Title: By: Date:

A detailed literature review of G-quadruplex ligand binding assays 3rd yr M.Pharm project Jamie Al-Nasir & Owez Madhani Kingston School of Pharmacy and Chemistry & St. Georges Hospital Medical School 14/12/2009

Abstract

G-Quadruplexes are secondary nucleic acid structures formed in guanine rich regions.They comprise of stacks of square-planar hoogsteen bonded guanine nucleotides known as
G-tetrads. G-Quadruplex conformations can be found to form in telomeric and oncogene regions and are implicated in the expression of growth factors. Hence they are potential targets for therapeutic agents. The search for these ligands that bind and stabilise
G-Quadruplexes is therefore an important one. To assess if these ligands are ideal drug candidates various ligand binding assays may be carried out. Such assays yield information on binding affinity, selectivity, stoichiometry, and conformation for ligand and G-Qudruplex interactions. In this literature we outline the fundamentals of SPR, PCR-stop, ESI-MS, Circular dichroism, FRET, TRAP, G4-FID and TRE assays with respect to G-Quadruplex ligand binding.


Introduction

G-Quadruplex type structures consists of Guanine rich nucleic acid first discovered by Davies et al in 1962.¹ In this paper we will look at significance of this type of structure and discuss various assays used to study its’ ligand binding potential. According to Dr Julian Huppert a Computational biologist at Cambridge university, “Nucleic acids are capable of forming a wide variety of different structures, far removed from the Watson-Crick double helix…Many of the alternative structures that can be formed have physiological functions, such as controlling gene expression via gene transcription or translation.” (Huppert, 2006). G-Quadruplexes are such structures and their importance is implicated in a variety of intracellular processes such as gene expression, telomerase regulation and signal transduction.

Introduction to the G-Quadruplex structure

The G-Quadruplex (i.e. Guanine-Quadruplex) type structure consists of a four-stranded nucleic acid formed from guanine nucleotides. As we shall see later G-Quadruplexes can form from a single strand or from multiple strands. A repeating motif in this four-stranded structure is the

 respectively. The constituent tetrads stacked on top of each other can co-ordinate metal cations such as Na⁺ and K⁺ as each pair of tetrads contains eight oxygen atoms. It has been shown that G-Quadruplexes exhibit conformational polymorphism that is influenced by the polarity of the parallel strands.

Figure 1.2 above depicts possible types of G-Quadruplex structure, intermolecular G4-DNA comprised of different strands (A and C), and intramolecular G4-DNA where the bonding occurs on the within the same strand (B).

Hoogsteen bonding

(Figure 2.0) - Watson-Crick Base pairing

(Figure 2.1) - Hoogsteen Base pairing

G-Quadruplex DNA: Potential therapeutic targets

Telomere activity

G-Quadruplexes are significant as there are numerous guanine rich regions in the genome that are involved in regulatory functions. For instance telomeres located at the terminal ends of chromosomes are non-coding portions of DNA rich in Guanine where G-Quadruplexes form. These telomeric regions of DNA shorten on each replication. However the number of successful replications before the telomeres are depleted, and the cell cycle stops is finite, and is known as the Hayflick limit. Telomerase counteracts this senescence (ageing) process by repairing the telomeres, by adding repeats of TTAGGG. This operation increases the number of divisions the DNA may take part in. In normal cells telomerase expression and hence activity is low and thus limits the life-span of the cell’s line.^[1]. However in cancerous cells, where DNA replication is uncontrolled, telomerase expression and activity are increased. As telomeres are rich in Guanin, G-Quadruplexes form and these are potential targets for ligands that bind the Quadruplexes and interfere with telomerase activity thereby reducing tumour growth.

The G-Quadruplex structure as a therapeutic target in the treatment of cancer is an exciting possibility given the extreme cytotoxicity and un-specificity of traditional alkylating anti-cancer agents. ^[1]

The telomeres contain various conformations of G-Quadruplex nucleic acid and this is of consequence to ligand binding. Conformation is related to environmental conditions and ¹²⁵I-radioprobe studies reveal chair and propeller conformations in K⁺ solution whilst NMR studies show that basket and anti-parallel basket type conformations predominate in Na⁺ solution. The basket conformation exists in both ionic solutions. Since intracellular K⁺ is general high compared to that of Na⁺ studies of Q-Quadruplex conformation K⁺ is therefore thought to be of more biological relevance.^[4]

Oncogene transcription

Oncogenes are genes, which if sufficiently expressed, change a normal cell into a cancerous one in which apoptosis (pre-programmed cell death) does not occur and DNA replication continues uncontrollably.

