Introduction

In recent years, G-quadruplexes, which consist of four-stranded nucleic acids and are described as presenting a high-order secondary DNA structure, have drawn attention regarding their potential use in anticancer therapies1,2. These guanine-rich sequences exist ubiquitously in significant regions of the eukaryotic genome, such as in the promoter regions of several oncogenes and the telomeres at the ends of the chromosomes3,4,5,6. It has been reported that telomerase is active and up-regulated in approximately 85% of tumor cells, which would lead to telomere elongation and contribute to cancer cell immortalization7,8. Thus, the telomeric G-quadruplex has been considered to be a potentially effective antitumor target1,2. The formation of a G-quadruplex would result in the inhibition of telomerase activity and would thereby terminate telomere maintenance9. Researches aimed at the stabilization of the G-quadruplex structure of certain sequences and efficiently inhibition of telomerase activity represent a rising field of research in anticancer drug design and development.

G-quadruplexes consist of stacked G-tetrads, which are formed from four guanine bases of G-rich sequences connected by Hoogsteen hydrogen bonds (Fig. 1a). Monovalent cations, such as potassium and sodium ions, can stabilize G-quadruplex structures, presumably via electrostatic interactions with the guanine carbonyl moieties. Previous studies addressing various G-quadruplex conformations have been reported and crystallographic and NMR data have revealed that G-quadruplexes can be classified according to different strand orientations as showing a parallel structure10 (Fig. 1b), mixed-type “(3 + 1) hybrid” structure11 (Fig. 1c) or basket-type “(2 + 2) hybrid” structure12 (Fig. 1d). These structural data provide important information for the development of G-quadruplex binders.

Figure 1
figure 1

Schematic representation of a G-tetrad and different conformations of G-quadruplexes; ORTEP views of dptmp and 1.

(a) a G-tetrad consists of four guanines, (b) the wild-type 22-mer crystal with K+ (parallel structure, PDB ID: 1KF1), (c) the modified 20-mer in K+ solution (mixed-type “(3 + 1) hybrid” structure, PDB ID: 2KZD), (d) the wild-type 22-mer in Na+ solution (basket-type “(2 + 2) hybrid” structure, PDB ID: 143D).

Full size image

In past decades, many small molecules that interact with G-quadruplexes have been reported13,14. Some synthetic simple organic molecules15,16,17,18,19,20, typically metal complexes21,22,23,24,25,26,27,28,29,30, have been well characterized as effective G-quadruplex binders. Platinum(II) complexes, as successful inorganic anticancer drugs, have long been considered important due to their interaction with nucleic acids. Recently, planar platinum(II) complexes24,25,26,27,28,29,30 have drawn increasing interests related to targeting telomeric G-quadruplex DNA. Some mononuclear platinum(II) complexes, including platinum(II)-phenanthroline24,25, platinum(II)-terpyridine26 and others, with positively charged side arms, which interact readily with the negatively charged phosphate backbones, grooves and loops of the quadruplexes, have been reported. Additionally, a chain-like bidentate dinuclear platinum(II) complex30 has been reported as a c-myc and hTel quadruplex binder. A kind of self-assembled tetranuclear platinum(II) complexes27,28,29 with an aromatic plane was studied, which turned out to be good G-quadruplex binders, displaying evident telomerase inhibition. These studies have suggested that a large electron-deficient π-aromatic surface, positively charged area and positively charged substituents represent characteristics of a good G-quadruplex binder.

We have previously reported a series of mononuclear and tetranuclear Pt(II) complexes as G-quadruplex binders28,29,31,32. To further investigate the structure-activity relationship of various Pt(II) complexes, herein we synthesized a new bridged ligand by quaternizing a branch of a trigeminal star-like molecule, 4-[4,6-di(4-pyridyl)-1,3,5-(2-triazinyl)]-1-methylpyridine-1-ium iodide (dptmp·I) (CCDC 894005) and its derivative V-shaped dinuclear platinum(II) complexes, {[Pt(dien)]2(dptmp)}(PF6)5 (1) (CCDC 894006) and {[Pt(dpa)]2(dptmp)}(PF6)5 (2). Besides, the structures of dptmp and 1 have been successfully characterized by X-ray crystallography (shown in Fig. 1 and Supplementary Table S1).

