Selective cytotoxicity and antifungal properties of copper(II) and cobalt(II) complexes with imidazole-4-acetate anion or 1-allylimidazole

The physicochemical properties of metal complexes determine their potential applications as antitumor agents. In this study, the antitumor properties of mononuclear cobalt(II) and copper(II) coordination compounds (stoichiometry: [Co(iaa)2(H2O)2]·H2O (iaa = imidazole-4-acetate anion), [Co(1-allim)6](NO3)2 (1-allim = 1-allylimidazole), [Cu(iaa)2H2O] and [Cu(1-allim)4(NO3)2]) and their ligands have been evaluated on human lung carcinoma A549 cells and normal bronchial BEAS-2B cells. Designing the chemical structure of new antitumor agents the possible interactions with macromolecules such as DNA or proteins should be take into account. PCR gene tlr4 product served as DNA model, whereas lysozyme and phage-derived endolysin (both peptidoglycan degrading enzymes) were applied as protein/enzyme model. The interactions were analysed using PCR-HRM and circular dichroism, FT-IR, spectrophotometry, respectively. Additionally, the antimicrobial properties of the complexes at a non-cytotoxic concentration were analyzed against S. aureus, E. coli, P. aeruginosa and C. albicans strains. The results obtained in this study showed the selective cytotoxicity of metal complexes, mainly [Cu(1-allim)4(NO3)2] towards tumor cells. From all tested compounds, only [Co(iaa)2(H2O)2].H2O non-covalently interacts with DNA. Cu(II) and Co(II) complexes did not affect the secondary conformation of tested proteins but modified the hydrolytic activity of enzymes (lysozyme and endolysin). Moreover, only [Co(iaa)2(H2O)2].H2O exhibited the antifungal properties. In conclusion, Co(II) and Cu(II) metal complexes bearing two imidazole-4-acetate ligands seemed to be promising antitumor and antifungal agents for future drug design and application.


Results
In the first step of this study, the antitumor properties of [Cu (1-allim) 4  To identify the cellular target of metal complexes conditioning an antitumor activity, the interactions with DNA (using PCR-HRM technique) and proteins (lysozyme and recombinant endolysin) were tested using biophysical techniques. As the model DNA molecule the PCR gene tlr4 product was used, whereas lysozyme and phage-derived endolysin served as protein/enzyme model. Lysozyme and phage endolysin degrade peptidoglycan (PG) in two different ways corresponding to their enzymatic specificity: muramidase or endopeptidase, respectively. Additionally, the antimicrobial properties of all tested metal complexes were analyzed only at non-cytotoxic concentrations to eukaryotic cells.
Cytotoxic effect of metal complexes on eukaryotic cells lines. The cytotoxicity of tested complexes was determined using Annexin V-FITC apoptosis detection kit and MTT test. Tables 1 and 2 present the percentage of early/late apoptotic and necrotic A549 or BEAS-2B cells following the incubation with cobalt, cooper complexes or metal ions alone. The A549 cancer cells turned out to be more sensitive to Co(II) and Cu(II) ions alone than normal BEAS-2B ones. Interestingly, this tendency was statistically more significant in the case of coordination complexes of metals, especially [Co(iaa) 2 (Table 1). [Cu(1-allim) 4 (NO 3 ) 2 ] generally possess stronger cytotoxic properties than copper ions alone and at the concentration above 60 μM and 125 μM exhibited the toxic activity on A549 cells and BEAS-2B, respectively. The ligands alone had no cytotoxic activity against both tested eukaryotic cell lines. Figure 1 shows the metabolic activity of A549 and BEAS-2B cells after the incubation with [Co(iaa) 2 2 ] complexes or their metals or ligands alone at the range of 7-250 μM concentrations measured by the MTT assay. The results showed that: (i) A549 cells are more sensitive to Co(II) and Cu(II) complexes or their ions alone compared to the normal BEAS-2B cells; (ii) [Cu(1-allim) 4 (NO 3 ) 2 ] is more cytotoxic than copper ions alone; and (iii) ligands are non-cytotoxic for both tested eukaryotic cell lines.
