Challenges in Application of Langmuir Monolayer ... - ACS Publications

Nov 20, 2017 - Challenges in Application of Langmuir Monolayer Studies To .... Low-Molecular-Weight Supramolecular Ionogel Based on Host–Guest Inter...
0 downloads 0 Views 3MB Size
Download PDF
Article Cite This: Langmuir 2017, 33, 14167−14174

pubs.acs.org/Langmuir

Challenges in Application of Langmuir Monolayer Studies To Determine the Mechanisms of Bactericidal Activity of Ruthenium Complexes B. Sandrino,*,† J. F. A. de Oliveira,‡,§ T. M. Nobre,† P. Appelt,∥ A. Gupta,⊥ M. P. de Araujo,∥ V. M. Rotello,⊥ and O. N. Oliveira, Jr.† †

São Carlos Institute of Physics, University of São Paulo, CP 369, 13560-970 São Carlos, São Paulo, Brazil National Laboratory of Synchrotron Light (LNLS), CP 6192, 13083-970 Campinas, São Paulo, Brazil § Institute of Chemistry, State University of Campinas (Unicamp), CP 6154, 13083-970 Campinas, São Paulo, Brazil ∥ Department of Chemistry, Federal University of Paraná, CP 19081, 81531-980 Curitiba, Paraná, Brazil ⊥ Department of Chemistry, University of Massachusetts Amherst, 01003 Amherst, Massachusetts, United States ‡

S Supporting Information *

ABSTRACT: The effects induced by antibiotics on the bacterial membrane may be correlated with their bactericidal activity, and such molecular-level interactions can be probed with Langmuir monolayers representing the cell membrane. In this study, we investigated the interaction between [Ru(mcbtz)2(PPh3)2] (RuBTZ, mcbtz = 2mercaptobenzothiazoline) and [Ru(mctz)2(PPh3)2] (RuCTZ, mctz = 2-mercaptothiazoline) with Langmuir monolayers of a lipid extract of Escherichia coli, an extract of lipopolysaccharides (LPSs), and a zwitterionic phospholipid, dioleoylphosphatidyl choline (DOPC). RuBTZ and RuCTZ had little effects on DOPC, which is consistent with their negligible toxicity toward mammalian cells that may be approximated by a zwitterionic monolayer. Also little were their effects on LPSs. In contrast, RuBTZ and RuCTZ induced expansion in the surface pressure isotherms and decreased the compressional modulus of the E. coli lipid extract. While the more hydrophobic RuBTZ seemed to affect the hydrophobic tails of the E. coli extract monolayer to a larger extent, according to polarization modulation infrared reflection absorption spectroscopy results, evidence of a stronger RuBTZ interaction could not be confirmed unequivocally. Therefore, the interaction with the E. coli cell membrane cannot be directly correlated with the observed higher bactericidal activity of RuBTZ, in comparison to that of RuCTZ. This appears to be a case in which Langmuir monolayer studies do not suffice to determine the mechanisms responsible for the bactericidal activity. catalysts in biological11 and nonbiological systems.12 Ruthenium has unique properties desirable for drug design, including its ability to bind nucleic acids and proteins, and a range of accessible oxidation states.4 Furthermore, ruthenium drugs present low toxicity owing to their iron-mimicking property when bound to biomolecules.4,6 Some ruthenium complexes also show antibacterial properties, which should be related to their effects on the bacterial membrane, but the molecular mechanisms involved are not known in detail.4 Four mechanisms have been suggested in the literature to govern the antibacterial action. In three of those mechanisms, there is participation of membrane proteins, more specifically with the inhibition or regulation of enzymes relevant for

1. INTRODUCTION The indiscriminate use of antibiotics has contributed to the emergence and development of drug-resistant microorganisms,1 which has brought the need to develop antimicrobials with novel structures for controlling bacterial and microbial infectious diseases.2,3 Metal-based drugs, for instance, have been studied in medicinal chemistry,4,5 with many transition metal complexes displaying biological activity. Metal complexes are advantageous because they can provide more stereochemical variability owing to their range of coordination geometries, thus allowing for introducing different organic molecules that may facilitate interaction and biorecognition. For example, platinum compounds have been used in cancer treatment,6 silver compounds possess antibacterial properties,7,8 and gold compounds can be used for treating rheumatoid arthritis.6 Ruthenium-based compounds, in particular, exhibit significant anticancer9 and antifungicidal10 activity, in addition to serving as © 2017 American Chemical Society

Received: June 29, 2017 Revised: October 18, 2017 Published: November 20, 2017 14167

