Small RNA-Seq of crude and purified extracellular vesicles from venous and arterial human sera

Circulating extracellular vesicles (EVs) play central physiological and pathophysiological roles in intercellular communication. Biomarker studies addressing common disorders such as cancer or cardiovascular diseases often focus on microRNAs (miRNAs) found in EVs. Depending on the type of disease and clinic routine, patient specimens may be either sampled from arterial or venous sites, which makes comparisons between studies difficult. Thus, it is essential to test whether circulating cell-free miRNA profiles depend on the respective sampling site. In this study, we assessed potential differences in arterial and venous miRNA profiles in a cohort of 20 patients scheduled for cardiac surgery. Prior to surgery, blood was simultaneously sampled from the radial artery and from the internal jugular vein, and crude EVs were precipitated from both sera. We then performed expression profiling by small RNA sequencing (RNA-Seq) and compared arterial and venous miRNA profiles. Using stringent filtering criteria for differential expression analysis, no significantly regulated transcripts were observed. Filtering with less stringent criteria, we detected four miRNAs slightly upregulated in arterial samples, one of which could be validated by reverse transcription real-time PCR. The applicability of these findings generated in crude EVs to purified EVs was subsequently tested in a subset of the initial study population. EVs were enriched from arterial and venous sera by precipitation and further purified using size-exclusion chromatography prior to small RNA-Seq. While the additional clean-up step reduced overall miRNA yield compared to crude EV samples, no transcripts with differential arteriovenous expression were detected. Particle characterization of crude preparations as well as characterization of EV markers in purified EVs resulted in highly similar characteristics for arterial and venous samples. Arterial vs. venous blood sampling should therefore not represent a likely confounder when studying differentially expressed circulating miRNAs and results from studies using arterial or venous sampling sites are probably comparable. The use of either arterial or venous serum EV samples should result in highly similar data on miRNA expression profiles for the majority of biomarker studies.