Results

Synthesis of dptmp and Pt(II) Complexes {[Pt(Lx)]2(dptmp)}(PF6)5

The compound dptmp·I was synthesized through the reaction of 2,4,6-tri(4-pyridyl)-1,3,5-triazine (tpt) with MeI. We tested several solvents (such as dichloromethane, trichloromethane, tetrachloromethane and N,N-dimethyl formamide) but could only obtain the pure product from trichloromethane. This was because dptmp·I precipitated from trichloromethane and could not undergo further substitution reactions. After heating the reaction mixture under reflux in trichloromethane for 48 h, it was filtered, the solvent was removed under reduced pressure and the residue was washed with trichloromethane. The isolated dptmp·I was characterized by1H and 13C NMR spectroscopy, mass spectrometry and elemental analyses (see Experimental Details). Subsequently, dptmp·PF6 was obtained through the reaction of dptmp·I with ammonium hexafluorophosphate.

The traditional method for synthesizing [Pt(dien)Cl]Cl·HCl involves the reaction of K2PtCl4 with diethylenetriamine (dien) dissolved in water (pH 3.0) by heating under reflux for 72 h followed by concentration until the yellow product precipitates out33. We employed a similar method, described by Annibale et al.34, using [Pt(COD)Cl2]35 as an intermediate, which was found to be more convenient and resulted in a very high yield of 96.7%.

The complex [Pt(Lx)Cl]Cl was first treated with AgNO3 to eliminate the two chlorides, then filtered to remove AgCl and transferred to a stoppered flask, followed by the addition of dptmp·PF6 and heated at 90°C for 72 h under N2. Finally, excess ammonium hexafluorophosphate was added to the reaction mixture and the obtained precipitate was filtered out. The two dinuclear Pt(II) complexes were fully characterized via 1H, 13C and 195Pt NMR spectroscopy and elemental analyses. Additionally, dptmp·I and 1 were structurally characterized by X-ray crystallography (Fig. 1).

Selectively stabilize hTel G-quadruplex DNA structure

Fluorescence resonance energy transfer (FRET) studies were conducted to investigate the binding abilities of the three compounds to human telomeric, promoter G-quadruplex DNA sequences (hTel, c-myc and bcl2) and a duplex DNA sequence. Reliable FRET melting curves (Fig. 2) and stabilization temperature (ΔTm) values (Table 1) were obtained36. As shown in Table 1, it was clear that the three compounds exhibited ΔTm values >17°C with the human telomeric (hTel) G-quadruplex; in contrast, they presented ΔTm values <6°C with c-myc, ΔTm values <3°C with bcl2 and ΔTm values <1.5°C with duplex DNA. At the same time, the two dinuclear Pt(II) complexes displayed high ΔTm values with the hTel G-quadruplex of 35.4°C and 30.5°C at 0.5 μM, which were higher than that obtained for the quaternary ammonium ligand-dptmp (17°C at 0.5 μM).

Table 1 Stabilization temperatures, ΔTm (°C) for hTel, c-myc, bcl2 and duplex DNA stabilized by the three compounds determined from FRETa
Full size table
Figure 2
figure 2

Upper: FRET-melting curvesa for experiments carried out with hTel, c-myc, bcl2 and duplex DNA separately with the three compounds (dptmp, 1 and 2). Lower: dose-dependent inhibition of PCR amplification by the different concentrations of the compounds (dptmp, 1 and 2) in the PCR-stop assaysb.

a All experiments were conducted at a DNA concentration of 400 nM and with a 0.5 μM concentration of the compounds, in Tris-HCl buffer (10 mM, pH 7.4) containing 60 mM potassium cacodylate buffer (DNA sequence: black; red: dptmp; blue: 1, green: 2). b The amplified PCR products were then analyzed on 15% non-denaturing polyacrylamide gels (100 V, 30 min) in 1× TBE buffer, followed by silver staining.

