Ruthenium(IV) Complexes as Potential Inhibitors of Bacterial Biofilm Formation

With increasing antimicrobial resistance there is an urgent need for new strategies to control harmful biofilms. In this study, we have investigated the possibility of utilizing ruthenium(IV) complexes (H3O)2(HL1)2[RuCl6]·2Cl·2EtOH (1) and [RuCl4(CH3CN)2](L32)·H2O (2) (where L1-2-hydroxymethylbenzimadazole, L32-1,4-dihydroquinoxaline-2,3-dione) as effective inhibitors for biofilms formation. The biological activities of the compounds were explored using E. coli, S. aureus, P. aeruginosa PAO1, and P. aeruginosa LES B58. The new chloride ruthenium complexes were characterized by single-crystal X-ray diffraction analysis, Hirshfeld surface analysis, FT-IR, UV-Vis, magnetic and electrochemical (CV, DPV) measurements, and solution conductivity. In the obtained complexes, the ruthenium(IV) ions possess an octahedral environment. The intermolecular classical and rare weak hydrogen bonds, and π···π stacking interactions significantly contribute to structure stabilization, leading to the formation of a supramolecular assembly. The microbiological tests have shown complex 1 exhibited a slightly higher anti-biofilm activity than that of compound 2. Interestingly, electrochemical studies have allowed us to determine the relationship between the oxidizing properties of complexes and their biological activity. Probably the mechanism of action of 1 and 2 is associated with generating a cellular response similar to oxidative stress in bacterial cells.

L 2 2: 267°C; for complex 2: 277°C. Studies have shown that the L 2 2 formed is tautomer of 3hydroxy-2-quinoxalinecarboxylic acid. The tautomeric structure of the obtained product was confirmed by X-ray crystallography and IR experiment. L 2

Physical measurements
The elemental analysis (C, H and N) was performed on a Vario Micro Cube Elemental Analyser CHNS. The IR spectra were recorded on a Nicolet 380 FT-IR spectrophotometer in the spectral range 4000 -500 cm −1 using the ATR-diffusive reflection method. UV-Vis spectra of the solid state of ligand -2-hydroxymethylbenzimidazole and complex 1 were recorded on a Shimadzu 2101 PC scanning spectrophotometer equipped with an ISR-260 attachment. The Kubelka-Munk function (F(R∞)) [2] was used to convert reflectance measurements into equivalent absorption spectra using the reflectance of BaSO4 as a reference. The multi-peak fitting analysis of the reflectance spectrum (complex 1) was applied using OriginPro8.5.1 program (OriginLab, Northampton, MA, USA). UV-Vis measurements in aqueous solutions were performed on a V-630 UV-Vis spectrophotometer from Jasco using 1 cm cuvettes against water as reference solutions. The absorbance measurements were recorded ca. 22°C and the concentrations were: 1.13·10 -4 M (for L1), 1.23·10 -4 M (for complex 1), 9.80·10 -5 M (for L 2 2) and 9.23·10 -5 M (for complex 2). The luminescence spectra were measured with an Infinite M200 PRO microplate reader (Tecan) with the xenon flashlamp as a light source (at room temperature). Magnetic measurements were carried out on a magnetic susceptibility balance (Sherwood Scientific) at room temperature by Gouy's method, using Hg[Co(NCS)4] as a calibrant. The data were corrected for diamagnetic contributions, which were estimated from Pascal's constants [3]. Molar conductivities of freshly prepared 110 -3 moldm -3 EtOH solutions were measured using Jenway. Voltammetric experiments were performed using a Model M161E electrochemical analyser connected with Model M162 preamplifier (mtm-anko, Poland) and controlled via a S4 Pentium computer using mEALab 2.1 software (mtm-anko, Poland). The details of the procedure have been described previously [4]. Electrochemical investigations of ruthenium complexes and free ligands were performed in a mixture of CH3CN -EtOH (3 : 2, v/v) containing 1 mM compound with 0.1 M tetrabutylammonium hexafluorophosphate (TBAPF6) (from Fluka, electrochemical grade) as a supporting electrolyte. The electrochemical properties of complexes 1 and 2 were studied by cyclic voltammetry (CV) on glassy carbon electrode (GCE) (2 mm in diameter A = 0.0314 cm 2 (Mineral, Warsaw)).
Some experiments were performed with the use of differential pulse voltammetry (DPV) on carbon fiber (CF) disk microelectrode (33 µm in diameter (BASi, United Kingdom)). DPV voltammograms were registered using a pulse amplitude of 20 mV, pulse width of 80 ms and scan rate of 20 mV s -1 . This technique is considered a convenient method because of its good sensitivity selectivity and resolution of the signals, limited influence of adsorption phenomena on recorded curves and thus excellent reproducibility [5].

