Systematic selection of reference genes for normalization of circulating RNA transcripts in pregnant women based on RNA-seq data

RNA transcripts circulating in peripheral blood represent an important source of non-invasive biomarkers. To accurately quantify the levels of a circulating transcript, one needs to normalize the data with internal control reference genes, which are detected at relatively constant levels across different blood samples. A few stably-expressed reference gene candidates have to be selected from transcriptome data before validation of their stable expression by reverse-transcription quantitative polymerase chain reaction. However, there is a lack of transcriptome, let alone whole-transcriptome, data from maternal blood. To overcome this shortfall, we performed RNA-seq on blood samples from women presented with preterm labor. Of 11215 exons detected in the maternal blood whole-transcriptome, we systematically identified a panel of 395 genes comprising exons that were detected at a coefficient of variation (CV) ranging from 7.75%-17.7%. Their levels were considerably less variable than any GAPDH exon (minimum CV, 27.3%). Upon validation, selected genes from this panel remained as more stably expressed than GAPDH in maternal blood. This panel is over-represented with genes involved with actin cytoskeleton, macromolecular complex and the integrin signaling pathway. This groundwork provides a starting point for systematically selecting reference gene candidates for normalizing the levels of circulating RNA transcripts in maternal blood.