Design and application of small circular RNAs with antisense function

Abstract

Circular RNAs (circRNAs) belong to a class of noncoding RNAs (ncRNAs) and are particularly characterized by their covalently closed ring structure. The biogenesis of circRNAs is based on a kind of alternative splicing mechanism, also known as backsplicing. Functionally, however, the vast majority of circRNAs are still largely uncharacterized, with the exception of a microRNA sponge function, which could only be experimentally confirmed in a few cases. Based on high-throughput sequencing technologies, the number of circRNAs described in the literature is steadily growing, emerging as novel and attractive biomarkers. However, since endogenous trans-spliced transcripts can lead to false-positive detection of circRNA-junction sequences, it is crucial that not only bioinformatic analyses alone are used for validation. In the context of this work, the important need for quality criteria for the validation and characterization of circular RNAs was therefore pointed out. In addition to bioinformatic analysis, classical biochemical methods are essential for the characterization of circRNAs and the convincing confirmation of their circularity (Pfafenrot and Preußer 2019, see section I). Considering these validation criteria, various biochemical methods were summarized that are particularly suitable for investigating circRNA-protein complexes (circRNPs). Two categories of methods were presented, focusing on either a specific RNA or a specific protein, which can be applied on a gene-specific as well as on a global level. Based on this, the advantages and challenges of the available approaches have been discussed (Ulshöfer et al. 2021, see section J). Among other RNAs, circular RNAs have been described as a component of platelet-derived extracellular vesicles (EVs), but the underlying selective sorting mechanism is currently unknown. In this context, the sequence dependence, size distribution, and the influence of the biogenesis and processing pathways of different linear reporter and endogenous RNAs on their secretion into EVs were investigated. This analysis revealed that short RNAs are secreted more efficiently than longer ones, and that Pol III transcripts are also packaged more efficiently into EVs compared to Pol II transcripts. However, a general enrichment of the investigated RNAs in EVs compared to the corresponding cellular levels could not be confirmed (Mosbach et al. 2021, see section K). Since circRNAs do not contain free 5' and 3' ends, they have a very high intracellular stability compared to their linear counterparts. Thus, circRNAs represent an attractive basis for the development of RNA-based therapeutics in the field of molecular medicine. In order to exploit this potential, the focus was placed on the establishment of small antisense circRNAs (AS-circRNAs), which could be used as molecular tools to specifically regulate translation initiation in eukaryotes. In a proof-of-principle study, a luciferase construct containing the β-globin (HBB, hemoglobin subunit beta gene) 5'-UTR and a series of synthetic AS-circRNAs demonstrated a dose-dependent decrease in translational activity. In this context, the cap- as well as the start codon-proximal regions were identified as the most effective target regions. Based on these findings, an endogenous mRNA (MDM2; mouse double minute 2 homologue) was subsequently used as a target for further investigations to verify this approach. A reduction of the translational activity at the protein level by Western blot as well as a reduction of the mRNA stability at the mRNA level by RT-qPCR could be demonstrated (see section M). In the context of the current COVID-19 pandemic with its global impact on human health and the economy, the established AS- circRNA technology from the previous studies was also applied to SARS-CoV-2. Based on a series of AS-circRNAs targeting specific 5′-UTR regions of the SARS-CoV-2 genome and subgenomic RNAs, the cap- proximal region (including part of the 5′- leader sequence) was identified as the most effective target region. Here, a reduction of viral translation of up to 90% could be achieved. Based on luciferase and infection assays, it was further shown that the inhibitory activity of circular RNAs is consistently more effective than that of linear RNAs and that the antiviral effect of AS-circRNAs persists for up to 48 h longer. A comparison with modified antisense oligonucleotides produced according to the state-of-the-art also confirmed the significantly stronger effect of AS-circRNAs. This approach to investigate the efficacy of AS-circRNA against SARS-CoV-2 is based on circRNA transfection followed by viral infection, which in turn corresponds to prophylactic treatment. In this context, however, it could also be confirmed that in the reverse order, i.e. viral infection followed by circRNA transfection, the efficiency and durability of the antiviral effect remains. This suggests that the AS-circRNA approach is useful not only for prophylactic strategies, but also for protection against viral infections and for antiviral therapy (Pfafenrot et al. 2021, see section L). In summary, the results indicate that the designer AS-circRNAs established in this work represent a new generation of versatile and adaptable RNA therapeutics with great potential.