Synthesis and Pharmacological Evaluation of Modified Adenosines Joined to Mono-Functional Platinum Moieties

The synthesis of four novel platinum complexes, bearing N6-(6-amino-hexyl)adenosine or a 1,6-di(adenosin-N6-yl)-hexane respectively, as ligands of mono-functional cisplatin or monochloro(ethylendiamine)platinum(II), is reported. The chemistry exploits the high affinity of the charged platinum centres towards the N7 position of the adenosine base system and a primary amine of an alkyl chain installed on the C6 position of the purine. The cytotoxic behaviour of the synthesized complexes has been studied in A549 adenocarcinomic human alveolar basal epithelial and MCF7 human breast adenocarcinomic cancer cell lines, in order to investigate their effects on cell viability and proliferation.


Introduction
The discovery of the anti-proliferative properties of cisplatin [1] marked the beginning of modern chemotherapy based on the use of metal complexes capable of blocking the replication of cancer cells targeting the nuclear DNA [2,3]. In fact, it is widely accepted that cisplatin, once inside the cell, may form a highly reactive species [4,5] that can react with DNA through the formation of intra-strand linkages [6,7]. Such linkages alter the secondary structure of DNA, resulting in an inhibition of transcription and replication, ultimately leading to cell death [4]. However, the poor solubility in biological fluids [8], the serious side effects [9,10], and, more importantly, the intrinsic and acquired resistance of many types of tumors [5], have limited its use in the clinic. Carboplatin and oxaliplatin, second and third generation anti-neoplastic agents, respectively, are able to enhance the quality of life of patients in terms of dosage and drug administration. However, they can still trigger mechanisms of resistance (intrinsic and/or acquired) during repeated cycles of oncological treatment [5]. It was then discovered that the presence of more charged metal centres separated by unbranched alkylamine chains leads to complexes capable of being uptaken by cells via active import [11] and of overcoming the intrinsic and/or acquired resistance in some tumors [6,[12][13][14][15]. Based on the understanding that they may exert anti-tumoral activities through inter-strand linkages with DNA that cannot be repaired by enzymes [16], many multinuclear platinum complexes have recently been synthesized, by varying the diamine backbone chain length, or by introducing modified linkers [8,[10][11][12]17,18]. In addition, new Pt-based compounds have been obtained by conjugating biologically important substances or drugs to Pt-containing subunits [19][20][21][22][23][24] and novel drug-delivery based methodologies have been explored to vehiculate platinum-based anti-cancer complexes directly against tumors [25,26].
Nucleoside and nucleotide anti-metabolites and their base analogues are able to inhibit specific pathways of the cancer cell metabolism by blocking the biosynthesis or the function of nucleic acids. For example, the combination of cisplatin and 5-fluorouracil, a chemotherapeutic agent that inhibits thymidylate synthase, has been extensively used in clinical practice to treat various types of cancer. Acyclovir, a guanosine nucleoside analogue containing an open-chain sugar surrogate, has been used as a ligand for platinum and the corresponding complex exhibits high in vitro activity against various herpes viruses [27]. Furthermore, the synthesis and the preliminary pharmacological activity of novel modified adenosines and thymidines, employed as N-donor ligands of platinum(II) dichloride complexes, have been recently reported [28,29].
Recently, we have reported the solid-phase synthesis and the pharmacological activity of the first examples of bis-platinated nucleoside complexes in which the mono-functional metal is linked both to N-7 of the purine nucleus of inosine and to the terminal amino-group of a hexylamine side chain installed on N-1 (compounds 1-3, Figure 1) [30]. The amino-alkyl chain was introduced on the purine base system through a chemical strategy recently developed by us [30][31][32][33][34][35][36]. These complexes showed very good water solubility thanks to the charged platinum centres and to the ribose hydroxyl groups. They were tested against four different human tumor cell lines and, in particular, the complex bearing two monofunctional cisplatin units was revealed to be more cytotoxic than cisplatin against the MCF7 cancer cell line in a short-term exposure assay [30].  [30].
In the light of the above results, to investigate the importance of nucleoside scaffolds in the construction of novel platinum complexes we have speculated about the antitumoral activity of cisplatin-adenosine complexes. As a result, in this paper we report on the synthesis of four dinuclear platinum complexes 8a,b and 10a,b (Scheme 1) carrying N 6 -(6-aminohexyl)adenosine (compound 5, Scheme 1) or a 1,6-di-(adenosin-N 6 -yl)-hexane (compound 9, Scheme 1) respectively, as ligands of mono-functional cisplatin or monochloro(ethylene diamine)platinum(II) (compounds 7a and 7b, respectively, Scheme 1).
Platinum complexes, endowed with particular structural strains, have been designed to evaluate alternative interactions with DNA strands [8,12,17,37,38]. The novel prepared complexes 8a,b and 10a,b are characterized by a rigid planar purine substructure joined to a flexible hexylamine chain. On this basis, it could be expected that they may alter the binding mode with DNA through, for example, base-stacking or by the formation of multiple hydrogen bonds. Therefore, we have investigated the preliminary effects of these compounds on viability and proliferation in two different human cancer cell lines.