DNA binding by metal complexes. The interaction of metal complexes with DNA molecule as a common target of antitumor agents, was analysed using PCR-HRM and cisplatin was used as the positive control. In this experiment we used a high resolution melting (HRM) procedure (PCR-based technique), to analyse the melting temperature of amplicon/tested chemical agent complex formed after DNA incubation with metal complexes. This application of PCR-HRM technique was described in detail in our previously work 20 Table 2. Percentage of early and late apoptotic and necrotic A549 or BEAS-2B cells following treatment with copper complexes measured by flow cytometry; mean of three independent experiments ± SD. IP -propidium iodide. Bolded font when p < 0.05 in comparison to proper control (normal cells).
www.nature.com/scientificreports www.nature.com/scientificreports/ the type of DNA-metal interaction (Fig. 2). In our study cisplatin has been chosen as a positive control having a well-known DNA binding mechanism of creating the intra-strand and inter-strand DNA cross-links 21 . Figure 3 shows the melting curve of DNA after the incubation with cisplatin for 1 h at 37 °C. DNA cross-linked with cisplatin (displace of the SYBR Green I molecules corresponds to the lower level of fluorescence) requires a higher temperature for its denaturation in comparison to the free DNA control. The melting peak was shifted to a higher temperature range in a dose-depended manner. The lower peak is observed on the melting curve of DNA incubated with [Co(iaa) 2  Interaction of proteins with Cu(II) and Co(II) metal complexes. Phage recombinant endolysin (KP27 endopeptidase) and commercially available lysozyme were chosen as an enzyme model for interactions with  www.nature.com/scientificreports www.nature.com/scientificreports/ Cu(II) and Co(II) metal complexes analysis. The changes in the secondary structure of enzymes by metal complexes (measured by circular dichroism, CD) were determined in correlation to PG degradation (measured by spectrophotometry). The secondary structure of endolysin and lysozyme was measured in the wavelength range from 260 to 195 nm (Fig. 4). In the CD spectra, no changes in the secondary structure were observed for endolysin and lysozyme, at any molar ratio of metal complexes presence (Fig. 4).
Additionally, we have tested PG degradation by selected enzymes after the incubation with metal complexes. The maximum of PG degradation occurred 15-20 minutes after lysozyme (lysozyme:complex) administration (Fig. 5). Because PG is not soluble in water, the maximum of absorbance showed up after 20 minutes and decreased very slowly. Not significant difference was observed in lysozyme and lysozyme:complexes activity, except for [Cu(1-allim) 4 (NO 3 ) 2 ], where the absorbance was still increasing (Fig. 5A). The results presented in It means that generally all tested complexes changed the kinetics or inhibited the enzymatic activity of endolysin. In the case of endolysin, Cu(II) and Co(II) complexes with imidazole-4-acetate anion or 1-allylimidazole   www.nature.com/scientificreports www.nature.com/scientificreports/ had a negative impact on enzymatic activity. Above results suggest a completely different mechanism of metal complex interactions with the molecule surface of these two proteins affecting catalytic site functioning 22 .

Discussion
The antitumor and antimicrobial properties of metal complexes are investigated for decades. Depending on the metal type in complex, the different biological properties can be observed. In this study, the Cu(II) and Co(II) complexes were tested. The antitumor activity was observed for Cu-complexes in contrast to none for Co-complexes. The results showed that [Cu(1-allim) 4 (NO 3 ) 2 ] at 125 µM concentration has selective cytotoxic properties towards tumor cells, higher than metals alone. No cytotoxic effect was detected for normal cell line treated at this concentration of metal complexes.
For cytotoxicity analyses we have used A549 tumor and BEAS-2B normal cell lines. The choice of model cell lines was made on the basis of previous report 23 . The authors characterized the expression of 380 genes encoding proteins involved in the metabolism of xenobiotics in commonly used lung cell lines and four primary cultures of human bronchial epithelial cells. This group of genes comprised 137 genes of phase I enzymes (including 56 P450s), 69 genes of phase II enzymes, 103 genes of transporters (including 31 ABC and 62 SLC transporters), 48 genes of nuclear receptors and transcription factors (including coactivators and corepressors), and 23 miscellaneous genes (including 9 metallothioneins). As the BEAS-2B cell line shows the highest homology in the gene expression pattern compared to the primary cells and the lowest number of dysregulated genes with nontumoral lung tissues, it could be used as a model for toxicological and pharmacological studies 24 . The A549 cells were chosen for this study as the most frequently used experimental model of tumor line in similar studies 25,26 .
