Screening essential oils for their antimicrobial activities against the foodborne pathogenic bacteria Escherichia coli and Staphylococcus aureus.

The application of essential oils as antimicrobials is a current subject of research and a promising approach in terms of natural food preservation. Due to the diversity of EO producing plant genera and the inconsistent use of susceptibility testing methods, information on the antibacterial potency of many EO varieties is fragmentary. This study was performed to assess the minimal inhibitory concentrations (MIC) of 179 EO samples from 86 plant varieties, using a single method approach, excluding emulsifying agents. MICs were acquired in a broth microdilution assay, using a dispersion based approach to incorporate EOs in a concentration range of 6400 to 50 μg/ml. Staphylococcus aureus and Escherichia coli were used as model bacteria. At concentrations below 400 μg/ml S. aureus was inhibited by 30, E. coli by 12 EO varieties. Azadirachta indica (50 μg/ml vs. S. aureus) and Litsea cubeba (50 μg/ml vs. S. aureus, 200 μg/ml vs. E. coli) essential oils were identified as promising new antimicrobial EO candidates with significant antimicrobial activity against the two foodborne pathogenic bacteria.

Introduction

Investigating the antimicrobial activities of plant essential oils (EO) has concerned many scientific studies within the last two decades. Besides a few general screenings [1, 2, 3], most studies were focused on one type of essential oil only, mainly Thymus vulgaris, Origanum vulgare or Cinnamomum species. Significant activities of these and other EOs against certain foodborne pathogenic bacteria such as Escherichia coli, Listeria monocytogenes, and Salmonella typhimurium have been demonstrated [4, 5, 6]. Furthermore, there has been continuative work regarding the identification of single active compounds from well investigated EOs, such as thymol, carvacrol or eugenol and partial elucidation of their cellular mechanism of action [7, 8, 9]. Most work on antimicrobial EOs was inspired by the idea of identifying alternative preservative agents with an overall “green” and “natural” or “bio-based” character for modern food technology applications.

The findings published to date, are still very fragmentary regarding a wide variety of essential oils [10]. Some EOs are still untested or negative results remain unpublished. In addition, it is difficult to reliably compare results from literature data, due to the strong variance of the used antimicrobial susceptibility testing methods. This complicates the selection of the most suitable EO candidates for further antimicrobial research and applications. Additionally, there are some constraints for EOs concerning food application, especially their sensory impact. EO components are considered aroma compounds (e.g. Thymol, Citral, Limonene, α-/β-Pinene, etc.), meaning they generally induce sensory activity [10]. Regarding food application it is therefore recommended to choose an essential oil based on the sensory profile of the targeted food product [11]. EOs with low inhibitory concentrations are considered advantageous to achieve antimicrobial effects in the product without affecting the sensory properties. For this study we defined a critical minimal inhibitory concentration (MIC) of 400 μg/ml in vitro, to select EOs with the most promising suitability for future application trials.

This study was performed to extent the available information on antibacterial activities of plant essential oils, intending to normalize the discussion on the antibacterial activity of plant essential oils with quantitative data for a large number of EOs. The comparability of the results is maximized by the use of a reproducible, quantitative method for MIC determination, combining the advantages of a dispersion approach, as developed by Remmal, Bouchikhi [12] and Friedman, Henika [13], with the reliability of a standardized broth microdilution assay [14]. Emulsifying agents and organic solvents were excluded when possible, as several authors argued that such additives distort susceptibility tests [12,15]. The trials included a wide variety of EOs, available from three different German essential oil retailers and was focused on the two foodborne pathogenic model bacteria: Escherichia coli and Staphylococcus aureus.

