CircRNP complexes: from nature to design

Supplementary Materials and methods In vitro synthesis of circRNA-based IMP3 sponge Previous work from our lab had revealed an array of IMP3 RNA-binding motifs, based on SELEX analysis (Schneider et al., 2019). A natural high-affinity target, exon 29 of human ANKRD17, which is processed to a circRNA (Schneider et al., 2016), contains this RNAbinding array within 121 nucleotides; this was taken as a comparison to a synthetic, SELEXbased 101-mer RNA (Schneider et al., 2019). In a control sequence, the five elements were mutated to UG-repeats (Figure 1A). Each of these templates was synthesized by T7 transcription, using synthetic double-stranded oligonucleotides (Sigma-Aldrich) and PCRextension for adding the 20-nucleotide stem-loop sequence (Figure 1A) and the T7 promoter. For P-labeling during in vitro transcription, RNA was synthesized with equal amounts of ATP, GTP and CTP (0.5 mM each), 0.04 mM UTP and 10 μCi of [α-P]-UTP. In addition, 10 mM DTT and a 4-fold excess of GMP (2 mM) were used. Following RQ1 DNase treatment, RNA was purified with Mini Quick Spin RNA columns (Roche). Transcripts were ligated with T4 RNA ligase at 16°C overnight, followed by gel extraction (Costar Spin-X centrifuge tube filters; Corning) and ethanol precipitation.


In vitro synthesis of circRNA-based IMP3 sponge
Previous work from our lab had revealed an array of IMP3 RNA-binding motifs, based on SELEX analysis (Schneider et al., 2019). A natural high-affinity target, exon 29 of human ANKRD17, which is processed to a circRNA (Schneider et al., 2016), contains this RNAbinding array within 121 nucleotides; this was taken as a comparison to a synthetic, SELEXbased 101-mer RNA (Schneider et al., 2019). In a control sequence, the five elements were mutated to UG-repeats ( Figure 1A). Each of these templates was synthesized by T7 transcription, using synthetic double-stranded oligonucleotides (Sigma-Aldrich) and PCRextension for adding the 20-nucleotide stem-loop sequence ( Figure 1A) and the T7 promoter.
For 32 P-labeling during in vitro transcription, RNA was synthesized with equal amounts of ATP, GTP and CTP (0.5 mM each), 0.04 mM UTP and 10 µCi of [α-32 P]-UTP. In addition, 10 mM DTT and a 4-fold excess of GMP (2 mM) were used. Following RQ1 DNase treatment, RNA was purified with Mini Quick Spin RNA columns (Roche). Transcripts were ligated with T4 RNA ligase at 16°C overnight, followed by gel extraction (Costar Spin-X centrifuge tube filters; Corning) and ethanol precipitation.
Binding reactions were analyzed on native 5% TBE gels (containing 5% glycerol), pre-run for 20 min. Gel electrophoresis was performed for 45 min with 45 mA at 4°C. Radioactive signals were visualized by the Typhoon FLA 9500 Phosphorimager system and intensities quantified. Curve fitting of raw data used the quadratic binding equation (Altschuler et al., 2013), and for Kd calculations from experimental triplicates, OriginPro was employed (OriginLab) (Figure 1C).

RNA design based on SELEX-derived data
Highly specific, SELEX-derived RNAs were designed, based on motif enrichment and spacing analyses (for a detailed description of SELEX-seq analysis, see Schneider et al., 2019). The L_12/10/12 RNA (65 nt) contains four RNA-binding motifs (ACAC, ACAC, ACAU, and ACAU) for binding to the four RRM domains of hnRNP L, with a 10-or 12-nucleotide spacing in between (Figure 2A). As a negative control (mut L_12/10/12 RNA), ACAC was mutated to UGUG, ACAU to UGUU. RNAs were transcribed by T7 RNA polymerase and used in linear form, or circularized by T4 RNA ligase, followed by gel purification.

RNA transfection in HeLa cells and splicing analysis.
HeLa cells were cultured in DMEM medium supplemented with 10% FBS at 37°C in an incubator with 5% CO2. For RNA transfection, 5 × 10 4 cells per well were seeded in a 24well plate 24 h before transfection. 1 µg of gel-purified circRNA (or corresponding linear RNA) was transfected with Lipofectamine 2000 Invitrogen) in OptiMEM medium (Gibco, 3 µl/µg of RNA), according to the manufacturer's instructions. Cells were harvested 24 h after transfection, and total RNA was extracted.
Cells were lysed (TRIzol, Ambion), and contaminating DNA was removed by RNasefree RQ1 DNase. Total RNA was extracted, and its concentration was measured at 260 nm (Nanodrop).