In vitro antibacterial activity of dinuclear thiolato-bridged ruthenium(II)-arene compounds

ABSTRACT

The antibacterial activity of 22 thiolato-bridged dinuclear ruthenium(II)-arene compounds was assessed in vitro against Escherichia coli, Streptococcus pneumoniae, and Staphylococcus aureus. None of the compounds efficiently inhibited the growth of the three E. coli strains tested, and only compound 5 exhibited a medium activity against this bacterium [MIC (minimum inhibitory concentration) of 25 µM]. However, a significant antibacterial activity was observed against S. pneumoniae, with MIC values ranging from 1.3 to 2.6 µM for compounds 1–3, 5, and 6. Similarly, compounds 2, 5–7, and 20–22 had MIC values ranging from 2.5 to 5 µM against S. aureus. The tested diruthenium compounds have a bactericidal effect significantly faster than that of penicillin. Fluorescence microscopy assays performed on S. aureus using the BODIPY-tagged diruthenium complex 15 showed that this type of metal compound enters the bacteria and does not accumulate in the cell wall of gram-positive bacteria. Cellular internalization was further confirmed by inductively coupled plasma mass spectrometry experiments. The nature of the substituents anchored on the bridging thiols and the compounds molecular weight appears to significantly influence the antibacterial activity. Thus, if overall a decrease of the bactericidal effect with the increase of compounds’ molecular weight is observed, however, the complexes bearing larger benzo-fused lactam substituents had low MIC values. This first antibacterial activity screening demonstrated that the thiolato-diruthenium compounds exhibit promising activity against S. aureus and S. pneumoniae and deserve to be considered for further studies.

IMPORTANCE

The in vitro assessment of diruthenium(II)-arene compounds against Escherichia coli, Streptococcus pneumoniae, and Staphylococcus aureus showed a significant antibacterial activity of some compounds against S. pneumoniae, with minimum inhibitory concentration (MIC) values ranging from 1.3 to 2.6 µM, and a medium activity against E. coli, with MIC of 25 µM. The nature of the substituents anchored on the bridging thiols and the compounds molecular weight appear to significantly influence the antibacterial activity. Fluorescence microscopy showed that these ruthenium compounds enter the bacteria and do not accumulate in the cell wall of gram-positive bacteria. These diruthenium(II)-arene compounds exhibit promising activity against S. aureus and S. pneumoniae and deserve to be considered for further studies, especially the compounds bearing larger benzo-fused lactam substituents.

INTRODUCTION

The increasing occurrence of multidrug-resistant (MDR) bacteria and difficult-to-treat infections associated with high morbidity and mortality have become important medical issues (1 – 3). Yearly, the number of estimated deaths caused by antibiotic-resistant infections is 33,000 in Europe and 35,000 in the United States (4). In 2019, on a global scale, 1.27 million deaths were directly attributable to drug resistance (5). Rapid and efficient solutions are required to counter potential outbreaks, with terrible life costs and economical damages (1, 6). A list of the world’s leading antibiotic-resistant bacteria with an urgent demand for new antibiotics was published by World Health Organization (WHO) (7), aiming not only to favor the surveillance and control of MDR bacteria but also to favor research of new active compounds. Antibiotic-resistant Enterobacteriaceae (encompassing Escherichia coli), Staphylococcus aureus as well as Streptococcus pneumoniae are included in the WHO’s list (7).

To efficiently face the problem of bacterial resistance to antibiotics [intrinsic and/or acquired (8)], the improvement of already existing antibiotics should be paralleled by the development of new classes of active compounds against which bacteria are less prone to develop resistances (9).

Metal complexes reached clinical trials for the treatment of various diseases and some compounds show also interesting antibacterial activity (10 – 12). However, the use of metal-based compounds as antibiotics is not widespread, and only a limited number of compounds have undergone clinical trials (13). Examples of metal complexes active against Gram-negative and Gram-positive bacteria (E. coli and S. aureus) are provided in Table S1 (13 – 32). Ruthenium is one of the metals considered for the development of metal-based antibiotics being generally associated to reduced toxicity (10). Some ruthenium complexes exert good activity against Gram-positive bacteria and low activity toward Gram-negative bacteria, notable exceptions being dinuclear polypyridylruthenium(II) complexes (31, 33) and ruthenium-based carbon-monoxide-releasing molecules (34 – 36).

Di- and trithiolato-bridged dinuclear ruthenium(II)-arene complexes have been investigated for more than a decade. This type of diruthenium complex is generally stable (37, 38) and showed promising in vitro and in vivo anticancer activity (37, 39) and in vitro antiparasitic activity against Toxoplasma gondii (40 – 44) and Trypanosoma brucei (45). Their exact mode of action has not yet been elucidated, but a profound alteration of the structure and activity of mitochondria has been identified (40, 41, 45).

