Nanopore sequencing-based profiling of circular RNAs in multiple myeloma uncovers circNFATC3 as a promising prognostic biomarker Free

Introduction: Circular RNAs (circRNAs) are a distinct class of non-coding RNAs with regulatory roles and emerging value as non-invasive biomarkers in cancer, including multiple myeloma (MM). While advances in nanopore sequencing have enabled in-depth characterization of circRNAs, the circular transcriptome of MM remains poorly defined. In this study, our goal was to apply high-accuracy nanopore sequencing to investigate the circRNA structural diversity and expression patterns in MM and to assess the prognostic potential of selected circRNAs.

Methods: Ten MM cell lines (L-363, H929, U266, RPMI 8226, KMS-12-BM, LP-1, OPM-2, SK-MM-2, MM.1S, and AMO-1) were cultured. Following total RNA isolation from each cell line and ribosomal RNA (rRNA) depletion, all RNA extracts were incubated with RNase R to selectively degrade linear transcripts. The remaining RNA was reverse-transcribed using random hexamers through a rolling-circle amplification approach. Sequencing libraries were then prepared and sequenced on the PromethION 2 platform, with high-accuracy basecalling activated. All sequencing reads were aligned to the GRCh38 human genome, and data analysis was performed using the CIRI-long 1.1.0 pipeline alongside custom in-house bioinformatics tools. Additionally, bone marrow aspirates were obtained from 175 adults diagnosed with plasma cell dyscrasias. Plasma cells were enriched via CD138+ magnetic selection from mononuclear cells, followed by RNA extraction and cDNA synthesis. Then, we selected 4 circRNAs with high numbers of sequencing reads across all cell lines to explore their clinical relevance. Target circRNAs and the reference gene GAPDH were pre-amplified using a first-round PCR, and their levels were quantified through real-time quantitative PCR. In both reactions for circRNA amplification, divergent PCR primers were used to achieve high specificity and accuracy. Expression levels were calculated as relative quantification units (RQU).

Results: Analysis of the circRNA-enriched sequencing data resulted in the identification of 94,048 distinct full-length circular RNA sequences in the 10 MM cell lines. Remarkably, these circRNAs originated from 14,313 different genes, underscoring the extensive alternative splicing in MM, with features such as novel exon usage and intron retention events. Noteworthy examples include CCND1, implicated in the t(11;14) translocation, which generates 18 distinct circRNAs; NSD2, associated with the t(4;14) translocation, generating 36 circRNAs; IRF4, which produces 45 circRNAs, and BRAF, which produces 49 circRNA isoforms. Next, 4 circRNAs, namely circNFATC3, circMTO1, circATAD2, and circSEC24B, were highly expressed in all cell lines. Interestingly, circNFATC3 and circMTO1 back-splice variants are both novel circRNAs, not yet deposited in established circRNA databases. We successfully designed and implemented a quantitative method to assess the expression of these circRNAs in patients' CD138+ plasma cells. The patient cohort included 110 newly diagnosed symptomatic MM cases, 35 smoldering MM (SMM) cases, and 30 cases with monoclonal gammopathy of undetermined significance (MGUS). circMOT1 was detected in 35 MM, 12 SMM, and 12 MGUS samples; circATAD2 was determined in 61 MM, 18 SMM, and 18 MGUS samples; circSEC24B was detected in 73 MM, 28 SMM, and 24 MGUS samples. circNFATC3 emerged as the most promising biomarker, being expressed in all MM patients' samples. The intracellular levels of circNFATC3 differed between patients' CD138+ plasma cells, showing a gradual increase from MGUS to SMM and to MM (P=0.014). Kaplan-Meier survival analysis did not show any significant correlation between circNFATC3 levels and progression-free survival; however, it uncovered the favorable prognostic value of circNFATC3 expression in MM patients' CD138+ plasma cells, with regard to overall survival (P=0.040).

Conclusions: Our circRNA profiling workflow of 10 MM cell lines revealed great structural diversity of the circular transcriptome in this disease. The integration of these findings with patient-derived data revealed circNFATC3, a novel circRNA, as a potential biomarker with differential expression across disease stages and its correlation with MM patients' overall survival, highlighting its value for future clinical research in larger cohorts of patients.