Synthesis and characterization of bismuth(III) complex with an EDTA-based phenylene ligand and its potential as anti-virulence agent

Synthesis of the ligand

The edtabz ligand was synthesized as previously described by Beltran-Torres et al. (2019). Specifically, 2.4 g (9.3 mmol) of EDTA dianhydride dissolved in 8 mL of dry dimethylformamide (DMF) was added to 2 mL (20 mmol) of aniline previously distilled. The resulting reaction mixture was left to stand overnight and any solids were filtered out. The filtrate was concentrated to 5 mL by a rotary evaporator, into which acetone was added. Precipitates formed were filtered off, washed with acetone several times until a colorless solid was obtained. Finally, it was vacuum-dried for 8 h at 25 °C. The purity was confirmed by IR and ¹H NMR spectra and determining the decomposition point.

Synthesis of the bismuth complex

The bismuth complex was synthesized by reaction of the appropriate nitrate salt and ligand. In 15 mL of water, 1 mmol of edtabz was suspended and solubilized by adding Li₂CO₃. This solution was mixed with 1 mmol of nitrate bismuth (10 mL aqueous solution, dissolved with a few drops of nitric acid), no visible change of color was observed. After stirring was continued for 30 min and the solution was heated at 40–50 °C. When it was concentrated to half the initial volume and stabilized to 25 °C, the product precipitated. A white powder was filtered off and vacuum-dried for 8 h at 25 °C. Yield: 0.4146 g, 54%. mp 238 °C (dec). Found: C, 33.18%; H, 3.12%; N, 10.49%. Calcd for BiC₂₂H₂₄O₆N₄ • 2NO₃: C, 33.39%; H, 3.31%; N, 10.62%. Mass spectrum (ESI−) m/z: 709.6 (100%), [(M − H)−] (The mass spectrum of the complex is given in Fig. S1).

Spectroscopic measurements

The bismuth complex was characterized in terms of ¹H and ¹³C NMR, UV-Vis, Fourier-transform infrared (FTIR), mass spectra of electrospray ionization (ESI/MS) and thermogravimetric analysis (TGA) (Beltran-Torres et al., 2019). The ¹H and ¹³C NMR spectra were obtained with a Bruker AVANCE 400 spectrometer for D₂O solutions at 25 °C. The internal reference was sodium 2,2-dimethyl-2-silanpentane-5-sulfonate (DSS). IR spectra were recorded on a Perkin-Elmer FT-IR Spectrometer Model Frontier equipped with an ATR accessory. UV-Visible spectroscopy was carried out using Perkin-Elmer Lambda 20. For the pH-variable experiment of the Bi-edtabz mixture, the sample was dissolved in 0.1 M NaCl, and the pH values of the sample solutions were adjusted with 0.1 M HCl or 0.1 M NaOH to keep the ionic strength and sample concentration constant. A quartz cuvette of the spectrometer was loaded with the basic solution of the Bi-edtabz, and proper amounts of the acid complex solution were added to have a pH gradient of 0.5 in the final reading of the acid complex solution.

Mass spectra of electrospray ionization (ESI/MS) were obtained on 6,130 Quadrupole LC/MS of Agilent Technologies in the negative ionization mode. Thermogravimetric Analysis (TGA) was carried out on a thermogravimetric analyzer Perkin Elmer Pyris 1 TGA to study the complex’s thermal stability and composition. Five mg of sample was set in a ceramic pan and analyzed at a 25−800 °C range under an O₂ atmosphere.

Antimicrobial activity of ligand and bismuth complex

The antibacterial activity of edtabz and the bismuth complex was evaluated as previously reported by Vázquez-Armenta et al. (2021). The tested pathogenic bacteria were Escherichia coli O157: H7 (ATCC 43895), Listeria monocytogenes (ATCC 7644), Pseudomonas aeruginosa (ATCC 10154), Salmonella enterica sub. enterica serovar Typhimurium (ATCC 14028) and Staphylococcus aureus (ATCC 6538). An inoculum of 1 × 10⁸ CFU/mL for each bacterium was obtained from exponential phase cultures in nutrient broth (Luria-Bertani, Brain Heart Infusion or Tryptic Soy broths). Then, 5 µL of inoculum were added to a sterile 96-well microplate (Costar 96), followed by 295 µL of each compound at different concentrations (0–1 mM) diluted in the corresponding nutrient broth to achieve a final inoculum level of 1 × 10⁶ CFU/mL. The microplate was incubated for 24 h at 37 °C. Bacterial growth in the presence of ligand or bismuth complex was inspected visually. The highest tested concentration (1 mM) did not inhibit the growth of the evaluated bacteria; thus, 0.25, 0.5 and 1 mM were selected for further analyze the anti-virulence activity.

