1. Introduction
Tuberculosis is still a global threat [
1], but the emergence of multi- and extremely-drug resistant
Mycobacterium tuberculosis (Mtb) strains makes the treatment of this disease much more difficult. Since circulating strains are resistant to various combinations of drugs and the development of new drug leads takes a long time, a different approach is to find adjuvants capable of enhancing the efficacy of antibiotics. An enhancement of microbes’ susceptibility to antibiotics was observed under the influence of natural products [
2], especially, essential oils [
3,
4,
5,
6,
7]. Essential oils are known for their positive antimicrobial action and many possible mechanisms of their activity against a wide range of bacterial strains have been proposed [
8,
9,
10,
11]. The studies on synergy between essential oils components [
12] as well as on the synergy between essential oils and antibiotics against wide range of microorganisms [
13] indicates the possible impact of these compounds on restoring the sensitivity of resistant bacteria to antibiotics. Positive, synergistic interactions (FICI ≤ 0.5) were previously described for the antibiotics ciprofloxacin, gentamicin and polymyxin B with
Eucalyptus camaldulensis essential oil against multidrug-resistant
Acinetobacter baumannii isolates [
3], for norfloxacin and
Pelargonium graveolens essential oil against
Bacillus cereus and
Staphylococcus aureus [
14], for ticarcillin, imipenem, gentamicin and tobramycin and
Origanum compactum essential oil against
Pseudomonas putida [
15], to name just a few. However, very little is known about the influence of combinations of antibiotics and essential oils constituents against Mtb. Only
trans-cinnamic acid with amikacin as well as
cis-cinnamic acid with rifampicin were mentioned to have positive action against isolates of multi-drug resistant
M. tuberculosis [
5,
6]. For this reason we selected several natural terpenes (α-pinene, bisabolol, β-elemene, (
R)-limonene, (
S)-limonene, myrcene, sabinene), which are the main constituents of the essential oil obtained from
Mutellina purpurea L. (syn.
Ligusticum mutellina), a plant with described antitubercular activity [
16], to investigate their interactions with antibiotics against different mycobacterial strains. A monoterpene α-pinene is the most frequently found in Nature [
17]. It is a main constituent in essential oils obtained from coniferous trees and shrubs [
18,
19], however it was also described in many other essential oils like
Rosmarinus officinalis,
Satureja montana [
20,
21],
Eucalyptus globulus [
22],
Melaleuca leucadendron, [
23], many other herbs and some citrus fruits essential oils.
Myoporum crassifolium and
Matricaria recutita are the most abundant sources of bisabolol, although it is present in many plants such as lavender or rosemary [
17,
24]. Rhizome of curcuma is used for β-elemene isolation, however this compound can be obtained also from
Michelia figo,
Alisma orientale,
Solidago decurrens or
Nigella damascena essential oils [
25]. Limonene is known from citrus essential oils, where it accounts up to 90%, and is present in both enantiomeric forms ((−)-(
S)-limonene and (+)-(
R)-limonene) in many other essential oils [
17]. Myrcene is other compound very widespread in Nature. It was described in most of the common highly fragrant herbs and spices, among which
Humulus lupulus,
Laurus nobilis,
Thymus vulgaris,
Cymbopogon schoenanthus or
Ocimum basilicum contain high levels of this compound [
17]. Sabinene is frequently present in essential oils. It is a major constituent in
Juniperus scopulorum essential oil [
26], it is also detected is higher amounts in
Myristica fragrans [
27],
Piper nigrum [
28] or
Daucus carota essential oils [
29].
The previous studies showed that α-pinene, myrcene, limonene, bisabolol and
M. purpurea essential oilalter the mycobacterial cell shape and homogeneity [
30], thereby they may enhance the action of antibiotics. Also the positive influence of these terpenes on activity of first line antibiotics against isolated Mtb was observed [
31]. Hence, in this work we aimed to investigateif natural terpenes show any synergistic action with antituberculous agents against multi-drug resistant MTb clinical strains.
3. Discussion
A synergistic action of drug-drug combinations is defined as greater activity than the activity of the sum of the individual components [
12]. This effect is desirable and leading to a better antimicrobial therapy efficacy. On account of emergence of multi- and extensively drug-resistant Mtb strains the combinations of drugs with natural compounds are tested to search for enhancers of antibiotics activity [
35]. Also our study aimed to investigate the possible interactions between terpenes, constituents of essential oils, and antimycobacterial antibiotics against drug-resistant clinical isolates and reference Mtb strains. We also performed cytotoxicity evaluation of tested compounds against eukaryotic cells.
