1.5. An overview of G-quadruplex DNA binding agent
1.5.3 Metal Schiff base complexes
One of the most widely investigated classes of compounds in coordination chemistry is metal complexes of Schiff base ligands. In these complexes, imine nitrogen atoms and deprotonated phenolic oxygen atoms coordinate strongly to a variety of metal ions in a tetradentate fashion.[124, 125] In recent years, complexes of salen (the abbreviation for salicylidene ethylenediamine) (e.g. Figure 1.18, entry 7) and salphen (salicylidene phenylenediamine) coordinated to various transition metals have emerged as promising G-quadruplex DNA binding agents. For example, the nickel salphen complex (54) was shown using FRET melting assays to produce a significant degree of stabilisation of a unimolecular G-quadruplex (Tm = 29.5 ± 0.4
oC), while TRAP assays revealed the complex could also inhibit telomerase (IC50 = 0.14 ± 0.01 M). These results, along with a high degree of binding selectivity in
36 favour of G-quadruplex over dsDNA, were attributed to the square planar geometry of the metal complex.[93, 94]
Entry M Ref X R Position
of R
1
NiII (54) CuII (55) ZnII (56) VIV=O (57)
[93], [94]
[126]
4
2 NiII (58) [127] 4
3 NiII (59) [128] 4
4 NiII (60) [126] 4
5 NiII (61) [95],
[129] 5
6 Niii (62)
CuII (63) [94] 4
7 NiII (64)
CuII (65) [130] 3
8 PtII (66) [96]
4
9
NiII (67) CuII (68) ZnII (69)
[97]
3
Figure 1.18 Metal Schiff base complexes investigated as G-quadruplex binding agents.
37 An attractive feature of this class of compounds is that the synthetic methods used to prepare them are often simple to perform, and can be readily modified by incorporating different starting materials to produce a wide range of potential drug molecules. Molecular modelling and X-ray crystallographic studies have shown that the metal ions in these square planar complexes can become aligned with the center of a G-tetrad when binding interactions with G-quadruplexes occur (Figure 1.19).[93, 94, 96] It has been suggested that coordination of metal ions to the Schiff base ligand would result in withdrawal of electron density from the latter, and strengthen
stacking interactions with an electron rich G-quartet.[57]
(a) (b)
Figure 1.19 Interactions between a nickel Schiff base complex and G-quadruplex DNA:
(a) Results of a molecular docking study involving (54) and a unimolecular parallel G- quadruplex DNA (Elements are represented by the following colours: grey: carbon; green:
nickel; red: oxygen);[93] (b) Solid state structure of (55) bound to a bimolecular parallel G- quadruplex DNA as revealed by X-ray crystallography.[94]
Using as a starting point (54), Vilar and co-workers have prepared a range of analogues in order to understand how varying the structure affects binding towards G-quadruplexes.[93, 94, 126] In general, NiII complexes have superior G- quadruplex-binding properties compared to their CuII counterparts. However, complex (55) has been shown to inhibit telomerase to a greater extent using the TRAP assay (telEC50 = 3.6 M) and exhibit superior anti-proliferative effects towards A549 lung cancer cells. These results suggest that some complexes may elicit anti-
38 proliferative effects by mechanism other than binding and stabilising telomeric G- quadruplexes, such as direct binding at the active site of the enzyme or targeting other DNA sequences such as promoter regions.[94]
A number of studies have now been carried out to explore the relationship between the G-quadruplex binding properties of Schiff base complexes involving different metal ions and their structures.[94, 97, 126, 131] Crystallographic studies showed that the NiII complexes examined (54), (58) – (62), (64), as well as PtII complex (66) and CuII complexes (55), (63), (66), and (68) possess square planar geometries. In contrast, the VIV oxo complex (57) and the ZnII complexes (56) and (69) exhibited square pyramidal geometries.[107, 126] The results of DNA binding studies conducted using CD spectroscopy, FID, SPR and FRET melting assays, showed that the square planar complexes were more effective than those with other geometries. For example, the results of FRET experiments showed that the melting temperature of a G-quadruplex increased by 29.5 and 19.2 °C when bound to the square planar complexes (54) and (55), respectively. In contrast, the melting temperature only increased by 1.4 °C in the presence of the square pyramidal complex (56).[93, 94, 126]
