The effects of nickel ( II ) complexes with imidazole derivatives on pyocyanin and pyoverdine production by Pseudomonas aeruginosa strains isolated from cystic fibrosis *

1Department of Microbiology, Institute of Biology, Jan Kochanowski University, Kielce, Poland; 2Institute of Chemistry, Jan Kochanowski University, Kielce, Poland; 3Department of Environment Protection and Modelling, Jan Kochanowski University, Kielce, Poland; 4Institute of Physics, Jan Kochanowski University in Kielce, Kielce, Poland; 5Department of Radiobiology and Immunology, Jan Kochanowski University, Kielce, Poland; 6Centre for Radiobiology and Biological Dosimetry, Institute for Nuclear Chemistry and Technology, Warsaw, Poland


INTRODUCTION
Cystic fibrosis (CF) is a heritable disease, characterized by chronic inflammation of the airways, leading to bronchiectasis and respiratory failure.It causes accumulation of a thick, dehydrated mucus in the lungs which increases the patient's susceptibility to bacterial infections.It is an optimal niche for microorganisms which induce chronic lung diseases in children and young adults.These bacterial infections lead to progressive pulmonary damage and emphysema.Eradication of bacterial biofilms formed in mucus is a crucial problem, because the distribution of classic antibiotics in biofilm/mucus structure is weak and their antibacterial activity might stimulate drug resistance.
Biophysical properties of biofilm structure exopolymer secretions (EPS) in gel-like mucus are directly associated with reduced susceptibility to antibiotics and limit the effective eradication of bacteria (Stewart & Costerton, 2003;Manago et al., 2015).It suggests that not only antibacterial, but also diffusion parameters of new drugs should be considered in the eradication of biofilm-forming bacteria.
P. aeruginosa is a common opportunistic pathogen causing respiratory diseases in patients with CF, and produces a diversity of pigments, such as pyocyanin and pyoverdine (Kolpen et al., 2014;Nguyen et al., 2014;Muller & Merrett, 2015).Pyocyanin (1-hydroxy-5-methyl-phenazine) is a redox-active compound having the ability to accept and donate electrons.In the sputum of CF patients, pyocyanin is detected at 100 µM (Wilson et al., 1988).In the cytoplasm of host cells, this pigment has the ability to transfer electrons from reduced compounds, e.g.glutathione, and leads to a reduction of antioxidants along with higher concentrations of strong oxidizing compounds, such as reactive oxygen species (ROS) and superoxide anions (Manago et al., 2015).Pyocyanin is able to damage the cilia of respiratory epithelium, increases IL-8 secretion by epithelial cells, induces apoptosis and inhibits T-cell proliferation (Gloyne et al., 2011).Pyoverdine, acting as a siderophore, binds and transports iron ions into bacterial cell.It was shown that pyoverdine plays a key role in biofilm formation regardless of the presence of iron (Meyer et al., 1996;Nadal Jimenez et al., 2010).
The aim of this study was to evaluate the effects of complexes of nickel(II) ([Ni(iaa) and NiCl 2 on pyocyanin and pyoverdine production by 23 strains of P. aeruginosa isolated from CF at non-cytotoxic range of concentrations against A549 cells.Moreover, the diffusion parameters of tested substances through P. aeruginosa PAO1 biofilm (used as a model of Pseudomonas sp.) were determined by laser interferometry and confocal microscopy.The physical parameters of PAO1 biofilm were determined by atomic force microscopy.
Pyocyanin and pyoverdine production by clinical strains of P. aeruginosa.Clinical P. aeruginosa strains were grown in Tryptic Soy Broth (TSB) in the presence of metal complexes or their ligands, at concentrations in the range from 7-500 µM, overnight (16-18 h) at 37°C under microaerophilic conditions (5% CO 2 ) in stationary culture.The absorbance of pyocyanin was determined spectrophotometrically at 691 nm in TSB growth medium (200 µl of bacterial cell free supernatant) from the culture described above.The fluorescence of pyoverdine was determined (λ ex = 398 nm; λ em = 460 nm) in growth medium under the same conditions with a Microplate Reader Infinite 200 PRO (TECAN, Switzerland).All samples were measured in one repetition, in two independent experiments.
The level of pyocyanin or pyoverdine production in the presence of tested substances, as well as non-treated control, was normalized (ND t or ND c , respectively) against the culture density measured spectrophotometrically at 600 nm according to the formulas: and where P is pyocyanin or pyoverdine absorbance/fluorescence, G is the liquid culture density respectively for the treatment group (T) and the control group (C).
The estimation of the nickel ions, complexes or their ligands, effects (E s ) on pyocyanin and pyoverdine production and bacterial growth were based on the measurement developed by the authors, according to the formula: Theoretically, one of three kinds of effects might be observed: E s >1 (increase of the pyocyanin or pyoverdine production); E s =1 (no effect on the pyocyanin or pyoverdine production) and E s <1 (decrease of the pyocyanin or pyoverdine production).
The statistical significance of E s was estimated by the Wilcoxon matched-pairs signed rank test of paired T-test, nonparametric.A P value of <0.05 was considered significant.Statistical analyses were performed using GraphPad Prism version 6 (GraphPad Software, USA).
Statistical analyses were carried out to check the strength of the E s values of all the substance concentrations, in comparison to the absolute value 1 taken as no effect.If the E s values of all substance concentrations were significantly below or above the value of 1, it meant that E s is statistically significant for all substance concentrations for a particular bacterial strain.
