Analysis of non-canonical three-and four-way DNA junctions

The development of compounds that can selectively bind with non-canonical DNA structures has expanded in recent years. Junction DNA, including three-way junctions (3WJs) and four-way Holliday junctions (HJs)


Introduction
Since the structure of DNA was first eluded in 1953 by Watson and Crick, 1 Franklin 2 and Wilkins, 3 our knowledge of the structure and function of DNA has greatly expanded.DNA is a major target for many small molecules and targeted therapies.Many of these therapeutics bind with double-stranded DNA (dsDNA) but lack specific genetic targeting.Consequently, there is significant interest in developing compounds that bind sequence specifically or with higher order nucleic acid structures.In particular, non-canonical DNA structures, such as three-way junctions (3WJs), 4 Holliday junctions (HJs), 5 G-quadruplexes (G4s) 6 and I-motifis, 7 have become a focus within the field of targeted therapeutics.[10] However, much less work has been reported on chemical compounds that bind with junction DNA.DNA junctions, including 3WJs and HJs, are unique branched structures that consist of several double stranded DNA sequences converging at a single point, which is known as the branch point 11,12 (Figure 1).These structures play important roles as intermediates during genetic rearrangement processes, including DNA replication 13 and homologous recombination, 14,15 which makes them important emerging therapeutic targets.3WJs are the simplest branched nucleic acid structure and consist of three strands of DNA that are partially complementary to each other.They are formed transiently during assembly of the replication fork during DNA replication and are therefore of interest as a biological target.Various compounds have been synthesised to selectively bind with, or stabilise 3WJs, including peptide-based supramolecular helicates 11,[16][17][18][19] , three-fold symmetric triptycene derivatives, 19 and small molecules. 20or this study, we selected a 3WJ DNA binding peptide helicate (ΛΛ-Fe II 2 LLD) 17 as the molecule of choice for our assay development.
HJs are sequence dependent 4-way junctions that play essential roles in regulating DNA functionality by mediating recombination and repair processes.HJs have two structural conformations, open-X and stacked-X, the latter of which has two isoforms 21 (Figure 2).The conformation adopted by the HJ is dependent on several variables including pH, salt concentration, and presence of recognition elements.Early work probing the HJ sought to understand the cleavage profile of metallodrugs with the branched structure, and subsequent analysis with dyes and porphyrins.Several recent HJ binding compounds have been reported, including peptides, 22 psoralen derivatives, 23 ATR inhibitors, 24 organometallic pillarplexes, 25 and bis-acridine compounds. 26In particular, a bis-acridine compound with a C6 linker (BA-C6) has been shown to efficiently bind with and induce HJ formation through molecular assays 27,28 and crystallography studies. 26However, the development of new in vitro molecular assays that take advantage of recent advancements in fluorescence and immobilisation-free binding methods have yet to be reported.
Herein, we report new multiplex PAGE and microscale thermophoresis methods used to characterise the recognition of 3WJ and HJ DNA.

Synthesis of LLD helicate
The LLD peptide ligand was synthesised using Fmoc solid-phase peptide synthesis protocols as previously described 17 and Fe(II) was allowed to coordinate it (2:1, Fe(II):LLD) giving rise to the corresponding dinuclear Fe(II) peptide helicate.

Synthesis of Bis-acridine
BA-C6 was synthesised using a method previously described 29,30 and purified using column chromatography.The tetrahydrochloride salt of the compound was obtained by solubilising the BA-C6 freebase in a 1:1 chloroform:methanol mixture and titrating HCl into the stirring mixture, followed by vacuum reduction.The subsequent yellow powder of the tetrahydrochloride salt was obtained by redissolving the compound in methanol followed by precipitation with ethyl acetate and isolated by vacuum filtration.

