Introduction

Isoniazid (INH) was first introduced in tuberculosis therapy in the 1950s (Judge et al. 2012). Since then, it is regarded as one of the most commonly used and efficient drugs in treatment of human tuberculosis (Bernardes-Génisson et al. 2013).

In spite of efforts to eradicate tuberculosis, its burden remains substantial (Dheda et al. 2016). Over one-third of the global population is infected by tuberculosis, which causes death in approximately two to three million people annually (WHO 2014). The increased incidence of multidrug-resistant (Prozorov et al. 2012), and more recently, extensively drug-resistant (Sotgiu et al. 2009) strains of Mycobacterium tuberculosis compromise the recurrent effective treatment and evidences the urgency for novel antituberculosis agents (Hoagland et al. 2016).

Modification of existing drugs and the development of novel active compounds have been some of the strategies to improve tuberculosis drug therapy (Wong et al. 2013; Ellis et al. 2014; Chaves et al. 2015). In this sense, INH has become the most researched antitubercular agent (Hearn et al. 2009). A range of isoniazid analogues has been studied for its antitubercular potential and a number of promising candidates has been described (Oludina et al. 2014; Parumasivam et al. 2013; Matei et al. 2013; Ramani et al. 2012; Kumar et al. 2014).

Recently, Ellis and co-workers (2014) noticed that the INH analogue 2-pyridinecarboxaldehyde isonicotinoyl hydrazone (HPCIH), an effective iron chelator, displays potent inhibition of mycobacterial growth, probably acting as a lipophilic vehicle for the transport of the intact INH moiety into the mycobacterium. In spite of its marked antimycobacterial action, it has been shown HPCIH presents limited antiproliferative activity against malignant SK-N-MC (neuroepithelioma) (Becker and Richardson 1999). In our previous work, we demonstrated the analogues 2-acetylpyridine-(HAPIH), 2-benzoylpyridine-(HBPIH), 2-pyridineformamide-(HPAmIH) and 2-pyrazineformamide-(HPzAmIH) isonicotinoyl hydrazones in general exhibit better growth-inhibiting properties towards MCF-7, OVCAR-8 and SF-295 cells than HPCIH (Amim et al. 2016).

Metal complexes of substituted hydrazones have been reported to hold therapeutic activity and have shown pharmacological applications (Lessa et al. 2012, 2013; Raman et al. 2008; Chang et al. 2015). It has been evidenced that the presence of an α(N)heterocyclic ring at azomethine scaffold of hydrazones plays a major role in extending their chelating and/or pharmacological properties (Beraldo and Gambino 2004).

Due to the effectiveness of HPCIH (Armstrong et al. 2003) and its analogues (Ababei et al. 2010; Chang et al. 2015) as tridentate chelating agents for transition metal ions, in this work we report the synthesis of four copper(II) complexes (5–8) with HPCIH (1), HAPIH (2), HPAmIH (3) and HPzAmIH (4) (Fig. 1). As part of our interest in the development of copper(II) complexes as antimycobacterial and anticancer agents, compounds 5–8 were assayed for their action toward Mycobacterium tuberculosis H37Rv ATCC 27294 strain and the human tumor cell lines OVCAR-8 (ovarian cancer), SF-295 (glioblastoma multiforme) and HCT-116 (colon adenocarcinoma).

Fig. 1
figure 1

Structures of α(N) heterocyclic isoniazid-derived hydrazones HPCIH (1), HAPIH (2), (HPAmIH) (3) and HPzAmIH (4)

Full size image

Materials and methods

Chemicals

Isoniazid, 2-pyridinecarboxaldehyde, 2-acetylpyridine, 2-pyridinecarbonitrile, pyrazinecarbonitrile and copper(II) chloride dihydrate were purchased from Aldrich and used without further purification.