Cytotoxicity of Ni(II) complexes with imidazole derivatives.A549 cells (adenocarcinoma human alveolar basal epithelial cells) were cultured at 37°C, in a humidified 5% CO 2 atmosphere, in plastic dishes, in F-12K medium supplemented with 10% heat inactivated fetal calf serum, 2 mM l-glutamine and antibiotics (100 units/ ml penicillin and 100 µg/ml streptomycin).A549 cells were treated with (NO 3 ) 2 and their ligands at concentrations of 125 μM or 250 μM, for 48 hours.After treatment, the frequencies of early apoptotic, late apoptotic and necrotic cells were evaluated with the Annexin V-FITC apoptosis detection Kit I (BD Pharmingen, USA), described elsewhere (Darzynkiewicz, 2003).The cells were briefly washed two times with cold PBS, and then resuspended in a 1× binding buffer at a concentration of 1×10 6 cells/ml.An aliquot of 125 µl of the cell suspension was incubated with 5 µl of annexin V-FITC, and 5 µl of propidium iodide (PI), at room temperature for 15 min in the dark.The cells were resuspended in 400 µl of 1× binding buffer.The fluorescence was determined using an LSR II flow cytometer (Becton Dickinson, USA).CellQuest Pro (Becton Dickinson, USA) was used for data acquisition and analysis.Data for 20 000 events were stored.A cell gate containing A549 was established on the basis of forward and side light scatter.Four different populations of cells were detected with the annexin V-FITC kit: normal cells that are annexin-negative and PI-negative and express no fluorescence, early apoptotic cells that are annexin-positive and PI-negative and express green fluorescence, late apoptotic/necrotic cells that are annexinpositive and PI-positive and express green and orange fluorescence, and necrotic cells that are annexin-negative and PI-positive and express orange fluorescence.
Atomic Force Microscopy of PAO1 biofilm.The PAO1 biofilm was formed for 72 h, at 37°C, in TSB medium on polyethylene terephthalate (PET) membrane with pore diameter of 1 µm, as an element of the Cell Culture Inserts (BD Biosciences, USA).The study of surface topography was performed using atomic force microscopy NanoCompact AFM (PHYWE, Germany).It was measured with the tapping mode, using the probe Tap190Al-G (BudgetSensors, Bulgaria).The mean surface roughness (S a ), root mean square (S q ), average value (S m ), valley depth (S v ), maximum peak (S p ) and height The effects of complexes on P. aeruginosa (S y ) were determined.Roughness parameters were determined using the Measure Nano software dedicated to this model of the microscope.
Laser interferometry.The measurements of 1-allylimidazole or imidazole-4-acetate anion diffusion through PAO1 biofilm was performed by the laser interferometry system (Arabski et al., 2007;2009a;2009b;2013b;Danis-Wlodarczyk et al., 2015).The system under study consists of two glass cuvettes separated by the horizontally located membrane (Nephrophane membrane with PAO1 biofilm formed for 72 h).The lower cuvette was filled with an aqueous 1-allylimidazole or imidazole-4-acetate anion solution at an initial concentration of 100 mmol/L, while in the upper cuvette there was pure water.With such a configuration of the measurement system, the solution in the upper cuvette remains undisturbed and a stable concentration boundary layer (CBL) of d thickness is created.The 1-allylimidazole or imidazole-4-acetate anion concentration is uniform in the planes parallel to the biofilm-solution interface and concentration gradients occur only in the vertical direction.The interferograms, which appear due to the interfer-ence of two laser beams, are determined by the refraction coefficient of the solute, which in turn depends on the concentration of the 1-allylimidazole or imidazole-4-acetate anion solution.With recording the interferograms with a given time-step (Dt=2 min.),one can reconstruct after computer analysis the interference fringe course of the spatio-temporal concentration distribution (i.e.concentration profile C(x,t)).The amount (mol) of 1-allylimidazole or imidazole-4-acetate anion, N(t), which diffuses in time t through PAO1 biofilm formed on Nephrophane membrane to water was calculated by integrating the concentration profile C(x,t) in the upper cuvette according to: where S denotes the surface of the biofilm-water interface (S=7×10 -5 m 2 ), and d the CBL thickness (determined interferometrically).All experiments were performed at a temperature of 37°C.The diffusion coeffi- (4) cient D of 1-allylimidazole or imidazole-4-acetate anion in PAO1 biofilm as well as in water was determined on the basis of Fick's first law and the second Kedem-Katchalsky equation (Arabski et al., 2007;Slezak et al., 2005).
Confocal microscopy of P. aeruginosa PAO1 biofilm.The thickness of the biofilm formed on the Nephrophane membrane was analyzed on the basis of pyoverdine fluorescence released into the matrix of PAO1 biofilm formed 72 h at 37°C.The membrane with formed biofilm was placed on a microscopic slide and analyzed by confocal microscopy.The value of the biofilm thickness was used for calculation of diffusion parameters of imidazole derivatives through biofilm (obtained by laser interferometry).The fluorescence analysis of PAO1 biofilm was performed by using confocal microscope A1R (Nikon, Japan).The specimen underwent excitation with laser beam having a wavelength equal to 405 and 488 nm.Images was taken with PlanAPO 10× DIC-L lens.Resolution of DU4 12-bit photomultiplier tube was set to X:1024 px Y:1024 px and the resolving power of microscope was 0.62 μm/px.Image acquisition was obtained within the range of 85 μm.The thickness of PAO1 biofilm was measured in 18 steps with Z-step (5 μm).The fluorescence intensity in the area of the biofilm was measured as the average value along a randomly chosen thin plane of biofilm.It was measured through all layers collected in the Z -axis scan.

