The fluorophore DFHBI and Baby Spinach RNA

DFHBI (Lucerna Technologies, USA) was dissolved in DMSO to get 40 mM stock solution and stored at − 20 °C. From that, each stock of diluted solution (200 µM) is made in HEPES buffer (10 mM pH 7.5) containing DEPC treated water, 50 mM KCl and 5 mM MgCl₂.

RNA (sequence: 5′- GGU GAA GGA CGG GUC CAG UAG UUC GCU ACU GUU GAG UAG AGU GUG AGC UCC -3′) was purchased from Dharmacon Inc in protected form. It was then de-protected following Dharmacon’s protocol using de-protection buffer, purified by HPLC, and lyophilized to dryness. The dry pellet was stored at − 80 °C and before use, suspended in HEPES buffer (10 mM pH 7.5) containing DEPC treated water, 50 mM KCl and 5 mM MgCl₂. The 2-ml Eppendorf tube containing RNA in buffer was put in a water bath (2 L), then heated to 65 °C for 3 min and left cooling overnight to room temperature to let the RNA slowly fold to the correct conformation. The tube was then stored at -30 °C fridge and used as RNA stock after thawing to room temperature. The integrity of the RNA was confirmed by gel electrophoresis experiment as shown in Fig. ^S9.

DFHBI at 200 µM stock was diluted in HEPES buffer (10 mM pH 7.5) containing DEPC treated water, 50 mM KCl and 5 mM MgCl₂ to get 2 µM DFHBI solution, which contains less than 0.1% DMSO. RNA aliquots were added from the RNA stock to the 2 µM DFHBI solution to achieve different ratios of RNA to DFHBI concentrations in the mixture. The mixture was vortexed for 1 min and left in the dark for 2 h for completion of the binding process. Spectroscopic experiments were carried out right after that. The formation and the integrity of the G-quadruplex binding pocket before and after forming the complex was confirmed by CD spectra as shown in Fig. ^S10.

All our measurements were performed under flow-cell conditions to minimize photo-bleaching unless otherwise stated. A total of 20 ml of the DFHBI/RNA mixture (at different ratios) was prepared in HEPES buffer (10 mM pH 7.5) containing DEPC treated water, 50 mM KCl and 5 mM MgCl₂. The solution was circulated between a 20-ml bottle and a flow-cell quartz cuvette by a Gilson Minipulse 3 peristaltic pump with a rotation speed of 48 rpm. Fig. ^S11 clearly shows photostability of bound DFHBI by using the flow-cell setup for recording fluorescence upon excitation at 460 nm with an irradiation power of 650 µW at an averaged photon flux density of 10¹⁴ photons/mm²s.

To determine the folding efficiency of the RNA, we employed a method which was reported in Folding assay section in Online Methods in Ref.⁶. That method compared fluorescence (excited at 469 nm) of the mixture under 2 extreme conditions: one in which the RNA is in excess relative to the DFHBI (0.1 µM DFHBI + 10 µM RNA), and one in which the DFHBI is in excess relative to the RNA (10 µM DFHBI + 0.1 µM RNA). For each condition, the signal from DFHBI without RNA was subtracted from each signal. The signal from the first condition (limiting RNA) was divided by the signal from the second condition (limiting dye) to determine the fraction folded. From Fig. ^S12, the fraction of properly folded RNA is estimated to be 91%. Identical CD spectra of the Baby Spinach before and after adding DFHBI (Fig. ^S10) confirmed the formation and the integrity of the G-quadruplex binding pocket before and after forming the complex.