The c-myc oncogene contains a region known as NHE III (nuclear hypersensitivity agent) which responsible for controlling 85-90% of transcription activation.
Stabilising G-Quadruplexes in this region with ligands has the effect of inhibiting transcription and therefore decreasing c-myc expression.

Proto-oncogenes such as bcl-2 are precursors to oncogenes, which in their normal state or their normal expressed level constitute part of the normal cell cycle and are not oncogenic. However, when altered or over expressed they result in a cancerous state of altered senescence or inhibited apoptosis which constitutes a lack of control over DNA replication. The oncogene Bcl-2 is yet another guanine rich region and is capable of forming a mixture of G-Quadruplex structures which are potential therapeutic targets. ^[1]  

Growth factors

VEGF (vascular endothelial growth factor) is involved in the angiogenesis of tumour growth, that is the vasculisation that supports tumour growth. This is another potential therapeutic target as the formation of G-Quadruplexes are implicated in VEGF expression. ^[1]

G-Quadruplex ligand binding assays

Now that the potential clinical significance of G-Quadruplexes has been established it is logical to study the interactions between this type of structure and it’s ligands. There are a variety of methods available, some utilise a range of routine analytical techniques, others involve more specific protocols.

Calorimetric techniques: ITC (Isothermal titration calorimetry)
and DSC (Differential scanning calorimetry)

Calorimetric techniques such as ITC and DSC provide a means of quantifying the thermodynamic properties and processes of G-Quadruplex ligand systems. The characteristic conformational polymorphism of G-Quadruplexes, which may alter on binding or un-binding is a physico-chemical process that can be monitored by DSC. For instance in DSC a sample of a G-Quadruplex-ligand complex is heated and any resulting conformational change results in a corresponding change in specific heat of the system, ΔCp which is measured and provides useful quantifiable data.

Stability data can be obtained from DSC because an equilibrium exists between different G-Quadruplex conformations or G-Quadruplex-ligand complexes and stability is proportional to the amount of heat required to alter the conformation. The transition mid-point, T_m is the temperature at which the equilibrium is a 1:1 mix of each conformation. ^[3]

Whilst DSC monitors the temperature changes of a single analyte as a whole (i.e. the G-Quadruplex-ligand complex), ITC works by introducing the two binding species, the ligand and the G-Quadruplex separately. On molecular binding of the species, heat is either absorbed or released and this is measured to provide detailed thermodynamic data in the form of an Isotherm which can be used to calculate various parameters such as binding enthalpy, K_b, binding stoichiometry, Gibbs free energy, and entropy change.

Polymerase Chain Reaction assays

PCR, Polymerase chain reaction, is a pioneering method used to synthesise DNA without the requirement for bacterial cells. (Karp cell biology, 2004) Developed in 1983 by Nobel laureate Kary Mullis, PCR is based on a similar polymerase reaction first proposed by Nobel laureate H. Gobind Khorana and Kjell Kleppe in 1967. Both methods involve a polymerase enzyme that generates DNA from single stranded oligonucleotide primers which are complementary to the DNA that is to be synthesised. PCR utilises Taq polymerase isolated from Thermus Aquaticus bacteria (a thermophilic species), and uses thermal cycling is used to “melt” the DNA into complementary single strands that are again replicated by the polymerase. Thermal cycling followed by the polymerase reaction results in an exponential production of DNA, and hence it is known as an “amplification” type procedure.

The PCR process consists of the following stages: -

• Denaturation

  • Heating ~93 °C for 20–30 seconds

  • cleaves hydrogen bonds between strands of DNA, yielding single stranded DNA

• Annealing

  • Temperature lowered to 60 °C for 20–40 seconds

  • Annealing of Primers complementary to the single stranded DNA

  • Polymerase binds and DNA synthesis commences

• Elongation

  • Temperature increased and held at optimum temperature for the polymerase used, normally 72 °C for Taq polymerase.

  • Synthesis of the DNA strand occurs in the 5’ to 3’ direction

  • Time taken to elongate the fragment is dependent on the length of the primer and of the polymerase and is normally in the order of 1kb/minute.

• Final elongation (after cycling of earlier stages is complete)

  • Temperature held at polymerises optimum temperature for 5-15 minute

  • To ensure that all fragments are fully elongated.

PCR-Stop assay

There are numerous derivations of the polymerase chain reaction, and one assay that is of particular significance to us is the PCR stop assay. PCR stop is used to study interference of ligands on polymerase activity. PCR stop assays are therefore an important means of assessing the binding of stabilising ligands on G-Quadruplex DNA. If bands of paused polymerase activity are found at guanine rich segments of electrophoresis assays on PCR products this indicates efficacy of ligand stabilisation, and the intensity of these paused bands is proportional to the inhibition of polymerase activity.