DOI: 10.1021/acs.langmuir.7b02247 Langmuir 2017, 33, 14167−14174

Article

Langmuir

Lipids. The LPSs (rough strains) from E. coli J5 (Rc mutant) were purchased from Sigma-Aldrich. All of these compounds were used without further purification. 2.2. Determination of Antimicrobial Activities of RuBTZ and RuCTZ. Experiments to determine the minimal inhibitory concentration (MIC) were conducted with CD-2 strain (E. coli). CD-2 was cultured overnight in minimal M9 medium at 37 °C and speed of 275 rpm until reaching the stationary phase and then harvested by centrifugation and washed three times with 0.85% sodium chloride solution. The optical density at 600 nm was used to determine the concentration of resuspended bacterial solution. M9 medium was used to make dilution of bacterial solution to a concentration of 1 × 106 cfu/ mL. Then, a volume of 50 μL of this solution was added into a 96-well plate and mixed with 50 μL of RuBTZ or RuCTZ solutions in M9, yielding a final bacterial concentration of 5 × 105 cfu/mL. For preparing the initial solutions with the complexes, 2 mg were dissolved in 1 mL DMSO, and then 128 μL of these solutions were mixed with M9 to complete 400 μL to obtain the concentration of 0.64 mg/mL for each complex. From this solution (0.64 mg/mL), a new one with half of concentration 0.32 mg/mL was obtained and then dilutions were made to have concentrations ranging from 160 to 1.25 μg/mL. Neat dimethyl sulfoxide (DMSO) (without any complex) was also tested to evaluate its possible bactericidal effect. A growth control group without Ru complexes and a sterile control group with only growth medium were used. All the experiments were conducted in triplicate and performed simultaneously. 2.3. Cytotoxicity Assays for Mammalian Cells. 3T3 cells were maintained in high glucose Dulbecco’s modified Eagle’s medium (ATCC 30-2002) supplemented with 10% fetal bovine serum (Fisher Scientific, SH3007103), 100 units/mL penicillin, and 100 μg/mL streptomycin (Gibco). The cells were grown in 96-well plates at 37 °C with 5% CO2 atmosphere incubation. Approximately 1.5 × 104 3T3 cells per well were seeded onto 96-well plates. After 24 h, the medium was replaced with a fresh one containing the Ru complexes. Incubation for 24 or 48 h followed, and then cell viability was determined using alamarBlue assay according to the manufacturer’s protocol (Invitrogen BioSource). After washing with a phosphate-buffered saline (PBS) solution three times, cells were treated with 120 μL of 10% alamarBlue in serum-containing media and incubated at 37 °C under an atmosphere of 5% CO2 for 3 h. The solution from each well was transferred into a 96well black microplate. The red fluorescence, resulting from the reduction of alamarBlue solution, was quantified (excitation/emission: 560 nm/590 nm) on a SpectraMax M5 microplate reader (Molecular Devices) to determine cell viability. Cells without any complex were considered as 100% viable. All experiments were conducted in triplicate, performed simultaneously, and the reported values are averages. The cell viability was expressed as a percentage relative to the control according to eq 1.

biosynthesis in the cell wall, protein synthesis, and nucleic acid metabolism and repair.13 The fourth mechanism is unique in the sense that it does not involve other membrane components rather than those responsible for the membrane structure. This action mechanism is based on disrupting the membrane structure,14 which is especially relevant to fight highly resistant bacteria because the latter cannot develop resistance against antibiotics that are capable of destroying or causing leakage in the membrane.15 For Gram-negative bacteria, this mechanism is rather complex because the antibiotics have to disrupt both the inner, lipidic membrane and the outer bacterial wall made of lipopolysaccharides (LPSs).15 To cross the LPS barrier, antibiotics use porin channels to reach the periplasmic space, where they can bind to their targets and develop their activity. Porins are beta barrel proteins that cross the membrane and are responsible for the passive diffusion of molecules, including the influx of antibiotics (mostly hydrophilic and smaller than 1500 Da). Testing the probable action of drugs via the membrane disruption mechanism is thus crucial for drug design. Experiments with whole cell membranes to verify such a mechanism are not possible with present technologies, and therefore, cell membrane models such as Langmuir monolayers16 and vesicles are used.17 Even though the Langmuir monolayers mimic only half a membrane, they are convenient owing to the possible control of lateral packing and membrane composition. Furthermore, studies can be made with drugs that are watersoluble or not, as the drug may be injected into the subphase or co-spread with the monolayer-forming material. In this study, we used Langmuir monolayers to simulate the inner plasmatic membrane and the outer membrane of Escherichia coli to investigate the effects from two ruthenium complexes, namely, [Ru(mcbtz)2(PPh3)2] complex, mcbtz = 2-mercaptobenzothiazoline, referred to as RuBTZ (Figure 1A) and [Ru-

Cell viability (%) =

Figure 1. (A) Ruthenium complex with 2-mercaptobenzothiazoline (RuBTZ) and (B) with 2-mercaptothiazoline (RuCTZ).

fluorescencesample fluorescencecontrol

× 100

(1)

where fluorescencesample is the sample fluorescence in the presence of the complex and fluorescencecontrol is the fluorescence of the control group (untreated cells). 2.4. Langmuir Monolayers. Langmuir monolayers were produced by spreading the solutions of the compounds onto the surface of Milli-Q water in a Mini KSV Langmuir trough (KSV Instruments Ltd., Helsinki, Finland) equipped with a Wilhelmy plate as the surface pressure sensor. DOPC, LPSs, and E. coli extract were dissolved in CHCl3. The solutions for the monolayers incorporating the Ru complexes were obtained by mixing appropriate volumes of stock solutions made of 10 mg/mL for the Ru complex solubilized in DMSO and 2 mg/mL for E. coli extract (or LPSs or DOPC) in CHCl3. The relative concentrations of 8 and 16% of the Ru complexes indicate the amount of material needed to replace the lipid or LPSs in the film, with a final solution concentration of 2 mg/mL. Aliquots of 40 μL of these solutions were then spread onto the water surface with a micrometric syringe. For LPSs, PBS at pH 7.4 was used as the subphase because stable LPS films could not be formed on ultrapure water. The solvent was allowed to evaporate for ca. 10 min before

