Introduction

Science efforts in the design and discovery of more efficient medicines are a key commitment for a longer life expectancy of mankind. Among new drugs for treatment of diseases such as cancer and tuberculosis, those based on transition metals will probably play an important role. Within this context, chemistry offers expressive opportunities for medicinal area, and the appearance of metal-based drugs is continuously changing from chance discovery to rational drug design by exploiting changes in metal oxidation states as well in their ligands (Bruijnincx and Sadler, 2008; Farrell, 2002; Fricker, 2007; Jakupec et al., 2008; Leyva et al., 2007; Seng and Tiekink, 2012; van Rijt and Sadler, 2009).

In fact, the discovery of biological properties of cisplatin (cis-diamminedichloroplatinum(II); cis-DDP) was one of the most important achievement for cancer chemotherapy. This square planar Pt(II) complex was approved for clinical use in 1978 and has become first-line therapy for the treatment of testicular cancer (Fricker, 2007; Jamieson and Lippard, 1999; Seng and Tiekink, 2012). Cis-DDP is currently one of the three most widely used drugs in chemotherapy and is highly effective in treating ovarian and testicular cancers (Jakupec et al., 2008). Carboplatin (cis-diammine(1,1-cyclobutyldicarboxylato)platinum(II)) has a similar clinical activity to cis-DDP; however, there is cross-resistance with cisplatin, which is not surprisingly since both have similar mechanism of action (Farrell, 2002; Fricker, 2007). The cytotoxicity of these drugs is originated by platinum–DNA monoadducts and intra- and interstrand adducts, which are formed following uptake of the drug into the nucleus of cells (Jamieson and Lippard, 1999; van Rijt and Sadler, 2009).

The success of platinum drugs has inspired the search for other cancer metallodrugs (Arnesano and Natile, 2009). A wide range of metals has been studied such as Pd, Ir, Rh, Ru, Cu and Au (Fricker, 2007; Meng et al., 2009; Rocha et al., 2010; Seng and Tiekink, 2012). Nowadays, it is known that many metal complexes with potential for antitumor activity do not behave like cisplatin (Farrell, 2002; Fricker, 2007; Jakupec et al., 2008; Seng and Tiekink, 2012; van Rijt and Sadler, 2009). This is the case of some bipyridyl complexes of Pt(II) containing thiourea which have been reported as a new class of efficient DNA intercalators, whose mode of action, based on non-covalent interactions, differs from that of cis-DDP (Marverti et al., 2008). There are also examples of metal-based drugs that target biological processes other than DNA replication (Barnard et al., 2004; Bruijnincx and Sadler, 2009). For instance, gold(I) carbene complexes with potential antitumor activity target mitochondrial membranes causing mitochondrial membrane permeability, possibly via interaction with the mitochondrial permeability transition pore (Barnard et al., 2004).

Cyclopalladated compounds display cytotoxic properties toward several tumor cell lines, and some of them are also effective against cells that are resistant to cisplatin (Caires, 2007; Cutillas et al., 2013; Karami et al., 2012; Moro et al., 2009; Oliveira et al., 2009; Omae, 2014; Quiroga et al., 1998; Spencer et al., 2009). Although some Pd(II) cyclometallated complexes have been reported to interact with DNA (Zamora et al., 1997), these organometallic compounds can also interact with other cellular sites (e.g., mitochondrial membrane thiol-groups) (Serrano et al., 2011) and inhibit important enzymes implicated in a number of mammalian diseases (e.g., cathepsin B) (Bincoletto et al., 2005; Caires, 2007; Spencer et al., 2009). These findings are of particular relevance since the development of new anticancer agents with different modes of action than those observed for cisplatin and analogues is of key importance to overcome clinical therapy resistance.