Effect of ligand and bismuth complex on biofilm formation of pathogenic bacteria

The capacity of bismuth complex and edtabz to prevent biofilm formation of pathogenic bacteria was evaluated by the crystal violet staining procedure as previously reported by Vázquez-Armenta et al. (2021). Bacterial inoculum was prepared at 1 × 10⁸ CFU/mL in their corresponding broth from an exponential phase culture. Two µL of the bacterial inoculum and 150 μL of each compound at different concentrations (0, 0.25, 0.5 and 1 mM) were taken and placed in sterile 96-well polystyrene microplates (Costar 96) and incubated for 24 h at 37 °C. After the incubation period, the culture was removed by aspiration, and the wells were washed three times with distilled water and let dry for 15 min. Subsequently, the formed biofilms were stained by adding 200 µL of 0.1% (w/v) crystal violet to each well (45 min at 25 °C). Then, wells were washed gently three times with distilled water to remove the unbound dye and dried for an additional 15 min. The bound dye to biofilms was solubilized by adding 200 µL of 20% acetic acid for 15 min. Finally, the optical density (OD) was measured at 580 nm in a FLUOstar Omega spectrophotometer (BMGLabtech, Chicago, IL, USA). Nutrient broth without any compound and bacteria was used as a blank, and the OD values were subtracted from treatment readings. Each experiment was carried out in triplicate, and results were expressed as percentages of inhibition compared with control samples (biofilms grown without compounds) following the formula:

$\mathrm{P}\mathrm{e}\mathrm{r}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{a}\mathrm{g}\mathrm{e}\mathrm{i}\mathrm{n}\mathrm{h}\mathrm{i}\mathrm{b}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}(\mathrm{\%})={\displaystyle \frac{O{D}_{positivecontrol}-O{D}_{treatment}}{O{D}_{positivecontrol}}\times 100}$

Effect of bismuth complex on swimming motility

The effect of bismuth complex and ligand on swimming motility of pathogenic bacteria was evaluated by exposing 10 μL of bacterial suspensions growth overnight (18 h, at 30 °C) to the presence of 1 mM of each compound. The treated bacteria were placed in the center of Petri dishes that contained soft agar (0.3% agar), and untreated bacteria were used as controls. The Petri dishes were incubated at 30 °C, and the diameter (mm) of bacterial motility halos was measured after 24 h of incubation (Bernal-Mercado et al., 2020). Each experiment was performed in triplicate.

Screening for anti-quorum sensing activity

The biomonitor strain Chromobacterium violaceum (ATCC 12472) was used for preliminary screening of anti-quorum sensing activity of the bismuth complex. For this purpose, the disk diffusion assay was employed where the inoculum of C. violaceum was grown aerobically in Lauria-Bertani (LB) broth at 30 °C for 18 h and adjusted to 1 × 10⁸ CFU/mL. Then LB agar plates were spread with 100 μL of C. violaceum inoculum, and 20 µL of 2 mM bismuth complex or ligand solution was loaded to the sterile disks and placed on the surface of inoculated LB agar plates. The plates were incubated upright for 24 h at 30 °C, and a colorless appearance indicated the inhibition zone of QS (Alvarez et al., 2014).

Experimental design and statistical analysis

A complete randomized design was performed, the evaluated factor was the concentration of compounds (0.25, 0.5 or 1 mM), and the response variable was biomass reduction of biofilms (%) and motility halos (mm). An analysis of variance (ANOVA) was carried out to estimate significant differences among treatments (p ≤ 0.05), and the Tukey-Kramer test was used for means comparison (p ≤ 0.05) in the software NCSS 2007.

