QN-302 demonstrates opposing effects between i-motif and G-quadruplex DNA structures in the promoter of the S100P gene

GC-rich sequences can fold into G-quadruplexes and i-motifs and are known to control gene expression in many organisms. The potent G-quadruplex experimental anticancer drug QN-302 down-regulates a number of cancer-related genes, in particular S100P. Here we show this ligand has strong opposing effects with i-motif DNA structures and is one of the most potent i-motif destabilising agents reported to date. QN-302 down-regulates the expression of numerous cancer-related genes by pan-quadruplex targeting. QN-302 exhibits exceptional combined synergistic effects compared to many other G-quadruplex and i-motif interacting compounds. This work further emphasises the importance of considering G-quadruplex and i-motif DNA structures as one dynamic system.

cacodylate and 100 mM KCl buffer at pH 5.5 while 5 μM G-rich S100P samples were annealed in 10 mM lithium cacodylate and 100 mM KCl buffer at pH 7.0.After annealing, the samples (250 µL) were transferred to quartz 10 mm cuvettes and stoppered to reduce evaporation.
For the UV melting/annealing experiments, the absorbance of the samples was recorded at every 1 °C increase/decrease in at least two cycles at 295 nm and 260 nm.Initially, the samples were held at 4 °C for 10 min followed by gradual increase to 95 °C at a rate of 0.5 °C/min (melting).When the temperature reached 95 °C, it was held for 10 min before the process was reversed (annealing).The average melting (T m ) and annealing temperature (T a ) were identified by the first derivative method for each measured cycle.
The thermal difference spectra (TDS) of the S100P i-motif was obtained by measuring the absorbance spectrum from 230 nm to 320 nm at 4 °C for the folded DNA structure and at 95 °C for the unfolded structure.The sample was equilibrated for 5 to 10 minutes at each of the two temperatures before recording the absorbance.The TDS signature was determined by subtracting the absorbance spectra of the folded structure from the unfolded structure, zero corrected at 320 nm, and then normalised to the maximum absorbance.
For the UV titrations, DNA samples were annealed at 200 μM in 10 mM lithium cacodylate and 100 mM KCl buffer at pH 5.5 for the C-rich S100P sequence and pH 7 for the G-rich S100P sequence.Firstly, 250 μL of 5 μM QN-302 ligand in 10 mM lithium cacodylate and 100 mM KCl buffer at pH 5.5 were placed in quartz 10 mm cuvette.During the titration, an aliquot of annealed 200 μM DNA solution was added to the cuvette and the sample scanned in wavelength range of 425 -580 nm, data interval 0.5 nm and the temperature 20 °C.Two different methods were employed for calculating the fraction bound.The first method was based on recording the change in absorbance at the wavelength of maximum absorbance before addition of DNA (520 nm).The second method was based on recording the change in wavelength at the maximum absorbance after every addition of the DNA.In both cases the data was then converted to fraction bound where 0 indicates unbound (no DNA added) and 1 fully bound (when the change in the spectrum had plateaued).The fraction bound was then plotted against the concentration of the DNA in the sample and these data was fitted to the Rectangular Hyperbola Function according to equation 1.
Where y = the bound fraction = concentration of the DNA, = stoichiometry and is the     association constant.The dissociation constant ( ) was calculated using equation 2. Final   values are given as the average and standard deviation of two repeats.

Circular Dichroism (CD) Spectroscopy
The CD spectra of the S100P C-rich sequence at different pH values, were recorded on a Jasco J-1500 spectropolarimeter using a 1 mm path length quartz cuvette under a constant flow of nitrogen.The samples were diluted to 10 μM in 10 mM lithium cacodylate and 100 mM KCl buffer at pH values ranging from 4.0 to 8.0 and annealed as described above (100 µL per sample).Four spectra scans were accumulated ranging from 200 nm to 320 nm for the buffer at each pH (blank) and DNA samples and measured at 20 °C with a data pitch at 0.5 nm, scanning speed of 200 nm/min with 1 second response time, 1 nm bandwidth, and 200 mdeg sensitivity.Data was zero corrected at 320 nm and transitional pH (pH T ) was determined from the inflection point of the Boltzmann sigmoidal fit for the measured ellipticity at 288 nm and pH range.
Circular dichroism spectroscopy was used to measure any ligand-induced effects on S100P the i-motif and G-quadruplex DNA structure.Scans were accumulated ranging from 230 nm to 320 nm.DNA samples (10 µM) were thermally annealed in 10 mM lithium cacodylate, 100 mM KCl pH 7 (for the G-quadruplex) and pH 5.5 (for the i-motif) at 100 µL final sample volume.
QN-302 was titrated step-wise via consecutive additions of ligand at the following concentration ranges: 0 -100 μM QN-302 for the G-quadruplex and 0 -110 μM QN302 for the i-motif.Titrations presented are the average of two repeats and were corrected for the solvent effect, smoothed using the Savitzki-Golay method at 10 points of window and then zero corrected at 320 nm.Analysis and processing of the data was performed using OriginLab data analysis software to plot the normalised ellipticity at 288 nm for the i-motif against ligand concentration.The sigmoidal curves of the ellipticity at 288 nm against ligand concentration were fitted to the Hill 1 equation to obtain the Hill coefficients (n) and the concentrations of ligand that are required to reach 50% reduction of the molar ellipticity ([D] 50% ).
Melting experiments for the S100P sequences were performed in the presence and absence

Figure S10 .
Figure S10.UV titration of QN-302 with S100P i-motif.Fraction bound was calculated based on the change in absorbance at the wavelength of maximum absorbance when the ligand is unbound -519.5 nm (method 1).

Figure S11 .
Figure S11.UV titration of QN-302 with S100P i-motif.Fraction bound was calculated based on the change in the maximum absorbance after every addition of the DNA (method 2).

Figure S12 .
Figure S12.UV titration of QN-302 with S100P G-quadruplex.Fraction bound was calculated based on the change in absorbance at the wavelength of maximum absorbance when the ligand is unbound -520 nm (method 1)

Table S1 .
Change in melting temperature (ΔT m ) of the DAP, hTelo and ILPR i-motif with QN-302 measured by CD melting experiments.

Table S2 .
Dissociation constants ( ) and stoichiometry ( ) of the S100P i-motif and G-quadruplex with the QN-302 ligand    as determined by equations 1 and 2 by calculating the fraction bound based on methods 1 and 2. Values are average of two repeats.