1 Introduction

In everyday microbiological practice, both Gram-positive and Gram-negative strains are used. These Gram-negative bacteria are in the so-called “S” form, with a smooth colony on the agar plate, and their lipopolysaccharides (LPS) are complete consisting of a lipid A core and an O-side chain polysaccharide. We know that after mutation in biosynthesis of LPS, sometimes the so-called rough “R” colonies can grow on agar plate, and in this case, the chains of LPS are shorter. These mutant strains may be more sensitive to antimicrobial drugs [1,2,3]. Currently, the antimicrobial activity of essential oils (EOs) is an outstanding research field due to the antibiotic resistance of microorganisms. The classical microbiological assays (e.g., disc diffusion and agar dilution) are not appropriate for the antimicrobial testing of EOs, because these plant extracts are volatile and nonwater soluble. Therefore, there is a need to find a method that is suitable for testing EOs with multicomponent composition. Some previous studies proved that thin-layer chromatography (TLC) is an applicable method for the detection of antimicrobial compounds in various matrices, e.g., in EOs [4]. Bioautography is a useful and effective method to find a compound with antimicrobial activity in a mixture extracted from plant material. The sensitivity and the success of this method depend mostly on the sensitivity of the test bacterium to the compound under screening. During direct bioautography (DB), microorganism cultures grow directly on the surface of TLC plate. Each step of the method is performed on the plate. After development of the TLC plate, it is dipped into a cell suspension of test microorganism and placed into a humid chamber to provide appropriate conditions for biological functions on the surface of the TLC plate. The inhibition zones of the antimicrobial compounds can be visualized by tetrazolium salt, e.g., 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT). On the TLC plate, metabolically active bacteria can convert MTT into formazan dye (blue or purple color). In the white inhibition zones produced by antibacterial separated components, the above-mentioned chemical reaction cannot occur due to the lack of dehydrogenase activity [5]. TLC‒DB can be expected to produce rapid results. Previously, we could demonstrate that some plant pathogen bacteria, Xanthomonas campestris pv. vesicatoria and Pseudomonas syringae pv. phaseolicola, provided fast result (after 2 h incubation time) and were highly sensitive to cinnamon oil and its main compounds (trans-cinnamic aldehyde and eugenol) [6]. It was clearly demonstrated that this analytical toolbox is beneficial in the discovery of antibiotic compounds, allowing a fast, purposive, and, thus, cost-effective tracking, isolation and identification of antibacterial compounds that can become promising drug candidates [7]. Now the aim of this study is to analyze the usefulness of rough mutant test bacteria with more sensitivity to antimicrobial compound in the bioautography assay. The antibacterial activity of three EOs (thyme, clove, and peppermint) were included in the present study due to their economical and therapeutic aspects. In our previous experiment, these oils exerted anti-Haemophilus activity [8]; therefore, we wanted to explore their “behavior” against rough bacterial mutant strains used in this study. As far as we know, these rough mutants have never been tested against thyme, clove and peppermint EOs yet.

2 Experimental

2.1 Essential oils and their components

The EO of thyme (batch number H3981/1509), clove (batch number H7352/1602), and peppermint (batch number H7101/1601) were obtained from a Hungarian company (AROMAX Zrt., Budapest, Hungary). The main components of the EOs (thymol, eugenol, and menthol) were bought from Sigma-Aldrich (Budapest, Hungary).

2.2 gas chromatography‒mass spectrometry (GC‒MS) and Gas chromatography‒flame ionization detection (GC‒FID)