Full size image

To quantitatively investigate the binding constants between compounds and DNA samples (the hTel G-quadruplex and duplex DNA sequences), surface plasmon resonance (SPR) experiments were conducted (Supplementary Fig. S8 and Table 2), which serve as a powerful technique for monitoring molecular reactions in real time37,38,39,40,41,42,43. Figure S8 shows the SPR sensorgrams generated for the three compounds binding to the immobilized hTel G-quadruplex and duplex DNA at different concentrations. Based on the obtained data, shown in Table 2, the two Pt(II) complexes displayed higher binding constants with the hTel sequence compared to that with dptmp (dptmp, 4.00 × 105; 1, 1.37 × 106 and 2, 1.40 × 106). Regarding the binding selectivities between the hTel quadruplex and duplex sequences, 1 showed a higher selectivity of 7.1 fold (5,170 vs. 731 nM), compared to 2 at 4.7 fold (3,390 vs. 712 nM), while dptmp showed the lowest selectivity of 1.2 fold (2,890 vs. 2,500 nM). We also found that the kinetic binding affinities (KA) of the three compounds with the hTel G-quadruplex were superior to those with the duplex DNA sequence.

Table 2 Kinetic parameters determined with SPR spectroscopyabc
Full size table

Stabilizing hTel G-quadruplex structures

To further evaluate whether the compounds effectively bound to the tested oligomer HTG21 and stabilized the G-quadruplex structure, polymerase chain reaction (PCR) stop assays were performed39,40,41. In these assays, in the presence of the compounds, the template sequence can form a G-quadruplex structure and PCR products of dsDNA will be undetectable.

Figure 2 indicates that the inhibitory effects of complexes 1 and 2 were enhanced distinctly as the concentrations of both complexes increased from 1.0 to 9.0 μM, with no PCR products being detected at 5.0 or 7.0 μM, respectively. However, dptmp showed a weaker ability to inhibit the appearance of the PCR products and some PCR products were detected at 180 μM. The results indicated that the two complexes exhibited higher binding affinities for the G-quadruplex and were effective G-quadruplex binders. Moreover, to avoid any possible interaction of rTaq polymerase with the compounds and subsequent inhibition of the PCR process, we conducted parallel experiments using the mutated oligomer HIG21 mu instead of HIG21 under the same experimental conditions. The HIG21 mu sequence was unable to form a G-quadruplex. The obtained results suggested that the compounds could not significantly affect the activity of rTaq polymerase at comparable concentrations (Supplementary Fig. S10).

Induction of the formation of mixed-type parallel/antiparallel hTel G-quadruplexes and stabilization of this structure

Circular dichroism (CD) spectroscopy was performed to roughly characterize the structural conversion of various conformations of human telomeric G-quadruplexes, which consist of intra- and inter-molecular in parallel, antiparallel-stranded and mixed arrangements, depending on the strand orientations37,44,45. It has been reported that the CD spectra of the human telomeric DNA sequence (AG22) may consist of a mixed-type parallel/antiparallel G-quadruplex conformation in the presence of potassium cations and usually show a strong positive peak around 290 nm, a small positive peak at 265 nm and a small negative peak at 240 nm46,47,48. On the other hand, in the presence of sodium cations, the CD spectra consist of a positive band near 295 nm and a negative band at 265 nm, which may be characteristics of a typical antiparallel G-quadruplex structure12,49. In the absence of metal ions, the CD spectra of the AG22 sequence indicate the coexistence of single strand, parallel and antiparallel G-quadruplexes31,37,38,44,45,46,47.

