Correlation between chemical composition and antibacterial activity of some Lamiaceae species essential oils from Tunisia

Background Lamiaceae family is one of the most diverse and common plant families in terms of ethnomedicine due to their potential therapeutic effects. The aim of this study is to investigate the correlation between the chemical composition and the antibacterial effect of five essential oils from this family against five reference bacterial strains responsible of nosocomial diseases and foodborne illnesses. Methods The commercial essential oils of Tunisian Rosmarinus officinalis, Thymus capitatus, Origanum majorana and Salvia officinalis were analyzed by GC/FID and GC-MS. Essential oils were evaluated for their antibacterial activities by disc diffusion and microbroth dilution methods against five reference bacterial strains: Pseudomonas aeruginosa, Escherichia coli, Salmonella enterica, Bacillus subtilis and Staphylococcus aureus. The inhibition zone diameter values and the twenty major compounds of the selected essential oils were subjected to PCA and HCA analysis. Results Analysis by GC/FID and GC/MS allowed the identification of ninety-one components representing 96.0 to 98.2% of the total oils. The different component contents varied according to the species. The main components were carvacrol, 1,8-cineole, α-thujone, α-terpineol and α-pinene. The PCA and the HCA of the selected essential oil components and the inhibition zone diameter (IZD) values identified four species groups and subgroups. Each essential oils group constituted a chemotype responsible for their bacterial inhibition ability. Thymus capitatus essential oil showed the strongest antibacterial activity with MBC ranging from 0.73 to 2.94 mg mL− 1. Conclusion Rosmarinus officinalis, Thymus capitatus, Origanum majorana and Salvia officinalis essential oils have shown promising antibacterial activities against reference bacterial strains responsible for nosocomial diseases and foodborne illnesses.


Background
In the last few years, the interest in natural medicine has been increasing in industrialized societies particularly against microbial agents because of the ever growing problem of antibiotic resistance [1]. Natural products are an important source of novel and chemically diverse molecules for drug discovery [2]. In fact, because of their bioactive compounds and their antimicrobial properties, the interest in medicinal plants has increased considerably [3]. Actually, 69% of currently employed antibiotics are derived from natural products [4]. Essential oils are volatile fractions obtained from medicinal and aromatic plants, they have been screened for their potential uses as alternative remedies for the treatment of many infections and as natural food preservatives [5][6][7][8]. Lamiaceae family is one of the most diverse and widespread plant families in term of ethnomedicine and its medicinal value is mainly based on the volatile oils composition [9]. It was reported that the leaves of T. capitatus were used for many purposes such as unguents for embalming by Egyptians [10], antispasmolytic, antiseptic, expectorant, anti-ulcers, anti-dermatitis and rheumatic pains [11]. T. capitatus showed good results in treatment of skin diseases and as tonic and antitussive agent [12]. S. officinalis is used for its antibacterial [13,14], antiviral [15] and antioxidant [16] activities. It is also used as a flavoring agent in perfumery and cosmetics [17]. R. officinalis has been traditionally used for both medical and culinary purposes [18]. Indeed, the whole fresh plant or its dried milled powder is used as a food flavoring and preservative agent [19]. Likewise, R. officinalis is known as a medicinal herb with significant activities against many illnesses especially against headaches, respiratory diseases and several neuropsychiatric disorders, according to traditional Tunisian medicine [20]. O. majorana L. is used as a medicinal plant; it has traditionally been used as stimulant, condiment, and tonic [21]. In addition, O. majorana L. essential oil is used for its antibacterial [22] and antioxidant activities [21].
Because of their interesting physicochemical characteristics of substantial value, the Lamiaceae essential oils have garnered research and industrial interests for their use as natural products [23]. However, in Tunisia, apart from R. officinalis EO, Lamiaceae EOs production is still weak despite the plants abundance and scientific researches on these EOs bioactivities [14,24,25]. Essential oils chemical composition is affected by several factors, including the species, part of the plant, season of harvesting, geographical origin, and also the extraction method and conditions which consequently impact on their bioactive properties [26][27][28][29]. Hence, the present work aimed to study the chemical composition of five commercial Lamiaceae EOs from Tunisian provenance and their antibacterial effects against five pathogen bacteria models. A correlation between 20 major compounds and their antibacterial ability was investigated.
The essential oils were extracted by hydrodistillation of fresh aerial parts. For each species, 3 Samples of the obtained EO were stored at 4°C until analysis was attempted.