The chemical and the percentage compositions of the EO samples were determined with GC‒MS. The parameters were the same as described in our previous paper [8]. An Agilent 6890N/5973N GC-MSD (Santa Clara, CA) system equipped with an Agilent SLB-5MS capillary column (30 m × 250 µm × 0.25 µm) was used. An aliquot of 10 µL EO was diluted in 1 mL ethanol, and from this solution, 1 µL was injected in split mode (the split ratio was 1:50). The injector temperature was 250 °C. The oven temperature was increased at a rate of 8 °C/min from 60 °C (3 min isothermal) to 250 °C, with a final isotherm at 250 °C for 1 min. High purity helium was used as the carrier gas (1.0 mL/min and 37 cm/s, in constant flow mode). The mass selective detector (MSD) was equipped with a quadrupole mass analyzer and was operated in electron ionization mode at 70 eV in full scan mode (41–500 amu at 3.2 scan/s). The data were evaluated using MSD ChemStation D.02.00.275 software (Agilent). The identification of the compounds was carried out by comparing retention times, linear retention indexes, and recorded spectra with the data of authentic standards, and the NIST 2.0 library was also used. Gas chromatography‒flame ionization detection (GC‒FID) was made using a Fisons GC 8000 gas chromatograph (Carlo Erba, Italy). An Rt-β-DEXm (Restek) capillary column, 30 m × 0.25 mm i.d., 0.25 μm film thickness, was used. The carrier gas was nitrogen at 6.8 mL/min flow rate. A 0.2 mL volume of a 0.1% solution of the oil was injected (1 μL EO in 1 mL chloroform). The splitless injection was carried out. The temperatures of the injector and detector were 210 and 240 °C, respectively. The oven temperature was increased at a rate of 8 °C/min from 60 to 230 °C, with a final isotherm at 230 °C for 5 min. Identification of peaks was made by retention data compared with data obtained by GC‒MS and data of standards (Fluka Analytical and Sigma-Aldrich); percentage evaluation was carried out by area normalization. Three parallel measurements were made, and relative standard deviation (RSD) percentages were below 4.5%.

2.3 Microbiological tests

2.3.1 Cultivation of test bacteria

In our microbiological experiments, the following strains were used: Salmonella minnesota Re595 was kindly supplied by O. Lüderitz (Freiburg, Germany). The Escherichia coli ReD31m4 strains were kindly supplied by N. Kato, T. Hasegawa (Nagoya University, Japan). Other microorganisms were isolated from our experiments when we studied the cell wall structure of Shigella sonnei strains, for example, S. sonnei phase I (this is a smooth “S” form) and phase II 4303 (this is a rough “R” mutant) [9] and S. sonnei R4341, R4350, and R4351 (these are “R” mutants). Endotoxins were isolated from S. sonnei rough mutants as well. The S. sonnei 4303 strain is a phase II type and is a spontaneous mutant of S. sonnei phase I [9]. After ethyl methanesulfonate treatment of S. sonnei 4303, we could isolate a great number of rough mutants, and we characterized these mutants by core specific phages [10]. The bacteria of these groups were maintained and cultivated in Mueller‒Hinton broth.

2.3.2 Sugar content of core region of “S” form and “R” mutant strains

Phenol‒water method was used for extraction of LPS of S. sonnei phase I (smooth mutant) after Westphal [11]. A phenol‒chloroform‒petroleum ether method was suitable to isolate the endotoxins of rough mutants (R, LPS and LOS, lipooligosaccharide) after Galanos [12] and then was precipitated and freeze-dried. Determination of monosaccharide components in core region of S and R endotoxin was performed with the following method: the LPSs and LOSs were hydrolyzed by sulfuric acid. and the monosaccharides were identified in an alditol acetate form by GC [13]. Table 1 presents these data. R mutant strains were maintained at the Department of Medical Microbiology and Immunology, University of Pécs (Hungary). In the case of the DB assay, bacteria were grown in 100 mL Mueller‒Hinton broth (Sigma Aldrich Ltd., Darmstadt, Germany) at 37 °C in a shaker incubator (at 60 rpm speed for 24 h) [14]. The bacterial suspension was diluted with fresh nutrient broth to an optical density at 600 nm (OD600) of 0.4, which corresponds to approximately 4 × 107 colony-forming units (cfu)/mL.

Table 1 Sugar content of core region of “S” form and rough “R” mutant strains
Full size table

2.3.3 Determination of minimum inhibitory concentration (MIC)

Macrotube dilution methods were used to detect the minimum inhibitory concentrations (MICs) of the test strains to synthetic antibiotics (cephalexin, gentamicin, tobramycin, and erythromycin). The MIC value is the lowest concentration of antibiotics could kill the 90% of test bacteria [15]. The data of results are presented in Table 2.

