Advertisement

JBIC Journal of Biological Inorganic Chemistry

, Volume 23, Issue 7, pp 1165–1183 | Cite as

Efficient copper-based DNA cleavers from carboxylate benzimidazole ligands

  • Víctor A. Barrera-Guzmán
  • Edgar O. Rodríguez-Hernández
  • Naytzé Ortíz-Pastrana
  • Ricardo Domínguez-González
  • Ana B. Caballero
  • Patrick Gamez
  • Norah Barba-Behrens
Original Paper
Part of the following topical collections:
  1. Alison Butler: Papers in Celebration of Her 2018 ACS Alfred Bader Award in Bioorganic or Bioinorganic Chemistry

Abstract

Four copper(II) coordination compounds from 2-benzimidazole propionic acid (Hbzpr) and 4-(benzimidazol-2-yl)-3-thiobutanoic acid (Hbztb) were synthesized and fully characterized by elemental analyses, electronic spectroscopy, FT-IR and mass spectrometry. The molecular structure for the four complexes was confirmed by single-crystal X-ray crystallography. The DNA-interacting properties of the two trinuclear and two mononuclear compounds were investigated using different spectroscopic techniques including absorption titration experiments, fluorescence spectroscopy and circular dichroism spectroscopy. Trinuclear [Cu3(bzpr)4(H2O)2](NO3)2·3H2O·CH3OH (2) and [Cu3(bzpr)4Cl2]·3H2O (3) bind to DNA through non-intercalative interactions, while for mononuclear [Cu(bzpr)2(H2O)]·2H2O (1) and [Cu(bztb)2]·2H2O (4), at minor concentrations in relation to the DNA, a groove binding interaction is favored, while at higher concentrations an intercalative mode is preferred. The nuclease properties of all complexes were studied by gel electrophoresis, which showed that they were able to cleave supercoiled plasmid DNA (form I) to the nicked form (form II). Compound 4 is even capable of generating linear form III (resulting from double-strand cleavage). The proposed mechanism of action involves an oxidative pathway (Fenton-type reaction), which produces harmful reactive species, like hydroxyl radicals.

Graphical abstract

Keywords

Copper(II) complexes 2-Carboxylate benzimidazoles ct-DNA pBr322 DNA DNA-cleaving properties 

Introduction

Metal coordination compounds capable of binding to DNA and/or showing nuclease properties have received much attention from the scientific community. Transition-metal complexes have shown enormous potential in this field, not only because of their distinctive spectral and electrochemical properties, but also because, by changing the ligand, one may adjust their affinity to DNA. Two DNA-binding modes are possible, namely through covalent or non-covalent interactions (e.g., via intercalation). Since the discovery of cisplatin [Pt(NH3)2Cl2], DNA has become an important target for the design of metal-based therapeutic agents [1]; platinum-based drugs are currently used for the treatment of various cancers, such as testicular, ovarian, head, neck and small cell lung cancers [2]. It is known that the molecular mechanism of cisplatin involves its coordination to DNA bases, principally at the N7-position of guanines, with such binding leading to apoptosis [3].

Cisplatin treatment usually causes unpleasant side effects. Moreover, cell resistance to the drug may occur. Therefore, the design and development of new metal coordination compounds with higher efficiencies, lower toxicities and target-specific properties are essential [4]. Copper is an important trace element in the human body; for instance, some vital enzymatic functions involve Cu(II)/Cu(I) redox processes [5], which have been studied by numerous researchers worldwide [6, 7, 8, 9]. Furthermore, a rich diversity of coordination structures is possible with copper that can adopt diverse geometries. A number of copper(II) compounds have been developed, which show interesting DNA-binding and cleavage properties, as well as remarkable antibacterial and cytotoxic activities [10, 11, 12, 13, 14, 15, 16].

Copper(II) can bind to electron-pair donors such as O, N and S, which can be found in bio-macromolecules such as DNA, RNA and a wide variety of enzymes and proteins [17, 18, 19]. As a matter of fact, numerous examples of copper complexes able to interact with DNA have been reported, i.e., copper(II) compounds from chelating ligands such as phenanthroline, porphyrins, Schiff bases and (thio)semicarbazones [20]. On the other hand, copper coordination compounds with macrocyclic polyaza ligands induce B → Z transitions of poly d(GC)DNA [21, 22]. Some other copper complexes have been reported that show binding properties toward active molecules such as enzymes and peptides [23, 24].

Polynuclear copper coordination compounds have recently been described, which induce DNA damage and are cytotoxic toward HeLa cells through apoptosis [25]. Some pyridine-based mononuclear, dinuclear and trinuclear copper compounds have been developed, whose DNA-cleaving efficiency depends on the number of metal centers present in the compound [26, 27].

Benzimidazoles and their acid derivatives are important bioactive compounds that are widely used for their potential biocide or antineoplastic properties [28, 29, 30, 31, 32, 33, 34]. Among these derivatives, some 2-substituted benzimidazoles, like 2-nitronyl nitroxide benzimidazole [35] and 2-mercaptobenzimidazole [36], may stabilize free radicals. Such N, O, S-containing tridentate ligands can bind different transition-metal ions, generating compounds of various geometries and coordination numbers, and with interesting properties. For instance, mononuclear cobalt(II) and nickel(II) compounds from 4-(benzimidazol-2-yl)-3-thiobutanoic acid have been reported that exhibit weak antiferromagnetic couplings, arising from intermolecular hydrogen bonding interactions [37].

In the present work, the biological activities of four copper(II) coordination compounds, obtained from 2-benzimidazole propionic acid (Hbzpr) and 4-(benzimidazol-2-yl)-3-thiobutanoic acid (Hbztb) (Scheme 1), were investigated. The interaction of the trinuclear and mononuclear compounds [Cu(bzpr)2(H2O)]·2H2O 1, [Cu3(bzpr)4(H2O)2](NO3)2·3H2O·CH3OH 2, [Cu3(bzpr)4Cl2]·3H2O 3 and [Cu(bztb)2]·2H2O 4 with calf thymus DNA (ct-DNA) was examined using several techniques, namely electronic absorption titration, fluorescence spectroscopy and circular dichroism spectroscopy. Their nuclease activities were also investigated by gel electrophoresis with pBR322 plasmid DNA.
Scheme 1

a Structure of 2-benzimidazole propionic acid; b structure of 4-(benzimidazol-2-yl)-3-thiobutanoic acid

Experimental

Materials and physical measurements

For the synthesis of the coordination compounds, the ligands 2-benzimidazole propionic acid (Hbzpr) and 4-(benzimidazol-2-yl)-3-thiobutanoic acid (Hbztb) were obtained from Sigma-Aldrich and used as received. The copper(II) nitrate and copper(II) chloride (Cu(NO3)2·2.5H2O, CuCl2·2H2O), and the methanol were obtained from Baker and used without purification.

For the biological measurements, the following compounds were acquired from Sigma-Aldrich: sodium cacodylate, NaCl, Tris–borate–EDTA (10× concentration), ethidium bromide and Hoechst 33258, and used without further purification. The pBR322 plasmid DNA (0.5 µg/0.5 µL−1) was obtained from Thermo Scientific and SYBR safe DNA gel stain from Invitrogen.

For the coordination compounds, FT-IR spectra were recorded in the range 4000–400 cm−1 at 298 K, using a Perkin Elmer FT-IR Spectrum 400 spectrophotometer equipped with a universal ATR sampling accessory. Electronic spectra of solid powdered crystalline samples were measured at 298 K, over the range 40,000–5000 cm−1, by the diffuse reflectance method on a Cary-5000 Varian spectrophotometer. Elemental analyses were carried out with a Fisons EA 1108 analyzer. Magnetic susceptibility measurements at room temperature of powered samples were recorded on a Johnson–Matthey DG8 5Hj balance, using the Gouy method.

Synthesis

2-Benzimidazole propionic acid (Hbzpr) compounds

General procedure

The coordination compounds 13 were synthesized in situ by first deprotonating the ligand Hbzpr in water (15 mL) with NaOH (1:1 ratio) and slightly heating the mixture until complete dissolution of Hbzpr, (final pH of 7). A solution of the metal salt in 15 mL of methanol was subsequently added to the ligand solution, and the resulting reaction mixture was slightly heated for 2 h. The mixture was then left unperturbed for the slow evaporation of the solvent; the crystals obtained were filtered under reduced pressure and washed with methanol and water.

[Cu(bzpr)2(H2O)]·2H2O (1)

A methanol solution of copper(II) nitrate (Cu(NO3)2·2.5H2O) (0.2325 g, 1 mmol) was added to an aqueous solution of the deprotonated Hbzpr (0.1902 g, 1 mmol). After 2 h of heating, the blue solution was allowed to stand unperturbed at low temperature for 2 weeks. Blue crystals of 1 were obtained that were suitable for X-ray diffraction analysis. (Yield 0.2035 g, 41%.) Anal. Calc. for C20H24N4O7Cu: C, 48.43; H, 4.88; N, 11.30. Found: C, 47.82; H, 3.34; N, 11.07%. IR: ν(N–H) 3182 cm−1, νas(COO) 1570 cm−1, νs(COO) 1454 cm−1. EM ESI negative (m/z): 441 ([Cu(bzpr)2]), 379 ([(bzpr)2]), 189 ([(bzpr)]). µeff = 2.74 B.M.

