Chemical Composition and In Vitro Antimicrobial Efficacy of Sixteen Essential Oils against Escherichia coli and Aspergillus fumigatus Isolated from Poultry

Escherichia coli and Aspergillus fumigatus are two pathogens largely present among poultry. They can cause mild or severe forms of disease, and are associated with significant economic losses. The aim of the present study was to investigate the chemical composition and the in vitro antimicrobial activity of sixteen essential oils (EOs) and five mixtures against E. coli and A. fumigatus strains previously isolated from poultry. The study was performed with the following EOs: Aloysia tryphilla, Boswellia sacra, Cinnamomum zeylanicum, Citrus aurantium, Citrus bergamia, Citrus limon, Citrus reticulata, Cymbopogon citratus, Eucalyptus globulus, Lavandula hybrida, Litsea cubeba, Ocimum basilicum, Melaleuca alternifolia, Mentha piperita, Pelargonium graveolens, and Syzygium aromaticum. Moreover, the following mixtures were also tested: L. cubeba and C. citratus (M1), L. cubeba and A. triphylla (M2), A. triphylla and C. citratus (M3), A. triphylla, C. citratus and L. cubeba (M4), S. aromaticum and C. zeylanicum (M5). One hundred and ninety-one compounds were identified in the tested EOs and mixtures. MIC determination found good anti-E. coli activity with C. zeylanicum (2.52 mg/mL), C. citratus (1.118 mg/mL), L. cubeba (1.106 mg/mL), M. piperita (1.14 mg/mL) and S. aromaticum (1.318 mg/mL) EOs. Among the mixtures, M5 showed the best result with a MIC value of 2.578 mg/mL. The best antimycotic activity was showed by A. triphylla (0.855 mg/mL), followed by C. citratus (0.895 mg/mL), while C. aurantium, M. piperita, B. sacra and P. graveolens did not yield any antifungal effect at the highest dilution. The mixtures exhibited no antifungal activity at all. This study shows promising results in order to use EOs in the environment for disinfection purposes in poultry farms and/or in hatcheries.


Introduction
Escherichia coli and Aspergillus fumigatus are two pathogen agents which are largely present in poultry flocks and can cause able to cause mild or severe forms of disease associated with economic losses. Improper hygiene conditions and animal immune system deficiencies would influence the severity of the infection outcome. Environmental stressors such as poor ventilation, warm temperature, excessive ammonia and moisture, degraded litters, long-term storage of feed may increase the concentration of fungal spores and bacteria in the farm environment [1][2][3].

Essential Oils
The antimicrobial activity of the following sixteen EOs, kindly provided by the producer (FLORA ® , Pisa, Italy), was investigated: lemon verbena (Aloysia tryphilla (L'Hèr.) Britton), incense (Boswellia sacra After preliminary results about the efficacy of each EO, mixtures have been made with the most active ones. Mixtures were prepared with 1:1 proportion (w/w) of L. cubeba and C. citratus (M1), L. cubeba and A. triphylla (M2), A. triphylla and C. citratus (M3), as well as for S. aromaticum and C. zeylanicum (M5) together with a mixture of three EOs (M4) obtained with A. triphylla, C. citratus and L. cubeba (1:1:1). All EOs and mixtures were maintained at 4 • C in dark glass vials until their use. Then they were microbiologically analyzed for quality control before antibacterial and antimycotic activity tests. At this purpose, a loopful of each EO was streaked onto a blood agar plate and the plates were incubated at 37 • C for 48 h.
Fractional Inhibitory Concentration Index (FICI) was calculated as reported by Doern [10] to evaluate the possible synergistic effect in the mixtures. The FICI was interpreted as: a synergistic effect when ≤0.5; an additive effect when >0.5-1; indifferent effect when 1-4 and an antagonistic effect when >4.

Essential Oils Analysis
All the selected EOs and mixtures were analyzed using Gas Cromatography-Mass Spectrometry (GC-MS) according to the method cited in a previous paper, as well as for the identification of the different compounds [11].

