Biologically active [ Pd 2 L 4 ] 4 + quadruply-stranded helicates : stability and cytotoxicity

Please note that technical editing may introduce minor changes to the text and/or graphics, which may alter content. The journal’s standard Terms & Conditions and the Ethical guidelines still apply. In no event shall the Royal Society of Chemistry be held responsible for any errors or omissions in this Accepted Manuscript or any consequences arising from the use of any information it contains. Accepted Manuscript Dalton Transactions


Introduction
Spurred on by the success of small molecular inorganic drugs 1 such as cisplatin 2 (cis-[Pt(NH 3 ) 2 Cl 2 ], CDDP) 3 there has been an upsurge of interest in the biological properties 4 of metallosupramolecular architectures. 5 This is driven by the hope that the different size and shape of these metallosupramolecular systems, compared to small molecule drugs, will lead to cytotoxic agents with novel biological mechanisms of action. The biological properties of a range of self-assembled architectures including metallo-macrocycles, 6 cages/prisms 7 and helicates have been examined. It is the metallo-helicates, 8 however, that have received particular attention. Building on early work from Lehn and co-workers, 9 which had shown that cationic doublehelical copper(I) complexes interact strongly with DNA, Hannon and co-workers carried out a series of pioneering studies on the biological properties of the triply-stranded [Fe 2 L 3 ] 4+ (L = N,N′-(methanediyldibenzene-4,1-diyl)bis[1-( pyridin-2-yl)methanimine]) helicate. 10 This tetracationic cylinder interacts strongly with duplex DNA, binding in the major groove, 11 but interestingly it can display alternative binding modes including non-covalent complexation at the center of three-way (Y-shaped) DNA 12 and RNA 13 junctions. The binding of [Fe 2 L 3 ] 4+ to DNA has also been shown to induce coiling of the nucleic acid structure. 14 Additionally, Hannon and coworkers found that these tetracationic iron(II) cylinders and related ruthenium(II) systems display both anti-cancer 15 and anti-bacterial 16 properties. The biological properties of other related triply-stranded helicate systems have also been examined. These molecules have been shown to bind G-quadruplex DNA, 17 inhibit Aβ amyloid aggregation 18 and also display cytotoxic (anti-cancer 19 and anti-bacterial 19b ) properties. While it is clear from these studies that cationic helicate architectures can strongly interact with DNA, a recent report from Scott and co-workers 20 has suggested that DNA binding is not always responsible for the observed biological activity.
Palladium(II) containing systems 21 represent the most common class of metallosupramolecular architectures. However, despite considerable interest in small palladium(II) complexes, 22 the biological properties of palladium-containing metallosupramolecular architectures have hardly been examined. 6i, 23 Herein, as part of our interests in metallosupramolecular architectures 24 and the biological applications of metal complexes, 25 we screened four quadruply-stranded, dipalladium(II) architectures for their antiproliferative effects against three different cancer cell lines, and compared this activity against non-malignant cells. It is shown that the most stable dipalladium(II) helicate displays low micromolar IC 50 values against all the tested cell lines. Furthermore, mechanistic studies indicate that the helicate does not cause cell death by interacting with DNA, rather it seems to disrupt the cell membrane. To the best of our knowledge this is the first reported cytotoxic study of quadruply-stranded helicate architectures.
It appears that both electronic and steric factors are responsible for the observed differences in the stability of the [Pd 2 (L) 4 ](BF 4 ) 4 architectures. The electron donor strength of the pyridyl and triazolyl ligands was determined using the palladium(II) probe complexes reported by Huynh and co-workers (ESI †). 28 The 13 C carbene carbon chemical shift of these probe complexes suggested that the triazolyl ligands are all more electron-rich and therefore "better" ligands than the pyridyl system (Table 1 and ESI †). The observed higher ligand donor strength of the triazole ligands relative to the pyridyl system undoubtedly contributes to the greater stability of the triazole based helicates relative to the pyridyl cage. Inspection of the  The difference in the stabilities of the three triazole-based helicates is more subtle. Each system has the same helicate structure and the 13 C carbene chemical shift of the probe complexes suggested that the donor strength of all the triazolyl ligands was very similar (Table 1 and ESI †). Despite these structural and electronic similarities, the stability testing showed that the [Pd 2 (hextrz) 4 ](BF 4 ) 4 complex was considerably more stable than the [Pd 2 (bntrz) 4 ](BF 4 ) 4 and [Pd 2 ( pegtrz) 4 ]-(BF 4 ) 4 helicates against the amino acid nucleophiles histidine and cysteine. We postulate that this enhanced stability is connected to the presence of the hydrophobic hexyl substituents in the [Pd 2 (hextrz) 4 ](BF 4 ) 4 complex. Presumably these hydrophobic chains aggregate together in the polar d 6 -DMSO/D 2 O solvent mixture used for stability studies and reduce the access of the nucleophiles to the palladium(II) ions of the helicate, thereby increasing the kinetic stability of this system relative to the other helicates.

