Contents

Metals based drugs Metals as Anti-cancer drugs Platinum complexes as anti-cancer drugs: Cisplatin Alternatives of cisplatin Ruthenium chemistry Ruthenium properties as anti-cancer agent

Ligand exchange Oxidation Iron mimicking

General mechanism of action of ruthenium based drugs Classical DNA interaction with ruthenium complexes Alternative interactions of ruthenium complexes Groove binding intercalation

Ruthenium complexes having anti-cancer activity Classical ruthenium complexes

Classical chemotherapy Classical complxes

Non-classical ruthenium complexes NAMI-A

Mechanism of interaction with DNA, serum proteins, & genes

KP1019 Mechanism of interaction

RAPTA Mechanism of action

Future trends toward ruthenium based anti-cancer drugs References

Precious metals have been used for medicinial purposes for at least 3500 years, when records show that gold was included in a variety of medicines in Arabia and China. At that time precious metals were believed to benefit health - because of their rarity - but research has now linked the medicinal properties of inorganic drugs to specific biological properties. The elucidation of a drug mechanism is however complex and the exact route of activity for many drugs remains unknown. The biological targets or mechanism of action of many metal drugs are now being resolved step by step, and this information is then used to design improved drugs with increased potency and reduced side effects.

A possible alternative to platinum therapy: ruthenium chemistry

In the search for drugs with improved clinical effectiveness, reduced toxicity and a broader spectrum of activity, other metals than platinum have been considered, such as rhodium and ruthenium. Non-platinum active compounds are likely to have different mechanisms of action, biodistribution and toxicities than platinum-based drugs and might therefore be active against human malignancies that have either an intrinsic or an acquired resistance to them. Ruthenium complexes are very promising, especially from the viewpoint of overcoming cisplatin resistance with a low general toxicity. Ruthenium has found its way into the clinic, where its properties are exploited for very miscellaneous uses. The radio physical properties of 97Ru can be applied to radio diagnostic imaging. Other ruthenium compounds have potential as immunosuppressant’s (cis-[Ru(III)(NH3)4(HIm)2]3+), antimicrobials (e.g. organic drugs coordinated to ruthenium centers, such as [Ru(II)Cl2(chloroquine)2] against malaria and others for the treatment of Chaga´s disease), antibiotics (ruthenium complexes of organic antibiotic compounds, e.g. the Ru(III) derivative of thiosemicarbazone against Salmonella typhi and Enterobacteria faecalis), nitrosyl delivery/scavenger tools (e.g. the Ru(III) polyaminocarboxylates known as AMD6245 and AMD1226 to treat stroke, septic shock, arthritis, epilepsy and diabetes), sodilator/vasoconstrictor agents and, as above mentioned, as drugs for cancer chemotherapy.

Ruthenium Properties Suited to Biological Applications:

There are three main properties that make ruthenium compounds well suited to medicinal application:

(i) rate of ligand exchange(ii) the range of accessible oxidation states and

(iii)The ability of ruthenium is to mimic iron in binding with certain biological molecules.

(i) Ligand Exchange

Many ruthenium complexes have been evaluated for clinical applications, particularly in the treatment of cancer, due in part, to Ru(II) and Ru(I1I) complexes having similar ligand exchange kinetics to those of PtQl) complexes. Lpnd exchange is an important determinant of biological activity, as very few metal drugs reach the biological target without being modified. Most undergo interactions with macromolecules, such as proteins, or small S-donor compounds and/or water. Some interactions are essential for inducing the desired therapeutic properties of the complexes. As the rate of hgand exchange is dependent on the concentration of the exchangmg hgands in the suirounding solution, diseases that alter these concentrations in cells or in the surroundmg tissues can have an effect on the activity of the drug.

The uxidation state changes of ruthenium in cancer and healthy cells. The reductive environment of cancer cells favours Ru(II). which ismore biologically active than Ru(III). Hence Ru( III) compounds are essentially prodrugs that become activated by reduction on reaching the cancer cell.

(ii) Oxidation State

Ruthenium is unique amongst the platinum group in that the oxidation states Ru(Il), Ru(I1I) and R u m are all accessible under physiological conditions. In these oxidation states the ruthenium centre is predominantly hexacoordinate with essentially octahedral geometry, and R u g complexes tend to be more

biologically inert than related Rum and (IV) complexes. The redox potential of a complex can be modified by varying the ligands. In biological systems glutathione, ascorbate and single electron transfer proteins are able to reduce Ru(IIl) and R u m , while molecular oxygen and cytochrome oxidase readily oxidise Rum. The redox potential of ruthenium compounds can be exploited to improve the effectiveness of drugs in the clinic. For example, the drug can be administered as relatively inert Ru(I1I) complexes, which are activated by reduction in diseased tissues.In many cases the altered metabolism associated with cancer and microbial infection results in a lower oxygen concentration in these tissues, compared to healthy ones, and this promotes a reductive environment. Cancer cells are known to have higher levels of glutathione and a lower pH than healthy tissues, creating a strongly reducing environment. If the active Rug9 complex leaves the low oxygen environment, it may be converted back to Ru(I1I) by a variety of biological oxidants.Proteins that can catalyse the reduction of Ru(I1I) to Ru(II) include mitochondrial and microsomal single electron transfer proteins. The mitochondrial proteins are of particular interest in drug design as apoptosis, the desired mechanism for cell death, can be initiated in the mitochondria, as well as by other pathways, for instance, by the Fas/FasL pathway. Transmembrane electrontransport systems can also reduce Ru(I1I) complexes outside of the cell and this is hghly relevant to the mechanism of action of a ruthenium based drug in clinical use which has anticancer activity independent of cell entry (vide inpa).

(iii) Iron Mimicking

The low toxicity of ruthenium drugs is also believed to be due to the ability of ruthenium to mimic iron in binding to many biomolecules, including serum transferrin and albumin.

