle links muscle wasting and the acute phase response in cancer cachexia

Abstract

Background: Cachexia, or weight loss despite adequate nutrition, significantly impairs quality of life and response to therapy in cancer patients. In cancer patients, skeletal muscle wasting, weight loss and mortality are all positively associated with increased serum cytokines, particularly Interleukin-6 (IL-6), and the presence of the acute phase response. Acute phase proteins, including fibrinogen and serum amyloid A (SAA) are synthesized by hepatocytes in response to IL-6 as part of the innate immune response. To gain insight into the relationships among these observations, we studied mice with moderate and severe Colon-26 (C26)-carcinoma cachexia.

Methodology/principal findings: Moderate and severe C26 cachexia was associated with high serum IL-6 and IL-6 family cytokines and highly similar patterns of skeletal muscle gene expression. The top canonical pathways up-regulated in both were the complement/coagulation cascade, proteasome, MAPK signaling, and the IL-6 and STAT3 pathways. Cachexia was associated with increased muscle pY705-STAT3 and increased STAT3 localization in myonuclei. STAT3 target genes, including SOCS3 mRNA and acute phase response proteins, were highly induced in cachectic muscle. IL-6 treatment and STAT3 activation both also induced fibrinogen in cultured C2C12 myotubes. Quantitation of muscle versus liver fibrinogen and SAA protein levels indicates that muscle contributes a large fraction of serum acute phase proteins in cancer.

Conclusions/significance: These results suggest that the STAT3 transcriptome is a major mechanism for wasting in cancer. Through IL-6/STAT3 activation, skeletal muscle is induced to synthesize acute phase proteins, thus establishing a molecular link between the observations of high IL-6, increased acute phase response proteins and muscle wasting in cancer. These results suggest a mechanism by which STAT3 might causally influence muscle wasting by altering the profile of genes expressed and translated in muscle such that amino acids liberated by increased proteolysis in cachexia are synthesized into acute phase proteins and exported into the blood.

Figure 1. Characterization of C26 cachexia.

A, B, Body weight changes for control mice and for mice injected with C26 tumor cells. Mice euthanized on day 19 had lost ∼10% body weight, considered moderate cachexia. Mice euthanized on day 24 had lost ∼15% body weight, considered severe cachexia. C, Significantly decreased quadriceps and gastrocnemius weights were observed with C26 cachexia. Differences were not significant in muscles from moderate versus severe cachexia. D, Liver weight and tumor weight both increased with severity/duration of cachexia. E, Representative Western blotting for Myosin Heavy Chain (MyHC) in quadriceps muscle from both control and tumor bearing mice. F, Densitometric quantification of the Western blotting for MyHC protein shows a marked reduction in both moderate (−69% vs. controls) and severe (−81% vs. controls) cachexia. G, Total protein content in quadriceps muscle from control and tumor-bearing mice. Protein content is significantly reduced in both moderate (−7% vs. controls) and severe (−13% vs. controls). n = 4–5 per group; **P<0.01, ***P<0.001.

Figure 2. Overlapping patterns of gene expression in skeletal muscle in C26 cachexia.

A, Heat map of quadriceps gene expression in C26 cachexia showing distinct clustering of genes by experimental groups. Samples are normalized to the controls. Blue indicates down-regulated genes, yellow up-regulated genes, and black no change. Only genes with P<0.05 by one-way ANOVA are shown. B, Comparison of gene expression in moderate and severe C26 cachexia (NextBio).

Figure 3. Expression of known genes in muscle growth regulation in C26 cachexia.

A, Expression of the ubiquitin ligases, Fbxo32/Atrogin-1 and Trim63/MuRF1 are induced by microarray and qPCR analysis. Expression of genes encoding muscle structural proteins, including ACTA1 and MHY8 are decreased. B, Gene expression patterns in C26 cachexia mapped onto a schematic of muscle growth regulation pathways. Genes with demonstrated growth-promoting activity in muscle are shown in green, growth inhibitory genes in red. The number on the left represents fold-change in moderate cachexia and right in severe cachexia, with the arrow indicating the direction of the change. Genes with no values were either not changed or not present in the dataset.

Figure 4. C26 cachexia pathway analysis.