::Example of a PCR stop assay

A reaction mixture consisting of Template DNA and primers (Texas red labelled primer) were heated to 95 C for 3 minutes in a reaction buffer (10mmol Tris-HCl, pH 8.30, 50mmol KCl, 1.5mmol MgCl2). This was then allowed to cool for 30 minutes. The ligand to be assayed was added and the reaction mixture incubated at ambient temperature for 1 hr. Taq polymerase was then added together with dNTPs and incubated at 48 for 30 minutes. Reaction products were then sequenced using an automatic sequencer. (ligand tandem)

TRAP (Telomere Repeat Amplification Protocol) assay

The TRAP assay utilises PCR and quantifies the activity of telomerase, which is directly proportional to the amount of TTAGGG repeats added to the telomere. TRAP uses two primers, M2 (also known as Telomerase substrate) which is the forward primer required for telomerase addition of TTAAGGG repeats, and CX which is the reverse primer.

M2 (TS) Primer:          5’-AAT CCG TCG AGC AGA GTT-3’
CX Primer:                  5’- CCC TTA CCC TTA CCC TTA CCC TAA-3’

The annealing of the repeats occurs at 25°C and PCR is then used to amplify the products. The resultant products are then electrophoresed using PAGE (Polyacrylamide gel). (childrens medical research institute)

::Example of a trap assay

Procedure

1.      Prepare reaction mixture according to above specifications.

2.      Aliquot 48 μl into thin-walled PCR tubes.

3.      Add 2 μl of cell or tissue extract (1 μg/ μl) per tube.

4.      Allow telomerase reaction to proceed for 30 minutes at R.T.P.

5.      Overlay with oil and place the tubes in PCR machine and run following thermal cycling program:

The PCR products are then electrophoresed on Sybergreen stained Polyacrylamide gel.

SPR (Surface plasmon resonance) assay

Is a type of photoelectric analysis technique exploiting surface plasmons, electromagnetic waves that occur near the surface of an adsorptive surface (~300nm). These waves alter the refractive index of the solution near the sensor and the extent of this modulation is directly proportional to the number of molecules within the vicinity. One of the ligands is immobilised at the sensor surface whilst the analyte is injected, as the analyte binds to the ligand the concentration of molecules increases around the sensor, thereby altering the refractive index. SPR is a fast and sensitive technique useful in screening libraries of small ligands and is ideally used to characterise interactions between ligands and macromolecules such as
G-Quadruplexes.^[1]

The equipment generates a plot of response vs time known as a sensorgram and another corresponding sensorgram is created for background activity. The sensorgram of background activity is then subtracted from the first in order to yield a measure of the effect of the binding interaction on the response (the change in refractive index). A sensorgram usually has a characteristic increase in response as the species bind known as the association phase and a corresponding decrease, the dissociation phase when the species depart from the sensor surface.

The unit of measurement for Response is the aptly named Response Unit, RU and a single RU is equivalent to the binding of 1pg of protein per mm² of sensor.

SPR is able to analyse equilibrium measurements such as binding affinity and enthalpy. G-Quadruplex-ligand interaction affinity constants determined by SPR also generally correlate well with other methods such as Fluorescent titration, TRAP and thermal melting studies. However equilibrium analysis requires multiple and sequential injections of analyte at different concentrations, and this limits its practical use to ligands that establish equilibrium within around 30 minutes. Analytes having a low dissociation constant (K_D < 10 nM) are highly affinitive and have very slow rates of dissociation, Koff and are therefore unsuitable for analysis. The use of SPR for equilibrium analysis is thus practically restricted to those ligands with a high dissociation constant (K_D > nM).^[2]

SPR provides data on the stoichiometry of the G-Quadruplex-ligand interaction. According to Redman, "The maximum response observed when all surface binding sites are saturated is proportional to the mass of bound analyte, which in turn is proportional to the molecular weight of the analyte, the number of binding sites per immobilized ligand, and the surface density of the ligand. The expected maximum response, RU_max, for every molecule of analyte bound per molecule of ligand can be calculated from the response obtained during loading of the chip with ligand, and the molecular weights of the ligand and analyte.".

Ligands designed to bind with Quadruplex DNA are often cationic and aromatic so as to stack with the terminal Tetrad segments of the Quadruplex. These properties post problems for SPR in that the solubility of aromatic compounds in aqueous solutions is problematic and the SPR sensor requires a predominantly aqueous solution. However, the use of water imiscible co-solvents such as DMSO in small amounts (as much as 1%) is often employed to overcome the issue.