Statistical Analysis
Regarding the huge variety of the selected EOs, obtained from plants belonging to different orders, families and species, a statistical analysis approach was applied to evaluate the results of their composition. Hierarchical Cluster Analysis (HCA) was performed using Ward's method and squared Euclidian distances for the measure of similarity. Moreover, the Principal Component Analysis (PCA) was done by 'Past 3 software package' version 3.15.

Bacterial Strain
An Escherichia coli strain, previously isolated in a case of poultry colibacillosis, was employed in the study. The isolate was typed using the API20E System (BioMérieux, Marcy l'Etoile, France) and stored in glycerol broth at −80 • C until used.

Agar Disc Diffusion Method
Antibacterial activity of each EO and mixture was tested by Kirby-Bauer agar disc diffusion method following the procedures previously described [12]. A 1:10 dilution in dimethyl sulfoxide (DMSO, Oxoid Ltd., Basingstoke, Hampshire, UK) of each EO and mixture was assayed. All tests were performed in triplicate.
The in vitro sensitivity of E. coli strain to amoxycillin-clavulanic acid (30 µg) (Oxoid) was evaluated by Kirby-Bauer method and the results were interpreted as indicated by Clinical and Laboratory Standards Institute [13].

Minimum Inhibitory Concentration
Minimum inhibitory concentration (MIC) was determined for all EOs and mixtures with the broth microdilution method, starting from a dilution of 10% (v/v) and following the guidelines of CLSI [14] and a protocol previously reported [12]. All tests were performed in triplicate. The MIC value was determined as the lowest concentration, expressed in mg/mL, of each EO and mixture at which bacteria show no visible growth.

Fungal Strain
An avian clinical isolate of A. fumigatus was employed for testing. The mold was maintained onto malt extract agar (MEA), and identification was accomplished on the basis of macroscopic and microscopic features on both MEA and Czapeck agar, following the keys provided by Raper and Fennel [15].

Minimum Inhibitory Concentrations
MICs were determined by a microdilution test carried out as reported elsewhere [12], starting from a dilution of 5% (v/v), following the methods described by CLSI [16] for molds. All tests were performed in triplicate and positive controls using a conventional antimycotic drug (voriconazole) were also performed by microdilution test.

Essential Oil and Mixture Composition
One hundred and ninety-one compounds were detected in the tested EOs and mixtures (Table 1). Table 2 shows the percentages of the main class of constituents in each EO tested. HCA showed two main groups (A and B), each of them also divided in two subgroups ( Figure 1). The first subgroup (A1) included the following species: A. triphylla, M. alternifolia, C. bergamia and mixtures M2, M3 and M4. It is important to underline that A. triphylla EO was one of EOs present in four mixtures. O. basilicum, L. cubeba, P. graveolens, E. globulus, C. citratus, L. hybrida, M. piperita together with M1 mixture were inserted in subgroup (A2). M1 mixture consisted of two EOs categorized in the same subgroup (L. cubeba and C. citratus).
The EOs obtained from the remaining botanical species and mixture M5 belonged to the second group (B), which was divided in two subgroups: (B1) with S. aromaticum, C. zeylanicum and mixture M5; while the subgroup (B2) gathered four out of the five EOs belonging to the Sapindales Order.
HCA analysis gave only limited information; therefore, a further statistical analysis was important to better understand this partition. PCA plot, where the two first axis explain for more than 98.3% of variability, grouped the EOs species among 3 main quadrants: two lower quadrants characterised by high amount of monoterpene hydrocarbons (MH) on the left and by an important percentage of oxygenated monoterpenes (OM) on the right. The upper left quadrant included two EOs (clove and cinnamon) and their mixture M5, in which the amount of phenylpropanoids (PP) was more than 60% ( Figure 2).
In detail, the EOs present in the lower left quadrant all belonged to Sapindales order. Among these species, C. reticulata, C. aurantium, C. limon and C. bergamia belonged to the same Rutaceae family. The first three species showed the highest percentage of MH (99.7%, 97.4% and 94.3%, respectively) while C. bergamia evidenced an equivalent amount between MH and OM (49.0% and 48.5%, respectively). Bergamot was therefore positioned in the middle between EOs characterised by an important percentage of MH and OM. B. sacra, which belongs to the Burseraceae family in the Sapindales order, showed a good percentage of MH (84.3%) and was near the other Citrus spp. Regarding Lamiales order, only L. hybrida and M. piperita were present in the lower right quadrant of PCA, due to the high amount of OM (85.0% and 87.8%, respectively), together with E. globulus, C. citratus and P. graveolens characterized by a similar high percentage of the same constituents (OM: 90.5, 86.3 and 83.4%, respectively). The other two EOs from species belonging to Lamiaceae, showed some distance from each other: in fact, O. basilicum presented a lower percentage of OM (56.1%) but also the highest amount of sesquiterpene hydrocarbons (SH) (20.0%) and was positioned in the upper right quadrant, while lemon verbena, which belongs to the Verbenaceae family (Lamiales order), was on the opposite side due to its relevant amount of MH (66.0%).
L. cubeba from Lauraceae family showed high percentage of OM (75.7%) and a good amount of MH (21.3%). This aromatic profile differed from the others in the same quadrant.         The position of the mixtures in PCA analysis followed the EOs that took part in their composition. In fact, M1 was located in the right lower quadrant as well as L. cubeba and C. citratus with a content of 80.0% of MH. Moreover, M2 was found in the middle between L. cubeba and A. triphylla, while M3 was positioned between A. triphylla and C. citratus. M4, which was obtained mixing three EOs (A. triphylla, C. citratus and L. cubeba, 1:1:1) was shifted on the right quadrant due to the high percentage of OM (56.0%).