To identify further biological applications of dinuclear thiolato-bridged ruthenium(II)-arene complexes, a library of 22 compounds was screened for potential antibacterial properties. This library included both previously reported and newly synthesized compounds with significant structural diversity: dithiolato compounds, symmetric and mixed trithiolato compounds, and conjugates containing one or more carbohydrate units, short lipid chains, benzo-fused lactams, and fluorophores. The minimum inhibitory concentration (MIC) values of the compounds were measured against various strains of clinically relevant Gram-negative (E. coli) and Gram-positive bacteria (S. pneumoniae and S. aureus). Additional experiments to evaluate the bacteriostatic and bactericidal properties of the compounds as well as their internalization and cellular localization were also performed.

RESULTS

Chemistry

Based on their structural features, the compounds included in this study were organized into four families.

Di- and trithiolato diruthenium compounds 1–8 (Family 1)

Family 1 comprises eight previously known diruthenium complexes (1–8, Fig. 1) and includes the neutral dithiolato compound 4, two symmetric trithiolato diruthenium complexes 5 and 6 with benzyl and phenyl substituents on the bridging thiols, and five mixed trithiolato diruthenium complexes 1–3, 7, and 8 in which various structural elements were varied. Some of these compounds have previously shown high anticancer and antiparasitic activity (44, 46, 47). For example, compound 1 exhibited nanomolar range IC₅₀ (half maximal inhibitory concentration) values against various parasites as Neospora caninum (41) and T. gondii (40).

Fig 1.

Structure of compounds 1–8 forming Family 1.

Compounds 9–12—diruthenium complexes conjugated to hexanoic acid and to carbohydrates (Family 2)

In this family, four trithiolato diruthenium compounds with organic molecules anchored on one of the bridge thiols were also selected. Compounds 9 and 10 (48) contain a hexanoic acid residue connected via an ester and, respectively, an amide bond to the diruthenium unit, and they were chosen considering that the lipophilic chain could influence membrane transfer (Fig. 2).

Fig 2.

Structure of diruthenium conjugates 9–12 forming Family 2.

Compounds 11 and 12 present acetyl-protected D-glucose and D-galactose units, respectively (Fig. 2), anchored to one of the bridge thiols via a triazole linker, and they were selected in view of a potentially facilitated bacteria internalization due to the presence of the carbohydrate unit. The synthesis, characterization, and T. gondii antiparasitic effects of compounds 9–12 have been previously described (48, 49).

Compounds 13–16—diruthenium complexes conjugated to BODIPY fluorophores (Family 3)

The development of fluorophore-labeled conjugates of metal-based drugs as traceable therapeutic agents has become extremely popular as they can provide important information relative to compounds cellular uptake, localization, and specific accumulation (42, 50). BODIPY (boron dipyrromethene, 4,4-difluoro-4-bora-3a,4a-diaza-s-indacene) is among the most attractive fluorophores tagged as they are generally non-toxic, photostable, and exhibit high fluorescence quantum yields (50).

Four fluorescent conjugates, 13–16, containing a BODIPY dye unit attached to the diruthenium moiety through linkers of various lengths via ester (compounds 13 and 15) or amide (compounds 14 and 16) bonds (Fig. 3), were selected. Although an important fluorescence quenching was observed after conjugating the BODIPY to the diruthenium unit, compounds 13–16 could be used as fluorescent tracers (51) and were recently investigated as potential anti-toxoplasma agents. Fluorescence microscopy investigations of these compounds showed a pattern of cytoplasmic, but not nuclear, localization in human foreskin fibroblasts (HFFs) (51).

Fig 3.

Structure of diruthenium-BODIPY conjugates 13–16 forming Family 3.

Compounds 17–22—diruthenium compounds with various thiol ligands (ortho-substituted arenes and benzo-fused lactams as pending groups; Family 4)

Six new thiolato diruthenium compounds 17–22 (Fig. 4) were synthesized to evaluate structurally different ligands. The description of the synthetic protocols and the analysis and characterization of these compounds and the corresponding ligands are provided in the Supporting information [Fig. 5 and 6 for the synthesis of the diruthenium complexes, and Schemes S1 and S2 (Supporting information) for the synthesis of the thiol ligands].

Fig 4.

Structure of compounds 17–22 forming Family 4. 23 is the benzo-fused lactam thiol ligand used in compounds 20 and 21 and tested against various bacteria.

Fig 5.

Synthesis of mixed trithiolato diruthenium complexes 17–20 and 22.

Fig 6.

Synthesis of the symmetric trithiolato diruthenium complex 21.

In compounds 17–19 (Fig. 5), three different groups, diphenylphosphonate (17), 2–4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl (18), and aldehyde (19) were introduced in the ortho position rather than in para position of one of the bridging thiol-ligand. The synthesis of the ligands 17a [diphenyl-(2-mercaptophenyl)phosphonate], 18a [2-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)benzenethiol], and 19a (2-mercaptobenzaldehyde; Schemes S1) is presented in detail in the Supporting information.

For compounds 20–22 (Fig. 5), two benzo-fused lactams containing thiol groups [6-mercapto-3,4-dihydroquinolin-2(1H)-one (22 a) and 7-mercapto-1,3,4,5-tetrahydro-2H-benzo(b)azepin-2-one (23)] were used as bridging ligands instead of 4-mercaptophenol in the parent trithiolato diruthenium compound 1. Their synthesis is provided in detail in the Supporting information and were selected because benzo-fused lactams exhibit various biological activities among which antimicrobial properties (52, 53).