::Examples of Quadruplex structures that have been immobilised onto SPR sensors

• Human Telomeric Quadruplex

• Tetrahymena telomeric Quadruplex

• cMyc promotor Quadruplex

• G2T4 Basket Quadruplex

ESI-MS (Electrospray mass spectrometry) assay

Electrospray ionisation mass spectrometry is a powerful analytical tool with numerous applications. Whilst a detailed discussion of its applications are beyond the scope of this paper, it suffices to say that the technique is a “soft” form of mass spectrometry in which very little fragmentation occurs and non-covalent interactions are preserved in the process. This is of particular biological significance and facilitates the analysis of proteins where it is desirable to preserve these non-covalent interactions within the gas phase.

In ESI-MS, the analyte is very slowly injected into the apparatus by means of a syringe driver at a rate of around 1µl/min. The solution passes through a needle and plate between which there is a large potential difference that accelerates the sample. Solvent evaporation of the sample occurs and the droplets break apart, a process termed couloumbic explosion, when their surface tension cannot support their charge (known as the Rayleigh limit). The particles then usually possess multiple charges distributed around different sites of the protein and enter the mass spectrometer for separation according to m/z ratio. ESI-MS uses a minimal amount of sample, is a fast technique and yields the stoichiometry of the interaction making it very useful tool for bimolecular analysis between two species.[bristol]

::Example of an ESI-MS assay ^[3]

The binding affinity and stoichiometry of three ligands (depicted below) to G-Quadruplex DNA of the human telomeric sequence AGGGTT was assessed investigated. The buffer used for the procedure was NH4Oac.

Equipment and apparatus settings
Finnigan LCQ XP Plus ion mass trap spectrometer
Infusion rate:               2μL/min
Spray Voltage:             2.0-2.5kV
Temperature:              100 C with a double sheath gas
Software:                    Xcalibur software

where

• Im = N-methylimidazole

• γ = γ-aminobutyric acid

• β = β-alanine

• Dp = N,N-dimethylpropyldiamine
  
  

The Quadruplex was mixed with the three ligands in molar ratios ranging from 1:1 to 1:8. Binding affinity was evaluated by comparing the ratio of abundances of the [complex] to [Quadruplex DNA].

The Base peak at m/z 1514 is the Q5^- ion and represents the G-Quadruplex DNA.

The results of the ESI-MS assay indicate the preferential binding of the ligands to the
G-Quadruplex DNA to be Tel01 > ImImImβDp > PyPyPyγImImImβDp.

CD (Circular Dichroism)

Circular dichroism utiliseses plane polarised light that is rotated about it’s axis to create a helical light wave from the source. The ability of chiral of stereocenters is exploited and measured as a change in Molar Ellipticity from which structural and conformation information about the analyte can be deduced. CD is a useful tool for characterising G-Quadruplex structures and can provide information on binding stoichiometry of G-Quadruplex-ligand interactions. Circular dichroism depicts the changes in conformation that such ligand interactions may cause to the G-Qudruplex structure.

::Example of a CD assay utilising Papaverine derivative ligands ^[9]

Oligonucleotides used

-                          dG₃(T2AG₃)₃ (htel21)

-                          dAG₃(T2AG₃)₃ (htel22)

-                          d(T₂AG₃)₄ (htel24)

Ligands studied

 (figure 5.0) - Papaverine derived ligands

2.5 uM of oligonucleotide was placed in a TRAP buffer and the test ligand added in a drug:quadruplex ratio of 10:1. CD spectra measurements in the range of 220-550nm were repeated at 5um of G-Quadruplex oligonucleotide. Results were obtained from averaging three scans with a scan of the buffer solution subtracted.

CD Titration experiments were also carried out in TRAP buffer. The spectra recorded of the oligonucleotides before ligands were introduced is characteristic G-Quadruplex in K+ conditionsm and this is depicted by a “distintive shoulder” at ~270nm and a smaller negative peak at ~230nm. The relative intensities of each peak differs between each of the oligonucleotides, however the overal pattern is the same.

Introduction of the ligands to the oligonucleotides results in conformational change confirmed by an alteration to the CD spectra. For ligand 1, the peak at 250nm disappeared whereas for ligand 2 this same peak was preserved. The major positive band (the distinctive shoulder) near 290 showed an increase in molar ellipticity. The alterations in the spectra of
G-Quadruplexes suggest that they are modified by complexation of the ligands.

At least two conformations of G-Quadruplex are suggested to occur in K+ solution, namely the basket-type and anti-parallel. A hybrid form is also likely to occur and is characterised by a shoulder near 270nm.

Destabilisation of the basket structure was observed with ligand 1 as the 250nm peak disappeared.  Ligand 2 however was found to stabilise the basket structure as the 250nm peak was preserved.