Materials and methods

Essential oils

Essential oils were provided by three different German essential oil manufacturers (Neumond GmbH, Raisting, Germany; Frey&Lau GmbH, Henstedt-Ulzburg, Germany; Düllberg Konzentra GmbH, Hamburg, Germany). Overall, essential oils from 86 plant varieties were accessible, whereof 38 equal varieties where available in triplicate and 20 in duplicate, but from the different sources, respectively. Altogether 179 commercial oil samples were tested. The investigated samples did not contain additives or solvents and were confirmed to be natural by the manufacturers. Furthermore, EOs were considered sterile. EO samples were stored in resealable vials at 5 °C in the dark, but were allowed to adjust to room temperature prior to investigation. The samples were sorted in terms of plant name and the respective plant parts they originated from according to the manufacturer information. Geographical origins were excluded as information was mostly not available.

Bacterial strains and growth conditions

The Gram-positive bacterium Staphylococcus aureus DSM 1104 and the Gram-negative bacterium Escherichia coli DSM 1103 were used as test organisms. The strains were obtained from the Leibniz Institute DSMZ German Collection of Microorganisms and Cell Cultures (DSMZ, Braunschweig, Germany) and are both recommended for antimicrobial susceptibility testing. Stock cultures were grown in 100 ml sterile tryptic soy broth (TSB) (Oxoid, Hampshire, United Kingdom) (pH = 7.2) in shaking flasks at 37 °C for 18 h until their early stationary growth phase. Prior to use, cells were washed twice in sterile ¼*Ringer's solution (Oxoid, Hampshire, United Kingdom) at 9000 rpm for 10 min, respectively. Cell count was adjusted to 1.0*108 cfu/ml in ¼*Ringer's solution by turbidity measurement at 620 nm wavelength, using a McFarland 0.5 standard. The cell counts were preliminarily validated by plate counts. Cells were kept in Ringer's solution for no longer than 15 min before assay medium inoculation.

Essential oil incorporation

Essential oil incorporation was optimized for the execution of a broth microdilution assay for antibacterial susceptibility testing [16]. To avoid the interfering influences of organic solvents or emulsifying agents, a dispersion approach, as first described by Remmal, Bouchikhi [17], was chosen. To increase their viscosity, Mueller-Hinton broth (MHB) (Merck-Millipore, Darmstadt, Germany) and deionized water were spiked with 0.15 % agarose (Merck-Millipore, Darmstadt, Germany) prior to sterilization. All media were adjusted to pH 7.0 ± 0.1. EO stock solutions were adjusted to 12.8 mg/ml in glass vials which were sealed and vigorously shaken for 30 s. Oil containing stock dispersions turned slightly white and were stable towards phase separation for up to 48 h. Air bubbles were driven out by slight vortexing.

A small number of EOs was found to be unstable in dispersion, due to complex-formation and clouding. For these EOs dimethyl sulfoxide (DMSO) was used to aid oil incorporation. Final DMSO concentration was 4.4 % (v/v) and had no growth inhibitory effects. Oils incorporated in DMSO are marked separately in the results tables.

Broth microdilution assay

Antibacterial susceptibility testing was performed according the CLSI laboratory standard for broth microdilution assays [16]. Concentrations tested in the assays ranged from 6400 to 50 μg/ml in bisecting dilution steps. Assays were performed using sterile 96 well microplates (transparent, F-bottom) (Greiner bio-one, Frickenhausen, Germany). Each concentration was tested in triplicate and a single blank (broth with corresponding EO-concentration). Three oils were tested in parallel per plate (n = 3). Ultimately, each well contained 100 μl of EO dispersion. Inoculum was prepared by a hundredfold dilution of the adjusted cell suspension in double concentrated (2*) MHB. Each well was spiked with 100 μl inoculated 2*MHB resulting in a final cell count of 1.0*105 cfu per well. Inoculation media were used immediately to avoid growth dependent shifts in the cell count. Growth, sterility and quality controls were analyzed for every culture, but on separate microplates. Chloramphenicol (2.0–0.008 μg/ml) served as control bacteriostatic to assess cell susceptibility for each batch culture (data not shown). Prior to incubation for 24 h at 37 °C, the inoculated plates were shaken orbital for 30 s on a plate mixer (Kisker, Steinfurt, Germany). To avoid loss of volatile essential oil and humidity during incubation each plate was sealed with a sterile, gas-permeable seal (BREATHseal, Greiner bio-one, Frickenhausen, Germany). Turbidity measurement was performed after incubation in a microplate reader (Tecan, Männedorf, Switzerland) at 595 nm. The seals were removed and plates were shaken orbital for 30 s. Each well was measured at nine spots with 5 flashes per spot. Obtained spot-OD-values of each well were averaged as well-OD-values. Values of parallel wells were averaged afterwards and blank-OD-values were subtracted. The minimal inhibitory concentration (MIC) was defined as the lowest concentration tested which did not allow cell growth within 24 h at 37 °C.

