Design of vanadium mixed-ligand complexes as potential anti-protozoa agents

Journal of Inorganic Biochemistry 103 (2009) 609–616

Contents lists available at ScienceDirect

Journal of Inorganic Biochemistry journal homepage: www.elsevier.com/locate/jinorgbio

Design of vanadium mixed-ligand complexes as potential anti-protozoa agents Julio Benítez a,1, Lucía Guggeri b,1, Isabel Tomaz c,d,1, Gabriel Arrambide a, Maribel Navarro e, João Costa Pessoa c, Beatriz Garat b,*, Dinorah Gambino a,* a

Cátedra de Química Inorgánica, Facultad de Química, Universidad de la República, Gral. Flores 2124, C. C. 1157, 11800 Montevideo, Uruguay Laboratorio de Interacciones Moleculares, Facultad de Ciencias, Universidad de la República, Iguá 4225, 11400 Montevideo, Uruguay c Centro de Química Estrutural, Instituto Superior Técnico, TU Lisbon, Av. Rovisco Pais, 1049-001 Lisbon, Portugal d Centro de Ciências Moleculares e Materiais, Faculdade de Ciências da Universidade de Lisboa, Campo Grande, 1749-016 Lisbon, Portugal e Instituto Venezolano de Investigaciones Científicas, Carretera Panamericana, KM 11, Altos de Pipe, Centro de Química, 1020-A, Caracas, Venezuela b

a r t i c l e

i n f o

Article history: Received 13 September 2008 Received in revised form 13 October 2008 Accepted 20 October 2008 Available online 7 November 2008 Keywords: Chagas’ disease Vanadium Salicylaldehyde semicarbazones Pyridyl ligands DNA

a b s t r a c t In the search for new therapeutic tools against Chagas’ disease (American Trypanosomiasis) four novel mixed-ligand vanadyl complexes, [VIVO(L2-2H)(L1)], including a bidentate polypyridyl DNA intercalator (L1) and a tridentate salycylaldehyde semicarbazone derivative (L2) as ligands were synthesized, characterized by a combination of techniques, and in vitro evaluated. EPR suggest a distorted octahedral geometry with the tridentate semicarbazone occupying three equatorial positions and the polypyridyl ligand coordinated in an equatorial/axial mode. Both complexes including dipyrido[3,2-a: 20 ,30 -c]phenazine (dppz) as polypyridyl coligand showed IC50 values in the lM range against Dm28c strain (epimastigotes) of Trypanosoma cruzi, causative agent of the disease, being as active as the anti-trypanosomal reference drug Nifurtimox. To get an insight into the trypanocidal mechanism of action of these compounds, DNA was evaluated as a potential parasite target and EPR, and 51V NMR experiments were also carried out upon aging aerated solutions of the complexes. Data obtained by electrophoretic analysis suggest that the mechanism of action of these complexes could include DNA interactions. Ó 2008 Elsevier Inc. All rights reserved.

1. Introduction According to World Health Organization (WHO), infectious and parasitic diseases are major causes of human disease worldwide [1,2]. Although representing a tremendous burden when compared to other communicable diseases that receive a high level of attention from health systems, a group of parasitic and infectious diseases has been characterized by historically low investment by the pharmaceutical industry. Among them, Chagas’ disease (American Trypanosomiasis) is the largest parasitic disease burden in the American continent, being endemic in 21 countries from Mexico to southern Argentina and Chile. The morbidity and mortality associated with this disease in America are more than one order of magnitude higher than those caused by malaria, schistosomiasis or leishmaniasis [1,3]. The disease, caused by the flagellate protozoan parasite Trypanosoma cruzi (T. cruzi), is transmitted to humans either by transfusion of in-

* Corresponding authors. Tel.: +598 2 9249739; fax: +598 2 9241906 (D. Gambino). E-mail addresses: [email protected] (B. Garat), [email protected] (D. Gambino). 1 Equally contributed to the work. 0162-0134/$ - see front matter Ó 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.jinorgbio.2008.10.018

fected blood, from an infected mother to her child, or by its most important vector, a blood-sucking insect of Triatominae family (called vinchuca, chipo, barbeiro, kissing bug, cone nose, or assassin bug, depending on the geographical region), which carries the parasite in its contaminated feces. Humans and a large number of species of domestic and wild animals constitute the reservoir of the parasite. Despite the progress achieved in the study of T. cruzi’s biochemistry and physiology, in which several potential new drug targets have been identified, the chemotherapy of this parasitic infection remains undeveloped and no effective method of immune prophylaxis is available. The treatment has been based on old and quite unspecific nitroaromatic drugs that have significant activity only in the acute phase of the disease and cause severe side effects [4,5,6–8]. The development of bioactive metal complexes appears to be a promising new approach in the search for a pharmacological control of Chagas’ disease [5,9–12]. Complexes have been designed by combining ligands bearing anti-trypanosomal activity and metals with pharmacological potential. This strategy takes advantage of the emerging medicinal chemistry paradigm on drug discovery to develop agents that could modulate multiple targets simultaneously with the aim of enhancing efficacy rela-

610

J. Benítez et al. / Journal of Inorganic Biochemistry 103 (2009) 609–616

of the complexes in solution was investigated by EPR and 51V nuclear magnetic resonance spectroscopy, and conductivity measurements. Their in vitro anti-T. cruzi activity was tested against Dm28c strain of the parasite and compared to that of the reference anti-trypanosomal drug Nifurtimox. Furthermore, to provide insight into the probable mechanism of anti-trypanosomal action, the compounds were tested for their DNA interaction ability on plasmid DNA by using gel electrophoresis experiments. Time course studies of the interaction have been also carried out.