For many years, our research has been focused on the synthesis of cyclopalladated compounds containing N-, P- or S-based molecules as co-ligands and the evaluation of their cytotoxic properties toward tumor cell lines and pathogenic microbes (Caires et al., 1999; Ferreira et al., 2012; Moro et al., 2009, 2012; Santana et al., 2011). We have recently described the synthesis of cyclometallated species of the type [Pd(C 2,N-dmba)(X)(tu)] (Hdmba = N,N-dimethylbenzylamine; X = Cl, Br; tu = thiourea) and their cytotoxicity activity against murine mammary adenocarcinoma (LM3) and lung adenocarcinoma cell lines (LP07) (Moro et al., 2009). Figure 1 summarizes the inhibitory concentrations against LM3 and LP07 cells found for [Pd(C 2,N-dmba)(X)(tu)] derivatives.

Fig. 1
figure 1

Cytotoxicity data (IC50) of the [Pd(C 2,N-dmba)(X)(tu)] (Hdmba = N,N-dimethylbenzylamine; X = Cl, Br; tu = thiourea) compounds against murine LM3 and LP07 tumor cell lines

Full size image

The [Pd(C 2,N-dmba)(Cl)(tu)] compound did not display a significant level of activity against the LM3 and LP07 cells, while substitution of the chloride by the more polarizable Br− anion increased the cytotoxic effects in both cell lines. Interestingly, the IC50 value toward LM3 cells obtained for [Pd(C 2,N-dmba)(Br)(tu)] was comparable to cisplatin. Based on these findings, we assumed that the strength of the Pd–X bond is important for cytotoxic activity of these cyclopalladated compounds. It is important to emphasize that the same trend in the cytotoxicity of cyclometallated compounds [Pd(C 2,N-dmba)(X)(tu)] (Moro et al., 2009) and [{Pd(C 2,N-dmba)(X)}2(μ-bpp)] (X = Cl, Br, NCO, N3; bpp = 1,3-bis(4-pyridyl)propane) (Moro et al., 2012) was also observed toward M. tuberculosis.

As a part of our ongoing studies on the coordination and biological chemistry of palladium(II) compounds (da Silva et al., 2014; Ferreira et al., 2012; Mauro et al., 1999; Moro et al., 2012; Netto et al., 2001, 2005; Santana et al., 2001, 2011), we have synthesized the new iodo-derivative [Pd(C 2,N-dmba)(I)(tu)] from the cleavage reaction involving the precursor [Pd(C 2,N-dmba)(μ-I)]2 and thiourea ligand aiming at investigating whether the increase of the anion softness would result in an enhancement of the cytotoxic activity against murine tumor cell lines (LM3 and LP07) and Mycobacterium tuberculosis. In addition, we have also evaluated the mutagenic potential of the organometallic [Pd(C 2,N-dmba)(μ-X)]2 and [Pd(C 2,N-dmba)(X)(tu)] compounds (X = Br, I) by the Ames test.

Experimental

Physical measurements

Elemental analyses of carbon, nitrogen and hydrogen were performed on an EA1110–CHNS-O microanalyzer from CE Instruments. Melting points (M.p.) were determined on a Microquímica apparatus. Infrared spectra were recorded on a Nicolet Impact 400 spectrophotometer in the spectral range 4000–400 cm−1 using the KBr pellets technique. 1H- and 13C{1H}-NMR spectra were obtained from DMSO-d 6 (Merck) solutions and are referred to the high-field Si(CH3)4 signal on a Bruker AC-200 spectrometer working at 200 MHz for 1H and at 50 MHz for 13C.

Materials

The reagents thiourea (Merck), N,N-dimethylbenzylamine (Aldrich), PdCl2 (Degussa) and KBr (Merck) were employed without further purification. Methanol, chloroform and n-pentane of analytical purity were purchased from Merck.

Synthesis of complexes

Complexes [Pd(C 2,N-dmba)(μ-Br)]2 (1), [Pd(C 2,N-dmba)(μ-I)]2 (2) and [Pd(C 2,N-dmba)(Br)(tu)] (3) were prepared according to published methods (Crociani et al., 1970; Maassarani et al., 1987; Moro et al., 2009).