Schematic representation of the selective uptake of transferrin by cancer cells. Ruthenium can mimic iron in binding to transferrin. Metal-loaded rransferrin is delivered to cells

according to the number of transferrin receptors on their surfaces. As most cancer cells have a higher number of transferrin receptors (left), compared to healthy cells (right), ruthenium is target

These two proteins are used by mammals to solubilise and transport iron, thereby reducing its toxicity. Since rapidly dividmg cells, for example microbially infected cells or cancer cells, have a greater requirement for iron, they increase the number of transferrin receptors located on their cell surfaces, thereby sequestering more of the circulating metaloaded transferrin. The exact increase in radio-labelled ruthenium compounds in cancer cells, compared to healthy cells, has been shown to range from 2-12 folds, depending on the cell type. As the drug is targeted to cancer cells, its toxicity is reduced because less of it will reach healthy cells.

Alternative ways of interaction between metallodrugs and DNAAnticancer therapy with classical ruthenium coordination compounds is

based on the capability of the metal to coordinatively bind to DNA. These ruthenium complexes are usually It functional and they mostly exert their action by forming intra- or interstrand Crosslinks with the DNA molecule. The cytotoxicity of these monofunctional complexes could also be related to coordination to DNA. Other ways of interaction with DNA are known, including backbone binding3 and recognition of DNA junction structures.

Groove bindingThe dinuclear complex [{Ru (apy) (tpy)}2{μ-H2N(CH2)6NH2}]+4 (1g),

interacts with DNA presumably via electrostatic and especially via groove-binding interactions. The activity displayed by this compound in a number of cell lines is comparable to cisplatin.Two strategies can be followed that are inspired by the above-described results.

(i).Electrostatically binding of homonunclear complexes:The first one consists on the synthesis of homodinuclear ruthenium(II)

complexes that are first electrostatically attracted to DNA, subsequently form a coordinative interaction with the latter, and finally interact with the DNA in the same way 1g does, i.e., by groove binding

Scheme depicting a homodinuclear, positively-charged Ru(II) complex being first

electrostatically attracted to DNA (left), coordinated to a nucleic base (middle) and finally binding to a DNA groove (right).

Heterodinuclear complexes:A second strategy deals with the synthesis of heterodinuclear Pt-Ru complexes usingsuch ligands. The Pt moiety can be chosen such that it will form a coordinative interaction with DNA, like transplatin, or it could even be an intercalator, vide infra, such as [Pt(tpy)]+2. Following the first approach, the homodinuclear ruthenium (II) compound [{Ru (tpy)Cl}2(μ-paa)](BF4)2(1h) was obtained, where tpy is 2,2´-6´2”-terpyridine and paa is 2-pyridinealdazine and some cell tests were subsequently performed.

Molecular structures of [{Ru(tpy)Cl}2(μ-paa)](BF4)2(1h, left) and[Ru(abpt) (bpy)4](PF6)2 (1i, right). Proton nu6mbering scheme as used in 1HNMR spectra

Intercalation:

Small, planar aromatic molecules can bind DNA through intercalation, as proposed already by Lerman in 1961. The base pairs and helical backbone extend and unwind to accommodate the molecule, which inserts into the resulting hydrophobic pocket. The intercalating surface is stabilized electronically in the helix by π-π stacking with the bases, thus the intercalator is rigidly held and oriented with the planar moiety perpendicular to the helical axis.

MetallointercalatorA decade later, the concept “metallointercalator” was introduced. The

platinum(II) complexes [Pt(tpy)(SCH2H2OH)]+ and [Pt(tpy)Cl]+, where tpy is 2,2´-6´,2”-terpyridine, were proven to bind strongly to DNA by intercalation between base pairs.7

Figure: Intercalated DNA with DNA

Although in principle the square-planar geometry of platinum (II) was thought to be essential for a metallointercalator, octahedral metal centres with large planar aromatic ligands were synthesized afterwards, which also displayed intercalative interactions with the DNA helix. While one of the planar units inserts between base-pair planes, the metal and additional co-ligands interact in one of the DNA grooves.

Figure: Insertion of platinum metal in DNA

Ruthenium intercalation:To date, many [Ru (bpy)2L]+2 and [Ru(phen)2L]+2 complexes have been

described, where L is an aromatic bidentate ligand, which have been proven to interact with DNA via intercalation. Even a dinuclear analogue with a large aromatic bridging ligand has been reported to very slowly bind to DNA via an intercalation process.

Figure: Above left - schematic representation of an unpaired or base bulge, where the duplex section of the DNA helix is interrupted by the inclusion on one strand of one or more bases that have no base(s) on the complementary strand with which to form a base-pair.

The groove binder distinguish from an intercalator is not straightforward, as illustrated by many discussions on the controversial case of [Ru(phen)3]+4, where phen is phenantroline.

It may be very interesting to synthesize ruthenium(II) polypyridyl ligands containing the ligands 4-amino-3,5-bis(2-pyridinyl)-1,2,4 triazole (abpt) and 3,5-bis(2-pyridinyl)-1,2,4 triazole (Hbpt), for several reasons. Firstly, some ruthenium complexes with π-deficient ligands behave as photo-oxidants, giving rise to photo-induced electron-transfer processes that lead to DNA cleavage. Moreover, the strong σ–donor properties of the triazole/triazolate groups make these ligands optimal for use as bridges in the synthesis of dinuclear and polynuclear complexes.An especially interesting feature of this kind of complexes is the luminescence displayed by some of them.32 Finally, the abpt and Hbpt ligands may behave as intercalators.The ruthenium(II) complex [Ru(abpt)(bpy)2](PF6)2 was synthesized and its anticancer activity was tested against some selected cell lines. Although this complex displayed an activity comparable to that of cisplatin in the cell line H226 and a reasonable activity in the cell line WiDR, it was found to be virtually inactive in the rest of the tested cell lines. The interaction of this compound with DNA remains to be studied.

