Are circular RNAs (circRNAs) translated? It is very intriguing to speculate that (at least some) circRNAs might encode functional peptides. So far however, evidence in support of this theory is weak. One might hypothesize that by non-canonical initiation once in a while a ribosome might indeed translate an open reading frame encoded by circular RNA, whether this yields significant amounts of a functional protein still remains to be demonstrated.
In a collaborative effort with the Bindereif lab at the Justus-Liebig-University of Giessen, we have analyzed the sedimentation of circRNAs in sucrose gradients. If circRNAs are indeed translated this could be revealed by a change of sedimentation behavior after treatment with a drug that releases elongating ribosomes from RNAs (Puromycin). In our experiments however, no significant change in sedimentation behavior of selected abundant circRNPs from HeLa cells could be observed, suggesting that these circRNAs are not (or at best only weakly) translated.
CircRNAs do however associate with proteins to form ribonucleoproteins (RNPs). In our recent publication ‘CircRNA-protein complexes: IMP3 protein component defines subfamily of circRNPs’ (published in Scientific Reports) we demonstrate that in HeLa cells, circRNAs form distinct, large RNPs. Moreover, we identify the RNA-binding protein and tumor marker IMP3 (IGF2BP3) as a protein component of numerous circRNPs.
(Picture taken from Schneider et al., 2016, Scientific Reports 6, 31313, doi:10.1038/srep31313, CC 4.0)