Physical measurements

Partial elemental analyses were performed on a Perkin Elmer CHN 2400 analyzer. Melting points were determined on Gehaka-PF1500 Farma, a Capillary Melting Point Apparatus. A CG 1800 Gehaka conductivity bridge (conductimetric cell constant 1 cm−1) was employed for molar conductivity measurements of compound solutions (1 × 10−3 mol L−1) in dimethylsulfoxide (DMSO). Infrared spectra were recorded on an attenuated total reflectance/Fourier transform infrared spectrometer (Varian FT-IR 660) in the 4000–600 cm−1 range. Electronic spectra were recorded on an Agilent Technologies 8453 spectrophotometer at room temperature, using a 10 mm beam path quartz cuvette and DMSO as solvent. Magnetic susceptibility measurements were carried out at 298.5 °C on a Johnson Matthey MSB/AUTO balance.

Synthesis of the hydrazones (1–4) and their copper(II) complexes (5–8)

Synthesis of HPCIH (1) (Armstrong et al. 2003), HAPIH (2) (Ababei et al. 2010), HPAmIH (3) (Amim et al. 2016), and HPzAmIH (4) (Glushkov et al. 2004) are described in the literature. Copper(II) complexes 5–8 were obtained by mixing, under reflux and stirring for 4 h, a methanol solution (15 mL) of the desired hydrazone (1 mmol) with CuCl2·2H2O in 1:1 ligand-to-metal molar ratio. The resulting solids were filtered off, then washed with methanol followed by diethylether and dried at 50 °C for 24 h. Single crystals of [Cu(HAPIH)Cl]Cl·H2O (6a) and [Cu(HPzAmIH)Cl2]·H2O (8a) were obtained from mother solutions of 6 and 8, respectively.

Dichloro(2-pyridinecarboxaldehyde-isonicotinoyl hydrazone)copper(II)] hydrate, [Cu(HPCIH)Cl2]· 0.4H2O (5)

Green solid. Anal. Calc. for C12H10.8Cl2CuN4O1.4 (FW = 367.89 g mol−1): C, 39.18; H, 2.96; N, 15.23. Found: C, 39.50; H, 2.79; N, 14.76 %. IR (ATR, cm−1): ν (N + py −H) 2555, ν (C = N) 1523, ρ(py) 644. Molar conductivity (1 × 10−3 mol L−1, DMSO) 46.0 Ω−1 cm2 mol−1. Effective magnetic moment: 1.91 MB. UV–vis (DMSO, λ in nm/ε in mol−1dm2): 386/3.02 × 105, 752/521. Yield 42 %.

Dichloro(2-acetylpyridine-isonicotinoyl hydrazone)copper(II) hidrate, [Cu(HAPIH)Cl2]·1.25 H2O (6)

Green solid. Anal. Calc. for C13H14.5Cl2N4O2.25Cu (FW = 397.23 g mol−1): C, 39.31; H, 3.68; N, 14.10 %. Found: C, 39.47; H, 3.43; N, 13.82 %. IR (ATR, cm−1): ν (N + py −H) 2569, ν (C = N) 1531, ρ(py) 645. Molar conductivity (1 × 10−3 mol L−1, DMSO): 33.4 Ω−1 cm2 mol−1. Effective magnetic moment: 1.85 MB. UV–vis (DMSO, λ in nm/ε in mol−1 dm2): 383/1.34 × 105; 743/136. Yield 77 %.

Dichloro(2-pyridineformamide-isonicotinoyl hydrazone)copper(II) hidrate, [Cu(HPAmIH)Cl2]·H2O (7)

Brown solid. Anal. Calc. for C12H13Cl2CuN5O2 (FW = 393.71 g mol−1): C, 36.61; H, 3.33; N, 17.79 %. Found: C, 36.65; H, 2.85; N, 17.33 %. IR (ATR, cm−1): ν (N–H) 3064, δ(NH2) 1658, ν (C = O), 1622 ν(C = N) 1531, ρ(py) 646. Effective magnetic moment: 1.89 MB. UV–vis (DMSO, λ in nm/ε in mol−1dm2): 398/2.59 × 105; 754/229. Yield 87 %

Dichloro(pyrazineformamide-isonicotinoyl hydrazone)copper(II), hidrate [Cu(HPzAmIH)Cl2]·1.25 H2O (8)