Helicate cytotoxicity
To assess biological activity the cytotoxicities (as half-maximal inhibitory concentrations (IC 50 )) of the individual ligands tripy, bntrz, hextrz and pegtrz, along with their respective [Pd(L) 4 ](BF 4 ) 4 architectures, were determined against four different cell lines: A549 (lung cancer), CDDP resistant MDA-MB-231 (breast cancer), 29 DU-145 ( prostate cancer), and MCF10A (immortalized non-malignant breast tissue) (  (Table 2) and also the highest stability (Table 1), it was possible that an intact [Pd 2 (L) 4 ](BF 4 ) 4 helicate structure was required for high cytotoxicity. A comparison of the ligand hextrz and [Pd 2 (hextrz) 4 ](BF 4 ) 4 cytotoxicity data supported this hypothesis. As the palladium(II) metallocenters are coordinated by four ligands, if ligand dissociation was required for cytotoxicity then the helicate IC 50 would be four-fold less than the IC 50 of the individual ligands (Fig. 2). While [Pd 2 (hextrz) 4 ](BF 4 ) 4 was consistently more toxic than hextrz, a four-fold ratio was only observed against A549 cells (Fig. 2). Additionally, the ratio of ligand to helicate toxicity varied substantially between cell lines, from approximately 2 for the MCF10A cells to 15 for the MDA-MB-231's ( Fig. 2), demonstrating that there was no correlation between ligand and helicate cytotoxicity.

Time course of helicate actiona comparison with CDDP
To further identify the mechanistic basis of the high cytotoxicity displayed by [Pd 2 (hextrz) 4 ](BF 4 ) 4 , a time course analysis was performed to determine the time required for the onset of cytotoxicity (Fig. 3). No further difference in cytotoxicity was observed between 2 and 48 hours, indicating that by 2 hours the administration of the helicate had resulted in the rapid onset of cell death; however, after 2 hours the helicate induced no further toxicity to the surviving cells. This cytotoxic mode of cell death is distinct from classical metal complexes such as CDDP, which display increasing cytotoxicity over time. 30  toxicity is a result of DNA binding, causing errors in the cell replication process which results in the initiation of programmed cell death via apoptosis. As CDDP toxicity is dependent upon the replication status of individual cells, it selectively kills cells as they undergo cell division, resulting in a staggered cell death over several days. 31 In contrast, the rapid onset of cell death with [Pd 2 (hextrz) 4 ](BF 4 ) 4 provides evidence that the helicate is killing cells by a mechanism unrelated to DNA damage.