Results

The MIC values of 179 essential oils from 86 plant species against E. coli DSM 1103 and S. aureus DSM 1104, determined in a broth microdilution assay, are presented in Table 1. The results show growth inhibitory activities for the majority of the tested EO samples. Inhibition was generally stronger against the Gram-positive bacterium S. aureus than against Gram-negative E. coli. At or below the preliminary defined critical concentration of 400 mg/ml 46 EO samples from 30 plant genera inhibited S. aureus, whereas E. coli was only inhibited by 22 samples from 12 plant genera. EO varieties from different providers rarely revealed identical MIC values. Nonetheless, Azadirachta indica, Backhousia citriodora, Cinnamomum cassia, Cinnamomum verum, Leptospermum scoparium, Litsea cubeba, Nardostachys jatamansi, Origanum vulgare, Pogostemon cablin, Santalum album, Thymus zygis and Vetiveria zizanoides were found to be the most inhibitory EOs against S. aureus with MIC values of 50 μg/ml. Only four EOs, Cinnamomum cassia, Cinnamomum verum, Origanum vulgare and Thymus zygis could exhibit inhibitory effect against E. coli at 50 μg/ml. But certain other EOs from Backhousia citriodora, Cupressus sempevirens, Cymbopogon citratus, Cymbopogon martini, Cymbopogon nardus, Litsea cubeba, Origanum majorana, Origanum vulgare, and Syzygium aromaticum still revealed promising activity with MICs between 100 and 400 μg/ml against E. coli.

Table 1

Minimal inhibitory concentrations (μg/ml) of essential oils from three different manufacturers (a, b, c) against Staphylococcus aureus DSM 1104 and Escherichia coli DSM 1103.

	plant botanical name	oil common name	extracted plant part	MIC (μg/ml)
				S. aureus			E. coli
				a	b	c	a	b	c
1	Abies alba	silver fir	branches	n. I.			n. I.
2	Abies procera	noble fir	branches		n. I.	n. I.		n. I.	n. I.
3	Achillea millefolium	yarrow	herb	6400			n. I.
4	Anethum graveolens	dill	seeds + herb		n. I.			n. I.
5	Angelica archangelica	garden angelica	roots	400	n. I.	n. I.	1600	1600	n. I.
6	Anthemis nobilis	roman chamomile	blossom	200*			n. I.*
7	Artemisia dracunculus	tarragon	leaves		n. I.			n. I.
8	Artemisia pallens	davana	herb	6400		n. I.	n. I.		n. I.
9	Azadirachta indica	neem	seeds	50*			1600*
10	Backhousia citriodora	lemon myrte	herb	50			200
11	Boswellia carterii	olibanum	resin	1600		n. I.	n. I.		n. I.
12	Cananga odorata	ylang-ylang	blossom	n. I.	n. I.	n. I.	n. I.	n. I.	n. I.
13	Canaricum luzonicum	elemi	resin	400		n. I.	6400		n. I.
14	Carum carvi	caraway	seeds	800	n. I.	3200	3200	6400	1600
15	Cedrus atlantica	atlas cedar	wood	n. I.			n. I.
16	Cinnammomum camphora	ravintsara	leaves	3200			6400
17	Cinnamomum camphora	camphor	branches	n. I.		n. I.	6400		6400
18	Cinnamomum camphora Sieb	ho	leaves	1600			800
19	Cinnamomum cassia	chinese cinnamon	branches	50	50	50	200	50	50
20	Cinnamomum verum	true cinnamon	bark	200	50	800	200	50	200
21	Cistus ladaniferus	cistrose	leaves + branches	400			n. I.
22	Citrus aurantifolia	lime	peel	1600	800	3200	6400	800	6400
23	Citrus aurantium amara	bitter orange	peel	1600	n. I.	3200	1600	n. I.	1600
24	Citrus aurantium amara	neroli	blossom	6400	800	3200	3200	800	6400