Brown solid. Anal. Calc. for C11H12.25Cl2CuN6O2.25 (FW = 394.70 g mol−1): C, 33.09 %; H, 3.16 %; N, 21.05 %. Found: C, 33.27 %; H, 2.90 %; N, 20.66 %. IR (ATR, cm−1): ν(N–H) 3055, δ(NH2) 1666, ν(C = O) 1616, ν(C = N) 1525, ρ(py) 654. Molar conductivity (1 × 10−3 mol L−1 DMSO): 33.1 Ω−1 cm2 mol−1. Effective magnetic moment: 1.91 MB. UV–vis (DMSO, λ in nm/ε in mol−1dm2): 332/2.60 × 104, 425/6.36 × 104, 764/229. Yield 91 %.

Crystallography

Single-crystal X-ray diffraction methods were used to determine the structures of [Cu(HAPIH)Cl]Cl·H2O (6a) and [Cu(HPzAmIH)Cl2]·H2O (8a). Data were collected at room temperature on a Bruker D8 VENTURE equipped with Mo Kα high-brilliance IµS radiation (λ = 0.71073 Å) and a PHOTON 100 CMOS detector. The instrument was controlled by the APEX2 software package (Bruker 2014). Data were processed using the integrate plug-in in the controlling software package (SAINT) and corrected for absorption by the multiscan semi-empirical method implemented in SADABS (Bruker 2014). Using Olex2 (Dolomanov et al. 2009) the structure was solved with the SHELXS-97 (Sheldrick 2008) structure solution program by means of Direct Methods and refined with the SHELXL-2013 (Sheldrick 2008) refinement package using Least Squares minimization. Positional and anisotropic atomic displacement parameters were refined for all non-hydrogen atoms. Hydrogen atoms were placed geometrically and the positional parameters were refined using a riding model.

In vitro biological activity assays

Cytotoxicity toward human tumor cell lines

The cytotoxic activity of compounds 1–8 was tested against SF-295 (glioblastoma multiforme), HCT-116 (colon adenocarcinoma) and OVCAR-8 (ovarian cancer), from National Cancer Institute (Bethesda, MD, USA). The cells were maintained in RPMI 1640 medium supplemented with 10 % fetal bovine serum, 2 mM glutamine, 100 μg/mL penicillin, and 100 μg/mL streptomycin at 37 °C/5 % CO2. Each compound was previously dissolved in DMSO (stock solution), whose final concentration in the RPMI culture medium was kept below 0.1 % (v/v). For initial cytotoxic activity evaluation, compounds 1–8 (5 μg/mL) were incubated with SF-295, HCT-116 and OVCAR-8 cell lines, for 72 h. Cell viability was determined by dye reduction 3-(4,5-dimethyl-2-thiazole)-2,5-diphenyl-2H-tetrazole bromide (MTT) assay to yield the formazan, which is detected by electronic spectroscopy (Mosman 1983). Compounds that inhibited the proliferation in more than 50 % were selected for determination of the half maximal inhibitory concentration (IC50). To this end, 5–0.009 μg mL−1 range for compound concentration was used. All experiments were performed in least three replicates per compound and results are shown as the average and 95 % confidence interval of three independent experiments.

Antitubercular activity

Antimycobacterial activities of compounds 1–8, as well as isoniazid and copper(II) chloride, were assessed against Mycobacterium tuberculosis H37Rv ATCC 27294 using the Micro plate Alamar Blue Assay (MABA) (Franzblau et al. 1998). This methodology is nontoxic, uses a thermally-stable reagent and shows good correlation with proportional and BACTEC radiometric methods (Tortoli et al. 2002; Kontos et al. 2004). The method is described as follows: 200 ml of sterile deionized water was added to all outer-perimeter wells of 96 sterile well plates (falcon, 3072: Becton–Dickinson, Lincoln Park, NJ) to minimize evaporation of the medium in the test wells during incubation. The 96 plates received 100 mL of the Middlebrook 7H9 broth (Difco laboratories, Detroit, MI, USA) and successive dilution of the compounds was performed directly on the plate. The final drug concentrations tested were 0.01–20.0 mg/mL. Plates were covered and sealed with parafilm and incubated at 37°C for 5 days. Twenty five milliliter of a freshly prepared 1:1 mixture of Alamar Blue (Accumed International, WestlakeOhio) reagent and 10 % tween 80 were then added to the plate and incubated for 24 h. A blue color in the well was interpreted as no bacterial growth, and a pink color was scored as growth. The minimal inhibition concentration (MIC) was defined as the lowest drug concentration, which prevents a color change from blue to pink.

