BIOLOGIJA. 2020. Vol. 66. No. 2. P. 80–93 © Lietuvos mokslų akademija, 2020

Evaluation of antimicrobial activity of synthesized 9H-alkylcarbazole and 10H-alkylphenothiazine derivatives on the cells of Salmonella enterica ser. Typhimurium, Saccharomyces cerevisiae, and Candida albicans

Simona Sutkuvienė 1, 2 *,

Sandra Sakalauskaitė1,

Neringa Kuliešienė1,

Lina Ragelienė1,

Rimantas Daugelavičius1

1Department of Biochemistry, Faculty of Natural Sciences, Vytautas Magnus University, Vileikos str. 8, 44404 Kaunas, Lithuania

2Department of Biochemistry, Faculty of Medicine, Lithuanian University of Health Sciences, Tilžės str. 18, 47181 Kaunas, Lithuania

10H-substituted phenothiazine and 9H-substituted carbazole derivatives are important because of a very wide range of applications and especially in medical chemistry due to their pharmacological activities. In this study, we synthesized 9H-alkylcarbazole and 10H-alkylphenothiazine derivatives with various lengths of alkyl chains and evaluated their antimicrobial and efflux inhibiting activities on the cells of Salmonella enterica ser. Typhimurium, Saccharomyces cerevisiae, and Candida albicans. Results of our study revealed that an increased length of alkyl chains of the carbazoles increased the accumulation of efflux indicator tetraphenylphosphonium (TPP+) ions. Cells of S. enterica efflux mutant ΔTolC had a considerable susceptibility to the synthesized compounds. The  compounds exerted synergy with fluconazole against S. cerevisiae yeast. Efflux pump mutant ΔPdr5 was hypersensitive to the investigated carbazole and phenothiazine derivatives. The inhibitory effect of the compounds with a shorter alkyl chain (10-methyl-10H-phenothiazine and 9-methyl-9H-carbazole) was the highest for Candida albicans cells.

Keywords: phenothiazine, carbazole, tetraphenylphosphonium ions, minimal inhibitory concentration, efflux pumps, Salmonella enterica, Saccharomyces cerevisiae, Candida albicans

INTRODUCTION

Antimicrobial resistance is a  worldwide problem in human and veterinary medicine. An extensive use of antimicrobials leads to the spread of resistant bacteria in animals and humans (Michael et al., 2014). Antimicrobial drugs from both hospital and agricultural sources can persist in soil or aquatic environments, and these compounds may affect the treatment of human diseases (Allen et al., 2010). The appearance of multiple resistant bacteria of human and animal origin is accompanied by co-contamination of the environment apparently leading to a great health concern (Kossow et al., 2017).

Microorganisms have the remarkable ability to pump antimicrobials out of cells. This protective feature is the most widespread form of resistance to many classes of antimicrobials. Over the  years, several solutions to solve this problem have been proposed (Brüssow, 2017). One of such solutions is efflux pump inhibitors, which could be therapeutic agents for restoration of sensitivity to antibiotics. Efflux pump inhibitors act against multidrug resistant efflux pumps with different substrates and they are expected to not only reverse resistance to a single drug but also to clinically beneficial antimicrobials. Some antimicrobial derivatives could act as enhancers of antibiotic efficiency. They may not have any antimicrobial properties alone, but when used with bacteria-resistant drug they could enhance the  effect of the  drug (Opperman et al., 2015). One of the best-studied peptidomimetic compounds – phenylalanine-arginine β-naphthylamide (PaβN) – was originally described in 1999 and characterized further in 2001 as a broad-spectrum efflux pump inhibitor (Lomovskaya et al., 2001), but as it possesses detrimental side effects, it was not introduced into clinical practice (Bolla, Brune, 2017).

Another relevant problem is resistance to antifungal compounds. New directions of treatment of fungal infections are needed, especially for infections caused by the yeast of the Candida family. ABC (ATP-binding cassette) transporters of pleiotropic drug resistance subfamily, like Cdr1p of C. albicans, are the most frequent cause of resistance to antifungal agents.

Contemporary attempts to find effective enhancers were relatively unsuccessful against Gram-negative bacteria, so it is important to continue the  search for new effective compounds.

Phenothiazines are the oldest synthetic antipsychotic drugs, which do not have analogues in the world of natural compounds. Phenothiazine was first synthesized by August Bernthesen in 1883 and for many years it was used in veterinary medicine as an anthelmintic drug. Having an amino alkyl side chain connected to nitrogen atom, these compounds are important in medicinal chemistry (Clausen et al., 2017; Carsuo et al., 2008; Conchon et al., 2008; Bouaziz  et  al., 2015; Indumati  et  al., 2012; Zhang et al., 2010; Kantevari et al., 2011; Asche, Demeunynck, 2007; Hieda et al., 2014; Humphries  et  al., 2016; Kaur  et  al., 2012; Rennison et al., 2016; Pluta, Morak-Młodawska, Jeleń, 2011; Sarmiento  et  al., 2011; Warman  et  al., 2013; Spengler et al., 2016; Liu et al., 2009; Mosnaim et al., 2006; Hou et al., 2019).

Investigation of substituted 10H-phenothiazines has had a strong growth during the recent years because of wide range of applications (Table). Carbazoles are also a large and interesting group of organic compounds: its moiety is frequent in structures of numerous drugs such as listed in the Table.