Fluorescent oligonucleotide sequences
Fluorescently labelled oligomers were required to visualise non-canonical structures that arise when partially complementary 3WJ and HJ strands hybridise upon incubation at 37 ºC.To visualise individual strands fluorophores without overlapping emissions were selected, such that multiplex gel electrophoresis could be performed.Fluorescence SpectraViewer (ThermoFisher Scientific) was employed to identify the excitation and emission profiles of suitable fluorophores (Figure 3).Fluorescein (FAM), carboxy-X-rhodamine (ROX) and cyanine-5 (Cy5) were selected as appropriate labels for the 3WJ strands (Y1-FAM, Y2-ROX and Y3-Cy5), while Alexa Fluor TM 350 (AF350), FAM, ROX and Cy5 were selected for the HJ (B-Cy5, H-AF350, R-ROX and X-FAM).HJ strands (B, H, R and X) are named according to literature nomenclature, which originated from unique restriction sites on each arm proximate to the junction's core within an 80 bp HJ sequence developed by Duckett et al., 31 which can be shortened by specific restriction enzymes (BamHI, HindIII, EcoRI and XbaI).A fluorescently labelled off-target DNA duplex was designed in parallel to visualise offtarget effects of 3WJ and HJ binding compounds.A spacer sequence of four thymine bases, shown in blue, was included on all strands to ensure the fluorophore did not interfere with ligand binding.Fluorophore labelled sequences were purchased from Integrated DNA Technologies, re-dissolved to 100 μM according to manufacturer guidelines, and their concentration was confirmed by measuring absorbance at 260 nm (Nanodrop 1000, Themo Fisher Scientific).

Gel electrophoresis
Multiplex polyacrylamide gel electrophoresis (PAGE) was performed to evaluate the non-canonical 3WJ and HJ structures that are formed.To correctly visualise both structures, it was necessary to independently optimise the PAGE conditions for the 3WJ and HJ.

Evaluation of non-canonical HJ structures
The HJ was probed for its ability to generate other higher order pseudo-duplex DNA.Pseudo-duplex structures were analysed using polyacrylamide gel electrophoresis (PAGE) using a 12% native Tris Borate EDTA (TBE) gel composition (1x TBE pH 6.0, adjusted to pH 6 with 6 M acetic acid-FisherScientific, 10031223).Each combination of fluorescent HJ strands (i.e.B, H, R, X, B-H, B-R, B-X, H-R, H-X, R-X, B-H-R-X) were mixed in H 2 O (nuclease-free) containing 10 mM phosphate, 150 mM NaCl, 2 mM MgCl 2 , at pH 6.0 at a concentration of 10 pmol, melted and annealed to room temperature prior to the addition of loading dye.Samples required no post-gel staining, and were run for 180 min at 50 V, and imaged using a multiplex assay (G:Box 9 mini-Syngene).AF350 (ex.UV, em.441 nm), FAM (ex.Blue, em.525), ROX (ex.Green, em.605 nm) and Cy5 (ex.Red, em.705 nm) images were captured separately (supplementary Figure S2), overlayed, and a false colour applied to distinguish each fluorophore labelled sequence (GeneSys software-Syngene).

Band densitometric analysis of HJ DNA formation induced by BA-C6
HJ DNA must be prepared from oligomers that are annealed on the same day as use.A Cy5-labelled sequence and three non-labelled strands (B-Cy5, H, R and X) were used in this study whereby Cy5 enabled HJ visualisation and subsequent quantification without interference from other fluorescent signals.Annealing was performed by heating samples to 95 ºC and cooled in 5 ºC increments to 4 ºC (annelaing was performed on a Mastercycler ® nexus-Eppendorf).Samples with HJ DNA (10 pmol) and 0.5-250 equivalents of BA-C6 (2.5-1250 pmol) were mixed in nuclease-free H 2 O and incubated at 37 ºC for 1 hour.A single stranded control (Cy5-B strand, 10 pmol), HJ control (10 pmol in nuclease-free H 2 O), and buffered HJ control (10 pmol, 10 mM phosphate, 150 mM NaCl, 2 mM MgCl 2 , pH 6.0) were prepared to serve as reference for compound-induced HJ formation.DNA loading dye (6X, 10 mM Tris-HCl, 0.03% bromophenol blue, 0.03% xylene cyanole FF, 60% glycerol, 60 mM EDTA-ThermoScientific, R0611) was added and the samples were analysed by 12% native PAGE (1x TBE-FisherScientific, 10031223-adjusted to pH 6.0 with acetic acid).The gel was run at 50 V for 180 min and visualised using red excitation together with a 705 nm filter (G:Box 9 mini-Syngene).Band densitometry was performed in triplicate to evaluate formation of the HJ (GeneTools software-Syngene).All samples were compared to control sequences run without BA-C6 with the relative intensities plotted and analysed in GraphPad Prism (supplementary Figure S5).

