The aim of functional genomics is to understand the role of specific genes in phenotypic variation. The forward genetics
approach has led to a large number of identified genomic regions, known as quantitative trait loci (QTL), influencing various phenotypes, including those for muscle weight [1
]. However, a bottle neck has developed in the transition from QTL to their causative quantitative trait genes (QTG) [5
]. Advanced intercross line strategy permits accumulation of recombinations and improves resolution of QTL mapping [6
], which in the case of muscle weight has led to major reduction in confidence intervals [1
]. Although appreciably refined, these QTL still harbour a number of genes. Thus, further efforts are needed to identify the QTGs that are the causative factors in complex traits.
It has been proposed that testing for the expression differences could identify genes underlying phenotypic differences [7
]. Implementation of such strategy led to several nominations of QTG’s [8
]. However, microarray technology, used as a tool for a high throughput expression analyses, has several limitations which might have interfered with a more productive contribution of this approach to the nomination of the candidate genes. Hybridization artefacts caused by SNP’s [10
], non linearity among probes, inability to detect splice variants and, importantly, the bias towards available information (i.e. only transcripts with corresponding probes on microarray can be examined), limit the utility of expression microarrays. Transcriptome analyses by means of a massive parallel sequencing technology, RNA-Seq, circumvents the above-mentioned limitations [11
], it is highly replicable [13
] and therefore a very attractive research method for an unbiased identification of differentially expressed genes.
Our QTL studies focused on muscle size, which is an important variable influencing health and quality of life particularly in the elderly which are affected by sarcopenia, age-related muscle wasting [14
]. In addition, skeletal muscle tissue is a major component of diet and a source of nutrients for the growing population of the planet. Genetic variation plays a substantial role in determining muscle size in mammals [1
] but the underlying genes remain largely unknown. Muscle mass is a function of the number and size of its fibers. The number of fibers in mouse is determined prenatally and remains stable throughout adulthood [17
], whereas cross sectional area (CSA) of the fibres increases during post-natal development [18
The LG/J and SM/J strains, which were selected for large and small body weight, respectively, in order to study processes related to growth [19
], is a promising model system for exploration of the genetic effects on muscle mass. These strains differ prominently in mass of several hind limb muscles (2-fold difference between them) and 22 QTL contributing to this difference were mapped [1
]. Subsequent analyses of the soleus muscle found that the number of fibres in the muscle of the two strains was similar, whereas CSA differed substantially, LG/J > SM/J [21
The phenotypic differences due to genetic variation are determined by the pattern of information flow through molecular networks [22
]. A mouse muscle Bayesian Network (MMBN) has been recently constructed based on genetic and gene expression data. The MMBN provides directional information about the relationship of gene expression and can be used as a tool for identification of the key drivers of gene expression signatures characterising various phenotypes [23
The main finding of the present study is that the integration of the gene expression signature with the QTL analysis and muscle gene network data can lead to the identification of plausible QTGs underlying phenotypic differences in muscle mass.