Comparison of RNA isolation and library preparation methods for small RNA sequencing of canine biofluids

Small RNA sequencing (RNA-seq) of biofluids is challenging due to the relative scarcity of microRNAs (miRNAs), limited sample volumes, and the lack of a gold standard isolation method. Additionally, few comparisons exist for the RNA isolation and sequencing methods of biofluids.

We aimed to compare the performance of six commercial RNA isolation kits and two library preparation methods for small RNA-seq using canine serum and urine.

Serum and urine were collected from seven dogs with protein-losing nephropathy, and the samples were pooled. Total RNA from serum (2 mL) and urine (10 mL) was isolated in triplicate using three methods each for serum (Zymo Direct-zol, mirVana PARIS, miRCURY Biofluids) and urine (Qiagen exoRNeasy, Norgen Urine Exosome, miRCURY Exosome). For each sample type, the two kits yielding the highest RNA concentration were selected, and small RNA-seq was performed using TruSeq and NEXTflex library preparations. Data were analyzed by CPSS 2.0 and DESeq2.

For serum, Zymo Direct-zol combined with NEXTflex was the only combination that enabled successful library preparation, while for urine, Qiagen exoRNeasy combined with NEXTflex outperformed other combinations for detecting miRNAs. The total number of miRNAs detected in serum and urine was 198 and up to 115, respectively. miRNA expression in serum was distinct from urine. Furthermore, the library preparation method introduced a higher variation of urine results than the RNA isolation method.

Different isolation and library preparation methods show significant differences in miRNA results that could affect biomarker discovery. Small RNA-seq provides an unbiased, global assessment to compare these methods in canine biofluids.