There are several potential mechanisms of antitumor activity of metal complexes, especially in the case of Cu-complexes. These types of activity could be associated with the transport of Cu(II) ions into the cell. The Cu(II) ions are introduced by specific copper transporters hCtr in the Cu(I) form. The presence of natural copper transport system is crucial from the clinical point of view because no additional exogenous drugs carriers is need. Thus, the cellular effects of metal complexes activity might be determined by their interaction with proteins (eg. receptors) or nucleic acids. Considering metal complexes activity not only the interactions with DNA but also with proteins should be analyzed as the targeted molecules in antitumor mechanisms. In our study, lysozyme and endolysin, two low molecular proteins were chosen as an protein/enzyme model. All tested complexes have no influence on the secondary structure of lysozyme and endolysin, but they selectively affected the activity of studied proteins. Only Cu-complexes decreased the activity of both enzymes suggesting the interaction with the catalytic center of enzymes. That might explain the specific ability of [Cu(1-allim) 4 (NO 3 ) 2 ] in antitumor activity.
The cytotoxicity of metal complexes can also rely on the interaction with DNA like observed for cisplatin 21 . However, in our case the Cu-complexes are not able to bind to DNA. Only [Co(iaa) 2 (H 2 O) 2 ] . H 2 O exhibited that tendency, but no antitumor activity was noticed.
The next tested property of metal complexes was the antimicrobial activity. There was no antibacterial effect seen after metal complexes treating. The antifungal properties were observed mainly for [Co(iaa) 2 (H 2 O) 2 ]H 2 O complex at non-cytotoxic concentrations for eukaryotic cell lines. The probable antifungal mode of action might be related to a metal complex-protein interactions. Proteins that play role in the lipid metabolism, cell wall biogenesis, protein metabolism, electron transport, ATP synthesis and cellular stress protein samples were indentified in the lipid rafts of C. albicans located on the cell surface 27 . These proteins might be targeted and modified by metal complexes.
Kasuga et al. as well as Nawar et al. found that mainly the chemically labile complex, which has ability to the ligand replacement, exhibits a bactericidal effect. Such effect is determined by the substitution reactions of released ligands or metal from their complex with the main residues of targeted bacterial cell components 28,29 . It means that the stable chemical structure of metal complexes used in our study limited their bactericidal effect.

Analysis of cytotoxicity of metal complexes on eukaryotic cells. The human lung normal BEAS-2B
and carcinoma A549 were selected to compare the antitumor activity of four tested Co(II) and Cu(II) complexes with imidazole derivatives using Annexin V-FITC apoptosis detection test.
The BEAS-2B cells exhibited the highest homology in gene expression pattern with primary cells and the lowest number of dysregulated genes compared with non-tumoral lung tissues (comparison of the expression profiles of 380 genes encoding proteins involved in the metabolism of xenobiotics in 10 commonly used lung cell lines and four primary cultures of human bronchial epithelial cells) 23  The BEAS-2B and A549 were selected for MTT cell metabolic activity test according to manufacturer instructions using EZcount MTT Cell Assay kit (HiMedia, India). The exponentially growing A549 and BEAS-2B cells (1 × 10 5 /ml) were incubated with above metal complexes and their components alone at 7-250 μM for 48 hours. All samples were tested in three independent experiments using Microplate Reader TECAN Infinite 200 PRO (Tecan Group Ltd., Switzerland).

Analysis of metal complexes interaction with DNA.
