Background

Circular RNAs (circRNAs) are covalently closed RNA molecules that exhibit increased stability due to their resistance to exonuclease cleavage [1]. circRNAs have been implicated in a variety of biological functions [2], including human diseases such as cancer [3] and neurological disorders [4], where circRNAs have shown potential as biomarkers [5]. Additionally, circRNAs are considered attractive candidates for RNA therapeutics because of their enhanced stability and protein production compared to linear RNAs [6,7]. Given the growing interest in circRNAs for these applications, accurate detection and sequence validation are crucial.

Numerous RT-PCR-based sequencing methods have been described for identifying natural circRNAs in cell extracts [8,9]. These are high-throughput methods designed to detect the circRNA transcriptome by utilizing RNA enrichment strategies, such as ribosomal RNA (rRNA) depletion and RNase R digestion (to eliminate linear RNAs) [10]. However, these approaches are too costly and time-consuming for the verification of individual circRNAs, such as those synthetically made or individual natural circRNAs. Currently, individual circRNAs are sequence-verified through RT-PCR followed by Sanger sequencing across the circRNA junction, or nanopore sequencing [6,7,11,12]. However, these approaches have several notable limitations. For small circRNAs (~100 nt), the resulting RT-PCR product is too short for Sanger and Nanopore sequencing, thus necessitating a cloning step prior to sequencing, as utilized by Du et al. [7]. This additional cloning step increases the time and cost of the verification process. Additionally, RT-PCR is typically performed using retroviral reverse transcriptases (RTs) [13,14], which stall on structured RNAs [15], restricting detection to unstructured circRNAs within a limited size window. While more highly processive RTs have been employed for circRNA detection [16], existing protocols require the use of high temperatures during cDNA synthesis that may shear and degrade RNA circles. Moreover, they have only been used for high-throughput sequencing of cellular circRNAs. Currently, there is no simple protocol for sequencing individual synthetic or cellular circRNA sequences by RT-PCR, irrespective of circRNA structure, sequence complexity, or length.

The method described here enables the straightforward validation of individual circRNA sequences, without the need for enrichment strategies or cloning, by using the highly processive UltraMarathonRT (uMRT) [17,18]. This approach utilizes RT-PCR with uMRT and commercial Nanopore or Sanger sequencing to rapidly and cost-effectively validate small, large, and structured circular RNAs at room temperature. Although this protocol focuses on validation of individual circRNAs, these highly processive RTs could be used for transcriptome-wide RNA sequencing of natural circRNAs or in sequencing approaches that circularize a transcriptome [19]. Applying uMRT to transcriptome-level approaches will require further experiments, but it has the potential to reveal novel circRNAs, including those containing structured regions and other complex elements.