Chemical analysis
Quantitative and qualitative data of all the essential oils were determined in triplicate by GC/FID and GC/MS respectively.

Gas chromatography analysis
One microliter sample of 10% solution of essential oil in purified hexane was injected for analysis using a HP 6890 chromatography apparatus equipped with flame ionization detector (FID) and HP5 column (30 m × 0.25 mm, film thickness 0.25 μm). The analytical conditions were as follows: injector and detector temperature were maintained at 250°C and 280°C respectively. Oven temperature was programmed to rise from 60°C to 250°C at 5°C/min, isothermal temperature 60°C for 1 min and 250°C for 3 min; carrier gas was N 2 with a flow rate of 1.2/min. The software HP Chemstation, allowing assimilation of the percentages of the peak areas to the percentages of the various constituents was used to calculate relative concentration. Kovats retention Indexes (KI) were obtained by running a series of aliphatic hydrocarbons (C9-C28) increasing number order of carbon atoms on the column.

Gas chromatography-mass-spectrometry analysis
The essential oils analysis was performed using a Hewlett Packard (HP) 5890 II GC equipped with a HP-5MS capillary column (30 m × 0.25 mm, film thickness 0.25 μm) and a HP 5972 mass selective detector. Helium was used as the carrier gas at a flow rate of 1.2 ml/min. For GC/MS detection, an electron ionization system, with ionization energy of 70 eV, a scan time of 1.5 s and mass range 40-300 amu was used. The GC/MS parameters were identical to those for the GC/FID analysis.

Compound identification
The compounds were identified by comparing retention indices (determined relatively to the retention time of aliphatic hydrocarbons (C9-C28)) of the mass spectra with those of authentic compounds by means of Wiley 275.L, 6th edition mass spectral library and NBS75K.L data bases. The identification is confirmed by comparing their retention indices with data published in the literature [30,31].

Antibacterial assay Bacterial strains
Three Gram negative bacteria (Gram (−)), Pseudomonas aeruginosa (CIP 82118), Salmonella enterica (CIP 80.39), Escherichia coli (CIP 53.126) and two Gram positive (gram (+)) bacteria Staphylococcus aureus (CIP 4.83) and Bacillus subtilis (CIP 52.62) were used to carry out this study. Microorganisms were obtained from the culture collection of the laboratory of transmissible diseases and biologically active substances, Faculty of Pharmacy, Monastir, Tunisia.

Agar disk diffusion method
To evaluate the antibacterial activity of the selected EOs, the agar disk diffusion assay [32] was carried out. Overnight cultures (24 h at 37°C) in MH broth were prepared. The bacterial inocula were obtained by diluting the overnight cultures in Mueller Hinton Broth medium to adjust the optical density at approximately 0.5 Mc Farland standards. Absorbent discs (Whatman disc No.3, 6 mm diameter) were impregnated with 10 μL of essential oil and then placed on the surface of inoculated Petri dishes. Positive control discs of standard antibiotic GENTAMICIN® (10 μg /disc) were tested. After 24 h of incubation at 37°C, the inhibition zones were measured and expressed in mm. All experiments were performed in triplicate.

