Exploring the C^N^C theme: Synthesis and biological properties of tridentate cyclometalated gold(III) complexes

Abstract

A family of cyclometalated Au(III) complexes featuring a tridentate C^N^C scaffold has been synthesized and characterized. Microwave assisted synthesis of the ligands has also been exploited and optimized. The biological properties of the thus formed compounds have been studied in cancer cells and demonstrate generally moderate antiproliferative effects. Initial mechanistic insights have also been gained on the gold complex [Au(C^N^C)(GluS)] (3), and support the idea that the thioredoxin system may be a target for this family of compounds together with other relevant intracellular thiol-containing molecules.

Introduction

Gold-based organometallic complexes as anti-cancer agents have become increasingly popular in recent years, and several reviews have been published highlighting their structural diversity and biological activity.¹ In fact, organometallic complexes display numerous attractive features. For example, while the organic ligand allows for the introduction of stereospecificity and alteration of the physicochemical properties by choosing different functional groups, the metal–carbon (M–C) bond provides strong trans influence and, in the case of π-bonded aromatic arene and cyclopentadienyl ligands, can act both as electron donors and acceptors. Furthermore, variations of the organic moiety are accessible, allowing the “fine-tuning” of the physiochemical properties of the respective metal compound. Finally, by choosing specific targeting moieties, either incorporation into the ligand or tethering to the metal centre can be achieved in a modular approach.

Typical classes of organometallics include metallocenes, metallo-arenes, metallo-carbonyls, metallo-carbenes (e.g. N-heterocyclic carbenes, NHC), alkynyl complexes, etc.² featuring both mononuclear and multinuclear scaffolds. All these compound families have been extensively applied in catalysis during the last decades.³ However, their medicinal use has been considered only much more recently.1(c), 2, 4

Concerning Au(I) organometallics, in 2008 Berners-Price et al. synthesized a series of mononuclear cationic Au(I) biscarbene complexes that show remarkable cytotoxicity in vitro and are able to induce mitochondrial damage.⁵ Since then, the effects of Au(I) NHC complexes on cell metabolism and their interference with pathways relevant to cancer cell proliferation have been studied in a broad number of cases.1(a), 4(b), 6

In this context, to explore the design of Au(III) compounds for biological applications, cyclometalation is a convenient method to stabilize the otherwise easily reduced Au(III) centre, and numerous cyclometalated scaffolds have been synthesized, including C^N, C^N^C, C^N^N and C^N^S.1(b), 1(c), 7 Cyclometalated Au(III) C^N compounds of general formula [(Au(damp)X₂] (Fig. 1) with anticancer properties were first investigated by Parish, Buckley et al. in 1996,⁸ featuring a 2[dimethylamino)methyl]-phenyl (damp) backbone. The compounds display cytotoxic activity which is comparable to that of cisplatin, against a variety cancer cell lines. This activity is also accompanied by high selectivity and cytotoxicity in vitro which translates to moderate activity in vivo.

In 1996, the synthesis of an Au(III) 2-benzylpyridine derivative [Au(py^b-H)Cl₂] (py^b-H = C^N cyclometalated 2-benzylpyridine) (Fig. 1) was reported by Cinellu et al.⁹ In 2015 Casini, Cinellu et al. synthesized a structural analogue of this complex replacing the chlorido ligand, trans to the nitrogen atom, with 1,3,5-triazaphosphaadamantane (PTA).¹⁰ The resulting compound displays good cytotoxic activity against various cancer cell lines such as A2780 (human ovarian adenocarcinoma). Furthermore, dinuclear oxo-bridged Au(III) C^N^N complexes of formula [(C^N^N)₂Au₂(µ-O)](PF₆)₂ (with C^N^N = 6-(1-methylbenzyl)-2,2‘-bipyridine or 6-(1,1-dimethylbenzyl)-2,2′-bipyridine) showing moderate cytotoxicity against various cancer cell lines were synthesized.¹¹ (Fig. 1).

