1 Introduction

G-quadruplexes (G4) are nucleic acid secondary DNA structures that have been found in G-rich regions of genome having biological significance such as human telomeres [1,2,3,4] and various oncogenic promoter regions, including c-myc [5,6,7], c-kit [8, 9], bcl-2 [10, 11] or RET, [12, 13]. The core structures in the G-quadruplex are two or three G-quartets, which can associate through Hoogsteen hydrogen bonding to form a square planar structure. In addition, the alkali metal ions such as Na+ and K+ are known to play important roles in the stabilization of G-quadruplex which coordinate with the eight electronegative O6 atoms of the adjacent stacked G-tetrads. Since K+ and Na+ are the physiologically relevant monovalent ions, G-quadruplex formation is favoured under physiological conditions. A G-rich sequence may adopt different structures in the presence of different cations. The K+ form in general is preferred over Na+ because of its better coordination with eight guanine O6s and a lower dehydration energy as well as its higher intracellular concentration (~ 140 mM) than that of Na+ (5–15 mM) [14]. G-rich DNA sequences can form intramolecular G-quadruplexes formed by single-stranded DNA and intermolecular G-quadruplexes between multiple DNA strands [15]. Different sequences can adopt distinct topologies, in addition, a given sequence can fold into a variety of different conformations, as in the case of the human telomeric DNA sequence G3(T2AG3)3. In addition, intermolecular multimeric G-quadruplexes can be formed by the association of two or more strands [16]. Intervening mixed-sequence nucleotides form loops of folded G-quadruplex structures which can adopt a variety of different topological forms [17, 18]. The polarity with respect to the two adjacent strands and the location and length of the loops would be expected to lead to a polymorphic G-quadruplex structures. When the alignment of guanine tracts is in same direction, the double-chain reversal (propeller) loops link two adjacent parallel strands to form a parallel structure [1]. When the alignment of guanine tracts is in opposite directions, the edgewise or diagonal loops link two antiparallel strands to form an antiparallel G-quadruplex [19]. When a single strand is oriented in a different direction from the others then it forms antiparallel hybrid or so-called mixed structures [20,21,22]. Recently, Marušič et al. reported a novel mixed type fold which exhibits a conformation in which all three loop types occur in one conformation: edgewise, diagonal, and double-chain reversal loops [23]. Intramolecular G-quadruplexes formed by single-stranded DNA are of intensive current research interest due to their potential formation in telomeres [24,25,26] and oncogene promoter sequences [27]. Intramolecular structures form quickly and are more complex, exhibiting great conformational diversity, folding topologies and their loop connectivity. With the extensive and thorough structural studies of G-quadruplexes, numerous rules for folding patterns for G-quadruplex have been recognized. However, to understand each G-quadruplex structure and its ligand interactions, the prediction of a G-quadruplex conformation is necessary although difficult. Reason is high degree of structural polymorphism regulated by various factors like length, base composition of G-rich sequence and physical factors like buffer, pH, and type of binding cation etc. [28,29,30,31,32]. G-quadruplex structures may form intramolecular G-quadruplexes within a single DNA strand, or intermolecular G-quadruplexes between multiple DNA strands [15]. Moreover, due to the various orientations of nucleic acids strands during folding, the G-quadruplex structures can further be divided into different conformations, including parallel, antiparallel, and hybrid conformations [22, 33,34,35]. Intramolecular G-quadruplexes are of intensive research interest currently due to their potential of forming G-quadruplex structure in biologically important regions. Telomeric DNA is a 3′ G-rich overhang of hexa-nucleotide repeat sequence 5′-(GGGTTA)n which forms highly polymorphic G-quadruplexes structures depending on the differences in the arrangements of loop, strand orientations, G-tetrad arrangements, and capping structures. Telomere is an attractive anticancer drug targets due to the stabilization of telomere G4s and blockage of telomerase activities and hence offers a new strategy for antitumor therapy. More importantly, telomerase activity is mainly expressed in aging, tumors, cell proliferation, and carcinogenesis that comes around 85–90%. Many researchers have explored many small synthetic drug molecules that induces the formation of G-quadruplex structure of telomere and stabilizes them or may act by binding with oncogenic promoter sequences and down regulate their expression [36, 37]. Synthetic drug molecules are very toxic and have many side effects. On the other hand, nature provides chemically distinct scaffolds in the form of various flavonoids like Quercitin, Myricetin and Genisten etc. which are readily available, less toxic and have better bioavailability [38, 39]. Therefore, from a drug-development point of view, research on screening the binding potential of quadruplex-interactive small natural compounds is expected to be highly active, with a particular emphasis on the selectivity toward polymorphic G-quadruplex structure. In the present study, we attempted to screen the binding potential of plant flavonol (quercetin) with different polymorphic telomeric sequences, cancer proto-oncogene targets and compared with RNA G-quadruplex using molecular docking approach. The purpose of the study is to select the most suitable binding mode of quercitin with respect to the G-quadruplex forming sequences as well as the conformation of G-quadruplex. Based on the calculated binding energies, our results suggest that quercetin binds with high affinity with parallel, antiparallel and mixed telomeric G-quadruplex structures. This highlights that quercetin is well suited natural ligand to target polymorphic telomeric quadruplexes. However, quercetin also binds with cancer protocogenes which indicates that the quercetin can be selected as a natural ligand to down regulate the gene expression of cancer proto-oncogenes. The binding energies for RNA G-quadruplex were in the range of − 14 kcal/mol which indicates that quercetin is not a suitable natural ligand to target the RNA G-quadruplexes in transcriptosome.

