Further evidence for redox activation of the plasmid – dirhenium(III) complexes interactions

The DNA-interactions in vitro are still necessary investigations for determination of the possible anticancer properties of the compounds, candidates for application in cancer therapy. The aim of the present work was to realize if the interaction of cis-dicarboxylates of dirhenium(III), with pivalato(I), isobutirato(II) and adamanthyl(III) ligands cleaves the plasmid in the same manner and what is the influence of the ligands on this process. For experiments we used the prokaryotic plasmid which is good model to analyze DNA-cleaving ability of different substances that exists in supercoiled conformation and turns to nicked and linear forms. It was shown that gradual conversion of the supercoiled Form I to a mixture of supercoiled (Form I) and nicked (Form II) DNA takes place and increasing amounts of Form II are produced with higher concentrations of I–III under increasing of concentration that showed the DNAcleaving abilities of all investigated dirhenium complexes. This process was taking place with different intensity in the range I ˃ II ˃ III, that demonstrates the influence of the organic radical on the cleaving activity of the dirhenium(III) complexes. Under hydrogen peroxide conditions, I and II showed close results, demonstrating more intensive process of cleaving, including formation of the linear plasmid (Form III) under higher concentration, witnessing about redox-activation of the DNA-cleaving reaction. Cleaving activity of III was approximately the same in all experiments, that was demonstrated only by decreasing of the supercoiled form I and increasing of the nicked form II of the plasmid and by absolutely absence of the linear form III of the plasmid. The electrophoresis mobility shift assays showed that rhenium cluster compounds have nuclease activity and confirmed that natural DNA may be their target in the living cells. The conclusion was made that the mechanism of DNA-cleavage reaction of the dirhenium(III) complexes is multiple in which the electron donating (withdrawing) effects of the ligands and catalytic activity of the metal core should be taken in consideration.


Introduction
Earlier using the method of UV-titration we have shown that the rhenium(III) quadruple bonding compounds interacted with supercoiled Calf Thymus DNA (CT DNA) (Paramonova et al., 2016;Polokhina et al., 2016) and that the mechanism of the interaction was redox-activated. Also, bis-dimethylsulfoxidecis-tetrachlorodi-µ-pivalatodirhenium(III) cis-Re2((CH3)3CCOO)2Cl4•2DMSO (I) was shown to cleave plasmid DNA by electrophoretic mobility shift assay (Shtemenko et al., 2013) in different manner depending from the redox state of the medium. The DNA-interactions in vitro are still necessary investigations for determination of the possible anticancer properties of the compounds, candidates for application in cancer therapy (Ismail et al., 2019). It is known, that anticancer properties of the dirhenium(III) dicarboxylates in vivo depended on the length of the alkyl ligands (Leus et al., 2012).
Thus, the aim of the present work was to realize if the interaction of other two dicarboxylates of dirhenium(III), with shorter (isobutirato-) and longer, bulkier (adamantyl-) ligands cleave the plasmid in the same manner and what is the influence of the ligands on this process.
Electrophoretic Mobility Shift Assays. The mobility assays were performed according to (Keck, Lippard, 1992;Mitra et al., 1997) by using aliquots (10 µl) of native supercoiled plasmid pUC18 (20 µg/ml) and increasing concentrations of complex (0 to 80 µM). Electrophoresis experiments were carried out using 1 % agarose gel and buffer solution (40 mM Tris-acetate and 1 mM EDTA, pH 8.2). The applied voltage was 250 V and the gels were run for a period of 20 min. After electrophoresis, the gels were stained with 0.5 mg/L ethidium bromide and imaged under UV light. The efficiency of the DNA cleavage was measured by determining the ability of the complex to form open circular or nicked circular DNA from its supercoiled forms. The same assays were also carried out in the presence of H2O2 (experiment b) and 2-mercaptoethanol (experiment c) (25 µM). Gel image analysis was made with the help of free LabImage 1D -Software for 1D Gels and Western Blots (https://www.kapelanbio.com/products/).

