Do two mutually exclusive gene modules define the phenotypic diversity of mammalian smooth muscle?

Smooth muscle cells (SMCs) are key components of all hollow organs, where they perform contractile, synthetic and other functions. Unlike other muscle cells, SMCs are not terminally differentiated, but exhibit considerable phenotypic variation. Such variation is manifested both across disease states such as asthma and atherosclerosis, and physiological states such as pregnancy and wound healing. While there has been considerable investigation into the diversity of SMCs at the level of morphology and individual biomarkers, less is known about the diversity of SMCs at the level of the transcriptome. To explore this question, we performed an extensive statistical analysis that integrates 200 transcriptional profiles obtained in different SMC phenotypes and reference tissues. Our results point towards a non-trivial hypothesis: that transcriptional variation in different SMC phenotypes is characterized by coordinated differential expression of two mutually exclusive (anti-correlating) gene modules. The first of these modules (C) encodes 19 co-transcribed cell cycle associated genes, whereas the other module (E) encodes 41 co-transcribed extra-cellular matrix components. We propose that the positioning of smooth muscle cells along the C/E axis constitutes an important determinant of SMC phenotypes. In conclusion, our study introduces a new approach to assess phenotypic variation in smooth muscle cells, and is relevant as an example of how integrative bioinformatics analysis can shed light on not only terminal differentiated states but also subtler details in phenotypic variability. It also raises the broader question whether coordinated expression of gene modules is a common mechanism underlying phenotypic variability in mammalian cells.