The landscape of microRNA, Piwi-interacting RNA, and circular RNA in human saliva

Abstract

Background: Extracellular RNAs (exRNAs) in human body fluids are emerging as effective biomarkers for detection of diseases. Saliva, as the most accessible and noninvasive body fluid, has been shown to harbor exRNA biomarkers for several human diseases. However, the entire spectrum of exRNA from saliva has not been fully characterized.

Methods: Using high-throughput RNA sequencing (RNA-Seq), we conducted an in-depth bioinformatic analysis of noncoding RNAs (ncRNAs) in human cell-free saliva (CFS) from healthy individuals, with a focus on microRNAs (miRNAs), piwi-interacting RNAs (piRNAs), and circular RNAs (circRNAs).

Results: Our data demonstrated robust reproducibility of miRNA and piRNA profiles across individuals. Furthermore, individual variability of these salivary RNA species was highly similar to those in other body fluids or cellular samples, despite the direct exposure of saliva to environmental impacts. By comparative analysis of >90 RNA-Seq data sets of different origins, we observed that piRNAs were surprisingly abundant in CFS compared with other body fluid or intracellular samples, with expression levels in CFS comparable to those found in embryonic stem cells and skin cells. Conversely, miRNA expression profiles in CFS were highly similar to those in serum and cerebrospinal fluid. Using a customized bioinformatics method, we identified >400 circRNAs in CFS. These data represent the first global characterization and experimental validation of circRNAs in any type of extracellular body fluid.

Conclusions: Our study provides a comprehensive landscape of ncRNA species in human saliva that will facilitate further biomarker discoveries and lay a foundation for future studies related to ncRNAs in human saliva.

Fig. 1. miRNA expression in CFS

(A), Example length distribution of a small RNA sequencing library from CFS. Library adapters have been trimmed. The read lengths of the major peak (22 nt) and minor peak (29 nt) are illustrated, which correspond to the known lengths of miRNAs and piRNAs, respectively. (B), Scatterplot of miRNA expression (log₂ RPM) across 2 individuals. Pearson correlation coefficient is shown. (C), Experimental validation of miRNA expression in exosome fraction (E) and exosome-free fraction (NE) by use of ddPCR. For each miRNA, the ddPCR fluorescence intensity is shown for E and NE samples in each individual. Negative control (no template) was run together with the actual samples in the same batch of experiments, and fluorescent signal was barely detected. (D), Histogram of ISI for all expressed miRNAs calculated with the 6 independent saliva samples (biological replicates not included). The distributions of ISI values of other public data sets are also shown for comparison. X2, J1, D1, M1, and S1 are study-assigned IDs for samples. Ctrl, control.

Fig. 2. Comparison of miRNA expression across different cell types and body fluids

Heat map of correlation (Kendall τ) of miRNA expression levels derived from public and in-house small RNA sequencing data sets. A subset of the samples including our CFS samples is shown here because of space limits, with the entire figure shown as online Supplemental Fig. 3. Hierarchical clustering was applied. Samples were named by their type, the laboratory that generated the data (via an arbitrary numerical ID), and GEO IDs (if available). Raw sequencing data were analyzed in exactly the same way except for data from lab 17, for which expression data of miRNAs in serum and CSF were directly obtained from their publication due to lack of raw data. Saliva_lab_16 represents CFS miRNA data generated in this study. All the data sets derived from extracellular body fluids [CFS, serum, CSF, and plasma (exosome-associated RNA)] clustered together, with relatively smaller distances compared with their distances to intracellular RNA samples.

Fig. 3. piRNA expression in CFS

(A), Scatterplot of piR expression (log₂ RPM) across 2 individuals. Pearson correlation coefficient is shown. (B), Histogram of ISI for all expressed piRNAs calculated with the 6 independent saliva samples (biological replicates were not included). The distributions of ISI values of other public data sets are also shown for comparison. (C), Experimental validation of piRNA expression in exosome fraction (E) and exosome-free fraction (NE), similar to Fig. 1C. X2, J1, D1, M1, and S1 are study-assigned IDs for samples. Ctrl, control.

Fig. 4. Comparison of piRNA expression across different cell types and body fluids

Heat map of piRNA expression levels derived from public and in-house small RNA sequencing data sets. z Scores of expression levels were calculated for each piRNA. A subset of the samples including our CFS samples is shown here because of space limits, with the entire figure shown as online Supplemental Fig. 5. Hierarchical clustering was applied. Samples were named similarly as in Fig. 2. CFS, ES cells, and skin cells were among those having the highest expression levels of piRNAs.

Fig. 5. circRNA expression in CFS

Experimental validation of circRNAs in CFS was carried out by RT-PCR. Primers were designed to amplify the circular junctions, as illustrated. CFS samples were collected from the same individuals, but on different days, as for RNA-Seq. PAGE gel images are shown to visualize the presence of the PCR products. Representative Sanger sequencing traces are shown, together with the nucleotide sequences and genomic coordinates (hg19) of the circular junction. All Sanger-based sequences were identical to the predicted circular junctions via RNA-Seq. (A), Canonical circRNAs. Sanger sequencing was conducted on gel-purified PCR products. (B), Noncanonical circRNAs. Sanger sequencing was conducted on random clones amplified after TOPO cloning of the PCR products. DOPEY2, dopey family member 2; UBAP2, ubiquitin associated protein 2; GSE1, Gse1 coiled-coil protein; ASXL1, additional sex combs like transcriptional regulator 1; UGP2, UDP-glucose pyrophosphorylase 2; WDFY1, WD repeat and FYVE domain containing 1. D1, M1, S1, and C1 are study-assigned IDs for samples. M, marker.