In the symmetric trithiolato diruthenium complex 21 (Fig. 6), the benzyl and phenyl bridging thiols in complexes 5 and 6 were replaced by three 7-mercapto-1,3,4,5-tetrahydro-2H-benzo[b]azepin-2-one units.

Asymmetric trithiolato diruthenium complexes 17–20 and 22 were synthetized by reacting the dithiolato intermediate 4 with an excess of the corresponding ligands 17a–19a, 22a, and 23 (in 1,2-dichloroethane -DCE- at 100°C or in dichloromethane at 60°C). In the case of complexes 17–19, excess of triethylamine was employed to ensure the completion of the reaction. All the compounds were isolated in moderate yields 41%–65% (Fig. 5).

The symmetric trithiolato diruthenium complex 21 was obtained by reacting the commercially available ruthenium dimer {[Ru(η ⁶-p-MeC₆H₄Pr ⁱ )Cl]₂Cl₂} with an excess of the thiol ligand 23 in refluxing EtOH in basic conditions (K₂CO₃) and was isolated in 56% yield (Fig. 6).

Stability of the diruthenium compounds

For the biological activity evaluation, 1 mM stock solutions of all compounds were prepared in dimethylsulfoxide (DMSO). ¹H-NMR spectra of several conjugates dissolved in DMSO-d ₆, recorded at 25°C 5 min, and 100 days after sample preparation showed no modifications, demonstrating very good stability of the diruthenium compounds in this highly complexing solvent (42, 44, 51).

Compounds 9, 11–13, and 15 present a carboxyl ester bond that can potentially be hydrolyzed in cellular growth media. Compounds 13 and 15 as well as other similar conjugates with coumarin and BODIPY fluorescent units have been recently studied (42, 51), and for these compounds, fluorescence measurements proved that the ester bonds remained intact for 48 h, but very limited hydrolysis of the ester bonds with the release of coumarin and BODIPY dyes after 168 h could be observed. It was concluded that diruthenium conjugates with ester linkers were stable under the conditions used for biological evaluations, and therefore, it was assumed that compounds 9, 11–13, and 15 are sufficiently stable during the in vitro evaluation.

Determination of the MIC values on E. coli

The MIC values (lowest concentration of antibiotic at which bacterial growth is completely inhibited) of the compounds were determined on three E. coli strains [P53.1R (54), P54.1T (54), and P54.2R (54)], the S. pneumoniae D39 (NCTC 7466) (55) strain, and the S. aureus 20 (54) strain. Of note, the P53.1R and P54.2R strains were extended-spectrum beta-lactamases resistant, while the P54.1T strain was colistin resistant. A table (Table S1) summarizing the nature and resistance profiles of the three strains is included in the Supplementary Information.

The first screening of the 22 diruthenium compounds was conducted on E. coli P53.1R, a colistin sensitive but extended spectrum cephalosporin resistant strain (Table 1). Most of the compounds exhibited no or very low antibacterial activity on this Gram-negative bacterium [MIC values over or at the limit of the concentration range (0.01–100 μM) used in the assays]. From this library, the only compound inhibiting E. coli P53.1R cells growth with medium activity (MIC value of 25 µM) was the symmetric diruthenium compound 5 (Table 1). The potency of compounds 2, 5, and 6 from Family 1 was further assessed for two additional E. coli strains P54.1T and P54.2R (Table 1). On the E. coli P54.1T and P54.2R strains, the lowest MIC values (of 25 and 20.8 µM, respectively) were obtained for the same compound 5 (Table 1), which inhibited the bacteria growth with medium efficiency compared with the reference drug colistin for which the MIC value is 0.8 µM.

TABLE 1.

MIC values measured for the diruthenium compounds 1–22, the ligand 23, and colistin and penicillin used as controls ^c