In the current studies, the AG22 sequence was used to study the structural induction and structural transition of G-quadruplexes by CD titration assays in the presence of the three compounds. As reported previously, the AG22 sequence is considered to show a summation spectrum involving multiple G-quadruplexes (mixed-type/basket-type) in K+ solution from NMR data obtained for AG2251. The sequences Tel26 and wild-type Tel26 which both contain the AG22 four-G-tract human telomeric core sequence and have been well-defined based on NMR and CD data obtained with mixed-type G-quadruplexes, exhibit a distinct CD spectrum including a strong positive peak around 295 nm, with a shoulder peak around 268 nm and a smaller negative peak at 240 nm50,51,52. When different concentrations of the compounds were added to the K+ solution of AG22, very similar CD spectra (Fig. 3) were generated to the Tel26 and wild-type Tel26 which had been confirmed the mixed-type G-quadruplex conformation, demonstrating that the compounds may induce the formation of mixed-type G-quadruplex. In the absence of any metal cations, the CD spectra consist of a positive band at 257 nm, a small positive band at 295 nm and a small negative peak near 240 nm, as shown in Figure 3. When the concentrations of the compounds were increased, there was an obvious decrease observed at 257 nm, accompanied by a sharp increase around 295 nm. These data suggest that all of the compounds, especially the two Pt(II) complexes, probably induced the conformation of a mixed-type parallel/antiparallel G-quadruplex structure and stabilized this structure38,53. These CD phenomena were consistent with the findings of Zhou et al.54. Similar conclusions were drawn based on a more identifiable conformation transition detected using 100 mM Na+ and parallel experiments also roughly demonstrated the selectivity of the compounds towards the mixed-type stranded topology.

Figure 3
figure 3

CD titration spectra of the human telomeric G-quadruplexes (3.0 μM) induced by the three compounds (from left to right: dptmp, 1 and 2) in 10 mM Tris-HCl, pH 7.4, at room temperature (r = CML/CDNA).

a: In the absence of metal ions; b: In 100 mM NaCl; c: In 100 mM KCl.

Full size image

Inhibition of telomerase activity

To obtain further information about whether the compounds could inhibit telomerase activity and affect telomere elongation by telomerase, we performed the telomeric repeat amplification protocol (TRAP) assays, which can qualitatively and quantitatively evaluate the inhibition of telomerase9,42,55,56,57. In these assays, the elongated telomeres were further amplified by PCR cycling and could be visualized with dye staining. From the results presented in Figure 4, it is clear that the three compounds inhibited telomere elongation in a concentration-dependent manner, while the amplification of the internal telomerase assay standard (ITAS) control was not obviously affected. Thus, it was the telomerase activity rather than the PCR procedure itself that was inhibited by the three compounds. The inhibition of telomerase by the two complexes was found to be due to their function as strong G-quadruplex binders, showing IC50 values of 0.113 ± 0.019 and 3.69 ± 0.53 μM, respectively. Thus, complexes 1 and 2 exhibited stronger telomerase inhibition than dptmp (IC50 = 12.3 ± 1.2 μM).

Figure 4
figure 4

TRAP assaya results for the three compounds (dptmp, 1 and 2), showing telomeric ladders produced by PCR amplification of the oligonucleotides generated by the action of telomerase on a TS primer. The lower band is an internal control primer (ITAS).

a Each single assay was dependent on 1.0 μL of telomerase extract (200 cells), prepared from HeLa cells with NP-40 lysis buffer. Negative controls 1 and 2, which accompanied every assay, involved either incubating approximately 1.0 μL of cell lysate at 85°C for 10 min before primer extension or incubating approximately 1.0 μL of NP-40 lysis buffer, respectively. The control (0 nM) assay was performed without adding any compound.