(mctz)2 (PPh 3 ) 2 ] complex, mctz = 2-mercaptothiazoline (RuCTZ, Figure 1B). The aim is to try and correlate the effects of these complexes on the cell membrane model with their inhibition of bacterial growth18 and bactericidal activity. As we shall show, in spite of a thorough comparative study with the complexes and three types of monolayer, the monolayer studies are not sufficient to establish such correlation.

2. EXPERIMENTAL SECTION 2.1. Materials. All chemicals used were of reagent grade or comparable purity, and the solvents were purchased from SigmaAldrich. The complexes RuBTZ and RuCTZ were synthesized through a well-established route.18 The E. coli extract (PE 57.5 wt %, PG 15.1 wt %, CL 9.8 wt % and unknown content 17.6 wt %) and 1,2-dioleoyl-snglycero-3-phosphocholine (DOPC) were acquired from Avanti Polar 14168

DOI: 10.1021/acs.langmuir.7b02247 Langmuir 2017, 33, 14167−14174

Article

Langmuir

Figure 2. (A) Optical density at 600 nm, which is correlated with the bactericidal activity of the Ru complexes and DMSO. Lower density means higher bactericidal effect. The gray bar corresponds to the control sample, for which no bactericidal effect exists. The values for RuBTZ (red bars), RuCTZ (blue bars), and DMSO (black bars) at different concentrations refer to bacterial growth inhibition after incubation for 24 h. (B) Cell viability for 3T3 mammalian cells for incubations for 24 and 48 h into solutions containing varied concentrations of RuBTZ (red), RuCTZ (blue), and DMSO (black). The data are compared to the control group with 100% viability. Data shown are the mean for each condition ± SD. Statistical analysis was performed according to the t-test. The letters on top of the bars (a, b and c) indicate mean P < 0.05 for the same group. Letters in common denote that the data are not statistically different. monolayer compression. The surface pressure versus area per molecule (π−A) isotherms were recorded with symmetrical compression of monolayers with two barriers at a constant speed of 10 mm min−1 (∼5 Å2 molecule−1 min−1). The experiments were performed at 20 ± 1 °C, and the isotherms were made in duplicate. The effect of neat DMSO on the membrane model was also evaluated, and the results are given in the Supporting Information. In this experiment, rather than using 16% of CHCl3 as in the monolayers with E coli, DOPC, and LPS, DMSO was used to verify its effect by mimicking the DMSO volume used for 16% mixed films. 2.4.1. Polarization Modulation Infrared Reflection Absorption Spectroscopy. Polarization modulation infrared reflection absorption spectroscopy (PM-IRRAS) measurements were taken with a KSV PMI 550 instrument (KSV Instruments Ltd., Helsinki, Finland). The experimental setup used was similar to that described by Geraldo et al.19 The Langmuir trough is placed in a way that the light beam reaches the monolayer at a fixed incidence angle of 80°, at which the intensity is maximum with a low level of noise. The incoming light is continuously modulated between s- and p-polarization at a high frequency (50 kHz), which allows for the simultaneous measurement of the spectra for the two polarizations. The difference between the spectra provides surfacespecific information, and the sum gives the reference spectrum. The effect from water vapor is reduced with the simultaneous measurements. The mixed films studied had 8% of Ru complexes in E. coli. 2.4.2. Brewster Angle Microscopy. The Brewster angle microscopy (BAM) images were obtained using an ellipsometer Accurion EP4 with ultra objective, camera SVS-Vistek eco 285, laser at 658 nm and 50 mW with 100% of power. The angle of observation was 53.1°, polarization at 2°, and analyzer at 10°. The principles of this technique are explained in ref 20. The geometric correction in the images was performed with software DataStudio, applying in Y-axis a factor of 1.6655 and producing images with 450 × 570 μm. During the measurements, the greyscale was adjusted depending on the reflectivity because the image of the interfacial film is formed by the contrast of regions with lower film concentrations (dark regionswithout reflection) and spots where the water surface is covered with high film concentrations (bright regions reflection).