Preparation of the [Pd(C 2,N-dmba)(I)(tu)] complex (4)

A 10 mL chloroform suspension of [Pd(C 2,N-dmba)(μ-I)]2 (0.10 g; 0.14 mmol) was mixed with a 5 mL methanol solution of thiourea (0.021 g; 0.28 mmol) at room temperature. The resulting solution was stirred for 1 h, the solvent was removed under reduced pressure, and the yellow solid obtained was washed with n-pentane and dried in vacuum. Yield: 75 %. M.p. >157.0 °C (dec.). Anal. Calcd. for C10H16IN3PdS (%): C, 27.1; H, 3.64; N, 9.47. Found: C, 26.5; H, 3.60; N, 9.51. 1H NMR: 7.77 br (NH2, 4H); 7.28 d (H3, 7.4 Hz, 1H); 6.94 dd (H4/H5, 7.0 Hz; 1.3 Hz; 2H); 6.90 − 6.77 m (H6, 1H); 3.97 s (N–CH2–, 2H); 2.73 s (–N(CH3)2, 6H). 13C{1H}-NMR: 148.76 (C2), 128.70 (C3), 124.70 (C4), 124.02 (C5), 121.90 (C6), 71.86 (N–CH2–), 51.49 (–N(CH3)2).

figure b

Cell culture and MTT assay

LM3 and LP07 cell lines were generously supplied by Prof. Elisa Bal De Kier Joffé from Cell Biology Department, Research Area, Instituto de Oncología “Angel H. Roffo,” Universidad de Buenos Aires, Buenos Aires, Argentina. LM3 and LP07 cells were maintained in MEM (Sigma) supplemented with 10 % heat-inactivated FBS, 2 mM l-glutamine and 80 µg mL−1 gentamicin, defined as complete medium, in plastic flasks (Corning) at 37 °C in a humidified 5 % CO2 atmosphere (Galli et al., 2003). Passages were made by trypsinization of confluent monolayers (0.25 % trypsin and 0.02 % EDTA in Ca2+–Mg2+-free PBS). Cell number was counted by the trypan blue dye exclusion method.

Test solutions of the compounds (1000 µM) were freshly prepared by dissolving the substance in 50 µL of DMSO completed with 4950 µL of culture medium. Afterwards, the tested compounds were diluted in culture medium to reach the final concentrations ranging from 140 to 10 µM. The DMSO solvent in the concentrations used in test did not reveal any cytotoxic activity.

The cytotoxicity evaluation of the tested compounds against LM3 and LP07 was determined in vitro by MTT assay (MTT = 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) (Mosmann, 1983). The IC50 was defined as the medium of three independent experiments through the equation of graphic line obtained (Microcal Origin 5.0™). Triplicates tests were performed for each concentration (Rocha et al., 2010).

Mutagenicity assay

A standard Ames test was carried out using each test agent at a concentration range of 2.5–20.0 µM (Maron and Ames, 1983). Salmonella typhimurium tester strain, TA98, was used to detect mutation by frameshift. The positive controls were 4-nitro-o-phenylenediamine (NPD) at a maximum concentration of 10.0 μg/plate for TA98 in the absence of S9 mix and 2-aminoanthracene (2AA) at a maximum concentration of 1.25 μg/plate for TA98 in the presence of S9 mix. DMSO was used as the negative control. Triplicate plates were run and results were shown as mean values.

The mutagenic index (MI) was also calculated for each concentration tested, this being the average number of revertants per plate with the test compound divided by the average number of revertants per plate with the negative (solvent) control. A test solution was considered mutagenic when a dose–response relationship was detected and a twofold increase in the number of mutants (MI ≥ 2) was observed for at least one concentration (Santos et al., 2006).