The field of Au(III) C^N^C complexes with anticancer properties has been closely examined by Che and co-workers1(b), 12 For example, dinuclear complexes of the type [Au_m(C^N^C)_mL]ⁿ⁺ (with HC^N^CH = 2,6-diphenylpyridine; m = 1–3; n = 0–3) showed higher cytotoxic activity against various cancer cell lines than their mononuclear counterparts.¹²

The highest cytotoxicities were observed for complex [Au₂(C^N^C)₂(μ-dppp)](OTf)₂ (Fig. 1), relating to the cytotoxicity of the free 1,2-bis(diphenylphosphino)propane (dppp) ligand. By replacing the phosphane moiety with an NHC ligand as in the cationic C^N^C stabilized complexes [Au_n(R–C^N^C)_n(NHC)]ⁿ⁺, a general decrease in cytotoxic effects has been noted,¹³ supporting the idea that indeed phosphine ligand-mediated cytotoxicity plays a major role in the overall anticancer properties. Nevertheless, within this series, the mononuclear complex [Au(C^N^C)(IMe)]CF₃SO₃ (IMe = 1,3-dimethylimidazol-2-ylidene) (Fig. 1) shows higher cytotoxic activity than its dinuclear analogue, and a high degree of selectivity towards human cancer cells compared to normal lung fibroblasts (CCD-19Lu). Through DNA interaction studies it was demonstrated that the compound induces DNA strand breaks and can cause subsequent cell death through the stabilization of Topoisomerase-linked DNA.¹⁴ Treatment of nude mice bearing PLC tumors (hepatocellular carcinoma) with [Au(C^N^C)(IMe)]CF₃SO₃ at 10 mg/kg/week for 28 days significantly suppressed (47%) tumor growth when compared with that of the vehicle control.¹³

Of note from the same group, is the application of supramolecular polymers, self-assembled from cyclometalated Au(III) C^N^C complexes, as anticancer agents. The mononuclear complex [Au(C^N^C)(4-dpt)]⁺ (C^N^C = 2,6-diphenylpyridine, 4-dpt = 2,4-diamino-6-(4-pyridyl)-1,3,5-triazine) (Fig. 1) was chosen due to the ability of the antiangiogenic 4-dpt ligand to form intramolecular hydrogen bonds and to establish π-π interactions, leading to supramolecular complex formation by self-assembly at ambient temperatures.¹⁵ The supramolecular polymer displays high cytotoxic activity towards B16 cells (murine cancer).

In terms of possible mechanisms of action, cytotoxic Au(III) complexes can interact with protein targets,¹⁶ including those constituting the thioredoxin system, often overexpressed in tumor cells and involved in maintaining the intracellular redox balance.¹⁷ Among the enzymes included in this system, the seleno-protein thioredoxin reductase (TrxR) contains a cysteine-selenocysteine redox pair at the C-terminal active site, and the solvent-accessible selenolate group, arising from enzymatic reduction, constitutes a likely target for “soft” metal ions such as gold. In addition, Au(III) complexes can interfere with thiols including glutathione, the most abundant intracellular reducing agent.

Within the scope of new medicinally relevant organometallic Au(III) compounds, two new cyclometalated Au(III) complexes of the general formula [Au(III)(C^N^C)X] (with C^N^C = 2,6-diphenylpyridine and X = Cl, thio-β-d-glucose-tetraacetate (GluS), 1,3,5-triazaphosphaadamantane (PTA)) were synthesized. The complexes were prepared starting from the literature known chlorido precursor [Au(III)(C^N^C)Cl]¹⁸ (Fig. 2, Fig. 1, 1), substituting the chlorido ligand with PTA (2) and thio-β-d-glucose-tetraacetate (GluSH) (3), respectively. The PTA ligand was chosen to increase the water solubility, while the GluS⁻ ligand was selected as modulator of the hydrophilic/lipophilic character, as well as a possible facilitator of the compound’s uptake via interactions with the GLUT1 transporter. A second series of analogous C^N^C complexes – Au(III)[(C^N^R^C)] – was synthesized starting from the corresponding 2,4,6-triarylpyridine ligands, but featuring different substituents in the para position of the phenylpyridine (R = OH, F, Br, NO₂) (Fig. 2, Fig. 4, Fig. 5, Fig. 6, 4–7) in order to evaluate the effects of this ligand modification on the biological properties. Notably, complex 5 has already been described by Che et al. in 2013 as a precursor for photoactive functionalized bis-cyclometalated alkynyl gold(III) complexes suitable as phosphorescent organic light-emitting diodes (OLEDs).¹⁹ However, the authors did not evaluate the biological activity of this complex. All the cyclometalated Au(III) complexes reported herein were studied for their antiproliferative effects against different human cancer cell lines, including some that are resistant to cisplatin. Moreover, compounds 1–3 were also tested for their inhibitory properties of TrxR on the purified protein and in cell lysates. The effects of the new complexes on the oxidation state of Trx were also investigated by Western blot analysis. Discrimination between the oxidation of the GSH/GSSG system and the Trx system can be very informative in terms of mechanisms of toxicity since different cellular pathways are controlled by GSH and Trx. Thus, estimation of the glutathione content was performed in treated cells.