Minimum inhibitory and minimum bactericidal concentrations
For each essential oil, the minimum inhibitory concentration (MIC) and minimum bactericidal concentration (MBC) values against the bacterial strains were determined according to the National committee for clinical Laboratory Standard [33]. An overnight culture (37°C) in Mueller Hinton broth medium of the tested strains was prepared by adjusting the turbidity of each bacterial culture to reach an optical density of 0.5 Mc Farland standards.
The broth dilution method was carried out in 96-well microtitre plates using bacterial reference strains. To obtain stable diffusion, EOs stock solutions were prepared aseptically in DMSO (80, 20%, v/v), transferred to sterile 96-well microtitre plates containing the Mueller Hinton medium and then diluted by two-fold serial dilution. Ten microliter of each strain inoculum was added to each well. Negative control wells contained only bacteria in the MH broth medium. Positive control wells contained 10 μg mL − 1 Gentamicin® antibiotic. After incubation for 24 h at 37°C, 10 μL of 0.02% (p/v) resazurin solution was added. Bacterial growth was evaluated by the observation of the wells color change from blue to pink. The MIC was defined as the lowest concentration that completely inhibit visible cell growth during a 24 h incubation at 37°C (blue colored). DMSO 20% (v/v) had no inhibition effect. To determine the minimum bactericidal concentration (MBC) values, 10 μL of each well with no visible growth were removed and inoculated in MH plates. After 24 h of incubation at 37°C, the number of surviving organisms was determined. MBC was defined as the lowest concentration at which 99% of the bacteria were killed. The experiments were performed in triplicate.

Statistical analysis
The data were subjected to statistical analysis using analysis of variance (ANOVA), and the significance of the difference between means was assumed at p ≤ 0.05 using Student Newman and Keuls (SNK) test. Twenty Compounds detected in the essential oil samples at an average concentration greater than 3.0% were selected and used to evaluate whether the identified compounds may be useful in reflecting chemotaxonomic and biological activities relationships. Data of antibacterial activity and components were subjected to principal components (PCA) and hierarchical cluster analysis (HCA). Statistical tests were performed using IBM® SPSS® 20 software (SPSS INC.Chicago,IL, USA).

Principal component analysis (PCA) and hierarchical cluster analysis (HCA) Chemical classes
The contents in different chemical classes for the five EOs was significantly different between the species (p < 0.05). The PCA horizontal axis explained 34.12% of the total variance and the vertical 21.26% (Fig. 2). The HCA (Fig. 3) based on the Euclidean distances between groups indicated the presence of three groups (A, B and C) identified by their contents in different chemical classes with a dissimilarity > 15. Group C was divided in two Subgroups (C 1 and C 2 ) with a dissimilarity > 5. The two species T. capitatus and O. majorana stand out in both HCA and PCA analyses, forming separate groups, group A and B, respectively. Group A formed by T. capitatus which EO was distinguished by the highest content in monoterpene phenols (56.6 ± 2.8%) and the smallest contents in monoterpene alcohols (4.6 ± 1.0%), monoterpene oxides (2.9 ± 4.6%) and monoterpene ketones

Major components
Twenty components of the essential oils with a minimum content of 3.0% in at least one species were selected for the PCA and the HCA to evaluate whether they may be useful in reflecting the chemotaxonomic relationships in the four Lamiaceae species ( Table 2). The contents of the selected oil components were significantly different between species  Table 2 Moumni (p < 0.05). The PCA horizontal axis explained 56.28% of the total variance while the vertical axis a further 16.86% (Fig. 4).
The HCA based on the Euclidean distance between groups indicated three species groups (A', B′ and C′) identified by their essential oil chemotypes with a dissimilarity > 10 ( Fig. 5) which was mainly due to the variation along the major axis in PCA analysis. With a dissimilarity > 5, group C′ was divided in two subgroups (C'1 and C'2) in both HCA and PCA analysis, which was affected essentially by the variation along the vertical axis. The T. capitatus EO of Group A' stand out forming separate group in both PCA and HCA analysis and was characterized by the highest content of carvacrol (56.1 ± 3.0%) and a relatively high contents of p-cymene (7.4 ± 0.8%) and ϒ-terpinene (10.8 ± 3.2%). However, it was very poor in 1,8-cineole (2.9 ± 4.6%), α-pinene (0.8 ± 0.0%) and camphene (0.2 ± 0.0%). Group B' correlated positively with axis1 and was represented by O. majorana, which EO was characterized by the highest contents of α-terpineol (27.4 ± 2.8%), ϒ-terpinene (16.3 ± 1.5%), p-cymene (9.7 ± 0.8%), α-terpinene (7.7 ± 1.3%), and by the lowest content of the compounds which were negatively correlated with axis 1, such as αhumulene and camphor (tr) .