Antimycobacterial assay

The activity of the tested compounds against M. tuberculosis H37Rv ATCC 27294 was determined in vitro by resazurin microtiter assay (REMA) (Palomino et al., 2002). The minimum inhibitory concentration (MIC) defined as the lowest concentration resulting in 90 % inhibition of growth of M. tuberculosis was determined by incorporating decreasing concentrations of the tested compounds dissolved in dimethylsulfoxide in Middlebrook 7H9 agar medium. The tested compound concentrations ranged from 0.15 to 250 μg mL−1 and the isoniazid from 0.015 to 1.0 μg mL−1. As a standard control, the MIC value of isoniazid was determined on each microplate (Collins and Franzblau, 1997). MIC values represent mean of three separate experiments.

Results and discussion

N,N-dimethylbenzylamine is readily orthometallated by Pd(II) salts via C(sp2)–H bond cleavage to yield the corresponding acetato- or halide-bridged dimers (Cope and Friedrich, 1968; Dupont et al., 2005; Omae, 2004; Ryabov et al., 1985; Ryabov, 1990). These complexes have been widely studied and employed as convenient starting materials of mononuclear and dinuclear cyclopalladated compounds. Particularly, the cleavage of halide bridges (X) in [Pd(C 2,N-dmba)(μ-X)]2 dimers by donor ligands (L) is a classical reaction thoroughly employed for the synthesis of monomeric [Pd(C 2,N-dmba)X(L)] species (de Lucca Neto et al., 1998; Dupont et al., 2005; Ryabov et al., 1995). In this work, the binuclear compounds [Pd(C 2,N-dmba)(μ-X)]2 {X = Br (1), I (2)} have been used as forerunners for the synthesis of mononuclear derivatives [Pd(C 2,N-dmba)X(tu)] {X = Br (3), I (4)}, respectively (Scheme 1).

Scheme 1
scheme 1

Preparation of the cyclopalladated compounds 3 and 4

Full size image

The synthesis and characterization of binuclear compounds [Pd(C 2,N-dmba)(μ-X)]2 {X = Br (1), I (2)} have already been described some decades ago (Crociani et al., 1970; Maassarani et al., 1987). Complex [Pd(C 2,N-dmba)(μ-Cl)]2 is converted into the corresponding halide-bridged analogues by treatment in acetone with aqueous sodium bromide or iodide to give 1 and 2, respectively, whose binuclear structures have already been confirmed by spectroscopic methods (Crociani et al., 1970; Maassarani et al., 1987). Our elemental analyses results were consistent with the empirical formula C9H12BrPd (1) and C9H12IPd (2), agreeing well with their expected binuclear composition. The confirmation of the molecular composition and structure of [Pd(C 2,N-dmba)(Br)(tu)] (3) has come from our spectroscopic studies and single-crystal X-ray diffraction determination (Moro et al., 2009), while compound [Pd(C 2,N-dmba)(I)(tu)] (4) has been prepared and characterized in this work by elemental analysis, infrared (IR) and 1H- and 13C{1H}-NMR spectroscopies. The data obtained from these techniques are discussed as follows. The elemental analyses are consistent with the empirical formula C10H16IN3PdS (4), which indicates a mononuclear structure.

IR and NMR spectra

The maintenance of the ortho-metallated C 2,N-dmba ring in 1–4 was inferred based on the number of bands assigned to γCH vibrational modes. Free N,N-dimethylbenzylamine (Hdmba) displays two γCH absorptions at 736 and 698 cm−1, which is typical of mono-substituted benzene rings (Socrates, 1997). The appearance of a single γCH band at ca. 743 cm−1 for 1–4 strongly supports the 1,2-disubstituted aromatic ring character of dmba (Fig. S1, Supplementary Material). In addition, the main vibrational modes of the cyclometallated ring at ca. 3047 cm−1 (νC–Har), 2985–2834 cm−1 (νC–Haliph.) and 1594 cm−1 (νC=C) in the IR spectrum of [Pd(C 2,N-dmba)(I)(tu)] (4) remain unchanged when compared with those observed for the precursor, a proof that the coordination of the thiourea to the Pd(II) atom did not affect the integrity of the orthometallated ring (de Lucca Neto et al., 1998; Moro et al., 2009). Among the physical techniques employed to evidence the coordination mode of thiourea-type ligands, IR spectroscopy is one of the most widely used methods. According to Swaminathan and Irving (1964), the electronic structure of thiourea may be denoted by a hybrid of the following three canonical forms (I, II and III), Scheme 2.