The effect of metal complexes on DNA was measured by polymerase chain reaction (PCR) with high resolution melting (HRM) using LightCycler 480 II (Hoffman-LaRoche, Switzerland) 20 . The PCR-HRM was performed according to manufacturer instructions using LightCycler 480 SYBR Green I Master (Hoffman-LaRoche, Switzerland). Sequences of primers used in the reaction: forward 5′-GCT GTT TTC AAA GTG ATT TTG GGA GAA-3′ and reverse 5′-CAC TCA TTT GTT TCA AAT TGG AAT G-3′ for the gene tlr4 with a amplicon length of 147 bp. After amplification, tested metal complexes at 125 μM or positive control (cisplatin at 5-25 μM) were added to the prepared amplicon and incubated for 1 h at 37 °C and subsequently followed by a melting curve HRM protocol. HRM ramps were generated by acquiring fluorescence data at the temperature ramp of 37 °C to 95 °C at 0.1 °C intervals. HRM curve analysis was performed by using the LightCycler 480 Gene Scanning Software (version 1.5).

Analysis of metal complexes interactions with proteins.
The changes in secondary structure of endolysin and lysozyme due to the interaction with complexes were tested using circular dichroism method. Circular dichroism (CD) was measured in the far-UV region (the range between 260 and 195 nm) using a J-815CD spectrometer (Jasco, Japan). The experiments were done in 10 mM sodium phosphate buffer (pH 7.4). The concentration of the endolysin was 1 μM at 20 °C. Quartz 1 cm path length cells (Helma) were used for all CD experiments. The recording scan speed was 50 nm/min. 3 scans were run for each sample, and concentration-dependent with increased metal complexes concentrations up to 1:20 molar ratio (protein:complex respectively) were taken. The CD spectra were corrected by subtracting the spectra (as baseline) obtained for complexes without protein. The mean residue ellipticity θ 35 expressed as mdeg × cm 2 × dmol −1 , was calculated 36 .
Analysis of metal complexes interactions with enzymes. PG was isolated from E. coli ATCC 8739 cells. Briefly, E. coli was cultivated under aerobic conditions in a fermenter (BioStat A, Sartorius Stedim Biotech) in nutrient broth (BTL, Poland) under controlled conditions (37 °C, pH 7.2-7.4, pO 2 70-86%). Cells were harvested at the end of the logarithmic growth phase, centrifuged (5000 g, 30 min), washed with distilled water, and lyophilized. PG was isolated in accordance with the method described by Bera et al. 35 . It seems to be important to determine the purity of PG suspension in water for lysozyme or endolysin activity testing in the presence of metal complexes. The purity of the PG was tested by FT-IR method using lysozyme as a standard protein with peptidoglycan enzymatic degradation properties. Figure 8 shows FT-IR spectra of PG treated with lysozyme for 1 h at 37 °C. The reduction of -C-O-C-stretching vibrations (1037 cm −1 ) is correlated with PG degradation by lysozyme (hydrolysis of 1,4-beta-linkages between N-acetylmuramic acid and N-acetyl-D-glucosamine residues) (Fig. 9). The results suggest a proper purity of PG for enzymatic assays.
The analysis of metal complexes interactions was performed on peptidoglycan-degrading enzymes (lysozyme and endolysin) as a model (Fig. 9). The degradation of peptidoglycan (PG) by recombinant phage endolysin (KP27 endopeptidase) or lysozyme was measured spectrophotometrically in the presence of Co(II) and Cu(II) complexes with imidazole derivatives. Lysozyme or KP27 endopeptidase were preincubated with metal complexes at 1:5 molar ratio (protein: complex respectively) for 15 min at 37 °C. Secondly, 0.25 mg/ml of PG was added and the kinetics of its degradation was measured at 560 nm for 110 min with 2 min time intervals at 37 °C using Microplate Reader TECAN Infinite 200 PRO (Tecan Group Ltd., Switzerland).
Antimicrobial properties of metal complexes. The ~2 × 10 5 cfu/ml of bacterial or fungal culture were grown for 24 h at 37 °C with 5% CO 2 in stationary culture in Tryptic Soy Broth (Oxoid, USA) or Sabouraud Broth (Oxoid, USA), respectively, in the presence of metal complexes at non-cytotoxic concentrations: 60 µM for [Co(iaa) 2