FRET (Fluorescence Resonance Energy Transfer) melting assay

A FRET melting assay can determine the ‘affinity’ and ‘selectivity’ of ligands by measuring the increase in melting temperature of a quadruplex induced by the linkage of ligands to G4 DNA.

Principle of FRET:

FRET is a fluorescence-based spectroscopic method which provides distance based information on structural changes of macromolecules ¹³.

A typical FRET experiment involves covalently attaching a polymer with two fluorophore probes, a donor and an acceptor. The pair should be such that the emission spectrum of one probe (donor molecule) coincides with the absorption spectra of the other (acceptor molecule). In which case, fluorescent energy is transferred from the donor to the acceptor in a non-radiational manner. This transfer is only possible when the pair is within a certain distance from each other. This distance within which the non-radiative energy transfer is possible is known as the ‘Förster distance’ (R₀) and is characteristic for the donor-acceptor pair. It is usually within 10 - 80 ⁰A ¹⁴.

When the probes are within this distance the fluorescence emission of the donor is lower than when they are further apart (because the energy is transferred to the acceptor). Now when the distance increases such that R > R₀, the energy would no longer be transferred to the acceptor and the energy peak in the emission spectra of the donor increases accordingly ¹⁴.

This principle is applied in a ‘FRET melting assay’ where increase in temperature leads to denaturing (melting) of the macromolecule. Denaturing causes the distance between the probes to increase leading to an increase in fluorescent energy.

::The FRET melting assay

Materials:

To conduct a FRET melting assay, a G-quadruplex forming oligonucleotide is labelled with a pair of fluorescent probes (donor-acceptor pair). To explain the working of the assay we have selected the fluorescent dye ‘fluorescein’ (FAM molecule) as our donor and ‘tetramethylrhodamine’ (TAMRA) as an acceptor dye and a single stranded guanine rich nucleotide chain with 21-bases as our quadruplex forming oligonucletide (QFO). Their structures are shown below:

5′-d-GGGTTAGGGTTAGGGTTAGGG3

(Fig 6.0) Structure of the QFO, donor and the acceptor molecules. (Image taken from Juskowiak B.14)

FAM and TAMRA are attached to the 5’ and 3’ ends of the oligonucleotide respectively as shown in the figure below. This FAM-TAMRA dually labelled oligonucleotide is called ‘F21T’.

(Fig 6.1) Structure of the dual probe labelled QFO. (Image taken from  Juskowiak B.1[4])

Using the FRET system shown above we can measure the distance between the donor (FAM) and the acceptor (TAMRA) probes in terms of the fluorescence of the donor (FAM).

Working:

As the theory states, at low temperatures the oligonucleotide is folded in a quadruplex structure and thus the distance between the probes falls within the Förster radius (R₀). This causes the emission radiation of FAM being quenched by TAMRA.

Now as the temperature is increased closer to the melting point (denaturation) of the oligonucleotide, it unfoldes and the distance between the probes increases beyond the förster radius (R₀) and the quenching effect of the donor emission (by the acceptor) decreases. This leads to increased donor emission.

It can be seen from the diagram that as the distance increases (or as the temperature, unfolding of the oligonucleotide increases) the fluorescence energy of the donor increases.

In order to quantify the effects of temperature, the fluorescence energy measurements of the dual probe coupled oligonucleotide are conducted with step wise increase in temperature.

(Fig 6.3) Graph on the left shows the increase in fluorescence with increasing temperature. Graph on the right is similar although the emission values have mean normalised to the scale of 0 to 1

It is somewhat impractical to calculate a true melting point for macromolecules like an oligonucleotide. Thus we use T ½ for comparison studies. T ½ is the temperature value for which the normalised emission is 0.5. It can be seen from the graph that its value for ‘F21T’ is approximately 53⁰C.

Ligand binding:

Now on addition of putative ligands which stabilise the folded G-quadruplex structure, the melting point (or the denaturation temperature) of the ligand bound oligonucleotide should increase. Thus the T ½ values should increase.

The extent of change in T ½ values (∆ T ½ ) on addition of ligands to F21T from the values obtained from F21T alone are indicative of stabilisation of the G-quadruplex structure due to ligand binding. The higher the ∆T ½ value for a ligand, the better is its affinity.

