Circular RNA Biology: Detection, Function and Pitfalls

Circular RNAs are produced by back-splicing, in which a downstream splice donor joins an upstream acceptor, creating a covalently closed loop. They are abundant in some tissues, stable because they lack free ends, and in a few well-documented cases functionally important.

They are also the subject of a large literature in which the molecule's existence, its abundance, and its function are established with unequal rigor. Three problems recur: detection artifacts (signals that look circular but are not), abundance confusion (copy numbers too low to support a stoichiometric function), and overexpression artifacts (phenotypes produced by non-physiological levels).

Flanking intronic repeats and RBP binding are the two features most predictive of whether a given exon circularizes — useful when deciding which candidate to pursue.

The validation combination that holds up: divergent primers confirmed by Sanger sequencing of the junction, RNase R resistance with a linear control, and — where available — long-read confirmation. Two of the three, not one.

The most common artifact is a genomic duplication or rearrangement that produces a junction indistinguishable from a back-splice event by PCR alone. Sequencing the junction and checking the genomic locus resolves it.

CircRNA function claims often require stoichiometry: a sponge must be present at comparable copy number to its target miRNA; a scaffold must be present at sufficient concentration to assemble complexes.

Practical measurements rarely reported:

• • Absolute copies per cell, from spike-in normalized quantification.
• • Ratio to the linear isoform from the same locus.
• • Ratio to the proposed interaction partner.

A circRNA present at a few copies per cell cannot sponge a miRNA present at hundreds. Papers that omit these numbers cannot support stoichiometric claims, which is the majority of the sponge literature.

The fifth row is the load-bearing one. Overexpression of a circRNA to high levels will produce phenotypes through sequestration of almost any abundant RNA-binding protein; that is a property of the experiment, not necessarily of the molecule.

• • Overexpression artifacts. Plasmid-driven circRNA expression can exceed physiological levels by orders of magnitude; phenotypes do not transfer.
• • Back-splice junction effects. Overexpressed minigene constructs can generate linear species from the same vector; verify what is actually produced.
• • siRNA against a junction. Knockdown of a circRNA with siRNA risks silencing the linear isoform if the siRNA also matches it; design junction-specific reagents and verify specificity.
• • Knockout of the locus. Deleting the genomic region removes the linear isoform too; interpret as a locus-level perturbation unless the circularization signal is specifically disrupted.
• • Translation claims. Detection of a junction-spanning peptide requires mass spectrometry with a synthetic standard, not just a predicted ORF.

1. Junction sequence confirmed (Sanger or long-read).
2. RNase R resistance with a linear control; genomic locus checked for rearrangement.
3. Absolute abundance and ratio to linear isoform.
4. Loss-of-function, not only gain-of-function.
5. For interaction claims, stoichiometry of both partners.
6. For translation claims, peptide-level evidence.

Papers that meet all six exist and are convincing. Papers that meet only the first two are describing expression, which is a legitimate but much smaller claim.

EDITGENE builds models for circRNA work: locus-level and circularization-specific perturbations, isogenic lines with verified junction sequences, and quantification with spike-in standards for absolute abundance.

Tell us the circRNA and the claim. Contact EDITGENE to request a project consultation or quote.

Detection performance varies by method, abundance, and sequence; junction-level confirmation is required before a circular species is reported.

Function claims require loss-of-function evidence and, for stoichiometric mechanisms, absolute abundance measurements for both partners.

• • Kristensen LS, et al. The biogenesis, biology and characterization of circular RNAs. Nature Reviews Genetics. 2019;20:675–691.
• • Jeck WR, et al. Circular RNAs are abundant, conserved, and associated with ALU repeats. RNA. 2013;19:141–157.
• • Szabo L, Salzman J. Detecting circular RNAs: bioinformatic and experimental challenges. Nature Reviews Genetics. 2016;17:679–692.
• • Chen LL. The expanding regulatory mechanisms and cellular functions of circular RNAs. Nature Reviews Molecular Cell Biology. 2020;21:475–490.