Mining, validating, and quantifying circular RNA transcriptome from total RNA as a biomarker or target

Circular RNAs (circRNAs) participate in complex diseases and have emerged as a novel biomarker and therapeutic target. It is challenging to identify circRNAs from total RNA-seq data and to validate and quantify their levels by qRT-PCR. We sought to characterize the circRNA transcriptome from RNA-seq with total RNA from septic peripheral blood mononuclear cells (PBMCs), assess/compare commonly used aligners, and optimize validation methods, considering sepsis is the leading cause of death in intensive care units (ICU) and there is no effective treatment. PBMCs were isolated from sepsis patients before and after intensive care, and RNA sequencing was conducted with total RNA. RNA-seq data for circRNAs were analyzed using four pipelines. A circRNA transcriptome has been generated, and four alignment and annotation pipelines have been evaluated. TopHat was found to be the most sensitive aligner while MapSplice was found to be the most accurate. Overall, circRNA expression in septic PBMCs were found to be more abundant and less diverse at ICU admission (ICU-AD) compared with ICU discharge (ICU-DC). There were three differentially expressed circRNAs between ICU-AD and ICU-DC. The addition of reverse primers to a reverse transcription reaction improved reproducibility and accuracy of qRT-PCR. Among these altered circRNAs, two different isoforms of circular RNA from the ASPH gene (circASPH(2,4), and circASPH(2,3,4) with the same backsplicing junction (BSJ) were identified and specifically quantified by qRT-PCR. All the two isoforms were highly expressed at ICU-AD and circASPH(2,3,4) levels positively correlated with the length of stay in ICU. The circRNAs predictably bind to miR-670, miR-7975, miR-6818, and miR-384 and interact with proteins TLR2 and ZC3H12D. In conclusion, it is feasible to identify circRNA transcriptome from RNA-seq with total RNA. Moreover, RT-PCR followed by gel electrophoresis is important to identify/ distinguish a new isoform of a circRNA with the same BSJ. CircRNA transcriptome in septic PBMCs evolves as infectious severity changes. CircASPH(2,3,4) has a clinical potential as an indicator of disease severity.