Synthesis and Characterization
The synthesis of the novel modified adenosine-based platinum complexes 8a,b and 10a,b is depicted in Scheme 1. 6-Cloropurine riboside 4 proved to be a good candidate to obtain the key intermediate 5 from which the construction of complexes 8a,b and 10a,b has been accomplished. As reported by Schammells et al. [39] seven equivalents of 1,6-diaminohexane have proved sufficient to convert 4 into 5 in a very good yield (86%), avoiding the formation of dimeric species, which would complicate the purification procedure. In fact, the recovery of 5 from the reaction mixture was easily performed thanks to its insolubility in EtOH upon cooling. Compound 9 was prepared by reacting 4 with a slight excess of 5. During the reaction, this substance precipitated from the reaction mixture as a white solid which was then collected in 93% yield by filtration. This represents a very good improvement on the reported yield for 9 (12%) [40].
Next, the platinum-containing moieties were installed. In particular, treatment of 5 and 9 with a seven-fold excess of the suitable platinating complex 7a,b, activated by overnight reaction with AgNO 3 (0.9 equiv.) in DMF, furnished the bisplatinated compounds 8a,b and 10a,b. Purification of these substances could be accomplished by reverse-phase HPLC only using a gradient of CH 3 CN in 0.1 M triethylammonium bicarbonate buffer (TEAB) as the solvent mixture, whereas complex chromatographic patterns were observed in the absence of the TEAB.
The structures of complexes 8a,b and 10a,b (yields 58%-62%) were supported by 2D-NMR and positive mode high resolution mass spectrometry (HRMS) data; whereas their purity was ascertained by CHN analyses.
In Figure 2 (panel A) the representative HRMS spectrum of complex 8a is reported; in particular, the isotopic pattern of the base peak (expansion, panel B) perfectly fits with that of the calculated one (expansion, panel C), confirming the presence of two Pt and two Cl atoms and a net 2+ charge.
In the 1 H-NMR spectra of 8a,b and 10a,b the downfield shift of H-8, compared with the resonance of the same proton in 5 and 7 (∆δ = 0.6 and 0.5, respectively), confirmed the presence of the N(7) → Pt bond in all these substances [41,42]. In Table 1 the differences between the 13 C-NMR shifts of the Pt-coordinated and not-coordinated purine carbon atoms (∆δ = δ complex − δ ligand , ppm) are listed: in the case of complexes 8a,b significant coordination shifts were found for the C-5 and C-8 atoms, whereas for the complexes 10a,b the C-5 atoms underwent the major shifts. Further evidence of metallation at N-7 come from the increased acidity of the H-8 as indicated by the reduced intensity of the pertinent signal in the 1 H-NMR spectra of 8a,b and 10a,b, when these complexes were dissolved in D 2 O and the spectra acquired after several hours [43]. 1 H-NMR analyses excluded also the presence of equilibria involving migrations of the platinum moieties from N-7 to N-1 of the nucleobases; such migrations would shift the H-2 resonances of complexes 8a,b and 10a,b to higher frequences in comparison to the same signals of compounds 5 and 9, respectively [44]. The resonances of the methylene protons belonging to the CH 2 -NH 2 -Pt moieties were seen as broad partly overlapped triplets in the range 2.3-2.7 ppm, in the proton spectra of 8b and 10b, likewise the protons of the methylene group geminal to platinum in 8a resonated as a triplet at δ 2.6. In 13 C-NMR spectra of 8a and 8b the 5 ppm downfield shift of the ω-methylene carbon could be a consequence of the NH 2 platination. In the UV spectra of the complexes 8a,b and 10a,b the maxima of absorption appeared red-shifted (∆λ = 10-13 nm) in respect to those of the free ligands, in accordance with literature data [45]. Comparison of the IR spectra of the ligands and complexes did not provide evidence about the bonding mode of the ligands in the complexes; in fact in the 4000-2700 cm −1 region very little information could be obtained about the nature of metal-ligand interaction, because of the presence of broad and intense bands belonging to hydrogen-bonded hydroxyls of the ribose moieties. The 1800-400 cm −1 region, where vibration frequencies of purine and ribose skeletons fall, did not furnish significant differences between platinated and not-platinated nucleosides. Furthermore, in the region 550-400 cm −1 , where the resonances of Pt-N could be expected [45,46], no distinctive bands were observed.