Crystal structure determination and refinement (XRD)
Diffraction intensity data for single crystal of complex 1 was collected at room temperature on a KappaCCD (Nonius) diffractometer with graphite-monochromated MoK radiation ( = 0.71073 A). Corrections for Lorentz, polarization and absorption effects [6,7] were applied. The structure was solved by direct methods using the program package SIR-92 [8] and refined using a full-matrix least square procedure on F 2 using SHELXL-2016/6 [9,10]. Anisotropic displacement parameters for all non-hydrogen atoms and isotropic temperature factors for hydrogen atoms were introduced. In the structure the hydrogen atoms connected to carbon atoms were included in calculated positions from the geometry of molecules. In the crystal lattice of 1, the presence of two hydronium cations was observed.
There is no indications in the difference density map as to the location of the H atoms belong to O(0). Diffraction intensity data for single crystal of complex 2 and L 2 2 were collected at 120 K on the Oxford Diffraction Super Nova diffractometer using monochromatic Mo Kα radiation, λ = 0.71073 Å. Cell refinement and data reduction were performed using firmware. [11] Positions of all of non-hydrogen atoms were determined by direct methods using SHELXL-2016/6 [9,10]. All non-hydrogen atoms were refined anisotropically using weighted full-matrix least-squares on F 2 . Refinement and further calculations were carried out using SHELXL-2016/6 [9,10]. All hydrogen atoms joined to carbon atoms were positioned with an idealized geometries and refined using a riding model with Uiso(H) fixed at 1.2 Ueq (Carom). The positions of all hydrogen atoms were constrained for all compounds S5 using AFIX (SHELXL) commands. The positions of hydrogen atoms of water molecules (complex 2) have been refined with restrains for ideal water molecules. The final structure models have been refined with constrains for positions of first time refined water hydrogen atoms with restrains. The crystallographic data and detailed information on the structure solution and refinement for the ruthenium complexes (1 and 2) and L 2 2 are given in Table   S9. The figures were made using DIAMOND [12] software. CCDC 1059868, 1996910 and 1996909 contain the supplementary crystallographic data for 1, 2 and L 2 2, respectively.
These data can be obtained free of charge from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
The XRD investigation was carried out on a DRON-2 (Russia) diffractometer connected to an IBM computer, stepwise, over the 2 angle range 10-75°, using CuK radiation. The products of decomposition were studied using X-ray powder method identified on the basis of ICDD using XRAYAN package [13].

Hirshfeld surface analysis
Molecular Hirshfeld surfaces calculations were performed using the Crystal Explorer package ver.
3.1 [14]. When the .cif file of the title compounds was entered into the Crystal Explorer program, all of the bond lengths to hydrogen were automatically modified to the standard neutron values (CH = 1.083 A). Hirshfeld surface analysis included the descriptor dnorm and the shape index [15]. The calculations and details of analysis were made as described in [4]. The molecular Hirshfeld surfaces of complexes 1 and 2 were generated using a standard (high) surface resolution with the 3D dnorm surfaces mapped over a fixed colour scale of −0.372 (for 1)/−0.313 (for 2) (red) to 1.385 (for 1)/1.108 (for 2) Å (blue). The shape index was mapped in the colour range of −1 to 1 (for 1 and 2).
The colour encodes normalized distance to nearest nuclei and thus conveniently illustrates the "strength" of all types of the intermolecular contacts present. In turn, the fingerprint plots provide a quantitative measure of the intermolecular interactions on the surface.

Minimal inhibitory concentration
A broth microdilution method was used to determine the minimum inhibitory concentrations of the tested samples of the compounds. The ruthenium complexes and the ligands were prepared by dissolving compounds in distilled water. The stock concentrations of tested compounds were 2 mM.
The serial two-fold dilutions were made in a concentration range from 1 mM to 0.0625 mM in the sterile 96-well microtiter transparent plates (Greiner, Monroe, NC, USA) containing nutrient broth.
After that, diluted suspensions were added to appropriate wells. The inoculated plates were incubated at 37°C for 24 h. The negative control (bacterial culture in the medium) and positive control (antibiotic controlstreptomycin) were used as references to determine the growth inhibition of bacteria. The MIC parameter was recorded as the lowest concentration of the compound at which the isolate was completely inhibited (as evidenced by the absence of visible bacterial growth). The experiments were performed using the Infinite M200 PRO microplate reader (Tecan, Männedorf, Switzerland). Tests were conducted as three independent repeats.

