Metabolomic Analyses Reveal Extensive Progenitor Cell Deficiencies in a Mouse Model of Duchenne Muscular Dystrophy

Abstract

Duchenne muscular dystrophy (DMD) is a musculoskeletal disorder that causes severe morbidity and reduced lifespan. Individuals with DMD have an X-linked mutation that impairs their ability to produce functional dystrophin protein in muscle. No cure exists for this disease and the few therapies that are available do not dramatically delay disease progression. Thus, there is a need to better understand the mechanisms underlying DMD which may ultimately lead to improved treatment options. The muscular dystrophy (MDX) mouse model is frequently used to explore DMD disease traits. Though some studies of metabolism in dystrophic mice exist, few have characterized metabolic profiles of supporting cells in the diseased environment. Using nontargeted metabolomics we characterized metabolic alterations in muscle satellite cells (SCs) and serum of MDX mice. Additionally, live-cell imaging revealed MDX-derived adipose progenitor cell (APC) defects. Finally, metabolomic studies revealed a striking elevation of acylcarnitines in MDX APCs, which we show can inhibit APC proliferation. Together, these studies highlight widespread metabolic alterations in multiple progenitor cell types and serum from MDX mice and implicate dystrophy-associated metabolite imbalances in APCs as a potential contributor to adipose tissue disequilibrium in DMD.

Keywords: Duchenne muscular dystrophy; adipose tissue; metabolomics; skeletal muscle; stem cells.

Figure 1

Metabolomic analysis of muscle satellite cells. (A) A principal component analysis (PCA) plot derived from nontargeted metabolite profiling of muscle satellite cells isolated from four WT control and four MDX mice. (B) A dendrogram depicting hierarchical clustering results of samples from (A). Clustering method: complete linkage. Distance measurement: Euclidean. (C) A volcano plot depicting differentially abundant metabolites (DAMs) between WT and MDX satellite cells (SCs). Horizontal and vertical dashed lines represent a threshold of p-value 0.05 and fold change 1.5, respectively. Seven-hundred-and-fifty-five features were identified to be significant from 11,151 features. (D) Pathway analysis of SC DAMs. Italics: pathways with multiple metabolites implicated. Red: metabolic pathways identified in both serum and muscle satellite cell analyses.

Figure 2

Serum metabolomic analysis. (A) A principal component analysis (PCA) plot derived from nontargeted metabolite profiling of serum isolated from four WT control and four MDX mice. (B) A dendrogram depicting hierarchical clustering results of samples from (A). Clustering method: complete linkage. Distance measurement: Euclidean. (C) A volcano plot depicting DAMs between WT and MDX serum. Horizontal and vertical dashed lines represent a threshold of p-value 0.05 and fold change 1.5, respectively. One-thousand-one-hundred-and-twenty-five features were identified to be significant from 9562 total features. (D) Pathway analysis of serum DAMs. Italics: pathways with multiple metabolites implicated. Red: metabolic pathways identified in both serum and muscle satellite cell analyses.

Figure 3

MDX mice exhibit decreased adipose tissue mass. (A) Images depicting WT and MDX epididymal fat pads. (B) A bar graph depicting epididymal fat pad weight normalized to total mouse weight. * p < 0.05. n = 5 mice in each experimental group. (C) Representative dual-energy X-ray absorptiometry (DEXA) images of WT and MDX mice. (D) Quantification of fat percentage based on DEXA image analysis. * p < 0.05. n = 4 mice in each experimental group.

Figure 4

Primary adipose progenitor cells (APCs) from MDX mice exhibit in vitro expansion defects. (A) Representative images of APC cultures derived from WT and MDX mice at 3d and 5d post-isolation. Shown are phase contrast images and corresponding mask overlays used for proliferation quantification analyses. (B) Line graphs depicting proliferation curves of WT (blue curves) and MDX (red curves) APCs over a 5d time-course. Shown are proliferation curves based on initial seeding densities of 2.5 K (light coloring), 5 K (medium coloring), and 10 K (dark coloring) APCs/well of a 96-well plate.

Figure 5

APC metabolomic analysis. (A) A principal component analysis (PCA) plot derived from nontargeted metabolite profiling of APCs isolated from four WT control and four MDX mice. (B) A dendrogram depicting hierarchical clustering results of samples from (A). Clustering method: complete linkage. Distance measurement: Euclidean. (C) A volcano plot depicting DAMs between WT and MDX APCs. Horizontal and vertical dashed lines represent a threshold of p-value 0.05 and fold change 1.5, respectively. One-hundred-and-fifty-three features were identified to be significant from 11,151 total features. (D) Pathway analysis of serum DAMs. Italics: pathways with multiple metabolites implicated. Red: metabolic pathways identified in common with both serum and muscle satellite cell analyses. (E) A bar graph quantifying the relative abundance of five acylcarnitine species. * p < 0.05, ** p < 0.01. n = 4 mice in each experimental group.

Figure 6

Exposure to selected acylcarnitines inhibits in vitro APC expansion. (A) A line graph depicting APC proliferation over a >10-day timecourse. Shown are representative proliferation traces of APCs treated with a vehicle control (water or MeOH), propionyl-l-carnitine (yellow), butyryl-l-carnitine (orange), octanoyl-l-carnitine (brown), palmitoyl-l-carnitine (light blue), or stearoyl-l-carnitine (dark blue). (B) A line graph depicting APC proliferation when exposed to increasing concentrations of stearoyl-l-carnitine.