For the purpose of reviewing we have selected the following shown ligand:

(Fig 6.4) Structure of the ligand used for the FRET melting assay (Image modified from Cian A D. et al 13)

After conducting controlled melting experiments as demonstrated above with mixtures F21T and the above shown ligand at different concentration strengths, the following results are obtained:

(Fig 6.5) Graph showing the extent of melting temperatures for F21T due to binding to ligands. Here the same ligand is used at different concentration strengths. Strengths moving from left to right following the curves are 0μM, 0.5μM, 1μM, 2μM, 4μM & 8μM. Double lines are from duplicate measurements. (Image modified from Cian A D. et al 13)

It can be seen from the graph above that the ligand concentration of 8μM leads to a ∆T ½ of (77 ⁰C – 53⁰C) = 24⁰C.

From literature, ligands causing ∆T ½ of >20⁰C at less than 1μM concentration are the best at stabilising G4 quadruplex. Our selected ligand (figure 6.4) require more than 3μM to cause ∆T ½ of 20⁰C. Thus it is not as effective, however, higher concentrations of 8μM as seen above give 24⁰C stabilisation. 

Ligand selectivity¹³:

Ligand selectivity is an important factor to consider when selecting a ligand as a potential drug target. FRET melting assays can also show ligand selectivity. It can be measured by adding a non-tagged (non-fluorescent) duplex DNA into the test solution containing ligand bound F21T oligonucleotide and then conducting the melting experiments.

Comparing the values obtained from this experiment with the values obtained from ones in which a competitor DNA is absent, one can measure the extent of ligand trapping (binding of ligand to the duplex DNA) by measuring the decrease T ½ values.

Advantages of FRET melting assays ^{13, 14, 15}:

·         It allows analysis to be conducted in a variety of ionic conditions.

·         The method is rapid and convenient.

·         It can easily be adapted for high-throughput screening.

Disadvantages of FRET melting assays ^{13, 14, 15}:

• We can only achieve semi-quantitative results for ligand affinities.

• The quadruplex forming oligonucleotide has to be modified.

• The method can generate false positives (quenching of the donor emission by the tested ligand) or false negatives (folding of the QFO with the ligand in such a configuration that the distance prevents quenching)

Newly proposed assays

G4-FID (G-quadruplex Fluorescent Intercalator Displacement) assay:

This assay is based on the displacement of a fluorescent probe by putative ligands from DNA matrices. This test measures the affinity by which the ligands bind and their selectivity over different types of DNA matrices.

The procedure measures the loss of fluorescence of the bound probe by its displacement due to a DNA binding ligand. For the purpose of describing the workings of this assay, we would select one ligand analogue and show how we can test its binding affinity to a quadruplex forming oligonucleotide (QFO) and its selectivity over duplex DNA.

The experimental data as presented below has been obtained from Monchaud D. et al ¹¹.

Materials:

•   Fluorescent probe: Thiazole Orange.
  This molecule binds to the quadruplex-forming oligonucleotide 22AG in a single-site manner, with high affinity (K_a = 3 × 10⁶ M⁻¹). Its fluorescent quantum yield is very low when free in solution (Φ_F = 2 × 10⁻⁴), however it increases by 500-1000 fold on binding to DNA. Hence its displacement by test ligands can be monitored by the reduction in TO fluorescence at λ_max = 539 nm.

• G-quadruplex structure: 22AG (a quadruplex forming oligonucleotide which mimics human telomeric sequence [5′-AG₃(T₂AG₃)₃-3′]) in K+ buffer

• Duplex DNA strand: It is a 17 base pair duplex stranded DNA (ds 17). The sequences of the two complementary strands are: [5′-CCAGTTCGTAGTAACCC-3′] & [5′-GGGTTACTACGAACTGG-3′]

• Ligand tested: A N-methylated quinacridine (MMQ16)

Experimental:

-  Titration of Thiazole orange (TO) with the solution of 22AG:

Titration is carried out in a 3 ml cell at 20⁰C in a 10nM sodium cacodylate buffer pH 7.3 and 100mM KCl. The addition of TO is followed by 22AG at 501nm excitation wavelength. Data is then collected by scanning a range of wavelengths of UV-Vis light from 510 to 750 nm. The graph obtained by plotting fluorescent intensity against wavelength (510 nm – 750 nm) would be similar to the one depicted in figure 1 below. The area under curve is FA₀ which is known as the Fluorescent Area; when no ligand exists in the test solution.

-  G4-FID assay (addition of MMQ16 to 22AG):

With similar experimental conditions as above (i.e. 3ml cell, 20⁰C temperature, buffer solution and KCl), 0.25 µM pre-folded 22AG is mixed with 0.50 µM TO. The ligand MMQ16 (different concentrations) is then added to the solution. After a 3 minute equilibrium period a fluorescence spectrum is obtained similar to figure 7.0 below. However, in this case the area under curve obtained (or the Fluorescent area, FA) is reduced due to the displacement of TO by the ligand MMQ16.

