Fenofibrate prevents skeletal muscle loss in mice with lung cancer

Abstract

The cancer anorexia cachexia syndrome is a systemic metabolic disorder characterized by the catabolism of stored nutrients in skeletal muscle and adipose tissue that is particularly prevalent in nonsmall cell lung cancer (NSCLC). Loss of skeletal muscle results in functional impairments and increased mortality. The aim of the present study was to characterize the changes in systemic metabolism in a genetically engineered mouse model of NSCLC. We show that a portion of these animals develop loss of skeletal muscle, loss of adipose tissue, and increased inflammatory markers mirroring the human cachexia syndrome. Using noncachexic and fasted animals as controls, we report a unique cachexia metabolite phenotype that includes the loss of peroxisome proliferator-activated receptor-α (PPARα) -dependent ketone production by the liver. In this setting, glucocorticoid levels rise and correlate with skeletal muscle degradation and hepatic markers of gluconeogenesis. Restoring ketone production using the PPARα agonist, fenofibrate, prevents the loss of skeletal muscle mass and body weight. These results demonstrate how targeting hepatic metabolism can prevent muscle wasting in lung cancer, and provide evidence for a therapeutic strategy.

Keywords: cachexia; fenofibrate; glucocorticoids; ketones; skeletal muscle.

Fig. 1.

Cachexia development in genetically induced lung cancer. (A) Weight normalized to peak weight over time following induction of lung cancer using adenovirus to deliver Cre recombinase (Cre, red boxes) and control mice treated with an empty adenovirus (Empty, black triangles, n = 6). Following Cre treatment, n = 12 mice developed cachexia (+cacs) and n = 8 mice maintained their weight (−cacs). (B) Food intake in mice from A treated with Empty or Cre virus. (C) Gastrocnemius (Gastroc) weight in Empty, Cre−cacs, and Cre+cacs mice. (D) Gonadal WAT weight in Empty, Cre−cacs, and Cre+cacs mice. (E) Liver weight in Empty, Cre −cacs, and Cre +cacs mice. (F) Kidney weight in Empty, Cre −cacs, and Cre +cacs mice. Bar graphs are mean ± SEM. Student’s t test comparisons with the Empty group. *P < 0.05.

Fig. 2.

Cachexia does not correlate with tumor burden or subtype of lung cancer. (A) Normal lung (Top, Left), adenocarcinoma subtypes identified histologically: papillary (Top, Right), acinar (Middle, Left), mucinous (Middle, Right), lymph node metastasis containing acinar adenocarcinoma (Bottom, Left), and lepidic subtype of adenocarcinoma in situ (Bottom, Right). (Scale bars: Right Top, Middle, and Bottom and Left Middle, 50 μm; Left Top and Bottom, 100 μm.) (B) Percent of total weight loss over the duration of the study vs. tumor burden, as defined by the percent of lung replaced by tumor in mice with (Cre +cacs, red circles) or without (Cre −cacs, black circles). Linear regression analysis (R², coefficient of determination, and p, P value).

Fig. 3.

Changes in serum metabolites and hormones in mice with and without cachexia. (A) Percent weight loss, serum glucose, BHB, NEFA, and TG in fed (Fed) and fasted (Fast) nontumor-bearing mice and tumor-bearing (Cre) mice. Values in each column are color-coded in a red (highest) to blue (lowest) heatmap. (B) Serum corticosterone (mean ± SEM) levels over time following induction of lung cancer in noncachexic (Cre −cacs, red open boxes, n = 5) and cachexic (Cre +cacs, red closed boxes, n = 12).

Fig. 4.

Changes in liver metabolites and gene expression in fasted and tumor-bearing mice. (A) Liver glycogen content in fed (Fed) and fasted (Fasted) nontumor-bearing mice and tumor-bearing mice with (+cacs) and without (−cacs) cachexia. (B) Pyruvate, OAA, citrate, palmitate (C16:0), stearate (C18:0), and BHB metabolite abundance relative to noncacs liver and gastrocnemius (Gastroc). n = 5 each. (C) Immunoblot of liver lysates for indicated proteins. (D) Normalized quantitation of ACOX1 intensity over β-actin intensity from the immunoblot in D normalized to the average value from the −cacs group. (E) PPARα immunoprecipitation from hepatic nuclear (Nuc) and cytoplasmic (Cyto) fractions taken from Fed, Fasted, fenofibrate (Feno)-treated, and tumor-bearing mice −cacs and +cacs. (F) FPKM from RNA-Seq for indicated genes in −cacs, +cacs, and Fasted mice. Box and whiskers plot with minimum and maximum whiskers. (G) Schematic view of pyruvate generated from different amino acids using genes from F. Red color denotes up-regulated genes. Bar graphs are mean ± SEM. Student’s t test compared with Fed mice in A and to −cacs in B, D, and E. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 5.

