Sequencing of salivary microRNA (miRNA) is rapidly becoming a means to explore disease biomarkers. In this study we demonstrate that stabilized saliva samples provide RNA of sufficient quality for detection of miRNA biomarkers, including miRNAs with reported links to cancer, head injury and mental health disorders.
Saliva was collected from 14 participants using the ORAcollect•RNA device (DNA Genotek). RNA was extracted using the miRNeasy Serum and Plasma kit (QIAGEN) and sequencing libraries were generated with the QIAseq miRNA Library Kit with UDIs (QIAGEN) without normalization of the input RNA. Libraries were sequenced on the NextSeq 550 (20 million reads per sample). Reads were annotated using GeneGlobe Design & Analysis Hub (miRbase and piRNAdb).
The mean number of reads per sample was 20.7 million and Q scores of Unique Molecular Index (UMI) reads were above 30 demonstrating a high sequencing performance in all libraries. Of the reads containing UMIs, a range of 10-51% of the reads in each sample were annotated with miRBase or piRNAdb. The range of annotated reads in saliva samples is expected given the variable miRNA input that is inherent to saliva samples. Despite the variable input, saliva covered 35-50% of the total number of annotation records in miRbase.
Recently a retrospective cross-sectional miRNA sequencing study utilizing 1225 saliva samples was published and gives greater insight into the expected expression of salivary miRNA’s from multiple cohorts (Sullivan R. et al., 2022). We compared a cohort of expression profiles from the Sullivan study to our data to discern concordance between the two. The most highly-expressed miRNAs in each study were distinctive, though there was a 50% concordance between the top 20 most-expressed miRNAs in the two studies.
These results demonstrate that saliva is a suitable sample type for miRNA biomarker detection using small RNA sequencing.