not statistically significantly different from the blank control due to precipitation of the complex, which has a limited solubility in water. At these higher concentrations, DMSO itself had a significant bactericidal effect, which was even higher than that for RuBTZ at the highest concentration tested. Therefore, the efficient bactericidal activity of RuBTZ in Figure 2A could be ascribed to some combination with the effect from DMSO. In summary, RuBTZ displayed significant bactericidal activity against E. coli, whereas RuCTZ had only minor effects, even at high concentrations. With regard to the possible cell toxicity of the ruthenium complexes, we performed cell viability assays using 3T3 mammalian cells. The data in Figure 2B show no significant toxic effect from either RuBTZ or RuCTZ, or DMSO, except at very high concentrations. We should note that the antibacterial activity of RuBTZ and RuCTZ against E. coli presented here does not agree with recently published data obtained with the disc diffusion assay method,18 in which inhibition halos of 9.0 and 15.0 cm were obtained for RuBTZ and RuCTZ, respectively. The discrepancy may be associated with the different methodologies to determine antibacterial activity. The disk diffusion assay is based on the amount of material that can diffuse from the filter paper into the agar containing the bacterial solution spread on the top. Therefore, the bactericidal activity of a particular compound depends on its solubility, aggregation, and ability to diffuse from the disk. According to Balouiri et al.,22 the agar disk diffusion method is not appropriate to determine the MIC because it is not possible to quantify the amount of material diffused from the disk into the agar. Moreover, inhibition of bacterial growth does not mean bacterial death, which implies that the method cannot distinguish bactericidal from bacteriostatic effects. In contrast, dilution methods such as the 96-well plate microdilution provide a more accurate value of MIC, with homogeneous bacterial distribution in the medium. 3.2. Interaction with Langmuir Monolayers Mimicking the E. coli Membrane. The mechanisms of action of antibiotics are normally related to either destruction of the membrane or pore formation that leads to leakage;23 so the higher bactericidal activity of RuBTZ can in principle be related to its interaction with the bacterial cell membrane. As already mentioned, in Gram-negative bacteria such as E. coli, there are two adjacent membranes, viz., an inner plasmatic lipid membrane and an external wall made mainly of LPSs. We therefore tested the hypothesis of membrane-related activity with Langmuir

3. RESULTS AND DISCUSSION 3.1. Bactericidal Activity of RuBTZ and RuCTZ. The data for the bactericidal effect against E. coli, obtained with the 96-well plate microdilution method, are shown in Figure 2A for RuBTZ, RuCTZ, and DMSO at distinct concentrations. Comparison with DMSO is important because the ruthenium complexes are not soluble in aqueous solutions, similarly to hydrophobic drugs such as amoxycillin, clarithromycin, and rifampicin.21 The bactericidal activity for RuBTZ was consistently higher, especially above 40 μg/mL, where the activity of RuCTZ was 14169

DOI: 10.1021/acs.langmuir.7b02247 Langmuir 2017, 33, 14167−14174

Article

Langmuir

Figure 3. (A) π−A isotherms for the membrane model with E. coli extract, LPS, DOPC, and the complexes RuBTZ and RuCTZ neat and the mixed films with DOPC (B) and LPS (C). The subphase was PBS pH = 7 for LPS measurements and Milli-Q water for DOPC, both at 20 °C. To guarantee consistency in the comparison, in subsidiary experiments, we spread DOPC and E. coli extract monolayers on a PBS subphase, and the isotherms were essentially the same as those for monolayers on ultrapure water (results not shown). The area values in A and C are given in cm2 (area occupied on the trough) because the molecular weight of E. coli extract and LPS is not known in detail. For the E. coli extract, for example, 17% of the weight corresponds to unknown compounds.

(see Figure 5), indicating a relatively high fluidity due to the presence of unsaturated lipids and protein residues in the extract. Clear formation of domains during compression was observed in the BAM images at the right hand side of Figure 4A for this Langmuir film. The complexes RuBTZ and RuCTZ were also able to form Langmuir films, which appear more condensed than for the E. coli lipid extract as shown in Figure 4. The collapse pressure was ca. 60 mN/m for both compounds, higher than that for other Ru phosphine complexes.25−27 On the basis of the maximum compressibility modulus, that is, Cs−1max of 91.0 ± 0.2 mN/m for RuBTZ and 77.4 ± 9.4 mN/m for RuCTZ, these films may be classified as liquid-expanded. For RuBTZ, there is no plateau in the π−A isotherm, consistent with the BAM image in Figure 4B, typical of a film without domains. In contrast, for RuCTZ, different phases coexisted with a plateau near 5 mN/m. This leads to a BAM image with domains at the corresponding surface pressure. Increasing surface pressure resulted in the coalescence of domains to form a film with a porous appearance, which may be correlated with non-monomolecular structures formed for this inorganic complex.27 One should remark that condensed domains were observed in spite of the liquidexpanded nature of the film according to the compressibility modulus, as the low value of the latter seems to arise from porous films with non-monomolecular structures. The incorporation of RuBTZ or RuCTZ into monolayers of E. coli extract led to an expansion in the surface pressure isotherms. The expanded area increased with the complex concentration, as indicated in Figure 5A,B and Table 1. Expansion should be expected to be larger for RuBTZ owing to its 2-mercaptobenzothiazoline ligand, which has an additional benzyl ring compared to the 2-mercaptothiazoline in RuCTZ. This was indeed observed for the lower concentration of 8% but not for 16%, as

