Muscle NAD+ depletion and Serpina3n as molecular determinants of murine cancer cachexia-the effects of blocking myostatin and activins

Abstract

Objective: Cancer cachexia and muscle loss are associated with increased morbidity and mortality. In preclinical animal models, blocking activin receptor (ACVR) ligands has improved survival and prevented muscle wasting in cancer cachexia without an effect on tumour growth. However, the underlying mechanisms are poorly understood. This study aimed to identify cancer cachexia and soluble ACVR (sACVR) administration-evoked changes in muscle proteome.

Methods: Healthy and C26 tumour-bearing (TB) mice were treated with recombinant sACVR. The sACVR or PBS control were administered either prior to the tumour formation or by continued administration before and after tumour formation. Muscles were analysed by quantitative proteomics with further examination of mitochondria and nicotinamide adenine dinucleotide (NAD⁺) metabolism. To complement the first prophylactic experiment, sACVR (or PBS) was injected as a treatment after tumour cell inoculation.

Results: Muscle proteomics in TB cachectic mice revealed downregulated signatures for mitochondrial oxidative phosphorylation (OXPHOS) and increased acute phase response (APR). These were accompanied by muscle NAD⁺ deficiency, alterations in NAD⁺ biosynthesis including downregulation of nicotinamide riboside kinase 2 (Nrk2), and decreased muscle protein synthesis. The disturbances in NAD⁺ metabolism and protein synthesis were rescued by treatment with sACVR. Across the whole proteome and APR, in particular, Serpina3n represented the most upregulated protein and the strongest predictor of cachexia. However, the increase in Serpina3n expression was associated with increased inflammation rather than decreased muscle mass and/or protein synthesis.

Conclusions: We present evidence implicating disturbed muscle mitochondrial OXPHOS proteome and NAD⁺ homeostasis in experimental cancer cachexia. Treatment of TB mice with a blocker of activin receptor ligands restores depleted muscle NAD⁺ and Nrk2, as well as decreased muscle protein synthesis. These results indicate putative new treatment therapies for cachexia and that although acute phase protein Serpina3n may serve as a predictor of cachexia, it more likely reflects a condition of elevated inflammation.

Keywords: APR; Activin receptor; C26; Cancer cachexia; Nrk2; OXPHOS.

Figure 1

Overall study design (A). Experiment 1. Mice were inoculated with C26 cells on day 0, and sACVR (5 mg/kg) or PBS vehicle were administered on days −11, −7, −3, 1, 5, and 9. Muscles, liver, and blood were collected 11 days after the C26 cell inoculation. In experiment 2, sACVR (10 mg/kg) or PBS was injected only on days 6 and 9 after C26 cell inoculation. Heat map representation of differential proteome changes in 3 conditions showing proteins with the strongest response to cancer and anti-cachexia treatment in muscle in experiment 1 (B). For stringency, 3 conditions with changes in protein expression displaying only high fold change ratios are visualised (FC>|2.0|, FDR corrected p value < 0.05 and ≥ 2 unique peptides used for quantitation). C26 + PBS vs control (a), C26 + sACVR/b vs C26 + PBS (b) and C26 + sACVR/c vs C26 + PBS (c). An expression-based heat map was drawn by implementing average linkage and the Spearman rank correlation distance measure by using Heatmapper.

Figure 2

Ingenuity pathway analysis reveals pathways and upstream regulators affected in muscle by cancer and anti-cachexia treatment in experiment 1. Canonical pathways analysis (A). Pathways with significant z-scores (>|2|) in one of 3 comparisons are presented, and if available, z-scores of those pathways are also shown from other comparisons. Heat map representation (B) of OXPHOS proteins with largest differential changes in their expression, in 3 comparisons: (a) C26 + PBS vs control, (b) C26 + sACVR/b vs C26 + PBS, and (c) C26 + sACVR/c vs C26 + PBS. The criteria used for assessment were as follows: fold change, FC>|1.3|, FDR corrected p value < 0.05, and ≥2 unique peptides used for quantitation. An expression-based heat map using average linkage and the Spearman rank correlation distance measures. Upstream analysis pinpointing upstream regulators predicted to be suppressed (C) or activated (D) in C26 + PBS vs CTRL comparison (z-score >|2|) and if available, z-scores of those pathways are also shown from other comparisons. Phosphorylated ACC at ser⁷⁹ as a proxy for AMPK activation (p-ACC; E) and p-Akt at Ser⁴⁷³ as a proxy for mTORC2 activation (F). ∗ = P < 0.05 using a two-tailed unpaired Student's t test.

Figure 3

Muscle mitochondrial activity in situ is in part altered in cancer in experiment 1. Percentage of fibres with high SDH (A) or COX (B) activities. ∗ = P < 0.05 using a two-tailed unpaired Student's t test.

Figure 4

NAD⁺ and its regulators are affected in TB-mice and in part rescued by anti-cachexia treatment in experiment 1. Contents of NAD⁺ (A) and NADH (B), and PARP activity (C). Gene expression of the regulators of NAD⁺ biosynthesis (D). ∗ = P < 0.05, ∗∗ = P < 0.01, and ∗∗∗ = P < 0.001 when using a two-tailed unpaired Student's t test.

Figure 5

Anti-cachexia treatment after tumour inoculation (experiment 2) reveals increased body mass, muscle masses, and muscle protein synthesis and repleted NAD⁺. Changes in body mass (BM; A), tissue masses (B), muscle protein synthesis (C), contents of NAD⁺ (D), and gene expression of the regulators of NAD⁺ biosynthesis (E) in control mice and in mice treated with sACVR or PBS. Epi. fat = epididymal fat pad. ∗ = P < 0.05, ∗∗ = P < 0.01, and ∗∗∗ = P < 0.001 when using a two-tailed unpaired Student's t test.

Figure 6

APR is activated in TB-mice independent of anti-cachexia treatment. Heat map representation of differential proteome changes of APPs in experiment 1. The analyses in 3 conditions with the highest fold change ratio are shown (FC>|1.3|, FDR corrected P value < 0.05 and ≥ 2 unique peptides used for quantitation), namely: (a) C26 + PBS vs control, (b) C26 + sACVR/b vs C26 + PBS, and (c) C26 + sACVR/c vs C26 + PBS. An expression-based heat map is using average linkage and the Spearman rank correlation distance measures (A). Phosphorylated STAT3 (p-STAT3; B) and Serpina3n quantifications in muscle (C) and in serum (D). Correlation analysis of Serpina3n in the muscle, quantified either by label free proteomics or with Western blotting and cachexia (a decrease in the body mass, BM; E). Serpina3n in muscle (F) and in serum (G) in mice treated with sACVR/PBS after tumour cell inoculation in experiment 2. ∗ = P < 0.05, ∗∗ = P < 0.01, and ∗∗∗ = P < 0.001 when using a two-tailed unpaired Student's t test.