Scheme 2
scheme 2

Resonance forms of thiourea

Full size image

In principle, thiourea is potentially capable of coordinating through both sulfur and nitrogen atoms. Both these possibilities will be reflected in the IR spectra of its complexes. Being a typical soft Lewis acid, Pd(II) shows a specific affinity to sulfur donors. When thiourea is S-coordinated to Pd(II), the contribution of the structures II and III will increase, resulting in an decrease in the νC=S frequency (737 cm−1 in free thiourea) due to the weakening of the C=S bond on coordination. The opposite effect is expected if a Pd–N bond is formed, with no appreciable change in νNH frequency because of the existence of hydrogen bonding interactions in the solid state. In this present case, the coordination of thiourea via thione sulfur atom was suggested by the decrease of intensity and shift to lower frequency of the νC=S band from 737 cm−1(free ligand) to 719 cm−1 (4). This is in agreement with a reduction in the double-bond character of the C=S bond on coordination mode (NH2)2C=S–Pd of thiourea (Arantes et al., 2011; Moro et al., 2006; Nadeem et al., 2010). Such observed shifts in the thiourea vibrations upon coordination with respect to the free ligand in 4 are similar to those observed for its analogous [Pd(C 2,N-dmba)(Br)(tu)] (3) (724 cm−1,νC=S), whose structure has been determined by single-crystal X-ray crystallography (Moro et al., 2009). The νN–C–N band is obscured by other strong absorptions over the 1510–1480 cm−1 spectral range.

The 1HNMR data (DMSO-d 6 ) also strongly support the cleavage of the dimer [Pd(C 2,N-dmba)(μ-I)]2 (2) into the monomeric compound 4 (Figure S2, Supplementary Material). In addition, comparison of the NMR spectrum of 4 with that of the structurally characterized [Pd(C 2,N-dmba)(Br)(tu)] (3) (Moro et al., 2009) clearly showed their structural similarity. The 1H NMR spectra of the cyclometallated compounds 3 and 4 share several features: (1) one doublet at 7.28 ppm (H3), one double doublets at 6.94 ppm (H4 and H5) and a multiplet between 6.90 and 6.77 ppm (H6) corresponding to the four remaining protons in the orthometallated ring (see numbering scheme in the ‘Experimental’ Section); (2) two singlets at 3.97 and 2.73 ppm associated with the methylene and methyl groups, respectively; and (3) a broad signal at 7.77 ppm attributed to NH2 groups of coordinated thiourea. The downfield shifting of the NH2 signal when compared with that of the free ligand (7.06 ppm) is assigned to the delocalization of π-electrons between N atoms as a consequence of the greater double-bond character of the CN bond as well as the weakening of C=S bond on complex formation (Ahmad et al., 2002; Yamaguchi et al., 1958). The typical signals of the ortho-metallated dmba were also detected in the 13C{1H} spectrum of 4 (Fig.S3, Supplementary Material), and the assignments are presented in Experimental Section.

Thus, the molecular structure of 4 consists of a palladium atom to which a chelating C 2,N-dimethylbenzylamine moiety is coordinated affording a five-membered cyclometallated ring. The square–planar coordination geometry around the metal is achieved by an iodo and a sulfur atom from the thiourea.

Cytotoxic activity against tumor cell lines

In our recent report (Moro et al., 2009), we have noticed significant differences in terms of relative cytotoxic potency between the compounds [Pd(C 2,N-dmba)(X)(tu)] (X = Cl, Br) toward LM3 and LP07 cells. This finding prompted us to synthesize and characterize new cyclopalladated compounds with the aim to establish useful structure–property relationships. Compound [Pd(C 2,N-dmba)(I)(tu)] (4) was evaluated in vitro for inhibition of cell proliferation against murine mammary (LM3) and lung adenocarcinoma (LP07), using cisplatin as a positive control. For comparison purposes, the cytotoxicity of the metal-free ligands (Hdmba and thiourea) along with the binuclear precursors [Pd(C 2,N-dmba)(μ-X)]2 {X = Br (1), I (2)} was also determined under the same experimental conditions. The results are given in Table 1 in terms of IC50 values (the concentration needed to inhibit 50 % of the cellular proliferation).

