Enhanced Tailored MicroRNA Sponge Activity of RNA Pol II-Transcribed TuD Hairpins Relative to Ectopically Expressed ciRS7-Derived circRNAs

Abstract

As key regulators of gene expression, microRNAs (miRNAs) have emerged as targets in basic experimentation and therapy. Administration of DNA-encoded RNA molecules, targeting miRNAs through base pairing, is one viable strategy for inhibiting specific miRNAs. A naturally occurring circular RNA (circRNA), ciRS-7, serving as a miRNA-7 (miR-7) sponge was recently identified. This has sparked tremendous interest in adapting circRNAs for suppressing miRNA function. In parallel, we and others have demonstrated efficacy of expressed anti-miRNA Tough Decoy (TuD) hairpins. To compare properties of such inhibitors, we express ciRS-7 and TuD-containing miRNA suppressor transcripts from identical vector formats adapted from RNA polymerase II-directed expression plasmids previously used for production of ciRS-7. In general, markedly higher levels of miR-7 suppression with TuD transcripts relative to ciRS-7 are observed, leading to superior miRNA sponge effects using expressed TuD hairpins. Notably however, we find that individual ciRS-7 transcripts are more potent inhibitors of miR-7 activity than individual TuD7-containing transcripts, although each miR-7 seed match target site in ciRS-7 is, on average, less potent than the perfectly matched target sites in the TuD motif. All together, our studies call for improved means of designing and producing circRNAs for customized miRNA targeting to match TuD hairpins for tailored miRNA suppression.

Keywords: TuD; ciRS-7; circular RNA; miRNA sponge; miRNA suppression.

Graphical abstract

Figure 1

Superior miR-7 Suppression by miR7-Specific TuDs Compared to ciRS-7 (A) Schematic representation of ciRS-7 and TuD-7 expressed from the CMV RNA pol II promoter. SA and SD refer to splice acceptor site and splice donor site, respectively. (B) Schematic representation of the psiCHECK reporter plasmid designated miR-7 target. The miR-7 target reporter vector encodes a perfect miR-7 target site in the 3′ UTR of an RLuc reporter gene, whereas the vector designated no miRNA target encodes an RLuc reporter gene without any miRNA target sites in the 3′ UTR. Furthermore, the psiCHECK reporter plasmid encodes a FLuc reporter gene used as a control for potential variations in transfection efficiencies. Dual-Glo luciferase assays evaluating the miR-7 suppression potentials of TuD-7 and ciRS-7 in HeLa (C) and HEK293 (D) cells. The transfections were made using equal molar amounts of plasmids. Data are depicted as mean ± SEM; ***p < 0.001.

Figure 2

Modest Enhancement of miR-7 Suppression by ciRS-7 Carrying a miR7-Specific TuD Dual-Glo luciferase assays evaluating miR-7 suppression by ciRS-7 carrying TuD-7 in either (A) MCS5 (ciRS7-MCS5-TuD) or MCS3 (ciRS7-MCS3-TuD), (B) an eGFP-WPRE-TuD-7 expression cassette (ciRS7-eGFP-WPRE-TuD), or (C) WPRE in either MCS5 (ciRS7-MCS5-WPRE) or MCS3 (ciRS7-MCS3-WPRE). Data are depicted as mean ± SEM; *p < 0.05; ***p < 0.001; ns, not significant.

Figure 3

Higher RNA Expression Levels of TuD-7 Than ciRS-7 (A) Schematic representation of ciRS-7 encoding primer and probe binding sites for eGFP-specific TaqMan qPCR (eGFP-qPCR) or probe binding sites for northern blot (eGFP-Northern). TaqMan qPCR (B) and northern blot (C) evaluating ciRS-7 and TuD-7 expression levels in HeLa cells. (D) Quantifications of band intensities from the northern blot shown. The same RNA samples were used for the eGFP-specific TaqMan qPCR and northern blot as well as the ciRS-7 specific TaqMan qPCR shown in Figure S2C. (E) Dual-Glo luciferase assay comparing miR-7 suppression mediated by ciRS-7, ciRS7-TuD7, and TuD-7 at varying plasmid dosages. Except for the Dual-Glo luciferase assay shown in (E), the transfections were made using equal molar amounts of plasmids. (B and E) Data are depicted as mean ± SEM; ***p < 0.001.

Figure 4

Superior RNA Expression Levels of TuD-7 Compared to ciRS-7 (A) Schematic representation of vectors encoding ciRS-7 flanked by complementary intron sequences derived from the endogenous sequence (ciRS-7) or introns from the Drosophila Laccase2 gene (Laccase-ciRS7). (B) Schematic representation of a miR-7 expression cassette stably integrated in the genome of HEK Flp-In T-Rex cells by Sleeping Beauty transposition (upper) and an expression cassette encoding mCherry with four canonical miR-7 binding sites in the 3′ UTR (mCherry-4xmiR7-target) inserted in the genomic FRT-site of HEK Flp-In T-Rex cells by Flp recombination (lower). Northern blot (C) and TaqMan qPCR (E) evaluating ciRS-7, Laccase-ciRS7, and TuD-7 expression levels in HEK Flp-In T-Rex miR7-4x miR7-target cells. (D) Quantifications of band intensities from the northern blot shown. For the comparison of ciRS-7, Laccase-ciRS7, and eGFP-WPRE-TuD-7 RNA levels, the expression of ciRS7-MCS3-eGFP-qPCR or ciRS7-MCS3-eGFP-Northern was used to normalize ciRS-7 and eGFP expression levels measured by qPCR or northern blot, respectively. The ciRS-7 inhibitor effect was unaffected by insertion of eGFP fragments used for quantification (Figure S4) (F) miR-7 suppression potential of ciRS-7, Laccase-ciRS7, and TuD-7 in HEK Flp-In T-Rex miR7-4xmiR7-target cells evaluated by flow cytometry. The transfections were made using equal molar amounts of plasmids. Data are depicted as mean ± SEM; ***p < 0.001; ****p < 0.0001; ns, not significant.

Figure 5

Relative miR-7 Inhibition Capacities of ciRS-7, Laccase-ciRS7, and TuD-7 Transcripts (A) Relative miR-7 inhibition by ciRS-7, Laccase-ciRS7, or TuD-7 normalized to the amount of plasmid. (B) Relative miR-7 inhibition capacity the three inhibitors presented as inhibition potential per RNA molecule. (C) Relative inhibition by ciRS-7, Laccase-ciRS7, or TuD-7 presented as inhibition capacity per miRNA binding site contained in ciRS-7 (73 miR-7 binding sites), Laccase-ciRS7 (73 miR-7 binding sites), and TuD-7 (2 miR-7 binding sites). Relative inhibition capacities are based on results shown in Figures 4D–4F.