tive to drugs that address only a single target [13]. In the case of the metal complexation approach, the new metal compounds could act through dual or even multiple mechanisms of action by combining the pharmacological properties of both the ligand and the metal in a cooperative effect [11,12]. Using this approach we have previously developed different bioactive metal complexes and studied their mechanism of action [14–22]. It is worth to mention that vanadium compounds have been scarcely investigated for the development of anti-parasitic drugs, the vanadyl complexes of N-oxide quinoxaline derivatives that we previously reported being the only ones showing anti-T. cruzi activity [18]. Therefore, we planned to further study the potential of vanadium for the development of novel bioactive anti-trypanosomal compounds. Metabolic pathways of kinetoplastid parasites (Leishmania and Trypanosoma parasites) are similar to those present in tumor cells leading to a correlation between anti-trypanosomal and anti-tumor activities. In addition, it has been proposed that compounds that efficiently interact with DNA in an intercalative mode could also show anti-trypanosomal activity [11,12,23]. To test this hypothesis we designed mixed-ligand vanadyl complexes (Fig. 1) that include in the VIV-coordination sphere polypyridyl chelators (Npy, Npy donors) capable of intercalating DNA. Among them, dipyrido[3,2-a: 20 ,30 -c]phenazine (dppz) was included taking into account the anti-Leishmania sp. activity previously shown by some of its gold and copper complexes [24,25]. Some semicarbazones, versatile ligands which present a wide range of bioactivity, were selected as coligands [26]. In particular, tridentate salicylaldehyde semicarbazone derivatives were selected since, according to our previous work, their vanadium(V) complexes, VVO2L, showed cytotoxic effect on kidney tumor cells (TK-10 cell line) [27,28]. Four novel vanadyl complexes, [VIVO(L2-2H)(L1)], where L1 = dppz or 2,20 -bipyridine (bipy) and L2 = salicylaldehyde semicarbazone (Salsem) or 5-bromosalicylaldehyde semicarbazone (BrSalsem), were synthesized, characterized and evaluated in in vitro experiments (Fig. 1). The complexes were characterized by elemental analysis, electrospray ionization mass spectrometry (ESI-MS), conductimetric measurements and infrared (FTIR) and electronic paramagnetic resonance (EPR) spectroscopies. Stability

N

2. Experimental 2.1. Materials All common laboratory chemicals were purchased from commercial sources and used without further purification. VIVO(acac)2, where acac = acetylacetonate, and 2,20 -bipyridine (bipy) were commercially available. Both semicarbazone ligands were synthesized from an equimolar mixture of the aldehyde (salicylaldehyde or 5-bromosalicylaldehyde) and semicarbazide using a previously reported procedure [27,28]. Dppz was also synthesized by a previously reported method [29]. 2.2. Syntheses of the mixed-ligand vanadyl complexes [VO(L2-2H)(L1)] [VIVO(L2-2H)(L1)] complexes, where L1 = dppz or bipy and L2 = salicylaldehyde semicarbazone (Salsem) or 5-bromosalicylaldehyde semicarbazone (BrSalsem) were synthesized by the following procedure: 0.375 mmol of L2 (67 mg Salsem or 88 mg BrSalsem) and 0.375 mmol of L1 (59 mg bipy or 110 mg dppz) were suspended in 15 mL of absolute alcohol previously purged with nitrogen for 10 min. VIVO(acac)2 (0.375 mmol, 100 mg) was suspended in 6 mL of absolute alcohol; this suspension was added to the previous one and nitrogen was kept passing through the solution for ca. 10 min. The suspension was refluxed under nitrogen for 3.5 h and the coloured solid formed was filtered off from the hot mixture, and solid was washed three times with 2 mL portions of EtOH:Et2O (1:1).

H N

R

N

NH2

N O

N

OH

N

Salsem BrSalsem

R=H R = Br

salicyladehyde semicarbazone 5-bromosalicylaldehyde semicarbazone

dppz = dipyrido[3,2-a:2´,3´-c]phenazine

L2 = BrSalsem or Salsem X

X = Br or H

O O

X

N

O

N

V

N

O

N N

bipy = 2,2´-bipyridine

2

O

N

N

N

L1 = bipy or dppz

1

[VO(L -2H)(L )]

Fig. 1. Scheme showing the designed mixed-ligand vanadyl complexes and the ligands included in their VIVO-coordination sphere.