In vitro cytotoxicity assays

The anticancer activity of [Ru(abpt)(bpy)2] (PF6)2was tested in vitro in several selected cell lines, following the experimental procedure described in chapter 4 of this thesis. The synthesis of groove-binder homodinuclear ruthenium (II) and heterodinuclear Pt-Ru complexes has been introduced. As former, the ruthenium (II) compound [{Ru(tpy)Cl}2(μ-paa)](BF4)2 (1h) was obtained. This dinuclear compound has two leaving groups, one per ruthenium atom, therefore a coordinative interaction with DNA is also possible, and even the formation of intra- and interstrand adducts might be expected. According to the results obtained in preliminary cell tests, complex 1h is moderately active in the L1210/2 cell line, although it displays virtually no activity in the human ovarian cancer cell lines A2780 and A2780R, in which the homodinuclear complex 1g was shown to be active.The ruthenium (II) complex [Ru(abpt)(bpy)2](PF6)2 was selected as the parent compound of a family of ruthenium(II) polypyridyl complexes to be tested for anticancer activity. Substitution of the bpy groups by other chelating polypyridyl ligands, such as 2,2´:6´,2”-terpyridine or phenantroline, or the more π-deficient 2,2´-bipyrazine, 1,4,5,8- tetraazaphenanthrene or ,4,5,8,9,12-hexaazatriphenylene, would yield a group of various relted ruthenium(II) complexes. The cytotoxicity of all these compounds should be tested, as well as their ways of interaction with DNA and their DNA cleavage ability. Some structure-activity relationships could be extracted from the differences in their properties and anticancer activities.Work in these compounds has not gone yet any further than the synthesis and testing of the chosen parent compound against some selected cancer

cell lines. The activity displayed by [Ru (abpt) (bpy)2] (PF6)2 was disappointing in most of the cell lines.

Current Uses of Ruthenium-Based Drugs

The array of clinical applications for some platinum group metals, have versatility of metallodrugs in the clinic. The activity of each compound is a function of the oxidation state of the metal and the nature of the attached lgands. These features dictate not only how the drug interacts with the disease target but also the biological transformations that occur en route. By manipulating these features activity can be fmetuned to maximise the potency but minimise the general toxicity of the drugs.

Anticancer Activity

Despite the success of platinun-based anticancer compounds in the dinic, there is still a need for new and improved metal-based anticancer drugs. The need for new drugs is fuelled by the inability of platinum compounds to tackle sometypes of cancer of hlgh social incidence and by the associated toxic side effects of the current platinum compounds in clinical use. Platinum anticancer drugs bind DNA, causing damage that prevents protein synthesis and replication causing cell death. The success of platinum anticancer drugs has biased the screening of new metal-based anticancer compounds towards looking for damage caused to DNA. Many Ru(Il), Ru(I1I) and R u m complexes with amine, dimethylsulfoxide, imine, polyaminopolycarboxy late, and N-heterocyclic ligands have been found to bind to DNA. However, many of these compounds are barely soluble in aqueous solution, which is necessary to allow efficient administration and transport. Solubility has been increased by using dialkyl sulfoxide derivatives, such as, NAMI-A, which is now recognised as the most successful ruthenium-based anticancer compound. Interestingly, although NAMI-A can bind DNA, in vivo DNA damage does not appear to be part of its anticancer mechanism. In general, anticancer activity hinges on the ability of a drug to bring about apoptosis (programmed cell death) of the tumour cells. Apoptosis is a complicated process by which the cell 'commits suicide' in a controlled manner such that there is no cell debris or damage done to surrounding cells. This process is perhaps best illustrated on a cellular level using neuroblastoma cell lines: growing cells are semi-differentiated, forming irregular shapes that can be clearly distinguished from the spherical apoptotic bodies. Alternatively cells can die by a process called necrosis, which is less controlled, and causes inflammation and damage to adjacent cells.

(a) DNA Damaging Agents

A number of ruthenium compounds have been shown to bind DNA in vjh and there is a direct correlation between this activity and the cytotoxicity of Ru (III) ammine complexes in tissue culture. The mechanism of DNA bin- has been probed and certain ruthenium complexes form cross-links between DNA strands – possibly favoured due to the steric restrictions imposed by the octahedral geometry of the complexes. This bin- mechanism differs from the intrastrand cross-links favoured by cisplatin, and consequently the cancer cell lines that have developed resistance to cisplatin by accelerating the rate of repair of intrastrand cross-links are still susceptible to ruthenium anticancer drugs. Interestingly, it has been demonstrated that RuO complexes are far more reactive towards DNA than Ru and it is therefore possible that the anticancer activity of Ru (III) involves initial reduction to Rum at the tumour site, promoted by the altered physicochemical environment in tumor cells. If this hypothesis is correct then R u m complexes are essentially prodrugs. However, there is growing evidence to show that protein interactions are also extremely important in the anticancer activity of ruthenium compounds and these interactions could occur with the ruthenium in either oxidation state.

(b) Radiosensitisers

Radiation therapy is routinely used against some types of cancer. This treatment can be enhanced by using nitroimidazoles and halogenated pyrimidine radiosensitisers (compounds that increase the irradiation sensitivity of the target cells). As the activity of these compounds depends on their proximity to DNA, coordinating radiosensitisers to metals that are able to bind to DNA, for example platinum and ruthenium, enhances the radiosensitising properties.The two key features of an effective radiosensitiset are the ability to bind DNA and the redox potential of the bound complex. Strong DNA bin- affinity is a feature of many ruthenium compounds, although not all have dosemitising activity. This activity depends upon the compound having a high reduction potential, which can be optimised by the use of appropriate ligands. For example, the nitroimidazole complex, RuCl2(DMSO)2(-4N021m)2 is one of the most effective radiosensitisers having both a hgher activity and a lower toxicity than 4-NO2Imadozole alone.

(c) Photodynamic Therapy

As with radiotherapy, photodynamic therapy uses chemicals targeted to diseased cells. The chemicals become cytotoxic when exposed to electromagnetic radiation. For example, nitrosylruthenium complexes release NO on reduction; the reduction may be triggered using photodynamic methods. Until recently, the application of photodynamic therapy was restricted by the poor accessibility to cancer cells, but a

unique approach using ruthenium complexes has been developed which overcomes these restrictions. This new method of therapy centres on the Mossbauer absorption of y-rays by ruthenium; this can induce emission of Auger electrons which damage the DNA to which the Ru is bound.

