Determination of qPCR reference genes suitable for normalizing gene expression in a novel model of Duchenne muscular...

Duchenne muscular dystrophy (DMD) is a X-linked neuromuscular disorder arising from mutations in the dystrophin gene, leading to a progressive muscle wasting and disability. Currently there is no universal therapy, and there is thus a strong interest in preclinical studies for finding novel treatments. The most widely used and characterized mouse model for DMD is the C57BL/10ScSn-Dmdmdx/J (BL10-mdx), but this model exhibits mild pathology and does not replicate key features of human disease. The D2.B10-Dmdmdx/J (D2-mdx) mouse is a more recent model which seems to better mimics the complex human DMD phenotype. However, the D2-mdx mouse remains less extensively characterised than its BL10-mdx counterpart. Quantitative PCR analysis of gene expression is an important tool to monitor disease progression and evaluate therapeutic efficacy, but measurements must be normalised to stably expressed reference genes, which should ideally be determined and validated empirically. We examined gene expression in the gastrocnemius (GC), diaphragm (DIA) and heart in the D2-mdx mouse, the BL10-mdx mouse, and appropriate strain-matched wild-type controls (D2-wt and BL10-wt), from 4 to 52 weeks of age, using a large panel of candidate references (ACTB, AP3D1, CSNK2A2, GAPDH, HPRT1, PAK1IP1, RPL13A, SDHA, and in the heart, also HTATSF1 and HMBS). Data was analyzed using GeNorm, Bestkeeper, deltaCt and Normfinder algorithms to identify stable references under multiple possible scenarios. We show that CSNK2A2, AP3D1 and ACTB represent strong universal reference genes in both GC and DIA, regardless of age, muscle type, strain and genotype, while HTATSF1 and SDHA are optimal for the heart. GAPDH, HPRT1 and RPL13A were conversely revealed to be poor references, showing tissue-, age- or disease-specific changes in expression. Our results illustrate the importance of determining appropriate reference genes for specific comparative scenarios, but also reconfirm that universal panels can nevertheless be identified for normalising gene expression studies in even complex pathological states.