Synthesis, Reactivity, and Biological Activity of Gold(I) Complexes Modified with Thiourea-Functionalized Tyrosine Kinase Inhibitors

Thiourea-modified 3-chloro-4-fluoroanilino-quinazoline derivatives have been studied as potential receptor-targeted carrier ligands in linear gold(I) complexes. The molecules mimic the epidermal growth factor receptor (EGFR) tyrosine kinase-targeted inhibitor gefitinib. Thiourea groups were either directly attached to quinazoline-C6 (compounds 4, 5, and 7) or linked to this position via a flexible ethylamino chain (compound 9). Compound 7 acts as a thiourea-S/quinazoline-N1 mixed-donor ligand, giving the unexpected dinuclear complex [{Au(μ-7-S,N)}2]X2 (X = Cl−, SCN−) (12a,b) (X-ray crystallography, electrospray mass spectrometry). Derivative 9 forms a stable linear complex, [Au(PEt3)(9-S)](NO3) (13). The biological activity of the carrier ligands and corresponding gold(I) complexes was studied in NCI-H460 and NCI-H1975 lung cancer cells. Compound 9 partially overcomes resistance to gefitinib in NCI-H1975, a lung cancer cell line characterized by a L858R/T790M mutation in EGFR (IC50 values of 1.7 and 30 μM, respectively). The corresponding gold complex (13) maintains activity in the low-micromolar concentration range similar to the metal-free carrier. Compound 9 and the corresponding [Au(PEt3)] complex, 13, inhibit EGFR kinase-mediated phosphorylation with sub-micromolar IC50 values similar to those observed for gefitinib under the same assay conditions. Potential mechanisms of action and reactions in biological media of this new type of hybrid agent, as well as shortcomings of the current design are discussed. Introduction Gold(I)-based complexes display a wide range of biological activities, which are thought to be mediated by the metal's reactions with (seleno)cysteine residues in proteolytic and redox-active enzymes, in particular, thioredoxin reductase (TrxR).1 While these interactions are often nonspecific due to relatively fast ligand exchange rates, fortuitous targeting of specific cysteine residues by gold(I) complexes containing sufficiently long-lived carrier ligands has also been reported.2 Several new lines of research suggest that gold(I) has a significant potential as a component of therapies designed to inhibit therapeutically relevant targets, such as zinc-finger domains3 and protein kinases.4 We were interested in the possibility of introducing reactive gold(I)-based electrophilic groups into inhibitors of tyrosine kinase (TKIs) with the goal of hijacking a thiophilic metal into the active site of this therapeutically important enzyme.5 One potential binding mechanism of these agents with epidermal growth factor receptor tyrosine kinase (EGFR-TK) would involve reaction of gold(I) with the solvent-accessible cysteine thiol residue (Cys-797) near the entrance to the enzyme's binding pocket.6 Such a mechanism might potentially inhibit binding of ATP and, consequently, kinase-mediated phosphorylation more effectively than reversible binding of the TKI alone. The desired pharmacophore would combine the high sulfur affinity of gold(I) with the selectivity and submicromolar binding affinity of an EGFR-TK-targeted ligand. Mutationally activated EGFR-TKs are considered a major driver of cancer cell survival and aggressive tumor growth.5 In nonsmall cell lung cancer (NSCLC), expression levels of EGFR are inversely correlated with survival of the disease.7,8 Somatic (activating) EGFR mutations sensitize cancer cells to small-molecule TKIs, which target the enzyme's ATP binding site and show potent antiproliferative properties.9 Gefitinib (Chart 1) is a quinazoline-based TKI indicated against cancers harboring aberrant EGFR, in particular, lung carcinomas.9 Unfortunately, the efficacy of this drug is limited by the emergence of acquired resistance as a consequence of a secondary mutation within the ATP binding pocket (T790M).10 One currently pursued approach to combatting this form of resistance observed for mutant EGFR and other clinically relevant kinases is to turn the reversible TKIs into irreversible inhibitors, such as the recently FDA-approved inhibitor afatinib (Chart 1).11 These molecules contain strategically positioned reactive electrophilic groups (usually Michael acceptors), which are able to form a covalent bond with accessible cysteine residues in the active sites of targetable enzymes. Several irreversible TKIs are currently being evaluated in the clinic.12 Chart 1 Structures of Gefitinib and Afatinib Showing Relevant Atom Numbering for the Quinazoline Scaffold Introduction of gold(I) as a cysteine-targeting moiety in place of a Michael acceptor can be considered a new strategy of improving the activity of TKIs against resistant disease. The purpose of the current study was to assess if the gefitinib-type quinazoline scaffold can be functionalized with a thiourea moiety without compromising the binding affinity of the TKI, and if the S-donor group can be modified with thiophilic (pseudo)halide and phosphine substituted gold(I) electrophiles. A new TKI derivative was identified that strongly binds to the tyrosine kinase ATP-binding pocket and partially overcomes resistance to gefitinib in NSCLC cell lines harboring wild-type and mutated EGFR kinase domains. In the corresponding gold(I) complex, the metal appears to have an effect on both the biological activity and kinase binding affinity of the inhibitor, but the linear thiourea–Au(I)–phosphine coordination appears to be too labile for the desired targeted application.