Table 1 Cytotoxicity data (IC50) of the tu and Hdmba ligands and their palladium (II) complexes against LM3 and LP07 tumor cell lines
Full size table

From the inspection of IC50 values in Table 1, some interesting trends can be observed. For LM3 cells, it was verified that the compound [Pd(C 2,N-dmba)(I)(tu)] (4) was approximately two times more potent (IC50 = 14.4 µM) than [Pd(C 2,N-dmba)(Br)(tu)] (3) and cisplatin (Moro et al., 2009). The cytotoxicity displayed by 4 was also considerably higher than that observed for its chloro-analogue [Pd(C 2,N-dmba)(Cl)(tu)] (IC50 = 72.4 µM) (Moro et al., 2009), suggesting that the replacement of chloro with a bulkier and softer iodo ligand lowered the IC50 value by a factor of ca. 5. With respect to the cytotoxic effects on LP07 cells, compound 4 was approximately 2.2 times less active than [Pd(C 2,N-dmba)(Br)(tu)] (3) but 1.5-fold more active than [Pd(C 2,N-dmba)(Cl)(tu)] (IC50 = 76.6 µM) (Moro et al., 2009).

From these results, it is possible to suggest a ranking of biological activity based on the halide ligands investigated: I > Br > Cl (LM3 cells) and Br > I > Cl (LP07 cells). These findings indicated that the substitution of X = Cl with a more polarizable and bulkier bromo or iodo ligand increased the cytotoxic effects in vitro, implying that a combination of HSAB (hard–soft acid–base) character and steric factors may be involved in the activity of this type of compounds against LM3 and LP07 cells. Nevertheless, the structure–activity relationships proposed in this study are only preliminary since they were based on only two murine cell lines and three Pd(II) compounds, and therefore, testing on further human lung and mammary tumor cell lines is required in order to confirm the structure–activity relationship.

On comparing the IC50 values found for the dimers [Pd(C 2,N-dmba)(µ-X)]2 {X = Br (1), I (2)}, it was noticed that compound 2 was 1.3- and 2.5-fold less cytotoxic against LM3 and LP07, respectively, than its bromo-bridged analogue 1. It appears that depending on the type of the bridged halide ligand (X) in the “Pd(µ-X)2Pd” metallocycle, a different chemical reactivity of these compounds may be achieved, which would affect their propensity to interact with biomolecular targets (Caires, 2007; Fricker, 2007; Seng and Tiekink, 2012).

The bromo-bridged [Pd(C 2,N-dmba)(µ-Br)]2 (1) and its product from the cleavage reaction with thiourea, [Pd(C 2,N-dmba)(Br)(tu)] (3), displayed comparable IC50 values (~29 μM toward LM3 and ~23 μM against LP07). Despite the cytotoxicity against LM3 cell line increased by 2.5-fold on cleaving the iodo-bridged [Pd(C 2,N-dmba)(µ-I)]2 (2) by thiourea, this structural modification did not result in a significant increase in activity toward LP07 cells. It is worth mentioning that the Pd(II) compounds 1–4 were considerably less active than cisplatin against LP07.

Aiming to get a more detailed data concerning the mode of action of the compounds, we decided to determine the mutagenic potential of the cyclopalladated complexes 1–4 using the standard Ames test.

Evaluation of mutagenic activity (Ames test)

Salmonella typhimurium/microsome assay (Ames test) is a widely accepted short-term bacterial assay for identifying compounds able to produce genetic damage that results in gene mutations. The Ames test, which is normally used to evaluate the mutagenic properties of new chemicals and drugs, can be also used to assess the ability of tested compounds to interact with DNA (Mortelmans and Zeiger, 2000). In addition, it is important to consider the mutagenic properties of new drugs to be used in humans, since mutagenicity may lead to the development of secondary tumors (Halámiková et al., 2008).