a – Neumond GmbH.

b – Frey&Lau GmbH.

c – Düllberg Konzentra GmbH.

blank space – EO not available.

n. I. – no inhibition.

* - contains DMSO.

Some EOs did not show any antibacterial activities. Cananga odorata, Cupressus sempervirens, Daucus carotta, Foeniculum vulgare, Juniperus communis, Pimpinella anisum oils were available from each provider, but did not show any activity against S. aureus or E. coli, respectively. Artemisia pallens, Boswellia carterii, Matricaria chamomilla, Pinus mugo, Piper nigrum, Pogostemon cablin, Pinus sylvestris and Vetiveria zizanoides only exhibited inhibitory activity against S. aureus. Growth of E. coli was not affected by these EOs. For Pogostemon cablin and Vetiveria zizanoides these findings were particularly distinct, as the three EO samples from the different providers revealed the same results for E. coli. Exclusive inhibitory potential against E. coli was only shown for samples from Cinnamomum camphora and Citrus sinensis.

Discussion

The number of scientific studies on the antimicrobial activity of plant essential oils has strongly grown over the last three decades. Due to the use of many different microbiological methods for susceptibility testing and different definitions of antimicrobial activity, the comparability of studies on essential oils is often critical. Many studies focus on selected EOs, providing insight into their activity against one or more microorganisms [4, 5, 7, 8,18], but only few publications compress information by testing multiplicities of essential oils with a defined single method [2, 3, 13]. Therefore, this study was conducted to investigate as many EOs as possible with a single quantitative microbiological method, to achieve maximum comparability.

As scientifically reviewed elsewhere, thyme, oregano and certain cinnamon EOs, as well as their major constituents, carvacrol, cinnamaldehyde and thymol, are known to possess outstanding inhibitory potential against gram-positive and gram-negative bacteria [3, 10, 13]. Therefore the MICs of thyme and oregano EOs can be set as the benchmark for other EOs. On this basis, it can be asserted that Azadirachta indica, Backhousia citriodora, Litsea cubeba and Nardostachys jatamansi exhibit comparably high antibacterial activity. These EOs were especially active against S. aureus, fully inhibiting its growth at 50 μg/ml (the lowest concentration tested). Furthermore, the EOs from these plants appear to be scientifically less well investigated in terms of antimicrobial activity, as only a small number of preliminary findings exist [19,20,21,22].

Backhousia citriodora and Litsea cubeba mainly consist of Citral [23,24], a monoterpenoid aroma compound which is generally known to possess strong antimicrobial potential [23,25,26]. Litsea cubeba inhibited S. aureus at 50 or 100 μg/ml and E. coli at 400 or 200 μg/ml, depending on the EO variety tested. In comparison to other essential oils Litsea cubeba EO is produced in high amounts and cheaply available. Further research is necessary to identify its full antimicrobial spectrum and to optimize its potential. As the tested Litsea cubeba EOs tested in this study fell below the defined food application limit of 400 μg/ml they might be considered a promising candidate for food preservative applications, also due to its unique, refreshing aroma [23,27]. The main active compound of L. cubeba EOs is the monoterpene Citral, which has been found to be positive in terms of sensory effects when used as an antimicrobial compound in food products [28].