Table. Frequency of 9H-carbazole and 10H-phenothiazine moiety in the drugs

Drug type Carbazole Phenothiazine
Antibacterial (Clausen et al., 2017; Carsuo et al., 2008; Conchon et al., 2008; Bouaziz et al., 2015; Indumati et al., 2012) (Pluta et al., 2011)
Antifungal (Zhang et al., 2010) (Sarmiento et al., 2011)
Antitubercular (Kantevari et al., 2011) (Warman et al., 2013)
Anticancer (Asche, Demeunynck, 2007) (Spengler et al., 2016)
Antioxidant (Hieda et al., 2014) (Liu et al., 2009)
Antidiabetic (Humphries et al., 2016) (Mosnaim et al., 2006)
Antipsychotic, sedatives (Kaur et al., 2012) (Hou et al., 2019)
Anthelmintic (Rennison et al., 2016) (Smith, 1942)
Antiviral (Gulza et al., 2014) (Mucsi et al., 2001)
Anti-inflammatory (Nalli et al., 2016) (Sharma et al., 2005)

A  slight change in the  structure of these compounds causes distinguishable differences in their biological activities (Jaszczyszyn et al., 2012). Applied as neuroleptic drugs, phenothiazine derivatives easily cross the  blood-brain barrier, since they exhibit a  strong affinity to lipid bilayers of the cell membranes in neurons and other lipid-rich tissues as the phenothiazine ring possesses a high degree of lipophilicity (Seelig et al., 1994). Depending on the structure of substituents in the side chain, the intensity of the neuroleptic action of phenothiazine derivatives could be ranked as follows: piperazine group  >  piperidine group  >  aliphatic chain (Jaszczyszyn  et  al., 2012). Among the  candidates for effective anti-MDR drugs, phenothiazines are worth further studying, because they are strong inhibitors of the Pgp transport function and exhibit several cancer chemopreventive actions (Jaszczyszyn et al., 2012).

Derivatives of carbazole and phenothiazine are considered as potential multidrug resistance (MDR) efflux pump inhibitors (Rodrigues et al., 2011). It is very important to discover molecules which could inhibit the efflux and to understand the mechanism of inhibition (Cuestas et al., 2012). Here we explore a possibility to use phenothiazine and carbazole derivatives as susceptibility enhancers to antimicrobials, which can act as competitive substrates in efflux of the drugs.

MATERIALS AND METHODS

Bacteria, yeast, and chemicals

Salmonella enterica  ser. typhimurium strain SL1344 wild type (WT) and ∆TolC mutant with ∆tolC channel deletion were obtained from Prof. Seamus Fanning (Institute of Food and Health, University College Dublin, Ireland). Saccharomyces cerevisiae strains W3031a (MATa) and ΔPdr5 (MATa, pdr5:HIS3) were obtained from Prof. Chuang-Rung Chang (National Tsing Hua University, Taiwan). Candida albicans ATCC 10231 were obtained from Dr Eglė Lastauskienė (Institute of Biosciences, Vilnius University).

10H-Phenothiazine 98%, 9H-carbazole 95%, 1-bromoethane 98%, iodomethane 99%, 1-bromohexane 98%, tetrabutylammonium hydrogensulfate 97%, potassium hydroxide 96%, potassium carbonate 96%, chloramphenicol 98% were purchased from Aldrich (St. Louis, MO) and used as received without additional purification. Polymyxin B (PMB) sulphate, Gramicidin D (GD), Luria-Bertani broth (LB) came from Sigma (St.  Louis, MO), tetraphenylphosphonium (TPP+) chloride from Fluka (St. Gallen, Switzerland), and ethylene diamine tetraacetic acid (EDTA), HCl and glucose  –  from Sharlau (Barcelona, Spain). Tris(hydroxymethyl) aminomethane (Tris) was obtained from Roth (Karlsruhe, Germany), ethidium bromide and fluconazole from Acros Organics (New Jersey, USA). Yeast extract and bacteriological peptone were from Oxoid (Hampshire, England), RPMI medium 1640 from Merck (WGK, Germany). PAβN (phenylalanine-arginine β-naphthylamide) was synthesized as described in Sutkuviene et al. (2013).

Instrumentation for the synthesized compounds

1H NMR spectra of the synthesized compounds were recorded using a  Bruker Ascend 400 (400 MHz) apparatus using chloroform-d. Mass spectra were obtained on Waters ZQ spectrometer. Synthesis reactions of 9H-carbazole and 10H-phenothiazine compounds were monitored by thin-layer chromatography (TLC) on pre-coated plastic sheets with 0.25 mm Merck silica gel 60F-254. Column chromatography of synthesized carbazole and phenothiazine compounds was carried out using Merck silica gel 60 (230–240  Mesh). Melting points were determined on a Stuart SMP11 apparatus (Bibby Scientific, Staffordshire, UK). TPP+ selective electrodes (Daugelavicius et al., 1997) were connected to the potential-amplifying system based on an ultralow input bias current operational amplifier AD549JH (Analog Devices, Norwood, Ma, USA). The  amplifying system was connected to a  computer through PowerLab  8/35 logger (ADInstruments, Oxford, UK). Agar salt bridges were used for indirect connection of reference electrodes (Orion model 9001; Thermo Fisher Scientific, USA) with cell suspensions in the vessels. The representative sets of curves from three independent series of measurements are presented in the Fig. 4.

Synthesis of N-alkylphenothiazine and N-alkylcarbazole derivatives

General procedure of synthesis and analysis of phenothiazine derivatives

Compounds 1–3 were synthesized by the  N-alkylation method (Simokaitiene et al., 2006). 10H-Phenothiazine (1.00 g, 5 mmol) was dissolved in 10 mL of dry toluene. Then potassium carbonate (0.346 g, 2.5 mmol), potassium hydroxide (0.842 g, 15 mmol), tetrabutylammonium hydrogensulfate (0.02 g, 0.06 mmol) and iodomethane (1.065 g, 7.5 mmol) or 1-bromoethane (0.817  g, 7.5  mmol) or 1-bromohexane (1.238  g, 7.5  mmol) were added, respectively (Fig. 1). The reaction mixture was refluxed for 12 hours. After TLC monitoring, the reaction mixture was cooled down at room temperature and filtered. The  solvent was removed under reduced pressure on a rotary evaporator. Compounds were recrystallized from ethanol and additionally purified by column chromatography (eluent: ethylacetate-hexane, 1:3). Yields, melting points, and spectral data of these compounds are given bellow.