[Cu3(bzpr)4(H2O)2](NO3)2·3H2O·CH3OH (2)

A methanol solution of copper(II) nitrate (Cu(NO3)2.2.5H2O) (0.0930 g, 0.4 mmol) was added to an aqueous solution of the deprotonated Hbzpr (0.0400 g, 0.2 mmol). After 2 h of heating, the blue solution was allowed to stand unperturbed at low temperature for 2 weeks. Blue crystals of 2, suitable for X-ray diffraction analysis, were obtained. (Yield 0.0801 g, 50%.) Anal. Calc. for C41H50N10O20Cu3: C, 41.25; H, 4.22; N, 11.74. Found: C, 40.95; H, 3.64; N, 12.05%. IR: ν(N–H) = 3464 cm−1, ν(C = C) = 1610 cm−1, νas(COO) 1454 cm−1, νs(COO) 1344 cm−1. EM Maldi-TOF (m/z): 569 ([Cu3(bzpr)2]), 505 ([Cu2(bzpr)2]), 443 ([Cu(bzpr)2]). µeff: 3.25 B.M.

[Cu3(bzpr)4Cl2]·3H2O (3)

This compound was prepared as above using copper(II) chloride (CuCl2·2H2O) (0.1704 g, 1 mmol) and deprotonated 2-benzimidazole propionic acid Hbzpr (0.1902 g, 1 mmol). After slightly heating for 2 h, the green solution was allowed to stand unperturbed at low temperature for 2 weeks. Green crystals of 3, suitable for X-ray diffraction analysis were obtained. (Yield 0.1563 g, 40%.) Anal. Calc. for C40H38N8O8.80Cl2Cu3: C, 46.50; H, 3.70; N, 10.84. Found: C, 46.31; H, 3.19; N, 10.83%. IR: ν(N–H) = 3123 cm−1, ν(C=C) = 1620 cm−1, νas(COO) = 1570 cm−1, νs(COO) = 1453 cm−1. EM Maldi-TOF (m/z): 505 ([Cu2(bzpr)2]), 443 ([Cu(bzpr)2]), 191 ([(bzpr) + 2H]. µeff = 3.60 B.M.

4-(Benzimidazol-2-yl)-3-thiobutanoic acid (Hbztb) compound

[Cu(bztb)2]·2H2O (4)

A solution of copper(II) nitrate (Cu(NO3)2·2.5H2O) (0.0465 g, 0.2 mmol) in methanol (5 mL) was added to a solution of 4-(benzimidazol-2-yl-9-3-thiobutanoic acid (Hbztb) (0.0888 g, 0.4 mmol). The resulting green solution was heated under reflux for 4 h. The green precipitate formed was isolated and recrystallized in methanol containing 1% of water. The solution was allowed to stand unperturbed at low temperature for 4 weeks. Green crystals of 4, suitable for X-ray diffraction analysis, were obtained. (Yield 0.054 g, 49.8%.) Anal. Calc. for C20H22N4O6S2Cu: C, 44.31; H, 4.09; N, 10.34. Found: C, 44.69; H, 3.66; N, 10.28%. IR: ν(N–H) = 3061 cm−1, ν(C=O) = 1656 cm−1, ν(C=C) = 1624 cm−1, ν(C=N) = 1572 cm−1, νas(COO) = 1452 cm−1, νs(COO) = 1293 cm−1, ν(C=S) = 744 cm−1. EM ESI negative (m/z): 505 ([Cu(bztb)2]), 443 ([(bztb)2]), 221 ([(bztb)]); µeff = 2.36 B.M.

Single-crystal X-ray crystallography

Single-crystal X-ray diffraction data for 1, 2 and 4 were collected using standard procedures on an Oxford Diffraction Gemini “A” instrument with CCD area detector using graphite-monochromated MoKα radiation at 293 K. Intensities were measured using φ + ω scans. All structures were solved using direct methods, with SHELX-97, and the refinement (based on F2 of all data) was performed by full-matrix least-squares techniques with Crystals 12.84. The diffraction intensity patterns from a single crystal of compound 3 were collected on a Bruker D8 Venture Geometry diffractometer 208039-3, equipped with a CCD-detector and using graphite-monochromated MoKα (λ = 0.71073 Å) radiation source. APEX2 v2012.10.0 (Bruker 2012) package was used for data collection and data integration. Absorption corrections were applied using analytical procedure. The structures were solved by direct methods using the package SHELXT 2014/5 and refined with an anisotropic approach for non-hydrogen atoms using the SHELXL-2016/6 program. All hydrogen atoms attached to C atoms were positioned geometrically as riding on their parent atoms, with C–H = 0.93–0.99 A and Uiso(H) =−1.2 Ueq(C) for aromatic and methylene groups [38, 39]. For compound 2, the unit cell contains eight methanol molecules which have been treated as a diffuse contribution to the overall scattering without specific atom positions by SQUEEZE/PLATON [40].

The crystallographic data and structure refinement are given as Electronic Supplementary Information (Table S1).

Supplementary crystallographic data for this paper have been deposited with the CCDC: 1841514, 1841515 and 1841516; these data can be obtained free of charge via http://www.ccdc.cam.ac.uk, or by contacting the Cambridge Crystallographic Data Centre, 12, Union Road, Cambridge CB2 IEZ, UK; fax: +44 1223 336033.

Electronic absorption titrations

The electronic absorption titrations (UV/vis spectra) were recorded on a Varian Cary 100 scan spectrophotometer at room temperature, employing 1 cm pathlength cuvette. Absorption titration experiments were performed in cacodylate–NaCl buffer (1 mM cacodylate/20 mM NaCl, pH 7.1). The metal complexes were dissolved in DMSO so that the final titration solutions contained 5% DMSO and [Complex] = 25 µM. The concentration of ct-DNA was determined spectrophotometrically at 260 nm by using the molar extinction coefficient value of 6600 M−1 cm−1 [41]. All samples (complex and DNA solutions) were incubated at 37 °C for 1 h. Absorption titration experiments were carried out using increasing concentrations of ct-DNA (0–50 µM), while keeping the complex concentration constant. Corrections were made using the solution of free ct-DNA (i.e., without copper complex). Before recording each spectrum, the samples were left for 5 min to achieve equilibrium.

To compare the DNA-binding strength of the different complexes, their intrinsic binding constant Kb was calculated using Eq. (1) [42]:
$$\frac{{\left[ {\text{DNA}} \right]}}{{\varepsilon_{\text{a}} - \varepsilon_{\text{f}} }} = \frac{{\left[ {\text{DNA}} \right]}}{{\varepsilon_{0} - \varepsilon_{\text{f}} }} + \frac{1}{{K_{\text{b}} \left( {\varepsilon_{0} - \varepsilon_{\text{f}} } \right)}},$$
(1)
where [DNA] is the concentration of ct-DNA in base pairs, εa is the extinction coefficient of the formed ct-DNA–Complex (Aobs/[Complex]), εf is the extinction coefficient of the free complex in solution, and ε0 is the extinction coefficient for the compound in the fully DNA-bound form. The intrinsic binding constants Kb were determined from the corresponding plots of [DNA]/(εaεf) vs [DNA].

Fluorescence spectroscopy

Emission intensity measurements were carried out using a spectrofluorophotometer. The ct-DNA (15 μM) was pretreated with ethidium bromide (EB, 75 μM) or Hoechst 33258 (2 μM,) and incubated for 30 min at 37 °C. The complex was added to this mixture in [Complex]/[DNA] ratios ranging 0.0–1.7, and the final samples were incubated for 1 h before the measurements. After incubation, the emission intensities were measured from 530 to 800 nm with an excitation wavelength at 514 nm (EB). For the measurements with Hoechst 33258, the emission intensities were measured from 380 to 650 nm applying an excitation wavelength of 350 nm. In both cases, a KONTRON SFM 25 spectrofluorometer was used. The quenching data were analyzed applying the linear Stern–Volmer equation [43]:
$$\frac{{I_{0} }}{I} = 1 + K_{\text{sv}} \left[ {\text{Complex}} \right],$$
(2)
where I0 and I represent the fluorescence intensities of the DNA–EB/Hoechst complex in the absence and presence of quencher, respectively. Ksv is the linear Stern–Volmer quenching constant and [Complex] is the quencher concentration. In the plot of I0/I versus [Complex], Ksv is given by the ratio of the slope to the intercept.

