Turning Ribozymes into Reusable Tools: New System Separates a Key Step in Large-Scale Circular RNA Production

Circular RNA lacks the two ends that are readily attacked by nucleases, making it more resistant to degradation than linear RNA and potentially enabling longer-lasting protein expression. However, producing these molecules with sufficient purity and stability at a controllable cost remains a manufacturing hurdle in moving them from the laboratory into vaccines and protein therapeutics. A study published in *Nature Communications* presents a “split intron-exon” (SIE) system that seeks to make the RNA catalyst responsible for circularization work repeatedly, like an industrial enzyme.

The conventional PIE method arranges a Group I intron ribozyme and the sequence to be circularized on the same precursor RNA. Each time one molecule is spliced, the ribozyme also becomes part of the reaction products, making it difficult to reuse and leaving impurities such as linear precursors and free introns that must be removed. SIE instead divides the two into separate RNAs: one retains the catalytic core, while the other carries the sequence to be circularized. The two pair through recognition sequences and then complete two transesterification reactions. The ribozyme remains intact after the reaction, potentially allowing it to enter another catalytic cycle.

The research team confirmed the products using electrophoresis, RT-qPCR, RNase R treatment, and splice-junction sequencing, and produced circular RNA as long as approximately 5.7 kb. Under the test conditions, the reaction approached a plateau in about 3 to 5 minutes. The system could also process substrates containing chemical modifications such as N⁶-methyladenosine or N¹-methylpseudouridine. Because the modifications occur only on the substrate and do not interfere with the separately prepared ribozyme, SIE retained clear circularization capacity when the corresponding starting materials were fully replaced with modified nucleotides, whereas the PIE and TRIC comparator systems in the study largely lost activity.

The design most directly connected to large-scale production involved immobilizing a ribozyme biotinylated at its 3′ end on streptavidin magnetic beads. After magnetic separation, ribozyme contamination in the reaction solution was almost undetectable. The immobilized ribozyme could still synthesize circular RNA after 20 consecutive cycles of use, although the yield had declined moderately. This indicates that recovery is not lossless: free introns may continue to bind to the ribozyme, and repeated regeneration may also compromise its stability. A more robust covalent immobilization method may be needed in the future.

The team further produced circular RNAs encoding an anti-PD-L1 antibody fragment and human erythropoietin. In human cells, protein expression from both products persisted longer than expression from the corresponding linear mRNAs. After encapsulation in lipid nanoparticles, the circular RNA encoding erythropoietin also increased the proportion of reticulocytes in mice. The liver and kidney function measures and TNF-α indicator examined in the study showed no significant differences, but this was only a limited animal proof-of-concept and cannot be regarded as a comprehensive toxicology or immune-safety assessment.

The published international patent application defines SIE as a two-component system comprising a ribozyme and a substrate, and lists substrates of different lengths, reaction conditions, and uses including vaccines and protein expression. A related Chinese patent seeks to circularize the linear catalytic ribozyme itself to improve stability and reuse efficiency. The paper’s authors also disclosed that five of them are inventors of SIE and its applications, and that the technology has received a Chinese patent and is the subject of a PCT application, indicating that the research has a clear translational and intellectual-property strategy.

However, the study has not demonstrated scaled-up batches compliant with pharmaceutical manufacturing standards, and it lacks comprehensive comparisons with current processes in terms of cost, recovery rate, impurity profile, and batch-to-batch consistency. Immune responses to circular RNA are often affected by residual linear RNA, triphosphate ends, and splicing traces. Although SIE can reduce ribozyme carryover, it cannot replace downstream purification and quality testing. At present, it is better viewed as a modular manufacturing prototype supported by experimental evidence. Considerable engineering and regulatory validation is still required to demonstrate that it can reliably supply therapeutic-grade products at industrial scale.