Circular RNAs are the predominant transcript isoform from hundreds of human genes in diverse cell types.
Salzman J, Gawad C, Wang PL, Lacayo N, Brown PO. · 2012
Read Circular RNAs are the predominant transcript isoform from hundreds of human genes in diverse cell types. by Salzman J, Gawad C, Wang PL, Lacayo N, Brown PO. (2012) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc3270023.txt.
Introduction
Julia Salzman and colleagues reported that circular RNAs are common products of human genes, in this 2012 paper in PLoS ONE. Textbook splicing joins a gene's exons in their genomic order to make a straight molecule. Deep sequencing of RNA from normal and cancerous human cells kept turning up transcripts whose exons appeared shuffled.
Statistical estimates and biochemical tests gave strong evidence that a substantial share of the spliced output from hundreds of genes is circular, with the RNA's ends joined into a loop. In many cases the circular form was the predominant version of the transcript. The authors conclude this unusual splicing is not an oddity. It is a general feature of how human genes are expressed. That opens questions about what these stable circles do in the cell.
Catalog record
| CREATOR | Salzman J, Gawad C, Wang PL, Lacayo N, Brown PO. |
|---|---|
| YEAR | 2012 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 59 KB |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Circular RNAs Are the Predominant Transcript Isoform from Hundreds of Human Genes in Diverse Cell Types By Julia Salzman, Charles Gawad, Peter Lincoln Wang, Norman Lacayo, Patrick O. Brown, Thomas Preiss Published in: PLoS ONE DOI: 10.1371/journal.pone.0030733 ABSTRACT Most human pre-mRNAs are spliced into linear molecules that retain the exon order defined by the genomic sequence. By deep sequencing of RNA from a variety of normal and malignant human cells, we found RNA transcripts from many human genes in which the exons were arranged in a non-canonical order. Statistical estimates and biochemical assays provided strong evidence that a substantial fraction of the spliced transcripts from hundreds of genes are circular RNAs. Our results suggest that a non-canonical mode of RNA splicing, resulting in a circular RNA isoform, is a general feature of the gene expression program in human cells. INTRODUCTION Deep sequencing of RNA from biological samples, “RNA-Seq”, is a powerful tool for discovering and cataloguing novel alterations in the expression, sequence, and structure of transcriptomes. In the present study, we used RNA-Seq in a deliberate search for transcripts that could not be accounted for by conventional splicing of primary transcripts from an unrearranged human genome. Although our initial goal was to discover cancer-specific chromosomal rearrangements by identifying the resulting fused or rearranged transcripts, we also investigated the possibility that the exon order specified by the genome sequence might be rearranged during RNA processing. We were surprised to find numerous examples of transcripts in which the exon order was a circular permutation of the order encoded in the genome. We hypothesized that these anomalous transcripts might the result of intramolecular but non-canonical splicing events that joined a splice donor to an upstream (i.e. toward the 5′ end of the transcript) splice acceptor to produce a circular RNA molecule. Indeed, for many genes, in both cancer and normal human cells, we found RNAs with circularly permuted exon orders at levels comparable to those of the canonical, linear mRNA. The first observation suggesting that eukaryotic RNAs can exist in circular form was made more than 30 years ago by electron microscopy [1]. 10 years later, human cytoplasmic RNA was reported to contain very low levels of transcripts of the DCC gene with exons spliced in non-canonical order (i.e. shuffled relative to the reference genome). These scrambled transcripts were estimated to comprise less than one one-thousandth of DCC transcripts, and the phenomenon was dubbed exon-scrambling [2]. Since that time, a handful of expressed mammalian genes have been shown to express circular RNA isoforms at low levels [3], [4]. Such examples include very low levels of human RNA transcripts with scrambled exons observed in several human genes, including MLL and ETS-1 [13], [14]. The best-characterized circular transcripts are in rodents. The mouse SRY gene, the sex-determining gene in males, consists of a single exon. During development, the RNA exists as a linear transcript that is translated into protein. In the adult testes, the RNA exists primarily as a circular product that is predominantly