Circular dichroism (CD) spectroscopy

For the circular dichroism titrations, a A JASCO J-815 circular dichroism spectropolarimeter was used to acquire the CD spectra, at room temperature with a scanning speed of 200 nm min−1 and a 0.5 cm-pathlength cuvette. Solutions of ct-DNA (100 μM) with [Complex]/[DNA] ratios of 0.0, 0.2, 0.6 and 1.0 were incubated at 37 °C for 1 h. The CD spectra of these solutions were then recorded from 200 to 400 nm using a quartz cuvette with an optical path length of 5 mm and a scanning rate of 200 nm min−1. The data were collected with a time constant of 1 s and a spectral bandwidth of 1.0 nm. The background signal due to the buffer was subtracted.

Gel electrophoresis

The DNA-cleaving properties of the complexes were investigated in the presence of ascorbic acid (H2ASC; reducing agent) using agarose gel electrophoresis. For the pBR322 plasmid DNA (0.5 μg μL−1), the final concentration in the gel tests was 15 µM in terms of nucleobase pairs. This plasmid was treated with the different copper(II) compounds (in the concentration range of 2.5–50 µM). 20 μL samples in cacodylate–NaCl (1 mM cacodylate/20 mM NaCl buffer, pH 7.25) were incubated at 37 °C for 1 h. Next, a loading buffer (4 μL) consisting of 0.25% w/v xylene cyanol (5 mM) and 30% v/v glycerol was added. The mixture was subjected to gel electrophoresis on 1% agarose in TBE buffer (40 mM Tris–borate and 1 mM EDTA) at 100 V for 1 h. After the electrophoresis, the gel was placed in a container containing 150 mL of TBE buffer and 15 μL of display buffer (SYBR Safe) and was agitated overnight. Subsequently, images were acquired using a Gel Doc EZ Imager instrument. The intensities of supercoiled DNA (Form I) were corrected, multiplying them by 1.22, since intercalation in Form I is weaker than that in nicked (Form II) or linear (Form III) DNA [44]. The DNA-cleaving efficiency was obtained from modified Eq. (3) [45] (see below). The percentage of DNA-cleaving activity derived from the volume (= band intensity × area) of supercoiled DNA (Form I) was obtained as follows:
$$\begin{aligned} \% {\text{ DNA}} - {\text{cleaving activity}} & = \, \left\{ {\left[ {\left( {{\text{volume of SC}} - {\text{DNA}}} \right)_{\text{control}} } \right.} \right. \\ & - \left. {\left( {v{\text{olume of SC}} - {\text{DNA}}} \right)_{\text{sample}} } \right]/\left. {\left( {{\text{volume of SC}} - {\text{DNA}}} \right)_{\text{control}} } \right\} \, \times \, 100. \\ \end{aligned}$$
(3)

Results and discussion

Spectroscopic characterization

The diffuse reflectance electronic spectra of compounds 1–4 characterize copper(II) ions in different coordination geometries [46] (Table 1). Mononuclear [Cu(bzpr)2(H2O)]·2H2O 1 presents a band at 14,429 cm−1, corresponding to a transition for a copper(II) ion in a square-pyramidal environment. The spectra of the trinuclear compounds [Cu3(bzpr)4(H2O)2](NO3)2·3H2O·CH3OH 2 and [Cu3(bzpr)4Cl2]·3H2O 3 exhibit a band with a shoulder at 14,601 cm−1/11,392 cm−1 and 14,314 cm−1/11,906 cm−1, respectively. These bands are in the region expected for electronic transitions of a copper(II) ion in an square-pyramidal geometry and an octahedral geometry. The spectrum of the coordination compound [Cu(bztb)2]·2H2O 4 shows a band at ν1 = 13,240 cm−1, for a copper(II) ion in a cis-distorted octahedral geometry.
Table 1

UV–Vis–NIR (reflectance) data and magnetic moments (μeff, BM) for compounds 1–4

Compound

ν1(cm−1)

µeff (BM)

[Cu(bzpr)2(H2O)]·2H2O 1

14,429

2.74

[Cu3(bzpr)4(H2O)2](NO3)2·H2O 2

14,601 (11,392) sh

3.25

[Cu3(bzpr)4Cl2]·3H2O 3

14,314 (11,906) sh

3.60

[Cu(bztb)2]·2H2O 4

13,240

2.36

The magnetic moment was determined for each complex. The mononuclear compounds 1 and 4 have a µeff of, respectively, 2.74 and 2.36 B.M. (Table 1), slightly above the range expected for a copper(II) ion. The magnetic moment, per molecule, for the trinuclear compounds 2 and 3 are indicative of antiferromagnetic coupling between the copper(II) ions in these molecules (Table 1) [47].

The IR spectra of the coordination compounds exhibit absorption bands from the respective ligands, assigned to ν(N–H), ν(C=C), ν(C=N), νas(COO), νs(COO) and ν(C=S) vibrations, which shift upon coordination to the metal ion (see Table 2). The absorption data suggest that the deprotonated ligand (bzpr) acts as a bidentate ligand, through a carboxylate oxygen atom and the benzimidazolic nitrogen atom. On the other hand Hbztb, upon deprotonation on the reaction synthesis, behaves as a tridentate ligand with its sulfur atom coordinated to the copper(II) ion. For the mononuclear compounds νas(COO) is observed in the region of 1570 cm−1, while νs(COO) is at 1350 cm−1 with a ∆ν (νasνs) = 220 cm−1, for compound 1 and for compound 4 this band is at 1293 cm−1 with a ∆ν = 279 cm−1; these ∆ν values are in accordance with a monocoordinated carboxylate. The trinuclear complexes present a νas(COO) at 1570 cm−1; however for νs(COO), three bands were observed (Table 2), indicative of different coordination modes for the carboxylate group, bridging, monodentate and bidentate [48].
Table 2

Selected IR bands for compounds 14 (cm−1)

Vibration (cm−1)

ν(N–H)

ν(C=C)

νas(COO)

νs(COO)

ν

ν(C=S)

bzpr (ligand)

3174

1635

1567

1411

1

3182

1570

1350

220

2

3464

1610

1570

1454, 1416, 1344

3

3123

1620

1569

1488, 1451, 1378

Hbztb (ligand)

3104

1621

1536

1446

755

4

3061

1624

1572

1293

279

744

X-ray diffraction analysis

2-Benzimidazole propionic acid (Hbzpr) coordination compounds

The ligand Hbzpr shows great coordination versatility with its carboxylate group acting as a bridging dinucleating unit through its two oxygen atoms or a μ2-bridging dinucleating unit. Hbzpr can also generate a seven-membered chelate ring involving the coordination of an oxygen and a nitrogen atom to the metal center.

[Cu(bzpr)2(H2O)]·2H2O 1
The mononuclear compound 1 contains two deprotonated bzpr ligands, coordinated in a bidentate fashion, and one water molecule is bound at the apical position of a square pyramid (Fig. 1a). Hence, the benzimidazolic nitrogen atoms and carboxylate oxygen atoms, N23, O34, N3 and O14, occupy the equatorial plane, and the water molecule (O1) is at the apical position (Fig. 1b). Selected bond distances and angles are listed in Table 3.
Fig. 1

a

ORTEP diagram of [Cu(bzpr)2(H2O)]·2H2O 1; displacement ellipsoids are drawn at 30% probability level; b illustration showing the plane formed by N23, O34, N3 and O14

Table 3

Selected bond distances (Å) and angles (°) for [Cu(bzpr)2(H2O)]·2H2O 1

Bond lengths (Å)

Angles (°)

Cu1−O34

1.9676(13)

O34−Cu1−N23

93.35(6)

Cu1−N23

1.9678(16)

O34−Cu1−O14

177.14(6)

Cu1−O14

1.9690(14)

N23−Cu1−O14

87.10(6)

Cu1−N3

1.9783(16)

O34−Cu1−N3

89.68(6)

Cu1−O1

2.3009(16)

N23−Cu1−N3

158.44(6)

  

O14−Cu1−N3

90.93(6)

  

O34−Cu1−O1

86.57(6)

  

N23−Cu1−O1

94.03(6)

  

O14−Cu1−O1

90.59(6)

  

N3−Cu1−O1

107.46(6)

The coordination of the nitrogen and oxygen atoms of the ligands generate seven-membered chelate rings, with angles in the range of 87.10−93.5° (Table 3). The Cu1−Ow1 bond length is 2.3009(16) Å, which is significantly longer than the equatorial ones (Table 3).

