N6-ISOPENTENYL ADENOSINE, A PROMISING ANTICANCER AGENT: SYNTHESIS OF NOVEL ANALOGUES, EVALUATION OF ANTIPROLIFERATIVE ACTIVITY AND INSIGHT INTO ACTION MECHANISM

Isopentenyl adenosine (N6-(∆2-Isopentenyl)adenosine, iPAdo) is an important member of the class of cytokinins, N6-substituted adenosines, endowed with hormonal properties in plants. iPAdo is the only known cytokinin existing in animals. Its bioactivity in mammalians has been investigated and the overall reports confirm promising in vitro antiproliferative activity, less pronounced or absent in vivo. The PhD project was initially aimed to investigate this aspect of iPAdo activity and the preparation of iPAdo-grafted gold nanoparticles (GNPs) was selected as a possible solution, due to the well-recognized, remarkable properties of GNPs, such as high stability and biocompatibility, as well as potentiality of enhanced delivery and intracellular uptake of a bound to them drug into the cancer cells. The approach consisted in the preparation of a 5'-ester of iPAdo with lipoic acid (LA), later reduced to a dithiol form (DHLA), that could bind to gold nanoparticles, thus forming iPAdo-DHLA-GNPs. The feasibility study for an efficient preparation of the ester linkage between the 5'-hydroxyl of iPAdo and LA has been performed on another cytokinin, the commercially available kinetin riboside (KR), structurally related to iPAdo. Initially, a selective formation of the 5'-O-ester linkage of KR has been approached both chemically and enzymatically, preparing a few LA esters. However, in absence of the desired 5'-O-selectivity a more traditional chemical approach was investigated. The protection of KR at the 2' and 3'-hydroxyls as acetonide and esterification of 5'-OH proceeded in good yields (88% from KR). However, hydrolysis of the 2', 3'-protecting group led to concurrent hydrolysis of the β-glicosidic bond and partial removal of LA moiety. Finally, reduction with sodium borohydride in water/methanol (required for the reduction of LA into DHLA prior to grafting to Au particles) partially hydrolyzed LA-ester linkage. The above-reported negative results were a preliminary rather convincing indication that also LA-esters of iPAdo could not resist to the sequence of chemical reactions necessary for the formation of GNPs. Thus the iPAdo delivery project has been put aside and the efforts were focused towards the synthesis of iPAdo analogues as a mean to overcome the in vivo catabolic pathway. Meanwhile, with the aim of deepening the knowledge of iPAdo bioactivity palette, additional studies on iPAdo antiproliferative activity and on its mechanisms of action were carried out. Detailed bioactivity studies showed iPAdo dose-dependent cytotoxic effects on MDA-MB-231 and MCF-7 human breast cancer cells and IC50 values of 6.2 and 12.2 μmol/L, respectively, were determined. For both cancer cell lines the assessment of the cell shape and cell morphology of iPAdo treated cells established the loss of adhesion, rounding, cell shrinkage and detachment from the substratum. The interaction of iPAdo with DNA and Bovine Serum Albumin (BSA, a model of the analogous human protein) has been investigated via UV spectroscopy. The value of binding constants for iPAdo–DNA and iPAdo-BSA complexes have been estimated to be KiPAdo–BSA=4.9 x 104 M-1 and KiPAdo–DNA=4.1 x 103 M-1 that are indicative of good iPAdo-protein interaction and suggest affinity of iPAdo for DNA complexation. The dose-dependent cell cycle arrest and apoptogenic effect of iPAdo on MCF-7 cancer cell line has been evaluated, as well. The cell cycle analysis of MCF-7 cells by mean of flow cytometry showed an increase in the amount of sub G1/G0 phase on iPAdo treatment, suggesting that the inhibition of cell growth could be related to a mechanism of apoptosis. However, when the apoptotic activity of iPAdo on MCF-7 and human leukaemia cells (HL-60 cell line) has been studied via caspase-3 or -7 