10-Methyl-10H-phenothiazine (1)

White solid. M.p: 99°C; Yield: 65%; 1H  NMR (CDCl3, 400 MHz, δ, ppm): 7.21–7.12 (m, 4H, Ar), 6.96–6.90 (m, 2H, Ar), 6.82 (d, 2H, Ar, J = 8 Hz), 3.38 (s, 3H, Ar-CH3); MS (ESI, m/z): 214.08 [M+H]+, 199.02.

10-Ethyl-10H-phenothiazine (2)

White solid. M.p: 102°C; Yield: 79%; 1H NMR (CDCl3, 400 MHz, δ, ppm): 7.18−7.08 (m, 4H, Ar), 6.96–6.82 (m, 4H, Ar), 1.42 (t, 3H, Ar-CH2-CH3, J = 6.9 Hz), 3.93 (b.s, 2H, Ar-CH2-CH3); MS (ESI, m/z): 228.10 [M+H]+, 199.02.

10-Hexyl-10H-phenothiazine (3)

Yellow solid. M.p: 72°C; Yield: 75%; 1H NMR (CDCl3, 400  MHz, δ, ppm): 7.18−7.10 (m, 4H, Ar), 6.96–6.82 (m, 4H, Ar), 3.84 (s, 2H, Ar-CH2-(CH2)4-CH3), 1.80 (p, 2H, Ar-CH2-CH2-(CH2)3-CH3, J = 7.4 Hz), 1.43 (p, 2H, Ar-(CH2)2-CH2-(CH2)2-CH3, J = 7.3 Hz), 1.35−1.18 (m, 4H, Ar-(CH2)3-CH2-CH2-CH3), 0.87 (t, 3H, Ar-(CH2)5-CH3, J  =  6.9  Hz); MS (ESI, m/z): 284.15 [M+H]+, 199.02.

General procedure of synthesis and analysis of N-alkylcarbazole derivatives

Compounds 4–6 were synthesized by the N-alkylation method (Simokaitiene et al., 2006). 9H-Carbazole (1.00 g, 6 mmol) was dissolved in 15 mL of dry toluene. Then potassium carbonate (0.4299 g, 3 mmol), potassium hydroxide (1.0037 g, 18 mmol)), tetrabutylammoni-um hydrogensulfate (0.02 g, 0.06 mmol) and iodomethane (1.2819  g, 9  mmol) or 1-bromoethane (0.9870  g, 9.1  mmol) or 1-bromhexane (1.4493  g, 8.8  mmol) were added into the reaction mixture, respectively. The reaction mixture was refluxed for 12–14 hours. After TLC monitoring, the  reaction mixture was cooled down at room temperature and filtered. The solvent was removed under reduced pressure on a rotary evaporator. Compounds were recrystallized from ethanol and additionally purified by column chromatography (eluent: ethylacetate-hexane, 1:3). Yields, melting points, and spectral data of these compounds are given bellow.

9-Methyl-9H-carbazole (4)

White solid. M.p: 89°C; Yield: 59%; 1H  NMR (CDCl3, 400  MHz, δ, ppm): 8.10 (d, 2H, Ar, J = 7.8 Hz), 7.51–7.46 (m, 2H, Ar, J = 8.1 Hz), 7.41 (d, 2H, Ar, J  =  8.1  Hz), 7.25–7.21 (m, 2H, Ar), 3.86 (s, 3H, Ar-CH3); MS (ESI, m/z): 182.05 + [M+H]+, 167.05.

9-Ethyl-9H-carbazole (5)

White solid. M.p: 68°C. Yield: 46%; 1H NMR (CDCl3, 400  MHz, δ, ppm): 8.12 (d, 2H, Ar, J  =  7.7  Hz), 7.49–7.45 (m, 2H, Ar), 7.42 (d, Ar, J  =  8.1  Hz), 7.25–7.21 (m, 2H, Ar), 4.40 (q, 2H, Ar-CH2-CH3, J = 7.2 Hz), 1.45 (t, 3H, Ar-CH2-CH3, J  =  7.2  Hz); MS (ESI, m/z): 196.04 + [M+H]+, 167.98.

9-Hexyl-9H-carbazole (6)

Yellow solid. M.p: 61°C. Yield: 47%; 1H  NMR (CDCl3, 400  MHz, δ, ppm): 8.11 (d, 2H, Ar, J = 7.8 Hz), 7.48–7.44 (m, 2H, Ar), 7.41 (d, 2H, Ar, J = 8.1 Hz), 7.25–7.20 (m, 2H, Ar), 4.31 (t, 2H, -CH2-, J = 7.3Hz), 1.90 (p, 2H, -CH2-, J = 7.6Hz), 1.43–1.26 (m, 6H, -(CH2)3-), 0.88 (t, 3H, -CH3 J = 7.0 Hz); MS (ESI, m/z): 252.11 + [M+H]+, 167.98.

Determination of antimicrobial activity to the synthesized compounds

To evaluate the antimicrobial activity of the synthesized compounds, we applied the broth dilution method. First, S. enterica SL1344 or ΔTolC cells were cultivated in fresh LB medium with aeration at 37°C for 18  hours to OD600 of 1. S. cerevisiae cells were cultivated in 10 mL fresh yeast extract peptone dextrose (YPD) medium (1% yeast extract, 2% peptone, 2% glucose) at 30°C for 18  h. The  procedure involved serial two-fold dilutions of the  antimicrobial compounds in a liquid growth medium in 96-well microtitration plates. Dilutions of our synthesized compounds were performed starting from concentration of 150  μM of ethanolic stock solution. Each well was inoculated with a microbial inoculum. The bacterial cells were inoculated to obtain concentration 5 × 105 and 1–5 103 cfu/mL for the yeast. Microplates were incubated without agitation at 37°C for bacteria and 30°C for yeast. The turbidity of the cell suspensions was measured using TECAN GENios Pro™ (Männedorf, Switzerland) plate reader after 16–20 hours of incubation for bacteria and after 48 hours for yeast. The plate was shaken 5 s before each registration point. Representative sets of at least three independent measurements are presented in the following chapter.

