First Evidence of a Combination of Terpinen-4-ol and α-Terpineol as a Promising Tool against ESKAPE Pathogens

Antimicrobial resistance is a major public health issue raising growing concern in the face of dwindling response options. It is therefore urgent to find new anti-infective molecules enabling us to fight effectively against ever more numerous bacterial infections caused by ever more antibiotic-resistant bacteria. In this quest for new antibacterials, essential oils (or compounds extracted from essential oils) appear to be a promising therapeutic option. In the present work, we investigate the potential antibacterial synergy between a combination of terpinen-4-ol and α-terpineol (10:1) compared to standard tea tree oil. The minimum inhibitory concentration (MIC) and the minimum bactericidal concentration (MBC) were determined. Then, time kill assays, in vitro cytotoxicity and bactericidal activity on latent bacteria (persisters) were investigated. Finally, an in silico study of the pharmacokinetic parameters of α-terpineol was also performed. Altogether, our data demonstrate that the combination of terpinen-4-ol and α-terpineol might be a precious weapon to address ESKAPE pathogens.


Introduction
In 2017, the World Health Organization (WHO) published a list of bacteria divided into three categories according to the urgent need for new antibiotics against these pathogens [1]. Among them, unsurprisingly, were Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacter spp. (i.e., ESKAPE) [2]. New antibiotics are therefore urgently needed to combat ESKAPE pathogens. Along with antibiotics, essential oils could be promising weapons among the different strategies currently being studied to fight those bacteria [3].
Melaleuca alternifolia essential oil (tea tree oil, TTO) and its compounds have been extensively studied for their biological properties, especially against bacteria [4][5][6]. In addition to its beneficial effects, TTO might also be irritating or toxic. We previously described a process for the fractionation of TTO, to separate toxic molecules from beneficial ones [7]. Interestingly, antimicrobial activity was maintained between TTO and the detoxified extract named Titroleane™.
α-terpineol is another monocyclic terpene alcohol, found in essential oils such as Citrus aurantium (aka neroli), Melaleuca alternifolia or Origanum vulgare, widely used in cosmetics, soaps fragrances and pharmaceutical industry [26]. α-terpineol is well documented for its bactericidal activity on both Gram-positive and Gram-negative bacteria [15,[27][28][29][30], in addition to other properties including antioxidant, anti-inflammatory, antiproliferative, and analgesic properties [31]. Remarkably, the molecule is also efficient against antibioticresistant bacteria [32]. Moreover, α-terpineol inhibits biofilm formation by destabilizing the cell membrane [19,[33][34][35]. Investigations on the combination of several terpene alcohols demonstrated that α-terpineol was bactericidal against S. typhimurium, Escherichia coli O157:H7 and Staphylococcus aureus in a synergistic way when associated with linalool, but not when associated with eucalyptol, as well as the association of linalool with eucalyptol [36]. In the same study, the authors also showed by scanning electron microscopy that α-terpineol caused permeability alteration of the outer membrane, alteration of cell membrane function and leakage of intracellular materials.
As mentioned above, we have previously demonstrated that our detoxified extract from tea tree oil (i.e., Titroleane™) maintained the same antibacterial activity as the original product, on a wide range of bacteria, including antibiotic-resistant ones [7]. Moreover, as terpinen-4-ol and α-terpineol are the two main compounds retrieved in Titroleane™, we then decided to focus on the antibacterial properties of a combination of these two molecules.

Antibacterial Activity of α-Terpineol and Terpinen-4-ol, Alone
It was previously shown that a commercial fraction of TTO named Titroleane™ exhibits a similar antimicrobial activity to TTO, although it was depleted from its monoterpenes which were expected to be the most important antimicrobial molecules in TTO [7]. We investigated the role of some major compounds of Titroleane™ to maintain the antibacterial activity.

Synergstic Antibacterial Activity between α-Terpineol and Terpinen-4-ol
A potential synergy between terpinen-4-ol and α-terpineol was then investigated, using the checkerboard method [38]. Combination assays of terpinen-4-ol and α-terpineol were performed three times independently, on the same bacterial strains. Respective FICs were reported on Figure 1.  Table 1 summarizes results of MICs for terpinen-4-ol and α-terpineol in % (v:v), mM and mg/mL. As percentage as a concentration unit is commonly used for essential oils and their compounds, this unit is used in the rest of the article [7,15,17,18,37].