J. Benítez et al. / Journal of Inorganic Biochemistry 103 (2009) 609–616

2.2.1. VIVO(Salsem-2H)(dppz), 1 Red solid, yield: 179 mg, 91%. Anal. calc. for C26H17N7O3V: C, 59.3; H, 3.2; N, 18.6. Found: C, 59.0; H, 3.5; N, 17.9. ESI-MS (MeOH) m/z: 526.5 (100%), 527.4 (30%) (M+, C, N, H, O isotope pattern), 528.0 (M+H+). KM(DMF): 9.6 lS/cm2. (Note that even though the elemental analysis value for N was somewhat off (by 0.7%), the chemical formula for 1 was well supported by the ESI-MS analysis.) 2.2.2. VIVO(Salsem-2H)(bipy), 2 Orange brown solid, yield: 40 mg, 27%. Anal. calc. for C18H15N5O3V: C, 54.0; H, 3.8; N, 17.5. Found: C, 53.7; H, 3.7; N, 17.1. ESI-MS (DMSO) m/z: 401.1 (M+H+). KM(DMF): 7.7 lS/cm2. 2.2.3. VIVO(BrSalsem-2H)(dppz), 3 Red brown solid, yield: 155 mg, 68%. Anal. calc. for C26H16N7O3BrV: C, 51.6; H, 2.6; N, 16.2. Found: C, 50.9; H, 3.0; N, 15.8. ESI-MS (MeOH) m/z: 604.2 (100%), 606.0 (89%), 605.2 (21%), 607.0 (36%), 608 (5%) (M+, C, N, H, O isotope pattern). KM(DMF): 11.1 lS/cm2. (Note that even though the elemental analysis value for C was somewhat off (by 0.7%), the chemical formula for 3 was well supported by the ESI-MS analysis.) 2.2.4. VIVO(BrSalsem-2H)(bipy), 4 Orange solid, yield: 47 mg, 26%. Anal. calc. for C18H14N5O3BrV: C, 45.1; H, 2.9; N, 14.6. Found: C, 45.2; H, 3.2; N, 14.9. ESI-MS (MeOH) m/z: 478.0 (100%), 480.3 (100%), 479.0 (20%), 481.4 (20%), 482.3 (5%) (M+, C, N, H, O isotope pattern). KM(DMF): 10.1 lS/cm2. 2.3. Physicochemical characterization C, H and N analyses were performed with a Carlo Erba Model EA1108 elemental analyzer. Conductimetric measurements were carried out at 25 °C in 103 M dimethylformamide (DMF) solutions using a Conductivity Meter 4310 Jenway [30]. Electrospray ionization mass spectra (ESI-MS) of methanol or DMSO solutions of the complexes were recorded on a 500-MS Varian ion trap mass spectrometer in the positive mode (capillary voltage: 80 V; needle voltage: 5 kV; nebulizer gas: nitrogen; nebulizer pressure: 35 psi; drying gas temperature: 350 °C; drying gas pressure: 10 psi; m/z range recorded: 100–1000). FTIR spectra (4000–400 cm1) of the complexes and the free ligands were measured as KBr pellets with a Bomen FTIR model MB102 instrument. 51V NMR spectra of ca. 2 mM solutions of the complexes in DMSO and DMF (p.a. grade) were recorded on a Bruker Avance III 400 MHz instrument 30 min. after dissolution, and after a 4-day period standing in aerobic conditions. 51V chemical shifts were referenced to a VOCl3 primary external reference. Inorganic vanadate solutions in DMSO and DMF were also used for comparison. EPR spectra were recorded at 77 K on a Bruker ESP 300 spectrometer coupled to a Bruker ER041 X-band frequency meter (9.45 GHz). Complexes were dissolved in DMSO or DMF p.a. grade, previously degassed by passing N2 for 10 min, using ultrasound to completely dissolve the solid. Complex concentrations were kept at 0.9–1 mM to prevent precipitation upon freezing to 77 K. Solutions were prepared at room temperature, and immediately frozen in liquid nitrogen prior to recording the EPR spectrum. 2.4. In vitro anti-T. cruzi activity (Dm28c strain) T. cruzi epimastigotes of the Dm28c strain were maintained in exponential growth at 28 °C in liver infusion tryptose (LIT) medium complemented with 10% fetal calf serum. The effect on cell growth was analyzed incubating an initial concentration of 1  106 cells/ mL with various concentrations of the compounds for 5 days. Compounds were added as stock DMSO solutions immediately after

611

their preparation. The percentage of cell growth was followed measuring the absorbance, A, of the culture at 595 nm and calculated as follows: % = (ApA0p)/(AcA0c)  100, where Ap = A595 of the culture containing the drug at day 5; A0p = A595 of the culture containing the drug at day 0; Ac = A595 of the culture in the absence of any drug (control) at day 5; A0c = A595 in the absence of the drug at day 0. The results represent averages ± SE. The final DMSO concentration in the culture media never exceeded 0.4% (vol/vol) and had no effect by itself on the proliferation of the parasites [17,18]. Nifurtimox (Nfx) was used as the reference trypanocidal drug. Dose-response curves were recorded and the IC50 (50% inhibitory concentration) values were determined. 2.5. Plasmid DNA interaction studies Plasmid DNA interaction studies were performed using a previously described procedure [19]. Plasmid DNA (pBSK II BlueScript, Stratagene; 300 ng per reaction) was obtained and purified according to standard techniques [31]. Briefly, Escherichia coli XL1 cells were transformed with pBSK II. Transformation was verified by polymerase chain reaction (PCR) and plasmidic DNA was purified (Qiagen Plasmid Maxi Kit). Spectrophotometric DNA quantification was carried out assuming a molar absorption coefficient at 260 nm of 0.02 lg1 mL cm1. The complexes were dissolved in less than 4% DMSO–H2O and incubated with the purified DNA (300 ng) for 24 h at 37 °C (final volume: 20 lL, reaction buffer: Tris (tris(hydroxymethyl)aminomethane hydrochloride 10 mM pH 7.4). No effect on DNA due to DMSO addition was observed [19] even for concentrations up to 50% (data not shown). Three molar ratios ri (ri = mol of complex:mol of base pair), 1.0, 3.0 and 6.0, were assayed. After incubation, reactions were stopped by the addition of loading buffer (25% bromophenol blue, 50% glycerol, 25 mM EDTA pH 8.0). Samples were run in 0.7% agarose buffered with 90 mM Tris–borate-EDTA (TBE) at 70 V for 2 h. The gel was subsequently stained with an ethidium bromide solution (0.5 lg/ mL) for 30 min and destained in water for 20 min. Bands were visualized under UV light and quantified using OneDSCAN.Time course of the interaction has also been studied. The complexes were incubated with the plasmid DNA as described at a single ri value (ri = 3.0) for different times.

3. Results and discussion Four new mixed-ligand VIVO-complexes including a bidentate polypyridyl ligand (dppz or bipy) and a tridentate salicylaldehyde semicarbazone derivative (Salsem or BrSalsem) in the V(IV) coordination sphere were synthesized with high purity and reasonable yields. All of them are neutral non conducting compounds in DMF. Analytical, ESI mass spectrometry and FTIR and EPR spectroscopic results are in agreement with the proposed formula, [VIVO(L2-2H)(L1)], with L1 = dppz or bipy and L2 = Salsem or BrSalsem (Fig. 1). ESI-MS experiments allowed the clear detection of the molecular ion for each complex. In the case of complexes containing the brominated ligand, two peaks were detected for M+, reflecting the isotope distribution of 79Br and 81Br. For all samples the minor m/z signals observed reproduced the pattern expected for each compound content in C, H, N and O elements due to their isotope relative abundance/distribution. 3.1. IR spectroscopic studies The simultaneous presence of the semicarbazone and the polypyridyl ligands in the coordination sphere lead to spectral complexity. In particular, several bands corresponding to m(C@C) and

612

J. Benítez et al. / Journal of Inorganic Biochemistry 103 (2009) 609–616

Table 1 Tentative assignment of selected IR bands of the VIVO-complexes. Bands for the free semicarbazone ligands are included for comparison [27,28]. Band positions are given in cm1.