The τ value, obtained applying Eq. (4) [49, 50], amounts to 0.31, thus indicating that 1 is in a distorted square-pyramidal environment.
The crystal structure of 1 shows an intricate network of intermolecular interactions. Two hydrogen-bonded water molecules connect two neighboring complexes, through the benzimidazolic nitrogen atom from one molecule and an oxygen atom from the carboxylate group of a second one (N1−H···O3−H···O2−H···O33; Fig. 2 top). The water molecule coordinated to the metal ion is hydrogen bonded to a carboxylate oxygen atom from an adjacent molecule and to a lattice water molecule (O1−H···O2 and O1−H···O34), as shown in Fig. 2, bottom left (Table 4). Finally, π-stacking interactions between two benzimidazolic rings are observed (Fig. 2, bottom right), characterized by a distance of 3.49 Å, which is in the expected range for this type of interaction.
Fig. 2

Intermolecular H bonding and π-stacking interactions observed in the crystal lattice of 1

Table 4

Hydrogen bonding contacts (Å) in the X-ray crystal structure of [Cu(bzpr)2(H2O)]·2H2O 1

D−H···A (Å)

(DHA) angle (°)

N1−H1···O3

2.780(2)

168(2)

O3−H10···O2

2.722(2)

170(3)

O2−H5···O33

2.675(2)

172(3)

O1−H2···O2

2.816(2)

161(3)

O1−H3···O34

2.894(2)

166(3)

[Cu3(bzpr)4(H2O)2](NO3)2·3H2O·CH3OH 2

This coordination compound has three copper(II) centers (Fig. 3a), two of them (Cu1 and Cu3) being in a square-pyramidal geometry (τ value of 0.16) and the third one presenting a distorted octahedral geometry. Each of the pentacoordinated metal ions, Cu1 and Cu3, has a similar coordination sphere, i.e., Cu1 is coordinated by a bidentate ligand (N3 and O13), forming a chelate seven-membered ring, two carboxylate oxygen atoms (O33 and O73) from two other ligands and an apical water molecule (O1). Hexacoordinated Cu2 is coordinated to two bidentate ligands bzpr, each of them forming a chelate ring (N23, O33 and N43, O53), and two carboxylate oxygen atoms from two different ligands (O13 and O74), generating a distorted octahedral geometry (Fig. 3).
Fig. 3

a

ORTEP diagram of [Cu3(bzpr)4(H2O)2](NO3)2·3H2O·CH3OH 2; displacement ellipsoids are drawn at 30% probability level; b square-pyramidal geometries of Cu1 and Cu3 and octahedral geometry of Cu2 with atom numbering scheme

In 2, the carboxylate groups present two distinct types of coordination modes, namely type 1 where only one oxygen atom (O33) bridges two metal centers (Cu1 and Cu2), and type 2, where two oxygen atoms (O13 and O14) bridge two metal centers (Cu2 and Cu3). These two binding modes are shown in Fig. 4.
Fig. 4

Different coordination modes of the carboxylate groups in 2: left, type 1; right, type 2

Selected bond distances and angles are listed in Table 5. It is worth noting that in octahedral Cu2, the bond distances Cu2−O53 (2.539 Å) and Cu2−O33 (2.476 Å) are significantly longer than the Cu2−O74 (1.9925 Å) and Cu2−O13 (1.979 Å) ones, due to Jahn–Teller distortion [51].
Table 5

Selected bond distances (Å) and angles (°) for compound [Cu3(bzpr)4(H2O)2](NO3)2·3H2O·CH3OH 2

Bond lengths (Å)

Angles (°)

Cu1−O13

2.447(3)

O33−Cu1−O73

91.26(8)

Cu1−O33

1.9281(19)

O33−Cu1−N3

166.50(9)

Cu1−O73

1.9415(19)

O73−Cu1−N3

93.46(9)

Cu1−N3

1.975(2)

O33−Cu1−O1

89.58(9)

Cu1−O1

1.964(2)

O73−Cu1−O1

156.54(9)

Cu2−O13

1.979(2)

O1−Cu1−N3

91.10(10)

Cu2−N43

1.990(2)

O13–Cu2–N43

88.75(9)

Cu2−O74

1.9925(19)

O13–Cu2–O74

90.07(8)

Cu2−N23

1.986(2)

N43–Cu2–O74

165.96(9)

Cu2−O33

2.476(3)

O13–Cu2–N23

164.52(9)

Cu3−O53

2.539(5)

N23–Cu2–N43

95.44(10)

Cu3−O14

1.965(2)

N23–Cu2–O74

89.39(9)

Cu3−O53

1.9354(19)

O53–Cu3–O2

89.88(9)

Cu3−O2

1.947(2)

O53–Cu3–O14

90.65(8)

Cu3−N63

1.970(2)

O2–Cu3–O14

165.67(9)

  

O53–Cu3–N63

169.20(9)

  

O2–Cu3–N63

90.64(10)

  

O14–Cu3–N63

91.50(9)

[Cu3(bzpr)4Cl2]·3H2O 3

Compound 3 has four bzpr ligands coordinated to three copper(II) centers, one being in a square-pyramidal environment and the other two exhibiting a distorted octahedral geometry (Fig. 5).
Fig. 5

a ORTEP diagram of compound [Cu3(bzpr)4Cl2]·3H2O 3; displacement ellipsoids are drawn at 30% probability level; b Octahedral geometries of Cu2 and Cu3, and square-pyramidal geometry of Cu1, with atom numbering scheme

The square-pyramidal geometry of Cu1 is characterized by a τ value of 0.16. The square base of the pyramid includes the two atoms from a bzpr ligand, i.e., the benzimidazole nitrogen atom (N3) and the carboxylate oxygen atom (O13), forming a seven-membered chelate ring. Two carboxylate oxygen atoms (O53 and O74) from different ligands complete the remaining positions of the basal plane, and the chloro atom (Cl1) occupies the apical position. Cu2 presents a similar coordination sphere than that of the Cu2 center in 2 (see Figs. 5b and 6), with two ligands forming two chelate rings (N23, O33 and N43, O53), and two carboxylate oxygen atoms (O13 and O73), belonging to different ligands, at the elongated octahedral positions. Octahedral Cu3 shows a different donor set (compared with that of Cu2), formed by a chelating carboxylate-O,O group from one ligand (type 3 binding mode, Fig. 7), a chelating ligand (nitrogen atom N63 and carboxylate oxygen atom O73), a carboxylate oxygen atom (O14) from an additional bzpr ligand and a chloro (Cl2) atom (Fig. 7). The type 3 binding mode of the {O33, O34}-carboxylate group gives rise to a small bite angle of 51.83° (O33−Cu3−O34), the corresponding bond distances being Cu3−O34 of 2.799 Å and Cu3−O33 of 1.933 Å (Table 7).
Fig. 6

Cu1 and Cu2 centers in [Cu3(bzpr)4Cl2]·H2O 3 with the different coordination modes of the carboxylate groups: type 1 (oxygen atom O53) and type 2 (oxygen atoms O74 and O73)

Fig. 7

Cu3 center in [Cu3(bzpr)4Cl2].H2O 3. Type 3 coordination mode of the carboxylate group {O34, O33}

Selected distances and angles for compounds 2 and 3 are listed in Tables 6 (distances) and 7 (angles), for comparison purposes. All distances and angles for 2 and 3 are given as Supplementary Information (Tables S2 and S3).
Table 6

Selected bond distances (Å) for [Cu3(bzpr)4Cl2]·H2O 3, [Cu3(bzpr)4(H2O)2](NO3)2·3H2O·CH3OH 2

Distances (Å) Compound 3

Distances (Å) Compound 2

Cu1−O13

2.554 (2)

Cu1−O13

2.447(3)

Cu1−O53

1.942(5)

Cu1−O33

1.928(19)

Cu1−O74

1.968(4)

Cu1−O73

1.942(19)

Cu1−N3

1.977(5)

Cu1−N3

1.975(2)

Cu1−Cl1

2.256(2)

Cu1−O1

1.964(2)

Cu2−O13

1.966(4)

Cu2−O13

1.979(2)

Cu2−N43

1.994(5)

Cu2−N43

1.990(2)

Cu2−O73

2.004(4)

Cu2−O74

1.993(19)

Cu2−N23

2.005(5)

Cu2−N23

1.986(2)

Cu2−O33

2.460 (5)

Cu2−O33

2.476(3)

Cu2−O53

2.650 (2)

Cu2−O53

2.539(5)

Table 7

Selected angles (°) for [Cu3(bzpr)4Cl2]·H2O 3 and [Cu3(bzpr)4(H2O)2](NO3)2·3H2O·CH3OH 2

Angles (°) Compound 3

Angles (°) Compound 2

O53−Cu1−O74

88.00(19)

O33−Cu1−O73

91.26(8)

O53−Cu1−N3

163.2(2)

O33−Cu1−N3

166.50(9)

O74−Cu1−N3

90.0(2)

O73−Cu1−N3

93.46(9)

O53−Cu1−Cl1

94.42(16)

O33−Cu1−O1

89.58(9)

O74−Cu1−Cl1

153.59(15)

O73−Cu1−O1

156.54(9)

N3−Cu1−Cl1

94.79(17)

O1−Cu1−N3

91.10(10)

O13–Cu2–N43

162.8(2)

O13–Cu2–N43

88.75(9)

O13–Cu2–O73

88.14(17)

O13–Cu2–O74

90.07(8)

N43–Cu2–O73

90.4(2)

N43–Cu2–O74

165.96(9)

O13–Cu2–N23

88.7(2)

O13–Cu2–N23

164.52(9)

N43–Cu2–N23

96.6(2)

N23–Cu2–N43

95.44(10)

