Identification of potential proteins translated from circular RNA splice variants

Circular RNAs (circRNAs) are covalently closed RNA molecules generated from precursor RNAs by the head-to-tail backsplicing of exons. Hundreds of studies demonstrated that circRNAs are ubiquitously expressed and regulate cellular events by modulating microRNA (miRNA) and RNA-binding protein (RBP) activities. A few circRNAs are also known to translate into functional polypeptides regulating cellular physiology. All these functions primarily depend on the full-length sequence of the circRNAs. CircRNA backsplice junction sequence is the key to identifying circRNAs and their full-length mature sequence. However, some multi-exonic circRNAs exist in different isoforms sharing identical backsplice junction sequences and are termed circRNA splice variants. Here, we analyzed the previously published HeLa cell RNA-seq datasets to identify circRNA splice variants using the de novo module of the CIRCexplorer2 circRNA annotation pipeline. A subset of circRNAs with splice variants was validated by the circRNA-rolling circle amplification (circRNA-RCA) method. Interestingly, several validated circRNAs were predicted to translate into proteins by the riboCIRC database. Furthermore, polyribosome fractionation followed by quantitative PCR confirmed the association of a subset of circRNAs with polyribosome supporting their protein-coding potential. Finally, bioinformatics analysis of proteins derived from splice variants of circCORO1C and circASPH suggested altered protein sequences and structures that could affect their physiological functions. Together, our study identified novel circRNA splice variants and their potential translation into protein isoforms which may regulate various physiological processes.     

Two newly identified exons in human GRM1 express a novel splice variant of metabotropic glutamate 1 receptor

To date, five human metabotropic glutamate (mGlu) 1 receptor splice variants (1a, 1b, 1d, 1f, and 1g) have been described, all of which involve alternative C-terminal splicing. mGlu1a receptor contains a long C-terminal domain (341 amino acids), which has been shown to scaffold with several proteins and contribute to the structure of the post-synaptic density. However, several shorter mGlu1 receptor splice variants lack the sequence required for these interactions, and no major functional differences between these short splice variants have been described. By using RT-PCR we have shown that two human melanoma cell lines express both mGlu1a and mGlu1b receptors. In addition, using 3′RACE, we identified three previously unknown mGlu1 receptor mRNAs. Two differ in the length of their 3′ untranslated region (UTR), and encode the same predicted protein as mGlu1g receptor—the shortest of all mGlu1 receptor splice variants. The third mRNA, named mGlu1h, encodes a predicted C-terminal splice variant of 10 additional amino acids. mGlu1h mRNA was observed in two different melanoma cell lines and is overexpressed, compared with melanoma precursor cells, melanocytes. Most importantly, this new splice variant, mGlu1h receptor, is encoded by two previously unidentified exons located within the human GRM1 gene. Additionally, these new exons are found exclusively within the GRM1 genes of higher primates and are highly conserved. Therefore, we hypothesize that mGlu1h receptors play a distinct role in primate glutamatergic signaling.