Detection of canonical A-to-G editing events at 3' UTRs and microRNA target sites in human lungs using next-generation sequencing

Abstract

RNA editing is a post-transcriptional modification of RNA. The majority of these changes result from adenosine deaminase acting on RNA (ADARs) catalyzing the conversion of adenosine residues to inosine in double-stranded RNAs (dsRNAs). Massively parallel sequencing has enabled the identification of RNA editing sites in human transcriptomes. In this study, we sequenced DNA and RNA from human lungs and identified RNA editing sites with high confidence via a computational pipeline utilizing stringent analysis thresholds. We identified a total of 3,447 editing sites that overlapped in three human lung samples, and with 50% of these sites having canonical A-to-G base changes. Approximately 27% of the edited sites overlapped with Alu repeats, and showed A-to-G clustering (>3 clusters in 100 bp). The majority of edited sites mapped to either 3' untranslated regions (UTRs) or introns close to splice sites; whereas, only few sites were in exons resulting in non-synonymous amino acid changes. Interestingly, we identified 652 A-to-G editing events in the 3' UTR of 205 target genes that mapped to 932 potential miRNA target binding sites. Several of these miRNA edited sites were validated in silico. Additionally, we validated several A-to-G edited sites by Sanger sequencing. Altogether, our study suggests a role for RNA editing in miRNA-mediated gene regulation and splicing in human lungs. In this study, we have generated a RNA editome of human lung tissue that can be compared with other RNA editomes across different lung tissues to delineate a role for RNA editing in normal and diseased states.

Keywords: 3’UTRs; RNA editing; RNA-seq; exome; microRNAs.

Figure 1. Identification of RNA editing sites in normal lung samples

A. Both canonical (A-to-G) and all other types of non-canonical events that overlapped between all three normal lung samples are shown. B. Genomic distribution of all overlapping RNA editing types in the three normal human lung samples. C. Venn diagram showing the overlap between edited sites identified in our analysis and DARNED.

Figure 2. Validation of A-to-G editing by Sanger sequencing

A. Integrative Genome view of Ctsc, RhoA and Adam19 is shown. The RNA- and DNA-seq traces are shown in top and bottom panel, respectively. B. Validation of editing sites in Ctsc, RhoA and Adam19 is shown. For Ctsc, three editing sites were validated (chr11:88,055,689, 88,055,690, 88,055,691). For RhoA, we validated one site (chr3:49,397,323). For Adam19, we validated five sites chr5: (156,905,567, 156,905,566, 156,905, 565 in the left panel and 156,905,561, and 156,905,560) in the right panel. The sites validated by Sanger are shown and the nucleotide changes are labelled. Samples were sequenced using both forward and reverse M13 primers. We show few representative images of each.

Figure 3. Identification of RNA editing in miRNA target binding sites in the 3′ UTR of genes

A. Genes containing A-to-G edited miRNA targets for all chromosomes are shown. On the x-axis, the chromosome number, total number of editing events, and total number of miRNA targets associated with it are listed. On the y-axis, the number of genes associated with editing events in the miRNA binding sites is depicted. B.-C. Validation of editing in potential miRNA binding sites by in silico analysis for Rbbp4 B. and Spcs3 C. The images show miRNA target sites and the complementary hg19 sequence. The edited site in the seed sequence of the mature miRNA is highlighted in bold. The hsa-miR-6849-3p binding site was located in Rbbp4 3′ UTR at chr1:33,149,343-586. The bold position is an A-to-G RNA-editing event located at chr1:33,149,443 in Rbbp4 (B). Similarly the hsa-miR-1303 binding site is located in Spcs3 3′ UTR at chr4:177,252,226-427. The bold positions are two A-to-G edited sites located at chr4:177,252,326 and 327 (C)..