O73–Cu2–N23

166.0(2)

N23–Cu2–O74

89.39(9)

Coordination compound of 4-(benzimidazol-2-yl)-3-thiobutanoic acid (Hbztb)

[Cu(bztb)2]·2H2O 4

Mononuclear copper(II) compound 4 has two deprotonated 4-(benzimidazol-2-yl)-3-thiobutanoic ligands (bztb), coordinated in a tridentate fashion, generating a fac-octahedral environment (Fig. 8a). The sulfur atoms (S11 and S31) are bound in trans positions, while the oxygen (O14 and O34) and the nitrogen atoms (N1 and N21) are in a cis arrangement, as observed in a previously reported copper(II) compound [52]. The oxygen and nitrogen atoms are in the equatorial plane (O14, N1, N21, O34), with bond distances in the range of 1.95−2.00 Å (Table 8). The bond distances Cu−S11 and Cu−S31 amount to 2.74 Å (Fig. 8b). The angles S31−Cu1−S11, N1−Cu1−O34 and O14−Cu1−N21 of 161°−169°, and the chelate angles, ranging from 80.17° (N1−Cu1−S11) to 114.71° (N1−Cu1−S31), are indicative of a distorted octahedral geometry.
Fig. 8

a

ORTEP diagram of compound [Cu(bztb)2]·2H2O 4; displacement ellipsoids are drawn at 30% probability level; b illustration showing the equatorial plane O14, N21, N1, O34

Table 8

Selected bond distances (Å) and angles (°) for [Cu(bztb)2]·2H2O 4

Bond lengths (Å)

Angles (°)

Cu1−O14

1.9543(19)

O14−Cu1−N21

169.30(8)

Cu1−N21

1.971(2)

O14−Cu1−N1

90.98(9)

Cu1−N1

1.985(2)

N21−Cu1−N1

92.24(9)

Cu1−O34

2.0017(19)

O14−Cu1−O34

89.88(8)

Cu1−S31

2.7409(8)

N21−Cu1−O34

89.13(9)

Cu1−S11

2.7456(8)

N1−Cu1−O34

167.78(8)

  

O14−Cu1−S31

89.10(6)

  

N21−Cu1−S31

80.28(7)

  

N1−Cu1−S31

114.71(6)

  

O34−Cu1−S3

77.49(5)

  

O14−Cu1−S11

79.64(6)

  

N21−Cu1−S11

110.97(7)

  

N1−Cu1−S11

80.17(6)

  

O34−Cu1−S11

87.99(5)

  

S31−Cu1−S11

161.67(3)

The crystal packing of 4 exhibits a 3D supramolecular framework generated through intermolecular hydrogen bonding interactions involving a lattice water molecule connecting three coordination compounds (Fig. 9). Hence, the water molecule O2 is hydrogen bonded to two carboxylate oxygen atoms from two metal complexes (O2−H···O14 and O2−H···O15), and to a benzimidazolic nitrogen atom belonging to a third molecule of 4 (N3−H···O2) (Table 9).
Fig. 9

Hydrogen bonding interactions between a lattice water molecule and three molecules of 4

Table 9

Hydrogen bonds (Å) in [Cu(bztb)2]·2H2O 4

D−H···A (Å)

(DHA) angle (°)

N3−H1···O3

2.806(3)

168(3)

O2−H···O15

2.760(3)

165(4)

O2−H···O14

2.897(3)

167(4)

DNA-binding studies

DNA is a classical pharmacological target of metal-based anticancer agents; for that reason, the study of the potential interaction of complexes with DNA is of great interest as molecules for medical applications may be discovered. There are four possible ways in which coordination compounds can bind to double-stranded DNA, namely:
  1. (a)

    Intercalation between two adjacent base pairs and perpendicular to the helical axis.

     
  2. (b)

    Outside-edge binding to the sugar-phosphate backbone of the helix through electrostatic interactions.

     
  3. (c)

    Groove binding with functional groups into either the major or minor groove.

     
  4. (d)

    Covalent interaction between DNA and metal complexes, involving nitrogen atoms of nucleobases.

     

Several characterization techniques have been employed to study the interaction of copper(II) compounds 1−4 with DNA. The stability of the copper(II) compounds in DMSO was studied by UV–Vis spectroscopy. UV–Vis titrations to determine the Kb were carried out in less than 2 h, in cacodylate–NaCl buffer containing just 5% DMSO. Within such a short period of time, and at that percentage of DMSO, the complexes and DNA remain unaltered on their own.

Spectra for 1−4 were recorded during 48 h, and no significant absorption changes were observed, indicating that the compounds are stable in DMSO and in the cacodylate–NaCl buffer (Tables S4, S5 and Fig. S1–S5).

UV–Vis spectrophotometric titrations

Intensity changes and wavelength shifts are observed in the electronic absorption spectra of 1−4 when the concentration of added DNA is increased, which suggest that these coordination compounds interact with the biomolecule.

The intercalation of small molecules into double-stranded DNA generally results in hypochromism, viz., a decrease in absorption intensity, associated with its redshift (bathochromic effect); these features arise from strong ππ stacking interactions between planar aromatic ligands and DNA base pairs [53]. On the other hand, non-intercalative interactions will enhance the absorption intensity (hyperchromism), which is ascribed to DNA conformation changes upon complex binding, leading to the alteration of the structure of the double helix. The results achieved for 1−4 (from experiments carried out in triplicate) are summarized in Table 10.
Table 10

Electronic absorption data regarding the titrations of compounds 1–4 with ct-DNA

Compound

λmax (nm)

Effect on intensity

Shift (nm)

Kb (M−1)

[Cu3(bzpr)4(H2O)2](NO3)2·3H2O·CH3OH 2

273

Hyperchromism

0

1.11 × 106

[Cu3(bzpr)4Cl2]·3H2O 3

272

Hyperchromism

+ 1

1.3 × 107

[Cu(bzpr)2(H2O)]·2H2O 1

273

Hypochromism

0

6.0 × 105

[Cu(bztb)2]·2H2O 4

279

Hypochromism

0

4.5 × 105

The electronic spectra of 2 and 3 (Fig. 10) display absorption bands at λmax = 273 and 272 nm, respectively, which intensify when increasing amounts of DNA are added. This hyperchromism, with a redshift of 1 nm for 3, suggests that both compounds most likely interact through a non-intercalative binding mode with DNA. These interactions may take place through (i) hydrogen bonding between DNA base pairs and the ligands, (ii) electrostatic attractions between the positively charged metal center and the negatively charged phosphate backbone of DNA or even (iii) binding in the major or minor groove [54]. The intrinsic binding constants (Kb) are 1.11 × 106 M−1 for 2 and 1.3 × 107 M−1 for 3; 3 thus interacts significantly stronger with DNA than 2. In Fig. 10, the linearity is shown up to a DNA concentration of 32 µM.
Fig. 10

Absorption spectra of 25 µM solutions of 2 (top) and 3 (down), in the absence (∙∙∙∙) and presence (__) of increasing amounts of ct-DNA (0−50 µM). The arrows show the intensity change upon increase of [ct-DNA], in cacodylate–NaCl buffer (pH 7.1) after incubation at 37 °C for 1 h

The absorption bands of mononuclear complexes 1 and 4 decrease in intensity (hypochromic effect) upon addition of ct-DNA (Fig. 11), without any shift. Such features are indicative of an intercalative binding mode (between DNA base pairs). The maximum absorbances at λmax = 273 and 279, for 1 and 4, respectively, were used to calculate the corresponding Kb values, i.e., 6.0 × 105 M−1 for 1 and 4.5 × 105 M−1 for 4. In Fig. 11, the linearity is shown up to a DNA concentration of 32 µM. These values again indicate the occurrence of strong interactions between the compounds and DNA.
Fig. 11

Absorption spectra of 25 µM solutions of 1 (top) and 4 (down), in the absence (∙∙∙∙) and presence (__) of increasing amounts of ct-DNA (0−50 µM). The arrows show the intensity change upon increase of [ct-DNA], in cacodylate–NaCl buffer (pH 7.1) after incubation at 37 °C for 1 h

Fluorescence spectroscopy

Ethidium bromide (EB) fluoresces when intercalated between DNA base pairs (λem = 592 nm); hence, the binding of an incoming molecule may lead to fluorescence quenching through displacement of EB and/or by photoelectron transfer from the excited state of EB (to the incoming molecule) [55, 56, 57]. For trinuclear compounds l 2 and 3, a decrease of the emission intensity of EB is observed (Fig. 12). Three possible binding modes for these compounds may be envisaged (as schematized below): (i) competition between the incoming molecule and EB for the same binding sites; (ii) interaction of the incoming molecule at distinct sites (from those of EB); (iii) displacement of EB by a change on conformation induced by the interaction of the incoming molecule [58, 59].
Fig. 12

Plot of the relative fluorescence quenching of [EB–DNA] in the presence of 2 (• • •) and 3 (__), at [Complex]/[DNA] ratios ranging from 0.0 to 1.6

  1. (i)

    Cu3L4 + DNA–EB ↔ DNA–Cu3L4 + EB K.