MIC (µM and μg/mL)
	E. coli			S. pneumoniae D39 (NCTC 7466) b	S. aureus 20 a
Compound	P53.1R a	P54.1T a	P54.2R a
Family 1
1	>80.0	n.d.	n.d.	2.5 ± 1.0 (2.5 ± 1.0)	17.5 ± 5.0 (17.3 ± 4.9)
2	100.0 ± 0.0 (98.9 ± 0.0)	100.0 ± 0.0 (98.9 ± 0.0)	100.0 ± 0.0 (98.9 ± 0.0)	1.8 ± 0.5 (1.8 ± 0.5)	12.5 ± 5.0 (12.4 ± 5.0)
3	>80.0 (>82.5)	n.d.	n.d.	2.6 ± 0.8 (2.7 ± 0.8)	5 ± 0.0 (5.2 ± 0.0)
4	>80.0 (>72.0)	n.d.	n.d.	17.5 ± 5.0 (15.8 ± 4.5)	12.5 ± 5.0 (11.3 ± 4.5)
5	25.0 ± 0.0 (21.9 ± 0.0)	25.0 ± 0.0 (21.9 ± 0.0)	20.8 ± 7.2 (18.2 ± 6.3)	1.3 ± 0.6 (1.1 ± 0.5)	4.4 ± 1.3 (3.9 ± 1.2)
6	100.0 ± 0.0 (83.4 ± 0.0)	100.0 ± 0.0 (83.4 ± 0.0)	66.7 ± 28.9 (55.6 ± 24)	1.3 ± 0.6 (1.1 ± 0.5)	2.5 ± 0.0 (2.1 ± 0.0)
7	>80.0 (>76.9)	n.d.	n.d.	6.7 ± 2.9 (6.4 ± 2.8)	5.0 ± 0.0 (4.8 ± 0.0)
8	>80.0 (>83.2)	n.d.	n.d.	10.0 ± 0.0 (10.4 ± 0.0)	6.7 ± 2.9 (7.0 ± 3.0)
Family 2
9	>80.0 (>87.0)	n.d.	n.d.	10.0 ± 0.0 (10.9 ± 0.0)	40.0 ± 0.0 (43.5 ± 0.0)
10	>80.0 (>87.0)	n.d.	n.d.	> 10.0 (>10.9)	80.0 ± 0.0 (87.0 ± 0.0)
11	80.0 ± 0.0 (115.5 ± 0.0)	n.d.	n.d.	10.0 ± 0.0 (10.4 ± 0.0)	10.0 ± 0.0 (14.4 ± 0.0)
12	80.0 ± 0.0 (115.5 ± 0.0)	n.d.	n.d.	10.0 ± 0.0 (10.4 ± 0.0)	8.3 ± 2.9 (12.0 ± 4.2)
Family 3
13	>80.0 (>112.3)	n.d.	n.d.	n.d.	80.0 ± 0.0 (112.3 ± 0.0)
14	>80.0 (>115.7)	n.d.	n.d.	n.d.	>80.0 (>115.7)
15	>80.0 (>114.6)	n.d.	n.d.	n.d.	80.0 ± 0.0 (114.6 ± 0.0)
16	>80.0 (>108.9)	n.d.	n.d.	n.d.	>80.0 (>108.9)
Family 4
17	>80.0 (>96.5)	n.d.	n.d.	n.d.	40.0 ± 0.0 (48.2 ± 0.0)
18	>80.0 (>88.0)	n.d.	n.d.	n.d.	8.8 ± 2.5 (9.7)
19	>80.0 (>80.1)	n.d.	n.d.	n.d.	15.0 ± 5.8 (15.0 ± 5.8)
20	100.0 ± 40 (105.7)	n.d.	n.d.	n.d.	3.8 ± 1.4 (4.0 ± 1.5)
21	>80.0 (>86.6)	n.d.	n.d.	n.d.	2.5 ± 0.0 (2.7 ± 0.0)
22	>80.0 (>83.4)	n.d.	n.d.	n.d.	3.1 ± 1.3 (3.2 ± 1.3)
Ligand and reference drugs
23	>240.0 (>46.4)	n.d.	n.d.	n.d.	>30.0 (>5.8)
Colistin	0.8 ± 0.6 (0.9 ± 0.5)	5.8 ± 2.0 (6.7 ± 2.3)	1.1 ± 2.0 (1.3 ± 2.4)	n.d.	n.d.
Penicillin	n.d.	n.d.	n.d.	<0.02 (0.01)	0.09 ± 0.03 (0.03 ± 0.01)

^a Information in reference (54).

^b Information in reference (55).

^c The reported values are the means of three biological replicates ± SD. “n.d.” stands for “not determined.” The lowest MIC values for each strain are highlighted in bold. The values in brackets are the values expressed in μg/mL

Determination of the MIC values on S. pneumoniae

In a second screening assay, the MIC values of diruthenium complexes 1–12 on S. pneumoniae D39 strain were determined, proving to be significantly lower compared to those measured on the three E. coli strains (Table 1). Compounds 5 and 6 were most effective (MIC value of 1.3 µM for both compounds) and slightly more active than compounds 1–3 (MIC values of 1.8–2.6 µM). The MIC values corresponding to compounds 7 and 8 were at the limit of the concentration used for the assay (0.625–10 μM; Table 1) and being more than five times higher compared to the MIC value of compounds 5 and 6. From Family 1, the dithiolato derivative 4 exhibited the lowest activity on S. pneumoniae D39. The MIC values of compounds 9–12 (Family 2) on S. pneumoniae D39 were higher than those found trithiolato compounds 1–3 and 5–7.

Determination of the MIC values on S. aureus

In the third screening, MIC values of the 22 compounds on S. aureus 20 strain were also evaluated (Table 1). In Family 1, compound 6, with an MIC value of 2.5 µM, was the most potent, followed by compounds 3, 5, and 7 (MIC values ranging from 4.4 to 5 µM). Hydroxy compound 1 was the least efficient (MIC values of 17.5 µM), followed by the amino derivative 2, and the dithiolato compound 4 (MIC value of 12.5 µM for both). Interestingly, compared to 1, slight structural modifications in compound 7 ( ⁱ Pr group instead of ^t Bu as substituents on two of the bridging thiols) lead to an improvement of activity against S. aureus 20 (MIC values of 17.5 and 5.0 µM for 1 and 7, respectively). A similar effect was observed in the case of compound 8 compared to 2 for which the presence of BF₄ ⁻ as counterion instead of Cl⁻ was associated to a decrease of the MIC value on S. aureus 20 from 12.5 to 6.7 µM for 2 and 8, respectively.

