Extracellular small non-coding RNA contaminants in fetal bovine serum and serum-free media

In the research field of extracellular vesicles (EVs), the use of fetal bovine serum (FBS) depleted of EVs for in vitro studies is advocated to eliminate the confounding effects of media derived EVs. EV-depleted FBS may either be prepared by ultracentrifugation or purchased commercially. Nevertheless, these preparations do not guarantee an RNA-free FBS for in vitro use. In this study we address the RNA contamination issue, of small non-coding (nc)RNA in vesicular or non-vesicular fractions of FBS, ultracentrifugation EV-depleted FBS, commercial EV-depleted FBS, and in our recently developed filtration based EV-depleted FBS. Commercially available serum- and xeno-free defined media were also screened for small ncRNA contamination. Our small ncRNA sequencing data showed that all EV-depleted media and commercially available defined media contained small ncRNA contaminants. Out of the different FBS preparations studied, our ultrafiltration-based method for EV depletion performed the best in depleting miRNAs. Certain miRNAs such miR-122 and miR-203a proved difficult to remove completely and were found in all media. Compared to miRNAs, other small ncRNA (snRNA, Y RNA, snoRNA, and piRNA) were difficult to eliminate from all the studied media. Additionally, our tested defined media contained miRNAs and other small ncRNAs, albeit at a much lower level than in serum preparations. Our study showed that no media is free of small ncRNA contaminants. Therefore, in order to screen for baseline RNA contamination in culturing media, RNA sequencing data should be carefully controlled by adding a media sample as a control. This should be a mandatory step before performing cell culture experiments in order to eliminate the confounding effects of media.