     
  2. (ii)

    Cu3L4 + DNA–EB ↔ EB-DNA–Cu3L4 K′.

     
  3. (iii)

    Cu3L4 + DNA–EB ↔ DNA–Cu3L4 + EB K′′.

     

The UV–Vis studies with trinuclear 2 and 3 indicated that they interact via a non-intercalative pathway(s) with DNA. Thus, the main quenching process of the DNA–EB complex by these compounds most likely involves the formation of a non-fluorescent [DNA–EB–copper] species or/and a so-called fluorescence static quenching process.

It is important to stress that if a quenching efficiency > 50% is observed with a [Complex]/[DNA] ratio < 1, then most likely the compound is interacting through an intercalative binding mode [60]. In the present study, [Complex]/[DNA] ratios of 0–1.6 were employed, for which EB–DNA emission intensity decrease by less than 50% was achieved (Fig. 12). These data therefore suggest that 2 and 3 do not interact with the double strand at the EB sites.

The comparative DNA-binding strength of each complex was evaluated by determining their respective Stern–Volmer quenching constant (Ksv). From the linear plots of I0/I versus [Complex] (insert in Fig. 13), Ksv values of 1.42 × 104 M−1 and 1.15 × 104 M−1 were obtained for 2 and 3, respectively. It must be mentioned here that Kb and Ksv represent distinctive features; indeed, while the Kb values characterize the DNA-binding abilities of the complexes, the Ksv values measure their competitive binding efficiencies to EB–DNA species.
Fig. 13

Fluorescence emission spectra of the EB–DNA complex in the absence (∙∙∙∙) and presence (__) of complexes 2 (a) and 3 (b). [EB] = 75 µM, [DNA] = 15 µM, [Complex] = 0−25 µM, λem = 592 nm

For mononuclear compounds 1 and 4, emission quenching of [DNA–EB] is observed; however, the data obtained did not allow getting a satisfactory linear fit applying the Stern–Volmer equation, due to the low quenching of the fluorescence by the adduct [DNA–EB]. For this reason, additional fluorescence-quenching experiments were performed using the DNA minor groove binder Hoechst 33258 (bisbenzimid dye, λem = 461 nm). The data achieved reveal a clear diminution of the fluorescence of the [Hoechst–DNA] adduct (Fig. 14); the interaction of compounds 1 and 4 gives rise to a distortion of the DNA double helix, which expels the dye, according to equation (iii). The corresponding Ksv values are 4.17 × 104 M−1 (linearity is shown up to a complex concentration of 20 µM) and 4.14 × 104 M−1, for 1 and 4, respectively (Fig. 14).
Fig. 14

Fluorescence emission spectra of the [Hoechst-DNA] adduct in the absence (∙∙∙∙) and presence (__) of compounds 1 (a) and 4 (b). [Hoechst 33258] = 2 µM, [DNA] = 15 µM, [Complex] = 0−30 µM, λem = 461 nm

Circular dichroism (CD) spectroscopy

Circular dichroism spectroscopy is a useful technique to assess wether nucleic acids undergo conformational changes as a result of complex-DNA formation. The CD spectrum of B-DNA displays a positive band at 277 nm attributable to base stacking, and a negative band at 245 nm that is a characteristic of the right-handed helicity of DNA [61]. Any modifications of the base-stacking pattern or helicity of the DNA strands will produce changes in the band position or/and intensity. It is known that simple electrostatic or groove binding interactions (non-intercalative interactions) of small molecules with DNA do not cause significant alteration of the intensity of the CD bands. On the other hand, a classical intercalator tends to enhance the intensities of both bands or even change their position, as a result of strong base-stacking interactions that increase the stability of the right-handed, B conformation of DNA.

The CD spectra of ct-DNA in the presence of compounds 2 and 3, using [Complex]/[DNA] ratios of 0.0, 0.2, 0.6 and 1.0, are depicted in Fig. 15. Upon increasing the complex concentration, both the positive and negative bands decrease in intensity, and a slight redshift is observed from 277 to 280 nm. This redshift is ascribed to ππ* transitions, which indicates that the complexes do alter the DNA base stacking, but without provoking significant changes in the supramolecular helicity. These CD data are consistent with a strong non-intercalative binding mode with DNA.
Fig. 15

CD spectra of ct-DNA (100 µM) in the absence (___) and presence of complexes 2 (a) and 3 (b) at [Complex]/[DNA] ratios of 0.2 and 0.6 (—) and 1.0 (∙∙∙∙)

Conversely, the CD spectra for compounds 1 and 4 interacting with DNA (Fig. 16) display a clear redshift of both bands, in addition to the intensity decrease. The positive band shifts from 277 to 285 nm for compound 4, and from 277 to 281 nm for compound 1. The negative band shifts from 245 to 249 nm for 4 and from 245 to 248 nm for 1. These results indicate that the DNA-binding mode of these mononuclear compounds is not similar to that of 2 and 3. It should be noted that in the case of these mononuclear compounds, for fluorescence and CD experiments, different concentrations of the complex are added to an excess of DNA, which favors the groove binding interactions. On the contrary, for the UV–Vis experiments, different concentrations of DNA were added to high concentration of the complexes, which favors the intercalative mode.
Fig. 16

CD spectra of ct-DNA (100 µM) in the absence (___) and presence of complexes 1 (a) and 4 (b) at [Complex]/[DNA] ratios of 0.2, 0.6 (—) and 1.0 (∙∙∙∙)

DNA cleavage studies

The nuclease activity of each complex was subsequently investigated by using electrophoresis with pBR322 plasmid DNA in the presence of ascorbic acid (H2Asc) as a reducing agent. Agarose gel electrophoresis is a common and simple technique which is used to visualize interactions of molecules with DNA. Different forms of plasmid DNA can be observed by electrophoresis; supercoiled DNA (Form I) migrates faster on the gel. If one-strand scission occurs, the resulting circular nicked form (Form II) exhibits a much slower migration rate. Finally, when both strands are cleaved, a linear form (Form III) is generated that migrates in between Forms I and II.

The DNA-interacting/cleaving properties of compounds 1−4 evaluated by the electrophoretic mobility of pBR322 plasmid DNA are shown in Fig. 17 (1−3) and Fig. 18 (4). The DNA-cleaving properties of complexes from redox metal ions (such as copper) are better assessed by adding a reducing reagent such as hydrogen peroxide or ascorbic acid. In the present study, H2Asc was used to mimic the reducing environment found inside cells; H2Asc will favor the occurrence of a Cu(I)/Cu(II) redox cycle that will allow the formation of reactive oxygen species (ROS), which are harmful to DNA. It can be noticed that free H2Asc slightly affects DNA, as some Form II is produced (compare lanes 1 and 2 in Fig. 17). CuCl2 and the free ligands do not damage DNA (Fig. 17, lane 3 and lanes 4−5, respectively).
Fig. 17

Agarose gel electrophoresis images of pBR322 plasmid DNA incubated for 1 h at 37 °C in cacodylate–NaCl buffer, with increasing concentrations of complexes 1–3, in the presence of a reducing agent, viz., ascorbic acid (H2ASC; 100 µM). Lane 1—control; lane 2—100 µM H2ASC; lane 3—5 µM CuCl2; lane 4—10 µM Hbzpr; lane 5—50 µM Hbzpr, lanes 6–20—H2ASC + [Complex] (2.5, 5, 10, 25 and 50 µM, respectively). Lanes 6−10—complex 1; lanes 11−15—complex 2; lanes 16−20—complex 3

Fig. 18

Agarose gel electrophoresis images of pBR322 plasmid DNA incubated for 1 h at 37 °C in cacodylate–NaCl buffer, with increasing concentrations of complexes 4 and [Cu(bza)2]. Lane 1—control; lane 2—100 µM H2ASC; lane 3—5 µM CuCl2; lane 4—10 µM Hbzpr; lane 5: 50 µM Hbzpr; lanes 6−10—100 µM H2ASC + [CuII(bztb)2]·2H2O (2.5, 5, 10, 25 and 50 µM); lanes 11−13—controls with CuCl2 and the ligand Hbza (10 and 50 µM); lanes 14−18—DNA + [CuI(bza)2] (2.5, 5, 10, 25 and 50 µM)

The effects of 1−3 on DNA are shown in lanes 6−20. The lowest complex concentration, namely 2.5 µM, does not seem to significantly affect pBR322 plasmid DNA (lanes 6, 11 and 16). On increasing the [Complex], a vanishing of both bands (ascribed to Form I and Form II of DNA) is noticed until complete disappearance, at [Complex] = 25 μM for 1 (lane 9) and [Complex] = 10 μM for 2 and 3 (lanes 13 and 18, respectively). These electrophoretic data indicate that the plasmid DNA is completely degraded into undetectable small pieces. It is important to mention that the enhanced cleavage of 2 and 3 with respect to 1 may be related to the higher Cu content per molecule (3:1 with respect to the mononuclear compound).

