Multi-level omics analysis in a murine model of dystrophin loss and therapeutic restoration

Abstract

Duchenne muscular dystrophy (DMD) is a classical monogenic disorder, a model disease for genomic studies and a priority candidate for regenerative medicine and gene therapy. Although the genetic cause of DMD is well known, the molecular pathogenesis of disease and the response to therapy are incompletely understood. Here, we describe analyses of protein, mRNA and microRNA expression in the tibialis anterior of the mdx mouse model of DMD. Notably, 3272 proteins were quantifiable and 525 identified as differentially expressed in mdx muscle (P < 0.01). Therapeutic restoration of dystrophin by exon skipping induced widespread shifts in protein and mRNA expression towards wild-type expression levels, whereas the miRNome was largely unaffected. Comparison analyses between datasets showed that protein and mRNA ratios were only weakly correlated (r = 0.405), and identified a multitude of differentially affected cellular pathways, upstream regulators and predicted miRNA-target interactions. This study provides fundamental new insights into gene expression and regulation in dystrophic muscle.

Figure 1.

Analysis of the dystrophic proteome. Protein lysates prepared from the TA muscles of 8-week-old C57 and mdx mice (n = 4) were analysed by LC-MS/MS. Significantly different (P < 0.01, q < 0.05, t-test) proteins were analysed by (A) unsupervised hierarchical clustering and (B) PCA (the first two components represent 90.9% of the data). (C) Identification of differentially expressed genes by volcano plot. (D) Schematic of the DAPC. Dystrophin, dystrobrevin, dystroglycans, sarcoglycans, laminins, syntrophin and neuronal nitric oxide synthase are significantly down-regulated. Red indicates up-regulated proteins and blue indicates down-regulated proteins. Grey indicates undetected proteins. Scale bars represent row Z-scores.

Figure 2.

Restoration of the dystrophic proteome by antisense oligonucleotide-mediated exon skipping therapy. (A) 12-week-old mdx mice (n = 4) were injected with a single 12.5 mg/kg intravenous dose of Pip6e-PMO and harvested 2 weeks later. Age-matched (14-week old) C57 and mdx controls (n = 2) were harvested in parallel. Protein was harvested from TA muscle sections and analysed by LC-MS/MS. (B) Heatmap showing protein-clustered protein expression data. Protein expression data were filtered based on inter-replicate variation, leaving 119 proteins. Treatment with Pip6e-PMO resulted in a shift in the dystrophic proteome towards a wild-type-like expression profile. Two clusters show clear proteomic restoration. (C) Restored proteins that were down-regulated in mdx muscle. Dmd and the DAPC component Nos1 are highlighted in red. (D) Restored proteins that were up-regulated in mdx muscle. Red and blue indicate up- and down-regulated protein expression, respectively. Green and light blue indicate intermediate protein expression levels. The scale bar represents the row Z-score. (E) Determination of therapeutic efficacy in individual-treated samples by western blot, LC-MS/MS and RT-qPCR. Values are mean + SEM.

Figure 3.

Therapeutic restoration of DAPC components. (A) A cluster of significantly differently expressed proteins (labelled † in Supplementary Material, Fig. S7) includes six proteins from the DAPC that shows restoration following therapeutic dystrophin re-expression. Red and blue indicate up- and down-regulated protein expression, respectively. Green indicates an intermediate protein expression level. The scale bar represents row Z-scores. (B) Plots of individual log₂ expression ratios for each protein of interest. Values are mean +/− SEM.

Figure 4.

Matched mRNA transcriptomics and proteomics data analysis. RNA extracted from C57, mdx and Pip6e-PMO-treated mdx TA muscles was analysed by microarray. (A) Differentially expressed genes were visualized by volcano plot and filtered so as to include only statistically significantly changed genes (P < 0.01, t-test) that were changed by more than 1.5-fold in mdx muscle. (B) Heatmap showing the results of hierarchical clustering analysis of filtered data. Two clusters of mRNAs that were restored towards wild-type levels following treatment with Pip6e-PMO are labelled ‘†’ and ‘‡’, respectively. Red indicates up-regulated proteins and blue indicates down-regulated proteins. The scale bar represents row Z-scores. (C) PCA of filtered data (the first two components represent 80.5% of the data). (D) Venn diagram showing overlap between detected mRNAs and proteins. (E) Scatter plot showing the correlation between protein and mRNA expression ratios (2456 mRNA–protein pairs). (F) mRNA and protein expression ratios for genes that were concordantly differentially expressed in mdx muscle. Genes shown were statistically significant at the P < 0.01 level and were differentially expressed by at least 2-fold by either microarray or LC-MS/MS. Values are mean + SEM.

Figure 5.

Analysis of the dystrophic microRNA transcriptome. RNA extracted from C57, mdx and Pip6e-PMO-treated mdx TA muscles was analysed by miRNA microarray. (A) Differentially expressed genes were visualized by volcano plot and filtered so as to include only statistically significantly changed genes (P < 0.05, t-test) that were changed by more than 3-fold in mdx muscle. (B) Heatmap showing the results of hierarchical clustering (gene-clustered) analysis of filtered data. Red indicates up-regulated miRNAs and blue indicates down-regulated miRNAs. The scale bar represents the row Z-score. (C) PCA of filtered data (the first two components represent 92.6% of the data). (D) miRNA microarray data for specific miRNAs-of-interest. Individual log₂ expression ratios for each replicate are shown. Mean +/− SEM values are indicated, **P < 0.01, ***P < 0.001, one-way ANOVA and Bonferroni post hoc test.

Figure 6.

Identification of miRNA–mRNA target interactions. Network of miRNA–mRNA interactions identified by IPA for (A) dystromiRs up-regulated in mdx muscle (miR-21, miR-31, miR-34c, miR-146b and miR-206), and (B) the dystromiR miR-29c, which is down-regulated in mdx muscle. Red indicates increased expression in dystrophic muscle and green indicates reduced expression.