Bünger et al. (1996) have demonstrated that PdCl2, K2PdCl4 and (NH4)2PdCl4 were non-mutagenic in the Ames test, contrary to their platinum analogues. On the basis of the above results, some authors consider that the use of Pd(II) centers in the design of new antitumor agents may decrease the potential mutagenic and carcinogenic risks (Bünger et al., 1996; Kuduk-Jaworska et al., 2004).

In the current study, we have decided to determine the mutagenic potential of Pd(II) complexes using the standard Ames test in order to determine whether they display genotoxicity and, consequently, are able to interact with DNA. Table 2 shows the MI observed in S. Typhimurium strain TA98, in the presence (+S9) and absence (−S9) of metabolic activation after the treatments with cisplatin and complexes [Pd(C 2,N-dmba)(µ-X)]2{X = Br (1), I (2)} and [Pd(C 2,N-dmba)(X)(tu)] {X = Br (3), I (4)}.

Table 2 Mutagenic index in the strain TA98 of S. typhimurium after treatment with various doses of cisplatin and palladium complexes 1–4, with (+S9) and without (−S9) metabolic activation
Full size table

In the absence and presence of the external metabolizing system, S9 mix, cisplatin shows an induction number of revertant colonies, with a MI higher than 2.0 for all tested concentrations (Table 2). Thus, cisplatin induces frameshift mutations in the strain (TA98). In fact, previous studies have concluded that normal cells in patients undergoing treatment with cisplatin were at an increased risk of genotoxicity and that this must be taken into consideration when deciding on cancer treatment with this agent (Cross et al., 1996).

Contrary to cisplatin drug, the results given in Table 2 showed that none of the Pd(II) compounds induced any increase in the number of revertant colonies relative to the negative control, indicating the absence of any mutagenic activity. These data suggest that the cytotoxicity mechanisms of the cyclopalladated compounds 1–4 may not necessarily involve interacting with DNA and the possibility that their bioactivity is associated with additional molecular targets (other than the DNA) has not been ruled out. In fact, cyclopalladated compounds bearing phosphines synthesized by Caires’ group displayed notable cathepsin B inhibitory activity (Bincoletto et al., 2005).

The absence of mutagenic effect by palladium complexes (1–4) against S. Typhimurium bacterial strain in the Ames assay may suggest that the eventual therapeutical use of these cyclopalladated complexes would not be limited by possible mutagenicity.

Antimycobacterial activities

Tuberculosis (TB), a chronic bacterial infection transmitted through the air, is caused by M. tuberculosis and mainly affects the lungs (pulmonary TB) (Carmo et al., 2011; Pavan et al., 2012). The first report of the internal administration of a palladium compound for the treatment of tuberculosis was in 1943 (Garrett and Prasad, 2004). The interest in exploring the therapeutic potential of palladium(II) compounds as antimicrobial agents has significantly increased over the past 15 years (Garoufis et al., 2009). We and others have been investigating the antimycobacterial activity of Pd(II) compounds (da Silva et al., 2014; de Souza et al., 2010; Ferreira et al., 2012; Maia et al., 2010). Among them, cyclopalladated complexes [Pd(C2,N-dmba)(X)(tu)] (X = Cl, Br) (Moro et al., 2009) and [{Pd(C2,N-dmba)(X)}2(µ-bpp)] (X = Cl, Br, NCO, N3; bpp = 1,3-bis(4-pyridyl)propane) (Moro et al., 2012) have displayed promising activity at micromolar range against M. tuberculosis virulent strain H37Rv. These findings have motivated us to prepare new analogous cyclometallated Pd(II) derivatives and investigate their inhibitory growth effects against MTB.