Antibacterial activity
The essential oils were tested for their supposable antibacterial activity against five model bacteria. Antibacterial effects are reported as inhibition zones, and "in vitro activity" as MIC and MBC. The results obtained using the disc diffusion method recorded in MH agar were summarized in Table 3. According to the value of the inhibition zone diameter (IZD) expressed in mm, results  Table 2 Moumni et al. BMC Complementary Medicine and Therapies (2020) 20:103   Table 2 Moumni et al. BMC Complementary Medicine and Therapies (2020) 20:103 were appreciated as follows: not sensitive (−) for diameter equal to 8 mm or below; sensitive (+) for diameter between 8 and 14 mm; very sensitive (++) for diameter 14 and 20 mm and extremely sensitive (+++) for diameter equal or larger than 20 mm [34,35].
To determine the antibacterial power of the studied EOs, the MBC/MIC ratio was calculated and interpreted based on the classification of Schwarz et al. (2010). The EO is considered bactericidal when this ratio is lower or equal to 4 whereas it's bacteriostatic when it's greater than 4 [36].
To evaluate the correlations between the antibacterial activities and the essential oils, all the mean values of the inhibition zone diameters were subjected to PCA and HCA analysis.
The PCA horizontal axis explained 73.18% of the total variance while the vertical axis a further 16.43% (Fig. 6). The HCA based on the Euclidean distances between groups indicated 3 groups (A″, B″ and C″) of species identified by their antibacterial inhibition growth with a dissimilarity ⩾10. The first principal axis separated the groups A″ and C″, whereas the second principal axis allowed the separation of the group B″ from the groups C″ in the ACP (Fig. 6) and the HCA (Fig. 7).
Besides, Table 4 summarizes the MIC and MBC values of the tested Lamiaceae EOs obtained by the dilution method in MH broth. T. capitatus showed the most important bactericidal effect against all the strains (MIC range: 0.73-2.94 mg mL − 1 ; MBC range: 0.73-2.94 mg mL − 1 ) while R. officinalis 2 exhibited the weaker antibacterial activity against the tested strains (MIC range: 45.50-91.00 mg mL − 1 ; MBC range: 91.00-182.00 mg mL − 1 ).