H N

R

NH2

N O OH Compound

m(VO)

m(C@O)

m(C@N)a

m(O–H)

m(N–H)

Salsem VO(Salsem-2H)(dppz) VO(Salsem-2H)(bipy) BrSalsem VO(BrSalsem-2H)(dppz) VO(BrSalsem-2H)(bipy)

– 961 961 – 964 959

1683 1610 1606 1686 1609 1612

1598 1602 1600 1581 1596 1598

3422 – – 3422 – –

3160 – – 3170 – –

a The bands assigned to m(C@N) (azomethyne) may not be pure, as they may be associated with the aromatic (C@C) stretching bands [33].

m(C@N) in heterocyclic compounds lie in the 1650–1550 cm1 region, interfering with the assignments [24,25]. Anyway, taking into account our previous knowledge on vibrational behaviour of semicarbazone and thiosemicarbazone metal complexes [14,17,21,27,28], vibration bands related with the semicarbazone ligand’s coordination mode were tentatively assigned. Selected vibration bands and their tentative assignments are depicted in Table 1. The non-observation of the m(C@O) bands, present in the ligands at 1683 (Salsem) and 1686 (BrSalsem) cm1, indicates the enolization of the amide functionality upon coordination to the VIV-centre. Instead strong bands at ca. 1606–1612 cm1 are observed, which can be attributed to the asymmetric stretching vibration of the

conjugated CH@NAN@C group, characteristic for the coordination of the enolate form of the ligands [32]. The shifts of m(C@O) and m(C@N) bands and the disappearance of m(NH) and m(OH) bands (both in the 3150–3500 cm1 region), are in agreement with tridentate coordination through the carbonylic oxygen (OOAC(NH2)@N), the azomethyne nitrogen (Nazomethyne) and the phenolic oxygen (Ophenolate), and with double deprotonation of the semicarbazone ligand [18,27,28,34]. The strong m(VO) band around 960 cm1 could be clearly identified for all the complexes. 3.2. Characterization of the complexes in solution As the in vitro tests were carried out in aerated diluted solutions and with incubating periods of several days, it is important to determine the nature of complexes in solution and to understand their stability towards hydrolysis and/or oxidation of VIV. In fact the active biological species may differ significantly from the complex that is initially dissolved. To characterize further fresh and aged solutions of complexes 1–4 (see Fig. 2 legend) we made some EPR and 51V NMR studies. 3.2.1. EPR studies EPR spectroscopy of vanadyl complexes is a powerful tool to obtain structural information on the binding mode of the species present in solution [35]. The additivity relationship, first proposed by Wuthrich [36] and later refined by Chasteen [37], correlates the hyperfine coupling constant, namely A||, to the electron-donating ability of the ligands present in the equatorial plane of the VIVOP [A||i(i = 1, . . . , 4)], A||i standing centre, and it can be expressed as A|| = for the specific contribution of each donor atom equatorially bound to the VIVO moiety. Care must be taken when applying this rela-

1

B I

2

II

5.00E+05 simul.

4

IIIa IIIb II

1.75

I

1.65

I II

2.3

2.1

g-value

1.9

1.55

1 2 3 4 2 4

IIIb

2.5

2

1.7

1.5

Fig. 2. (A) X-band 1st derivative EPR spectra of frozen solutions (77 K) of complexes 1–4 in DMF (black) and DMSO (grey): 1 = [VIVO(Salsem-2H)(dppz)], 2 = [VIVO(Salsem2H)(bipy)], 3 = [VIVO(BrSalsem-2H)(dppz)], 4 = [VIVO(BrSalsem-2H)(bipy)]. (B) High field range detail (3800–4400 G) of the spectra, showing an amplification of the signals, and a representative simulated spectrum (‘‘simul.”) for complex 4 in DMF. Species I is assigned to the parent complexes, species II–III are their solvolysis products.

613

J. Benítez et al. / Journal of Inorganic Biochemistry 103 (2009) 609–616

Table 2 Spin-Hamiltonian parameters obtained by simulation of X-band (77 K) EPR spectra for complexes [VIVO(L2-2H)(L1)] 1–4, where 1 = VIVO(Salsem-2H)(dppz), 2 = VIVO(Salsem2H)(bipy), 3 = VIVO(BrSalsem-2H)(dppz), 4 = VIVO(BrSalsem-2H)(bipy). Complex

Species*

Spin-Hamiltonian parameters gz

Az/104 cm1

gx,y

Ax,y/104 cm1

Assignment

Solvent

1

I (M) II (m, id) IIIa (m, id) IIIb (M)

1.952 1.951 1.95 1.94

160.1 165.7 173 179

1.982 1.981 n.a. n.a.

55.6 55.7 n.a. n.a.

[VO(L2-2H)L1]eq [VO(L2-2H)S]eq [VO(Npy)S3]eq [VO(S)4]eq

DMF DMF DMSO DMSO

2

I (M) II (m) I (M) II (m, id) IIIb (id)

1.952 1.951 1.948 1.944 1.93

159.7 166.2 158.1 169 178

1.982 1.981 1.979 n.a. n.a.

54.9 56.8 55.5 n.a. n.a.