2 Materials and methods

2.1 Ligand preparation and virtual screening

The three dimensional structure of Quercetin (PubChem CID: 5280343) was downloaded from the PubChem Database. Schrödinger LigPrep module was used to generate possible states of quercetin based on variations on ring conformations and tautomers [40]. OPLS 2005 force field was used to generate minimized ionization states at the physiological pH of 7.0 ± 2.0, as it can influence its binding affinity during interaction studies. Keto-enol tautomerism along with analogous sulphur and nitrogen tautomerization was also done. Eventually, structures were desalted for removal of any water molecules or counter-ions and stereoisomers were generated by retaining the specific chiralities of the original state. A total of 32 structures were generated inclusive of the original state. Initially, all the 32 variants were screened against the pre-processed structure files using GLIDE (i.e. Parallel, Mixed & Anti-Parallel) [41]. The complexes with most negative G scores (ligand binding free energy), were aligned for voting out the best-suited conformation for telomeric G-Quadruplex structures (i.e. Parallel, Anti-Parallel & Mixed) and were isolated in SDF files for MMGBSA (binding energy) estimation.

2.2 Receptor grid generation

All the structure files for Parallel, Mixed and Anti-Parallel were curated from the Protein Data Bank (PDB) and were processed by the removal of excess unwanted waters and cations beyond 5 Angstroms from Hetro groups using Maestro [42]. Missing hydrogens were added and bond orders were assigned to stabilize the raw structure files from PDB. For the NMR structure files, comparative reconstruction using Modeller was done by using the possible states as templates [43]. 3-D structure files of each Quadruplex structure were first optimized at neutral pH and then were minimized using SwissPDB viewer to achieve energetically stable structure for docking [44]. The Gasteiger atomic charges were assigned for all the G4 structures. Receptor grid of 20 × 20 × 20 Angstroms was constructed around the complexes in three dimensions to be sufficiently large to encompass the entire G4 structures using the Receptor Grid Generation module [40].

2.3 Molecular docking and MMGBSA calculations

Schrödinger Glide with OPLS 2003 force field was used with Xtra precision workflow to carry out the Molecular docking [45]. The scaling factor was set to the default of 0.80 with the partial charge cutoff at 0.15 and root mean square deviations for ligand geometries were also computed. We flexibly docked categorically isolated (i.e. Parallel, Anti-Parallel & Mixed) conformations of the quercetin with their respective categorical structure files (i.e. Parallel, Anti-Parallel & Mixed). Docking score for every receptor-ligand was calculated for setting up the Molecular Mechanics generalized Born and surface area continuum solvation (MM/PBSA) estimation. MMGBSA was calculated in VSGB to estimate the free energy of the binding of quercetin to telomeric G-Quadruplexes [40, 41, 45,46,47]. After, sorting the ligand-receptor poses based on Gibbs Free Energy (deltaG), bound states of quercetin were again superimposed to find the similarity of binding conformation between the member of Parallel, Anti-Parallel & Mixed telomeric G-Quadruplexes structures independently.

All figures were rendered with PyMOL (www.pymol.org).