(d) Antimitochondrial

Apoptosis can be initiated by more than one pathway, one being via the mitochondria (the sub cellular compartments associated with energy and heat generation). Consequently any compounds that target these structures are of great interest as anticancer drugs. Ruthenium red is routinely used by biologists to stain mitochondria selectively, as it binds to the calcium channels on their surfaces and has long been known to inhibit tumour cell growth, but its toxicity is too great for use in the clinic. It is possible that the anticancer mechanism of some other ruthenium compounds may involve mitochondtial interactions. However, many ruthenium compounds with putative antitumour activity are unlikely to act like ruthenium red.

(e) Antimetastases

A particularly challenging area of cancer therapy is the treatment of metastases. Metastasis occurs at a late stage of the disease and involves the escape of cells fiom the primary tumour and their reestablishment at distinct secondary locations.The metastases are normally dormant and are often suppressed by hormones secreted from the primary tumour. However, if the primary tumour is removed or there are genetic changes in the metastases, growth can begin. Tumour growth beyond about 1 mm3 requires a blood supply, and the formation of the necessary blood vessels is termed angiogenesis. Some Antimetastases drugs restrict angiogenesis, for example angiostatin, throbospondin and badmastat, but statistics show that once this process has occurred the chance of five-year survival drops by about 50 per cent, dependmg upon the type of cancer. The first ruthenium anticancer drug to progress through clinical trials, [hzm-RuCl+(DMSO)Im]-m, NAMI-A, , is strongly active against tumour metastases. NAMI-A appears to alter protein expression, either by binding to proteins or to RNA, causing thickening of the protein layer surroundmg turnouts and metastases. As a result the tumour becomes isolated, preventing escape of metastasing cells and reducing the blood flow, which ultimately suffocates it. Only a very small portion of the drug reaches the tumour target and its activity appears to be independent of its concentration in tumour cells. Rather, it appearsblastoma that had metastasised to the bone marrow.. In tissue culture, cells grow in the S-form, which is partially differentiated and characterised by the irregular shapes of the cells. These cells undergo

two responses to stress: full differentiation or cell death. Full differentiation is observed by the more‘dendritic’ appearance of the cells, which occurs 1-3 hours after the stress. Subsequently cells either dedifferentiate back to the S phase - if the stress can be overcome - or die. that NAMI-A has an extracellular mode of activity centring on interactions with proteins. In our own laboratory we have investigated the role of the llgands attached to ruthenium in the passive diffusion of the drug across cell membranes, facilitating the movement of the drug into and within cells. A series of drugs of formula[Ru(arene)Cl~PTA]R, APTA, been developed that interact with proteins and DNA in vjh, triggering apoptosis in human cancer cell lines. Thepcymene derivative, RAPTA-C, has activity against, for example, SK-N-SH neuroblastoma cells, inducing apoptosis in nanomolar concentrations. This cell line was derived from the metastases of human neurons.

How these drugs work: mechanisms of action

In the past two decades a new approach to treating cancer, known as targeted therapy, has started to emerge. While classical chemotherapy involves drugs interfering with replication and mitotic rocesses of tumour cells, their “target” being thus DNA, a more recent strategy involves targeting cellular signalling pathways of cancer cells, yielding highly effective cancer treatments with less severe side effects. The recent discovery of receptors and growth factors, such as epidermal growth factor receptor (EGFR), vascular endothelial growth factor (VEGF), or cyclin-dependent kinases (CDK) that are upregulated in cancer cells provides new possible targets for cancer therapy.166 The high specificity of targeted therapies accounts for a more manageable toxicity profile of the drugs. Its main drawback is that most targeted therapeutic drugs are only effective in specific types of cancer (e.g. Imatinib mesylate for chronic myelogenous leukaemia, Erlotinib for advanced non-small cell lung cancer, etc), which limits their applicability.165 In recent years, ruthenium-based drug research is moving from classical chemotherapy into the nonconventional approaches.

Classical ruthenium anticancer therapy

Based on the capability of ruthenium to coordinatively bind to DNA via some of the nitrogen atoms of the nucleic bases, in particular via the nitrogen N7 of guanine. This is also the action expected from the first ruthenium complexes designed as anticancer drugs, the ammine-chloro derivatives. The novelty is that these complexes are thought to act as ruthenium (III) prodrugs, which would be inactive until the ruthenium gets reduced in the cytosol.

Classical ruthenium complexes as anti-cancer drugs:

Poly pyridyl complexes

Binding to DNA via an additional mode was achieved when an intercalating polypyridyl ligand was added to the ruthenium system. Additional properties that make polypyridyl groups esirable ruthenium ligands are their photoluminescence, which makes them suitable as DNA probes, as well as the stability of the complexes that they can originate, amongst others. An often encountered problem is the poor water solubility of many of these complexes.

Octahedral ruthenium complexes, capable of photodynamic singlet oxygen production at near 100% efficiency, were shown to cause light-dependent covalent crosslinking of p53 and PCNA subunits in mammalian cells and cell lysates. Azide, a singlet oxygen quencher, greatly reduced the p53 photocrosslinking, consistent with the idea that singlet oxygen is the reactive oxygen species involved in p53 photocrosslinking. A photodynamically inactive ruthenium complex, [Ru(tpy)2]2+ (tpy = [2,2;6,2]-terpyridine), had no effect on p53 or PCNA photocrosslinking. Photodynamic damage to p53 has particular relevance since p53 status is an important determinant of phototoxicity and the effectiveness of photodynamic cancer therapy. The two photodynamic complexes studied, [Ru(tpy)(pydppn)]2+, where pydppn = (3-(pyrid-2-yl)-4,5,9,16-tetraaza-dibenzo[a,c]naphthacene, and [Ru(pydppn)2]2+, differed in their efficiency of p53 and PCNA photocrosslinking in cells, but showed similar efficiency of photocrosslinking in cell lysates, suggesting that they differ in their ability to enter cells. Photocrosslinking of PCNA by [Ru(tpy)(pydppn)]2+ increased linearly with concentration, time of uptake, or light exposure. Both [Ru(tpy)(pydppn)]2+ and [Ru(pydppn)2]2+ caused photodynamic protein-DNA crosslinking in cells, but [Ru(tpy)(pydppn)]2+ was more efficient. The efficiency of photodynamic protein-DNA crosslinking by [Ru(tpy)(pydppn)]2+ in cells increased with increasing levels of photodynamic damage. Photodynamic damage by [Ru(tpy)(pydppn)]2+ caused inhibition of DNA replication in a classical biphasic response, suggesting that DNA damage signaling and cell

cycle checkpoint pathways were still operative after significant damage to nuclear proteins.