Inhibition of biofilm formation
The inhibition effect of the tested compounds on biofilm formation by P. aeruginosa PAO1 and LES B58 strains was measured by crystal violet method using 96-well microtiter plates [16]. The amount of biofilm formed was determined as described previously [4]. Stock solutions of test compounds were prepared in distilled water. The final concentrations of compounds in the cell cultures were in the range 0.0625 -1 mM. Additionally, fresh medium was used as a negative control and streptomycin as a positive control. Absorbance of the eluted crystal violet was measured on an Infinite M200 PRO microplate reader at wavelength of 595 nm (Tecan, Männedorf, Switzerland). Assays were performed at least in three independent experiments.
The measurement results, expressed in absorbance units, were converted into percentages to allow the comparison of numerical data obtained in different experiments.

Live/Dead staining of the bacterial biofilm
Fluorescence microscopy was used to image live/dead cells in the P. aeruginosa PAO1 biofilm.
First, the P. aeruginosa PAO1 biofilm was cultivated in 6-well microtiter plates on glass coverslips S7 in TSB medium at 37°C for 24 h without shaking. Then, the culture was supplemented with solutions of the ruthenium complexes (concentration: 1 mM). After 24-hour incubation, the coverslips were carefully washed with sterile water in order to remove nonadherent cells.
Microcolonies formed on the glass surface were stained with a FilmTracer™ LIVE/DEAD® Biofilm Viability Kit (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol.
After 15 minutes incubation at room temperature in the dark, the samples were washed with water to remove the excess dyes. Images were collected with a ZEISS Axio Scope.A1 epifluorescence microscope. The experiments were repeated three times to obtain consistent results.

Statistical analysis
Statistical analysis was performed using one-way analysis of variance (ANOVA). Significance was set at p<0.05.

S17
The diffuse reflectance spectrum of HL1 is shown in Figure S2A. In the UV region, the high-intensity bands observed at approximately 213 and 269 nm in the benzimidazole derivative are due to intra-ligand transitions (π-π*) [19]. In the reflectance spectrum of complex 1 (Figure S2A), the relatively intense bands at approximately 209, 247 and 270 nm can be assigned to intra/inter-ligand transitions from the heteroaromatic moieties, whereas moderately intense absorption bands in the and visible part of the spectrum (354, 385 and 443 nm) are attributed to ligand-to-metal charge transfer (LMCT) from the chloride ion to the metal centre. The lower intensity broad band in the 500-600 nm region is probably assigned to the d-d transitions (547 nm) for ruthenium(IV) due to the state of the d 4 ion.
The data obtained for the reflectance spectrum of complex 1 are in agreement with the experimental data in aqueous solution. The electronic spectrum of HL1 displayed two broad absorption bands ( Figure S2B, Table S6) that can be assigned to π→π* transitions in the delocalized π-electron system [20]. The split bands at 268 and 275 nm appear as doublets due to the probable existence of a tautomeric structure [21], as supported by comparing our spectrum with that of benzimidazole derivatives [22]. The electronic spectrum of complex 1 displayed three distinct absorption bands in water ( Figure S2B). The bands at 226, 274 and 288 nm may be assigned to the low-energy π→π* transitions within the benzimidazole moieties [19]. The UV-Vis spectrum is also characterized by a band at ~360 nm and by a second weaker band at ~460 nm (Table S6, Figure S2B). The former bands can be ascribed to πCl→t2gRu LMCT (Ru-Cl) transitions involving the four coplanar chlorides [23]. The last absorption band in the visible part of the spectrum with a maximum at ~590 nm is attributed to the d-d transition. According to the Tanabe-Sugano diagram, the d-d transition has been assigned to the 3 T1g→ 3 Eg transition for low-spin Ru(IV) complex (t2g 4 eg 0 configuration) in an Oh environment.
More complicated electronic spectra were recorded for the ligand L 2 2 and complex 2 ( Figure   S3). Analysis of the UV-Vis spectrum of ligand L 2 2 indicated that strong or moderate intensity bands in the 200 -390 nm region are related to the intra-ligand π→π*/n→π* transitions. It is noteworthy that in the process of complexation, the ligand was transformed into a diketone, and the bands corresponding to the n→π* transitions are in the 280 -350 nm region in the spectrum of compound 2. The next bands (380, 393, 453 nm) in the electronic spectrum of the complex were assigned to CT transitions (π(L)→d(Ru)).