G-quadruplex deconvolution with physiological mimicry enhances primary screening: Optimizing the FRET Melt2 assay

Abstract

Non-B-DNA G-quadruplex (G4) structures have shown promise as molecular targets. Modulating G4 stability for oncogenic transcriptional control is a promising avenue for the development of novel therapeutics. Extracellularly, G4 stabilization can be mediated by alkali cations, modifying water content, or with molecular crowding. Intracellularly, G4 formation is mediated by negative superhelicity and transcriptional activity, and can be stabilized with small molecules or oligonucleotides. Numerous G4-stabilizing compounds have been identified that impact promoter activity in plasmids. These compounds, however, infrequently show activity in cells, are found to have non-G4-mediated mechanisms of action, or do not demonstrate activity in vivo. The G4 field requires enhanced predictive screening methods to identify compounds with G4-mediated in vitro activity and in vivo efficacy. Using the best characterized promoter G4 to date, MYC, we examined the effects of varying annealing conditions (rate of cool down and number of heat/cool cycles), co-solvents (glucose, acetonitrile, polyethylene glycol, dextran sulfate, sucrose, ficoll-70, glycerol) and nucleoplasm on G4 formation and compound screening. We observed a marked decrease in hit rates when shifting from simple buffer conditions to include potassium and glycerol, and utilizing two or more rapid annealing cycles; the difference in hit compounds coincides with previous findings of active, inactive, and non-G4-mediated activity, including NSC338258, Quindoline i, and TMPyP4; with these changes, we describe a modification of the primary FRET Melt screening assay - the FRET Melt². This understanding of physiological principles governing the above G4 formation will better inform future drug discovery efforts for this and other oncogenic promoters.

Keywords: FRET melt; G-quadruplex; MYC; Physiological conditions; Small molecules.

Fig. 1.

MYC G4 heating and annealing conditions affect stability and isoform distribution. The MYC G4 forming oligonucleotide, Pu46, was subjected to an array of heating and annealing conditions including a variation of number of cycles and of annealing temperatures. Thermal melt profiles were examined by ECD for the melting and re-annealing at one, five, and ten total cycles under slow (left) and rapid (right) cooling conditions; ECD mdeg was normalized to pre-melting values at 20 °C. Data were collected every 1 °C, all experiments were performed in triplicate. Five cycles of annealing using rapid cooling demonstrated the most consistent thermal profiles under melting and annealing conditions.

Fig. 2.

Co-solvents and nucleoplasm effect on MYC G4 formation and isoform distribution. ECD Spectra (left, measured at room temperature) and thermal stability (right) were determined for the MYC G4 in the presence of saturating co-solvents (A) or increasing nucleoplasm (B). The dashed lined in both A and B compares the T_M of the MYC G4 in the presence of 2% nucleoplasm (demonstrating the fewest isoforms) with the co-solvents. (C) G4 formation and isoform density were confirmed by EMSA, as compared to a linear knockout control oligonucleotide. From these studies, 10% glycerol and 20% ficoll-70 were determined to be most consistent with nucleoplasm in regards to G4 formation, thermal stability, and isoform distribution.

Fig. 3.

Melting profiles of the 5′ FAM and 3′ TAMRA labeled Pu46 FRET probe under varying conditions of co-solvent and melt cycles with 10 μM test compounds. Fluorescence as a function of temperature, and as an indirect measure of G4 thermal stability, was measure on a BioRad CFX96 real-time PCR machine in varying control and test conditions, as indicated. RFU was normalized to initial (20 °C) and final (100 °C) values. Negative test compounds are shown in red and positive test compounds are shown in green; separation of the red and green compounds is most evident in the assay run with 10% glycerol as a co-solvent.

Fig. 4.

Improved predictability of the FRET Melt² assay. The FRET assay was done under varying conditions of up to nine heat and anneal cycles, with buffer conditions in the presence of varying osmolytes, and at either ligand:G4 ratios of 0.5 (1 μM) or 5 (10 μM). Z′-factors were calculated and consistently improved with the inclusion of glycerol (A and B). The updated FRET Melt assay using conditions with Z′’ factors > 0.5 was performed with an unrelated glucocorticoid, prednisolone, and with two newer MYC G4 stabilizing compounds, DC-34 and D-089 (C, the inset legends for each graph indicate the ΔT_M).