CDDP
[Pd 2 (hextrz) 4 ](BF 4 ) 4 treatment causes a rapid loss of cell membrane integrity We next aimed to identify the potential biological targets of the [Pd 2 (hextrz) 4 ](BF 4 ) 4 helicate. As rapid cell death is often associated with a compromised membrane integrity, we evaluated whether helicate induced cell death was attributable to membrane damage. For these studies we measured the release of the intracellular enzyme lactate dehydrogenase (LDH) into the extracellular medium. Extracellular levels of LDH are usually present at low concentrations, attributable to a small population of cells undergoing spontaneous death in culture systems. 32 However, these extracellular LDH levels rapidly increase if the cell membrane is disrupted, due to the release of intracellular enzyme into the extracellular space. The assay therefore provides a convenient method for rapidly assessing membrane integrity. 32 Consistent with a mechanism involving membrane damage, a time and concentration dependent increase in extracellular LDH levels was observed immediately following [Pd 2 (hextrz) 4 ]-(BF 4 ) 4 administration (Fig. 4). At a relatively high concentration Fig. 2 [Pd 2 (hextrz) 4 ](BF 4 ) 4 cytotoxicity is not due to helicate decomposition. Cells were treated with [Pd 2 (hextrz) 4 ](BF 4 ) 4 or hextrz for 24 hours before cell viability analysis. (A) Against DU-145 cells the IC 50 for the ligand (28.5 µM) was 8 fold higher than for the helicate (3.4 µM); if helicate toxicity was due to rapid decomposition to release the ligand, then the stoichiometry of the complex would predict a 4-fold ratio in IC 50 . (B) A comparison of the IC 50 ratio between ligand and helicate across the four cell lines demonstrated no correlation between helicate and ligand toxicity. Data points are expressed as means ± SEM where n = 6.  of 50 µM (7 fold above the IC 50 ), a rapid elevation in LDH activity was observed, which peaked at 10 minutes. However, at lower concentrations of 5 and 10 µM, continuous increases in LDH release were observable at 10, 30 and 60 minutes, indicating that around the IC 50 concentration membrane damage was a continual process that occurred over a sustained time period (Fig. 4). To investigate the extent to which LDH release was a diagnostic indicator of the viability of the entire cell population, we next conducted double labelling experiments with the dyes Hoechst 33342 and propidium iodide. Hoechst 33342 is a membrane permeable dye used for cell counting as it effectively labels all cells. In contrast, propidium iodide is a cell membrane impermeable dye, which is only taken up by nonviable cells with compromised cell membrane integrity. In combination with Hoechst 33342 it can therefore be used to distinguish between dead and viable cells within a heterogenous population. 33 In A549 cells this double label analysis confirmed that, by 10 minutes, a 10 µM concentration of [Pd 2 (hextrz) 4 ](BF 4 ) 4 (slightly above the 24 hour IC 50 of 7 µM) was sufficient to have killed >85% of the cell population, while a 50 µM concentration induced 100% cell death (Fig. 5). The phase contrast image demonstrated that [Pd 2 (hextrz) 4 ](BF 4 ) 4 induced cell death was accompanied by a rounding of the cells, a characteristic feature of cell damage that can accompany necrosis 34 (Fig. 5). In contrast 5 µM [Pd 2 (hextrz) 4 ](BF 4 ) 4 (slightly below the 24 hour IC 50 ) only induced 1.5% cell death at 10 min, although rounding of the cell population was observable in the phase contrast image, indicating that cell death processes had potentially been initiated in a greater number of cells. Therefore a small switch in helicate concentration, from slightly below to slightly above the IC 50 value, could result in the sudden onset of cell death. The mechanism underlying this remarkably rapid switch in activity is currently unknown. In comparison, administration of the hextrz ligand alone, even at 100 µM (4-fold higher than the 24 hour IC 50 ), resulted in no cell death at 10 min, confirming that the helicate and its cognate ligand induced cytotoxicity via distinct molecular mechanisms (Fig. 5).

Conclusions
Herein the biological activities of four quadruply-stranded dipalladium architectures were studied. The palladium(II)based helicates exhibited a range of cytotoxic properties, with the most cytotoxic complex [Pd 2 (hextrz) 4 ](BF 4 ) 4 possessing low micromolar IC 50 values against all of the cell lines tested, while the other helicates displayed moderate or no cytotoxicity. Stability studies indicated that the cytotoxic effect of the dipalladium architectures correlated well with the stability of these systems in the presence of biological nucleophiles, suggesting that it is intact helicate that is responsible for the biological activity.
Disappointingly, the [Pd 2 (hextrz) 4 ](BF 4 ) 4 helicate displayed no selectivity towards cancerous phenotypes; the compound exhibited approximately similar IC 50 values against non-malignant MCF 10A cells as towards the tumour-derived cell lines. Against the MDA-MB-231 cell line, which is resistant to platinum-based drugs, [Pd 2 (hextrz) 4 ](BF 4 ) 4 was 7-fold more active than cisplatin and, when compared to the IC 50 values against the other cell lines, there was no evidence of resistance. This suggests that the helicate does not induce cell death in the same way as clinically used metal complexes such as cisplatin. Preliminary mechanistic investigations with the [Pd 2 (hextrz) 4 ]-(BF 4 ) 4 helicate revealed that it induced cell death within minutes, accompanied by a loss of cellular membrane integrity. While these observations are consistent with the helicate acting as a metallo-detergent, 35 there are several potential alternative mechanisms of membrane damage, including pore formation 36 and changes to the fluidity of the lipid bilayer. 37 With the current data the exact cause of the membrane damage is not clear. Because the cytotoxic mechanism of the helicate was loss of cellular membrane integrity, the compound exhibited no selectivity towards cancer cells compared to non-malignant cells. This lack of specificity towards cancer cells prevents the immediate application of this complex as an anti-cancer agent.
However, the current work suggests that suitably designed stable palladium(II)-containing metallosupramolecular architectures (especially quadruply-stranded helicates) can display biological activity. As such we are now examining other palladium(II)-containing architectures and quadruply-stranded helicates containing other metals ions in search of more selective anti-cancer agents.