¹H NMR spectra edtabz and Bi-edtabz

The edtabz ligand shows the typical derivative EDTA signals, Ha, Hb and Hc. Due to the symmetry of the ligand, only five signals in the ¹H spectrum are seen (Fig. 1). Once the bismuth ion is coordinated, the equivalence of the aliphatic protons and the aromatic ones is lost. In Fig. 1A, protons from the aromatic ring in the free ligand are chemically equivalent, and just two signals for both the aniline rings are shown. Meanwhile, in the Bi-edtabz, the two aromatic rings can be distinguished (Fig. 1B). The ¹³C NMR spectra Bi-edtabz complex also confirms such a phenomenon. Nine carbon signals are observed in the ¹³C spectrum of the edtabz ligand (Fig. 2A), while seventeen signals are shown in the complex spectrum (Fig. 2B). To further understand the chemical environment, coordination mode of the edtabz with Bi³⁺ and the proper protons assignment, a Heteronuclear Simple Quantum Coherence (HSQC) experiment was performed (Full HSQC spectrum is given in Fig. S2, and the integrated signals of the complex is given as well, Fig. S4).

In the y-axis, the aromatic carbons (Fig. 3), Cortho, Cmeta and Cpara are shown. Six signals can be seen, three for each aniline ring present in the ligand. In the x-axis, a series of multiplets are shown, which are difficult to assign, but the 2D experiment makes it easier. The multiplets that seem to appear as a doublet at ~7.5 ppm come from the aromatic protons in the ortho position. The multiplet at 7.43 ppm is a contribution of ortho as well as the meta protons. Following the dots in the HSQC experiment, the protons from para and meta form the multiplet at 7.28 ppm, and finally, the multiplet at 7.20 ppm comes from para protons.

The aliphatic protons can be observed in the graph’s x-axis in the region from 5.2 to 3.5 ppm (Fig. 4). As previously mentioned, the number of signals observed, and the pattern displayed can be attributed to a system that has gone from a flexible backbone to a rigid one. The proton at 3.65 and 4.50 ppm are assigned to Ha protons. The Hb protons are assigned to the signals at 4.29 and 4.37 ppm. These protons are sign to the methylene protons neighboring the carboxylate group in the acid; based on previous information, it can be speculated that one arm coordinates from above while the other coordinates from below or a side (Beltran-Torres et al., 2019). These protons are the ones that suffer the most from the complexation; hence the pattern of the doublet with uneven height can be seen. At 4.02, and 5.05 ppm, the doublets are attributed to Hc protons (J = 16 Hz). A doublet can be seen at 3.84 ppm with a J = 16 Hz, but no correlation in the spectrum can be seen. The small doublets signals at 3.86 and 3.41 ppm (J = 12 Hz) can be attributed to an equilibrium of species when instead of the two carbonyl groups are coordinating, only one is.

Thermal analysis

The thermal stability of the Bi-edtabz was analyzed by TG analysis. The complex weight loss was monitored while the temperature increased at 10 °C min⁻¹ from 25 °C up to 800 °C, under an air atmosphere (Fig. 5). The thermal analysis of the bismuth complex included four decompositions steps (180–562 °C). The first decomposition step (180–220 °C) involves a 7.4% mass loss due to the evolution of a nitrate molecule. The second decomposition step (220–242 °C) involves a loss mass of 7.4%, which could be attributed to another nitrate ion. This result indicates that one NO₃ molecule is part of the outer sphere of the complex (the nitrate molecule in the first decomposition step), and the other nitrate molecule is part of the inner sphere coordination; therefore, the second decomposition step is attributed to this nitrate molecule (Taha et al., 2011; Radecka-Paryzek, Pospieszna-Markiewicz & Kubicki, 2007; Anjaneyulu, Prasad & Swamy, 2010). The third and fourth stages of decomposition (242–374 °C and 374–562 °C) with 38.6% and 16.7% mass loss are attributed to the ligand evolution. The remaining 29% loss corresponds to bismuth residues.