Antibacterial activity
The essential oils antibacterial activity showed considerable variation among the different Lamiaceae species and bacterial strains. T. capitatus EO antibacterial activity was even higher than that of the standard antibiotic Gentamicine®. The IZD of T. capitatus ranged from (15.0 ± 1.0 to 30.0 ± 1.0 mm) and MIC ranged from (0.73 to 2.94 mg/mL). In fact, T. capitatus EO was characterized by the most important mean percentage of the monoterpenoid phenol carvacrol (56.1 ± 3.0%), and important contents of hydrocarbon monoterpenes p-cymene (7.4 ± 0.8%) and ϒ-terpinene (10.8 ± 3.2%). Previous study proved that the thyme essential oil activity was mainly due to the presence of phenolic compounds and monoterpene hydrocarbons [46]. It was reported that an antimicrobial action of phenolic compounds was related to the inactivation of certain cellular enzymes. It depends on membrane permeability changes and consequently on the rate of penetration into the cell [39]. Increased membrane permeability is a major factor in the mechanism of antimicrobial action of phenolic compounds, which may disrupt membranes and cause a loss of cellular integrity and eventual cell death. Activity amongst the phenols and alcohols is at least partly due to the hydroxyl group, which has intrinsic antimicrobial activity and contributes to the relatively greater solubility of these components in biological membranes. Furthermore, the ability of components to release or accept protons has been postulated as an important factor in antimicrobial activity [5,47,48].
Another study attributed the activity of carvacrol to its hydroxyl group. In fact, they confirmed the importance of this group in term of activity in the phenolic structure when they compared carvacrol to its methyl ether [49] it was demonstrated in the same study that p-Cymene, the third major element according to percentage, does not show antibacterial efficacy when used alone. Moreover, other studies [50,51] have shown that EOs exhibit stronger antimicrobial activity than that of their major constituents or their mixtures which suggests synergistic effects of the minor components, but also the importance of all components in relation to the biological activity of EOs.
The activity was decreased with O. majorana oil with MIC values ranging from 22.5 to 45.0 mg/mL. This EO contained a very lower content of carvacrol and higher mean percentage of p-cymene and ϒ-terpinene. O. majorana oil was also characterized by the highest amount of monoterpene hydrocarbons, and alcohol monoterpene α-terpineol. The combination of these results permitted as to deduce that the main antibacterial activity against the studied strains might be attributed to the richness of the essential oils in the phenol carvacrol.
It is interesting to note the susceptibility to the T. capitatus EO of P. aeruginosa, which is known to be a very resistant bacterium even to synthetic drugs, with MICs ranging from 0.73 to 2.94 mg/mL. This result confirm those found with EL Jalel et al. (2018) [41].
S. officinalis, R. officinalis 1 and R. officinalis 2 were placed in the same group in both antibacterial ACP and HCA analyses, but within chemical ACP and HCA analysis, they were classified in two different subgroups. S. officinalis was characterized by the most important amount in monoterpene ketones (48,9 ± 4.1%) while R. officinalis were characterized by the highest amounts of monoterpenes oxides (47.2 ± 0.5 and 37.6 ± 0.5%).
These results suggested that the major components from these two chemical classes could be responsible for producing a similar inhibition. Our results were confirmed  [17].
The R. officinalis EOs was noted to be active against all the bacterial strains but in different degrees. In fact, R. officinalis 1 was more potent with MICs values ranging from 11.38 mg/mL to 22.75 mg/mL when R. officinalis 2 MICs ranged from 45.00 mg/mL to 91.00 mg/mL. These results could be explained by the difference in the mean percentages of some components in these EOs. Actually, R. officinalis 1 EO was richer in oxygenated monoterpene active compounds such as 1,8-cineole (47.2 ± 0.5% Vs. 37.6 ± 0.5%) and camphor (13.3 ± 2.9% Vs. 7.1 ± 1.3%). However, R. officinalis 2 , was characterized by higher amounts of hydrogenated monoterpene compounds, like αpinene (19.4 ± 5.4% Vs. 7.0 ± 5.4%) and camphene (6.5 ± 1.6% Vs. 2.7 ± 1.4%). These finding confirm the results of Griffin et al. (1999) and Van de Vel et al. (2017). These authors suggested that the water solubility and the hydrogen bonding capacity are the main factors influencing MIC values of terpenoids against P. aeruginosa, S. aureus, and E. coli [52,53], since 1,8-cineole and camphor are distinguished by a higher water solubility and hydrogen bonding capacity than αpinene and camphene. Thereby, we can confirm that the qualitative and quantitative differences between R. officinalis EOs chemical composition could affect their antibacterial activity [54].
The MBC/MIC ratio ranged from 1 to 4 for all the EOs, indicating a bactericidal activity for these lamiaceae EOs against the tested bacterial strains. Other Moroccan studies showed that R. officinalis had a bacteriostatic effect on S. aureus [44], whereas O. majorana had a bactericidal effect on E. coli and was bacteriostatic on S. aureus and B. subtilis [55]. The variation in the EOs antibacterial activity could be attributed to the variation of the chemical composition, which is depending of some factors including species, the plant origin, climatic conditions, distillation conditions, adaptive metabolism of plant [56,57] as well as to the bacterial strains susceptibility [58].

Conclusion
The PCA and the HCA of the chemical composition of the studied Lamiaceae EOs separated them in four groups and sub-groups with different chemotypes. T. capitatus was the richest species in carvacrol, S. officinalis was the richest species in α-thujone, and R. officinalis 1 was the richest species in 1,8-cineole. The EOs antibacterial activity varied significantly within species and within strains. In general, EOs have shown promising antibacterial activities we noted a pronounced activity of T. capitatus against P. aeruginosa. These results may represent useful and valuable leads for the development of new disinfectant products.