Recent Developments in Gold Drugs
Renato Skerjl, Beth R Cameron, Renee Mosi, Yongbao Zhu,
The ability of metals to react with a variety of biological ligands has been exploited by medicinal chemists and there are a variety of metal-based drugs from the simple antacid magnesium and aluminium oxides to the anti-tumour platinum drugs cisplatin and carboplatin. Gold drugs, which have been used for the treatment of rheumatoid arthritis since the mid 20th century, are of the disease-modifying antirheumatic drugs (DMARD) class. Methotrexate is often the preferred DMARD, and with the advent of new agents such as the anti-TNF antibody Infliximab, gold drugs are now used infrequently. Since the introduction of auranofin in 1985 there has been no new clinically approved gold drug for either rheumatoid arthritis or any other disease.
In spite of this apparent lack of progress there has been a great deal of interest in the use of gold compounds for cancer therapy. There were early indications that auranofin had limited anti-tumour activity in in vitro systems and in the P388 leukaemia in vivo, however it was inactive against solid tumours. More promising indications were achieved with a series of digold phosphine complexes. The lead compound [dppe(AuCl)2] was shown to rearrange to give the more stable tetrahedral complex [Au(dppe)2]+. The proposed mechanism of action for [Au(dppe)2]Cl was the formation of DNA-protein cross-links, the lack of affinity for Au(I) for O and N containing ligands resulting in poor reactivity with the bases of DNA . Though this compound had marked activity against P388 it had limited activity against solid tumour models. This compound was not entered for clinical trials, however, due to problems with cardiotoxicity highlighted during pre-clinical toxicology studies.
Gold(III) complexes have not been as thoroughly investigated as gold(I) complexes, primarily because of their reactivity. Gold(III) is isoelectronic (d8) with platinum(II) and likewise forms square planar complexes. It is therefore tempting to speculate that such complexes would have similar antitumour activity to cisplatin. In order to stabilise the gold(III) oxidation state in a reducing biological milieu complexes were synthesised with a single mononegative bidentate ligand, damp, (2-[(dimethylamino)methyl]phenyl), and two monodentate anionic ligands e.g. Cl or acetate (OAc). The damp ligand forms part of a five-membered chelate ring in which the nitrogen of the amine group and the carbon of the aryl ring bond to the metal. The monodentate ligands are readily hydrolysed and are available for substitution. These compounds had limited anti-tumour activity in in vivo models using human tumour xenografts in immune-deprived mice. Other groups have adopted a similar approach. Marcon et al have reported the cytotoxicity of Au(III) complexes such as [Au(phen)Cl2]+ , [Au(en)2]3+, and [Au(terpy)Cl]2+ and Dinger and Henderson have reported on the in vitro biological activity of Au(III) (damp) thiosalicylate and salicylate complexes. In both of these cases in vivo biological activity has still to be demonstrated. Studies in our laboratory however demonstrated that the mechanism of the Au(III)(damp) complexes was different to cisplatin. We hypothesise that an alternative molecular target for these compounds is biologically important thiol-containing molecules.
The cysteine protease cathepsin B has been proposed to play a role in tumour metastasis and progression. Other potential thiol-containing molecular targets include cathepsin K (bone disease), L and S (inflammation), caspases (cancer), thioredoxin reductase (cancer), and transcription factors such as NFk-B and its regulator IKK (cancer and inflammation). Gold compounds have been shown to be inhibitors of enzyme activity and transcription. With this in mind we have adopted two parallel approaches to drug discovery, a mechanism-based approach, and a disease-oriented (cancer) approach.
Mechanistic studies focussed on identifying compounds with activity against cysteine proteases. To this end we cloned and isolated the cathepsins B, K, L, and S, and developed assays for enzyme inhibition. The Au(III)damp complexes were first tested against cathepsin B and were found to be moderate inhibitors. Modifications were made to the Au(III) damp structure to give compounds with a six-membered chelate ring comprised of a phenyl and pyridyl rings linked by a linker X with monodentate ligands L=Cl or acetate. Further modifications were explored including changing the linker (X) and the substituents (R) on the phenyl or pyridine ring. The change from a methylene group to other linkers such as oxygen, sulfur or even a carbonyl group significantly alters the activity against cathepsin B whereas the activity with cathepsin K is influenced by the addition of bulky lipophilic groups on the phenyl ring such as butyl or phenyl. A methyl group at the 6-position on the pyridyl ring increased the potency against all the cysteine proteases. Kinetic analysis with compounds where L=Cl or OAC showed that inhibition was time dependent. In addition, this inhibition could be partially reversed by incubation with cysteine indicating that the compounds are acting as tight binding reversible inhibitors.
The disease-oriented approach involved screening compounds for cytotoxicity against a panel of human tumour cell lines. These cell lines were chosen as being representative of different tissue types. In addition they were tested for intracellular cathepsin B and thioredoxin reductase activity. Compounds with L=Cl were found to be unstable in a biological matrix containing plasma therefore a series of complexes were synthesised where L= dithiolate or thiosalicyate. The more stable dithiolate complexes were less active in the short-term enzyme assays but were cytotoxic towards the tumour cells over a 72 hour incubation period. Three compounds, two dithiolates and one thiosalicylate, were selected for in vivo evaluation against the most sensitive tumour line, HT29 colon carcinoma, grown as a xenograft in immune-deprived mice. All three compounds showed modest anti-tumour activity but with no marked improvement over the parent Au(III)damp complexes.
Initial indications are that Au(III) compounds targeted against disease-specific thiol-containing biological molecules have potential as drugs, particularly for cancer. However, further work is required to produce a target-specific drug with a suitable pharmacological activity and toxicity profile.
Gold drugs, Cathepsin B, Cancer