Full size image

Discussion

Guanine-rich sequences, which can form high-order G-quadruplex secondary structures, have been reported to be widely distributed in the human genome. Because of the potentially significant role of G-quadruplexes in the development of anticancer drugs, further research has been conducted on these compounds, for example, addressing G-quadruplex binders and telomerase inhibitors. In our studies, the interactions between the three compounds and hTel G-quadruplex sequences were investigated using FRET, SPR, CD and PCR stop assays. From the obtained FRET data, we conclude that the ΔTm values of the two V-shaped dinuclear Pt(II) complexes were much higher than those previously reported for a modified phenanthroline-platinum(II) complex (20°C at 1.0 μM, FRET)24 and [Pt(dppz-COOH)(N-C)]CF3SO3 (14°C at 20 μM, UV melting study)58 and similar to those of tetranuclear platinum(II) complexes (34.5°C at 0.75 μM, 33.5°C and 32.1°C at 0.5 μM, FRET)27,28. Thus, the three compounds exhibited specific recognition abilities for different sequences of hTel G-quadruplex DNA. Furthermore, the obtained extremely low ΔTm values demonstrate their poor stabilities with promoter sequences (c-myc and bcl2) and duplex DNA. The results obtained from SPR studies demonstrate the selectivity differences of the three compounds, with dptmp displaying little impact on G-qudruplex interaction, while 2 had a moderate selectivity and 1 had the highest selectivity. Furthermore, both Pt(II) complexes show preferential quadruplex/duplex DNA selectivity compared to corresponding dptmp, highlighting the importance of the platinum(II) moiety in defining their binding ability. This might be attributed to the platinum(II) moiety that can interact with both the grooves and loops of the hTel quadruplex and their negatively charged phosphate backbones.

The data obtained from PCR stop assays were consistent with previous results, demonstrating that both Pt(II) complexes 1 and 2, especially 1, exhibited an excellent stability towards hTel G-quadruplex, binding more effectively than the corresponding ligand-dptmp. We were able to obtain the exact conformation of the mixed-type stranded topology induced by the three compounds from the CD assays. Another result that should be paid attention to is that the two Pt(II) complexes were considered to show reasonable inhibition of telomerase activity based on the TRAP studies.

In conclusion, we have prepared a water-soluble, quaternized trigeminal ligand and its two derivative Pt(II) complexes. X-ray diffraction analysis showed that the dinuclear Pt(II) complex exhibited a V-shaped structure. Further biochemical assays confirmed that the ligand and the two complexes, especially complex 1, coordinated by straight-chain polyamine, can selectively stabilize the human telomeric G-quadruplex, likely induce the formation of mixed-type parallel/antiparallel G-quadruplex structure and significantly inhibit the activity of telomerase, probably due to interactions with the sugar-phosphate backbone of the G-quadruplex. Thus, the Pt(II) complexes exhibit a potential promise for antitumor therapies and need to be further investigated to clarify their specific biochemical details.

Methods

Materials

All chemicals and solvents were obtained from commercial sources and used without further purification.

Synthesis of 4-[4,6-di(4-pyridyl)-1,3,5-(2-triazinyl)]-1-methylpyridine-1-ium iodide C19H15N6·I (dptmp·I): A mixture of tpt (0.62 g, 2.00 mmol) and MeI (0.85 g, 6.00 mmol) in trichloromethane (240 mL) was stirred at 61°C for 48 h in darkness. The solution gradually became red and a red precipitate formed. The resulting slurry was collected by filtration and the residue was washed with trichloromethane (2 × 10 mL) to remove tpt. The obtained product was dried under vacuum to give a red powder (yield: 0.79 g, 86%). Yellow crystals were grown from an aqueous solution standing in air at room temperature and characterized by X-ray structure analysis. 1H NMR (Supplementary Fig. S2) (300 MHz, D2O): δ 8.92 (d, J = 6.3 Hz, 2H), 8.72 (d, J = 6.0 Hz, 2H), 8.40 (d, J = 4.8 Hz, 4H), 8.00 (d, J = 4.8 Hz, 4H), 4.538 (s, 3H); 13C NMR (Supplementary Fig. S3) (75 MHz, (CD3)2SO): δ 171.8, 168.7, 151.5, 149.7, 147.6, 142.5, 126.9, 123.1, 49.0. ESI-MS (Supplementary Fig. S1) (H2O) anal. calcd. for C19H15N6, [L-I−]+ (m/z) 327.4, found 327.36. Elemental analysis (anal. calcd., found for C19H15N6I·1.5H2O): C (47.41, 47.72), H (3.77, 3.97), N (17.46, 17.44).

