Guanine binding of a cytotoxic platinum-acridin-9-ylthiourea conjugate monitored by 1-D 1H and 2-D [1H,15N] NMR spectroscopy: Hydrolysis is not the rate-determining step

The reaction of [PtCl(en)(ACRAMTU)](NO3)2 (PT-ACRAMTU, 1; ACRAMTU = 1-[2-(acridin-9-ylamino)ethyl]-1,3-dimethylthiourea, en = ethane-1,2-diamine) and the [15N]-en labeled analogue, 1′, with 2′-deoxyguanosine (dG) was studied by 1H NMR and two-dimensional [1H,15N] HSQC (heteronuclear single quantum coherence) spectroscopy. Reactions were performed in phosphate buffered solution at 37 °C at various ratios and total concentrations of reactants. The 1H NMR data suggest that the hydrolyzed form of the drug, [Pt(H2O)(en)(ACRAMTU)]3+ (1a), forms at a rate (k1) similar to that observed in classical platinum chloroam(m)ines but to only a minor extent (∼15%). Attempts to detect and characterize 1′a by two-dimensional NMR spectroscopy, however, were unsuccessful, and 1′ and dG∗ were the only species observed in the HSQC spectra. Reaction of the putative aqua intermediate 1a with dG to yield [Pt(en)(dG-N7)(ACRAMTU)]3+ (dG∗) is slow and is highly dependent on the initial concentrations of the reactants. This unusual observation is consistent with a mechanism in which a second-order term becomes rate-determining (k2 < k1) (the opposite situation is usually observed in cisplatin-type complexes). The possibility of direct substitution of chloride in 1 by dG-N7 (k3) and the implications of the data acquired in this model system for the binding of the conjugate to double-stranded DNA are discussed.