Dysregulation of metabolic-associated pathways in muscle of breast cancer patients: preclinical evaluation of interleukin-15 targeting fatigue

Abstract

Background: Breast cancer patients report a perception of increased muscle fatigue, which can persist following surgery and standardized therapies. In a clinical experiment, we tested the hypothesis that pathways regulating skeletal muscle fatigue are down-regulated in skeletal muscle of breast cancer patients and that different muscle gene expression patterns exist between breast tumour subtypes. In a preclinical study, we tested the hypothesis that mammary tumour growth in mice induces skeletal muscle fatigue and that overexpression of the cytokine interleukin-15 (IL-15) can attenuate mammary tumour-induced muscle fatigue.

Methods: Early stage non-metastatic female breast cancer patients (n = 14) and female non-cancer patients (n = 6) provided a muscle biopsy of the pectoralis major muscle during mastectomy, lumpectomy, or breast reconstruction surgeries. The breast cancer patients were diagnosed with either luminal (ER⁺ /PR⁺ , n = 6), triple positive (ER⁺ /PR⁺ /Her2/neu⁺ , n = 5), or triple negative (ER^- /PR^- /Her2/neu^- , n = 3) breast tumours and were being treated with curative intent either with neoadjuvant chemotherapy followed by surgery or surgery followed by standard post-operative therapy. Biopsies were used for RNA-sequencing to compare the skeletal muscle gene expression patterns between breast cancer patients and non-cancer patients. The C57BL/6 mouse syngeneic mammary tumour cell line, E0771, was used to induce mammary tumours in immunocompetent mice, and isometric muscle contractile properties and fatigue properties were analysed following 4 weeks of tumour growth.

Results: RNA-sequencing and subsequent bioinformatics analyses revealed a dysregulation of canonical pathways involved in oxidative phosphorylation, mitochondrial dysfunction, peroxisome proliferator-activated receptor signalling and activation, and IL-15 signalling and production. In a preclinical mouse model of breast cancer, the rate of muscle fatigue was greater in mice exposed to mammary tumour growth for 4 weeks, and this greater muscle fatigue was attenuated in transgenic mice that overexpressed the cytokine IL-15.

Conclusions: Our data identify novel genes and pathways dysregulated in the muscles of breast cancer patients with early stage non-metastatic disease, with particularly aberrant expression among genes that would predispose these patients to greater muscle fatigue. Furthermore, we demonstrate that IL-15 overexpression can attenuate muscle fatigue associated with mammary tumour growth in a preclinical mouse model of breast cancer. Therefore, we propose that skeletal muscle fatigue is an inherent consequence of breast tumour growth, and this greater fatigue can be targeted therapeutically.

Keywords: Breast cancer; Cachexia; Fatigue; IL-15; Transcriptome.

Figure 1

Skeletal muscle transcriptome in breast cancer. (A) Principle components analysis plot showing that muscle biopsies from breast cancer patients (n = 14) cluster together regardless of tumour subtype, while biopsies from non‐cancer patients (n = 6) are more variable. This plot was the rationale for analysing all breast cancer samples together instead of dividing based on the specific breast tumour subtype. (B) Normalized gene expression heat map showing the differential expression patterns of the 40 most differentially expressed genes between breast cancer patients and non‐cancer patients. There is a down‐regulation of gene expression in muscle biopsies from breast cancer patients compared with muscle from non‐cancer patients. BC, breast cancer; CON, non‐cancer control.

Figure 2

Canonical pathways of oxidative phosphorylation and mitochondrial dysfunction. The canonical pathways of oxidative phosphorylation and mitochondrial dysfunction were dysregulated in the muscle samples from breast cancer patients compared with non‐cancer controls. The shading pattern of the individual electron transport chain complexes is used as an indicator of the degree of predicted dysfunction, with a greater percentage of light shading suggestive of greater dysfunction. Individual genes that were significantly down‐regulated are identified within the specific electron transport chain complex. Specific dysregulated genes were identified using DESqe2, while the canonical pathways of oxidative phosphorylation and mitochondrial dysfunction were identified using Ingenuity Pathway Analysis software. ATP, adenosine triphosphate; IMM, Inner Mitochondrial Membrane.