Cytotoxicity Studies
The synthesized complexes were subjected to preliminary cell viability and proliferation assays. In particular, the cytotoxic behaviour was studied in A549 adenocarcinomic human alveolar basal epithelial and MCF7 human breast adenocarcinomic cancer cell lines by MTT and BrdU assays, to investigate the potential effects of platinum complexes on cell viability and proliferation, respectively.   The compounds 8b, 10a and 10b (50, 100 and 200 µM) modified the viability and proliferation of MCF7 cells at higher concentrations, according to the published evidence that these cells have been associated with cisplatin resistance [47]. The compound 8a proved to reduce the proliferation of the MCF7 cell line with a greater ability than the other compounds. Strikingly, compound 8a was able to inhibit the cell proliferation slightly better than cisplatin, but only at 50 µM concentration (Figure 4).

MTT Viability Assay
The cells were plated in 96 culture wells (20 × 10 4 cells/well) and allowed to adhere overnight. Thereafter, the medium was replaced with fresh medium and the cells were incubated in the absence or presence of cisplatin, 8a, 8b, 10a and 10b (50, 100 and 200 µM). After 72 h, the cell viability was determined by using 3-(4,5-dimethylthiazol-2yl)-2,5-diphenyl-2H-tetrazoliumbromide (MTT) conversion assay [48]. Briefly, 10 µL of MTT (5 mg/mL) were added to the cells and incubated for an additional 3 h. After this time point, the cells were lysed and the dark blue crystals solubilized with 150 µL of a solution containing 50% (v:v) DMF, 20% (w:v) SDS with an adjusted pH of 4.5. The optical density (OD) of each well was measured with a microplate spectrophotometer (Multiskan MCCC/340, Titertek, Huntsville, AL, USA) equipped with a 620 nm filter. The experiment was performed twice in triplicate. The vitality inhibition, induced by each compound at the indicated concentrations, was expressed as a percentage versus the untreated cells (the control). The viable cells were also counted by the trypan blue exclusion assay and light microscopy.

BrdU Cell Proliferation Assay
The cells were plated onto 96-well plates (2 × 10 4 cells/well) overnight. Then, the medium was replaced with fresh medium and the cells were incubated in the absence or presence of cisplatin (50, 100 and 200 µM), 8a, 8b, 10a and 10b (50, 100 and 200 µM). After 72 h, 5-bromo-2'-deoxyuridine (BrdU; 10 µM) was added and the cells were cultured for a further 12 h. The mitogenic activity was determined according to the manufacturer's instructions (BrdU cell proliferation assay kit, Cell Signaling). The experiment was performed three times in triplicate. The proliferation inhibition, induced by each compound at the indicated concentrations, was expressed as a percentage versus the untreated cells (the control).

Statistical Analysis
The results are expressed as the means ± SEM of n experiments. The statistical significance was calculated by one-way analysis of variance (ANOVA) and Bonferroni-corrected p-value for multiple comparison testing. The level of statistically significant difference was defined as p < 0.05.

Conclusions
In summary, in this paper we have reported the synthesis of four novel platinum complexes embodying a N 6 -(6-aminohexyl)adenosine or a 1,6-di-(adenosin-N 6 -yl)-hexane as ligands of monofunctional cisplatin or monochloro(ethylene diamine)platinum(II) moieties, starting from commercially available 6-chloropurine riboside. The designed synthetic route allowed us to obtain the desired platinum complexes without blocking the ribose hydroxyl groups and through easy purification steps.
The effects of these platinum complexes on cell viability and proliferation have been studied in A549 and MCF7 cell lines. Our data demonstrate that these compounds are able to inhibit the survival and proliferation of sensitive cancer cells. Nevertheless, cisplatin remains the most active molecule under our experimental conditions. Further experiments are necessary to understand the mechanism by which 8a penetrates into the cells and inhibits cell proliferation in the MCF7 cell line.

Acknowledgments
This work was financially supported by "Progetto FARO 2011" (Finanziamento per l'Avvio di Ricerche Originali). We are grateful to Luisa Cuorvo for her technical assistance and to CSIAS (Centro di Servizio Interdipartimentale di Analisi Strutturale) for the NMR facilities.