activation assay, apoptosis could be demonstrated only for HL-60 cells. For the planned preparation of the iPAdo analogues with improved IC50 value and endowed with a higher then iPAdo in vitro activity, isopentenyl arysteromycin (iPAry), a carbocyclic iPAdo analogue, was selected as synthetic target, since the absence of the glycosidic bond constituted a reasonable promise of resistance to the action of catabolic enzymes. iPAry has been firstly synthesized transforming D-(-)-ribose into the carbocyclic key-intermediate (4R,5R)-(-)-4,5-O-Isopropylidene-2-cyclopentenone (8 steps, 7.3% overall yield). Seven additional steps of fine organic conversions were required to obtain iPAry (7 steps, 7.1% yield from (4R,5R)-(-)-4,5-O-Isopropylidene-2-cyclopentenone). In our hands, the total 0.52% yield of iPAry from D-(-)-ribose has been the result of 15 overall steps. In alternative, commercial sources of arysteromycin (Ary) were considered. The rare and expensive compound was available only from a very few dealers in mg amounts and from a few milligrams of Ary, iPAry has been efficiently prepared (63% yield) by a N1/N6-alkylation and subsequent Dimroth rearrangement. Dose-dependent experiments with iPAry established no notable effects on MCF-7 cells proliferation up to 3 days of treatment. Differently from iPAdo, iPAry was not found to induce apoptosis in HL-60 cells, even when a 10-fold augment of the 10 M effective iPAdo concentration was used. Thus, replacing ribose moiety of iPAdo with cyclopentyl ring in iPAry, the in vitro activity of iPAdo could not be ameliorated and the in vivo test was excluded. The lack of bioactivity of iPAry was surprising, considering the subtle structural difference between iPAry and iPAdo. A hypothesis to explain the modest antiproliferative activity of iPAry on MCF-7 could be related to a receptor-mediated effect. Among other possibilities, A2 or A3 adenosine receptors could be involved in iPAdo antiproliferative activity, as suggested by previous pharmacogenomic studies on iPAdo transcriptional profile. Molecular modeling studies were carried out in order to identify the amino acid residues crucial to the iPAdo / iPAry interaction with h2B and hA3 human adenosinic receptors. The docking studies on iPAdo and iPAry with above receptors suggested that, although with different mechanisms, the carbocyclic moiety of iPAry causes detrimental effects on the complex set of favorable interactions that characterize the ribose group of iPAdo within the two receptors. However, the comparison of the two putative complexes in discussion enlightened that both in hA2b and hA3, iPAry conserves the hydrophobic contacts generated by isopentenyl chain and some of the key interactions stabilized by adenine. For this reason, we concentrated our attention on the synthesis of some other N6-substituted adenosine analogues: N6-isopentenyl tubercidin, 2',3'-O isopropylidene N6-iPAdo, N6-isopentenyl-L-adenosine and N6-isopentenyl derivative of 5´-N-ethylcarboxamido adenosine. These compounds were prepared in acceptable yields (45-74%) via the treatment of the corresponding adenosine analogue with γ,γ-dimenthylallyl bromide and subsequent alkaline rearrangement, namely Dimroth rearrangement. The synthesis of N6-propargyladenosine, N6-Methyl-N6-propargyladenosine and N6-allyloxyadenosine was achieved in good yields by using the interaction of 6-chloro-9 ribofuranosylpurine with the corresponding amine in the presence of triethylamine. All the afore-mentioned synthetic iPAdo analogues, along with commercial 1,N6-ethenoadenosine were screened for their in vitro antiproliferative activity on MCF-7 breast cancer cells by AlamarBlue assay. Only N6-propargyladenosine and N6-Methyl-N6-propargyl adenosine caused inhibition of cell proliferation close to that of iPAdo. Investigations on iPAdo and its analogues uptake in the cell, mechanisms of action and metabolic pathways involved are in progress.