The MIC values of [Pd(C 2,N-dmba)(μ-X)]2 (X = Br (1), I (2)) and of [Pd(C 2,N-dmba)(X)(tu)] (X = Br (3), I (4)) against M. tuberculosis are shown in Fig. 2. Isoniazid, employed extensively for tuberculosis treatment (Khanye et al., 2011), was used as standard antitubercular drug.

Fig. 2
figure 2

Influence of the Pd(II) compounds 1–4 on M. tuberculosis H37Rv. Each column represents the minimal inhibitory concentration (MIC) of 1–4 against M. tuberculosis. Numbering Scheme of [Pd(C 2,N-dmba)(I)(tu)] (4)

Full size image

Thiourea and Hdmba ligands displayed poor inhibition ability against M. tuberculosis (MIC values >250 μg mL−1), but their coordination to palladium (II) caused an expressive increase of their antitubercular activities (Moro et al., 2009). We speculate that the formation of the “Pd(C 2,N-dmba)” metallocycle may cause a polarity reduction of the complex molecule as a whole, in comparison with the free ligands, by partial sharing of their charges within the complexes. As a consequence, their permeation through the lipid layer of the cell membrane is facilitated, hence promoting a better cell uptake of the active species (Maia et al., 2010).

It was noticed that antiproliferative activity of the compounds of the type [Pd(C 2,N-dmba)(X)(tu)] was also affected by the type of coordinated halide group X. The replacement of bromo by iodo lowered the anti-TB activity by a factor of ca. 2.5. The bioactivity against M. tuberculosis virulent strain H37Rv increased by sixfold on cleaving the bromo-bridged [Pd(C 2,N-dmba)(µ-Br)]2 (1) with thiourea. On the other hand, the iodo-bridged [Pd(C 2,N-dmba)(µ-I)]2 (2) and its product from the bridge-splitting reaction with thiourea, [Pd(C 2,N-dmba)(I)(tu)] (4), displayed comparable MIC indexes (60 μgmL−1). All Pd(II) complexes tested were less effective than isoniazid against the M. tuberculosis (MIC value of 0.030 μg mL−1) (Collins and Franzblau, 1997). However, the activities of complexes 2, 3 and 4 are comparable or better than pyrazinamide (MIC value of 50–100 μg mL−1) (de Souza et al., 2010; Moro et al., 2009), another drug employed for tuberculosis treatment.

Conclusions

In the challenging field of drug discovery, chemists must keep using all data available for a better drug design (da Silva et al., 2013; Despaigne et al., 2010; Fricker, 2007; Marchesi et al., 2012). Learning from experience, we can effectively devise new strategies and paradigms, aiming at the end with compounds that could enhance healthcare. In this line of reasoning, this study illustrates in a convincing way how changes in the compounds nuclearities and in their coordination spheres influence the biological activities. Thus, one major goal was to design an effective rational route via cleavage of dimeric cyclopalladated compounds [Pd(C 2,N-dmba)(μ-X)]2 (X = Br (1), I (2)) to obtain monomeric species containing thiourea coordinated to palladium [Pd(C 2,N-dmba)(X)(tu)] (X = Br (3), I (4)), which have shown in some instances better activities against cell tumor lines and M. tuberculosis than their precursors.

Our experiments illustrate that the cytotoxic activity of simple, structurally related Pd(II) compounds can be modulated from 14 to >50 μM simply by changing the type of halide ligand. The absence of mutagenic effect by palladium complexes (1–4) detected by the Ames assay is an important result since the absence of mutagenic effects by palladium complexes (1–4) infers that mutagenicity would not be an obstacle for an eventual therapeutic use of these cyclopalladated compounds (Kuduk-Jaworska et al., 2004; Quilles et al., 2013).

Further investigations on this class of compounds are underway in our laboratories in order to rationalize the IC50 values in terms of structure–activity relationship as well as to understand the molecular basis for the cytotoxicity of the cyclopalladated derivatives.

Supplementary information

Supplementary information contains the IR and1H and 13C NMR spectra of compound 4 (Figures S1, S2 and S3, respectively).