[VO(L2-2H)L1]eq [VO(L2-2H)S]eq [VO(L2-2H)L1]eq [VO(L2-2H)S]eq [VO(S)4]eq

DMF DMF DMSO DMSO DMSO

3

I (M) II (m, id) IIIa (m, id)

1.952 1.951 1.93

159.8 165.8 172

1.982 1.981 n.a.

54.8 55.3 n.a.

[VO(L2-2H)L1]eq [VO(L2-2H)S]eq [VO(Npy)S3]eq

DMF DMF DMSO

4

I (M) II (m) I (M) II (m, id)

1.952 1.951 1.950 1.94

160.0 166.1 160.0 166

1.982 1.981 1.98 n.a.

55.6 56.9 53 n.a.

[VO(L2-2H)L1]eq [VO(L2-2H)S]eq [VO(L2-2H)L1]eq [VO(L2-2H)S]eq

DMF DMF DMSO DMSO

M = major species (present in higher concentration); m = minor species (present in lower concentration); id = ‘‘ill defined species”; being present at low concentrations, the spectrum lines are difficult to assign with certainty; n.a. = parameters not available and/or inconsistent due to the very low concentration and/or severe line overlap. * L2-2H = semicarbazone ligand: (Ophenolate, Nazomethyne, OOAC(NH2)@N); L1  (Npy, Npy) base: bipy or dppz; S  solvent/H2O.

tionship as it may be difficult to distinguish between different donor groups with similar A||i values [38]. Contributions of interest to this work are: A||(Ophenolate) = 38.9  104 cm1 [37], A||(Npy) = 39.0  104 cm1 [38] and A||(DMF) = 43.7  104 cm1 [35]. The contribution of DMSO is taken as A||(DMSO) = 45.0  104 cm1. The A||(Nimine or Nazomethyne) values reported range from 38.5 to 44.0  104 cm1 [35,39,40] and were taken here as 40.0  104 cm1. The EPR spectra of the frozen solutions of the four VIVO-mixedligand complexes (Fig. 2) exhibit a hyperfine pattern typical of VIVO-complexes, consistent with the presence of monomeric 1 VIVO-bound species with dxy ground-state configuration. The EPR spectra exhibit no significant rhombicity and were analyzed as axial spectra. Although soluble at room temperature at concentrations of 3 mM, the solubility of these complexes decreased substantially at lower temperatures, and partial precipitation occurred upon freezing the samples whenever the complex concentration was above 1 mM. This fact limited the intensity of EPR spectra obtained which exhibited a not very favourable signal-tonoise ratio, especially for the minor species present. Nevertheless valuable structural information could still be extracted. Table 2 summarises the EPR parameters obtained for all samples in both solvents. The spectra displaying several paramagnetic species were analyzed with an iterative procedure using equations proposed by Chasteen [37] and corrected by Casella et al. [41], to estimate the hyperfine coupling constants and g-values. Some spectra were simulated using a computer program developed by Rockenbauer and Korecz [42]. Overall four paramagnetic species (labelled I, II, IIIa and IIIb) could be detected. Species I is normally the major species and assigned to [VIVO(L22H)(L1)] species. The spin-Hamiltonian parameters for species I of all four complexes are the same within experimental error regardless the solvent used, indicating the same binding mode for all four complexes, and are very close to those recently reported for [VIVO(L)(bipy or phen)] complexes (H2L = 2-hydroxy-4-methoxybenzaldehyde nicotinic acid hydrazone or 2-hydroxy-4-methoxyacetophenone nicotinic acid hydrazone) [43]. Semicarbazone ligands in mixed-ligand vanadium(IV) or (V) complexes form chelate rings which normally are practically planar, the donor atoms binding equatorially [27,28,32,33,44,45]. Therefore, we expect similar binding modes in complexes 1–4,

the L2-2H ligands coordinating by the Ophenolate, Nazomethyne, Ocarbonyl donor atoms, occupying the equatorial plane in a (5 + 6) chelate ring arrangement, and L1 binding to the VIVO moiety in an axial-equatorial arrangement (Fig. 1). The contribution to A|| of the carbonyl oxygen in the donor set  OAC(NH2)@N of semicarbazone ligands was not previously reported, and that rules out the calculation of an ‘‘expected” value for A|| for I from the additivity rule. However, knowing the contribution of all other donors enables the back calculation of a value for A||(OAC(NH2)@N). In fact, using the contributions mentioned before, A||(species I) = 160.0  104 cm1 yields the value of 42.1104 cm1 for A||(OAC(NH2)@N). Species II and III (a and b) are identified as decomposition products of the parent complexes. Species II is observed in all cases, and indicates partial solvent attack through the displacement of the Npy bidentate coligand by the solvent. The corresponding spinHamiltonian parameters are consistent with the binding mode [(Ophenolate, Nazomethyne, OOAC(NH2)@N)(DMSO)]equatorial. In fact, using 42.1104 cm1 for the OAC(NH2)@N contribution, we obtain: ¼ 164:7  104 cm 1 in DMF and Aexpected ¼ 166:0  Aexpected jj jj 4 10 cm 1 in DMSO. Therefore, we assign II to [VIVO(L2-2H)(DMF or DMSO)] complexes, with the semicarbazone ligand coordinating as tridentate in the equatorial plane. Species IIIa and IIIb have spin-Hamiltonian parameters that are too high for what is expected for the donors involved (A||  173– 179  104 cm1), and are assigned to different degrees of hydrolysis/decomposition of the parent complexes. For these species (observed only in DMSO) the spin-Hamiltonian parameters are obtained with a high level of uncertainty due to the very low intensity of the mI(5/2)|| and especially mI(7/2)|| spectral lines. Species IIIa is consistent with a [VIVO(DMSO)3(Npy)]eq binding mode, with an equatorial-axial geometry for the donors of the coligand L1, and the full displacement of the semicarbazone ligand. Species IIIb is assigned to the [VIVO(DMSO)4]eq donor set. In summary, the EPR data indicate that all four VIV-complexes are sensitive to solvent attack, and partial solvolysis occurs upon their dissolution, solvolysis being less extensive in DMF. Both complexes with dppz as a coligand (1 and 3) seem to be more sensitive to solvolysis in DMSO than in DMF when compared with the complexes with bipy (2 and 4). Nevertheless, the main complex species detected in freshly prepared solutions is [VIVO(L2-2H)(L1)] and as is