Synergstic Antibacterial Activity between α-Terpineol and Terpinen-4-ol
A potential synergy between terpinen-4-ol and α-terpineol was then investigated, using the checkerboard method [38]. Combination assays of terpinen-4-ol and α-terpineol were performed three times independently, on the same bacterial strains. Respective FICs were reported on Figure 1. FIC and FICI values of α-terpineol and terpinen-4-ol from three independent assays on Staphylococcus aureus, methicillin-resistant Staphylococcus aureus, and Escherichia coli clearly showed a synergistic antibacterial effect of the two molecules for all combinations tested ( Table 2). The same synergy was also observed for two out of three assays with Pseudomonas aeruginosa. FIC and FICI values of α-terpineol and terpinen-4-ol from three independent assays on Staphylococcus aureus, methicillin-resistant Staphylococcus aureus, and Escherichia coli clearly showed a synergistic antibacterial effect of the two molecules for all combinations tested ( Table 2). The same synergy was also observed for two out of three assays with Pseudomonas aeruginosa. Table 2. Bactericidal synergy between α-terpineol and terpinen-4-ol. FIC values of α-terpineol and terpinen-4-ol were obtained from three independent assays on the four bacteria reported in Figure 1. The effect deduced from the calculated FICI are detailed. Antibiotics used as controls are ampicillin for Staphylococcus aureus and MRSA, amoxicillin for Escherichia coli, and gentamycin for Pseudomonas aeruginosa. FIC: fractional inhibitory concentration; FICI: fractional inhibitory concentration index.

Comparison of the Bacteriostactic Activities of the α-Terpineol and Terpinen-4-ol Mixture and Tea Tree Oil
There is no clear ratio emerging for an optimal antibacterial activity on all four strains tested. According to the standard composition of TTO [39], terpinen-4-ol ranges from 35 to 48% of the global composition (m:m), and α-terpineol ranges from 2 to 5%. Hence, the ratio between terpinen-4-ol and α-terpineol is close to 10:1 (m:m). We thus designed a mixture of terpinen-4-ol and α-terpineol named Synterpicine™, with the same ratios of 10:1 (m:m) of terpinen-4-ol and α-terpineol. Interestingly, this 10:1 ratio is also similar to what was observed in the composition of Titroleane™ [7] with terpinen-4-ol ranging from 71 to 75% of the global composition (m:m), and α-terpineol ranging from 5 to 9%.
Antibacterial activity of Synterpicine™ was compared to standard TTO against the same bacterial strains (Table 3).

Bactericidal Activity of Synterpicine™
In addition to MICs, in order to better understand the effects of Synterpicine™ on some bacteria of major interest in health, minimum bactericidal concentrations (MBCs) were studied (Table 5). In our MBC assays, Klebsiella oxytoca and Klebsiella pneumoniae appeared to be the more sensitive bacteria, with MBCs of 0.31-0.62% and 0.62%, respectively. The MBC of Synterpicine™ on Escherichia coli was 0.65-2.50%. Enterobacter cloacae, Pseudomonas aeruginosa and Acinetobacter baumannii had a similar sensitivity to Synterpicine™, with MBCs of 1.25-2.50%. In our experiments, we were not able to define the MBCs of Staphylococcus aureus (MIC = 2.50->2.50%), MRSA (MIC > 2.50%) and Enterococcus faecium (MIC > 2.50%).

