Comparative analysis of EV isolation methods for mRNA profiling using the nCounter platform in breast cancer

Extracellular vesicle-derived mRNAs (EV-mRNAs) are emerging analytes for liquid biopsy research, but their reliable detection is limited by low RNA abundance and co-isolated nucleic acids. We compared five EV isolation approaches to identify a suitable workflow for EV-mRNA profiling using the NanoString nCounter platform in breast cancer. EV preparations from pooled plasma were evaluated for particle yield, RNA recovery, and transcript detection using the nCounter Breast Cancer 360 panel with pre-amplification. RNase A pretreatment and post-extraction DNase I treatment were used to assess extra-vesicular RNA and residual DNA contributions. DNase I treatment markedly reduced total counts and detected transcripts, highlighting the importance of minimizing residual DNA-derived signals in pre-amplified EV-mRNA analysis. Among the tested approaches, miRCURY (MIR) retained the highest number of detected transcripts after nuclease treatment and was selected for downstream profiling. This workflow was applied to plasma samples from healthy donors ( n = 36), breast cancer patients without recurrence ( n = 36), and breast cancer patients with recurrence ( n = 31), followed by RT-qPCR assessment in an independent cohort. CCL5 and HIST1H3H were identified as exploratory breast cancer-associated EV-mRNA candidates, with CCL5 also showing a recurrence-associated pattern. These findings support MIR-based isolation with DNase I treatment as a practical workflow for NanoString-based EV-mRNA profiling in clinical plasma samples.