614

J. Benítez et al. / Journal of Inorganic Biochemistry 103 (2009) 609–616

described below no VV-species are detected in these solutions by 51 V NMR. 3.2.2. NMR studies Having detected some solvolysis of the complexes in solution by EPR, 51V NMR spectroscopy was used to detect probable oxidation products after ageing the aerated solutions. Spectra of 2 mM solutions of the complexes in DMSO and DMF were measured at room temperature after dissolution (ca. 20–30 min.), and after a 4-day standing period in the presence of oxygen. Although freshly prepared solutions did not show any peaks, after 4 days all solutions exhibited one broad single peak for each sample in the 51V NMR spectra (d/ppm in DMSO: 543.6 for 1; 543.9 for 2; 543.4 for 3; 540.8 for 4; d/ppm in DMF: 537.3 for 1; 537.7 for 2; 536.6 for 3; 536.9 for 4). There are several publications with 51V NMR data of VV-L-L1 complexes (L = several hydrazonato-type ligands and L1 = a secondary ligand such as bipy or 8-hydroxyquinoline). The chemical shifts for the VVO(L)(L1) species are normally in the range 470 to 480 ppm [32,46], and in some cases, where L1 are non-innocent ligands, at very different d values. In contrast, the corresponding VOþ 2 -L complexes display resonances at ca. 505 to 550 ppm [44,45], the d values being often solvent dependent. The d values observed for aged (4 days) aerated solutions of complexes 1–4 in DMF or DMSO suggest the presence of VV bound 51 V NMR spectra indicates the presence of as VV Oþ 2 units, i.e. the V 2 [V O2(L -2H)(DMF or DMSO)] complexes where the L2-2H ligands are coordinated as tridentate donors to the VV Oþ 2 cation, while the solvent may occupy the vacant coordination position. Conductimetric measurements show an increase in conductivity of DMSO and DMF aerated solutions after several days in presence of oxygen, indicating that oxidation modifies charge and probably coordination sphere. The in vitro tests show that solutions of complexes 1–4 display interesting anti-trypanosomal activity. However, these EPR and 51V NMR results indicate that it is not possible at this stage to correlate this activity with a particular VIV- or VV-species formed by complexes 1–4 upon their dissolution in the incubation media. 3.3. In vitro anti-T. cruzi activity The complexes were evaluated in vitro for their anti-T. cruzi activities against epimastigotes of Dm28c strain. Results were compared to that of the reference drug Nifurtimox. Cell growth percentages in respect to control at different doses are shown in Fig. 3. Fifty percent inhibitory concentrations (IC50) obtained from these dose-response curves are depicted in Table 3. Salsem and BrSalsem were tested up to 100 lM and 10 lM doses, respectively, through the same in vitro test not showing inhibitory effect on T. cruzi epimastigotes. Instead, the bipy ligand showed an IC50 value of ca. 70 lM. Due to the low solubility of dppz in aqueous solution at the concentrations needed for the analysis of growth inhibition, the IC50 for this ligand could not been properly determined. All the complexes were active in vitro against epimastigote form of T. cruzi (Dm28c strain). Those complexes including dppz in the coordination sphere resulted significantly more active than those with bipy as coligand, showing IC50 values of the same order than Nifurtimox. 3.4. Plasmid DNA interaction studies The effect of the four vanadium complexes on supercoiled DNA was studied using different ri at 37 °C during a 24 h incubation period (Fig. 4). In the assayed conditions, plasmid DNA is visualized as supercoiled DNA. The corresponding linear and nicked DNA forms

Fig. 3. Dose-response curves for the four vanadyl complexes. The inhibition of exponential growth of an initial concentration of 1  106 Dm28c T. cruzi epimastigotes/mL in the presence of the indicated concentration of complexes was analyzed after 5 days of incubation following the A595 as indicated in Section 2. Upper panel, (A) VO(L2-2H)(bipy) complexes: 2, } – continuous line and 4, s – dashed line; Lower panel and (B) VO(L2-2H)(dppz) complexes: 1, h – dashed line, and 3, 4 – continuous line. Each point represents the average of three to six experiments ± SE.

Table 3 In vitro biological activity of the vanadyl complexes. IC50 values on T. cruzi (Dm28c strain). Nifurtimox has been included for comparison. Compound

IC50/lM

VO(Salsem-2H)(dppz) VO(Salsem-2H)(bipy) VO(BrSalsem-2H)(dppz) VO(BrSalsem-2H)(bipy) Nifurtimox

13 73 19 84 6

would have had lower mobility. Since the gel was run in the absence of ethidium bromide (or other intercalating compounds), it is possible to evaluate the effect that incubation with the complexes has on DNA tertiary structure. All the vanadium complexes introduce conformational changes in DNA following the same pattern but differing in the extent of the effect. The interaction of these complexes with DNA gives rise to forms with lower mobility than the supercoiled form that could be attributed to the nicked and linear forms. Complex 4, VO(BrSalsem-2H)(bipy), produced only for the highest assayed dose a unique new DNA species, in contrast the other three complexes produced two new DNA forms. Complexes 1 and 2, VO(Salsem-2H)(dppz) and VO(Salsem2H)(bipy), showed this effect even at the lowest dose assayed. Fig. 4 shows that the effect of 1, VO(Salsem-2H)(dppz), on the production of topological isomers of DNA is dose dependent. No appreciable effect of complex 3 on plasmidic DNA is observed at the lowest dose assayed (ri = 1). Nevertheless, incubation with the higher doses (ri = 3 and 6) shows the faint formation of the two reduced mobility DNA species.