The behavior of complex [Cu(bztb)2].2H2O 4 was compared with that of an analogous copper(I) compound, namely [Cu(bza)2] (bza = 2-sulfonyl acetic acid). For compound 4, DNA Form II is observed at a concentration of 10 µM (lane 8), and Form III (resulting from double-strand break(s)) appears when [Complex] ≥ 25 (lanes 9 and 10). Actually, Form I (supercoiled DNA) is completely absent at these high concentrations. In contrast, the copper(I) compound [Cu(bza)2] (lanes 14−18) does not induce any damage to the biomolecule.

The intensities of the bands were quantified for compounds 1−4, with the objective of determining the respective amounts of each form of DNA. Subsequently, the percentages of DNA-cleaving activity were calculated using Eq. (1). The corresponding data shown in Fig. 19 reveal that compounds 1−4 considerably damage the DNA, their efficiency following the order 3 > 2 > 4 > 1.
Fig. 19

Percentages of DNA cleavage induced by compounds 14

The data suggest that all the complexes oxidatively cleave DNA, by means of reactive oxygen species (ROS), such as hydroxyl radicals (HO), superoxide anions (O2) or dihydrogen peroxide (H2O2). The copper(II) compounds thus act as oxidation catalysts, generating extremely damaging oxidative species through a CuI/CuII redox cycle, under aerobic conditions in the presence of ascorbic acid as a reducing agent [62]:
$$\begin{aligned} {\text{H}}_{ 2} {\text{ASC}}\, + \,{\text{H}}_{ 2} {\text{O }} \leftrightarrow {\text{ HASC}}^{ - } \, + \,{\text{H}}_{ 3} {\text{O}}^{ + } \quad \left( {\text{a}} \right) \hfill \\ {\text{HASC}}^{ - } \, + \,2{\text{Cu}}\left( {\text{II}} \right)\, \to \,{\text{ASC}}\, + \,2{\text{Cu}}\left( {\text{I}} \right)\, + \,{\text{H}}^{ + } \quad \left( {\text{b}} \right) \hfill \\ 2{\text{Cu}}\left( {\text{I}} \right)\, + \,{\text{O}}_{2} \, + \,2{\text{H}}^{ + } \, \to \,2{\text{Cu}}\left( {\text{II}} \right)\, + \,{\text{H}}_{ 2} {\text{O}}_{2} \quad \left( {\text{c}} \right) \hfill \\ {\text{HASC}}^{ - } \, + \,{\text{O}}_{2} \, + \,{\text{H}}^{ + } \, \to \,{\text{ASC}}\, + \,{\text{H}}_{ 2} {\text{O}}_{ 2} \quad \left( {{\text{b}}\,{ + }\,{\text{c}}} \right). \hfill \\ \end{aligned}$$

(a) Hydrolysis of ascorbic acid (H2ASC), generating ascorbate (HASC), (b) reduction of the Cu(II) coordination compound to a Cu(I) compound by HASC, (c) production of H2O2 by reaction of copper(I) with dioxygen, (b + c) autoxidation of ascorbate.

From this last step, a Fenton reaction takes place (step (d) below), where hydrogen peroxide reacts with another equivalent of Cu(I) species, producing ROS (e.g., hydroxyl radicals) that cause oxidative DNA damage [63].
$$\left( {\text{d}} \right){\text{ Cu}}\left( I \right)\, + \,{\text{H}}_{ 2} {\text{O}}_{ 2} \, \to \,{\text{Cu}}\left( {\text{II}} \right)\, + \,{\text{HO}}^{ - } \, + \,{\text{HO}}^{ \cdot } .$$

Conclusions

Trinuclear and mononuclear copper(II) coordination compounds from benzimidazole carboxylate ligands were prepared and characterized, namely [Cu(bzpr)2(H2O)]·2H2O 1, [Cu3(bzpr)4(H2O)2](NO3)2·3H2O·CH3OH 2, [Cu3(bzpr)4Cl2]·3H2O 3 and [Cu(bztb)2]·2H2O 4. In these compounds, the Cu(II) ions present two types of coordination environments, i.e., distorted octahedral and square-pyramidal geometries. Compounds 1−3 illustrate the great versatility of the ligand Hbzpr as various coordination modes are observed, arising from the carboxylate group that can act as a (i) bridging μ2-(O,O′), (ii) (bifurcated) bridging μ2-(O,O) or (iii) chelating μ1-(O,O’) ligand. In compound 4, the bztb ligand is coordinated in a tridentate fashion, stabilizing a fac-octahedral geometry.

The DNA-interacting properties of the compounds have been studied with various techniques. The binding mode of mononuclear compounds 1 and 4 is in accordance with their concentration in relation to the DNA; at minor concentrations of the complexes a groove binding interaction is favored, while at higher concentrations an intercalative mode is preferred. From their X-ray crystal structures, these results may be related to the size of the coordination compound, which is in the range of ca. 11.3 Å for complex 1 and 9.5 Å for complex 4. At minor concentrations of these complexes, they may interact with the minor groove, via electrostatic interactions. On the other hand, at higher concentrations an intercalative mode is preferred that may be through π-stacking, with the benzimidazole rings, and/or hydrogen bonding. On the other hand, the trinuclear complexes 2 and 3 are larger, ca. 12.1 Å, and their aromatic rings are not exposed, favoring a groove interaction through electrostatic interactions.

All complexes exhibit nuclease properties and damage the DNA through an oxidative pathway, their cleavage efficiency following the order 3 > 2 > 4 > 1. Thus, the trinuclear compounds are more effective than the mononuclear ones, and mononuclear compound 4, containing a sulfur donor atom, is a better DNA cleaver than compound 1.

Notes

Acknowledgements

The financial support from CONACYT, grant CB2012-178851 and DGAPA-UNAM for grant IN224516 is acknowledged. V.A.B.-G. thanks a CONACYT scholarship. P.G. acknowledges the financial support from the Ministerio de Ciencia, Innovación y Universidades (projects CTQ2015-70371-REDT and CTQ2017-88446-R AEI/FEDER, UE). We thank P. Fierro for technical support.

Supplementary material

775_2018_1598_MOESM1_ESM.docx (262 kb)
Supplementary material 1 (DOCX 262 kb)