FTIR spectra of Bi-edtabz

The ligand and the complex’s infrared spectrum were taken to understand further the structure of the complex and functional groups involved in the bismuth coordination. Some informative bands of the ligand and its complex are given in Table 1. In the free ligand, the strong characteristic stretching band C=O of carboxylic acid appeared at 1,700 cm⁻¹ and moved to the lower wavenumber side by 80 cm⁻¹ upon complexation. This band shift indicates that the carboxylate groups participated in the formation of the complex. Other bands displaced to lower wavenumber sides upon complexation are the vibrational bands of amide C=O and C−N bonds, which indicates that the amide groups were also involved in the coordination metal ion (The FTIR spectra of bismuth complex and free ligand are given in Fig. S3).

UV-Visible spectra of Bi-edtabz

At 242 nm, the edtabz ligand shows a band attributed to transition π–π* proper from the aromatic rings. In the bismuth complex, there are two bands, one attributed to the interaction of the ligand’s aromatic rings, which shows a hyperchromic effect at 242 nm and 263 nm at a pH of 1.0 (Fig. 6A). The band at 263 nm is attributed to the metal-ligand transfer charge. The tendency of bismuth to form hydroxides prevented the attempt to titrate bismuth nitrate to ligand to corroborate the molar ratio of complex because the bismuth hydroxide precipitated as a white powder. In Fig. 6B, the pH-variable experiment is presented. At acid pH, the complex showed two shoulders, and as the pH became basic, the metal-ligand transfer charge band decreased, and the spectrum at pH 12 is the same as the free ligand. A dissociation occurred, which could be to the formation of bismuth hydroxides.

Effect of Bi-edtabz and edtabz on pathogenic bacteria in planktonic state and biofilm

Visual inspection of the bacterial cultures that grew in the presence of ligand or bismuth complex showed that both compounds at the highest concentration evaluated (1 mM) did not affect the viability of the bacterial cells. Thus, concentrations of 0.25, 0.5 and 1 mM were chosen to evaluate its effect on biofilm formation of pathogenic bacteria using the crystal violet assay where the amount of bound dye is proportional to the total biomass of biofilms (Burton et al., 2007). Figure 7 shows the percentages of biomass reduction of biofilms formed in the presence of edtabz or Bi-edtabz complex. Both compounds inhibited biofilm development in a dose-dependent manner and presented different inhibition patterns among bacterial species. Bi-edtabz complex at 1 mM showed higher biomass reduction against E. coli O157: H7 (p = 0.0029) and S. aureus (p = 0.014) (58% and 55%, respectively) than the ligand (43% and 37%, respectively). The ligand edtabz caused higher biomass reduction in L. monocytogenes (p = 0.000067), P. aeruginosa (p = 0.0021) and Salmonella Typhimurium (p = 0.0409) than Bi-edtabz complex. At 1 mM, edtabz completely prevented biofilm formation of L. monocytogenes and reduced by 88% and 67% the biomass of P. aeruginosa and Salmonella Typhimurium biofilm.

Effect of Bi-edtabz and edtabz on swimming motility

In Table 2, the effect of edtabz and Bi-edtabz on the swimming motility of pathogenic bacteria is presented. After 24 h of incubation at 30 °C untreated pathogenic bacteria displaced on the surface of soft agar showing motility zones from 40.9 to 78.1 mm. The presence of Bi-edtabz complex in the culture media impaired the motility of E. coli O157: H7 (p = 0.00029) and S. aureus (p = 0.00001) with reductions of 12.5% and 47.2%, respectively. While the motility of L. monocytogenes, P. aeruginosa and Salmonella Typhimurium was not affected (p > 0.05) by the bismuth complex. On the other hand, edtabz only affected (p = 0.00001) the motility of Salmonella Typhimurium, showing 41.9% inhibition compared to untreated bacteria.

Anti‑ quorum sensing activity of Bi-edtabz and edtbaz in Chromobacterium violaceum model

To explore the anti-quorum sensing activity of the bismuth complex as preliminary screening, biomonitor strain C. violaceum was used. This bacterial model produces a purple pigment (violacein) during the quorum-sensing activation; this production is regulated in response to self-produced acyl-homoserine lactones (AHL). Thereby, analyzing the pigment production in the bismuth complex presence, it is possible to evaluate the interference of a given substance in this process (Alvarez et al., 2014). As shown in Fig. 8B, Bi-edtabz inhibited the violacein production of C. violaceum as evidenced by the colorless halo (11.9 ± 0.4 mm), absent in edtabz treatment (Fig. 8A).