Ruthenium (II) dimethylsulfoxide complexes

As water-soluble versions of the above-mentioned ammine-chlorido derivatives, the good antimetastatic activity of trans-[Ru (II) Cl2 (dmso)4 ] soon became apparent, as well as its capability toovercome cisplatin resistance in certain cell lines. These two observations suggested a mechanism of action different to the by then widely accepted mechanism of cisplatin.

The use of polyaminocarboxylate ligands in metallopharmaceutical applications seemed a logical option due to their resemblance to biological molecules. Several of these complexes turned out to be antitumour active with low systemic toxicity. Some of them were proven to bind to DNA, alter its conformation and even induce DNA cleavage. In addition, several of these complexes were found to be effective NO scavengers and protease inhibitors, thus they could be used to treat various diseases or serve as antiviral agents.

RUTHENIUM POLYAMINOCARBOXYLATE COMPLEXES:

Ruthenium (Ru) complexes containing polyaminocarboxylate (pac) ligands (Ru-pac) have features which indicate they may be suitable for biological

applications. For instance, Rupac complexes can bind to biomolecules through a rapid and facile aquo-substitution reaction, and Ru-pac has a range of accessible oxidation states. Ru-pac also has some notable catalytic properties that mimic enzymatic hydrocarbon oxidation by cytochrome P-450 in homogeneous conditions. This is of immense significance towards developing Ru-pac based agents for oxidative cleavage of DNA and artificial nuclease in DNA foot-printing experiments. This review aims to highlight the scope of Ru-pac complexes as metallopharmaceuticals, and outlines their potential for certain biological applications.

Prospects for Ru-pac Complexes as Antitumour Agents

Although in cell culture studies a correlation has been observed between the cytotoxicity of some ruthenium complexes and their DNA binding ability (18), the mechanism of the drug action of these ruthenium complexes is still largely unknown. Octahedral Ru (III) and Ru (II) complexes containing ligands, such as ammines, N-heterocycles and dimethylsulfoxides, exhibited various degrees of biological activity, including antitumor action in vivo. Considering that the above Ru(III) complexes are more inert than the

corresponding Ru(II) analogues, an ‘activation by reduction’ mechanism was proposed to explain the antitumour activity of such complexes Ru-pac complexes, due to their lability towards aquo-substitution, bind DNA constituents in afacile and straightforward manner and thus have oncological significance. Antitumour activity has been reported for labile Ru(IV)-cdta (cdta = 1,2-trans-diaminocyclohexan-N,N,N',N'- tetraacetate) (23, 24), while cis-[RuIII(pdta)Cl2] (Pdta = ropylenediaminetetraacetate) is known to damage nuclear DNA and inhibit DNA recognition by enzyme restriction. A cross linking with a guanine base unit of DNA has been proposed as an explanation for the observed activity. However, the generation of a superoxide radical in NADPH oxidase, triggered by the presence of the Ru-pdta complex, may be another reason for the observed cytotoxity .Mixed-ligand complexes of Ru(II)-pac with a series of DNA bases have been reported by Shepherd’s group (15c) and the binding sites of the DNA constituents have been discussed with regard to their significance in chemotherapy. They reported a 2-coordination mode for Ru-pac (pac = hedtra, ttha; ttha = riethylenetetraaminehexaacetate) at the C5=C6 olefinic double bonds of uridine- and cytidine-related bases, along with coordination at the normal binding sites (N3 and N1). Although the ‘pac’ environment favours -donation by the ruthenium centre, no experimental evidence for 2-attachment was observed in the case of thymidine. This assumes the pyrimidine structure is important for 2-coordination. The affinity of Ru(II)-pac complexes to the 2-pyrimidine site was shown to be linked to a balance between electronic and steric factors, and thus Rupac could be significant as a DNA crosslinking agent .

Ruthenium arene complexes:

DNA also seems to be a target for the organometallic arene-ruthenium complexes. The coordination of the ruthenium atom to the nucleic bases was seen to be enhanced through H-bonding interactions or weakened because of steric interactions, suggesting the possibility to design compounds to target specific nucleotides.

The binding of the complex to DNA appeared to be promoted by hydrophobic arene-purine base π-π stacking interactions when large ring systems were used.Finally, the photoreactive ruthenium compounds can also be considered within the classical therapy, as well as most of the dinuclear ruthenium compounds aboved described, as their target is still DNA.

Photoreactive

Organometallic ruthenium(II)−arene (RA) compounds combine a rich structural diversity with the potential to overcome existing chemotherapeutic limitations. In particular, the two classes of compounds [Ru(II)(η6-arene)X(en)] and [Ru(II)(η6-arene)(X)2(pta)] (RA-en and RA-pta, respectively; X = leaving group, en = ethylenediamine, pta = 1,3,5-triaza-7-phosphaadamantane) have become the focus of recent anticancer research. In vitro and in vivo studies have shown that they exhibit promising new activity profiles, for which their interactions with DNA are suspected to be a crucial factor. In the present study, we investigate the binding processes of monofunctional RA-en and bifunctional RA-pta to double-stranded DNA and characterize the resulting structural perturbations by means of ab initio and classical molecular dynamics simulations. We find that both RA complexes

bind easily through their ruthenium center to the N7 atom of guanine bases. The high flexibility of DNA allows for fast accommodation of the ruthenium complexes into the major groove. Once bound to the host, however, the two complexes induce different DNA structural distortions. Strain induced in the DNA backbone from RA-en complexation is released by a local break of a Watson−Crick base-pair, consistent with the experimentally observed local denaturation. The bulkier RA-pta, on the other hand, bends the DNA helix toward its major groove, resembling the characteristic DNA distortion induced by the classic anticancer drug cisplatin. The atomistic details of the interactions of RA complexes with DNA gained in the present study shed light on some of the anticancer properties of these compounds and should assist future rational compound design.