Synthesis of 4-[4,6-di(4-pyridyl)-1,3,5-(2-triazinyl)]-1-methylpyridine-1-ium hexafluorophosphate C19H15N6·PF6 (dptmp·PF6): The red powder of dptmp·I (0.16 g, 0.35 mmol) was dissolved in a sufficient amount of water and ammonium hexafluorophosphate (0.50 g, 3.06 mmol) was then added to the solution and a pale solid precipitated immediately. The product was filtered, washed from the filter twice with 2 × 10 mL ethanol and then dried under vacuum (0.17 g, quantitative yield). Elemental analysis (anal. calcd., found for C19H15N6PF6): C (48.31, 48.23), H (3.20, 3.45), N (17.79, 17.76).

Synthesis of {[Pt(dien)]2(dptmp)}(PF6)5 (1): [Pt(dien)Cl]Cl·HCl was synthesized through a method similar to that described by Giuliano Annibale34, using [Pt(COD)Cl2]35 as an intermediate. [Pt(COD)Cl2] (0.25 g, 0.67 mmol) in methyl alcohol (about 50 mL), diethylenetriamine (68.9 mg, 0.67 mmol) was added with stirring and the mixture was heated at 40–50°C. After 15 min, the solid dissolved and a clear, pale yellow solution was obtained, which was cooled to room temperature and then filtered to remove any unreacted COD complex. Water was then removed under reduced pressure, leaving a pale yellow solid, which was collected and washed thoroughly with diethyl ether and then air dried (yield: 0.24 g, 96.7%). [Pt(dien)Cl]Cl·HCl (87 mg, 0.22 mmol) and AgNO3 (0.11 g, 0.65 mmol) in water (9 mL) were stirred in a stoppered flask in darkness for 48 h at 45°C and then filtered to remove AgCl. The clear filtrate was subsequently transferred to another stoppered flask, to which dptmp·PF6 (47 mg, 0.10 mmol) was added, followed by heating at 90°C for 72 h under N2. The whole experiment was conducted in darkness. The resulting solution was collected by filtration, after which excess ammonium hexafluorophosphate was added to the filtrate and a white product precipitated immediately. Finally, the resulting slurry was filtered and the product washed twice from the filter with 2 × 10 mL of ethanol, then dried under vacuum (0.13 g, 72%). Pale crystals were grown from an acetonitrile-water (1:5) mixture standing in air at room temperature and characterized by X-ray structure analysis (Fig. 1 and Table S1, S2). 1H NMR (Supplementary Fig. S4) (300 MHz, (CD3)2SO): δ 9.39 (d, J = 6.0 Hz, 2H), 9.30 (d, J = 6.0 Hz, 2H), 9.09 (d, J = 6.0 Hz, 4H), 8.98 (d, J = 6.0 Hz, 4H), 7.45 (broad, 2H), 6.16 (d, J = 6.0 Hz, 4H), 6.03 (d, J = 6.0 Hz, 4H), 4.52 (s, 3H), 3.17–2.77 (m, 16H); 13C NMR (Supplementary Fig. S5) (100 MHz, (CD3)2SO): δ 170.2, 168.9, 154.2, 149.1, 147.1, 144.8, 126.9, 126.0, 54.6, 50.3, 48.9; 195Pt NMR (1H 400 MHz, (CD3)2SO): −1,196.23 ppm and K2PtCl4 was used as an internal reference (δ = 0). Elemental analysis (anal. calcd., found for {[Pt(dien)]2(dptmp)}(PF6)5): C (19.67, 19.43), H (2.51, 2.76), N (10.19, 10.11).

