The DNA interference pathways exhibited by cisplatin and related anticancer active metal complexes have been extensively studied. Much less is known to what extent RNA interaction pathways may operate in parallel, and perhaps contribute to both antineoplastic activity and toxicity. The present study was designed with the aim of comparing the reactivity of two model systems comprising RNA and DNA hairpins, r(CGCGUUGUUCGCG) and d(CGCGTTGTTCGCG), towards a series of platinum(II) complexes. Three platinum complexes were used as metallation reagents; cis-[PtCl(NH3)2(OH2)]+ (1), cis-[PtCl(NH3)(c-C6H11NH2)(OH2)]+ (2), and trans-[PtCl(NH3)(quinoline)(OH2)]+ (3). The reaction kinetics were studied at pH 6.0, 25 degrees C, and 1.0 mM < or = I < or = 500 mM. For both types of nucleic acid targets, compound 3 was found to react about 1 order of magnitude more rapidly than compounds 1 and 2. Further, all platinum compounds exhibited a more pronounced salt dependence for the interaction with r(CGCGUUGUUCGCG). Chemical and enzymatic cleavage studies revealed similar interaction patterns with r(CGCGUUGUUCGCG) after long exposure times to 1 and 2. A substantial decrease of cleavage intensity was found at residues G4 and G7, indicative of bifunctional adduct formation. Circular dichroism studies showed that platinum adduct formation leads to a structural change of the ribonucleic acid. Thermal denaturation studies revealed platination to cause a decrease of the RNA melting temperatures by 5-10 degrees C. Our observations therefore suggest that RNA is a kinetically competitive target to DNA. Furthermore, platination causes destabilization of RNA structural elements, which may lead to deleterious intracellular effects on biologically relevant RNA targets.

The anticancer active complex cisplatin interacts preferentially with the common, G-C rich, wobble base pair region of both tRNA(Ala) and Mh(Ala) in a reaction that at pH 6.3 is rate limited by the acid hydrolysis of the metal complex.

Nuclear DNA is a well characterized target for many low molecular metal-based drugs, with cisplatin and related antineoplastic compounds as typical examples. Much less is known concerning to what extent targeting of RNA may influence the activity spectrum of these types of drugs. In a preliminary communication by us (Papsai et al., Dalton Trans., 2006, 3515) we were able to show that the folded, three-dimensionally well defined structure of tRNA(Ala) readily interacts with cisplatin. In the present study we have further analyzed the binding preferences within the preferentially targeted stem region (sMh(Ala)) by modulation of the sequence around the G-U wobble base-pair and the net charge of the 3' and 5' ends. Our data show that the adduct profile is strongly influenced by the presence of the 5' end phosphate group. Further, the adduct formation reaction can be prevented by replacement of the G-U wobble base-pair with the fully complementary G-C pair. To further investigate the influence from local sequence on the platination process, a model of the anticodon region (acMh(Ala)) was also investigated. In the absence of consecutive guanine-residues in the stem- and anticodon regions, preferential platination was found to take place at the terminal AG-site in the stem region. However, after introduction of a GG-pair in the anticodon loop, platination was observed also here. At 37 degrees C, pH 6.3 and C(Pt) = 0.10 mM the rate of platination was determined to be ca. 1 x 10(-4) s(-1), with the most rapid reaction observed for interaction with the anticodon model carrying two adjacent guanines in the single-stranded loop. Together, these data show that platination of RNA is highly sequence- and structure-dependent.

Wnt-5a is a secreted glycoprotein which has been shown to be involved in the regulation of cell adhesion and motility, processes which are of importance in metastasis formation by cancer cells. We here present an initial study aiming at evaluating whether small interfering RNA (siRNA) in combination with cisplatin can be used to modulate protein expression levels under in vitro conditions. For this purpose, an AU-rich region corresponding to the initial 260 bases of the Wnt-5a 3' untranslated region was chosen as the target. The effect of four different siRNAs was evaluated by analysis of protein suppression levels in rabbit reticulocyte lysate (RRL) and an immortalized noncancerous mammary epithelial (HB2) cell line by monitoring the activity of transiently expressed luciferase. The specificity and kinetics for hybridization of the siRNA with the messenger RNA target were followed by digestion techniques and analysis by polyacrylamide gel electrophoresis. Specific and temperature-dependent hybridization was observed, with a half-life of approximately 0.5 h at 4 degrees C. Significant downregulation of luciferase activity was obtained in the micromolar and nanomolar range, for RRL and HB2, respectively. In addition, the downregulation of protein production caused by addition of cisplatin could be further potentiated by addition of siRNA in a selective manner. The latter observation suggests that combined use of cisplatin and siRNA could be a method to decrease therapeutically used cisplatin concentrations. Thus, toxic side effects could be minimized while key proteins are targeted in a highly specific manner.