References

  1. 1.
    Tullius TD (1989) Metal-DNA chemistry. ACS Symposium Series, vol 402. American Chemical Society, Washington DCCrossRefGoogle Scholar
  2. 2.
    Loehrer PJ, Einhorn LH (1984) Ann Intern Med 100:704–713CrossRefGoogle Scholar
  3. 3.
    Lippert B (1999) Cisplatin: chemistry and biochemistry of a leading anticancer drug. VHCA & Wiley-VCH, ZurichCrossRefGoogle Scholar
  4. 4.
    Cuello-Garibo JA, James CC, Siegler MA, Bonnet S (2017) Chem Sq 1:2CrossRefGoogle Scholar
  5. 5.
    Stern BR (2010) J Toxicol Environ Health Part A 73:114–127CrossRefGoogle Scholar
  6. 6.
    Lippard SJ, Berg JM (1994) Principles of bioinorganic chemistry. Mill Valley, CaliforniaGoogle Scholar
  7. 7.
    Jagadeesh M, Kalangi SK, Krishna LS, Reddy AV (2014) Spectrochim Acta, Part A 118:552–556CrossRefGoogle Scholar
  8. 8.
    Sayen S, Carlier A, Tarpin M, Guillon E (2013) J Inorg Biochem 120:39–43CrossRefGoogle Scholar
  9. 9.
    Duff B, Thangella VR, Creaven BS, Walsh M, Egan DA (2012) Eur J Pharmacol 689:45–55CrossRefGoogle Scholar
  10. 10.
    Ali I, Wani WA, Saleem K, Hseih M-F (2013) Polyhedron 56:134–143CrossRefGoogle Scholar
  11. 11.
    Li G-Y, Du K-J, Wang J-Q, Liang J-W, Kou J-F, Hou X-J, Ji L-N, Chao H (2013) J Inorg Biochem 119:43–53CrossRefGoogle Scholar
  12. 12.
    Silveira VC, Benezra H, Luz JS, Georg RC, Oliveira CC, Ferreira AMC (2011) J Inorg Biochem 105:1692–1703CrossRefGoogle Scholar
  13. 13.
    Patel MN, Dosi PA, Bhatt BS, Thakkar VR (2011) Spectrochim Acta Part A 78:763–770CrossRefGoogle Scholar
  14. 14.
    Kellett A, Howe O, Connor MO, McCann M, Creaven BS, McClean S, Kia AF-A, Casey A, Devereux M (2012) Free Radical Biol Med 53:564–576CrossRefGoogle Scholar
  15. 15.
    Kashanian S, Khodaei MM, Roshanfekr H, Shahabadi N, Mansouri G (2012) Spectrochim Acta Part A 86:351–359CrossRefGoogle Scholar
  16. 16.
    Grau J, Renau C, Caballero AB, Caubet A, Pockaj M, Lorenzo J, Gamez P (2018) Dalton Trans 47:4902–4908CrossRefGoogle Scholar
  17. 17.
    Schreiber JP, Deune M (1969) Biopolymers 8:139–152CrossRefGoogle Scholar
  18. 18.
    Chikira M, Inue M, Negane R, Harada W, Shindo H, Antholine WE (2000) J Inorg Biochem 78:243–249CrossRefGoogle Scholar
  19. 19.
    Morrow JR, Iranzo O (2004) Curr Opin Chem Biol 8:192–200CrossRefGoogle Scholar
  20. 20.
    Erxleben A Interactions of copper complexes with nucleic acidsGoogle Scholar
  21. 21.
    Spingler B, Da Pieve C (2005) Dalton Trans 1637–1643Google Scholar
  22. 22.
    Medina-Molner A, Rohner M, Pandiarajan D, Spingler B (2015) Dalton Trans 44:3664–3672CrossRefGoogle Scholar
  23. 23.
    Harada W, Nojima T, Shibayama A, Ueda H, Sindo H, Chikira M (1996) J Inorg Biochem 64:273–285CrossRefGoogle Scholar
  24. 24.
    Negane R, Chikira M, Oumi M, Shindo H, Antholine WE (2000) J Inorg Biochem 78:243–249CrossRefGoogle Scholar
  25. 25.
    Li D-D, Tian J-L, Gu W, Liu X, Zeng H-H, Yan S-P (2011) J Inorg Biochem 105:894–901CrossRefGoogle Scholar
  26. 26.
    Suntharalingam K, White AJP, Vilar R (2010) Inorg Chem 49:8371–8380CrossRefGoogle Scholar
  27. 27.
    Suntharalingam K, Hunt DJ, Duarte AA, White AJP, Mann DJ, Vilar R (2012) Chem Eur J 41:4955–4965Google Scholar
  28. 28.
    Skinnerm WA, Schelstraete MGM, Baker BR (1959) J Org Chem 24:1827CrossRefGoogle Scholar
  29. 29.
    Biron KK (2006) Antivir Res 71:154–163CrossRefGoogle Scholar
  30. 30.
    Middleton T, Lim HB, Montgomery D, Rockway T, Tang H, Cheng X, Lu L, Mo H, Kohlbrenner WE, Molla A, Kati WM (2004) Antivir Res 64:35–45CrossRefGoogle Scholar
  31. 31.
    Labanauskas L, Brukštus A, Udre˙ naite˙ E, Gaidelis P, Bucˇinskaite˙ V (2003) Chemija (Vilnius) 14:49Google Scholar
  32. 32.
    Sari H, Covington AK (2005) J Chem Eng Data 50:1425–1429CrossRefGoogle Scholar
  33. 33.
    Meaney M, Allister J, McKinstry B, McLaughlin K, Brennan GP, Forbes AB, Fairweather I (2007) Parasitol Res 100:1091–1104CrossRefGoogle Scholar
  34. 34.
    Mirskova AN, Levkovskaya GG, Mirskov RG, Voronkov MG (2008) Russ J Org Chem 44:1478–1485CrossRefGoogle Scholar
  35. 35.
    Luneau D, Rey P (2005) Coord Chem Rev 249:2591–2611CrossRefGoogle Scholar
  36. 36.
    Kabatc J, Jurek K (2012) Polymer 53:1973–1980CrossRefGoogle Scholar
  37. 37.
    Yoe F, Flores-Álamo M, Morales F, Escudero R, Cortés-Hernández H, Castro M, Barba-Behrens N (2014) Inorg Chim Acta 423:36–45CrossRefGoogle Scholar
  38. 38.
    Sheldrick GM (2015) Acta Cryst. A71:3–8Google Scholar
  39. 39.
    Hübschle CB, Sheldrick GM, Dittrich B (2011) J Appl Crystallogr 44:1281–1284CrossRefGoogle Scholar
  40. 40.
    Spek AL (2015) Acta Cryst. C71:9–18Google Scholar
  41. 41.
    Reichmann MF, Rice SA, Thomas CA, Doty P (1954) J Am Chem Soc 76:3047–3053CrossRefGoogle Scholar
  42. 42.
    Wolfe A, Shimer GHJ, Meehan T (1987) Biochemistry 26:6392–6396CrossRefGoogle Scholar
  43. 43.
    Lakowicz JR, Weber G (1973) Biochemistry 12(21):4161–4170CrossRefGoogle Scholar
  44. 44.
    Shubsda MF, Goodisman J, Dabrowiak JC (1997) J Biochem Biophys Methods 34:73–79CrossRefGoogle Scholar
  45. 45.
    Patel MN, Dosi PA, Bhatt BS (2010) Polyhedron 29:3238–3245CrossRefGoogle Scholar
  46. 46.
    Lever ABP (1968) J Chem Ed 45:711–712CrossRefGoogle Scholar
  47. 47.
    Valderrama-Negrón AC, Alves WA, Cruz ÁS, Rogero SO, de Oliveira Silva D (2011) Inorg Chim Acta 367:58–92CrossRefGoogle Scholar
  48. 48.
    Nakamoto K (1986) Infrared and Raman Spectra of Inorganic and Coordination Compounds. Wiley-InterscienceGoogle Scholar
  49. 49.
    Addison AW, Rao TN, Reedijk J, van Rijn J, Verschoor GCJ (1984) Chem Soc 7:1349–1356Google Scholar
  50. 50.
    Klein A, Neugebauer E, Krest A, Lüning A, Garbe S, Arefyeva N, Schlörer N (2015) Inorganics 3:118–138CrossRefGoogle Scholar
  51. 51.
    Jahn HA, Teller E (1937) Proc R Soc Lond Ser A161:220–235CrossRefGoogle Scholar
  52. 52.
    Matthews CJ, Heath SL, Elsegood MRJ, Clegg W, Leese TA, Lockhart JC (1998) J Chem Soc Dalton Trans 12:1973–1977CrossRefGoogle Scholar
  53. 53.
    Kelly JM, Tossi AB, McConnell DJ, OhUigin C (1985) Nucleic Acids Res 13:6017–6034CrossRefGoogle Scholar
  54. 54.
    Meenongwa A, Chaveerach U, Siriwong K (2011) Inorg Chim Acta 366:357–365CrossRefGoogle Scholar
  55. 55.
    Chaveerach U, Meenongwa A, Trongpanich Y, Soikum C, Chaveerach P (2010) Polyhedron 29:731–738CrossRefGoogle Scholar
  56. 56.
    Marmur J (1961) J Mol Biol 3:208–218CrossRefGoogle Scholar
  57. 57.
    Meenongwa A, Brissos RF, Soikum C, Chaveerach P, Gamez P, Trongpanich Y, Chaveerach U (2015) N J Chem 39:664–675CrossRefGoogle Scholar
  58. 58.
    Nyarko E, Hanada N, Habib A, Tabata M (2004) Inorg Chim Acta 357:739–745CrossRefGoogle Scholar
  59. 59.
    Silveira VC, Benezra H, Luz JS, Georg RC, Oliveira CC, Ferreira AMC (2011) J Inorg Biochem 105:1692–1703CrossRefGoogle Scholar
  60. 60.
    Ling X, Zhong W, Huang Q, Ni K (2008) J Photochem Photobiol B 93:172–176CrossRefGoogle Scholar
  61. 61.
    Ivanov VI, Minchenkova LE, Schyolkina AK, Poletayev AI (1973) Biopolymers 12:89–110CrossRefGoogle Scholar
  62. 62.
    Collins CH, Lyne PM (1970) Microbiological methods. University Park Press, BaltimoreGoogle Scholar
  63. 63.
    Sigman DS, Chen C-HB (1986) Acc Chem Res 19:180–186CrossRefGoogle Scholar

Copyright information

© SBIC 2018

Authors and Affiliations

  • Víctor A. Barrera-Guzmán
    • 1
  • Edgar O. Rodríguez-Hernández
    • 1
  • Naytzé Ortíz-Pastrana
    • 2
  • Ricardo Domínguez-González
    • 3
  • Ana B. Caballero
    • 4
    • 5
  • Patrick Gamez
    • 4
    • 5
    • 6
  • Norah Barba-Behrens
    • 1
  1. 1.Departamento de Química Inorgánica, Facultad de QuímicaUniversidad Nacional Autónoma de México, Ciudad UniversitariaMexico CityMexico
  2. 2.Departamento de QuímicaCinvestavMexico CityMexico
  3. 3.Facultad de QuímicaUniversidad Nacional Autónoma de México, Ciudad UniversitariaMexico CityMexico
  4. 4.Departament de Química Inorgánica y OrgánicaUniversitat de BarcelonaBarcelonaSpain
  5. 5.Institute of Nanoscience and Nanotechnology (IN2UB)BarcelonaSpain
  6. 6.Catalan Institution for Research and Advanced StudiesBarcelonaSpain

Personalised recommendations