Most RNAs in human cells are linear because they are synthesized in a liner fashion using DNA template. However circular RNAs can be generated through back-splicing of internal exons, and recently biologists found that circular RNAs are quite abundant in cells, which is a surprise to many people. Previous circRNAs were known as noncoding RNA, whose main function is to control gene expression by acting as decoy to bind miRNAs or RNA binding proteins. In a recent Cell Research paper, Dr. Zefeng Wang’s group from the CAS-MPG Partner Institute for Computational Biology (PICB) in Shanghai Institutes for Biological Sciences (SIBS) and his collaborators from Zhejiang University and University of California, Los Angeles (UCLA) reported that circRNAs with modified RNA base are often translated to generate new proteins.
Generally all human mRNAs contain a 5’-cap, which is required for its translation. However, in some special cases like stress condition, translation can start in a cap-independent fashion from the middle of mRNA through a sequence called IRES. Many viruses use IRES site for their protein translation. Wang’s lab has previously showed that a circRNA can be efficiently translated using an engineered IRES site. Now they found that a common modification of RNA base, m6A, has similar activity like IRES to drive translation from circular RNA. “Our study shows m6A modification are even more common in circRNAs compared to linear RNAs, and m6A can drive circRNA translation”, said Zefeng Wang, PhD, principal investigator and director in PICB, “combining these two findings, we reached a reasonable conclusion that the extensive m6A modification in circRNA enables a large fraction of circRNAs to be translated in cells, suggesting that there are a large number of unknown new proteins coded by circular RNA in human cells”. This conclusion has big implication, which significantly expanded our known universe of human proteins.
Wang and his colleagues also studied the mechanism of circular RNA translation, and found that reader protein YTHDF3 can bind to m6A site in circRNAs and then recruit eIF4G2 and other translation initiation factors to drive circRNA translation. They also found a large number of circRNAs are associated with polysomes, which is a characteristic of RNAs undergoing translation. They even identified some new peptides encoded by human circRNA using a technique called mass spectrometry.
Although many circRNAs may still be non-coding RNAs, this study suggests that many of them likely function as mRNAs. “This finding blurs the definition of coding and non-coding RNAs”, according to Wang, “we’d like to think circular mRNAs as a new type of mRNAs in addition to the conventional linear mRNAs.”
An important question remain unanswered is what are the functions of these
circular RNA encoded protein. This paper raises some interesting possibility. Under normal condition, protein translation mostly happens in a cap-dependent fashion. However under stress or in some cancers, cap-dependent translation is repressed and cap-independent translation will take over. Since circRNAs do not have cap and all translation of circRNA is cap-independent, the authors indeed found that the circRNA encoded protein translation is evaluated under stress. This finding may suggest that circular mRNA play an important role in stress response by producing stress-induced proteins. Similarly, protein encoded by circRNAs may also play important roles during cancer development when the cap-independent translation is increased.
Finally this research and previous studies both demonstrated m6A can function as an IRES to initiate translation. Therefore Wang further proposed in this paper that one mRNA may generate more than one protein, “just like multiple mRNA isoforms are produced from same gene by alternative splicing.”
Other authors involved in this work are scientists from Institute of Biochemistry and Cell Biology of CAS, National Center for Protein Science, East China University of Science and Technology, Dalian Medical University
Cell Research : http://www.nature.com/cr/journal/vaop/ncurrent/abs/cr201731a.html