615

J. Benítez et al. / Journal of Inorganic Biochemistry 103 (2009) 609–616

ri

control

1

3

6

VO(Salsem-2H)(bipy)

1

3

6

1

3

6

VO(BrSalsem-2H)(dppz) VO(Salsem-2H)(dppz)

1

3

6

VO(BrSalsem-2H)(bipy)

Fig. 4. Plasmid DNA interactions with vanadium complexes. The different complexes used are indicated at the bottom of each panel. All the reactions were incubated in 10 mM Tris.HCl pH 7.4 using 300 ng of pBSK II Bluescript plasmid (Stratagene) in a final volume of 20 lL for 24 h at 37 °C. Control Plasmid incubated in the same conditions in the absence of added complex. The ri values (mol of complex/mol of base pair) are indicated at the top of each panel. Electrophoresis was carried out in the absence of ethidium bromide. Topological isoforms are indicated by arrows.

t=0

VO(Salsem-2H)(bipy)

VO(Salsem-2H)(dppz)

Fig. 5. Time course of DNA interactions with vanadium complexes. The complexes indicated at the bottom of each panel (ri = 3) were incubated in 10 mM Tris.HCl pH 7.4 using 300 ng of pBSK II Bluescript plasmid (Stratagene) in a final volume of 20 lL for different times: 0, 0.5, 1, 2, 2.5, 3, 4 and 5 h. Electrophoresis was carried out in the absence of ethidium bromide. Topological isoforms are indicated by arrows.

The effect of the complexes on DNA migration could be attributed to formation of linear and nicked DNA forms or even conformational modifications such as loss of negative supercoils in the tertiary structure of DNA. Further competition studies in the presence of intercalators and non-intercalator minor groove binders would solve this issue [47]. Since complexes 1 and 2 showed the sharpest effect we selected these metal complexes to study the interaction in more detail. The time course of conformational changes introduced by 1 and 2 on supercoiled DNA is shown in Fig. 5. Conformational changes are already evident after 30 min of incubation for complex 1. Instead, for complex 2 the topological isoform produced is only hardly seen after 2 h. The intensity of the shifted bands produced by complex 1 increases with time, there is also a concomitantly decrease in the intensity of the supercoiled DNA band almost reaching a plateau after 5 h. 4. Conclusions Novel vanadyl complexes including potential DNA intercalating ligands in their coordination sphere showed in vitro anti-trypanosomal activity. All of them showed interaction with DNA, suggesting that this biomolecule could be one of their probable targets in the parasite. Further work is in progress in order to get more potent and stable related vanadyl compounds and to better identify the active species in solution. Acknowledgments

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]

[11]

[12] [13] [14] [15]

[16]

[17]

[18]

[19] [20]

The authors are grateful to the Fundo Europeu para o Desenvolvimento Regional, Fundação para a Ciência e Tecnologia, the POCI 2010 Programme (Project PPCDT/56949/QUI/2004). I. Tomaz thanks the grant SFRH/BPD/34695/2007.

[21]

, . D. Engels, L. Savioli, Trends Parasitol. 22 (2006) 363–366. J. Urbina, Expert Opin. Ther. Patents 13 (2003) 661–669. H. Cerecetto, M. González, Curr. Top. Med. Chem. 2 (2002) 1185–1190. R.L. Krauth-Siegel, H. Bauer, R.H. Schirmer, Angew. Chem. Int. Ed. 44 (2005) 690–715. S. Croft, M. Barret, J. Urbina, Trends Parasitol. 21 (2005) 508–512. M. Ceaser, The Lancet Infect. Dis. 5 (2005) 470–471. Y. Yamagata, J. Nakagawa, Adv. Parasitol. 61 (2006) 129–165. C. Zhang, S. Lippard, Curr. Opin. Chem. Biol. 7 (2003) 481–489. N.P. Farrell, Metal complexes as drugs and chemotherapeutic agents, in: J.M. McCleverty, T.J. Meyer (Eds.), Comprehensive Coordination Chemistry II, vol. 9, Elsevier, 2003, pp. 809–840. R.A. Sánchez-Delgado, A. Anzellotti, L. Suárez, in: H. Sigel, A. Sigel (Eds.), Metal Ions in Biological Systems, vol. 41, Marcel Dekker, New York, 2004, pp. 379– 419. R.A. Sánchez-Delgado, A. Anzellotti, Mini-Rev. Med. Chem. 4 (2004) 23– 30. R. Morphy, Z. Rankovic, J. Med. Chem. 48 (2005) 6523–6543. L. Otero, P. Noblia, D. Gambino, H. Cerecetto, M. González, R. Di Maio, J. Ellena, O.E. Piro, Inorg. Chim. Acta 344 (2003) 85–94. E. Cabrera, H. Cerecetto, M. González, D. Gambino, P. Noblía, L. Otero, B. Parajón-Costa, A. Anzellotti, R. Sánchez-Delgado, A. Azqueta, A. López de Ceráin, A. Monge, Eur. J. Med. Chem. 39 (2004) 377–382. L. Otero, G. Aguirre, L. Boiani, M. González, A. Denicola, C. Rigol, C. Olea-Azar, J.D. Maya, A. Morello, D. Gambino, H. Cerecetto, Eur. J. Med. Chem. 41 (11) (2006) 1231–1239. L. Otero, M. Vieites, L. Boiani, A. Denicola, C. Rigol, L. Opazo, C. Olea-Azar, J.D. Maya, A. Morello, R. Luise Krauth-Siegel, O.E. Piro, E. Castellano, M. González, D. Gambino, H. Cerecetto, J. Med. Chem. 49 (2006) 3322–3331. C. Urquiola, M. Vieites, G. Aguirre, A. Marín, B. Solano, G. Arrambide, M.L. Lavaggi, M.H. Torre, M. González, A. Monge, D. Gambino, H. Cerecetto, Bioorg. Med. Chem. 14 (2006) 5503–5509. L. Otero, P. Smircich, M. Vieites, M. Ciganda, P. Cardoso Severino, H. Terenzi, H. Cerecetto, D. Gambino, B. Garat, J. Inorg. Biochem. 101 (2007) 74–79. L. Otero, C. Folch, G. Barriga, C. Rigol, L. Opazo, M. Vieites, D. Gambino, H. Cerecetto, E. Norambuena, C. Olea-Azar, Spectrochim. Acta Part A 70 (2008) 519–523. M. Vieites, L. Otero, D. Santos, D. Gajardo, J. Toloza, R. Figueroa, E. Norambuena, C. Olea-Azar, G. Aguirre, H. Cerecetto, M. González, A. Morello, J.D. Maya, B. Garat, D. Gambino, J. Inorg. Biochem. 102 (2008) 1033–1043.