Experimental design
Microscale thermophoresis experiments were developed using a fluorescently labelled target (3WJ or HJ DNA) and the non-labelled ligands (ΛΛ-Fe II 2 LLD or BA-C6).Therefore, 3WJ, HJ, and the dsDNA off-target each contained one Cy5 labelled strand as outlined below.Binding affinity experiments were performed using the red LED channel with automatic power (i.e.80%) and the temperature was set at 25 ºC.The Cy5-3WJ target (20 nM) was sufficiently luminescent and a maximum ΛΛ-Fe II 2 LLD ligand concentration of 50 μM was employed.ΛΛ-Fe II 2 LLD (20 μL, 100 μM) was added to sample 1 and then a serially diluted with 0.1% Tween-80 HEPES buffer (0.1% Tween-80, 10 mM HEPES, 750 mM NaCl, 50 mM MgCl 2 , pH 7) for samples 2 to 16. Tween was included in the buffer to mitigate the effects of DNA condensation by the cationic ligand.Cy5-3WJ (10 μL, 40 nM) was added to all samples and solutions were mixed but not vortexed.Samples were centrifuged to remove air bubbles and incubated at 37 ºC for 5 min.Each sample was then loaded into a glass capillary (Monolith standard capillary-NanoTemper Technologies GmbH, MO-K022) and placed into the sample tray.It is important not to touch the centre of the capillaries as this can impede accurate analysis.Controls with non-metallated LLD peptide ligand, (NH 4 ) 2 Fe(SO 4 ) 2 •6H 2 O (Mohr's salt), and Cy5-3WJ alone were also performed, and each assay was repeated in triplicate.

MST with ΛΛ-Fe II 2 LLD and dsDNA
Next, the binding affinity for ΛΛ-Fe II 2 LLD with dsDNA was investigated.Here, two discrete experiments were performed; i) ΛΛ-Fe II 2 LLD with Cy5-dsDNA; and ii) ΛΛ-Fe II 2 LLD with Cy5-dsDNA and non-labelled 3WJ.

Data analysis for ΛΛ-Fe II 2 LLD and 3WJ
MST data was imported into the MO.Affinity Analysis software (NanoTemper Technologies GmbH) and the dissociation constant (K d ) was calculated.During MST analysis fluorescence inhomogeneity and the presence of aggregates were observed (Figure S6), which is due to DNA condensation, a similar effect which was observed during PAGE analysis of ΛΛ-Fe II 2 LLD and 3WJ DNA (vide infra Figure 5).Due to this effect, the K d was calculated based on the initial fluorescence, rather than at a specific time-point during the MST analysis.This analysis was performed in triplicate on all samples and the data plotted in GraphPad Prism, with standard deviation shown as error bars.
The MO.Affinity Analysis software (NanoTemper Technologies GmbH) uses the K d fit model to describe a molecular interaction with 1:1 stoichiometry according to the law of mass action.

Where
is the fraction bound at a given concentration c;  (  ) Unbound is the F norm signal of the target alone; Bound is the F norm signal of the complex; K d is the dissociation constant (or binding affinity); c t is the final concentration of target in the assay.

