A novel biomarker of fibrofatty replacement in dystrophinopathies identified by integrating transcriptome, magnetic resonance imaging, and pathology data

Abstract

Abstract Background We aimed to analyse genome‐wide transcriptome differences between Duchenne muscular dystrophy (DMD) and Becker muscular dystrophy (BMD) patients and identify biomarkers that correlate well with muscle magnetic resonance imaging (MRI) and histological fibrofatty replacement in both patients, which have not been reported. Methods One hundred and one male patients with dystrophinopathies (55 DMD and 46 BMD) were enrolled. Muscle‐derived genome‐wide RNA‐sequencing was performed in 31 DMD patients, 29 BMD patients, and 11 normal controls. Fibrofatty replacement was scored on muscle MRI and histological levels in all patients. A unique pipeline, single‐sample gene set enrichment analysis combined with Spearman's rank correlations (ssGSEA‐Cor) was developed to identify the most correlated gene signature for fibrofatty replacement. Quantitative real‐time PCR (qRT‐PCR) analysis, western blot analysis, and single‐nucleus RNA‐sequencing (snRNA‐seq) were performed in the remaining patients to validate the most correlated gene signature. Results Comparative transcriptomic analysis revealed that 31 DMD muscles were characterized by a significant increase of inflammation/immune response and extracellular matrix remodelling compared with 29 BMD muscles (P < 0.05). The ssGSEA‐Cor pipeline revealed that the gene set of CDKN2A and CDKN2B was the most correlated gene signature for fibrofatty replacement (histological rs = 0.744, P < 0.001; MRI rs = 0.718, P < 0.001). Muscle qRT‐PCR confirmed that CDKN2A mRNA expression in both 15 DMD (median = 25.007, P < 0.001) and 12 BMD (median = 5.654, P < 0.001) patients were significantly higher than that in controls (median = 1.101), while no significant difference in CDKN2B mRNA expression was found among DMD, BMD, and control groups. In the 27 patients, muscle CDKN2A mRNA expression respectively correlated with muscle MRI (rs = 0.883, P < 0.001) and histological fibrofatty replacement (rs = 0.804, P < 0.001) and disease duration (rs = 0.645, P < 0.001) and North Star Ambulatory Assessment total scores (rs = −0.698, P < 0.001). Muscle western blot analysis confirmed that both four DMD (median = 2.958, P < 0.05) and four BMD (median = 1.959, P < 0.01) patients had a significantly higher level of CDKN2A protein expression than controls (median = 1.068). The snRNA‐seq analysis of two DMD muscles revealed that CDKN2A was mainly expressed in fibro‐adipogenic progenitors, satellite cells, and myoblasts. Conclusions We identify CDKN2A expression as a novel biomarker of fibrofatty replacement, which might be a new target for antifibrotic therapy in dystrophinopathies.

Keywords

• Becker muscular dystrophy
• Duchenne muscular dystrophy
• Fibrofatty replacement
• Transcriptome