616

J. Benítez et al. / Journal of Inorganic Biochemistry 103 (2009) 609–616

[22] M. Vieites, P. Smircich, B. Parajón-Costa, J. Rodríguez, V. Galaz, C. Olea-Azar, L. Otero, G. Aguirre, H. Cerecetto, M. González, A. Gómez-Barrio, B. Garat, D. Gambino, J. Biol. Inorg. Chem. 13 (5) (2008) 723–735. [23] K. Kinnamon, E.A. Steck, E.S. Rane, Antimicrob. Agents Chemother. 15 (2) (1979) 57–160. [24] M. Navarro, E.J. Cisneros-Fajardo, M. Fernandez-Mestre, D. Arrieche, E. Marchan, J. Inorg. Biochem. 97 (2003) 364–369. [25] M. Navarro, C. Hernández, I. Colmenares, P. Hernández, M. Fernández, A. Sierraalta, E. Marchán, J. Inorg. Biochem. 101 (2007) 111–116. [26] H. Beraldo, D. Gambino, Mini-Rev. Med. Chem. 4 (2004) 31–40. [27] P. Noblía, M. Vieites, B. Parajón-Costa, E.J. Baran, H. Cerecetto, P. Draper, M. González, O.E. Piro, E.E. Castellano, A. Azqueta, A. López, A. Monge-Vega, D. Gambino, J. Inorg. Biochem. 99 (2) (2005) 443–451. [28] P. Noblía, E.J. Baran, L. Otero, P. Draper, H. Cerecetto, M. González, O.E. Piro, E.E. Castellano, T. Inohara, Y. Adachi, H. Sakurai, D. Gambino, Eur. J. Inorg. Chem. 2 (2004) 322–328. [29] J.E. Dickeson, L.A. Summers, Aust. J. Chem. 23 (1970) 1023–1027. [30] W.J. Geary, Coord. Chem. Rev. 7 (1971) 81–91. [31] F.M. Ausubel, R. Brent, R.E. Kingston, D.D. Moore, J.G. Seidman, J.A. Smith, K. Struhl (Eds.), Short Protocols in Molecular Biology: A Compendium of Methods from Current Protocols in Molecular Biology, Wiley, New York, 1999. [32] S. Nica, M. Rudolph, H. Gorls, W. Plass, Inorg. Chim. Acta 360 (2007) 1743– 1752. [33] T. Ghosh, B. Mondal, M. Sutradhar, G. Mukherjee, M.G.B. Drew, Inorg. Chim. Acta 360 (2007) 1753–1761.

[34] D. Lin-Vien, N.B. Colthup, W.G. Fateley, J.G. Grasselli, The Handbook of Infrared and Raman Characteristic Frequencies of Organic Molecules, Academic Press, Boston, 1999. [35] T.S. Smith II, R. LoBrutto, V.L. Pecoraro, Coord. Chem. Rev. 228 (2002) 1–18. [36] K. Wuthrich, Helv. Chim. Acta 48 (1965) 1012–1017. [37] N.D. Chasteen, in: J. Lawrence, L. Berliner, J. Reuben (Eds.), Biological Magnetic Resonance, vol. 3, Plenum Press, New York, 1981, pp. 53–119. [38] P. Buglyó, J. Costa Pessoa, A.I. Tomaz, D. Sanna, G. Micera, E. Garribba, FIGIPAS 9, July 2007, Vienna, Austria. [39] J. Costa Pessoa, I. Cavaco, I. Correia, I. Tomaz, M.T. Duarte, R.D. Gillard, R.T. Henriques, F.J. Higes, C. Madeira, Inorg. Chim. Acta 293 (1999) 1–11. [40] J. Costa Pessoa, M.J. Calhorda, I. Cavaco, I. Correia, M.T. Duarte, V. Felix, R.T. Henriques, M.F.M. Piedade, I. Tomaz, J. Chem. Soc. Dalton Trans. (2002) 4407– 4415. [41] L. Casella, M. Gullotti, A. Pintar, S. Colonna, A. Manfredi, Inorg. Chim. Acta 144 (1988) 89–97. [42] A. Rockenbauer, L. Korecz, Appl. Magn. Reson. 10 (1996) 29–34. [43] E.B. Seena, N. Mathew, M. Kuriakose, M.R.P. Kurup, Polyhedron 27 (2008) 1455–1462. [44] M.R. Maurya, S. Khurana, C. Schulzke, D. Rehder, Eur. J. Inorg. Chem. (2001) 779–788. [45] M.R. Maurya, S. Agarwal, C. Bader, M. Ebel, D. Rehder, Dalton Trans. (2005) 537–544. [46] M.R. Maurya, S. Khurana, W. Zhang, D. Rehder, J. Chem. Soc. Dalton Trans. (2002) 3015–3023. [47] K. Storl, G. Burckhardt, J.W. Lown, C. Zimmer, FEBS Lett. 334 (1993) 49–54.