Binding of BA-C6 to HJ DNA
In a similar manner to the 3WJ assay, BA-C6 (100 μM, 20 μL) was added to sample 1 and then a serial dilution with water was performed for samples 2 to 16. Pre-annealed Cy5-HJ (40 nM, 10 μL) was then added to all samples and solutions were mixed, but not vortexed.Samples were centrifuged to remove air bubbles and incubated at 37 ºC for 60 min.Next, samples loaded into a glass capillary (Monolith standard capillary-NanoTemper Technologies GmbH, MO-K022) and placed into the Monolith sample tray.Care was taken to avoid touching the centre of the capillary.MST binding affinity was performed, as described above, with BA-C6 with HJ, 9-amino-acridine (9-NH 2acridine; Sigma-Aldrich (Merck)-92817) with HJ and HJ alone.A dsDNA control with BA-C6 was also performed.MST experiments were repeated in triplicate.

Data Analysis for the Holliday Junction
The MST data for BA-C6 with Cy5-HJ was analysed using the MO.Affinity Analysis software (NanoTemper Technologies GmbH), but the Hill model with EC 50 fit was used to analyse cooperative binding instead of the K d model used for the 3WJ.The EC 50 was more applicable in this situation as there is state change in the target whereby the HJ changes from the open-X to the stacked-X form upon binding of the BA-C6 ligand.
Where is the fraction bound at a given concentration c;  (  ) Bound is the F norm signal of the complex; EC 50 is half of the maximal concentration; n Hill is the Hill coefficient and describes the cooperativity of the reaction.

Multiplex fluorescent PAGE
Fluorophore labelled sequences (section 2.2) were designed to facilitate evaluation of 3WJ or HJ binding ligands.The 3WJ sequences were labelled with FAM, ROX and Cy5, while AF350 was also included for the HJ.Each strand of the 3WJ, the sequences that form pseudo duplexes, and the combined 3WJ strands and a control dsDNA were incubated at 37 ºC for 2 hours prior to PAGE separation.The gel was then visualised using a multiplex assay that enabled the emission of each strand to be imaged independently (Figure 4a).The colour applied to each fluorophore enables the visualisation of each sequence, and the subsequent overlay of images indicates the presence of two or more strands where blue (FAM) and red (Cy5) produce purple hybridised strands, and green (ROX) and red (Cy5) produce orange pseudo-duplexes.The 3WJ can be visualised as pinkish-yellow colour from hybridisation of blue (FAM), green (ROX), and red (Cy5).This assay was also applicable for the analysis of the non-canonical structures formed by the HJ strands.Here, the four sequences (B, H, R and X, shown in red, yellow, green, and blue, respectively) were analysed and their ability to form pseudoduplexes.As expected, not all strands form duplexes (Figure 4b) due to their complementarity.For example, B-R and H-X do not form pseudo-duplexes as they are not designed to bind within the HJ.On the contrary B-H, B-X, H-R and X-R, which have complementarity, can form pseudo-duplexes, and can be visualised by the overlaid colours-orange (B-H), pink (B-X), light green (H-R) and turquoise (X-R)-in these samples (Figure 4b).Finally, when the four HJ sequences (B, H, R and X) were incubated together we observed a light-yellow band for formation of the 4-way HJ, which is an overlay of the four fluorescent signals (Figure 4b).B-H-R-X).AF350 (H), FAM (X), ROX (R) and Cy5 (B) filters were employed for visualisation of each sequence (Figure S2) and the structures they form.