{[Pt(dpa)]2(dptmp)}(PF6)5 (2) was synthesized with a high yield of 95.3% as above synthesis method of 1, except that the first procedure of treating with AgNO3 should be 60°C for 40 h. The complex of [Pt(dpa)Cl]Cl.2H2O was synthesized according to a procedure described in the literature59. 1H NMR (Supplementary Fig. S6) (300 MHz, (CD3)2SO): δ 9.47–8.75 (m, 12H), 8.29 (t, J = 6.0 Hz, 4H), 7.89 (d, J = 6.0 Hz, 4H), 7.73 (t, J = 6.0 Hz, 4H), 7.49 (t, J = 6.0 Hz, 4H), 5.01–4.52 (m, 11H); 13C NMR (Supplementary Fig. S7) (100 MHz, (CD3)2SO): δ 172.1, 170.5, 169.9, 169.2, 168.9, 167.0, 154.6, 151.7, 149.1, 147.6, 145.6, 142.3, 127.6, 126.8, 126.1, 123.9, 122.9, 59.4, 48.9. 195Pt NMR (1H 400 MHz, (CD3)2SO): −770.2 ppm and K2PtCl4 was used as an internal reference (δ = 0). Elemental analysis (anal. calcd., found for {[Pt(dpa)]2(dptmp)}(PF6)5·CH3CN): C (28.72, 28.85), H (2.36, 2.52), N (9.68, 9.46).

X-ray diffraction measurements

The intensity data were recorded using an Agilent Xcalubur Nova CCD diffractometer with graphite monochromated Cu Kα radiation (1, λ = 1.54178 Å) at 293(2) K, a Rigaku R-AXIS SPIDE IP system with Mo Kα radiation (dptmp, λ = 0.71073 Å) at 298(2) K and a Brucker Smart 1000 CCD diffractometer with Mo Kα radiation ([Pt(dpa)NO3]PF6, λ = 0.71073 Å) at 173(2) K. The structures of dptmp, 1 and [Pt(dpa)NO3]PF6 (CCDC 904100) were solved via direct methods and refined using a full-matrix least-squares technique with SHELXTL program package. All non-hydrogen atoms were refined using anisotropic displacement parameters. All hydrogen atoms on organic ligands were generated in riding mode. The obtained crystallographic data and the details of the refinement of dptmp, 1 and [Pt(dpa)NO3]PF6 are listed in Supplementary Table S1 and the bond lengths and bond angles are given in Supplementary Tables S2–S4.

Fluorescence resonance energy transfer (FRET) studies

The employed FRET probes included a fluorescently labeled hTel oligonucleotide (5′-FAM-[GGGTTAGGGTTAGGGTTAGGG]-TAMRA-3′, mimicking the human telomeric repeat, FAM: 6-carboxyfluorescein, TAMRA: 6-carboxytetramethylrhodamine, Sangon), three promoter sequences (c-myc: 5′-FAM-[TG4AGGGTGGGGAGGGTGGGGAAGG]-TAMRA-3′; bcl2: 5′-FAM-[AGGGGCGGGCGCGGGAGGAAGGGGGCGGGAGCGGGGCTG]-TAMRA-3′, Sangon) and a duplex DNA sequence (5′-FAM-[TATAGCTATA-HEG-TATAGCTATA]-TAMRA-3′, HEG linker: [(-CH2-CH2-O-)6], Sangon). The experiment procedure was as previously reported method29. Final analysis of the data was carried out using Origin 8.0 (OriginLab Corp.).