Time Kill Assays
In order to evaluate the effects of Synterpicine™ on the growth kinetics of bacteria, time kill assays were performed on Staphylococcus aureus, MRSA, Escherichia coli and Pseudomonas aeruginosa. Bactericidal activity is defined as a drop of more than 99.9% (3 Log10) of viable bacteria compared to the inoculum [40].
At a 4×MIC, Synterpicine™ had bactericidal activity against Staphylococcus aureus and MRSA in less than 1 h (Figure 2b,d). As a control, the antibiotic of reference ampicillin was efficient against Staphylococcus aureus, but only reached the bactericidal limit of a 3 Log10 decrease in 24 h ( Figure 2a). As expected, MRSA could grow in the presence of ampicillin, though with less efficiency than in the absence of ampicillin (Figure 2c). Escherichia coli was sensitive to Synterpicine™, with a decrease of more than 3 Log10 in less than one hour (Figure 2f). With the antibiotic amoxicillin as a control, a decrease of more than 3 Log10 was observed only after 4 h (Figure 2e). Pseudomonas aeruginosa was sensitive to Synterpicine™, with a decrease of more than 3 Log10 in less than one hour (Figure 2h). With the antibiotic gentamicin as a control, a reduction of more than 3 Log10 was achieved only after 2 h (Figure 2g).

Time Kill Assays
In order to evaluate the effects of Synterpicine™ on the growth kinetics of bacteria, time kill assays were performed on Staphylococcus aureus, MRSA, Escherichia coli and Pseudomonas aeruginosa. Bactericidal activity is defined as a drop of more than 99.9% (3 Log10) of viable bacteria compared to the inoculum [40].
At a 4×MIC, Synterpicine™ had bactericidal activity against Staphylococcus aureus and MRSA in less than 1 h (Figure 2b,d). As a control, the antibiotic of reference ampicillin was efficient against Staphylococcus aureus, but only reached the bactericidal limit of a 3 Log10 decrease in 24 h ( Figure 2a). As expected, MRSA could grow in the presence of ampicillin, though with less efficiency than in the absence of ampicillin (Figure 2c). Escherichia coli was sensitive to Synterpicine™, with a decrease of more than 3 Log10 in less than one hour (Figure 2f). With the antibiotic amoxicillin as a control, a decrease of more than 3 Log10 was observed only after 4 h (Figure 2e). Pseudomonas aeruginosa was sensitive to Synterpicine™, with a decrease of more than 3 Log10 in less than one hour (Figure 2h). With the antibiotic gentamicin as a control, a reduction of more than 3 Log10 was achieved only after 2 h (Figure 2g).

Bactericidal activity of Synterpicine™ on Dormant Bacteria
Exponentially growing bacteria and dormant bacteria were compared for their sensitivity to Synterpicine™, using Escherichia coli as a model. DMSO (carrier) and amoxicillin (antibiotic) were used as negative and positive controls, respectively. All experiments were performed three times independently. Results are reported in Table 6.
Growth controls were performed on dormant versus exponentially growing Escherichia coli. After 24 h, exponentially growing bacteria increased by more than 3 Log10 (5.45 ± 0.05 Log10 CFU/mL at the beginning of the experiment versus 8.88 ± 0.05 Log10 CFU/mL after 24 h), while dormant bacteria numeration was almost unchanged (5.32 ± 1.37 Log10 CFU/mL at the beginning of the experiment versus 5.45 ± 0.65 Log10 CFU/mL after 24 h). This confirmed that the design of the experiment made it possible to study exponentially growing bacteria compared to dormant/non-growing bacteria in MBC experiments.

Bactericidal Activity of Synterpicine™ on Dormant Bacteria
Exponentially growing bacteria and dormant bacteria were compared for their sensitivity to Synterpicine™, using Escherichia coli as a model. DMSO (carrier) and amoxicillin (antibiotic) were used as negative and positive controls, respectively. All experiments were performed three times independently. Results are reported in Table 6. Table 6. MBC of Synterpicine™ on latent and exponentially growing Escherichia coli. Concentrations ranged from 0.048 to 25 µg/mL for amoxicillin, and from 0.0049% to 2.50% for Synterpicine™ and DMSO. Growth controls on exponential and latency/non-growing bacteria were performed at 0 (T0) and 24 h (T24). DMSO (carrier) and amoxicillin (antibiotic) were used as negative and positive controls, respectively (n = 3).