In conclusion, ruthenium drugs are particularly important in the clinic due to their low toxicity. These complexes appear in some cases to function in a different way to classical chemotherapies. For this reason the conventional tests used to screen new compounds for anticancer activity should be treated with caution, and new assays for potential drug candidates are needed. Methods are required to rapidly locate drug interactions with key protein targets. Finally, even when metal drugs are not found directly active, they may interact with the proteins that regulate apoptosis, thereby modifying cell behaviour.

Non-classical ruthenium complexes:

Non classical ruthenium complexes comprise of recently synthesized drugs, in which different ligands are replaced by some other ligands. This class of ruthenium includes these complexes:

Albumin

Serum albumin is the most abundant plasma protein. It plays a key role in a number of physiological functions, such as the control of osmotic blood pressure; transport, metabolism and distribution of various compounds; radical deactivation, and delivery of amino-acids after hydrolysis for the synthesis of other proteins.

Transferrin

The transferrins are a class of iron-binding and transporting proteins, widely distributed in the extracellular fluids of vertebrates. Most of the transferrins consist of a single polypeptide chain with a molecular weight of around 80 kDa, constituted by two remarkably similar amino acid sequences, each accounting for half of the molecule and each carrying an iron-binding site.Binding of iron is dependent on concomitant binding of carbonate, hydrogencarbonate or some other synergistic anion, which serves as a bridging ligand between protein and metal. The role of the bridging anion may be to prevent water from binding in the coordination sphere of the metal, locking it tightly to the protein and avoiding hydrolysis. Iron binding is strong enough to resist hydrolysis in the extracellular fluids, but still allows iron to be released within specific intracellular compartments. The metal binding site with its associated anion-binding site is a characteristic of all transferrins.The iron-binding cleft in the C-lobe is closed, both in the presence and in the absence of the metal. However, the cleft in the N-lobe is wide open in apotransferrin, exposing three basic side chains, which are buried within it in the iron-loaded transferrin. These side chains are Arg 121, Arg 120 and Lys 301; they may serve to attract the carbonate anion as the first step in binding.

Transferrin receptors are present in all dividing cells, in a number varying from several tens of thousands to almost a million. This number increases when a cell is in need of iron. Transferrin receptors are continuously traveling between the surface and the interior of the cell.At the slightly alkaline extracellular pH of 7.4, transferrin can bind 1 or 2 ferric ions,and 2 iron-bearing transferrin molecules can bind the dimeric transferrin receptor. Iron-free transferrin is not recognized by the receptor at this pH.

Transferrin is thought to release its iron within the cell in an endosomal compartment which has a pH of 5.5. Then the apotransferrin-transferrin receptor complex travels back to the membrane, and the apotransferrin is released again in the extracellular medium.

Cytochrome c

Cytochrome c is a mitochondrial peripheral membrane protein. Its function in the respiratory chain in the inner mitochondrial membrane consists on electron transfer from cytochrome c reductase to cytochrome c oxidase. In 1996 it has been reported that, when released into the cytosol, cytochrome c activates a programmed cell death cascade (apoptosis).

Other proteins

Haemoglobin is a globular tetrameric protein consisting of four subunits (two α- and two β-polypeptide chains) bound through non-covalent interaction. Each protein subunit carries a haeme group including a Fe(II) as the central atom.Haemoglobin is in charge of O2 and CO2 transport in the blood.Ubiquitin is a small cytoplasmic protein which has two potential binding sites for cisplatin. It was chosen as a model protein to study the formation of protein-cisplatin adducts.Another familiy of essential metal-transporting serum proteins are the γ-globulins.

Interactions between metallodrugs and serum proteins

Protein interactions with platinum drugs, amongst which cisplatin and carboplatin, have been studied thoroughly, using various techniques. The influence of these interactions in the distribution and pharmacokinetics of the drugs has been recognised.

Albumin

Cisplatin binds preferentially to haemoglobin, followed by albumin. The efficient binding to the latter can be explained by the high affinity of platinum to sulfur. Hence, the most likely binding point of cisplatin to albumin is the cysteine-34 residue. Cisplatin irreversible binding leads to cleavage of albumin disulfide bonds, inducing changes in the structure of the protein, thus affecting its activity. Other platinum compounds, such as oxaliplatin , display the same behavior with albumin; the interaction between albumin and transplatin is reported to be not very significant.

Transferrin

In an analogous way, cisplatin binds to sulfur-containing residues of transferrin, although the exact interaction position is a subject of debateThis interaction was proven to be determinant of properties such as cytotocixity, in vivo distribution of the drug and tumour-specificity.

Cytochrome c

The results obtained with various techniques indicate that the binding of the ruthenium(III) complex KP1019 to cytochrome c induces conformational changes in the protein. A loss of tertiary structure is experienced, together

with changes in the haeme group and an increase in the α–helical content of Apo cytochrome c. These conformational changes are expected to have an influence in the biological activity of cytochrome c, and subsequently, in its ability to induce cell apoptosis.

Other proteins:

The binding of different platinum complexes to the serum proteins haemoglobin, ubiquitin and γ-globulins has been widely studied and a review of these interactions is available.37 On the other hand, the studies involving ruthenium(III) complexes have been mainly focused on the interactions between these drugs and transferrin or cytochrome c. Interactions between Ru (II) polypyridyl complexes and serum transport proteins

