An estimate of the total number of true human miRNAs

Abstract

While the number of human miRNA candidates continuously increases, only a few of them are completely characterized and experimentally validated. Toward determining the total number of true miRNAs, we employed a combined in silico high- and experimental low-throughput validation strategy. We collected 28 866 human small RNA sequencing data sets containing 363.7 billion sequencing reads and excluded falsely annotated and low quality data. Our high-throughput analysis identified 65% of 24 127 mature miRNA candidates as likely false-positives. Using northern blotting, we experimentally validated miRBase entries and novel miRNA candidates. By exogenous overexpression of 108 precursors that encode 205 mature miRNAs, we confirmed 68.5% of the miRBase entries with the confirmation rate going up to 94.4% for the high-confidence entries and 18.3% of the novel miRNA candidates. Analyzing endogenous miRNAs, we verified the expression of 8 miRNAs in 12 different human cell lines. In total, we extrapolated 2300 true human mature miRNAs, 1115 of which are currently annotated in miRBase V22. The experimentally validated miRNAs will contribute to revising targetomes hypothesized by utilizing falsely annotated miRNAs.

Figure 1.

Workflow of the analysis to estimate the number of true human miRNAs. Samples containing NGS data were collected from SRA, TCGA and data uploaded to miRMaster, and the obtained samples and reads were filtered. Three sets of miRNAs were created: miRBase high-confidence (HC), miRBase low-confidence set (LC) and other (Other). Using in silico high-throughput validation and experimental low-throughput validation steps, the probabilities that a miRNA will pass the validation procedure have been calculated for each miRNA set respectively. Finally, the number of true miRNAs was estimated using the original miRNA counts and the computed probabilities.

Figure 2.

Northern blots of endogenous miRNAs. (A) The 11 analyzed cell lines that are indicated on top of the figure were derived from lung (A549), liver (HUH-7), bone marrow (SHSY-5Y), keratinocytes (HaCaT), cervix (HeLa), mammary gland (MCF7), placenta (JEG-3), Testis (Tera-1), prostate (PC-3), B-lymphocytes (DG-75) and T-lymphocytes (Jurkat). HEK-293T RNA was used as a reference to compare signals to exogenously expressed miRNAs. The endogenous mature forms are shown for the miRNAs indicated on the left side of the figure. (B) The number of mature and premature forms of the endogenous miRNA expressed in the 12 cells lines as indicated in Figure 2A.

Figure 3.

Representative NB results for a positive, questionable and negative miRNA in HEK 293T cells. (A) NB for hsa-miR-155–5p from high-confidence set A showing distinct bands for its precursor (p) and mature (m) form. (B) Hybridization against hsa-miR-1260a (low-confidence set B) detects two small RNA fragments with similar signal intensities for the control. (C) Probing for hsa-miR-6776–5p (low-confidence set B) did not result in any specific bands. Ethidium bromide staining of RNA gels was used as a loading control.

Figure 4.

Comparison of exogenous and endogenous miRNA expression by northern blotting. The mature form is indicated by ‘m’ and the premature form by ‘p’. The left part of the figure shows exogenous expression of miR-148a-3p in HEK 293T cells. HEK 293T cells that were transfected with an empty vector are shown as a control (ctrl). The right part of the figure shows endogenous expression of miR-148a-3p in 12 cells lines as specified in Figure 2A. To show endogenous miRNAs, the signal intensity of the endogenous miRNAs apparently is enhanced as compared to the signal intensity of the exogenous miRNAs due to the very strong signal of overexpressed miR-148a-3p (compare backgrounds of exo- and endogenous northern blots and see the comparison between HEK 293T cells that were used as a control for the transfection analysis (ctrl) shown in the left part of the figure and the HEK 293T cells that were compared with other cells shown in the right part of the figure). As described in the ‘Materials and Methods’ section, contrast and brightness were adjusted by the software during scanning according to the darkest spot on the blot.