Interestingly, for compounds 1 and 2, important differences of activity are observed on the two tested Gram-positive bacteria S. aureus 20 and S. pneumoniae D39.

In Family 2, conjugates 9 and 10, with a hexanoic acid residue attached on one of the bridging thiols via ester or amide bonds, showed poor activity on S. aureus 20 (MIC values of 40 and 80 µM, respectively). Interestingly, the acetyl-protected glucose and galactose hybrids 11 and 12 exhibited lower MIC values (11 µM and 8.3 µM, respectively) but are up to four times less efficient in inhibiting the bacteria growth compared to 6.

All the diruthenium compounds conjugated to BODIPY fluorophores (Family 3, compounds 13–16) exhibited low activity against S. aureus 20 (MIC values over 80 µM).

Contrasting MIC values were obtained for compounds 17–22 (Family 4). From compounds 17–19, bearing ortho-substituted ligands, derivative 17 with a diphenyl-(2-mercaptophenyl)phosphonate bridging ligand, exhibited reduced efficacy on S. aureus 20 (MIC value of 40 µM), while the boron-containing compound 18 and aldehyde compound 19 exhibited significantly lower MIC values (8.8 µM and 15 µM, respectively). Compounds 20–22, bearing at least one benzo-fused lactam unit on the bridging thiols, exhibited low MIC values of 3.8, 2.5, and 3.1 mM, respectively.

The antibacterial activity of ligand 23 [7-mercapto-1,3,4,5-tetrahydro-2H-benzo(b)azepin-2-one] was also evaluated on S. aureus, using three times the concentration employed in the tests with compound 21, but no bacteria growth inhibition was measured within the concentration range of the assay (7.5–240 μM).

The average MIC values of the 22 diruthenium compounds on the tested bacteria as a function of their respective molecular weight are shown in Fig. S10, as well as the linear regression for each bacterial species. Interestingly, Fig. S10 suggests a slight inverse correlation between MIC values and molecular weight for all bacterial species investigated in the present study. While this trend hardly exists for the E. coli phenotype P53.1R (only five compounds are considered), it appears evident for S. pneumoniae D39 and especially for S. aureus 20).

Toxicity

The toxicity of compounds 1–16 against HFFs has been previously evaluated (44, 48, 49, 56). The data have been compiled in Table S2. Briefly, compounds 3, 5, 6, 12, 13, 15, and 16 were classified as very promising, with no toxic effect on HFF cells at a high concentration of 2.5 µM, (survival rate of HFF >90%) compounds 1, 2, 7, and 8 were moderately toxic (survival rate >50%), and compounds 9 and 10 were very toxic on HFF (survival rate 0%).

Evaluation of the bacteriostatic and bactericidal properties

The bacteriostatic and bactericidal properties of selected diruthenium compounds 1–3, 5–8, and 11–12 were investigated on S. aureus 20 (Fig. 7). Interestingly, the compounds exhibited a significantly faster effect compared to penicillin, even when the penicillin concentration was increased to eight times its MIC value (250 ng/L). The sharp decrease of the optical density at 600 nm (OD₆₀₀) after the addition of all diruthenium compounds and the complete absence of cell growth within 24 h suggests a bactericidal rather than bacteriostatic effect.

Fig 7.

Growth curves of S. aureus 20 after the addition of the diruthenium compounds at MIC when OD₆₀₀ = 0.5 is reached (dotted lines). (A) S. aureus treated with compounds 1–2 and 5–6 at MIC upon reaching OD₆₀₀ = 0.5. (B) S. aureus treated with compounds 8, 11, and 12 at MIC upon reaching OD₆₀₀ = 0.5. (C) S. aureus treated with compounds 3 and 7 at MIC upon reaching OD₆₀₀ = 0.5. Treatment with penicillin at four times its MIC was used as a bactericidal control.

Fluorescence microscopy

To determine whether the compounds preferentially accumulate inside the bacteria or on their membrane, conjugate 15 bearing a fluorescent BODIPY tag was used to investigate the localization of the diruthenium compounds by fluorescence microscopy. The hexanoic ester conjugate 9 was also tested in the fluorescence microscopy experiments as a negative control (the fluorescence of 9 at 525/550 nm is shown in Fig. 8, panels B). In parallel with compounds 9 and 15, DAPI (4,6-diamidin-2-phenylindol) stain was used for the visualization of nuclear DNA.

Fig 8.

Fluorescence microscopy images of S. aureus 20 cells stained with DAPI after 3 h of incubation, with or without treatment with diruthenium conjugates 9 and 15. (A) Untreated control sample stained with DAPI. (B) The sample was treated with 9 for 3 h and stained with DAPI for visualization. (C) The sample was treated with 15 for 3 h and stained with DAPI for visualization. 15 is visible after excitation at 440/470 nm with emission at 525/550 nm. White arrows indicate cells with overlapped fluorescence of DAPI and BODIPY diruthenium conjugate 15. Green arrows indicate emission in 525/550 nm (corresponding to conjugate 15) that do not overlap with DAPI nor were cells visible under Brightfield.