monolayers from an E. coli extract and from LPS, representing the plasmatic membrane and the outer membrane, respectively. For comparison, the cell membrane of mammalian cells was mimicked with the zwitterionic phospholipid 1,2-dioleoyl-snglycero-3-phosphocholine (DOPC). Monolayers could be formed at the air/water interface for all the compounds (E. coli extract, LPS, and DOPC), in addition to the two complexes RuBTZ and RuCTZ, as indicated in the surface pressure (π−A) isotherms in Figure 3A. Note that the area in Figure 3A,C is given in cm2, rather than in area per molecule, because the total composition of E. coli lipid extract and LPS samples is not known. Figure 3B shows that RuBTZ and RuCTZ had very little effect on the isotherms of DOPC, where a small expansion was noted for RuCTZ, indistinguishable from the effect of DMSO, and a small condensation occurred for RuBTZ. The effect of RuCTZ was almost negligible on the LPS monolayer, again indistinguishable from DMSO, while RuBTZ caused some condensation of the monolayer, as shown in Figure 3C. As discussed below, the Ru complexes induced much larger changes on the monolayers made with the E. coli extract, and therefore we concentrate our attention to this system. The surface pressure (π−A) isotherm in Figure 4A shows that the E. coli extract formed a film with a limiting area of 136 cm2 (obtained by extrapolating the pressure−area curve in the condensed phase to zero pressure) and collapse pressure near 40 mN/m. Analogously to Figure 3A,C, the area in the axis is given in cm2. A phase transition from the liquid-expanded to the liquidcondensed state was observed at 25−28 mN/m, confirmed by BAM images. These results are similar to those of Tian et al.24 for E. coli total lipid extract on 0.1 M PBS, pH = 7.0. However, the expanded nature of the isotherm led to a maximum compressibility modulus (Cs−1max) of only 56.2 ± 1.2 mN/m 14170

DOI: 10.1021/acs.langmuir.7b02247 Langmuir 2017, 33, 14167−14174

Article

Langmuir

example, near 30 mN/m, which is believed to correspond to the pressure in a real cell membrane,28 Cs−1 decreased from 57.5 ± 7.5 mN/m for E. coli monolayer to 42.5 ± 0.5 and 30.9 ± 5.9 mN/m for the mixed E. coli + RuBTZ and E. coli + RuCTZ monolayers, respectively. It seems that incorporation of the complexes makes the E. coli films down more fluid, with a slightly larger effect for RuCTZ. In summary, one cannot infer unequivocally which of the complexes induced larger changes in the E. coli extract monolayer, in spite of the higher hydrophobicity of RuBTZ, whose effects were captured in the PM-IRRAS measurements below. The vibrational spectroscopy information from PM-IRRAS allows one to infer how the head group and tail of E. coli extract were affected by interaction with RuBTZ or RuCTZ. The PMIRRAS spectra for the Langmuir monolayers are shown in Figure 6, and the assignments for the main bands are listed in Table 2. The spectra were divided in two regions to distinguish between interactions with the headgroups and tails to highlight the main results, and the entire spectrum is shown in Figure S2 of the Supporting Information. The band intensity observed depends not only on the presence of a chemical group but also on its orientation. Under the conditions used in the experiments, for instance, upward bands mean that the corresponding group is oriented preferentially parallel to the air/water interface. Hence, the overall effect from RuBTZ and RuCTZ on the E. coli extract monolayer can only be analyzed qualitatively. For the E. coli lipid extract, bands are observed at 1559, 1642, and 1680 cm−1, assigned to residual proteins in the lipid extract. The bands at 1559 and 1642 cm−1 are assigned to amide II vibrations (C−N bending mode) and amide I vibrations (CO stretching mode), respectively, the band at 1680 cm−1 assigned to amide I vibrations can also be attributed to the secondary structure of proteins.29 In the nonpolar region, the E. coli lipid extract shows four intense bands, and the bands at 2857 and 2923 cm−1 are due to symmetric and antisymmetric stretching, respectively,30 of CH2 present in the lipids (saturated and insaturated) of the extract. For the mixed film of E. coli extract with RuBTZ, the spectrum shows larger effects in the tail region, indicating that the nonpolar region of the lipids has been affected. For instance, the band 2857 cm−1 for E. coli, assigned to CH2 symmetric stretching, was shifted considerably to the intense band at 2849 cm−1, while another band appeared at 2886 cm−1 owing to RuBTZ or owing to CH3 symmetric stretching. This shift to a lower wavenumber (2849 cm−1) can be related to more ordered CH2 groups. The band at 2945 cm−1 attributed to CH3 asymmetric stretching29 also shifted to 2957 cm−1. The effects from RuBTZ on the lipid polar region are less visible in Figure 6 (left), with the band attributed to the secondary structure of proteins at 1680 cm−1 shifting to 1686 cm−1. The presence of the 2-mercaptobenzothiazoline ligand, which is more hydrophobic than 2-mercaptothiazoline, governs the observed behavior. In RuCTZ, effects were seen in both polar and nonpolar regions of the spectra for the E. coli membrane. The CO stretching band is redshifted to 1735 cm−1, which may suggest an increase in the hydration of ester groups, consistent with the red-shift for PO asym in E. coli from 1236 to 1222 cm−1. This effect is not observed in the case of RuBTZ because only the band intensity changed. In fact, owing to its higher hydrophobicity RuBTZ interacts preferentially with the acyl chains. Also, the interaction with the headgroups could lead to a new Ru complexphospholipid structure because some ligands in the Ru complex can be labilized, thus allowing for a new coordination with the phospholipid head group, according to a previous study.31 The