2.4 Circular dichroism spectroscopy

CD spectra were carried out on JASCO-715 spectropolarimeter using a quartz cuvette of 1 cm path length. All the spectra were recorded in the range of 200–350 nm wavelengths at a scanning rate of 100 nm min−1. Before measurement, the samples were heated to 95 °C in water bath and slowly cooled till water attains room temperature and incubated at 4 °C overnight to avoid any non-equilibrium structures. Average scans of the DNA samples were subtracted from the buffer scan and data was normalized as a function of DNA strand concentration and pathlength of the cuvette. The CD curve was plotted between ellipticity as a function of wavelength.

2.5 Thermal melting analysis

UV absorbance of different samples were recorded with a Shimadzu 1800 spectrophotometer (Shimadzu, Tokyo, Japan) equipped with a temperature controller. Melting curves of DNA structures were obtained by measuring the UV absorbance at 295 nm in buffer pH 7.0 [100 mM KCl, and 0.5 mM EDTA] in the presence or absence of quercetin at DNA:quercetin ratio (1:0), (1:1), (1:2), (1:5), and (1:10). The Tm values for 4 μM DNA structures were obtained from the UV melting curves as described previously17. The heating rates were 0.5 °C min−1. The thermodynamic parameters were evaluated from the fit of the melting curves to a theoretical equation for an intramolecular association as described previously [48, 49]. Before measurement, the samples were heated to 95 °C in water bath and slowly cooled till water attains room temperature and incubated at 4 °C overnight to avoid any non-equilibrium structures. Experiment has been repeated in triplicates to reproduce the data.

3 Results and discussion

Guanine-rich nucleic acid sequences can fold into diverse topologies depending upon the length of the G tract, loop length and the syn- and anti-conformation of Guanines. Notably, not only the sequences but also the surrounding conditions determine the conformation of DNA G-quadruplexes. In particular, different monovalent and divalent cations coexisting in solution are known to induce diverse G-quadruplex structures with distinct thermodynamics [1, 19, 20, 50]. Hence, first we retrieved the crystal structures of different telomeric sequences and their PDBs from NCBI database and segregated them based on the folding topology into parallel, antiparallel and mixed G-quadruplexes. To understand better the binding mode of quercetin in detail we have also retrieved the G-quadruplex sequences of available human proto-oncogenes as well as one sequence of RNA G-quadruplex. All selected DNA and RNA sequences along with their PDB ID’s are given in Table 1.

Table 1 Retrieval of G-quadruplex forming sequences of telomeres, cancer proto-oncogenes and RNA G-quadruplex from NCBI data base (https://www.ncbi.nlm.nih.gov/Nucleotide)
Full size table

3.1 Superimpose conformers of Quercitin

As we have chosen to compare the binding efficiencies of quercitin with different topologies of G-quadruplex. Therefore, to make it more systematic, first we tried to identify the superimpose conformers of quercitin for parallel G-quadruplex, anti-parallel G-quadruplex as well as for mixed G-quadruplex separately as shown in Fig. 1 using Maestro alignment tool of Schrodinger software. The basic idea was to identify structure of most relevant conformers of the quercitin molecule based on the smart patterns for the atoms with each distinct G-quadruplex topology and to understand the conformational change using the software based drug discovery approach.

Fig. 1
figure 1

All quercetin conformers in Parallel G-quadruplex sequences (a) Antiparallel G-quadruplex sequences (b) and Mixed G-quadruplex sequences (c)