(Figure 7.0). Fluorescent spectrum of TO bound to 22AG obtained by scanning UV-Vis wavelengths from 510nm to 750nm. The scan shown is not up to scale and shows the area under curve or the fluorescent area (FA0, blackened) before the addition of the ligand MMQ16.

(Figure 7.1). Fluorescent spectrum of TO bound to 22AG obtained by scanning UV-Vis wavelengths from 510nm to 750nm. The scan shown is not up to scale and shows the reduction in the area under curve or the fluorescent area (FA, blackened) after the addition of the ligand MMQ16.

Results can similarly be obtained for ds 17 (Duplex DNA) following the above procedures.

Evaluation of the results:

From the results obtained from the assay, the percentage displacement of TO by the ligand at different ligand concentrations can be calculated as follows:

(Figure 7.2): Graph showing the % TO displacement from 22AG in K+ and ds 17 at different MMQ16 concentrations. Experimental conditions: [oligonucleotide] = 0.25 μM, [TO] = 0.5 μM for 22AG and TBA, 0.75 μM for ds26, cacodylate buffer. (Image modified from Monchaud et al.11)

Affinity of a ligand for the DNA matrix (22AG or ds 17) can be obtained as a concentration value at which the ligand leads to 50% of TO displacement. From the graph (Fig. 7.2), the affinity values for MMQ16 can be obtained as follows:

• Affinity to G-quadruplex forming 22AG is the concentration of MMQ16 which causes 50% of TO to displaced from the binding sites on the quadruplex. 
  The value from the graph is G4DC50 (22AG K+) = 0.14 µM.

• Affinity to duplex DNA (ds 17) can be similarly obtained from the graph. 
  Which in this case is dsDC50 (ds 17) > 2.5 µM (as it only causes 43.2% TO displacement)
  

Selectivity of MMQ16 for G-quadruplex DNA over duplex DNA can be obtained by the ratio of dsDC50 / G4DC50 values.

As from above, a G4-FID assay can evaluate a ligand’s affinity and selectivity to G-quadruplex DNA over duplex DNA. Affinity values show how strongly a ligand binds to the DNA whilst selectivity values show a ligand’s tendency to bind to G-quadruplex DNA over duplex DNA.

Advantages of G4-FID:

• The test doesn’t require modified oligonucleotides. Thus a variety of DNA matrices (different G-quadruplex conformations, duplex DNA etc.) can be tested ^[11]

• No specific requirements (Compounds used are readily available and equipments generally include a standard spectrofluorimeter for fluorescence measurements and a standard spectrophotometer for UV-Vis measurements) ^[12]

• The materials used in the assay are inexpensive and the procedure is neither technically demanding nor time consuming ^[11,12]

• A broad diversity of ligands can be evaluated by this assay. 

• Physiological conditions of temperature and cationic environment can be maintained during the assay ^[12]

• It can be applied to high-throughput screening as its run isn’t time consuming ^[12]

Drawbacks of G4-FID:

• One of the flaw is in its methodology. An indirect competition between the ligand and the probe when their binding sites are different leads to skewed results. The TO displacement values would be an underestimate of that ligand’s affinity ^[11]

• Absorption characteristics of the studied molecule should not overlap with the absorption or the emission spectra of the probe ^[11]

• G-quadruplex in telomere DNA can adopt numerous conformations which alters the binding sites and hence a ligand’s affinity in vivo ^[12]

TRE (Telomere Repeat Elongation)

Telomerse Repeat Elongation utilises SPR and quantifies the elongation of an immobilised primer on the SPR sensor.

::Example of a TRE assay

Telomeric sequence used:     5’-biotin-AATCCGTCGAGCAGAGTTAG(GGTTAG)4 

dNTP containing buffer:          (10 mM HEPES, pH 7.4, 10mM MgCl₂, 10mM NaCl, 2.5mM dNTP, 10mM EGTA)

The telomeric sequence was immobilised onto a Biacore SA sensor chip. Cell extracts that were telomerase positive were diluted in dNTP buffer and injected over the SENSOR at 5μL/min at 37°C for periods of 1, 5, 10 and 30 minutes to extend the telomeric oligonucleotide. The protein was then removed from the sensor by injection of 1% SDS in HEPES buffer. In order to quantify the extension of the oligonucleotide the baseline level was compared to that prior to the injection of cell extract. A measure termed the elongation factor was defined as the interval between the two aforementioned baseline responses. The elongation factor increased linearly with incubation time.^[Redman]

In theory the TRE assay could be used to quantify inhibition of telomerase activity of G-Quadruplex stabilising ligands. Additionally, as it does not utilise PCR amplification the TRE assay does not suffer from variations in efficiency arising from the binding of ligands to duplex DNA.  