Table 2 Minimum inhibitory concentration (MIC) values of smooth (“S”) and rough (“R”) Gram-negative strains depend on the cell wall structure of bacteria and chemical composition of antimicrobials
Full size table

2.3.4 Thin-layer chromatography

For bioautography 5 cm × 10 cm silica gel 60 F254 aluminum sheet TLC plates (Merck, Darmstadt, Germany) were used. The positive control antibiotics, cephalexin and gentamicin (see in chapter 2.3.5), were tested with DB. After this experiment, the antibacterial effect of thyme (Thymus vulgaris L., Lamiaceae), clove (Syzygium aromaticum (L.) Merr. & L. M. Perry, Myrtaceae), and peppermint (Mentha × piperita L., Lamiaceae) EOs was investigated against the most sensitive rough bacterium S. minnesota Re595 and against the least sensitive rough mutant E. coli D31m4. Ethanol solutions (100 µL oil in 500 mL absolute ethanol) of EOs were prepared. From these solutions, 0.2, 0.4, 0.5, and 1.0 µL were applied to the TLC plate. The antibacterial activity of the main EO components (thymol, eugenol, and menthol) were also investigated by TLC‒DB. Ethanol solutions of thymol, eugenol, and menthol (20 mg/mL) were made, and an aliquot of 0.5 µL (10 µg) of the stock solution was applied to the TLC plates. The position of the starting line was 1.5 cm from the bottom and 1.5 cm from the left side. The mobile phase toluene‒ethyl acetate (95:5, V/V) was used for chromatographic separation [16]. Ascendant development in a saturated twin-trough chamber (CAMAG, Muttenz, Switzerland) was performed at room temperature (20 °C). Then, the TLC plates were dried at 90 °C for 5 min to remove the solvent completely. The visualization of the separated compounds was detected by ethanolic vanillin‒sulfuric acid reagent [5]. The identification of the separated compounds was done based on retardation factor (RF) values, color of the standards, and under ultraviolet (UV) light at 254 nm. The photos of the plates were done with digital camera (Lumix DMC-LX100, Panasonic, Osaka, Japan). It should be highlighted that the TLC plates prepared for bioautography were not treated with vanillin‒sulfuric acid, because this step interferes with the microbiological steps in bioautography.

2.3.5 Direct bioautography

The positive controls were cephalexin (0.5 mg/mL, Hungaro-Gal Kft. Kaposvár, Hungary) and gentamicin (Gentamicin 40 mg/mL, Sandoz Hungary Kft., Hungary) antibiotics. These antibiotics were dissolved in sterile water. From the cephalexin solution, 0.2, 0.4, 0.8, and 1.6 µL were applied to the 5 cm × 10 cm silica gel adsorbent with automatic pipette (Merck, Darmstadt, Germany). An aliquot of 0.2 µL from gentamicin was applied to the TLC plate. Motic Images Plus 2.0 software (Motic, Hong Kong, China) was used to measure the inhibitory zones (expressed in mm) produced by cephalexin and gentamicin. In this case, chromatographic separation was not done. In the case of DB with antibiotics or with EO samples, the TLC plates were dipped into a 100 mL of bacterial suspension to assure a homogeneous distribution and adhesion of bacteria onto the surface of the layers. Immersion time was 10 s, and then, the layers were placed into a low-wall horizontal chamber (chamber dimension: 20 cm × 14.5 cm × 5 cm) and incubated at 37 °C for 2 h. Then, the TLC plates were immersed into the aqueous solution of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT, 50 mg/85 mL) (Sigma Aldrich Ltd., Darmstadt, Germany) using CAMAG immersion device for 5 s to visualize the antibacterial spots on the silica plates. Incubation of the layers (37 °C for 4 h) was the next step. The metabolically active bacteria can convert the tetrazolium salt, MTT, into formazan dye. Inhibition zones as white spots will appear against the bluish-violet background if there is no dehydrogenase activity on the plate due to the antibacterial activity of the tested EOs or their main compounds [17, 18]. The inhibition zones (expressed in mm) of the tested antibiotics without separation were measured with Motic Images Plus 2.0 program (ver. 2.0., Motic, Hong Kong, China). The experiments were prepared in six parallel repetitions.