DNA binding studies
Next, the DNA binding profile of a peptide helicate and a bis-acridine derivative with 3WJ and HJ DNA, respectively, was analysed by native PAGE.The ΛΛ-Fe II 2 LLD helicate was previously studied by Gómez-González et al. (2021) and 3WJ-binding was reported. 17,18Similarly, the binding of BA-C6 29 with HJ DNA was reported by the Cardin 26 and Searcey 27 groups, who performed X-ray crystallography and PAGE analysis to evaluate the interactions of BA-C6 with the HJ.
The ability of ΛΛ-Fe II 2 LLD to bind with the labelled 3WJ was assessed.Here, 3WJ DNA was incubated with ΛΛ-Fe II 2 LLD (1-1000 equivalents) at 37 ºC for 2 hours and the samples were analysed by PAGE (Figure 5a).The 3WJ forms in the presence of ΛΛ-Fe II 2 LLD and is observed as a pinkish-yellow band-arising from an overlay of FAM (blue), ROX (green) and Cy5 (red) fluorescent signals.At higher ΛΛ-Fe II 2 LLD concentrations, the intensity of the 3WJ decreases in a manner consistent with condensation of the 3WJ.
BA-C6 was previously shown to bind with and induce HJ DNA formation 26,27 and here this analysis is expanded to multiplex gel electrophoresis with samples prepared in water rather than buffer-for clear visualisation of the HJ.The fluorophore labelled HJ strands (B-Cy5, H-AF350, R-ROX and X-FAM) were incubated in water (without salt or buffer) and with increasing concentrations of BA-C6, HJ formation and then condensation was observed.The HJ can be observed across all samples but the intensity of the band increases with additional BA-C6, and, at 20 equivalents an upward shift for the HJ is observed (Figure 5b).This signifies binding of BA-C6 and a change in HJ conformation.A change in the pseudo-duplex band is also observed as the DNA adopts the junction conformation.In the presence of higher concentrations of BA-C6 we observed DNA condensation, which is not surprising as multiple bisacridine molecules have bound with HJ DNA and formed aggregates.Finally, band densitometry was performed with the Cy5-labelled HJ to determine the relative concentration of HJ present in each sample (supplementary Figure S5).An increase in HJ was observed, reaching its maximum at 30 equivalents of BA-C6, followed by a stepwise decrease in HJ DNA from 40-250 equivalents, which appears due to DNA condensation.

Microscale thermophoresis
Thermophoresis describes the movement of a molecule through a temperature gradient, and this movement is dependent on the size, charge, and hydration shell of the molecule. 323][34][35][36][37] MST is performed in thin glass capillaries and requires one partner in the interaction (i.e.ligand or target molecule) to be fluorescent, 37 which in this study is the junction DNA.Cy5-labelled 3WJ (Figure 6) and HJ DNA (Figure 7) were designed, with Cy5 attached to a short poly-thymine linker.

Binding constant of ΛΛ-Fe II 2 LLD with 3WJ DNA
To evaluate the binding affinity of ΛΛ-Fe II 2 LLD with the 3WJ, a serial dilution of the Fe II helicate-starting at the highest concentration at 50 μM-was performed in MST buffer (0.1% Tween-80 HEPES buffer, pH 7).Cy5-3WJ (20 nM) was added and the samples were incubated at 37 ºC for 5 min.Samples were then loaded into glass capillaries and MST was performed.Aggregates were observed at higher concentrations of ΛΛ-Fe II 2 LLD, an effect which was also observed during the gel electrophoresis experiments (Figure 5a).However, DNA condensation impacts MST analysis as it reduces the fluorescent signal, so methods to limit aggregate formation were tested.Buffers at different pH, concentration ranges, and incubation times were investigated, and it was found that 0.1% Tween-80 within HEPES buffer was the most suitable.While condensation of the 3WJ DNA was still evident, it was reduced to a level compatible with MST.Since aggregates were present, the fluorescent signal was not uniform across the tested sample range (supplementary Figure S6), and the MST traces could not be used to determine the dissociation constant (K d ).Consequently, the K d was calculated from a change in initial fluorescence (Figure 6).This change in initial fluorescence of Cy5-3WJ DNA treated with ΛΛ-Fe II 2 LLD was plotted and the K d value was calculated as 3.033 x10 -8 M. Controls of LLD peptide ligand, an Fe II salt, and 3WJ alone were also evaluated but no significant binding was detected.
Next, the binding of ΛΛ-Fe II 2 LLD with Cy5-dsDNA was investigated.Here, two different experiments were performed: i) ΛΛ-Fe II 2 LLD and Cy5-dsDNA; and ii) ΛΛ-Fe II 2 LLD and Cy5-dsDNA in the presence of 3WJ.In the second experiment the 3WJ was not fluorescently labelled and therefore binding to the 3WJ would not be detected (or would be silent) by MST and determined indirectly by changes in the Cy5-dsDNA sequence present.Samples were prepared as previously described and the K d was evaluated using initial fluorescence.Binding was observed for ΛΛ-Fe II 2 LLD with Cy5-dsDNA (K d = 3.743 x 10 -9 M) in the absence of 3WJ junction.However, when Cy5-dsDNA and the non-fluorescent 3WJ were both incubated with ΛΛ-Fe II 2 LLD, no binding of Cy5-dsDNA was detected (Figure 6).This shows that although the Fe IIhelicate can bind with both 3WJ and dsDNA, the helicate preferentially binds with 3WJ DNA, indicating that ΛΛ-Fe II 2 LLD is a selective binder of replication foci in cells.This result is therefore in excellent agreement with earlier reports. 17  .This demonstrates that ΛΛ-Fe II 2 LLD can bind both 3WJ and dsDNA, but that is preferentially binds with the 3WJ.