Exponential Latency
Bacterial count (Log10 CFU/mL) T0 5 Growth controls were performed on dormant versus exponentially growing Escherichia coli. After 24 h, exponentially growing bacteria increased by more than 3 Log10 (5.45 ± 0.05 Log10 CFU/mL at the beginning of the experiment versus 8.88 ± 0.05 Log10 CFU/mL after 24 h), while dormant bacteria numeration was almost unchanged (5.32 ± 1.37 Log10 CFU/mL at the beginning of the experiment versus 5.45 ± 0.65 Log10 CFU/mL after 24 h). This confirmed that the design of the experiment made it possible to study exponentially growing bacteria compared to dormant/non-growing bacteria in MBC experiments.
Escherichia coli was not sensitive to DMSO in the range of concentrations used (up to 2.50%), whatever its growing state. Exponentially growing bacteria were sensitive to amoxicillin at a MBC of 3.12-12.5 µg/mL, while non growing/latent bacteria were not sensitive to the antibiotic up to 25 µg/mL. Synterpicine™ MBC was 0.31% on exponentially growing Escherichia coli, and 0.62-1.25% on non-growing/dormant bacteria.

Cytotoxicity of Synterpicine™
Synterpicine™ cytotoxicity was evaluated in vitro on human foreskin fibroblasts (HFF) as previously described [7], and compared to TTO containing a similar ratio of terpinen-4-ol and α-terpineol (10:1). All experiments were performed with 0.30% DMSO, and the latter was used as a negative control of cytotoxicity. SDS (1%) was used as a positive control of cytotoxicity. After 24 h, no Synterpicine™ cytotoxicity was observed on HFF at a concentration of 0.025% (Figure 3). In the same conditions, TTO was slightly cytotoxic (i.e., viability = 81%) at a concentration of 0.006%.

In Silico Study of Pharmacokinetic Parameters
In silico analysis of the pharmacokinetic parameters of terpinen-4-ol has already been conducted [18], using the online Program SwissADME [41]. The same online Program was

In Silico Study of Pharmacokinetic Parameters
In silico analysis of the pharmacokinetic parameters of terpinen-4-ol has already been conducted [18], using the online Program SwissADME [41]. The same online Program was used to evaluate the theoretical pharmacokinetic parameters of α-terpineol (Table 7). The molecular mass of α-terpineol is 154.25 g/mol, and is in accordance with Lipinski's rule. However, this value is well below the optimal range for Ghose (180 ≤ MW ≤ 480), so the molecule violates this parameter.
Veber presents two essential parameters for molecules to be administrated orally: TPSA (Topological Polar Surface Area) and number of rotatable connections. α-terpineol has a TPSA of 20.23 Å 2 , and one rotatable connection. Therefore, the molecule has a high prospect of being used orally. α-terpineol has one hydrogen acceptor and one hydrogen donor, and is in accordance with Lipinski's rule and has an excellent theoretical oral bioavailability. The Egan method and bioavailability score also support a good theoretical oral bioavailability.
The partition coefficient between n-octanol and water (Log Po/w) is commonly used to evaluate lipophilicity and the ability to cross plasma membranes. The Log Po/w consensus is 2.58 for α-terpineol. It meets the rules according to Lipinski (Log Po ≤ 5), Ghose (Log Po/w ≤ 5.6) and Egan (Log Po ≤ 5.6). It slightly differs from terpinen-4-ol consensus (2.60). This difference is due to one of the five predictive models, i.e., XLOGP3, an atomistic method based on the fragmental system of Wildman and Crippen [42].
The Log S-determined by the Ali method, ESOL model and SILICOS-IT-reflects the water solubility, which is key for the absorption and distribution of a molecule in the body. Log S for α-terpineol with these methods is, respectively, −2.78, −3.36 and −1.91, indicating that α-terpineol is soluble in water.
The SwissADME program also estimates other pharmacokinetic data. According to the online Program, gastrointestinal (GI) absorption is high, and the molecule can cross the blood-brain barrier (BBB). The molecule is not a P-glycoprotein (P-gp) substrate, and is not an inhibitor of Cytochrome P450 (CYP1A2, CYP2C19, CYP2C9, CYP2D6, CYP3A4).