3 Results

3.1 Chemical composition of EOs

Chemical analyses of EOs were performed by GC‒MS technique. The identified compounds and percentage evaluation of the oils are presented in Table S1 (Supplementary Material). Thyme oil contained thymol in the highest percentage (39.8%). Eugenol was the dominant compound (78.8%) in the clove essential oil. In the peppermint oil, menthol (50.4%) was identified as the main component [8].

3.2 Sugar content of core region of “S” form and rough “R” mutant strains

To know exactly the sugar composition of the test bacteria included in this experiment, GC was performed in the case of “S” form and rough “R” strains. Table 1 presents that S. sonnei phase I has a complete LPS consisting of O-side chain, core monosaccharides, and lipid A. From the S. sonnei phase II, the O-side chain of LPS is missing. From the next LPSs, more exactly from LOSs, monosaccharides are missing including: galactose, glucose, ld-heptose. S. sonnei R4351 is an interesting mutant with d-glycero-d-mannoheptose in its LOS. This monosaccharide is precursor of l-glycero-d-mannoheptose. The S. sonnei Re4350, S. minnesota Re595, and E. coli ReD31m4 are so-called absolute “R” rough mutants; in their LOS, lipid A can only be found.

3.3 Minimum inhibitory concentration (MIC)

The MIC values of rough strains were measured by macrotube dilution methods. Table 2 presents the MIC values that divided into two groups: the MIC values for aminoglyosides (gentamicin and tobramycin) was a low level of 0.25‒2.00 µg/mL. The target for these antibiotics is the ribosomal protein synthesis. These MIC values were nearly independent from mutation in cell wall. The MIC values for the other group of antibiotics (erythromycin and cephalexin) were impressively influenced by mutation in LOS, in the outer membrane of Gram-negative mutants (125.00‒7.80 µg/mL). After these, the MIC values we decided to use rough, more exactly absolute rough mutants in DB.

3.4 Antibacterial activity of cephalexin and gentamicin by direct bioautography

In DB, the activity of the antibiotics was tested against “R” mutants of E. coli D31m4, S. minnesota Re595, S. sonnei phase II 4303, and S. sonnei 4341, 4350, and 4351. Figure 1 shows the antibacterial activity of the antibiotics (cephalexin and gentamicin). The diameter of the inhibition zones was expressed in mm (Fig. 1). From the stock solution (0.5 mg/mL) of cephalexin, 0.2, 0.4, 0.8, and 1.6 µL (equivalent to 0.1, 0.2, 0.4, and 0.8 µg antibiotic) were applied, and from the stock solution (40 mg/mL) of gentamicin, an aliquot of 0.2 µL (equivalent to 8 µg antibiotic) was applied on the TLC plate. Gentamicin (8 µg) resulted in quite similar inhibition zones for each bacterium (4.1–6.9 mm) (Fig. 1). The most sensitive bacterium was S. minnesota Re595 (inhibition zone 6.9 mm), and the most resistant bacteria were S. sonnei 4303 and E. coli D31m4 (inhibition zones 4.1 mm and 4.3 mm). The cephalexin treatment resulted in different inhibition zones. We experienced that if we increased the quantity of cephalexin, the diameter of the inhibition zones became higher (concentration dependent activity). Among the test bacteria, S. minnesota Re595 was the most sensitive to the cephalexin (inhibition zone 0.1 µg‒7.7 mm and 0.8 µg‒12.1 mm). The most resistant bacteria were E. coli D31m4 and S. sonnnei 4303 because these did not show inhibition zones at 0.1 µg of cephalexin. The 0.8 µg cephalexin resulted in inhibition zones for all bacteria: E. coli D31m4 (2.6 mm), S. sonnei 4303 (5.5 mm), S. sonnei R4341 (6.9 mm), S. sonnei Re4350 (7.8 mm), and S. sonnei 4351 (8.2 mm) (Fig. 1). The results are illustrated in the case of 4303 S. sonnei phase II and S. minnesota Re595 in Fig. 2.