Full size image

3.2 Comparative analysis of molecular docking studies of parallel g-quadruplex structures with quercitin

Molecular docking calculations was performed to investigate the binding mode of quercitin with Human Parallel telomeric G4 (1KF1) using Schrodinger. The details of atoms involved in hydrogen bonding between the parallel G-quadruplexes with Quercetin are summarized in Table 2. The crystal structure of 1KF1 was used as a template processed by removal of excess unwanted waters and cations beyond 5 Angstroms from Hetro groups using Maestro [42]. Here, during the initial optimization, the planarity restriction was not applied, and the final structures are obtained at intrinsic planar and nonplanar arrangements. The size of grid box was 20 × 20 × 20 in three dimensions. The grid was set to be sufficiently large to cover significant portions of the active sites and the result is depicted in Fig. 2a. After the optimization of the quercitin conformer for the parallel G-quadruplex, it interacted with G4. Results suggest that each G-tetrad is stabilized by Hoogsteen hydrogen bonding between hydrogen bond donor and acceptor atoms. Quercitin adopts a groove binding conformation in which hydrogen atom of the C-2 atom with A7, C-8 forms a hydrogen bond with T5. Another two hydrogen bonds formed between oxygen atom of C-13 and hydrogen atom of C-14 of quercitin with G4 and with G10 respectively (Table 1). In drug designing, the calculation of binding energies is an important parameter to evaluate the affinity of binding of the drug with G4. We calculated free energy of the binding of quercetin to telomeric G-Quadruplexes using the Molecular Mechanics generalized Born and Surface area which was found to be − 40.24 kcal/mol (Table 1). The negative sign of delta G indicates the favourable binding of quercetin with 1KF1 which is due to the formation of three hydrogen bonds formation during the interaction of quercetin with 1KF1. Next, we compared the binding mode of quercetin with 1K8P which is another parallel telomeric G-quadruplex formed in the presence of K+ [51] (Fig. 2b). As it can be seen in figure that quercetin binds in groove in which hydrogen atom of the C-7 atom of quercetin forms hydrogen bond with G3 and oxygen atom of C-8 atom. Second hydrogen bond is established between the hydrogen atoms of C-13 with G5. The estimated binding energies for the quercitin binding with 1K8P was estimated as − 33.06 kcal/mol (Table 1). Next, we checked the binding mode of another parallel telomeric G-quadruplex 352D (Fig. 2c). Here, quercetin during the interaction forms five hydrogen bonds with G4. Three hydrogen bond formed between hydrogen atom of C-2, C-7 and C-8 of quercitin with G4 (annotated by software as G25 in Fig. 2). Another two hydrogen bonds also formed between hydrogen atom of C-6 and oxygen atom of C-8 of G5 (annotated by software as G35 in Fig. 2). The binding energies calculated for this interaction studies -31.93 kcal/mol (Table 1). Next, we compared another parallel G-quadruplex (3CCO) for its binding mode with quercetin (Fig. 2d). As can be seen in the Figure that quercetin forms six hydrogen bonds. During 3CCO and quercitin interaction, hydrogen bond formed between the hydrogen atom of C2, C6, C8, C13, C14 and oxygen atom of C13 of quercitin bonded with A8 (annotated as A1008 in Fig. 1d), with G9 (annotated as A1009 in Fig. 2d), with G11 (annotated as A1011 in Fig. 2d) and oxygen atom of C13 of quercitin bonded with G5 (annotated as G1005 in Fig. 2d). The binding energies of this interaction calculated as − 29.54 kcal/mol (Table 1). We checked the binding of another parallel structure (3CDM) to find out the binding site of quercitin (Fig. 2e). As can be seen in the figure that 3CDM formed five hydrogen bonds between the hedrogen atom of C-2, C-6, C-7, C-13 and C-14 with G15, G3, G15, G9 and G3. The binding energies of this interaction studies calculated as − 23.17 kcal/mol (Table 1). Further, another parallel G4 (1O0K) was docked with quercetin (Fig. 2f). During the interaction this G4 formed three hydrogen bonds by involving hydrogen atom of C1, C13 and C14 of quercetin with G11 and G10 (annotated as G1011 and G1010 in Fig. 2f) respectively. The binding energies for this interaction is calculated as − 17.11 kcal/mol (Table 1). This data highlights the important points like quercetin binds with parallel telomeric G-quadruplex mainly by groove binding. Secondly, the quercetin recognized and binds with parallel G-quadruplex in a distinct manner. We observed the differences in the binding modes of quercetin with parallel conformation of G-quadruplexes, confirmed by the difference in the binding energies of the various complexes discussed above. We propose that the differences in binding energies is mainly due to the differences in the bond length of the hydrogen bond, shorter the bond length more will be the stability and longer the bond length less will be the stability.

Table 2 Retrieval of G-quadruplex forming sequences of telomeres, cancer proto-oncogenes and RNA G-quadruplex from NCBI data base (https://www.ncbi.nlm.nih.gov/Nucleotide)
Full size table
Fig. 2
figure 2

Lowest—energy-docked conformations representing the interaction profiles of quercetin with Parallel G-quadruplexes. The binding parameters are given in Table 1

Full size image

3.3 Comparative analysis of molecular docking studies of anti-parallel g-quadruplex structures with quercitin

We have selected two telomeric crystal structures which forms antiparallel G-quadruplexes 2KF8 [52] and 1V3N [53] and docked with quercetin. The details of atoms involved in hydrogen bonding between the anti-parallel G-quadruplexes with Quercetin are summarized in Table 2.