A. The top 20 Broad MSigDB canonical pathways significantly up-regulated in moderate and severe cachexia versus controls. Pink bars (upper axis) represent the significance of the overlap. The triangles (moderate cachexia) and circles (severe cachexia) (lower axis) denote the number of genes differentially expressed in that pathway. Red arrows indicate pathways related to IL-6/STAT3/inflammation. B, The top 20 Broad MSigDB canonical pathways significantly down-regulated in moderate and severe cachexia versus controls. Green bars (upper axes) represent the significance of the overlap. The triangles (moderate cachexia) and circles (severe cachexia) (lower axis) as above.

Figure 5. STAT3 and its target genes are activated in muscle in C26 cachexia.

A: Heat map of gene expression changes of Stat3 and co-cited gene products, as identified by Genomatix Bibliosphere. Blue indicates down regulated genes, yellow up regulated, and black no change. Only genes with P<0.05 by one-way ANOVA are shown. B: A subset of STAT3 target genes identified through the literature are differentially regulated in moderate and severe cachexia versus controls. C: STAT3 and its target genes SOCS3 and CEBPD are increased at the mRNA level by microarray and qPCR. D: Protein levels for p-STAT3 and STAT3 in protein extracts from quadriceps, gastrocnemius and liver evaluated by Western blotting analysis. E: SOCS3 protein levels. F: quantitative analysis of p-STAT3/STAT3 and p-STAT3/GAPDH ratio (expressed as fold-change vs. controls). G: quantitative analysis of SOCS3 protein levels (expressed as fold-change vs. controls). GAPDH was used as an internal reference to confirm equal loading. n = 3–5 per group; *P<0.05, **P<0.01, ***P<0.001 vs. Controls, ^$P<0.05 vs. moderate.

Figure 6. STAT3 nuclear localization is increased in the muscle of C26-bearing mice.

A, p-STAT3 protein levels are increased in nuclear extracts prepared from quadriceps of severely cachectic C26 tumor-bearing mice compared to controls. n = 5 per group. B, Immunofluorescence analysis performed reveals increased pSTAT3 localization in myonuclei of gastrocnemius from severely cachectic C26 tumor-gearing mice (green). Nuclear staining is shown in blue (DAPI).

Figure 7. Robust expression of acute phase response proteins in skeletal muscle versus liver in C26 cachexia.

A. Western blotting and quantitation of fibrinogen levels in control and C26 quadriceps and liver. Data (mean ± SEM) are expressed as relative densitometry value. **P<0.01, ***P<0.001. B, Western blotting analysis of fibrinogen standard proteins and quadriceps and liver extracts for control, CHO-IL6 injected nude mice and C26 injected CD2F1 mice. Quantitation was performed on the band indicated by the arrow. Data (means ± SEM) are expressed as ng fibrinogen / µg protein. *P<0.05, **P<0.01, ***P<0.001. C, Western blotting analysis demonstrates significantly increased fibrinogen and SAA1 protein levels in quadriceps and gastrocnemius in moderate and severe C26 cachexia. *P<0.05, **P<0.01, ***P<0.001.

Figure 8. Fibrinogen is produced and released by the skeletal muscle following activation of the IL-6/STAT3 pathway.

A, Western blotting analysis and quantitation of fibrinogen in C2C12 myotubes infected with Ad-cSTAT3-GFP or Ad-GFP as control. Fibrinogen expression was increased consistent with the increase in the levels of STAT3. **P<0.01, ***P<0.001 vs. GFP. B, Western blotting analysis and quantitation of fibrinogen expression in C2C12 treated with IL-6 (100 ng/ml) for 1, 24, 48 h. GAPDH was used as loading control. Increased expression of fibrinogen was observed at each time point after IL-6 treatment. Data (means ± SEM) are expressed as relative densitometry value. ***P<0.001 vs. respective controls. C, Fibrinogen levels by ELISA of the conditioned medium of C2C12 exposed to IL-6 for 30 min, 1, 6, 24, 48 h. Fibrinogen levels were significantly elevated after 6, 24 and 48 h of IL-6 treatment. Data (means ± SEM) are expressed as ng/ml. **P<0.01, ***P<0.001 vs. controls (C).