Interaction of BA-C6 with HJ DNA
MST was also employed to evaluate the interaction of BA-C6 with HJ DNA.A serial dilution from 50 μM BA-C6 was performed, Cy5-HJ DNA (20 nM) was added, and the samples were incubated at 37 ºC for 1 hour.MST was performed, but in a similar manner to the Fe II helicate, BA-C6 induced DNA condensation at higher concentrations.Mitigation efforts to reduce DNA condensation were investigated including, lower drug-DNA incubation times, detergents and systems comprising methyl cellulose, but these had little impact on the DNA condensation effects observed.Consequently, the change in initial fluorescent intensity was also employed to investigate the binding of BA-C6 with the HJ (Figure 7).Control experiments including a mono-acridine type ligand, 9-amino-acridine, and dsDNA were also performed and compared to BA-C6 with HJ DNA.Here, the change in normalised fluorescent signal were plotted and fit to the Hill model for ligand-target interactions (Figure 7).EC 50 values of 4.359 x10 -7 M and 2.406 x10 -7 M were obtained for the interaction of BA-C6 with Cy5-HJ DNA and a Cy5-dsDNA control, respectively, which shows that BA-C6 binds HJ and dsDNA with similar affinity.Previous studies of BA-C6 with HJ [26][27][28] and dsDNA 29 indicate that the compound can also bind dsDNA, with many acridine derivatives displaying similar properties.Furthermore, in this work a 9amino-acridine monomer was screened for its HJ-binding interaction, but at the same concentration range it failed to fully bind to the HJ and the data was subsequently unable to fit using the Hill model (Figure 6).Open-X HJ is incubated with BA-C6 and upon binding it induces the formation of stacked-X HJ (Iso I or Iso II can form).BA-C6 induces changes in initial fluorescence when incubated with Cy5-HJ (red circle) and Cy5-dsDNA (blue square), but 9-NH 2 -acridine (green triangle) is not capable of binding with Cy5-HJ under the same concentration range and the profile observed is comparable to HJ alone (yellow hexagon).

Conclusion
A multiplex fluorescent PAGE assay was combined with microscale thermophoresis to elucidate the recognition of 3WJ and HJ DNA binding agents.ΛΛ-Fe II 2 LLD and BA-C6 were selected for this study as earlier work 17,18,[26][27][28] identified their 3WJ and HJ DNA recognition.Here, we reported the design and application of a multiplex PAGE assay that enables the independent visualisation of each strand of junction DNA.This multiplex assay significantly expands the resolution of previous fluorophore-labelled PAGE experiments enabling the elucidation of both the junction formation and critical intermediates.6][37][38][39][40] To our knowledge, these results are the first examples of MST being used to probe the binding affinity of 3WJ-and HJbinding agents.The adaptation of the MST conditions to overcome ligand-induced fluorescent changes will enable the application of MST for a wider range of ligandtarget interactions.In summary, the multiplex PAGE and MST assays reported here demonstrate the selectivity of ΛΛ-Fe II 2 LLD and BA-C6 for binding with junction DNA and can aid the discovery and design of future therapeutics targeting non-canonical nucleic acid structures.

