Balancing yield, purity and practicality: a modified differential ultracentrifugation protocol for efficient isolation of small extracellular vesicles from human serum

• Published in: RNA biology 2019-01, Vol.16 (1), p.5-12
• Main Authors: Langevin, Scott M, Kuhnell, Damaris, Orr-Asman, Melissa A, Biesiada, Jacek, Zhang, Xiang, Medvedovic, Mario, Thomas, Hala Elnakat
• Format: Article
• Language: English
• Subjects: Biomarkers; Centrifugation, Density Gradient; Exosomes - metabolism; Exosomes - ultrastructure; Extracellular Vesicles - metabolism; Extracellular Vesicles - ultrastructure; Female; Humans; Male; MicroRNAs; RNA, Untranslated; Serum - chemistry; Technical Paper; Ultracentrifugation - methods
• ISSN: 1547-6286; 1555-8584
• DOI: 10.1080/15476286.2018.1564465

Ultracentrifugation remains the gold standard for isolation of small extracellular vesicles (sEV), particularly for cancer applications. The objective of this study was to determine if a widely used ultracentrifugation protocol for isolation of serum sEV could be modified to reduce the number of ultracentrifugation cycles and increase efficiency, while maintaining equal or better sample purity and yield. Serum was obtained from two healthy subjects. sEVs were isolated from 1 mL aliquots using three different ultracentrifugation protocols. Co-isolation of RNA carrier protein was assessed by performing Western blots for ApoA-I, ApoB, and Ago2. Small RNA-sequencing was performed on the sEV isolates, and differential detection of small ncRNA was compared across isolation protocols. Reduction from three- to two-ultracentrifuge cycles with no sucrose cushion resulted in a much higher sEV yield but also had the highest levels of lipoprotein and Ago2 contamination. However, the two-ultracentrifugation cycle protocol that incorporated a 30% sucrose cushion into the first cycle resulted in slightly higher sEV yields with lower levels of protein contamination compared to the lengthier three-ultracentrifugation cycle approach, therefore presenting a more efficient alternative approach for isolation of serum sEVs. It was also notable that there were some differences in sEV ncRNA cargo according to protocol, although it was less than expected given the differences in co-isolated RNA carrier proteins. Our results suggest that use of the modified serum sEV isolation protocol with two ultracentrifugation cycles and incorporating a 30% sucrose cushion offers a more efficient approach in terms of efficiency and purity.