Figure 3a shows the binding mode of 2KF8 with quercetin in which hydrogen atom of C-6, C-7, C-8, C-13 and C-14 formed bond with G21, G14, G14, G20 and G20 respectively. The binding energies for this interaction in calculated as − 35.73 kcal/mol (Table 1). Next, we compared another crystal structure 1V3N docked with quercetin (Fig. 3b) in which hydrogen of C-2, C-6, C-7, C-8, C-13 and C-14 formed bond with G5, A6, G5 (annotated as G13 in Fig. 3b), G5 (annotated as G13 in Fig. 3b), G7 and G7 respectively. Here, the binding energies calculated for this interaction studies were − 29.27 kcal/mol (Table 1). This comparative study highlights the distinctive recognition of quercetin due to its bended structure and the glyosidic conformation of the guanine base and the differences in the bond length.

Fig. 3
figure 3

Lowest—energy-docked conformations representing the interaction profiles of quercetin with Anti-Parallel G-quadruplexes. The binding parameters are given in Table 1

Full size image

3.4 Comparative analysis of molecular docking studies of mixed G-quadruplex structures with Quercitin

Next, we analysed the different telomeric crystal structures of mixed G-quadruplex. The details of atoms involved in hydrogen bonding between the mixed G-quadruplexes with Quercetin are summarized in Table 2. First crystal structure selected for mixed telomeric G-quadruplex was 2JSL which was docked with quercetin (Fig. 4a). Here, the hydrogen atom of C2, C7, C13 and C14 bonds with G3, G3, A2 and T1 respectively. The calculated binding energies of this interaction was − 32.68 kcal/mol (Table 1). Further, another mixed G-quadruplex (2HY9) was docked with quercitin (Fig. 4b) to know the details of binding sites and their binding energies. We observed that hydrogen atom of C2, C6 and oxygen atom of C13 forms bond with G11, G10 and A3 respectively. The binding energies calculated for this interaction—30.85 kcal/mol (Table 1). Next, another mixed G-quadruplex (2AQY) was docked with quercetin and it was observed that quercetine binds with groove (Fig. 4c). During binding the hydrogen atom of C-1, C-3, C-7, C-8, C-13 and C-14 of quercetin bonded with G7, T16, G7, G7, A6 and A6 of 2AQY mixed G-quadruplex respectively. The binding energies of this interaction was calculated as − 28.67 kcal/mol (Table 1). Next, we performed the molecular docking with another mixed G-quadruplex (2GKU) with quercetin (Fig. 4d). It was observed that hydrogen atom of C-2, C-6, C-13 and C-14 of quercetin formed hydrogen bonds with T7, T6, G11 and G11 of 2GKU respectively. The binding energies for this interaction was calculated as − 24.79 kcal/mol (Table 1). Further, another mixed G-quadruplex (2JPZ) was docked with quercetin (Fig. 4e). It was found that quercetin binds with grooves of G4 with available binding sites in which hydrogen atom of C-6 and C-8 of quercetin bonded with A3 and G22 respectively. The binding energies calculated as − 23.13 kcal/mol. Next, another G-quadruplex (2JSM) was docked with quercetin (Fig. 4f). It was found that during binding in the major groove of 2JSM, hydrogen atom of C-2, C-7, C-8, C-13, C-14, and oxygen and hydrogen atoms of C-6 of quercetin bound with A2, G3, G3, A2, T9, G9 and G13 respectively. The binding energies of this interaction energies was calculated as − 22.47 kcal/mol (Table 1).

Fig. 4
figure 4

Lowest—energy-docked conformations representing the interaction profiles of quercetin with Mixed G-quadruplexes. The binding parameters are given in Table 1