Figure 1 .
Figure 1.Crystal structure of 3WJ and HJ DNA.3WJ and HJ were adapted from PDB: 1F44 and 3CRX, respectively.

Figure 2 .
Figure 2. Single strands of HJ DNA and the open-X HJ are in equilibrium in water, but the addition of MgCl 2 enables the HJ to adopt the stacked-X forms.

Figure 5 .
Figure 5. Native PAGE analysis of the binding of ΛΛ-Fe II 2 LLD and BA-C with 3WJ and HJ DNA, respectively.a) Lane 1: Y1; Lane 2: Y2; Lane 3: Y3; Lanes 4-12: 3WJ (Y1-Y2-Y3) with 1-1000 equivalents of ΛΛ-Fe II 2 LLD.DNA condensation is observed after 25 eq.Fe II helicate with complete condensation at 100 eq.Individual images can be found in Figure S3.b) Lane 1: B; Lane 2: H; Lane 3: R; Lane 4: X; Lane 5: Negative control of HJ (B, H, R and X) in H 2 O; Lane 6: Positive control of HJ (B, H, R and X) incubated in phosphate buffer (10 mM phosphate, 150 mM NaCl, 2 mM MgCl 2 ); Lanes 7-12: HJ (B, H, R and X) with 0.5-40 equivalents of BA-C6.HJ formation and condensation were observed in the presence of the bis-acridine derivative, with almost complete condensation in the presence of 40 eq.BA-C6.The control experiments were performed as a comparison for the profile of ssDNA (B, H, R and X), pseudo-duplexes (control 1), open-X HJ DNA, and the unbound stacked-X HJ in the presence of salts (control 2).

Figure 6 .
Figure 6.Binding affinity of ΛΛ-Fe II 2 LLD with Cy5-3WJ and Cy5-dsDNA.Top: The change in initial fluorescence of Cy5-3WJ across a serial dilution of ΛΛ-Fe II 2 LLD was plotted (red circle) and the K d determined.Controls with LLD (blue square), Fe II SO 4 (green triangle) and 3WJ-alone (yellow hexagon) were also plotted.Upon binding of ΛΛ-Fe II 2 LLD with Cy5-3WJ a decrease in fluorescence was observed, no significant change in raw fluorescence was detected for LLD or Fe II SO 4 , which indicates that no binding has occurred with the Cy5-3WJ.Bottom: Change in initial fluorescence of Cy5-3WJ (i), Cy5-dsDNA (ii) and Cy5-dsDNA with non-labelled 3WJ (iii) in the presence of ΛΛ-Fe II 2 LLD.The Fe-helicate binds with Cy5-3WJ (red circle) and Cy5-dsDNA (blue square) with similar affinities (change in initial fluorescence), but ΛΛ-Fe II 2 LLD does not bind with Cy5-dsDNA in the presence of 3WJ (green triangle).This demonstrates that ΛΛ-Fe II 2 LLD can bind both 3WJ and dsDNA, but that is preferentially binds with the 3WJ.

Figure 7 .
Figure 7. MST analysis of BA-C6 and controls with Cy5-HJ DNA.Open-X HJ is incubated with BA-C6 and upon binding it induces the formation of stacked-X HJ (Iso I or Iso II can form).BA-C6 induces changes in initial fluorescence when incubated with Cy5-HJ (red circle) and Cy5-dsDNA (blue square), but 9-NH 2 -acridine (green triangle) is not capable of binding with Cy5-HJ under the same concentration range and the profile observed is comparable to HJ alone (yellow hexagon).