In Fig. 8 (panels C), an overlap of the blue fluorescent DAPI marker and the green fluorescence of BODIPY conjugate 15 inside the cells is clearly visible. Interestingly, emissions from 15 which do not overlap with those of DAPI are also observed (Fig. 8, panels C). The results only validate the nuclear DNA as intracellular target of DAPI. The relative low fluorescence signal of compound 15 (51) does not allow to clearly distinguish the exact intracellular compound localization which might be the cytoplasm or the inner/outer membrane. Additional images are presented in Fig. S12 to S15 in Supporting information.

ICP-MS and cell counting

The cellular uptake of compounds 5, 6, 9, 15, and 21 in the S. aureus 20 strain was quantified using inductively coupled plasma mass spectrometry (ICP-MS) measurements, the results being summarized in Table 2. The bacteria were incubated with the selected ruthenium compounds at their respective MIC value for 1, 2, and 3 h. A clear increase of the ruthenium content of treated S. aureus 20 cells compared to the untreated control [mean value (6.6 ± 2.2) × 10⁻¹⁷ µM/CFU(colony-forming unit)] was measured, indicating cellular uptake of the diruthenium compounds (Table 2).

TABLE 2.

Content of ruthenium in S. aureus 20 cells measured by ICP-MS ^a

Compound	Incubation time (h)	Average compound concentration per S. aureus cells (µM/CFU)
5	1	(7.7 ± 2.0) × 10−13
	2	(4.2 ± 1.5) × 10−13
	3	(4.1 ± 1.3) × 10−13
6	1	(4.0 ± 1.2) × 10−13
	2	(1.4 ± 1.4) × 10−12
	3	(1.4 ± 1.1) × 10−12
9	1	(2.8 ± 1.9) × 10−13
	2	(2.4 ± 1.8) × 10−13
	3	(2.6 ± 1.3) × 10−13
15	1	(2.2 ± 1.7) × 10−13
	2	(3.6 ± 2.7) × 10−13
	3	(3.4 ± 2.6) × 10−13
21	1	(6.6 ± 0.6) × 10−14
	2	(5.9 ± 0.3) × 10−14
	3	(6.1 ± 1.2) × 10−14

^a Values are the means of the three replicates ± SD.

The amount of compound internalized by the cells appears to be independent of the treatment duration (Table 2). If overall, within the treatment time frame of 3 h, no significant time-dependent modifications were observed for the same compound, important variations between the tested compounds were noticed. For example, the measured ruthenium concentration in S. aureus 20 cells after treatment with compound 6 (1.4 × 10⁻¹² µM/CFU) was more than 20 times higher compared with the value for compound 21 (6 × 10⁻¹⁴ µM/CFU).

The results suggest a possible correlation between the amount of diruthenium compound internalized by the bacteria and the molecular weight, but no dependence on the respective MIC values (Fig. S11). Compound 21, bearing three 7-mercapto-1,3,4,5-tetrahydro-2H-benzo(b)azepin-2-one ligands as bridging thiols, is particularly interesting. 21 exhibits a low MIC value (2.5 µM, Table 1) but less efficient cellular uptake. These data suggest that 21 is highly toxic to bacteria, and this compound can be considered as a lead for further compound development through appropriate chemical modifications.

DISCUSSION

In this work, the in vitro properties of a library of 22 trithiolato-bridged dinuclear ruthenium(II)-arene complexes as potential antibacterial agents against relevant bacterial species were investigated. The in vitro antiparasitic and anticancer activity of compounds 1–16 have been previously studied (40, 41, 44, 57). This type of diruthenium complex is known to target in the mitochondria (40, 41, 44, 45, 58) an organelle sharing many features with bacterial cells (59).

Following the trend of most antibiotics [being more effective against Gram-positive than on Gram-negative pathogens (8, 12)], the results of this study revealed that the MIC values of the 22 tested compounds were higher on E. coli strains than on S. pneumoniae and S. aureus. Moreover, compared to colistin and penicillin tested as reference drugs, even the most effective compounds in this library were only moderately active, showing MIC values more than 10 times higher.

For comparison, the MIC values of the ruthenium compounds cis-α-[Ru(phen)bb₁₂]²⁺ and cis-β-[Ru(phen)bb₁₂]²⁺ against the S. aureus MSSA ATCC 25,923 strain were 0.5 µM, and against the E. coli ATCC 25922 strain they were 8.3 and 16 µM, respectively (24). On the other hand, the MIC values of the ruthenium compound [Ru(bpy)₂(methionine)]²⁺ against the S. aureus MSSA ATCC 25,923 strain and against the E. coli ATCC 11303 strain were of 73.3 and 586.4 µM, respectively (21).