Electrochemical measurements

To monitor the interaction of the synthesized compounds with bacteria, overnight culture of S. enterica SL1344 cells was diluted 1:50 in fresh LB medium and grown with aeration at 37°C to OD600 of 1. The cells were collected by centrifugation at 4°C for 10  min at 3000  ×  g (Heraeus Megafuge 16R, Thermo Fisher Scientific, Walthman, Ma, USA). Pelleted cells were re-suspended in 100 mM Tris/HCl (pH 8.0) to obtain 1/150 of the original cell culture volume, kept on ice and used within 4 h. The measurements of TPP+ concentration were performed simultaneously in two vessels. 2.5 × 109 cfu/ml were added to thermostated (37°C) and magnetically stirred reaction vessels containing 5 mL of Tris/HCl, pH 8.0 supplemented with 3 μM TPP+ (Daugelavicius et al., 1997). The solutions of 1–6 compounds in ethanol were used. Representative sets of at least three independent measurements are presented below.

RESULTS AND DISCUSSION

Synthesis of N-alkylphenothiazine and N-alkylcarbazole compounds

N-alkylphenothiazines (1–3) and N-alkylcarbazoles (4–6) were synthesized by the  reaction of 10H-phenothiazine/9H-carbazole with alkyl (ethyl, methyl, hexyl) iodide or bromide in the  presence of tetrabutylamonium hydrogensulfate, potassium carbonate, potassium hydroxide and dry toluene, stirred at 110°C for 12–14 hours (Fig. 1). The 1H NMR of compounds 1 and 4 showed the singlets at δ 3.38 and 3.86 ppm, respectively, showing the presence of the  CH3 group. This clearly indicates the  formation of 10-methyl-10H-phenothiazine (1) and 9-methyl-9H-carbazole (4). Compound 2 showed a triplet at δ 1.42 ppm and a broad singlet at δ 3.93 ppm, compound 5 showed quadruplet at δ 4.40  ppm and triplet at 1.45  ppm showing the presence of CH2 and CH3 groups. This indicates the formation of 10-ethyl-10H-phenothiazine (2) and 9-ethyl-9H-carbazole (5). Spectra of compound 3 showed peaks at δ 3.84, 1.80, 1.43, 1.35−1.18, 0.87 ppm showing the presence of all CH2 and CH3 groups present in 10-hexyl-10H-phenothiazine (3). Spectra of compound 6 showed peaks at δ 4.31, 1.90, 1.43−1.26, 0.88 ppm showing the  presence of all CH2 and CH3 groups present in 9-hexyl-9H-carbazole (6). All other aromatic protons were observed at expected regions. Furthermore, mass spectra data are in accordance with the expected structure of the obtained compounds.

img

Fig. 1. Synthesis scheme and structures of compounds 1–6

Evaluation of antimicrobial activity of the synthesized materials

Determination of susceptibility of S. enterica cells with N-alkyl phenothiazines (1–3) and N-alkyl carbazoles (4–6)

Susceptibility of S.  enterica wild type SL1344 and ΔTolC mutant cells to the synthesized compounds was determined. S. enterica ΔTolC cells lack the channel-forming porin TolC, an important component of RND family efflux pumps. The experiments were conducted in the 96-well plates, the initial concentration of compounds 1–6 was 150 μM (Fig. 2). The screening results showed that the  presence of compounds 1–6 in the  LB medium did not affect the  growth of wild type cells, but ΔTolC mutant cells were more sensitive to these compounds. Higher sensitivity of ΔTolC cells was chosen because it possesses higher sensitivity to the experimental compounds. This difference occurred because of the nonfunctioning RND type efflux pumps in ∆TolC mutant cells.

RND family pump inhibitor phenylalanylarginyl-B-naphtylamide (PAβN) was used for deeper exploration of efflux. Screening results showed that in the presence of PaβN, the synthesized compounds at concentration of 37.5 μM decreased the optical density of bacteria suspension indicating the negative effect on the viability of both S. enterica cells types. 9-Hexyl-9H-carbazole (6) with the  longest N-alkyl chain was the  most active against ΔTolC mutant cells. Therefore, we can conclude that compounds 1–6 are RND efflux pump substrates. The data presented above are in correlation with the data presented by other scientists. It is known that PAβN could bind to AcrB binding sites. The results suggest how an inhibitor, even when it is pumped out itself, can reduce the pumping of other substrates.

PAβN inhibited the  efflux of other drugs by binding to the bottom of the distal binding pocket, the  so-called hydrophobic trap, and also by interfering with the  binding of other drug substrates to the upper part of the binding pocket. However, its mechanism of inhibition is not clear (Kinana et al., 2016).

In the other part of our research, chloramphenicol (Cm), which is known as substrate of MDR efflux pumps (Sun et al., 2014), was used. The combined system of the synthesized carbazole and phenothiazine compounds with Cm was used to determine if activity of the MDR efflux pump substrate could be enhanced.

The concentration of compounds 1–6 was 20 µM as a result of the previous experiments (Fig. 2 A, B). The solutions of this concentration did not influence the growth of the cells, but the growth was decreased by the combination of PAβN and compounds 1–6.

img

Fig. 2. Impact of compounds 1–6 and efflux inhibitor PAβN on the growth of S. enterica SL1344 (A and C) and ΔTolC (B and D) cells. The experiments were performed in LB-Medium. The cells were incubated for 20 hours at 37°C. In B and D the medium contained 32 µM PAβN. The initial cell concentration was 5 × 105 cfu/mL

To determine if synthesized compounds could influence the growth of the cells in composition with antibiotic  –  Cm, we performed the following experiments.

The results of these experiments showed that the  synthesized compounds increase the efficiency of Cm on the cells of both strains (Fig. 3). The growth inhibition results showed that mutant cells were more sensitive compared to WT cells.

Our results indicate that both phenothiazines (compounds 1–3) and carbazoles (compounds 4–6) enhanced the sensitivity of S. enterica cells to Cm. The effect of the synthesized compounds was more expressed on ∆TolC mutant, i.e., the  cells with nonfunctioning RND type of efflux pumps. The  sensitivity effect did not depend much on the compounds, but we could still conclude that the  sensitivity of the cells for phenothiazine having longer alkyl chain are higher comparing with the phenothiazine having the shortest chain.