Results and discussion

Formation of the copper(II) complexes

Microanalyses suggest the formation of [Cu(HPCIH)Cl2]·0.4H2O (5), [Cu(HAPIH)Cl2]·1.25H2O (6), [Cu(HPAmIH)Cl2]·H2O (7) and [Cu(HPzAmIH)Cl2]·1.25H2O (8). For complexes 5 and 6, the hydrazones coordinate as zwitterionic forms (based on infrared spectroscopy), whereas neutral ligands are attached to the metal center in 7 and 8. Molar conductivities of 5–8 weren’t determined in commonly used solvents (Geary 1971) due to its low solubility. Instead, molar conductivities were measured in DMSO whose values suggest the compounds behave as weak electrolytes in solution (Zianna et al. 2016). This behavior is probably consequence of either the labile nature of chloro ligands or metal coordinating feature of the solvent. At room temperature, powdered samples of complexes display effective magnetic moments (μeff) in the 1.85–1.91 BM range, which is higher than the spin-only value. Such divergence, which is not quite uncommon in mononuclear copper(II) complexes, is due to mixing-in of some orbital angular momentum from the closely lying excited states via spin–orbit coupling (Bhattacharyya et al. 1996).

Infrared spectra

Absorptions observed between 1545 and 1557 cm−1 assigned to the ν(C = N) in the IR spectra of free hydrazones (1–4) shift to 1523–1531 cm−1 in the spectra of the copper(II) complexes (5–8), suggesting coordination through azomethine nitrogen (Parrilha et al. 2014). Absorptions attributed to ρ(py) observed at 613–665 cm−1 for 1–4 exhibit pronounced shift in spectra of complexes 5–8 (644–654 cm−1), suggesting the pyridine nitrogen coordination (Ferraz et al. 2013).

Bands attributed to ν(N–H) of secondary amines are absent in the spectra of 5 and 6. On the other hand, a fairly strong and very broad absorption attributed to the ν(N+–H) stretching vibration of the pyridinium group (Lessa et al. 2011) is observed around 2600 cm−1. Moreover, the ν(C = O) absorptions at 1665–1677 cm−1 in the spectra of the uncomplexed hydrazones disappear in those ones of complexes 5 and 6, in agreement with coordination of an enolate oxygen (Mondal et al. 2013). Thus, for 5 and 6 the hydrazones are attached to copper(II) as zwitterionic species. Nonetheless, for complexes 7 and 8, ν(C = O) vibrations are present at 1616–1622 cm−1, which are shifted in relation to the free hydrazones, in accordance with coordination through a keto oxygen (Mishra and Sharma 2009).

Electronic spectra

The electronic absorption spectra of hydrazones (1–4) and their complexes (5–8) were recorded at room temperature using DMSO as the solvent. The absorption spectra of the ligands are characterized by one band and a sholder or by two bands in the 297–377 nm region, which are assigned to n → π* and π → π* transitions of azomethine and the carbonyl groups (Cohen and Flavian 1967; Gegiou et al. 1996; Sorrell 1989). In the UV–vis spectra of 5–8 these absorptions are shifted. Also, the complexes show a single broad band centred in 743–764 nm, which is typical for d–d transition of Jahn–Teller distorted copper(II) complexes in square pyramidal geometries (Tabbì et al. 2013).