Fig. 1
figure 1

The inhibition zones of cephalexin (Ceph) and gentamicin (Gent) antibiotics determined with DB. The diameter of the inhibition zones was expressed in mm

Full size image
Fig. 2
figure 2

Antibacterial activity of cephalexin (Ceph) and gentamicin (Gent) determined with DB using Shigella sonnei phase II (4303) (A and B) and Salmonella minnesota Re595 (Re595, absolute rough mutant) (C and D)

Full size image

3.5 Antibacterial activity of the main components of EOs by TLC‒DB method

We investigated the antibacterial effect of thyme, clove, and peppermint EOs and their main components after TLC separation against the most sensitive S. minnesota Re595 and the least sensitive E. coli ReD31m4 based on TLC‒DB performed without separation. In this case, we experienced similar result, in which S. minnesota Re595 was the most sensitive and the E. coli strain was the most resistant to the EOs and their components. Eugenol (RF = 0.52), as the main compound of the clove oil, was active (Fig. 3). In the peppermint oil, several compounds had antibacterial activity at the tested concentration. Menthol (RF = 0.31) in the peppermint oil inhibited the growth of bacteria (Fig. 3). Other active compounds of peppermint oil include 1,8-cineole, isomenthone, menthone, and isomenthyl acetate. These components were identified as active compounds according to GC‒MS result and Wagner and Bladt [18]. In the oil of thyme, thymol–carvacrol and the standard, thymol, had antibacterial activity (Fig. 3). At RF = 0.33, linalool was identified as an active compound according to the GC‒MS result and Wagner and Bladt, too [18].

Fig. 3
figure 3

Antibacterial components of the EOs used in this study by TLC‒DB in case of “R” mutants. A TLC plate (UV 254 nm). B TLC plate after treatment with vanillin‒sulfuric acid reagent (Vis). C TLC‒DB plate, bacterium: “R” form E. coli ReD31m4. D TLC‒DB plate, bacterium: “R” form S. minnesota Re595 (bright zones indicate antibacterial effects) Cl clove oil, Pp peppermint oil, Th thyme oil, S1 eugenol standard, S2 menthol standard, S3 thymol standard

Full size image

Based on Fig. 3, our results showed, that the peppermint is a promising antibacterial EO against S. minnesota Re595. This oil resulted in the most antibacterial components. After testing the EOs on “R” mutant form of E. coli ReD31m4 and S. minnesota Re595, we wanted to compare the effectiveness of peppermint EO on the “S” form and “R” mutant bacterium. S. sonnei phase I “S” form and “R” mutant of S. sonnei R4351 were used. In the case of peppermint oil, the difference between “S” and “R” strains is clearly visible in Fig. 4. By the S. sonnei “S” form, the inhibition zones are not so clear; moreover, only the main compound of peppermint oil (menthol, RF = 0.31) formed the inhibition zone, and the minor components did not show antibacterial effect (the experimental parameters were the same).

Fig. 4
figure 4

Antibacterial components of the peppermint oil used in this study by TLC‒DB using A S. sonnei “R” mutant and B S. sonnei “S” form. A TLC‒DB plate, bacterium: “R” form S. sonnei R4351. B TLC‒DB plate, bacterium: “S” form S. sonnei phase I (bright zones indicate antibacterial effects of peppermint oil components, see Fig. 3)