Discussion
In a previous work, we physically separated TTO into two fractions, the first one containing monoterpenes, and a second one named Titroleane™ containing sesquiterpenes and terpene alcohols. The latter fraction is composed of more than twenty molecules, the most abundant ones having been already described [7]. In the present study, we have focused on major monoterpene alcohols identified in Titroleane™ -, i.e., terpinen-4-ol and α-terpineol, which are well used in the cosmetic and pharmaceutical industries [8,26].
Combining essential oils appears to be a promising strategy for antimicrobial treatments [46]. The combination of essential oils or their compounds can lead to synergistic, antagonisms, or additive antimicrobial effects (reviewed in [47,48]). Extreme antagonism between essential oils or their compounds has been described [49,50], highlighting the need to carefully look at potential interactions when combining them. Synergies have been described for TTO, against Streptococcus mutans when mixed with Citrus limon and Piper nigrum essential oils [51], Streptococcus agalactiae when combined with Lavandula officinalis, and Staphylococcus aureus when associated to Origanum vulgare [37]. Functional interactions between some TTO compounds have also been studied by TS Yang et al. [52]. These authors investigated the potential synergies/antagonisms between linalool, terpinen-4-ol, α-terpineol and p-cymene, the major active compounds of Glossogyne tenuifolia, also present in TTO in significant amounts. Combinations were tested on E. coli O157:H7, Salmonella enterica, Vibrio parahaemolyticus, Staphylococcus aureus, Listeria monocytogenes, Streptococcus mutans, and Streptococcus sanguinis, using the checkerboard analysis. Linalool was tested in combination with terpinen-4-ol and α-terpineol, and p-cymene with linalool, terpinen-4-ol and α-terpineol. No synergy or antagonism was observed in these experiments, but in-triguingly, the terpinen-4-ol/α-terpineol combination has not been tested. One explanation should be that terpinen-4-ol and α-terpineol are two isomers of terpineol, and the authors considered these molecules possessed similar biological activities and might only present additive effects. Nevertheless, for the first time, the present work clearly establishes the synergistic antibacterial activity of terpinen-4-ol and α-terpineol combination against six major pathogenic bacteria responsible for thousand infections every year. Such a synergy suggests that in addition to common mode of action of the two molecules increased membrane permeability and enhanced cell wall and cell membrane integrity [19], additional specific process (ses) might occur. This also indicates that the doses of these two molecules might be decreased when associated. The consequence reduced side effects such as toxicity, allergenicity, possible resistance adaptation, or even fragrance while maintaining the antibacterial properties.
In 2017, the World Health Organization (WHO) published a list of bacteria with three categories depending on the urgency to develop new antibiotics against these pathogens (the most critical ones being Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacter spp. (acronym: ESKAPE) [54]. MICs results of Synterpicine™ on ESKAPE pathogens an MRSA clearly highlight the potential benefits of the mixture to prevent or even treat major infections.
Bactericidal activity is defined as a decrease of more than 1000 times (3 Log10) of viable bacteria compared to the inoculum [40]. Time kill assays confirm the bactericidal properties of Synterpicine™ on Staphylococcus aureus, Escherichia coli, MRSA and Pseudomonas aeruginosa in less than 1 h at 4×MICs, which is faster than ampicillin on Staphylococcus aureus, amoxicillin on Escherichia coli and gentamicin on Pseudomonas aeruginosa in standard conditions. These results are in accordance with previously published results with terpinen-4-ol [16,56] and α-terpineol [35] alone. Time kill assays thus support standard MBC assays, and show that Synterpicine™ is an efficient bactericide against Staphylococcus aureus, MRSA, Escherichia coli and Pseudomonas aeruginosa, making this mixture of terpinen-4-ol and α-terpineol a good antibacterial natural candidate.