Cachexia promotes Type IIx/b myofiber atrophy. (A) EDL and soleus muscle weight in nontumor-bearing fed and fasted mice in comparison with tumor-bearing mice with (+cacs) and without (−cacs) cachexia. (B) Representative photomicrograph of skeletal muscle cross-section following myosin heavy-chain fiber typing using immunofluorescence on EDL (Left) and soleus (Right). Cy5 (pink, type I), GFP (green, type IIA), unstained (black, type IIx), and dsRed (red, type IIB). (Scale bar, 320 μm.) (C) Type I and (D) type IIA myofiber CSA from noncachexic (black, −cacs) and cachexic (red, +cacs) EDL (Left) and soleus (Right) muscles. (E) Type IIx (Left) and type IIb (Right) EDL myofiber CSA from noncachexic (black, −cacs) and cachexic (red, +cacs) mice. Myofiber CSA from n = 3 mice were pooled to obtain data for C–E. Mean ± SEM indicated in each plot. Student’s t test comparing −cacs and +cacs groups, *P < 0.05, ****P < 0.0001.

Fig. 6.

Changes in skeletal muscle metabolites and gene expression in fasted and tumor-bearing mice. (A) Unbiased clustering of RNA-Seq using gastrocnemius from nontumor-bearing fasted mice compared with tumor-bearing cachexic (+cacs) and noncachexic (−cacs) mice. n = 5 each. (B) Venn diagram displaying the proportion and examples of differentially expressed genes in +cacs samples alone, fasted samples alone, and genes contained in both conditions. (C) Relative expression of −cacs, +cacs, and fasted gastrocnemius mRNA using qPCR normalized to the average value from the −cacs group. n = 5. (D) Immunoblot of tibialis anterior lysates for indicated proteins. (E) Relative abundance of amino acids in gastrocnemius muscles from +cacs, −cacs, and fasted mice normalized to the average value from the −cacs group. Bar graphs are mean ± SEM. Student’s t test comparing +cacs to −cacs mice. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 7.

Fenofibrate treatment prevents cachexia in tumor-bearing mice. (A) Weight normalized to peak weight over time following induction of lung cancer using adenovirus to deliver Cre recombinase in mice fed dietary fenofibrate (Cre+feno, orange open diamonds, n = 7) starting at 6.5 wk (black arrow) compared with Cre-treated (Cre) mice fed normal chow (n = 20) and nontumor bearing mice fed dietary fenofibrate (Fed+feno, blue closed diamonds, n = 4). (B) Food intake in mice from A. Red boxes represent Cre data reproduced from Fig. 1B. (C) Relative expression of PPARα target genes using qPCR from livers of nontumor-bearing (Fed, black bars, n = 5) and tumor bearing (Cre, n = 10) mice on a normal chow diet compared with nontumor bearing (Fed+feno, blue bars, n = 3) and tumor-bearing (Cre+feno, orange bars, n = 7) mice fed dietary fenofibrate (feno). (D) Serum BHB (Left) and NEFA (Right) from mice described in C. (E) Serum corticosterone levels over time following induction of lung cancer in nontumor-bearing (Fed+feno, blue diamonds, n = 4) and tumor-bearing (Cre+feno, orange diamonds, n = 7) mice fed dietary fenofibrate starting at 6.5 wk. (F) Gastrocnemius (Gastroc) weight from nontumor-bearing (Fed, black triangles, n = 5) and tumor bearing (Cre, red boxes, n = 16) mice on a normal chow diet compared with nontumor-bearing (Fed+feno, blue diamonds, n = 3) and tumor-bearing (Cre+feno, orange diamonds, n = 7) mice fed dietary fenofibrate (feno). (G) Gonadal WAT from mice described in F. (H) Changes in intermediary metabolism following onset of cachexia: excess lipolysis in adipose tissue increases serum NEFA, which can be used by the liver and skeletal muscle. Defects in hepatic ketogenesis reduce serum ketone levels resulting in excess glucocorticoid production. Glucocorticoids modulate hepatic metabolism and induce skeletal muscle degradation to enhance gluconeogenesis. Fenofibrate restores hepatic ketogenesis and reduces serum glucocorticoids thereby preventing skeletal muscle degradation. Bar graphs are mean ± SEM. Student’s t test compared with Fed mice: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Student’s t test compared with Cre mice: ^#P < 0.05, ^##P < 0.01, ^###P < 0.001, ^####P < 0.0001.