The physico-chemical properties of the bridging thiol ligands are very important, their influence appearing particularly important in compounds 20–22 bearing benzo-fused lactams substituents, which exhibited low MIC values, and to a lesser extent, for carbohydrate conjugates 11–12. The neutral dithiolato complex 4 had medium MIC values against E. coli and S. pneumoniae, following the trend observed with other dithiolato complexes against cancer cells and parasites (40, 47, 57). The lack of activity against cancer cells and parasites was attributed to the lability of the chloride ligands and the general lower stability of dithiolato compounds (60, 61). Yet, the MIC value of 4 against S. aureus was relatively low and comparable to that of the mixed trithiolato compounds 1 and 2, compounds that were otherwise much more active than dithiolato complexes against cancer cells and protozoan parasites (40, 62). This difference is appealing and needs further investigations.

The cellular uptake of the BODIPY-diruthenium conjugate 15 by S. aureus was confirmed by fluorescence microscopy investigations. The images presented in Fig. S15 (Supporting information) indicate that compound 15 does not have a specific cellular localization. Fig. S15 clearly shows that compound 15 does not accumulate in membranes or cell walls but concentrates inside cells, forming aggregates that potentially lead to cell death.

In conclusion, the results show that the lower molecular weight diruthenium compounds tend to have lower MIC values against Gram-positive bacteria, but the presence of specific substituents, especially the 1,3,4,5-tetrahydro-2H-benzo(b)azepin-2-one unit, on the bridge thiols strongly influence the antibacterial activity. The 7-mercapto-1,3,4,5-tetrahydro-2H-benzo(b)azepin-2-one is an interesting ligand to be considered for the development of other potent antibacterial metal complexes. Finally, the localization and the cellular uptake of the compounds assessed using fluorescence microscopy and ICP-MS, respectively, suggest that the cellular walls as well as the nucleic acid or protein synthesis are not the main targets of this type of compounds.

MATERIALS AND METHODS

Materials

All new diruthenium complexes were synthesized, analyzed, and characterized at the Department of Chemistry, Biochemistry, and Pharmaceutical Sciences of University of Bern according to the methods provided in the Supporting information. Colistin, penicillin G potassium salt, and gentamicin sulfate were purchased from Sigma-Aldrich and used as received.

Instrumentation

The OD₆₀₀ of cell cultures and of the tubes used for the assays were measured with a Helios Epsilon spectrophotometer. In the MIC assays, the absorbance values for wells plates were obtained using a Varioskan Flash spectral scanning multimode plate reader using the SkanIt software. Cell lysis was realized using a Branson Ultrasonics Sonifier 250. Fluorescence microscopy observations were conducted using a ZEISS AXIO Imager M1, fluorescence microscope. ICP-MS values were determined with a NexION 2000B ICP Mass Spectrometer using Ir-193 as internal standard.

Antibacterial activity

Cell culture

In this work were used three extended-spectrum cephalosporin-resistant E. coli strains (P53.1R, P54.1T, and P54.2R) with varied colistin sensitivity, a penicillin-sensitive strain of S. pneumoniae (D39) and a penicillin-sensitive strain of S. aureus (20). The E. coli and S. aureus strains were inoculated in Luria-Bertani agar plates and incubated overnight at 37°C. S. pneumoniae was inoculated in Columbia blood sheep agar (CSBA) plates and incubated overnight at 37°C and 5% CO₂. Individual colonies of E. coli were then subcultured overnight at 37°C under constant agitation in Luria-Bertani medium for initial screening and in Brain Heart Infusion (BHI) for the other assays. After the incubation period, the bacterial cultures were adjusted to a final concentration of 2 × 10⁷ CFU/mL. Individual colonies of S. pneumoniae and S. aureus were subcultured in BHI at 37°C until they reached the appropriate concentrations of 1 × 10⁸ and 2 × 10⁸ CFU/mL, respectively. The final concentration for inoculation in the assays was 5 × 10⁵, 5 × 10⁶, and 1 × 10⁷ CFU/mL, respectively.

Determination of the MIC

The assays were conducted in sterile Nunclon 96 microwell plates and Nunc 384 well polystyrene plates purchased from Thermo Fisher Scientific. DMSO stock solutions of the diruthenium complexes (1–5 mM) were diluted in H₂O to concentrations ranging from 2 to 200 µM according to compounds’ potency and dispensed into the plates with the previously adjusted bacterial suspension. The plates with E. coli or S. aureus were then incubated at 37°C for 18 h, and then the OD at 450 nm was measured using a Varioskan Flash, spectral scanning multimode plate reader. The plates with S. pneumoniae were incubated at 37°C for 22 h with agitation and measurement every 30 min. The MIC was determined as the lowest compound concentration showing a complete inhibition of visible bacterial growth. For comparison, all assays included antibiotic controls: colistin, 8 µg/mL (E. coli P53.1R and P54.2R), 64 µg/mL (E. coli P54.1T), and penicillin 0.5 µg/mL (S. pneumoniae and S. aureus), as well as DMSO controls to account for the lethality of the organic solvent.