Figure 4. π−A isotherms and BAM images for (A) E. coli and (B) RuBTZ and (C) RuCTZ Langmuir films spread on Milli-Q water at 20 °C. The greyscale had to be adjusted in images with excessive brightness, as in image 3 in B.

indicated in Table 1. The most striking change occurred for the monolayer with 16% RuBTZ, with plateau in the isotherm of E. coli extract being formed, typical of a new organization of the film. The monolayer expansion caused by the complexes was reflected on the BAM images (see Figure 5), and to achieve a similar coverage with condensed phase domains, higher surface pressures are required than for the neat E. coli monolayer. The DMSO combined effect could also be responsible for the isotherm profile but adding DMSO yielded only a small film contraction at typical membrane pressures (30−35 mN/m), as indicated in Figure S1 of the Supporting Information. Therefore, the expansion in the E. coli isotherms induced by RuCTZ and RuBTZ is not an artifact caused by the use of DMSO to obtain solutions with Ru complexes. The Ru complexes also affected the compressional modulus of the E. coli monolayer (Table 1). For 14171

DOI: 10.1021/acs.langmuir.7b02247 Langmuir 2017, 33, 14167−14174

Article

Langmuir

Figure 5. π−A isotherms, Cs−1 and BAM images for pure E. coli extract and their mixed monolayers with (A) RuBTZ and (B) RuCTZ at 2 M ratios spread on Milli-Q water at 20 °C. The greyscale had to be adjusted in images with excessive brightness, as in image 3 in A and B.

Table 1. Experimental Area (Aexp) and Maximum Compressibility Modulus (Cs−1) for the Neat Compounds and Their Mixtures with 8 and 16% of the Complexes in an E. coli Membrane Modela Aexp Cs−1max Cs−1at 30 mN/m a

E. coli

RuBTZ

8% RuBTZ

16% RuBTZ

RuCTZ

8% RuCTZ

16% RuCTZ

132.5 ± 3.5 56.2 ± 1.2 57.5 ± 7.5

95.0 ± 3.0 91.0 ± 0.2 93.5 ± 1.5

164 ± 1.0 50.3 ± 1.3 50.7 ± 5.7

182 ± 0.5 60.7 ± 1.7 42.5 ± 0.5

68.5 ± 0.5 77.4 ± 9.4 58.2 ± 7.2

151 ± 0.5 42.6 ± 1.6 39.3 ± 0.3

215 ± 3.0 51 ± 5 30.9 ± 5.9

The area was calculated in cm2 for the pressure at 30 mN/m by extrapolating the pressure−area isotherm to zero pressure.

other shifts for the polar groups in the E. coli membrane are shown in Table 2.

First of all, with membrane models as used here it is only feasible to interrogate mechanisms of antibacterial activity based on membrane disruption, with no consideration of the remaining mechanisms that depend on proteins and other cell constituents. In addition, the membrane models represented by Langmuir films are very simplified, and therefore, the effects observed for a

4. BIOLOGICAL IMPLICATIONS Prior to any analysis of possible biological implications, one has to consider the limitations of a study with cell membrane models. 14172

DOI: 10.1021/acs.langmuir.7b02247 Langmuir 2017, 33, 14167−14174

Article

Langmuir

Figure 6. PM-IRRAS spectra for monolayers at 30 mN/m of the pure compounds and for E. coli extract mixed with 8% of RuBTZ or RuCTZ. The baseline is the line of zero level of the PM-IRRAS signal.

Table 2. IR Modes (cm−1) for IRRAS Analysis of Phospholipids of E. coli Extract Neat and Mixed with 8% of Ru Complexes E. coli

E. coli + RuBTZ

Protein Vibrations 1559 and 1551 and 1642 absent amide I (related to the secondary 1680 1686 structure) Nonpolar Region Vibrations CH2 sym. stretching 2857 2849 CH2 asym. stretching 2923 2923 CH3 asym. stretching 2945 2957 Polar Region Vibrations PO2− asym. stretching 1236 1236 PO2− sym. stretching 1096 1096 CO stretch of esters 1742 1742 amide II and amide I

outer layer normally required for the bactericidal activity could not be captured in the monolayer studies. (iii) Both RuBTZ and RuBTZ affect the E. coli extract monolayer, but it was not possible to infer which of the complexes had a stronger effect. Hence, an unequivocal correlation between monolayer effects and the higher bactericidal activity of RuBTZ cannot be established.

E. coli + RuCTZ 1551 and absent masked

5. CONCLUSION In a comparative study with the complexes RuBTZ and RuCTZ and three types of monolayer, we tested the hypothesis that monolayer effects could be correlated to the bactericidal activity of these ruthenium complexes. Neither of the complexes induced significant changes in a DOPC monolayer, which was chosen to mimic a mammalian cell membrane, and this would correlate well with the lack of toxicity of RuBTZ and RuCTZ toward 3T3 cells. In contrast, the higher activity against E. coli observed for RuBTZ could not be correlated with its effects on the monolayers representing the two adjacent membranes of Gram-negative bacteria. No significant changes were observed on the LPS monolayer upon incorporation of RuBTZ and RuBTZ. As for the E. coli extract monolayer, there were indeed considerable changes induced by RuBTZ, but they were not unequivocally larger than for RuCTZ. In the compressional modulus, for example, RuCTZ induced a larger change at surface pressures typical of a cell membrane. It is therefore concluded that the monolayer studies were not sufficient to interrogate the molecular mechanisms responsible for the bactericidal activity of RuBTZ and RuCTZ. It is likely that such an activity may require pore formation in the bacterial membrane, and this type of information is not accessible in monolayer experiments.