Structural study of [Cu(HAPIH)Cl]Cl·H2O (6a) and [Cu(HPzAmIH)Cl2]·H2O (8a)

Crystal data and structure refinement for 6a and 8a are summarized in supplementary information (online resource). Selected bond distances and angles for 6a and 8a are shown in Table 1. Compounds 6a and 8a crystallized in the triclinic and monoclinic systems, respectively. ORTEP (Farrugia 1997) drawings (Figs. 2, 3) display the hydrazones tridentate to the metal ion through NazoNaromO system for both complexes, giving rise to two five-membered rings. Despite the similarities in ligand structures, metal ions adopt different coordination numbers and geometries, that is, 6a shows a distorted square planar geometry, in which one chloride is attached to the metal center, whereas 8a is a pentacoordinated compound, in which two chloride ions complete its coordination. According to Addison and co-workers (1984), for distorted pentacoordinate structures, the parameter τ (τ = (β−α)/60°, where α and β are the largest angles around the metal center) can be used to rationalize its geometries. Value τ is 0 for perfectly square-pyramidal geometry and 1 for perfectly trigonal–bipyramidal geometry. For 8a, τ is 0.36, which suggests it is likely a distorted square–pyramidal compound.

Table 1 Selected bond lengths (Å) and angles (°) for [Cu(HAPIH)Cl]Cl·H2O (6a) and [Cu(HPzAmIH)Cl2]·H2O (8a)
Full size table
Fig. 2
figure 2

ORTEP diagram for 6a (elipsoids at 50 % at probability)

Full size image
Fig. 3
figure 3

ORTEP diagram for 8a (elipsoids at 50 % at probability)

Full size image

Hydrazone ligands are nearly planar whose rms deviation of atoms from the least-squares plane is 0.0782 Å for 6a and 0.0588 Å for 8a. Metal ions lay close onto this plane (at 0.0481(10) and 0.0715(10) Å for 6a and 8a, respectively), as well as chlorine ligand in 6a (0.638(2) Å).

The N3–C8 bond lengths 1.320(2) and 1.306(3) Å found for hydrazones in 6a and 8a, respectively, is shorter than the similar free hydrazone HBPIH (1.3601 (1) Å) (Ababei et al. 2010). Furthermore, C8–O1 bond is marked longer in 6a (1.280(2) Å) and 8a (1.288(2) Å) in comparison with HBPIH (1.2152 (1) Å). The C8–O1 bond is most likely to change from a double to a predominantly single bond and N3–C8 acquires some double bond character when hydrazones are attached to copper(II) in the enolate form (Despaigne et al. 2012; Mondal et al. 2013).

Besides, the enolate ligands are protonated at the para-substituted pyridine nitrogen in both complexes, indicating it is attached to the metal ion in the zwitterionic form. Complex 8a is also protonated in the powder, whereas zwitterionic form of 6a was obtained only through the crystallization process.

The dihedral angles C2−C7−N2−N3 and N2−N3−C8−O1 are 179.87(17) and 2.5(3), respectively, for 6a as well as −179.59(16) and 1.8(3) respectively for 8, which are in accordance with EZ conformation adopted by the hydrazones when attached to copper(II).

Interactions in crystal packing for 6a and 8a are described in Table 2. The interaction between Cu1 and Cl1(1−x, 2−y, 1−z) (2.6375(6) Å) is the main contact in structure of compound 6a, which results in the formation of a dimeric arrangement. Study of hydrogen bonds reveals a chain along the [1–2 1] direction. For compound 8a, a three-dimensional hydrogen-bonding network is observed connecting water molecules and coordination compound (see supplementary information).

Table 2 Hydrogen bond distances (Å) and angles (°) for [Cu(HAPIH)Cl]Cl·H2O (6a) and [Cu(HPzAmIH)Cl2]·H2O (8a)
Full size table

Cytotoxicity against tumor cell lines

Figure 4 reports growth inhibition of the human tumor cell lines OVCAR-8, SF-295 and HCT-116 induced by compounds 1−4 (Amim et al. 2016) and its complexes 5–8. INH and copper chloride dihydrate were also tested for comparison.