Effect of nickel complexes on P. aeruginosa strains
Studies were carried out under microaerophilic conditions specific for the CF respiratory system.Statistical studies enabled the independent assessment of the growth and production of pigments: pyocyanin and pyoverdine by Pseudomonas aeruginosa in the presence of nickel(II) complexes and their ligands.The growth of all tested P. aeruginosa strains was positively correlated with pyocyanin production, in contrast to diversified secretion of pyoverdine by bacteria after 24 h, at 37°C under microaerophilic conditions (5% CO 2 ) (results not shown).
The nickel(II) complexes and their ligands had no effect on pyocyanin production by all tested bacterial strains (results not shown).Figure 1 shows the summary of effects of nickel(II) complexes with 1-allylimidazole or imidazole-4-acetate anion (% of strains; p<0.05) on P. aeruginosa growth in correlation with pyoverdine production.The presence of nickel ions(II) or imidazole-4-acetate anion, in contrast to their complexes, resulted in a decrease in pyoverdine secretion in 37% and 46% of tested P. aeruginosa strains, respectively.The ligand 1-allylimidazole alone and in the [Ni(1-allim) 6 ](NO 3 ) 2 complex, inhibited pyoverdine production in 42% and 37% of bacterial strains, respectively.

Cytotoxicity of nickel(II) complex
The Annexin/IP assay allowed to determine the frequency of apoptotic and necrotic A549 cells in the presence of nickel(II) complexes or their ligands.Table 1 shows that [Ni(iaa) 2 (H 2 O) 2 ]H 2 O induced late apoptosis, as well as necrosis, at the concentration of 250 μM in A549 cells after 48 hours, in contrast to the [Ni(1-allim) 6 ](NO 3 ) 2 complex.Both tested nickel (II) complexes at the concentrations of 125 μM, as well as their ligands, did not induce apoptosis in A549 cells.

Diffusion of ligands through PAO1 biofilm
Figure 2 shows the laser interferometry analysis of 1-allylimidazole or imidazole-4-acetate anion diffusion through PAO1 biofilm formed on Nephrophane membrane for 72 h at 37°C, 5% CO 2 .The amount of imidazole-4-acetate anion transported through P. aeruginosa PAO1 biofilm formed on the membrane (3.80 × 10 -6 mol) was similar to 1-allylimidazole (2.55 × 10 -6 mol) at initial concentration of 100 mmol/L after 60 min.It seems that this small difference is determined by specific diffusion properties of the membrane alone for each ligand.For 1-allylimidazole, it was observed that the membrane is saturated with this compound after 12 min, in contrast to imidazole-4-acetate anion.It was decided to calculate the diffusion coefficients for the elimination of the membrane effect (scaffold for biofilm).On the basis of confocal microscopy analysis (Fig. 3) of PAO1 biofilm thickness (65 μm), the diffusion coefficients of both tested imidazole derivatives were calculated.The diffusion coefficient of imidazole-4-acetate anion in PAO1 biofilm (2.1 × 10 -11 m 2 /s) in PAO1 biofilm was twice as good as 1-allylimidazole (1.1 × 10 -11 m 2 /s).