Comparing carbazole and phenothiazine compounds for the wild type cells, phenothiazines have a higher impact for the cell sensitivity to chloramphenicol. This effect is evident for both types of the cells except in the case of ΔTolC.

We applied electrochemical measurements to determine the  interaction of the  synthe-sized compounds with another efflux pump substrate  –  tetraphenylphosphonium (TPP+). Potentiometric measurements allowed us to follow the  permeability, depolarization of plasma membrane and efflux activity in real time. TPP+ accumulates inside the cells due to plasmamembrane voltage (ΔΨ) and releases out of the  cells after depolarization of inner membrane. TPP+ selective electrode did not show potential change when the measurements were performed with TPP+ and synthesized compounds mixtures without bacteria. These experiments show that there is no interaction between TPP+ and synthesized compounds.

img

Fig. 3. Effects of chloramphenicol and the synthesized compounds on the growth of S. enterica SL1344 (A) and ΔTolC (B) cells. The experiments were performed in LB-Medium. The cells were grown in 96-well plates for 20 hours at 37°C. Concentration of compounds 1–6 was 20 µM. The initial cell concentration was 5 × 105 cfu/mL

In the process of electrochemical measurements S.  enterica  ser. Typhimurium SL1344 cells were added to Tris/HCl buffer and EDTA was used to permeabilize the  bacterial outer membrane (OM). Due to low permeability of the OM to lipophilic compounds and activity of the efflux pumps, Salmonella enterica cells bind low amounts of TPP+ ions. To increase the influx of TPP+ ions, EDTA was used.

The cells accumulated TPP+ after EDTA addition and equilibrium distribution of this indicator ion was achieved in 3 min (Fig. 4A). TPP+ accumulation remained constant over the time without adding synthesized compounds. Increasing concentrations of the  synthesized phenothiazines induced additional uptake of TPP+. 10-Methyl-10H-phenothiazine (1) and 10-ethyl-10H-phenothiazine (2) additions induced stronger uptake of lipophilic TPP+ cation than 10-hexyl-10H-phenothiazine (3), whereas efficiency of efflux inhibition of compounds 4–6 depended on the length of the alkyl chain. The results of the experiments with carbazoles (Fig. 4B) showed that compounds 4 and 5, bearing methyl and ethyl chains, had a weaker effect on TPP+ accumulation than 9-Hexyl-9H-carbazole (6) with the longest alkyl chain. The results demonstrate that efflux-inhibiting efficiency of carbazoles correlate with the length of N-alkyl chain in the molecules. Polycationic antibiotic PMB was used to permeabilize the outer membrane of Gram-negative bacteria and depolarize the cytoplasmic membrane at high concentrations. In order to evaluate the total amount of accumulated TPP+, we used 100 µg/ml addition of PMB which caused the depolarization of cytoplasmic membrane and TPP+ ions were released back to the incubation medium.

img

Fig. 4. Influence of phenothiazines and carbazoles on the accumulation of TPP+ ions in S. en-terica SL1344 cells. The experiments were performed in 100 mM Tris/HCl buffer, pH 8, containing 0.1% glucose, at 37°C. Cells were added to OD600 of 1, EDTA – to 0.1 mM, polymyxin B (PMB) – to 100 μg/ml. Concentrations of compounds 1–6 in μM are indicated in the figures

Determination of susceptibility of S. cerevisiae, C. albicans cells with N-alkylphenothiazines (1–3) and N-alkylcarbazoles (4–6)

Susceptibility of S. cerevisiae wild type W3031a (Fig. 5A) cells and ΔPdr5 (Fig. 5B) mutant to the synthesized phenothiazines and carbazoles was tested (Fig. 5A, B). The ΔPdr5 cells do not contain plasma membrane ABC transporter, which is involved in the resistance against xe-nobiotics (Belofsky  et  al., 2013). The synthe-sized compounds at the  concentrations of up to 160 μM did not affect the growth of S. cerevisiae WT cells. Also, the efflux of mutant cells was not affected at the same range of concentrations and only compound 4 had a minor effect on the growth of mutant cells (Fig. 5A and 5B). The growth of C. albicans cells was more sensitive to the  synthesized compounds, both phenothiazines and carbazoles (Fig. 5C). Methyl derivatives of the synthesized compounds had the  strongest inhibitory effect, and hexyl derivatives the lowest.

img

Fig. 5. Impact of compounds 1–6 on the growth of S. cerevisiae W303-1a (A), ΔPdr5 mutant cells (B) and C. albicans ATCC10231 (C). The screening was carried out in the 96-well plates, the initial concentration of compounds 1–6 was 350 μM (A, B) or 160 μM (C). Cells were grown in YPD (S. cerevisiae) or RPMI-1640 (C. albicans) media at 30°C (S. cerevisiae) or 37°C (C. albicans) for 24 h. The initial cell concentration was 1–5 × 103 cfu/mL

To find out if carbazoles and phenothiazines can work as efflux inhibitors and enhance efficiency antifungal compounds, low concentrations of the synthesized compounds were tested in combination with the well-known drug fluconazole (Flu). This compound is widely used to treat a variety of fungal and yeast infections (Shi  et  al., 2018). It belongs to a  class of anti-fungals called azoles which arrest the growth of the cells by several mechanisms. The best known mechanism includes overexpression or mutations in Erg11p (CYP51p); others change ergosterol metabolism and increase expression of energy-dependent drug efflux (Coste et al., 2007).