The in-vitro cytotoxicity determination for terpenes (α-bisabolol, (
S)-limonene, (
R)-limonene, α-pinene and sabinene) and literature data (myrcene, β-elemene) [
33,
34] showed that they are non-toxic against eukaryotic cells (green monkey kidney cells: Vero or GMK cell line).
The antimycobacterial activity of tested terpenes was described for a panel of Mtb strains (including clinical isolates) for a first time. However several of these compounds (sabinene, myrcene and limonene) were previously screened against reference virulent strain H37Rv. The MIC values obtained in our study for sabinene, myrcene and limonene were significantly higher than previously described by Andrade-Ochoa et al. [
36] but this can be explained by different method applied for evaluation of inhibitory activity of tested compounds (visual evaluation, 21 day of incubation in our study and Alamar Blue technique, 8 days of incubation in the cited paper). Regardless the method used the activity of these three compounds was on the similar level. Among other terpenes, the interesting activity was shown by thymol and carvacrol (0.78 and 2.02 μg/mL, respectively), while the highest MIC values were obtained for
p-cymene and β-caryophyllene (91 and 100 μg/mL, respectively) [
36].
Taking into account the results of antimycobacterial activity of tested terpenes against different mycobacteria strains, it is difficult to correlate their activity with the structure, because the MIC values varied even by several orders for different strains, and the activity could not be described by any particular pattern. The quantitative structure–activity relationship studies performed by Andrade-Ochoa and coworkers revealed that the number of conjugated carbons, the number of phenolic and hydroxyl groups and the number of acceptor atoms of hydrogen bonds are the most important structural descriptors in the antimycobacterial activity of terpenes [
36]. Compounds tested in this study were simple terpenes thus phenolic and hydroxyl groups nor oxygen atoms did not contribute to their antimycobacterial activity. The lowest MIC values observed for two-compounds combinations were descried for monocyclic sesquiterpene—bisabolol having a long side chain and for monocyclic monoterpene—(
R)-limonene—with methyl and propylene groups. What is more interesting, there were significant differences in MIC values obtained for enantiomeric forms of limonene against mycobacterial strains (with exception of the avirulent strain H37Ra) and for combinations of limonenes with antiobiotics. (
R)-Limonene occurs commonly in many plants but is especially abundant in citrus fruits, it has citrus odor, while the less often found (
S)-limonene is a precursor in the biosynthesis of (−)-menthol and its odor is pine-like [
37]. The distinction between these compounds is a different spatial arrangement of the propylene and methyl groups and it seems that more often naturally occurring isomer has also better antimicrobial properties, probably via specific interactions with mycobacterial cell envelope
The results from antimycobacterial assay showed that enhanced activity of ethambutol and rifampicin against H37Ra strain under the influence of studied terpenes may be correlated to the capability of essential oils constituents to modify the bacteria resistance mechanisms in general. Three possible scenarios can take place: inhibition of multidrug efflux pumps (farnesol in
Mycobacterium smegmatis) [
38], cell wall and membrane disturbance (β-elemene and R-limonene in
M. tuberculosis) [
39] and alterations of quorum-sensing (Rose, geranium, lavender, clove essential oils in
Escherichia coli) [
40]. These may lead to better influx and retention of antibiotics inside the cells resulting their better efficacy. Our previous observations have shown that studied terpenes cause the changes in mycobacterium H37Ra cells shape, cell wall thickness and cytoplasm homogeneity in terms of uniformity and consistency [
30]. Bacteria exposed to terpenes became filamentous (continuing to elongate but not to divide) what usually occurs under the oxidative stress, nutrient limitation or DNA damage and alters DNA replication and cell division [
41]. Also other authors described similar morphological changes in pathogenic and spoilage-forming bacteria cells structures. The loose of regular cell shape and changes in cytoplasm and membrane integrity were observed under the influence of tea tree oil and some terpenes [
40,
42,
43]. These findings suggest that cell wall and membrane disturbances caused by essential oils constituents may mainly contribute to the better penetration of antibiotics into the cells and their enhanced activity when tested in combinations.