Full size image

3.5 Comparative analysis of molecular docking studies of cancer proto-oncogenes with quercitin

We also screened the binding potential of quercetin with G-quadruplex structure formed by c-myc proto-oncogene (1XAV) which forms parallel G-quadruplex (Fig. 5a). The details of atoms involved in hydrogen bonding between the G-quadruplexes formed by cancer proto-oncogenes with Quercetin are summarized in Table 2. Our molecular docking results suggested that oxygen and hydrogen atom of C-6, oxygen atom of C-1 and hydrogen atom of C-8 and C-13 formed hydrogen bond with G9, G13, G10, G15 and A22 respectively (Fig. 5a). The binding energies calculated for this interaction energies were − 38.67 kcal/mol (Table 1). Next, we screened c-kit2 which forms parallel G-quadruplex in K+ (2KQH) which binds with quercetin with its grooves in which hydrogen atom of C-2, C-6, C-8, C-13, C-14 and oxygen atom of C-7 with G14, G12, G8, G15, G15 and G7 respectively (Fig. 5b). The binding energies calculated for this interaction energies were − 24.75 kcal/mol (Table 1). Further, we screened the binding potential of quercetin with Human c-myc promoter which formed parallel G-quadruplex in K+ (6AU4) (Fig. 5c). Here, hydrogen atom of C-2, C-6, hydrogen atom of C-14 and oxygen atom of C-1 formed hydrogen bond with A21, A22 and A21 respectively. The binding energies calculated for this interaction energies were − 25.38 kcal/mol (Table 1). Next, we docked another Parallel G-quadruplex formed by VEGF promoter (2M27) with quercetin (Fig. 5d). We observed that quercetin binds with grooves of 2M27 in which hydrogen atom of C-14, C-2, C-7, C-8, and hydrogen and oxygen atom of C-14 with A21, G15, G15, C10, C10 and G9 respectively. The binding energies calculated for this interaction were − 12.97 kcal/mol (Table 1). These results are consistent with previously published reports of quercetin binding with c-myc promoter which stated that quercetin stacks at a 5′ and 3′-G-tetrads of c-myc promoter G-quadruplex, stabilized the structure by π-π interactions [54, 55].

Fig. 5
figure 5

Lowest—energy-docked conformations representing the interaction profiles of quercetin with G-quadruplexes formed by cancer-proto-oncogenes. The binding parameters are given in Table 1

Full size image

3.6 Comparative analysis of molecular docking studies of RNA G-quadruplex with quercitin

To confirm whether quercetin may bind to RNA G-quadruplex or not, we selected two RNA G-quadruplex forming motifs one is 1P79 and another one is RGG-motif of FMRP (PDB I.D. 5DE5) docked with quercetin to know the binding sites and binding energies. The details of atoms involved in hydrogen bonding between the RNA G-quadruplexes with Quercetin are summarized in Table 2. Figure 6a showed the molecular docking results of 1P79 with quercetin in which hydrogen atom of C-6, C-7, C-8 and oxygen atom of C-7 bonded with G4, G5, U6 and U6 respectively. The binding energies calculated for this interaction were − 14.8 kcal/mol (Table 1). Next, we docked another RNA quadruplex (5DE5) with quercetin (Fig. 6b). Binding results suggested that hydrogen atom of C-2, C-6, C-8, C-13, C-14 and oxygen atom of C-8 forms hydrogen bond with G4, G31, G4, C5, C5, and C5. The binding energies calculated for this interaction were − 14.6 kcal/mol (Table 1). The obtained results shows that the quercetin drug interacted with RNA G-quadruplex but interaction energies are less and close to similar energy values for both the RNA G-quadruplexes.

Fig. 6
figure 6

Lowest—energy-docked conformations representing the interaction profiles of quercetin with G-quadruplexes formed by RNA G-quadruplex. The binding parameters are given in Table 1

Full size image

3.7 Structural studies on the binding of quercetin with Human Telomeric and Bcl-2 G-quadruplex structures with and without quercetin in the presence of 100 mM K+ under molecular crowding conditions

