In-Depth Analysis Reveals Production of Circular RNAs from Non-Coding Sequences

Abstract

The sequencing of total RNA depleted for ribosomal sequences remains the method of choice for the study of circRNAs. Our objective was to characterize non-canonical circRNAs, namely not originating from back splicing and circRNA produced by non-coding genes. To this end, we analyzed a dataset from porcine testis known to contain about 100 intron-derived circRNAs. Labelling reads containing a circular junction and originating from back splicing provided information on the very small contribution of long non-coding genes to the production of canonical circRNAs. Analyses of the other reads revealed two origins for non-canonical circRNAs: (1) Intronic sequences for lariat-derived intronic circRNAs and intron circles, (2) Mono-exonic genes (mostly non-coding) for either a new type of circRNA (including only part of the exon: sub-exonic circRNAs) or, even more rarely, mono-exonic canonical circRNAs. The most complex set of sub-exonic circRNAs was produced by RNase_MRP (ribozyme RNA). We specifically investigated the intronic circRNA of ATXN2L, which is probably an independently transcribed sisRNA (stable intronic sequence RNA). We may be witnessing the emergence of a new non-coding gene in the porcine genome. Our results are evidence that most non-canonical circRNAs originate from non-coding sequences.

Keywords: circular junction; exonic circRNA; intron; intron circle; intronic circRNA; intronic lariat; non-coding; pig testis; sisRNA; sub-exonic circRNA.

Figure A1

The clustering of 544,011 CCRs mapped on porcine autosomes SSC1 to SSC18 led to the characterization of 148,505 LACs. The pie chart shows the proportion of LAC supported by 1, 2, 3, 4 or 5 and more CCRs.

Figure A2

Eight thresholds including the possible consideration of distinct CCRs were tested (A,B). The seven first thresholds are presented in order of increasing stringency, and the last is the previously used threshold (‘5 CCRs’) [13]. The number of LACs obtained for each threshold is reported. As only 12 genes between 1 and 2 Mb in size have been reported in pig and the length of the longest human gene is 2.3 Mb, we considered that a LAC greater than 3 Mb could be suspicious, and their number is thus reported. (B) In each of the eight thresholds explored, we evaluated the proportion of LACs corresponding to the exact back splicing between (one or) two known exons (Ensembl coding genes).

Figure A3

CircRNA production was detected in this region by identifying two clusters of CCRs. The first (in blue) and the second (in purple) contain six and five CCRs, respectively. After clustering of CCRs by perfectly identical genomic coordinates, two loci associated with circRNA production (LAC) were observed, but with a threshold of ‘5 CCRs including 4 distinct CCRs’ only the LAC#2 indicated in purple was retained.

Figure 1

CircRNA production was detected in this region by identifying two clusters of circular chimeric reads (CCRs). The first (in blue) and the second (in purple) contain six and five CCRs, respectively. The two fragments of each CCR are mapped in inverted order on the chromosome. The clustering of CCRs by identical genomic coordinates leads to two loci associated with circRNA production (LAC). Two LACs were thus observed, even if their genomic coordinates differed by only one nucleotide. Contrary to most authors, we do not consider that each LAC defines a distinct circRNA. When we examined the sequence contents of CCRs from different LACs, we observed that several LACs can describe a single circularization event (a single circRNA).

Figure 2

Intron-derived circRNAs. There are two intron-containing genes on this pseudo chromosome. The two multi-exonic genes contain four exons (blue boxes and purple boxes respectively). The analysis of reads (drawn above the chromosome) mapped in this region identified several reads containing a circular junction (CCR). Clustering CCRs allowed us to define several LACs suspected of being associated with different circRNAs. Concerning the gene in blue, three LACs were identified, and their genomic coordinates appeared to be compatible with the production of three intronic circRNAs. Nevertheless, regarding the number of introns able to produce these circRNA, the number was only one. Concerning the gene in purple, one LAC was identified, and its genomic coordinates appear to be compatible with the production of an intron circle. To emphasize the fact that in intronic circRNAs, the circular junction is a 2′-5′ bond, we avoided representing them as a simple circle like intron circles.

Figure 3

The height of the bars represents the size of the 127 intron-derived circRNAs identified in the Testis-31 dataset. A total of 123 lariat-derived intronic circRNA (in blue) and 4 intron circles (in purple) were detected in the dataset. The circRNAs were divided into groups according to their size (x axis) and the number of circRNAs concerned is shown on the y axis.

Figure 4

introLCirc-200. (A) The intronic sequence concerned by production of this circRNA. This intron is framed by two exons of 213 bp (ENSSSCE00000205149) and 190 bp (ENSSSCE00000199966 or the 3′ UTR exon). (B) 10 LACs were characterized. When we examined the sequence of CCRs and their mate pairs, four distinct ‘junction-sequences’ (js#1-4) were identified at the circular junction (C,D). They are presented in decreasing order of frequency. For details, see Supplementary Doc. 3A1.

Figure 5

Sub-exonic circRNA. Clustering CCRs allowed us to define several LACs suspected of being associated with different circRNAs. On this pseudo chromosome, we show a mono-exonic gene able to produce several sets of sub-exonic circRNAs. Six LACs were identified, whose genomic coordinates appear to be compatible with the production of several sub-exonic circRNAs. When we examined the sequence contents of CCRs from different LACS, we saw that several LACs described a single circularization event (a single circRNA) or that one LAC can define two or three circRNAs.

Figure 6

Analyses of sequence at the circular junction of a set of sub-exonic circRNA produced by RNase_MRP. The set is defined by five LACs of 145 nt. When we examined the sequence of CCRs and their mate pairs, two distinct ‘junction-sequences’ were identified at the circular junction, respectively. This is an example of several LACs defining a single circRNA. For details, see Supplementary Doc. 3A2.