Determination of the bacteriostatic/bactericidal effects

The assay was conducted in test tubes read with a Helios Epsilon spectrophotometer. DMSO stock solutions of the diruthenium complexes were diluted in H₂O to concentrations 50 times their MIC values, before being added into S. aureus bacterial suspension in the test tube at 1:50 of the total volume, thus reaching MIC. The tubes were incubated at 37°C in a water bath for 32 h. At the beginning of the assay, the OD of each sample was measured every hour until an OD₆₀₀ = 0.5 was reached. At this point, the previously prepared solution of diruthenium complexes and antibiotics was quickly added to their respective tubes and the negative growth controls before continuing incubation. Subsequently, the OD was measured for all samples every half hour until a total incubation time of 8 h was reached. Between 8 and 24 h of incubation, the OD was not measured and then measured sporadically between 24 and 32 h of incubation to observe potential changes until the end of the assay.

Cell counting

First, in a test tube, 5 mL BHI was added to 266.6 µL of a suspension of S. aureus in BHI and incubated at 37°C until an OD₆₀₀ = 0.5 was measured. Subsequently, sequential dilutions were prepared, by adding 900 µL of a 0.85% NaCl aqueous solution to 100 µL from the suspension of S. aureus in BHI. This dilution procedure was repeated until a dilution of 10⁻⁹ was reached. 100 µL of the resulting 10⁻³ to 10⁻⁹ solutions were then put on CSBA plates and incubated at 37°C and 5% CO₂ for 18 h. The CFU/mL of the original suspension was then estimated by counting the number of colonies in each plate and using a calibration curve. Each measurement was realized in triplicate.

Fluorescence microscopy

Samples were prepared by incubating S. aureus bacterial suspensions for 3 h at 37°C in a water bath. Then 5 µM of compounds 9 and 15 were added, and the suspensions were further incubated for 1, 2, and 3 h in 5 mL of BHI. Next, the samples were sedimented by centrifugation at 1,788 g for 12 min at 4°C and washed twice with 0.85% aqueous NaCl. Then, the pellets were re-suspended in 3 mL 0.85% aqueous NaCl and cooled on ice before adding 2 µL of DAPI [DAPI Solution (1 mg/mL), Thermo Scientific, 4′,6-diamidino-2-phenylindole, a fluorescent stain that binds strongly to adenine–thymine-rich DNA regions] stock solution. The samples were finally studied using a fluorescence microscope with filters for DAPI (excitation 365 nm and emission 445/50), GFP (green fluorescent compound, excitation 440/70, and emission 525/5) and brightfield. To ensure observation of the cell internal, Z-stacks pictures were taken with an interval of 320 nm, using an objective 100×, 0.065 µm/pixel.

Inductively coupled plasma mass spectrometry

The samples were prepared by incubating S. aureus bacterial suspensions (~4.3 × 10⁸ CFU/mL) for 3 h at 37°C in a water bath. Then 1 mM solution of compounds 5, 6, 9, 15, and 21 in DMSO was added, and the suspensions were incubated for 1, 2, and 3 h. A 150 µg/mL penicillin solution in distilled water was used to ensure cell death. The samples were then sedimented by centrifugation at 1,788 g for 12 min at 4°C and washed with 0.85% aqueous NaCl twice. Afterward, the pellets were resuspended in 3 mL H₂O and frozen at −20°C overnight before undergoing heat shock and sonication for cell lysis. The resulting samples were then filtered using 0.22 µm Millipore syringe filters and analyzed by ICP-MS.

Determination of cellular uptake of the ruthenium compounds

Using the data obtained from the cell counting and ICP-MS experiments and assuming comparable losses during filtration for both samples and controls, the intracellular ruthenium content was calculated using Equation 1:

$${\displaystyle \mu \mathrm{g}\mathrm{R}\mathrm{u}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{c}\mathrm{e}\mathrm{l}\mathrm{l}\mathrm{s}=\frac{\mathrm{S}\mathrm{a}\mathrm{m}\mathrm{p}\mathrm{l}\mathrm{e}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\left(\mathrm{p}\mathrm{p}\mathrm{b}\mathrm{R}\mathrm{u}=\mu \mathrm{g}\mathrm{R}\mathrm{u}/\mathrm{L}\right)/\mathrm{F}\mathrm{i}\mathrm{l}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{l}\mathrm{o}\mathrm{s}\mathrm{s}\mathrm{e}\mathrm{s}\left(\mathrm{\%}\right)}{10\mathrm{m}\mathrm{L}/3\mathrm{m}\mathrm{L}\ast \mathrm{O}\mathrm{D}600\ast 4.3\cdot {10}^8\mathrm{C}\mathrm{F}\mathrm{U}/\mathrm{m}\mathrm{L}}}$$ (1)

A calibration curve using ruthenium solutions of 0.02 ppb, 0.1 ppb, 1 ppb, 5 ppb, and 50 ppb in 2% HNO₃ was first established, and Iridium 193 was used as internal standard.

DATA AVAILABILITY

All the data are presented in this study and in the corresponding Supporting information.

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/spectrum.00954-23.

Supplemental material. spectrum.00954-23-s0001.docx.

Figures, tables, and spectra.

Table S3. spectrum.00954-23-s0002.xlsx.

Nature and resistance profiles of the three E. coli strains used in the study.

OPEN PEER REVIEW. reviewer-comments.pdf.

An accounting of the reviewer comments and feedback.

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.