Some bacteria can adapt to a dormant form under stresses such as nutrient starvation, antibiotic exposure, acid, or oxidative stress [57,58]. This subpopulation of dormant cells, also called persisters, can adopt a metabolic inactivity and become stress-tolerant (review in [59]). In addition, it has been suggested that chronic infections were probably caused by revivified persister bacteria [60,61], thus highlighting the importance of developing antibacterial molecules efficient against these dormant cells. Consequently, the observed bactericidal effect of Synterpicine™ on Escherichia coli persisters makes the mixture a promising candidate to fight chronic infections. Further experiments are needed to confirm the antipersister activity on other bacterial species, and extend applications of Synterpicine™.
The cytotoxicity of terpinen-4-ol and α-terpineol has been investigated for a long time, mostly on immortalized cell lines such as breast, lung, pancreatic, prostate, colorectal, gastric and leukemia cell lines [62,63]. Interestingly, it appeared that cancer cells were more sensitive to terpinen-4-ol than primary cells [64,65], as was observed with TTO [66]. Thus, terpinen-4-ol and α-terpineol might be good candidates as anticancer drugs [62,63,67,68]. TTO as well as its components terpinen-4-ol and α-terpineol were evaluated as Generally Recognized As Safe (GRAS) [69]. Though, α-terpineol has been described to exhibit antiproliferative effects when associated to other molecules [70]. We thus performed cytotoxicity assays on human primary cells (HFF) to ensure that Synterpicine™ was kept under TTO cytotoxic levels. Our experiments clearly show that Synterpicine™ is less cytotoxic than TTO. Interestingly, measured survival of the cells was similar with 0.05% Synterpicine™ (83%) and 0.006% TTO (81%), suggesting that Synterpicine™ is ten times less cytotoxic than TTO.
Terpinen-4-ol and α-terpineol are two isomers sharing the same chemical formula C 10 H 18 O. Thus, they are supposed to share very close pharmacokinetic parameters. In silico analysis of the pharmacokinetic parameters of terpinen-4-ol has already been performed [18], using the online Program SwissADME [41], and the present results obtained for α-terpineol confirmed this hypothesis. Results are similar for both isomers, only differing in one of the five methods, i.e., XLOGP3, which is an atomistic method based on the fragmental system of Wildman and Crippen [42] to evaluate lipophilicity and ability to cross plasma membranes.
The partition coefficient between n-octanol and water (Log Po/w) is commonly used to evaluate lipophilicity and ability to cross plasma membranes. The Log Po/w consensus is 2.58 for α-terpineol. It meets the rules according to Lipinski (Log Po ≤ 5), Ghose (Log Po/w ≤ 5.6) and Egan (Log Po ≤ 5.6). Interestingly, it slightly differs from the terpinen-4-ol consensus (2.60). This difference is due to one of the five predictive models, i.e., XLOGP3, an atomistic method based on the fragmental system of Wildman and Crippen [42]. In summary, α-terpineol and terpinen-4-ol share very close in silico pharmacokinetic parameters. The pharmacokinetic profile suggests good theoretical oral bioavailability, adequate absorption and distribution of α-terpineol in vivo. In silico pharmacokinetic results should be confirmed in vivo. Anyway, results from in silico analysis support the potential of α-terpineol as a drug candidate, alone or in association with terpinen-4-ol.