localized to the cytoplasm and is apparently not translated [5], [6]. Studies have demonstrated that inverted repeats in the genomic sequence flanking the SRY exon direct transcript circularization [5], [7], [8]. The sodium transporter NCX1 and the rat cytochrome P450 2C24 gene are two other well-studied examples of mouse transcripts with circular isoforms that are expressed at relatively low levels [9]–[11]. The circular isoform of the NCX1 gene is thought to encode a protein, although this possibility has not been conclusively demonstrated. Examples of exon scrambling have also been found in Drosophila [12]. All examples of circular transcripts reported to date in humans have been found to be expressed at low levels compared to the dominant canonical linear isoform, requiring sensitive nested PCR experiments for detection; these examples were discovered inadvertently or in an effort to characterize the structure of oncogenes. Circular RNAs have also been reported to be rare isoforms of the human Cytochrome P-450 2C18, and dystrophin transcripts [11], [15]. Most recently, a circular isoform of the non-coding RNA ANRIL was found to be expressed at very low levels; its expression was correlated with INK4/ARF expression, and with risk for developing atherosclerosis [16]. This finding suggests that circular RNAs may have both biochemical and phenotypic consequences. Because of their apparent rarity, however, circular RNAs are generally viewed as minor RNA structural variants, perhaps attributable to transcriptional noise [17]. Two studies have suggested that scrambled exons may occur in more than a handful of transcripts. Roughly 150 sequences representing exon scrambling were found in EST databases; in each case, the authors argued that these sequences were unlikely to derive from circular RNA molecules, but rather represented bona fide examples of exon scrambling within a linear transcript [18], [19]. The lack of direct experimental evidence makes interpretation of these studies difficult: EST database studies are fraught with confounding of RT-PCR artifacts and selection bias. If template switching occurs during reverse transcription, given a typical human exon length of 150 (the average of the lower 90th quantile), 1 out of 22500 template switches would be at exon-exon boundaries, and hundreds could be expected to be detected in an EST database. Indeed, it has been recently suggested that apparent examples of exon scrambling in human EST databases are primarily explained as RT artifacts. More recently, a single-read RNA-Seq study of poly-adenylated human RNA from a wide variety of human tissues identified 176 genes with 205 scrambled isoforms, achieving a 64% validation rate with RT-PCR, and showing that many such transcripts had high expression levels compared to the canonical transcripts [20]. These authors postulated that the RNA molecules containing scrambled exons were linear and poly-adenylated. RESULTS Hundreds of Putative Intragenic Rearrangements in Pediatric Acute Lymphoblastic Leukemia We recently reported evidence suggesting that local structural rearrangements in the human genome may be under-recognized genetic lesions contributing to the progression of human cancers [21]. To test this hypothesis, we explored whether intragenic rearrangements were prevalent in childhood cases of Acute Lymphoblastic Leukemia (ALL) by performing RNA-Seq on ribosomal RNA-depleted total RNA from the diagnostic bone marrow of 5 children between the ages of 2 and 6 with hyperdiploid B-precursor ALL. We searched for transcripts with scrambled exon order (i.e. where exon X is upstream of exon Y, but X≥Y) that might represent intragenic rearrangements by comparing paired-end RNA-seq data with a database of human transcriptome sequences (UCSC knowngene hg19). Reads that could not be aligned to the UCSC knowngene transcriptome were subsequently aligned to a custom database of exon-exon junctions comprising all possible pairs of exons annotated in hg19 RefSeq. A paired-end read is considered ‘junctional’ if read 1 aligns to the gene and read 2 aligns to an exon-exon junction with more than 10 mismatch-free bases flanking each side of the junction. The essential features of this analysis are described in the methods section and have been described in greater detail previously [21]. This analysis identified hundreds of genes with evidence of exon scrambling: in all, more than 1232 such genes had more than one junctional read. In addition, more than 700 isoforms with scrambled exons were estimated to comprise more than 10% of all transcript isoforms for a comparable number of genes. Genes with