Circular RNAs as stable age and experience markers: new insights from Drosophila.

Circular RNAs (circRNAs) arise via backsplicing, linking a downstream splice donor to an upstream acceptor and forming a covalently closed loop that is resistant to exonuclease digestion1. In a recent publication from Kirio et al, researchers profiled circRNAs longitudinally in Drosophila melanogaster brains and found that circRNA abundance increases linearly with age, whereas bulk mRNA levels remain largely steady2. Remarkably, many circRNAs had inferred half-lives exceeding 20 days, which contrasts sharply with the much shorter lifetimes of typical linear mRNAs in flies.

For example, measurements in Drosophila embryos estimate a mean life of ~35 minutes for mRNA3. This extreme stability explains why circRNAs accumulate over time in post-mitotic neurons. Intriguingly, in a temperature-stress paradigm, subsets of circRNAs became transiently upregulated, and their elevated levels persisted for weeks after returning to baseline conditions, behaviors consistent with the notion of circRNAs as “molecular recorders” of environmental experience.
 

Methodological considerations

The study of circRNAs requires careful attention to library preparation, sequencing strategy, and downstream bioinformatics. In the study of Kirio et al, the authors extracted total RNA and performed rRNA depletion. The depleted samples were used as input for the NEXTFLEX™ Rapid Directional RNA-Seq 2.0 Kit, a setup that captures both linear and circular species while preserving orientation information required to identify backsplicing junctions. This kit is well-suited for this purpose, as it tolerates partially degraded RNA, a condition frequently encountered in Drosophila brain samples, which are highly abundant in ribonucleases.

In other studies, samples are digested with RNase R to remove most linear RNAs, leaving behind circular forms. This approach has the caveat that structured linear RNAs may resist digestion, while some circRNAs (particularly long ones) are sensitive to degradation. Therefore, enrichment improves detection but not in a quantitative way.

To avoid biases authors sequenced all RNA molecules present in the sample at high depth (≥50-60 million reads/sample). The distinction between linear and circular RNA is made primarily at the bioinformatic level. circRNAs were distinguished by mapping reads that span the back-splice, which are unique to them. Then they applied stringent filtering, removing spurious junctions and comparing circRNA junction counts with linear mRNA counts from the same host gene. Some candidates were further validated y RT-qPCR across back-splice junctions, confirming the computational calls.
 

Why circRNAs matters

circRNAs are now recognized as a broad class of regulatory RNAs with diverse biological functions, not merely byproducts of splicing. Their extraordinary stability makes them natural long-term molecular traces in post-mitotic tissues, where linear RNAs decay too rapidly to preserve history. In neuroscience and aging research, circRNAs offer a unique means to reconstruct cellular trajectories over time. This suggests promise as biomarkers for diseases where early detection and longitudinal tracking are critical. Importantly, their interpretation requires rigorous methodology and validation, but with optimized library prep and analysis strategies, circRNAs stand out as one of the most stable and informative RNA classes in biology.