Some ruthenium (III) complexes are hypothesised to act as inactive prodrugs, which may get activated by reduction to ruthenium (II) once they entered the cells, vide supra. Serum transport proteins, such as transferrin, might be involved in this cellular uptake process. Hence the interest in studying the interactions between these proteins and the anticancer active ruthenium(III) complexes. However, while a number of ruthenium(II)

complexes are known that display a considerable activity in cell tests, one of important aspect is the interaction between these complexes and transferrin. Therefore, a preliminary experiment was carried out to explore whether or not such interactions could occur.Two 5 μM solutions of [Ru(apy)(tpy)(H2O)](ClO4)2·2H2O in phosphate buffered saline (PBS) were prepared. Human serum transferrin (Invitrogen) was added to one of them to give a 1 μM concentration. Both solutions were incubated for 3 hours at 37 °C. Both samples were ultrafiltered (Millipore centricon 10,000 MWCO) and the filter was washed four times with PBS. The unbound ruthenium complex should have been recovered after going through the filter in both cases. The portion that did not go through the filter should contain no ruthenium in the control experiment, and the transferrin-bound ruthenium, in the sample containing the protein. The four portions were analyzed for ruthenium by inductively coupled plasma (ICP).70% of the initial ruthenium was recovered in the portion of the control experiment that went through the membrane filter. The detected ruthenium in the portion that did not pass through the filter was negligible. From the sample that contained transferrin, the portion that went through the filter contained 34% of the initial ruthenium (unbound ruthenium), while the portion that did not go through the filter contained 35% of the initial ruthenium. This implies that after just 3 hours in PBS at 37 °C, at least 35% of the initial ruthenium was bound to transferrin.The results obtained clearly encourage further studies of the interactions between transferrin and other ruthenium(II) polypyridyl complexes,

Some confusions are still unresolved as whether this interaction has an influence in the cytotoxicity and tumour-selectivity of the compounds, or to what extent the results obtained in the performed cell tests are valid, without the involvement of serum transferrin in them.

Ruthenium complexes and metastasis

The existence of ruthenium drugs which, despite showing no significant activity against the primary tumour (and no in vitro cytotoxicity), do yield an important activityagainst metastases, illustrates the importance of testing ruthenium complexes not only against cancerous cell lines, but also for antimetastatic activity.Well-known in vitro methods for antimetastatic ability determination are migrationand invasion assays. However, since apoptotic cells do not migrate and not all cancerous cells are invasive, cytotoxic compounds are not susceptible to these studies, nor is every type of cell lines. The ability of a drug to diminish migration of a malignant cell from the initial tumour to another tissue can be measured in experiments involving Boyden chambers. On the other hand,

the invasion of basement membranes by tumour cells, a property which is characteristic of metastatic cells, can be studied by using Matrigel, a reconstituted membrane.In conclusion, a new testing routine is necessary for potential anticancer/antimetastatic ruthenium complexes. Not only should the interactions of these compounds with proteins be studied, which could lead to both selective apoptosis and a decrease in resistance to the drug, but also the antimetastatic ability of these drugs should

NAMI-A

(ImH)[trans-RuCl4(DMSO-S)(Im)], (Im=imidazole, DMSO-S=S-bonded dimethylsulfoxide, NAMI-A, is the first anticancer ruthenium compound that successfully completed Phase I clinical trials. NAMI-A shows a remarkable activity against lung metastases of solid tumors, whilst is not effective in the reduction of primary cancer.

Interaction with serum proteins

One of the most successful ruthenium-based anticancer drugs to date, NAMI-A, displays a unique behaviour. Its lack of cytotoxicity in vitro, together with its in vivo ability to reduce metastases weight while the primary tumour remains unaffected, appear to exclude DNA as the primary target. NAMI-A binds strongly to serum proteins, including the iron transporter transferrin, and it induces cell arrest in the premitotic G(2)-M phase.

Anti-metastatic activity

(a) Antimetastases

A particularly challenging area of cancer therapy is the treatment of metastases. Metastasis occurs at a late stage of the disease and involves the escape of cells fiom the primary tumour and their reestablishment at distinct secondary locations.The metastases are normally dormant and are often suppressed by hormones secreted from the primary tumour. However, if the primary tumour is removed or there are genetic changes in the metastases, growth can begin. Tumour growth beyond about 1 mm3 requires a blood supply, and the formation of the necessary blood vessels is termed angiogenesis. Some Antimetastases drugs restrict angiogenesis, for example angiostatin, throbospondin and badmastat, but statistics show that once this process has occurred the chance of five-year survival drops by about 50 per cent, dependmg upon the type of cancer. The first ruthenium anticancer drug to progress through clinical trials, [hzm-RuCl+(DMSO)Im]-m, NAMI-A, , is strongly active against tumour metastases. NAMI-A appears to alter protein expression, either by binding to proteins or to RNA, causing thickening of the protein layer surroundmg turnouts and metastases. As a result the tumour becomes isolated, preventing escape of metastasing cells and reducing the blood flow, which ultimately suffocates it. Only a very small portion of the drug reaches the

tumour target and its activity appears to be independent of its concentration in tumour cells. Rather, it appearsblastoma that had metastasised to the bone marrow.. In tissue culture, cells grow in the S-form, which is partially differentiated and characterised by the irregular shapes of the cells. These cells undergo two responses to stress: full differentiation or cell death. Full differentiation is observed by the more‘dendritic’ appearance of the cells, which occurs 1-3 hours after the stress. Subsequently cells either dedifferentiate back to the S phase - if the stress can be overcome - or die. that NAMI-A has an extracellular mode of activity centring on interactions with proteins. In our own laboratory we have investigated the role of the llgands attached to ruthenium in the passive diffusion of the drug across cell membranes, facilitating the movement of the drug into and within cells.

Studies carried out with NAMI-A analogues suggest that the imidazole fragment is not essential for the antimetastatic activity. On the other hand, the reinforcement of the axis dmso-Ru-N-donor ligand by using N-containing heterocycles that are less basic than imidazole reduce the loss of dmso from the complex, increasing at the same time the antitumour action.