Analysis of PAO1 biofilm by AFM microscopy
The diffusion coefficients in PAO1 biofilm was measured in biofilm matrix formed on the Nephrophane mem-brane.For physical analysis of PAO1 biofilm structure formed for 72 h at 37°C, 5% CO 2 , the PET membrane was chosen, from the methodological (AFM) point of view.Figure 4 shows the topography of the PET membrane alone and with PAO1 biofilm formed for 72 h at 37°C, 5% CO 2 measured by AFM.The PAO1 biofilm smooths the rough surface of the PET membrane (Table 2) and is formed on their surface (Danis-Wlodarczyk et al., 2015).

DISCUSSION
Transition metal complexes with derivatives of imidazole might be considered as new antibacterial agents.The antibacterial properties of transition metals depend on the physical and chemical properties of such molecules, which may be transported through the cell membranes by diffusion, via calcium and iron channels (in interaction with pyoverdine as a siderophore) and by phagocytosis (Kasprzak et al., 2003;Simonetti et al., 2001).For example, the presence of imidazole-4-acetic acid anion in synthesized metal complexes may promote the effective transport of these molecules by changing the permeability of the cell membrane and making the bactericidal effect of antibiotics more intense.Moreover, metal complexes with imidazole, in contrast to metal ions alone, should (i) better diffuse in biofilm, because the neutral charge of metal complexes reduces their interaction with biofilm matrix components, like nucleic acids or proteins, in contrast to metal ions, (ii) reduce the redox potential in the environment of free radical reactions (chronic bacterial infections), and (iii) protect against formation of complexes with chloride ions, as it is important in the case of cystic fibrosis (Sekhon, 2010).
The aim of this study was to evaluate the effects of complexes of nickel(II) with imidazole derivatives and their ligands alone, on pyocyanin and pyoverdine production by 23 strains of P. aeruginosa isolated from CF, at non-cytotoxic concentration against eukaryotic cells.The antibacterial properties of the above newly synthesized complexes, were tested for the first time under conditions specific for eukaryotic cell line growth.The decrease of pyocyanin production was not observed in the presence of all tested compounds, as well as pyoverdine secretion by P. aeruginosa strains by [Ni(iaa) II), and both imidazole ligands (imidazole-4-acetate anion and 1-allylimidazole), decreased the production of pyoverdine in c.a. 40% of the tested P. aeruginosa strains.The antibacterial effect of nickel(II) complexes with imidazole derivatives, as well as ligands alone, were tested at a non-cytotoxic range of concentrations (<125 μM) against A549 cells.
Imidazole-4-acetate anion and 1-allylimidazole have good diffusion properties through mature PAO1 biofilm, as measured by laser interferometry and confocal microscopy.The value of imidazole-4-acetate anion diffusion coefficient in biofilm was two-times higher than 1-allylimidazole, 2.1 × 10 -11 m 2 /s and 1.1 × 10 -11 m 2 /s, respectively.Moreover, the diffusion coefficient of ciprofloxacin in PAO1 biofilm, calculated on the basis of data from Arabski et al., 2013a, is equal to 1.75 × 10 -11 m 2 /s.It indicates that diffusion properties of ciprofloxacin used in CF infection treatment and imidazole derivatives are similar, and all analyzed substances might penetrate the biofilm matrix at the same level.It was concluded that imidazole-4-acetate anion and 1-allylimidazole might be considered in chemical synthesis with other metals, for example with those that act as iron antagonists, like cobalt or gallium.

Figure 1 .
Figure 1.The effects of nickel(II) complexes with 1-allylimidazole or imidazole-4-acetate anion, and their ligands alone, on P. aeruginosa growth in correlation with pyoverdine production.White bars denote that the tested compound decreased (p<0.05) the secretion of pyoverdine in comparison to untreated control, independent of bacterial growth (E s <1).Black bars denote that the tested compound increased (p<0.05) the secretion of pyoverdine in comparision to untreated control, independent of bacterial growth (E s >1).Grey bars denote no effect on pyoverdine level in growth medium in the presence of the tested compounds.The percent (%) of bacterial strains with proper effect on pyoverdine level are presented.

Figure 3 .
Figure 3.The confocal microscopy analysis of PAO1 biofilm thickness formed on Nephrophane membrane for 72 h, at 37°C, 5% CO 2 .The fluorescence intensity in the area of the biofilm was measured as the average value along a randomly chosen thin plane of biofilm (higher panel).The thickness of PAO1 biofilm was measured in 18 steps with Z-step (5 μm) (lower panel).

Figure 4 .
Figure 4. Topography and amplitude of PET membrane alone and with PAO1 biofilm formed for 72 h, at 37°C, 5% CO 2, measured by AFM microscopy.

Table 1 . Percentage of early and late apoptotic and necrotic A549 cells treated with nickel complexes and their ligands; the mean of two independent experiments ± S.D. Cytotoxic activity calculated by ANOVA test.
*p<0.05The effects of complexes on P. aeruginosa