Although the  synthesized derivatives of phenothiazines and carbazoles alone affected the  growth of S.  cerevisiae cells at concentrations higher than 180  μM, 20  μM of these compounds increased considerably the  sensitivity of S.  cerevisiae cells to Flu (Fig. 6). In the  case of WT cells, the  most efficient was 10-Methyl-10H-phenothiazine (1). An even stronger Flu supporting effect of compound 1 was observed in experiments with efflux mutant. In general, in the lack of Pdr5 transporter S. cerevisiae cells were much more sensitive to Flu, but the synthesized compounds additionally increased the efficiency of this antifungal. In the case of the mutant strain, the antifungal efficiency of Flu was the highest in the presence of 10-Methyl-10H-phenothiazine (1) and also compounds 4 and 5. In the case of opportunistic fungal pathogen C. albicans, the synthesized compounds enhanced the efficiency of Flu at the concentration of 2.5 μM almost equally.

img

Fig. 6. Impact of fluconazole (Flu) and compounds 1–6 on the growth of S. cerevisiae W3031a (A), ΔPdr5 (B), and C. albicans ATCC10231 (C). Cells were grown in YPD (S. cerevisiae) or RPMI-1640 (C. albicans) media at 30°C (S. cerevisiae) or 37°C (C. albicans) for 24 h. Concentration of compounds 1–6 was 20 µM in experiments with S. cerevisiae or 2.5 µM – with C. albicans. The initial cell concentration was 1–5 × 103 cfu/mL

CONCLUSIONS

In our experiments, the synthesized phenothiazine and carbazole derivatives (1–6) did not have strong bactericidal and fungal properties against the cells of S. enterica, S. cerevisiae, and C.  albicans. However, S.  enterica ΔTolC mutant cells were more sensitive to these compounds than wild type ones. Popular RND family efflux pump inhibitor PAβN increased the  susceptibility of S.  enterica WT cells to the synthesized phenothiazine derivatives and the inhibition level did not depend on the attached alkyl chain. However, in contrast to the  synthesized phenothiazine derivatives, efficiency of the  synthesized carbazoles was dependent on the length of the attached alkyl chain: 9-hexyl-9H-carbazole (6) was the most active. The  results of experiments with antibiotic chloramphenicol confirmed that synthesized compounds acted as competitive inhibitors of efflux and increased the efficiency of antibiotics, especially in the case of S. enterica ΔTolC mutant strain. N-alkylphenothiazines increased the  efficiency of the  antibiotic, but the length of the chain did not produce a significant change on the efficiency.

Potentiometric measurements of the amount of cell accumulated TPP+ confirmed the  conclusion that the inhibitory efficiency of phenothiazines does not depend on alkyl chain length while the efficiency of carbazoles does: the increase in TPP+ accumulation was the highest in the presence of 9-hexyl-9H-carbazole (6). Results of experiments with S. enterica cells indicated that phenothiazine and carbazole derivatives acted as competitive inhibitors in the case of TPP+ efflux.

The following results of our experiments with the  yeast cells showed that the  synthe-sized compounds were non-toxic to the cells of S. cerevisiae, but slightly toxic to those of C. albicans. The data of S. cerevisiae ΔPdr5 mutant cells confirmed significantly the role of efflux in the interaction with phenothiazines and carbazoles. In the presence of the synthesized compounds, the cells of S. cerevisiae and especially of C. albicans were much more sensitive to fluconazole. We can conclude that the synthesized compounds acted as efflux inhibitors in yeast and they work synergistically with fluconazole.

Experiments with yeast cells revealed the  dependence of the  inhibitory activity of the compounds on the attached chain length: in the presence of fluconazole, the methyl groups possessing 10-methyl-10H-phenothiazine (1) and 9-methyl-9H-carbazole (4) had the strongest effect against S.  cerevisiae cells. Combinations of fluconazole and compounds 1–6 showed high activity and effectively inhibited the growth of C. albicans.

This study demonstrated the differences in the  efficiency of the  synthesized compounds against yeast and gram-negative bacteria with evolved effective barriers.

As carbazole and phenothiazine derivatives have wide pharmaceutical activity, it could be considered as potential multidrug resistance (MDR) efflux pump inhibitors (Mahmood et al., 2016), but it requires further investigation.

AKNOWLEDGEMENTS

We thank Prof. Seamus Fanning of University College Dublin for his generous gift of S. enterica strains. We are grateful to Dr Eglė Lastauskienė (Institute of Biosciences, Vilnius University) for Candida albicans ATCC 10231 cells, and to Greta Neveckaitė, Kristina Kacevičiūtė, Ieva Kubiliute, and Justė Krivoščenkaitė for technical assistance.

This study was supported by the  Research Council of Lithuania, funding grants Nos. MIP-040/2015 and TAP LLT-3/2016, and the Ministry of Science and Technology of Taiwan.

CONFLICT OF INTEREST

The  authors declare they have no conflict of interest.

Received 10 February 2020

Accepted 10 March 2020

References

1. Allen  HK, Donato  J, Wang  HH, Cloud-Hansen KA, Davies J, Handelsman J. Call of the wild: antibiotic resistance genes in natural environments. Nat Rev Microbiol. 2010; 8: 251–9.

2. Asche C, Demeunynck M. Antitumor carbazoles. Anticancer Agents Med Chem. 2007; 7(2): 247–67.

3. Brüssow H. Infection therapy: the problem of drug resistance – and possible solutions. Microb Biotechnol. 2017; 10(5): 1041–6.

4. Bolla JM, Brune JM. Enhancing antibiotic activity to combat resistant Gram-negative bacteria: what’s next? Future Med Chem. 2014; 6: 17.

5. Bouaziz  Z, Issa  S, Gentili  J, Gratz  A, Bollacke A, Kassack M, Jose J, Herfindal L, Gausdal  G, Døskeland  SO. Biologically active carbazole derivatives: Focus on oxazinocarbazoles and related compounds. J Enzym Inhib Med Chem. 2015; 30: 180–8.

6. Belofsky  G, Kolaczkowski  M, Adams  E, Schreiber  J, Eisenberg  V, Coleman  CM, Zou  Y, Ferreira D. Fungal ABC transporter-associated activity of isoflavonoids from the root extract of Dalea formosa. J Nat Prod. 2013; 76: 915–25.