A number of research groups have examined the effect on G-quadruplex affinity and selectivity of replacing the phenylenediamine group in (54) by other moieties, resulting in the compounds shown in Figure 1.18.[93, 95, 96, 125, 126, 127, 129] These changes have in some instances resulted in significant changes to the G-quadruplex DNA binding properties of the nickel complex. For example, complex (59), which contains a phenanthrenediamine moiety instead of the phenylenediamine present in (54), showed greater affinity towards dsDNA, and
39 therefore a lower G-quadruplex DNA/dsDNA selecivity factor.[128] In contrast, steric hindrance from the meso-diphenylethylenediamine group in (58) resulted in a much lower affinity towards dsDNA, and notable selectivity for both tetramolecular and unimolecular G-quadruplexes.[107, 127]
Each of the nickel complexes shown in Figure 1.18 contain two pendant groups that either feature a permanent positive charge, or are likely to undergo protonation in aqueous solution, leading to a positive charge. One of the functions of these groups is to increase the solubility in water of the complexes. In addition, it is believed that these pendant groups participate in additional interactions with the loops and grooves of G-quadruplex that may affect overall affinity.[93, 94, 126, 132]
There has been significant effort devoted to optimising the G-quadruplex binding affinity of metal Schiff base complexes through modification of these pendant groups. This has included, for example, changing the length of the alkyl groups present in the pendant groups,[95, 126, 127] (Figure 1.18, entry 2), or altering their positions on the periphery of the Schiff base ligand scaffold (Figure 1.18, entry 5, 7, 9).[95, 96]
To date there have only been a small number of investigations into cellular uptake of metal Schiff base complexes, and the effects of cellular incorporation. In 2009, Mandal and co-workers demonstrated that MnIII salen and salphen complexes can influence cell viability and induce apoptosis in MCF7 breast cancer cells.[133]
The IC50 values of the complexes with MCF7 ranged from 12 to 55 M, which indicated that they are only weakly cytotoxic.[133] In another study, Che et al.
designed a series of PtII salphen and salen complexes including (66), and examined their interaction with c-myc G-quadruplex DNA. This study was motivated by the
40 knowledge that suppressed transcription of c-myc is one of the hallmarks of many types of cancers.[134] Complex (66) showed a capacity to influence the activity of the c-myc gene promoter (IC50 = 4.4 M), as well as inhibit c-myc transcription in human hepatocellular carcinoma (HepG2) cells (IC50 = 8.07 ± 0.11 M).[96]
Recently, Barone and co-workers synthesised a group of new Schiff base complexes, including (67), using naphthalendiamine as one of the starting materials.
[97] This complex was shown using a PCR stop assay to induce formation of a G-quadruplex structure by an oligonucleotide, Pu22myc, which corresponds to the nuclease hypersensitive element (NHE) III1 region of the c-myc promoter. The interaction between (67) and Pu22myc resulted in a significant degree of inhibition of amplification of Pu22myc (IC50 ~ 0.2 M). In addition, the metal complex proved cytotoxic towards MCF7 and HeLa cancer cells in a concentration dependent manner, with Growth Inhibition 50% (GI50) values (after 48 hours) of 1.42 and 0.31
M, respectively.[97]
In contrast to the numerous investigations into the DNA binding ability of other classes of metal complexes of ligands such as porphyrins and its analogues, there have still only been a relatively small number of similar studies involving complexes of Schiff base ligands. Despite this, it has already been demonstrated that the latter are a promising class of compounds, which have the ability to bind to and stabilise G-quadruplex structures, as well as inhibit telomerase and interfere with the regulation of the activity of different types of oncogenes. Further exploration of the DNA binding properties of new examples of this class of compounds is therefore warranted. In the following sections some of the different methods which have been
41 used to examine the G-quadruplex binding properties of metal complexes are discussed.