Samples Preparation
To enhance the solubility of TTO, terpinen-4-ol and α-terpineol, dimethyl sulfoxide (DMSO) (Sigma Aldrich, Saint-Quentin-Fallavier, France) was used as a solvent. Oils were diluted at 10% final volume (v:v) in 10% DMSO. All experiments were performed simultaneously with a control solution of 10% DMSO.

Growth Conditions and Inoculum Preparation
All strains were cultivated on trypticase soy agar and broth (236920 and 211825, BD, France) at 37 • C, except Acinetobacter baumannii, which was grown at 22 • C, as previously described [7]. The bacteria were grown on the plate and then a single colony was transferred to a second plate. The inoculum was prepared from isolates of the second agar plate in TPS solution (Tryptone 0.1% (211705, BD, Le Pont-de-Claix, France) and sodium chloride 0.85% (S9888, Sigma Aldrich, Saint-Quentin-Fallavier, France)), and was then diluted in adequate broth at 5 × 10 5 -1 × 10 6 CFU/mL.

The Minimum Inhibitory Concentration (MIC) and the Minimum Bactericidal Concentration (MBC) Determination
MIC values were measured using the microdilution broth method, according to the Clinical and Laboratory Standards Institute guidelines [43], with modification on the broths used to fit with organism requirements for growth [7]. Briefly, 96-well microplates were prepared with three controls-growth control without product, negative control without bacteria, and solvent control (DMSO from 2.50% to 0.0049%). Ampicillin (10 µg/mL), amoxicillin (25 µg/mL) and gentamycin (25 µg/mL) were used as positive controls. Wells were prepared by serial dilutions from 2.50% to 0.0049% with 50 µL of product and 50 µL of suspension; microplates were then incubated in growth conditions defined above, then MICs were read visually. A volume of 50 µL of each well with no visible growth was plated on Tryptone Soja agar plates, then incubated for 24 h before counting. The MBC was defined as the lowest concentration without growth after 24 h. A twofold variation between two MICs, corresponding to one dilution, was considered as non-significant. For persister studies, the same protocol was used except that bacteria were inoculated in PBS 3 h before the experiment to slow the growth and reach stationary growth due to the lack of nutrients. Counting was performed 3 h after inoculation (inoculum) and 24 h post-assay (MBC).

Latent Bacteria (Dormant, Persister Cells)
Latent bacteria were obtained by a successive combination of two previously described protocols [71,72]. Mueller-Hinton broth was depleted from its essential nutrients by successive cultures of Escherichia coli for 48 h. After each cycle, the culture was centrifuged at 7500× g. The supernatant was filtered on a 0.22 µm filter, then re-inoculated, until no growth could be observed. The obtained depleted MH broth was then inoculated with an inoculum of Escherichia coli at 5 × 10 5 -1 × 10 6 CFU/mL to determine the standard MBC, as described above.

Synergy
Synergy evaluation between terpinen-4-ol and α-terpineol was performed using the checkerboard method as previously described [38] using the standard MIC protocol described above, and serial dilutions of order 2. The assays were performed on Staphylococcus aureus, MRSA, Escherichia coli and Pseudomonas aeruginosa. MH broth was used as negative control. Inoculum alone was used as positive control. In order to assess the interactions between two compounds A and B, fractional inhibitory concentrations (FICs) and the fractional inhibitory concentration index (FIC) are calculated. The FICs of each of the A and B compounds are calculated as follows: is the MIC of the combination of A and B, MIC A is the MIC of compound A, and MIC B is the MIC of compound B. The fractional inhibitory concentration index (FICI) is the sum of FIC A and FIC B. A FICI of ≤0.5 indicates synergism, between 0.5 and 1 indicates additive effects, between 1 and 2 is considered as indifference, and ≥2 is antagonism [73].

Time Kill
Time kill experiments were adapted from [74,75]. Compounds were tested at concentrations corresponding to the MIC, the 2×MIC and the 4×MIC. DMSO was tested in parallel, at the same concentrations than the terpinen-4-ol and α-terpineol mix. Antibiotic concentrations for MRSA were identical to S. aureus. Bactericidal experiments were performed in round-bottomed 96-well plates. Bacteria were inoculated at 5 × 10 5 CFU/mL in 100 µL final volume of MH broth. Each test was performed in triplicates, then pooled for counting on TS agar broth.