7. Carsuo  A, Chiret  ASV, Lancelot  JC, Sinicropi  MS, Garofalo  A, Rault  S. Efficient and Simple Synthesis of 6-Aryl-1, 4-dimethyl-9H-carbazoles. Molecules. 2008; 13: 1312–20.

8. Clausen  JD, Kjellerup  L, Cohrt  KO, Hansen JB, Dalby-Brown W, Winther AL. Elucidation of antimicrobial activity and mechanism of action by N-substituted carbazole derivatives. Bioorg Med Chem Lett. 2017; 27(19): 4564–70.

9. Conchon E, Anizon F, Aboab B, Golsteyn M, Leonce S, Pfeiffer B, Prudhomme M. Synthesis, checkpoint kinase 1 inhibitory properties and in vitro antiproliferative activities of new pyrrolocarbazoles. Eur J Med Chem. 2008; 43: 282–92.

10. Cuestas ML, Castillo AI, Sosnik A, Mathet VL. Downregulation of mdr1 and abcg2 genes is a  mechanism of inhibition of efflux pumps mediated by polymeric amphiphiles. Bioorg Med Chem Lett. 2012; 22(21): 6577–9.

11. Daugelavicius  R, Bamford  JK, Bamford  DH. Changes in host cell energetics in response to bacteriophage PRD1 DNA entry. J Bacteriol. 1997; 179: 5203–10.

12. Gulza N, Klussmann M. Synthesis of antiviral tetrahydrocarbazole derivatives by photochemical and acid-catalyzed C-H functionalization via intermediate peroxides (CHIPS). J Vis Exp. 2014; 20(88): 51504.

13. Hieda Y, Anraku M, Choshi T, Tomida H, Fujioka H, Hatae N, Hori O, Hirose J, Hibino S. Antioxidant effects of the highly-substituted carbazole alkaloids and their related carbazoles. Bioorg Med Chem Lett. 2014; 24(15): 3530–3.

14. Humphries PS, Bersot R, Kincaid J, Mabery E, McCluskie  K, Park  T, Renner  T, Riegler  E, Steinfeld T, Turtle ED, Wei ZL, Willis E. Carbazole-containing sulfonamides and sulfamides: Discovery of cryptochrome modulators as antidiabetic agents. 2016; 26(3): 757–60.

15. Hou Bioorg Med Chem Lett. Y, Che D, Wei D, Wang C, Xie Y, Zhang K, Cao J, Fu J, Zhou N, He  H. Phenothiazine antipsychotics exhibit dual properties in pseudo-allergic reactions: Activating MRGPRX2 and inhibiting the H1 receptor. Mol Immunol. 2019; 111: 118–27.

16. Indumati  T, Fronczek  FR, Prasad  KJR. Synthesis of 2-amino-8-chloro-4-phenyl-5, 11-dihydro-6H-pyrido[2,3-a]carbazole-3-carbonitrile: structural and biological evaluation. J Mol Struct. 2012; 1016: 134–9.

17. Jaszczyszyn  A, Gąsiorowski  K, Świątek  P, Malinka  W, Cieślik-Boczula  K, Petrus  J, Czarnik-Matusewicz B. Chemical structure of phenothiazines and their biological activity. Pharmacol Rep. 2012; 64(1): 16–23.

18. Kossow  A, Kampmeier  S, Willems  S, Berdel  WE, Groll  AH, Burkhardt  B, Rossig  C, Groth C, Idelevich EA, Kipp F, Mellmann A, Stelljes  M. Control of multidrug resistant Pseudomonas aeruginosa in allogeneic hematopoietic stemcell transplant recipients by a novel bundle including remodeling of sanitary and water supply systems. Clin Infect Dis. 2017; 65(6): 935–42.

19. Kantevari S, Yempala T, Surineni G, Sridhar B, Yogeeswari P, Sriram D. Synthesis and antitubercular evaluation of novel dibenzo[b,d]furan and 9-methyl-9H-carbazole derived hexa-hydro-2H-pyrano[3,2-c]quinolines via Povarov reaction. Eur J Med Chem. 2011; 46: 4827–33.

20. Kaur H, Kumar S, Vishwakarma P, Sharma M, Saxena KK, Kumar A. Synthesis and antipsychotic and anticonvulsant activity of some new substituted oxa/thiadiazolylazetidinonyl/thiazolidinonylcarbazoles. Eur J Med Chem. 2012; 45(7): 2777–83.

21. Kinana  AD, Vargiu  AV, May  T, Nikaido  H. Aminoacyl β-naphthylamides as substrates and modulators of AcrB multidrug efflux pump. Proc Natl Acad Sci USA. 2016; 113(5): 1405–10.

22. Lomovskaya O, Warren MS, Lee A, Galazzo J, Fronko  R, Lee  M, Blais  J, Cho  D, Chamberland  S, Renau  T, Leger  R, Hecker  S, Watkins W, Hoshino K, Ishida H, Lee VJ. Identification and characterization of inhibitors of multidrug resistance efflux pumps in Pseudomonas aeruginosa: novel agents for combination therapy. Antimicrob Agents Chemother. 2001; 45(1): 105–16.

23. Liu  ZQ, Tang  YZh, Wu  D. Antioxidant effects of phenothiazine, phenoxazine, and iminostilbene on free-radical-induced oxidation of linoleic acid and DNA. J Phys Org Chem. 200922(10): 1009–14.

24. Michael CA, Dominey-Howes D, Labbate M. The  antibiotic resistance crisis: causes, consequences, and management. Front Public Health. 2014; 2: 145.

25. Mosnaim AD, Ranade VV, Wel ME, Puente J, Valenzuela AM. Phenothiazine molecule provides the basic chemical structure for various classes of pharmacotherapeutic agents. Am J Ther. 2006; 13: 261–73.

26. Mucsi  I, Molnár  J, Motohashi  N. Combination of benzo[a]phenothiazines with acyclovir against herpes simplex virus. Int J Antimicrob Agents. 2001; 18(1): 67–72.