Results and discussion
To study cleavage ability of metal-containing substances and their cleavage pathway, an assay is widely used, which employs a special type of DNA, a plasmid DNA (Wang et al., 2015;Kettenman et al., 2018;Kadoya et al., 2019). Single strand DNA cleavage (SSC) is the process, in which one scission is introduced in one of the two DNA strands in one cleavage event. Double strand DNA cleavage (DSC) denotes the process, in which two scissions are introduced to each of the strands of the DNA double helix in a single cleavage event. Although both pathways introduce double strand breaks ultimately, it is important to note that these are two inherently different processes. The pathway that is responsible for the induction of strand breaks is of more interest, as the direct double strand cleavage pathway is considered to be responsible for the cytotoxicity of the initiating substances to cancer cells.
Typically, the quaternary structure of plasmids is superhelical, which means that the double stranded DNA forms a compact folded structure. When one of the DNA strands is broken, a swivel, i.e. a free rotation site in the complementary strand opposite to the break, relieves the strain in the twisted molecule. This causes the supercoiled structure to relax into an uncoiled form, which is referred to as nicked DNA. When both strands of the DNA are cleaved opposite to each other, linear DNA is observed. Supercoiled DNA is often referred to as form I, nicked DNA as form II and linear DNA as form III. These forms are good resolved by electrophoresis (Sears et al., 2013;Kadoya et al., 2019). We used these good described phenomena to investigate cleaving activity of the investigated substances II and III ( Fig. 1, 2) and to compare the results with previously obtained data for the substance I.
The intensity of the bands obtained by scanning procedure is presented on the Fig. 3. As it is clear from obtained results, the gradual conversion of the supercoiled Form I to a mixture of supercoiled (Form I) and nicked (Form II) DNA takes place and increasing amounts of Form II are produced with higher concentrations of I-III under increasing of concentration that showed the DNAcleaving abilities of all investigated dirhenium complexes (Experiment a). This process was taking place with different intensity in the range I ˃ II ˃ III, that demonstrates the influence of the organic radical on the cleaving activity of the dirhenium(III) complexes. This range coincides with the binding constants, obtained from the UV-titration experiments (Paramonova et al., 2016;Polokhina et al., 2016).
Under hydrogen peroxide conditions (Experiment b) I and II showed close results, demonstrating more intensive process of cleaving, including formation of the linear plasmid (Form III) under higher concentration, witnessing about redox-activation of the DNA-cleaving reaction. Cleaving activity of III was approximately the same in Experiments a-c, that was demonstrated only by decreasing of the supercoiled form I and increasing of the nicked form II of the plasmid and by absolutely absence of the linear form III of the plasmid. In the presence of mercaptoethanol (Experiment c), the activation of the cleaving abilities of I-III are also evident but no Form III is observed as compared to the activity of I-III only.
The presence of the slowest moving bands in Fig. 1, 2 (slower than Form II) indicates formation of high molecular weight adducts that may be explained by the ability of unwound plasmids to polymerize (Vilfan et al., 2006) or to form unwound DNA-I intra-strand adducts similar to cisplatin and dirhodium compounds (Dunham et al., 2005).
The mechanism of the SSC and DSC of plasmids was in detail studied for bleomycin (BLM), an iron-containing natural nuclease (Decker et al., 2006). The experimental and computational studies presented here on the reaction coordinate of activated BLM strongly support the low-spin (BLM)Fe III -OOH complex as the active oxidizing species and a direct hydrogen-atom abstraction as the reaction mechanism for the initial attack on DNA. The mechanism for double-strand cleavage by a single BLM molecule would then involve the Fe IV intermediate generated by the first reaction as the active species for the initial second H-atom abstraction reaction damaging the second DNA strand. A hydroperoxo species of the dicopper(II) was also identified as the active species and spectroscopically identified in the reaction of the complex with plasmids (Kadoya et al., 2019). Analogical investigations for the dirhenium(III) complexes are absent, but acceleration of the nuclease reaction by hydrogen peroxide shown by us makes possible to think that the same hydroperoxo species may play definite role in the investigating processes with rhenium substances.
It is well-known that the microenvironment in the cancer cells is significantly different from that in normal cells (Zhou et al., 2014) and the concentration of reactive oxygen species (ROS) including H2O2 is relatively higher in cancer cells than in normal cells (Nogueira, Hay, 2013), that is mainly caused by mitochondrial malfunction (Aykin-Burns et al., 2009). Thus, redox activating of the process of cleavage by H2O2 shown by us in the Experiment b, is of great value as could work in cancer cells but not so extensively in normal cells and this phenomena could be used to reduce the side-effects of anticancer drugs. The influence of ligands surrounding the active center of metal-containing artificial nucleases was investigated on the copper -dipicolamine complexes (Wang et al., 2015). It was shown that replacement of a hydrogen atom of the dipicoyl ligand with a benzyl group enhanced the efficiency both of SSC and DSC cleavage; also introducing an electron donating group resulted in further increase in efficiency, whereas the presence of electron withdrawing group reduced the efficiency; the introducing of different ROS-inhibitors decreased the intensity of cleavage. In our experiments I (pivalic derivative) and II (isobutiric derivative) are homologues, I has two longer alkyl ligands, that we may note that they are more electron donating and thus agree with previously made conclusions about possible formation of the intermediate complex for which electron donation is a stabilization factor. As for III: recently we have shown that some dirhenium complexes with diadamanthyl ligands had catalase-like activity (Shamelashvili et al., 2016); thus, III could act as inhibitor of ROS-formation in the nuclease event. Nuclease activity of the dirhenium(III) clusters to our mind is of great interest and requires additional deep investigations.