In Vitro Cytotoxicity
Human foreskin fibroblast cell lines were purchased from ATCC collection SCRC-1041. Cells were grown in MEM (10370-047, Gibco, Paisley, UK) with the addition of 10% heatinactivated fetal bovine serum (10270, Gibco, Paisley, UK), 2 mM of L-glutamine (P04-80100, Pan biotech, Aidenbach, Germany), 100 µg/mL ampicillin (A9518, Sigma-Aldrich, Madrid, Spain), 0.10 mU/mL of penicillin-streptomycin (P4333, Sigma-Aldrich, Burlington, MA, USA) and 2.50 µg/mL of amphotericin B (P06-01050, PAN biotech, Aidenbach, Germany). Cells were maintained at 37 • C in a 5% CO 2 humidified atmosphere. The cytotoxicity was carried out by Neutral Red coloration assay for investigating cell viability [76]. Cells were seeded on 96-well plates, at a concentration of 10 5 -10 6 cells per mL in completed MEM, and incubated for 48 h before the addition of the sample to be tested. Various concentrations of samples, all prepared in 0.3% DMSO (final concentration in wells v:v) and sonicated at room temperature for 10 min at 45 kHz (Transsonic TIH5, ELMA, Düsseldorf, Germany) before use, were then added to wells with a final volume of 200 µL. Negative (no treatment), solvent (DMSO 0.30%), and positive (0.10% SDS, BP166, Fisher) controls were included. Each condition was tested three times per assay. Exposure periods of 24 h were chosen for determining and comparing the in vitro cytotoxicity potential of samples. After incubation, the supernatant was removed and cells were washed three times with PBS (10010-015, Gibco, Paisley, UK) before adding Neutral Red (229810250, ACROS organics, Mumbai, India) and solution prepared in completed MEM at 0.005% (v:v). The plate was then incubated for an additional 3 h. The Neutral Red solution was washed three times in PBS, the coloration was solubilized in acetic acid 1%:ethanol 50% (v:v) and the absorbance was measured at 540 nm using a microplate reader (Infinite M200 pro, Tecan, Crailsheim, Germany). The cell survival rate expressed in percentage was determined by comparing the absorbance values obtained with treated cells and with DMSO.

Statistical Analysis
The data were analyzed using a one-way ANOVA Dunnet statistical analysis, which allows for the comparison of data to a control-here the solvent control [77]. Statistical analysis was carried out with the free online SAS version (SAS Studio version 9.4), with the use of Oracle VM VirtualBox version 5.1, and VMware Player version 5.0 (SAS University Edition).

In Silico Pharmacokinetic Parameters
In silico analysis of the pharmacokinetic parameters of α-terpineol was performed using the free online SwissADME Program (www.swissadme.ch, (accessed on 5 August 2020) Swiss Institute of Bioinformatics ® , Lausanne, Switzerland). This program evaluates lipophilicity, water solubility, and druglikeness. It makes it possible to predict whether a drug candidate violates the rules of Lipinski [78], Ghose [79], Veber [80] and Egan [81]. Lipophilicity is classically estimated by evaluating the partition coefficient between noctanol and water (Log p/o). Five predictive models are used in SwissADME: XLOGP3 [82], WLOGP [42], MLOGP [83,84], SILICOS-IT and iLOGP. Water solubility is estimated using the two topological methods derived from the ESOL model [85] and from Ali et al. [86]. A third predictor for solubility developed by SILICOS-IT is also used by SwissADME. The skin permeability coefficient (Kp) is a multiple linear regression model adapted from Potts and Guy [87] which aims at predicting skin permeation, based on molecular size and lipophilicity. The BOILED-Egg model allows the prediction of blood-brain barrier (BBB) permeation and passive gastrointestinal (GI) absorption [88]. The machine learning algorithms developed for SwissADME also allow to evaluate if a molecule can be a substrate for permeability glycoprotein (P-gp), and the five major isoforms of Cytochrome P450 (CYP1A2, CYPC19, CYP2C9, CYP2D6, and CYP3A4) [41]. The results are binary: either Yes (substrate) or No (not a substrate). Drug-likeness is evaluated according to the Lipinski [78], Ghose [79], Veber [80], and Egan [81] methods, and the Abbot bioavailability score [89].

Conclusions
We demonstrated that terpinen-4-ol and α-terpineol act in synergy to kill bacterial pathogens. Particularly, the 10:1 ratio between terpinen-4-ol and α-terpineol-that we named Synterpicine™-has bactericidal effects on ESKAPE pathogens, and on the dormant bacteria. In vitro, Synterpicine™ is also less cytotoxic than standard TTO, a GRAS plant extract. In silico pharmacokinetic parameters suggest that Synterpicine™ is water soluble, can be used orally, gastrointestinal absorption is high, can cross the plasma membranes and the blood-brain barrier.
Taken together, the in vitro and in silico data make the combination of terpinen-4ol and α-terpineol a very promising candidate to help us to fight bacterial pathogens, including ESKAPE ones. Confirmation will come from in vivo experiments.

Patents
Patent application was submitted (application FR2012343).