Nardostachys jatamansi and Azadirachta indica might also become interesting regarding food preservation applications, as both plants are important in traditional Indian medicine and consequently have histories of safe use [29,30]. Nardostachys jatamansi is dominated by different sesquiterpenes which, to our best knowledge, have not been investigated concerning antimicrobial activities yet [30]. The essential oil from Azadirachta indica, commonly known as neem-tree essential oil, mainly consists of the compounds Azadirachtin and Nimbin. The compounds are known to possess antimicrobial activity, but are predominately used as spermicides [29,31].

By comparing the results of EOs from a single plant species, but from different manufacturers, it becomes evident that simple postulations regarding the antibacterial effect of a certain EO cannot be made easily. As shown in Table 1, most EO varieties revealed differing MIC values when purchased from another manufacturer. Possible reasons may be versatile, as chemical composition is affected by various external factors, such as geographic origin, environmental conditions, point of harvest or other processing dependent influences [32,33,34]. These findings are in line with the results from previous works [35,36] and enhance the often requested need for chemical characterizations of antimicrobial EOs to identify the active compounds and their interdependencies [3]. Consequently, EO optimization and standardization regarding antimicrobial activity appears to be inevitable for application.

This study also revealed results which differ greatly from those reported by others. The lacking growth inhibition by Cananga odorata, Cupressus sempervirens, Juniperus communis or Pimpinella anisum may be due to the comparably low concentrations used. Authors, who found these oils to be inhibitory, used way higher concentrations, revealing MICs ranging from 12.5 mg/ml for Cupressus sempervirens EO against E. coli [37] to 40 %v/v for Juniperus communis EO versus S. aureus [38]. In regard of future applications in food systems and the strong sensory impact of essential oils on food these EO varieties appear to be unsuitable. Given the focus of application it is apparent to select essential oils with very low MICs. As described before, we defined a critical MIC level of 400 μg/ml in order to identify EO varieties with a greater applicability concerning food preservation. In general it is recommended to couple food application studies with sensory profiling trials. Another peculiarity in this study is the fact, that none of the Foeniculum vulgare EOs showed any inhibitory activity. In previous studies fennel seed essential oil was found to be bactericidal at comparable concentrations between 20 μl ml and 80 μl/ml against E. coli and S. aureus by Dadalioglu and Evrendilek [39]. In this case clarification can only be achieved by chemical analysis of the respective oils which has not been performed as part of our investigations, due to the more broadened approach. On the other hand it was once more affirmed, that Daucus carotta essential oils completely lack antimicrobial activity, as also stated by Hammer, Carson [3].

Conclusion

In summary, this study provides insight into the in vitro antibacterial activity of a wide variety of essential oils from many different plant genera against E. coli and S. aureus. The data contributes to the ongoing scientific investigation regarding the application of essential oils as natural food preservative agents. As the comparison of MICs from different studies is most often difficult, due to the use of varying quantitative or semi-quantitative methods, this study aimed to normalize the discussion by testing a wide variety of plant essential oils with a single, standardized quantitative method for MIC detection. After benchmarking EOs from thyme and oregano as the most active, EO varieties from Azadirachta indica and Litsea cubeba were identified as promising candidates concerning possible applicability in food.

Declarations

Author contribution statement

Julian Thielmann: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data; Wrote the paper.

Pamina Kazman: Performed the experiments; Analyzed and interpreted the data.

Peter Muranyi: Analyzed and interpreted the data; Wrote the paper.

Funding statement

This work was partially funded by the Federal Ministry of Economics and Technology (BMWi) via the industrial collective research program of the German Federation of Industrial Research Associations (AiF) under contract number IGF 99EN/1.

Competing interest statement

The authors declare no conflict of interest.

Additional information

No additional information is available for this paper.