2862 2923 2945 1222 1100 1735 shoulder

particular drug may not take place on real cells. That is to say, conclusions drawn from a study with this type of membrane model allow one to establish correlations between molecularlevel mechanisms and actual antibacterial activity but do not amount to irrefutable proof. Having said that, using Langmuir films may still provide useful information due to the unique capability of probing molecular-level interactions, especially with a combination of experimental techniques. The implications of the experimental observations with Langmuir films in this study may be summarized as follows: (i) RuBTZ and RuCTZ caused almost negligible changes in DOPC monolayers, which could mimic the membrane of mammalian cells. This correlates well with the lack of toxicity of either Ru complexes in 3T3 mammalian cells. (ii) The effect of RuBTZ on the LPS monolayer was only minor, whereas RuCTZ seemed to have no effect. Therefore, the pore formation or disruption of the LPS

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.langmuir.7b02247. 14173

DOI: 10.1021/acs.langmuir.7b02247 Langmuir 2017, 33, 14167−14174

Article

Langmuir

■

π−A isotherms for E. coli extract soluble in chloroform and in chloroform/DMSO 16% spread on Mili-Q water at 20° C and PM-IRRAS spectra of polar region of the monolayers(PDF)

(14) Shai, Y. Mechanism of the Binding, Insertion and Destabilization of Phospholipid Bilayer Membranes by α-Helical Antimicrobial and Cell Non-Selective Membrane-Lytic Peptides. Biochim. Biophys. Acta, Biomembr. 1999, 1462, 55−70. (15) Stokes, J. M.; Macnair, C. R.; Ilyas, B.; French, S.; Côté, J.-P.; Bouwman, C.; Farha, M. A.; Sieron, A. O.; Whitfield, C.; Coombes, B. K.; et al. Pentamidine Sensitizes Gram-Negative Pathogens to Antibiotics and Overcomes Acquired Colistin Resistance. Nat. Microbiol. 2017, 2, 17028. (16) Stefaniu, C.; Brezesinski, G.; Möhwald, H. Langmuir Monolayers as Models to Study Processes at Membrane Surfaces. Adv. Colloid Interface Sci. 2014, 208, 197−213. (17) Sessa, G.; Weissmann, G. Phospholipid Spherules (Liposomes) as a Model for Biological Membranes. J. Lipid Res. 1968, 9, 310−318. (18) Appelt, P.; Fagundes, F. D.; Facchin, G.; Gabriela Kramer, M.; Back, D. F.; Cunha, M. A. A.; Sandrino, B.; Wohnrath, K.; de Araujo, M. P. Ruthenium (II) Complexes Containing 2-Mercaptothiazolinates as Ligands and Evaluation of Their Antimicrobial Activity. Inorg. Chim. Acta 2015, 436, 152−158. (19) Geraldo, V. P. N.; Pavinatto, F. J.; Nobre, T. M.; Caseli, L.; Oliveira, O. N. Langmuir Films Containing Ibuprofen and Phospholipids. Chem. Phys. Lett. 2013, 559, 99−106. (20) Hénon, S.; Meunier, J. Microscope at the Brewster Angle: Direct Observation of First-Order Phase Transitions in Monolayers. Rev. Sci. Instrum. 1991, 62, 936. (21) Andrews, J. M. Determination of Minimum Inhibitory Concentrations. J. Antimicrob. Chemother. 2001, 48, 5−16. (22) Balouiri, M.; Sadiki, M.; Ibnsouda, S. K. Methods for in Vitro Evaluating Antimicrobial Activity: A Review. J. Pharm. Anal. 2016, 6, 71−79. (23) Neu, H. C.; Gootz, T. D. Antimicrobial Chemotherapy. Medical Microbiology; University of Texas Medical Branch: Galveston, TX, 1996. (24) Tian, C.; Tétreault, E.; Huang, C. K.; Dahms, T. E. S. Electrostatic Interactions of Colicin E1 with the Surface of Escherichia Coli Total Lipid. Biochim. Biophys. Acta, Biomembr. 2006, 1758, 693−701. (25) Schurch, S.; Schurch, D.; Curstedt, T.; Robertson, B. Surface Activity of Lipid Extract Surfactant in Relation to Film Area Compression and Collapse. J. Appl. Physiol. 1994, 77, 974−986. (26) Schürch, S.; Green, F. H. Y.; Bachofen, H. Formation and Structure of Surface Films: Captive Bubble Surfactometry. Biochim. Biophys. Acta, Mol. Basis Dis. 1998, 1408, 180−202. (27) Wohnrath, K.; dos Santos, P. M.; Sandrino, B.; Garcia, J. R.; Batista, A. A.; Oliveira, O. N. A Novel Binuclear Ruthenium Complex: Spectroscopic and Electrochemical Characterization, and formation of Langmuir and Langmuir-Blodgett films. J. Braz. Chem. Soc. 2006, 17, 1634−1641. (28) Marsh, D. Lateral Pressure in Membranes. Biochim. Biophys. Acta, Rev. Biomembr. 1996, 1286, 183−223. (29) Yu, C.; Irudayaraj, J. Spectroscopic Characterization of Microorganisms by Fourier Transform Infrared Microspectroscopy. Biopolymers 2005, 77, 368−377. (30) Mendelsohn, R.; Mao, G.; Flach, C. R. Infrared Reflection− Absorption Spectroscopy: Principles and Applications to Lipid−Protein Interaction in Langmuir Films. Biochim. Biophys. Acta, Biomembr. 2010, 1798, 788−800. (31) Sandrino, B.; Wrobel, E. C.; Nobre, T. M.; Caseli, L.; Lazaro, S. R.; Júnior, A. C.; Garcia, J. R.; Oliveira, O. N.; Wohnrath, K. Interaction between Active Ruthenium Complex [RuCl3(dppb)(VPy)] and Phospholipid Langmuir Monolayers: Effects on Membrane Electrical Properties. Chem. Phys. Lett. 2016, 649, 29−36.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone: + 55 16 33739825. ORCID