CD spectroscopy was employed to investigate the changes on the conformation of 23-mer IND DNA sequence 5′-d(TAGGGTTAGGGTTAGGGTTAGGG)-3′ upon quercetin binding. The structure of each DNA strand is in 30 mM sodium cacodylate buffer pH (7.0), 100 mM K+ and 40 wt% PEG 200 in presence and absence of different concentrations of quercetin (Fig. 7a). CD spectrum of IND is characterized by a positive peak at 292 nm and negative peak at 240 nm typically observed for an antiparallel G-quadruplex in the presence of K+ [56]. Next, IND was titrated with an increasing concentration 0.2 μM, 0.4 μM, 0.8 μM, 2 μM, 4 μM and 8 μM of quercetine. We observed a slight increment of CD intensity at 292 nm and small shoulder around 260 nm and the negative peak at 240 nm shifted towards 236 nm upon the titration of quercetine. However, these changes are very small but the overall CD spectra showed that quercetine binds with IND G-quadruplex. In contrast, HTPu in the presence of K+ (Fig. 7b), exhibits a strong positive peak around 290 nm with a minor shoulder around 260 nm, 219 nm and a small negative peak at 234 nm, indicating a formation mixed G-quadruplex, consistent with the previously published report [56]. On titrating HTPu with quercetin, we observed slight increment of CD signal at 290 nm and. shoulder at 260 nm became sharper with an increase in intensity at 219 nm. These changes indicate that the binding of quercetin inducing structural change. Next, we recorded the CD spectra of Bcl2 in similar solution conditions and titrated with an increasing concentration of quercetin. We observed prominent negative peak at 298 nm, 219 nm and negative peak at 233 nm indicating a formation of antiparallel G-quadruplex [50, 57] (Fig. 7c). After the addition of quercetin, we observed a decrease in CD intensity at 298 nm and shoulder towards 260 nm started emerging. There was also increase in CD intensity at 219 nm while negative peak in control without quercetin at 233 nm shifted towards 237 nm. We proposed that the gradual decrease in CD intensity on increasing the quercetin concentration at 298 nm and development of the emerging shoulder at 260 nm shows the structural transition from antiparallel to parallel G-quadruplex, although this may require extended studies with more increasing concentration of quercetin to reach on any conclusion. These observed changes are due to the quercetin binding may be due to the two way binding of the quercetin with G-quadruplex structure, one is hydrogen bonding involving the hydrogen bonding sites of quercetin with G-bases and another is due to the intercalation of the aromatic ring within G-quartet core, consistent with the previously published reports [59].

Fig. 7
figure 7

CD spectra of normalized molar ellipticity of 4 μM IND (a), 4 μM HTPu (b) and 4 μM Bcl-2 (c) in buffer containing KCl (100 mM), 0.5 mM EDTA without any additive (black line) and titrated with an increase in concentration of quercetine. Control DNA of each sequence along with highest concentration of quercetin used in this study is shown in inset separately

Full size image

3.8 Thermodynamic analysis of the Human Telomeric and Bcl2 G-quadruplex with and without Quercetin

Next, we explored the thermal stability of the DNA structures with and without quercetin. Figure 8a shows normalized UV melting profile of 4µM IND in the buffer containing 100 mM KCl and 40 wt% PEG 200 in the absence and presence of quercetin. The ratios of DNA:quercetin were (1:0, 1:0.1, 1:0.2, 1:1, 1;2 and 1:3) respectively (Fig. 3). The melting temperature (Tm) was evaluated by a curve fitting procedure as described previously [48, 49]. The Tm of the IND G4 was increased from 67.3 °C, 69.1 °C, 70.3.0 °C, 70.3 °C, 70.3, 71.5 and 75.6 °C respectively. The melting curves with a single transition and overall 8.3 °C difference in the Tm values in different DNA:quercetin ratios. These results are consistent with the previously published reports on the stabilization of telomeric G-quadruplex on the quercetin binding [58, 59]. Next, we recorded the data with HTPu G4 in the buffer containing 100 mM KCl and 40 wt% PEG 200 in the absence and presence of quercetin with similar ratios. Tm of the HTPu G4 was varied from (68.5 °C, 69.0 °C, 69.0 °C, 69.0 °C, 69.0 and 69.5 °C) respectively (Fig. 8b). These results indicated that overall change in 1 °C these G4s possess similar thermal stability upon quercetin binding. On the other hand, the Tm value of the Bcl-2 G4 was increased from 65.9 °C, 66.5 °C, 67.5 °C, 68.0 °C, 68.0 °C and to 68.5 °C in similar solution conditions respectively (Fig. 8c). Therefore, the overall stabilization in Tm was of 3 °C upon quercetin binding, although further systematic studies are required. These results are consistent with our CD results which state the binding of antiparallel G4 with quercetin and previously published reports [60,61,62,63,64].