Fig. 4
figure 4

Growth inhibition (GI,  %) of OVCAR-8, HCT-116 and SF-295 cells promoted by 1−8, copper(II) chloride and isoniazid (INH)

Full size image

According to results, isoniazid proved to be poorly effective against the three cell lines. Copper chloride, in turn, reduced in 75 % the SF-295 cells growth, whereas it presented moderate to low activity against the other cell lines.

All copper(II) complexes were more effective in reducing growth of HCT-116 and SF-295 cells than the respective free hydrazones. Coordination led to significant higher cytotoxity of 7 to OVCAR-8 cells than hydrazone 3. 2 as well as its complex 6 also strongly inhibited OVCAR-8 cells growth. It is noteworthy that complexes 6 and 7 were able to inhibit the growth of all cell lines in more than 90 %.

The most potent compounds HAPIH (2), [Cu(HAPIH)Cl2]·1.25H2O (6) and [Cu(HPAmIH)Cl2]·H2O (7) were selected to determine the concentration which inhibits 50 % of cell growth (IC50) (Table  3 ). Complex 6 was found to be the most active compound against all strains, whose activity is superior to the free hydrazone 2. Besides, 6 is as potent as the anticancer drug doxorubicin. Thus, coordination of 2 to copper(II) was an efficient approach to obtain compound with improved action against tumor SF-295, OVCAR-8 and HCT-116 cell lines.

Table 3 Cytotoxic activity (IC50) against SF-295, OVCAR-8 and HCT-116 cell lines of 2, 6 and 7 in comparison with doxorubicin
Full size table

Antimycobacterial activity

Determined values of minimum inhibitory concentrations (MIC) of hydrazones 1−4, their complexes 5–8, INH and copper(II) chloride salt against Mycobacterium tuberculosis H37Rv (ATCC 27294) are listed in Table 4.

Table 4 Minimal inhibitory concentration of 1−8, copper(II) chloride and isoniazid (INH) against M. tuberculosis H37Rv (ATCC 27294)
Full size table

Selected hydrazones display different behavior toward M. tuberculosis. HPCIH (1) shows moderate activity, whereas substitution of hydrogen at imine carbon by a methyl group in HAPIH (2) increases the antimycobacterial potency. HAPIH (MIC = 2.60 µM) is as effective as the reference antitubercular drug INH (MIC = 2.27 µM) in inhibit M. tuberculosis. Nonetheless, the presence of formamide in HPAmIH (3) as well as pyrazine substituent in HPzAmIH (4) leads to moderate and lost of action, respectively. In general, coordination of hydrazones to copper(II) promotes reduction in MIC values. [Cu(HPCIH)Cl2]·0.4H2O (5), for example, exhibited sub-micromolar MIC value (0.85 µM) and was around 15-fold more effective than HPCIH (1) (13.79 µM) in inhibit growth of M. tuberculosis. Complexes 5 and [Cu(HAPIH)Cl2]·1.25H2O (6) displayed high activity against M. tuberculosis, as compared with isoniazid, which suggests the compounds are attractive candidates as antitubercular drugs.

Conclusion

In this work, copper(II) complexes 5–8 were evaluated toward three tumor cell lines (OVCAR-8, SF-295 and HCT-116). In most cases, chelation with metals gave rise to enhancement of the ligands activity against the tested cells. 6 and 7 were appointed as lead cytotoxic complexes. Additionally, 6 has proved to be as effective as the anticancer drug doxorubicin. Further work will be needed to understand the mechanism whereby the complex disturbs cellular proliferation.

Upon coordination to copper(II), activity against Mycobacterium tuberculosis H37Rv growth significantly improved except for 7. Copper(II) chloride is poorly effective, suggesting the action is probably due to the complex per se. Coordination of HPCIH (1) to copper(II) was an efficient strategy to produce a compound (5) with improved antimycobacterial action. Complex 5 was also more active than isoniazid, suggesting it is a promising compound, which should be considered for further studies aiming to confirm its potential as novel antitubercular drug candidate.