Full size image

4 Discussion

DB offers a great possibility when we want to determine antimicrobial activity of a component in an extract isolated from plants [4,5,6,7,8]. The sensitivity of this method depends mostly on the sensitivity of test bacterium to the detectable component. Test strains used generally in this method include Gram-positives, e.g., Staphylococcus aureus ATCC 29213, Micrococcus luteus ATCC 9341, and Bacillus subtilis ATCC 6633, or Gram-negatives, e.g., E. coli ATCC 25922, Pseudomonas aeruginosa ATCC 15442, and so on. These strains are so-called “S” smooth types—their colonies have a smooth surface and their lipopolysaccharides (LPSs) are complete, containing the O-specific side chain. After mutation in LPS biosynthesis, the so-called “R” rough mutants can be isolated. The surface of their colonies is rough, and core monosaccharides are missing from their LPS [19]. This type of cell wall is more penetrable for different compounds, and sometimes these mutants are more sensitive to antimicrobials than the original smooth strains [1, 2]. There were two types of drugs (antibiotics) in our study: (1) aminoglycosides, such as gentamicin, have targets in ribosomes and their sensitivity is not influenced by cell wall mutation, and (2) other antimicrobials, such as β-lactams, can penetrate through the mutated cell wall easily; therefore, the strain is more sensitive. First, our rough mutants were analyzed by classic GC (see Table 2). Later the sugar content of LPS extracted from these rough mutants was checked by more sophisticated methods: GC‒MS and LC‒MS [20,21,22]. In our study, the antibacterial activity of three EOs (thyme, clove, and peppermint) was tested by TLC‒DB method against rough “R” or smooth “S” type bacterial strains to increase the sensitivity of this assay and reduce the time of the experiment. First, two positive controls (gentamicin and cephalexin) were tested against S. sonnei “R” mutant and “R” mutants of S. minnesota and E. coli in DB. Based on the diameter of inhibition zones (expressed in mm), the most sensitive strain was S. minnesota (Figs. 1 and 2). Then, the EO samples of thyme, clove, and peppermint were tested against the most sensitive S. minnesota and the least sensitive E. coli D31m4 in TLC‒DB. In this case, chromatographic separation was done. We also realized that S. minnesota showed the highest sensitivity against EOs (Fig. 3). Finally, the peppermint oil was tested against “S” form and “R” mutant of S. sonnei to compare their “behavior” in the TLC‒DB assay. The peppermint oil produced more inhibition zones in case of “R” mutant of S. sonnei (Fig. 4). Our findings in the TLC‒DB assay could support the previous hypothesis that “R” mutant Gram-negative strains are more sensitive to antibiotics, plant EOs, and their main compounds tested in this study than “S” form of bacterial strains.

This finding was supported by the fact that “R” mutants have different cell-wall composition; therefore, the antibiotics and the components of the tested oil samples can penetrate more easily through the cell wall. General experience is that Gram-positive bacteria are more sensitive to different compounds including synthetic or plant-origin as well than Gram-negative bacteria. However, if we have an “R” mutant Gram-negative strain, this statement could be opposite. We supported this fact in our study. The explanation is based on the composition of the cell wall with a changed structure. In terms of the mechanism of action of EOs, their most important point of action is their ability to influence the permeability of the bacterial cell membrane, thus interfering with ion transport (K+, Ca2+, Na+), and they can also induce protein denaturation processes [16]. As a result, enzymatic processes are inhibited [23]. Cell death occurs as a result of the change in cell membrane permeability, disruption of ion transport, inhibition of mitochondrial processes, and disruption of the intracellular–extracellular adenosine triphosphate (ATP) balance [24].

5 Conclusions

In our study, we describe and compare the sensitivity of rough “R” mutants to antibiotics using a TLC‒DB system. In addition, we identify the antibacterial effect of thyme, clove, and peppermint EOs and their components, thymol, eugenol, and menthol, in the case of “R” mutants. We realized that our Gram-negative rough “R” mutants are more sensitive to antimicrobial drugs than smooth “S” mutants. In the TLC‒DB method, we suggest to use rough strains, mostly S. minnesota Re595, as this is a well-known strain in the microbiological laboratories all over the world. Our results offer a fast (4 h incubation time), sensitive (with “R” Gram-negative mutant strain), and cost-effective (TLC‒DB) screening method to receive information on the antibacterial activity of the mixture of hydrophobic, volatile compounds, e.g., EO samples.