KP1019

The only ruthenium drug other than NAMI-A currently undergoing clinical trials, KP1019 is significantly cytotoxic in vitro against colorectal cell lines SW480 and HT29 by inducing apoptosis.The drug was also found to be highly effective in in vivo tests in which cisplatin had been inactive. The mechanism of action of the “Keppler-type” complexes is thought to be due to at least

two factors, namely the“activation-by-reduction” process and the transferrin-mediated transport into the cells. Certain anticancer ruthenium(III) complexes, such as indazolium trans-[tetrachloridobis(indazole)ruthenate(III)], KP1019, were also proven to bind to both albumin and transferrin. Particularly the interaction of KP1019 with the latter suggested the theory that this ruthenium(III) complex could act as a virtually non-toxic prodrug that enters the cell when it is bound to transferrin. This prodrug would then be activated by intracellular reduction to a ruthenium(II) complex, which would be the actual cytotoxic drug. This mechanism would also account for a selective entrance of the drug in the tumorous cells, which express an increased number of transferrin receptors in their membranes, due to their higher iron requirements.A study of the ability of ruthenium(III) cytotoxic compounds to bind to transferrins was carried out in 1996. The presence of a large water-filled cavity in the interdomain cleft of each transferrin lobe, in which the metal- and anion-binding site is found, apparently allows some flexibility in the species that can be bound, while domain closure is still possible. Cell-culture experiments have given evidence that the antitumour capacity of some ruthenium(III) complexes is enhanced by binding to transferrin,and so the role of serum transferrin in the accumulation of ruthenium(III) complexes in tumours is suspected to be important. The ruthenium complex binds via a coordinative interaction with a histidine residue in the N-lobe of transferrin. The heterocyclic ligands remain bound to ruthenium, and this is presumably essential for antitumour activity following the release of the complex.

KP1019 is capable of forming crosslinks with DNA that are different to those originated by cisplatin. DNA is not completely excluded as a target for

KP1019. However, it induces apoptosis in colorectal cell lines mainly via the intrinsic mitochondria pathway. An increase in the number of indazole ligands of these complexes improved significantly the in vitro cytotoxicity in several cell lines, allegedly because the cellular uptake is facilitated and the reduction potential is increased.

MODES OF ACTION OF KP1019:

The ruthenium compound acts like a Trojan horse and is accumulated in the tumour via transferrin and partly also via albumin under the reductive and hypoxic conditions in tumours the compound is activated and treatment of tumours is possible with very low side effects. A clinical phase I study has been finished. All patients came in with progressive disease, five of six patients changed to stable disease.

The gallium compound KP 46 interacts with the iron metabolism. An inhibition of ribunocleotid reductase is involved in the mode of action, the compound is applied orally as tablets and in a phase I study three of four renal cell carcinoma patients showed response

Although DNA appears to be a target for the organometallic arene-ruthenium complexes (vide supra), the RAPTA complexes constitute a particular case (see Fig.1.16).

General formula of two groups of organometallic ruthenium(II) complexes with modified arene ligands. On the left, [(η6-arene)Ru(II)(en)X]+, where the arene can be benzene, p-cymene, biphenyl, 5,8,9,10-tetrahydroanthracene or 9,10-dihydroanthracene. X is Cl or I. On the right, [(η6-arene)Ru(II)(pta)XY] (RAPTA complexes). R1, R2 are alkyl groups; X and Y can be Cl or different μ-dicarboxylate ligands.

RAPTA

Parting from the observation that the complex RAPTA-T displayed a similar in vivo activity to NAMI-A, albeit with lower systemic toxicity, a group of derivatives from this parent compound was synthesised, which were then tested in vitro for interactions with different biological molecules and in vivo for antitumour and antimetastatic activity. Several of these complexes showed a reduction in lung metastases in mice, while leaving the primary tumour mostly unaffected. Moreover some specific protein-binding interactions were detected.

A proteomic-based analytical approach based on 2D PAGE and laserablation inductively-coupled mass spectrometry (ICP-MS) appears to be a promising tool to identify the specific proteins interacting with ruthenium-based drugs.

The reaction of metallothionein-2 (MT-2) with the organometallic antitumour compound [Ru(6-p-cymene)Cl2(pta)], RAPTA-C, was investigated using ESI MS and ICP AES. The studies were performed in comparison to cisplatin and significant differences in the binding of the two complexes were observed. RAPTA-C forms monoadducts with MT-2, at variance with cisplatin, that has been observed to form up to four adducts. These data, combined with ICP AES analysis, show that binding of both RAPTA-C and cisplatin to MT-2 requires the displacement of an equivalent amount of zinc, suggesting that Cys residues are the target binding sites for the two metallodrugs. The competitive binding of RAPTA-C and cisplatin towards a mixture of ubiquitin (Ub) and MT-2 was also studied, showing that MT-2 can abstract RAPTA-C from Ub more efficiently than it can abstract cisplatin

References:

1. Prof. Dr. J. Reedijk Prof. Dr. E. Alessio, Prof. Dr. J. Brouwer,Dr. J.G. Haasnoot ,Dr. A.C.G. Hotze , 2007, ‘’Ruthenium polypyridyl complexes with anticancer properties’’. Thesis: 1-146

2. Claire S. Allardyce and Paul J. Dyson, 2001, ‘’Ruthenium in medicine; current clinical use and future perspective,’’j. platinum metals , rev.. 45(2):61-69

3. Yaw Kai Yan,{ Michael Melchart, Abraha Habtemariam and Peter J. Sadler,2005,’’ Organometallic chemistry, biology and medicine: ruthenium arene anticancer complexes’’.j.The Royal Society of Chemistry 2005: 4761-4776

4. Wee Han Ang [a]Paul J. Dyson*[a] .2009, ‘’Classical and non-classical ruthenium based anti-cancer drugs : Towards Targeted Chemotherapy’’ ,J. Inorg. Chem. 2006: 4003-4018

5. Debabrata Chatterjee* and Anannya Mitra, 2006, ‘’Ruthenium polyaminocarboxylate complexes’’.j.platinum metals rev,.2006,50,(1): 1-12

6. Gianni sAVA',5, Ilaria cAPOZZ14, Alberta BERCAM04, Renato GACLIARD14, Moreno COCCHIETT04,

7. Laura MASlER03, Maurizio ONIST03, Enzo ALESSIO',G iovanni MESTRONi and Spiridione GARBlSA3,1996, Down-Regulation Of Tumour Gelatinase/Inhibitor Balance And Preservation Of Tumour Endothelium

8. By An Anti-Metastatic Ruthenium Complex, int.j.cancer, 68:60_66

9. Angela Casini, Andrei Karotki, Chiara Gabbiani, Francesco Rugi, Milan Va ák, Luigi Messori and Paul J. Dyson, Reactivity of an antimetastatic organometallic ruthenium

compound with metallothionein-2: relevance to the mechanism of action ,j. Metallomics, 2009, 1, 434