Surface plasmon resonance (SPR) studies

SPR measurements were conducted in a ProteOn XPR36 Protein Interaction Array system (Bio-Rad Laboratories, Hercules, CA, USA) using a Neutravidin-coated GLM sensor chip. Biosensor experiments28,29 were conducted in filtered and degassed running buffer (Tris-HCl 50 mM pH 7.4, 300 mM KCl, 0.005% Tween-20) at 25°C. The biotinylated oligonucleotides used in these assays were Biotin-hTel (Quadruplex: 5′-biotin-[AGGGTTAGGGTTAGGGTTAGGG]-3′) and Biotin-duplex (5′-biotin-[CGAATTCGTCTCCGAATTCG]-3′). DNA samples were captured (approximately 1,200 RU) in flow cells dptmp, 1 and 2, leaving the fourth flow cell as a blank. Data were analyzed with ProteOn manager software, using the two states model for fitting kinetic data.

Polymerase chain reaction (PCR) stop assays

The oligonucleotide sequence (HTG21, 5′-[GGGTTAGGGTTAGGGTTAGGG]-3′, Sangon) and the corresponding complementary sequence (HTG21rev, 5′-[ATCGCTTCTCGTCCCTAACC]-3′, Sangon) were used in this assay. The experiment procedure was as previously reported method29,31. A parallel experiment was performed using a mutated oligomer, HTG21mu (5′-[GAGTTAGAGTTAGAGTTAGAG]-3′, Sangon) with its corresponding complementary sequence HTG21murev (5′-[ATCGCTTCTCGTCTCTAACT]-3′, Sangon) instead of HTG21 and HTG21rev under the same conditions.

Circular dichroism (CD) measurements

CD spectra were measured using a J-810 spectropolarimeter (JASCO, Japan) with a 1 cm cell length quartz cell, over a wavelength range of 220-360 nm at a scan speed of 200 nm/min with five acquisitions at room temperature. The AG22 oligomer (5′-[AGGGTTAGGGTTAGGGTTAGGG]-3′, Sangon) was resuspended in Tris-HCl buffer (10 mM, pH 7.4) containing 100 mM K+, 100 mM Na+ or no metal cations. The experiment procedure was as previously reported method29,31 and final analysis of the data was carried out using Origin 8.0 (OriginLab Corp.).

Telomeric repeat amplification protocol (TRAP) assays

A telomerase extract (1.0 μL, 200 cells) was prepared from HeLa cells with NP-40 lysis buffer (10 mM Tris-HCl, pH 8.0, 1.0 mM MgCl2, 1.0 mM EGTA, 1.0% NP-40, 0.25 mM sodium deoxycholate, 10% glycerol, 150 mM NaCl, 5.0 mM beta-mercaptoethanol (β-ME), 0.1 mM PMSF). Each reaction was performed in a final volume of 20 μL in a reaction mixture containing of 2.0 μL of 10× TRAP buffer (200 mM Tris-HCl, pH 8.3, 15 mM MgCl2, 630 mM KCl, 0.5% Tween-20 and 10 mM EGTA in DEPC-treated water), 1.6 μL of dNTP mix (2.5 mM, TaKaRa), 0.4 μL of TS primer (100 ng/μL, 5′-[AATCCGTCGAGCAGAGTT]-3′, Invitrogen), 0.8 μL Primer mix (ACX reverse primer, 100 ng/μL, 5′-[GCGCGGCTTACCCTTACCCTTACCCTAACC]-3′ and NT primer 100 ng/μL, 5′-[ATCGCTTCTCGGCCTTTT]-3′, Invitrogen), 2.0 μL TSNT internal control primer (4.0 × 10−11 M, 5′-[AATCCGTCGAGCAGAGTTAAAAGGCCGAGAAGCGAT]-3′, Invitrogen), 0.4 μL of RNase inhibitor (2 U/μL, TaKaRa), 0.16 μL of Taq polymerase (5 U/μL, TaKaRa), 5.0 μL of different concentrations of the compounds and 6.64 μL of DEPC-treated water. Two negative controls (designated 1 and 2) accompanied with each assay: negative control 1 involved incubating approximately 1.0 μL of cell lysate at 85°C for 10 min prior to primer extension, while negative control 2 was also included involved the incubation of approximately 1.0 μL of NP-40 lysis buffer. The experiment procedure was as previously reported method28,29.