Comprehensive Identification of Translatable Circular RNAs Using Polysome Profiling

Background

Circular RNAs (circRNAs) are a special type of RNA without 5’- and 3’-ends, which form a covalent close loop. They were first discovered in viroids four decades ago (Sanger et al., 1976). Recently, many studies showed that circRNAs were widely present in eukaryotes and expressed in a tissue-specific manner (Rybak-Wolf et al., 2015; Chen, 2020). Several types of circRNAs are found in cells, including exonic circRNA, EIciRNA, and ciRNA (Kristensen et al., 2019); however, the majority of them are exonic circRNAs that are generated through back-splicing of pre-mRNAs (Salzman et al., 2012 and 2013; Memczak et al., 2013; Ashwal-Fluss et al., 2014; Zhang, X. O. et al., 2014). Thus, most exonic circRNAs sequences are the same as those of linear mRNAs, and the back-splicing junctions are unique sequences for circRNA detection.

CircRNAs regulate biological processes through their activities as miRNA sponges, protein scaffolds, and transcription regulators (Kristensen et al., 2019). However, the functions of most circRNAs are still unclear. Recently, several studies showed that many circRNAs can be translated through cap-independent translation from IRESs or m6A RNA modifications (Legnini et al., 2017; Pamudurti et al., 2017; Yang et al., 2017). In addition, some circRNA encode peptides reported to influence cancer proliferation or lifespan in flies (Weigelt et al., 2020; Yang et al., 2018; Zhang, M. et al., 2018), indicating that the translation products from circRNAs play important biological roles. However, the roles of translatable circRNAs are still unknown. Therefore, transcriptome-wide identification of translatable circRNAs is an important way to study their biological function.

Two different types of omics approaches are usually used to comprehensively measure the translation of mRNAs at RNA (translatome) and protein (proteome) levels. Ribosome footprinting is an emerging technique to globally measure mRNA translation via high-throughput sequencing of ribosome-protected mRNA fragments; however, it’s difficult to capture the translating circRNAs. Since the ribosome-protected mRNA fragments are short (around 28-35nt), the back-splicing junctions are easy to be missed by this method. Proteomic analysis is a technique to directly determine the translation products of mRNAs through identification of enzymatically digested peptides. Similar to ribosome footprinting, the peptides coded by back-splicing junctions are difficult to identify through tandem mass spectrometry.

We developed a new method to identify translatable circRNAs using polysome profiling and circRNA-seq. The polysome-associated RNAs can be separated using sucrose gradients and then isolated from different fractions. The polysome-bound circRNAs are enriched by an RNase R treatment and used to generate the circRNA-seq libraries. The resulting libraries are pair-end sequenced, and the circRNAs are identified by the back-splicing junction reads. Therefore, our method is easier and more sensitive than ribosome footprint and proteomic analyses for identifying translatable circRNAs.