Discussion

The assays mentioned herein involved a variety of analytical techniques that can be used to study G-Quadruplex ligand binding. Calorimetric techniques provide a thermodynamic profile of ligand:complex interactions from which various stability data can be obtained.

ESI-MS can be used to quantify the relative abundances of ligand, G-Quadruplex and ligand:Quadruplex in an interaction and clearly depicts preferential binding.

The interference and inhibition of Telomerase activity by potential G-Quadruplex ligands can be studied by PCR stop and TRAP assays. The TRAP assay whilst extremely useful suffers from reduced efficency that the PCR amplification component introduces as ligands may bind to duplex DNA produced by PCR. TRE is a new assay which utilises SPR and does not require PCR amplification.

Circular dichroism provides information on the effects of ligand interaction on the conformation of G-Quadruplex structures. On the addition of different ligands, the alteration in response (molar ellipticity) for a corresponding wavelength can be used to deduce these effect and characterise them as specific alterations in specific regions of the Spectra. From this information the effects of various ligands on the stabilisation and destabilisation of the G-Quadruplex can be studied. 

FRET and G4-FID assays both use fluorescence spectroscopy in their working. They involve attachment of fluorescent probes to the G-quadruplex forming oligonucleotide. G4-FID measures the decrease in fluorescence energy and the % displacement of the fluorescent dye whereas; FRET measures the increase in fluorescence on increasing temperature.

Both assays are tailored here to evaluate the binding properties of ligand molecules to
G4-DNA. Although these techniques only give semi-quantitative data (not giving real parametric values e.g. affinity constants), the information provided in invaluable to researches finding a potential target to stabilise the G-quadruplex DNA and help them chose the best ligands for further analysis. Both methods are easy to carry out and can be adapted to high-throughout screening of ligands. These screenings test several molecule libraries at the same time. G4-FID does not require specialised equipments; however, FRET may require several protocols to use the PCR apparatus to measure melting temperatures. FRET is a more commonly used technique when accessing G4-DNA ligand binding than G4-FID.

Selection of assay will depend on a number of factors such as the ligands to be tested,
G-quadruplex structure (i.e. inter- vs intramolecular), quantity of sample, ligand-quadruplex stoichiometry and the type of binding information sought.


References

1. Tian-miao et al, 2008 G-Quadruplexes: Targets in Anticancer Drug Design 2008. ChemMedChem, 3 , pp60-713

2. van der Merwe, P. A, Surface plasmon resonance. Oxford university 2009

3. MicroCal (2009) – ITC (Isothermal Titration Calorimeter) product catalogue

4. Jiang Zhou and Gu Yuan (2007) Specific Recognition of Human Telomeric G-Quadruplex DNA with Small Molecules and the Conformational Analysis by ESI Mass Spectrometry and Circular Dichroism Spectropolarimetry, Chem. Eur. J. 2007, 13, pp5018 – 5023

5. Bai L.P. et al Ligand Binding to Tandem G Quadruplexes from Human Telomeric DNA, ChemBioChem 2008, 9, pp2583 – 2587

6. Redman, J.E. Surface plasmon resonance for probing quadruplex folding and interactions with proteins and small molecules. Science Direct, Methods 43 (2007) pp302–312

7. Colgin, L. (2009) TRAP (Telomere Repeat Amplification Protocol). Children's Medical Research Institute, Australia

8. Gates, P. A, Surface plasmon resonance. Bristol university 2009

9. Monchaud, D. et al Ligands playing musical chairs with G-quadruplex DNA: A rapid and simple displacement assay for identifying selective G-quadruplex binders. (2008)

10. Galezowskaa, E. Spectroscopic study and G-quadruplex DNA binding affinity of two bioactive papaverine-derived ligands. International Journal of Biological Macromolecules 41 (2007), pp558–563

11. Monchaud D. et al. Ligands playing musical chairs with G-quadruplex DNA: A rapid and simple displacement assay for identifying selective G-quadruplex binders. Biochimie 2008; 90 (8): pp.1207-1223.

12. Cian A D. et al. Fluorescence-based melting assays for studying quadruplex ligands. Methods 2007; 42(2): pp. 183-195.

13. Juskowiak B. Analytical potential of the quadruplex DNA-based FRET probes. Analytica Chimica Acta. 2006; 568 (1-2) pp. 171-180.

14. Mergny J-L and Maurizot J-C. Fluorescence Resonance Energy Transfer as a probe for G-Quartet Formation by a Telomeric repeat. ChemBioChem. 2001; 2(2) pp. 124-132.