Figure 3

Regulatory gene network upstream of beta‐oxidation in mitochondria. Multiple genes involved in energy homoeostasis and mitochondrial metabolism were down‐regulated, while genes associated with muscle differentiation and stem cell function were up‐regulated in muscle from breast cancer patients. This specific gene network predicts an impairment in beta‐oxidation within the mitochondria. Blue shading/arrows indicate down‐regulation, and orange shading/lines indicate up‐regulation.

Figure 4

Mammary tumour‐induced skeletal muscle dysfunction in wild type mice. (A) Representative live in vivo bioluminescent images of luciferase‐containing E0771 tumour cells through 4 weeks of tumour growth in the mammary fat pads of wild type mice. (B) Bioluminescent quantification of luciferase‐containing E0771 tumour cells through 4 weeks of tumour growth in the mammary fat pads of wild type mice. (C) Changes in body weight in wild type after 4 weeks of tumour growth in the mammary fat pads of wild type mice. These weights do not include the weight of the tumour, which was resected. (D) Differential expression of the Il15 gene in skeletal muscles of wild type mice with no tumour, and following 2 and 4 weeks of tumour growth in the mammary fat pads. (E) Ex vivo skeletal muscle fatigue curves from wild type mice with no tumour, and following 2 and 4 weeks of tumour growth in the mammary fat pads. The leftward shift of the fatigue curve in the 4WK mice is indicative of a greater rate of muscle fatigue. (F) Quantification of the area under the fatigue curve in muscles from wild type mice with no tumour, and following 2 and 4 weeks of tumour growth in the mammary fat pads. (G) Isometric force output at increasing stimulation frequencies in muscles from wild type mice with no tumour, and following 2 and 4 weeks of tumour growth in the mammary fat pads. *P < 0.05; **P < 0.001; ***P < 0.0001. CON, control mice; EDL, extensor digitorum longus; IL‐15, interleukin‐15.

Figure 5

Skeletal muscle function in IL15TG mice following mammary tumour growth. (A) Representative live in vivo bioluminescent images of luciferase‐containing E0771 tumour cells after 4 weeks of tumour growth in the mammary fat pads of B6 littermate control and IL15TG mice. (B) Bioluminescent quantification of luciferase‐containing E0771 tumour cells through 4 weeks of tumour growth in the mammary fat pads of B6 littermate control and IL15TG mice. (C) Changes in body weight in wild type after 4 weeks of tumour growth in the mammary fat pads of wild type mice. These weights do not include the weight of the tumour, which was resected. (D) Ex vivo skeletal muscle fatigue curves from B6 littermate control mice with no tumour, and following 4 weeks of tumour growth in the mammary fat pads. The leftward shift of the fatigue curve in the CON tumour mice is indicative of a greater rate of muscle fatigue. (E) Quantification of the area under the fatigue curve in muscles from B6 littermate control mice with no tumour, and following 4 weeks of tumour growth in the mammary fat pads. (F) Isometric force output at increasing stimulation frequencies in muscles from B6 littermate control mice with no tumour, and following 4 weeks of tumour growth in the mammary fat pads. (G) Ex vivo skeletal muscle fatigue curves from IL15TG mice with no tumour, and following 4 weeks of tumour growth in the mammary fat pads. (H) Quantification of the area under the fatigue curve in muscles from IL15TG mice with no tumour, and following 4 weeks of tumour growth in the mammary fat pads. (F) Isometric force output at increasing stimulation frequencies in muscles from IL15TG mice with no tumour, and following 4 weeks of tumour growth in the mammary fat pads. In Figures C, E, and F, solid lines represent the average value from wild type non‐tumour‐bearing mice, and dotted lines represent the average value from wild type mice following 4 weeks of tumour growth. *P < 0.05; **P < 0.001; ***P < 0.0001. EDL, extensor digitorum longus.