However the important observations from this study are differences in avirulent and virulent bacteria susceptibility to terpenes tested individualy and in combinations with antibiotics. A comprehensive description of possible genetic variations, like multiple mutations, between virulent strain H37Rv and avirulent one, was described by Zheng and coworkers [
44]. The changes found in H37Ra may account for its attenuation of virulence and various other phenotypic changes, like changes in bacteria susceptibility to antibiotics [
44].Since synergistic interaction between antibiotics (ethambutol/rifmapicin) and terpenes was described only against avirulent strain H37Ra, the higher susceptibility of this strain to tested compounds may be related to the mutation in phoP gene, which is one of the factors leading to the inability of bacteria growth in human and murine macrophages [
45,
46,
47]. The phoP gene encodes PhoP protein playing a key regulatory function in synthesis or transfer of metyl branched fatty acyl substituents found in polyketide-derived acyltrehaloses (sulfolipids SL, diacyltrehaloses DATs and polyacyltrehaloses PATs) [
48,
49]. The absence of these lipids in the cell wall was described in H37Ra mutant strain [
47,
50] and this may contribute to the better action of terpenes themselves and in combination with antibiotics. The antibacterial activity of terpenes leading to the increased membrane permeability is dependent on the bacterial membrane net surface charge [
9], which in physiological conditions is negative [
9,
51,
52,
53]. However during interaction of bacteria with essential oils constituents it becomes less negative [
9]. In case of H37Ra mutant strain, the lack of sulfolipids in cell wall may be responsible for the enhanced antibacterial activity since sulfation gives an anionic form to a complex lipids of mycobacterial envelope and improves the lipid solubilization [
54]. What is more, the experiments performed by Trombetta et al., suggest a real transfer of terpenes through cell membrane and possibility of their interaction with intracellular elements [
11] but this action is strongly dependent on the lipid composition of the cell envelope [
9,
11,
55,
56]. Hence the observed lack of synergistic action of combinations of terpenes and antibiotics in case of all virulent strains might be a consequence of their richer composition of cell envelope [
11], especially regarding the presence of midchain methyl-branched fatty acids (10-methylhexadecanoic(16:1me(10)), 10-methyloctadecanoic acid (18:1me(10)), and 10-methylnonadecanoic acid (19:1me(10)) which are associated with tolerance mechanisms to xenobiotics and with the maintenance of membrane fluidity under different stress conditions [
57,
58].
The literature data show synergistic interactions among antibiotics: isoniazid, rifampicin, and ethambutol (18.1% of combinations with FICI = 0.6) or ofloxacin, rifampicin, and ethambutol (91.3% of combinations with FICI = 0.31–0.62) against
M. tuberculosis H37Rv [
59] or between antibiotics and natural products: oleic acid (in combination with isoniazid, rifampicin and ethambutol, FICI in a range 0.09–0.36), 7-methyljuglone (in combination with isoniazid, FICI = 0.2 and with rifampicin, FICI = 0.5), usnic acid (with rifampicin, FICI = 0.25–0.38) against
M. tuberculosis H37Rv strain and some Mtb drug-resistant clinical isolates [
2,
60,
61], however little is known about their mechanism of action. Apart from positive results, the lack of synergy against different bacterial species was also documented. In the study performed by Rey-Jurado et al., the indifference (FICI = 1.5–3) was observed for all three-drug combinations of second-choice antituberculous antibiotics tested against multidrug-resistant Mtb isolates, showing all these combinations to be equally effective [
62]. Also the combination of carvacrol, eugenol and cinnamaldehyde encapsulated within lipid nanocapsules and doxycycline tested against
Acinetobacter baumannii SAN,
A. baumannii RCH,
Klebsiella pneumoniae,
Escherichia coli and
Pseudomonas aeruginosa resulted in indifference (FICI = 0.7–1.30) [
63]. Nevertheless, despite the lack of synergy in these species, the scanning electron microscopy images showed holes in bacterial envelope and leakage of cellular contents after exposure to tested combination [
63] indicating their influence on the bacterial cell wall/membrane which probably can be attributed to the action of essential oil constituents. What is more, we also observed that MIC values obtained for rifampicin was lowered at least two-fold in the presence of bisabolol and (
R)-limonene for drug resistant isolates and reference strains although the synergy in action wasn’t observed for these strains. This observation may suggest that disturbances in a cell envelope integrity may facilitate the intake of antibiotic also in virulent strains, however the extent of changes was too low to produce synergy. The other interesting observation from our study and from literature data [
60,
61] showed that rifampicin produce positive interactions with natural compounds more often than INH, EMB or SM. This may be correlated with different target points of rifampicin (lipophilic compound inhibiting DNA-dependent RNA polymerase) [
64] and terpenes (influencing cell wall and membrane), however it was not explained yet. Additional studies are needed for better understanding of interactions taking place in bacteria when antibiotics are combined with essential oils or its constituents.