Fig. 8
figure 8

UV melting curves of 4 μM IND (a), 4 μM HTPu (b) and 4 μM Bcl-2 (c) in buffer containing KCl (100 mM), 0.5 mM EDTA without any additive (green line) and titrated with an increase in concentration of quercetine. in buffer containing KCl (100 mM), 0.5 mM EDTA, without any additive (light green line); with 0.2 μM quercetin (black), 0.4 μM quercetin (red), 0.8 μM quercetin (pink), 2 μM quercetin (green), 4 μM quercetin (majenta), 8 μM quercetin (blue) respectively. Experiment has been repeated in triplicates to reproduce the data

Full size image

To assess the origin of the observed stabilities of IND, HTPu and Bcl-2 G4 upon the complex formation with quercetin, the thermodynamic parameters of their formations, such as the enthalpy change (ΔH°), the entropy change (ΔS°), and the free energy change at 25 °C (ΔG°25) of the IND, HTPu and Bcl-2 G4 G4 formation were estimated in the presence and absence of (1:0, 1:0.1, 1:0.2, 1:1, 1;2 and 1:3) (summarized in Table 3). The interaction of IND-quercetin complex on increasing the quercetin concentration in a buffer containing 100 mM KCl and 30 mM sodium cacodylate buffer (pH 7.0), ΔH° decreased -37.17 kcal/mol, -47.50 kcal/mol, -48.15 kcal/mol, − 50.75 kcal/mol, -54.88 kcal/mol, -57.04 kcal/mol, TΔS° decreased from -12.76 kcal/mol, -16.18 kcal/mol, -16.40 kcal/mol, -17.20 kcal/mol, -18.69, -19.39 kcal/mol. The free energy (ΔG°) at 298 K follows the same order. ΔG°25 decreased − 0.27 kcal/mol, − 0.28 kcal/mol, − 0.30 kcal/mol, − 0.35 kcal/mol, − 0.38 kcal/mol, − 0.51 kcal/mol (Table 3). The interaction of HTPu-quercetin complex on increasing the quercetin concentration in a buffer containing 100 mM KCl and 30 mM sodium cacodylate buffer (pH 7.0), ΔH° decreased − 48.81 kcal/mol, − 49.84 kcal/mol, − 51.50 kcal/mol, − 56.52 kcal/mol, − 56.67 kcal/mol, − 73.19 kcal/mol, TΔS° decreased from − 16.67 kcal/mol, − 17.00 kcal/mol, − 17.58 kcal/mol, − 19.34 kcal/mol, − 19.39, − 25.00 kcal/mol. The free energy (ΔG°) at 298 K follows the same order. ΔG°25 decreased − 0.28 kcal/mol, − 0.26 kcal/mol, − 0.26 kcal/mol, − 0.33 kcal/mol, − 0.36 kcal/mol, − 0.41 kcal/mol respectively (Table 3). However, the interaction of Bcl2− quercetin complex on increasing the quercetin concentration in a buffer containing 100 mM KCl and 30 mM sodium cacodylate buffer (pH 7.0). ΔH° decreased − 29.17 kcal/mol, − 31.31 kcal/mol, − 33.40 kcal/mol, − 38.33 kcal/mol, − 45.21 kcal/mol, − 45.88 kcal/mol, − 51.99 kcal/mol, TΔS° decreased from − 10.17 kcal/mol, − 10.64 kcal/mol, − 11.40 kcal/mol, − 13.11 kcal/mol, − 15.72 kcal/mol, − 15.73 kcal/mol and − 17.83 kcal/mol (Table 3). Therefore, These thermodynamic parameters suggested that the stabilization of the G-quadruplex by the binding of quercetine is promoted by a favourable an enthalpic contribution exceeding an unfavorable entropy change. Accordingly, specific intermolecular hydrogen bonding between the quercetin and G4, as well as the stacking interactions of aromatic rings may contribute this enthalpic stabilization of G4. These enthalpic stabilization effects on G4 derived from specific interactions have been reported for small molecular ligands effects G4s [50, 56,57,58,59,60,61,62,63,64].

Table 3 Thermodynamic parameters for IND interaction with quercetin in the presence of 100 mM K+ and 40 wt% PEG 200
Full size table

4 Conclusion

In this study, we have examined the binding sites of flavonoid (quercetin) with polymorphic telomeric G-quadruplex DNA structures (parallel, antiparallel and mixed) and calculated their binding energies. We also examined the binding potential of quercetin with cancer proto-oncogenes and with RNA G-quadruplexes. The binding energies of quercetin calculated for each structure separately for each G-quadruplex structure gives an idea that Quercetin may be used as a lead molecule to target polymorphic telomeric G-quadruplex as well as cancer proto-oncogenes and hence a suitable natural drug molecule for anti-cancer therapeutics. Our study highlighted the structural aspects of the binding of Quercetin with different polymorphic biological targets of G-quadruplexes and we believe that this is the first report for the identification of selection of suitable nucleic acid target for quercetin binding.