27. Mahmood  HY, Jamshidi  S, Sutton  JM, Rahman KM. Current Advances in Developing Inhibitors of Bacterial Multidrug Efflux Pumps. Curr Med Chem. 2016; 23(10): 1062–81.

28. Nalli Y, Khajuria V, Gupta S, Arora P, Riyaz-Ul-Hassan S, Ahmed Z, Ali A. Four new carbazole alkaloids from Murraya koenigii that display anti-inflammatory and anti-microbial activities. Org Biomol Chem. 2016; 14(12): 3322–32.

29. Opperman  TJ, Nguyen  ST. Recent advances toward a molecular mechanism of efflux pump inhibition. Front Microbiol. 2015; 6: 421.

30. Pluta  B, Morak-Młodawska  M, Jeleń  M. Recent progress in biological activities of synthe-sized phenothiazines, Eur J Med Chem. 2011; 46: 3179–89.

31. Rennison D, Guereta SM, Laitaa O, Blandb RJ, Sutherlandb  IA, Boddyc  IK, Brimble  MA (2016) Substituted carbazoles  –  a  new class of anthelmintic agent. Aust J Chem; 69(11): 1268–1276.

32. Rodrigues L, Aínsa AJ, Amaral L, Viveiros M. Inhibition of drug efflux in mycobacteria with phenothiazines and other putative efflux inhibitors. Recent Pat Antiinfect Drug Discov. 2011; 6(2): 1–10.

33. Sarmiento  GP, Vitale  RG, Afeltra  J, Moltrasio GY, Moglioni AG. Synthesis and antifungal activity of some substituted phenothiazines and related compounds. Eur J Med Chem. 2011; 46(1): 101–5.

34. Spengler  G, Csonka  Á, Molnár  J, Amaral  L. The anticancer activity of the old neuroleptic phenothiazine-type drug Thioridazine. Anticancer Res. 2016; 36(11): 5701–6.

35. Smith  LE. Synthetic organic compounds as potential insecticides. Ing Eng Chem. 1942; 34: 499–501.

36. Sharma S, Srivastava VK, Kumar A. Synthesis and anti-inflammatory activity of some heterocyclic derivatives of phenothiazine. Die Pharmazie. 2005; 60(1): 18–22.

37. Seelig  A, Gottschlich  R, Devant  RM. A  method to determine the ability of drugs to diffuse through the  blood-brain barrier. Proc Natl Acad Sci USA. 1994; 91: 68–72.

38. Sutkuviene S, Mikalayeva V, Pavan S, Berti F, Daugelavicius  R. Evaluation of the  efficiency of synthesized efflux pump inhibitors on Salmonella enterica ser. Typhimurium cells. Chem Biol Drug Des. 2013; 82: 438–45.

39. Simokaitiene  J, Danilevicius  A, Grigalevicius S, Grazulevicius JV, Getautis V, Jankauskas V. Phenotiazinyl-based hydrazones as new hole-transporting materials for electrophoto-graphic photoreceptors. Synth Met. 2006; 156: 926–31.

40. Sun  J, Deng  Z, Yan  A. Bacterial multidrug efflux pumps: mechanisms, physiology and pharmacological exploitations. Biochem Biophys Res Commun. 2014; 453(2): 254–67.

41. Shi J, Li S, Gao A, Zhu K, Zhang H. Tetrandrine enhances the antifungal activity of fluconazole in a murine model of disseminated candidiasis. Phytomedicine. 2018; 46: 21–31.

42. Warman  AJ, Rito  TS, Fisher  NE, Moss  DM, Berry NG, O’Neill PM, Ward SA, Biagini GA. Antitubercular pharmacodynamics of phenothiazines. J Antimicorb Chemother. 2013; 68(4): 869–80.

43. Zhang FF, Gan LL, Zhou CH. Synthesis, antibacterial and antifungal activities of some carbazole derivatives. Bioorg Med Chem Lett. 2010; 20(6): 1881–4.

* Corresponding author. Email: simona.sutkuviene@vdu.lt

Simona Sutkuvienė, Sandra Sakalauskaitė, Neringa Kuliešienė, Lina Ragelienė, Rimantas Daugelavičius

SALMONELLA ENTERICA SER. TYPHIMURIUM, SACCHAROMYCES CEREVISIAE IR CANDIDA ALBICANS LĄSTELIŲ SUSINTETINTŲ 9H-KARBAZOLO IR 10H-FENOTIAZINO DARINIŲ ANTIMIKROBINIO AKTYVUMO ĮVERTINIMAS

Santrauka

Šiame tyrime mes susintetinome įvairaus ilgio alkilo pakaitus turinčius 9H-karbazolo ir 10H-fenotiazino darinius bei įvertinome jų antimikrobinį ir daugiavaisčio atsparumo siurblius slopinantį poveikį Salmonella enterica ser. Typhimurium, Saccharomyces cerevisiae ir Candida albicans ląstelėms. Šie dariniai yra svarbūs dėl plataus jų pritaikymo spektro, ypač medicininėje chemijoje vertinamas jų farmakologinis aktyvumas. Mūsų tyrimo rezultatai atskleidė, kad didėjantis karbazolų alkilo grandinių ilgis padidino tetrafenilfosfonio (TPP+) jonų kaupimąsi ląstelėje. S.  enterica mutantinės ΔTolC ląstelės buvo jautresnės susintetintiems junginiams. Junginiai sustiprino flukonazolo poveikį S.  cerevisiae mielių ląstelėse. Mutantinės padermės ΔPdr5 ląstelės buvo jautresnės tiriamiems karbazolo ir fenotiazino dariniams. Junginiai su trumpesne alkilo grandine (10-metil-10H-fenotiazinas ir 9-metil-9H-karbazolas) efektyviausiai slopino Candida albicans ląstelių gyvybingumą.

Raktažodžiai: fenotiazinas, karbazolas, tetrafe-nilfosfonio jonai, minimali slopinančioji koncen-tracija, išmetimo siurbliai, Salmonella enterica, Saccharomyces cerevisiae, Candida albicans