B. Sandrino: 0000-0001-7511-0077 M. P. de Araujo: 0000-0002-5845-2451 V. M. Rotello: 0000-0002-5184-5439 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by FAPESP, CNPq, CAPES (Brazil). B.S. and J.F.A.d.O. thank FAPESP for the fellowships (2104/ 12567-8 and 2015/21918-1, respectively).

■

REFERENCES

(1) Masadeh, M. M.; Mhaidat, N. M.; Alzoubi, K. H.; Hussein, E. I.; AlTrad, E. I. In Vitro Determination of the Antibiotic Susceptibility of Biofilm-Forming Pseudomonas Aeruginosa and Staphylococcus Aureus: Possible Role of Proteolytic Activity and Membrane Lipopolysaccharide. Infect. Drug Resist. 2013, 6, 27−32. (2) Ansari, M. A.; Khan, H. M.; Khan, A. A.; Malik, A.; Sultan, A.; Shahid, M.; Shujatullah, F.; Azam, A. Evaluation of Antibacterial Activity of Silver Nanoparticles against MSSA and MRSA on Isolates from Skin Infections. Biol. Med. 2011, 3, 141−146. (3) Fayaz, A. M.; Balaji, K.; Girilal, M.; Yadav, R.; Kalaichelvan, P. T.; Venketesan, R. Biogenic Synthesis of Silver Nanoparticles and Their Synergistic Effect with Antibiotics: A Study against Gram-Positive and Gram-Negative Bacteria. Nanomed. Nanotechnol. Biol. Med. 2010, 6, 103−109. (4) Li, F.; Collins, J. G.; Keene, F. R. Ruthenium Complexes as Antimicrobial Agents. Chem. Soc. Rev. 2015, 44, 2529−2542. (5) Mjos, K. D.; Orvig, C. Metallodrugs in Medicinal Inorganic Chemistry. Chem. Rev. 2014, 114, 4540−4563. (6) Allardyce, C. S.; Dyson, P. J. Ruthenium in Medicine: Current Clinical Uses and Future Prospects. Platinum Met. Rev. 2001, 45, 62−69. (7) Dal Lago, V.; França de Oliveira, L.; de Almeida Gonçalves, K.; Kobarg, J.; Borba Cardoso, M. Size-Selective Silver Nanoparticles: Future of Biomedical Devices with Enhanced Bactericidal Properties. J. Mater. Chem. 2011, 21, 12267. (8) Gorup, L. F.; Longo, E.; Leite, E. R.; Camargo, E. R. Moderating Effect of Ammonia on Particle Growth and Stability of QuasiMonodisperse Silver Nanoparticles Synthesized by the Turkevich Method. J. Colloid Interface Sci. 2011, 360, 355−358. (9) Levina, A.; Mitra, A.; Lay, P. A. Recent Developments in Ruthenium Anticancer Drugs. Metallomics 2009, 1, 458−470. (10) Balakrishnan, S. K.; Dass, S. K.; Rajendran, H. A. D.; Savrimuthu, I. Cytotoxicity and Antimicrobial Activity of Mono-, Di- and Trinuclear ruthenium(II) Polypyridine Complexes. Int. J. Pharm. Pharm. Sci. 2015, 7, 317−321. (11) Tomás-Gamasa, M.; Martínez-Calvo, M.; Couceiro, J. R.; Mascareñas, J. L. Transition Metal Catalysis in the Mitochondria of Living Cells. Nat. Commun. 2016, 7, 12538. (12) Gorgoll, R. M.; Harano, K.; Nakamura, E. Nanoscale Control of Polymer Assembly on a Synthetic Catalyst−Bilayer System. J. Am. Chem. Soc. 2016, 138, 9675−9681. (13) Walsh, C. T. Molecular Mechanisms That Confer Antibacterial Drug Resistance. Nature 2000, 406, 775−781. 14174

DOI: 10.1021/acs.langmuir.7b02247 Langmuir 2017, 33, 14167−14174