Antibacterial Activity
Agar disc diffusion method revealed growth inhibition zones with the following EOs: C. zeylanicum, C. citratus, L. cubeba, M. piperita, O. basilicum, P. graveolens and S. aromaticum. Antibacterial activity was revealed also testing the five mixtures. No inhibition zone was observed with the remaining oils and the negative control. E. coli tested against amoxycillin-clavulanic acid resulted sensitive with an inhibition zone of 20 mm and a MIC value of 0.008/0.004 mg/mL. MIC determination found good anti-E. coli activity with C. zeylanicum (2.52 mg/mL), C. citratus (1.118 mg/mL), L. cubeba (1.106 mg/mL), M. piperita (1.14 mg/mL) and S. aromaticum (1.318 mg/mL) EOs, whereas O. basilicum and P. graveolens resulted effective to E. coli, but with higher MIC values (9.15 mg/mL and 17.8 mg/mL) ( Table 3). Among the tested mixtures, M5 showed the best result with a MIC value of 2.578 mg/mL. FICI calculation highlighted an indifferent effect between the two EOs present in M5. The remaining mixtures did not show good antibacterial activity, as also established by FICI that revealed antagonistic effect between the mixtures components.

Antimycotic Activity
The selected EOs showed different patterns of efficacy. A. triphylla appeared to be the most effective (0.855 mg/mL) followed by C. citratus (0.895 mg/mL), while C. aurantium, M. piperita, B. sacra and P. graveolens did not yield any antifungal effect at the highest dilution. The fungal isolate resulted sensitive to 1 mg/L of voriconazole. The mixtures exhibited no antifungal activity at all, when tested undiluted, as indicated by FICI that showed antagonistic effects among the mixtures components.

Discussion
The present study takes into account the efficacy of a conspicuous number of EOs against two phylogenetically distant organisms, both involved in impairing poultry health and breeding hygiene. Wild type bacterial and fungal strains were selected to better resemble field conditions.
Our results showed that selected EOs exhibited different antimicrobial activity against the tested pathogen agents.
Cinnamon and clove EOs evidenced good antibacterial activity, when used alone or in combination (M5). This action may be related to the main compounds present in these EOs, eugenol (77.9%) and its acetate form (12.2%), as suggested in other studies, too [17].
Zhang et al. [18] observed that cinnamon EO induces damage on permeability and integrity of membrane with consequent loss of inner cell materials. Similar effect against E. coli was found with clove EO by Rhayour and coworkers [19] who observed a damage as holes in both cell wall and membrane.
L. cubeba EO has a relevant anti-E. coli activity, as previously observed by Li et al. [20], who documented holes and gaps on outer and inner membranes of E. coli cells treated with this EO, mainly attributed to the presence of aldehydes as geranial (36.4%) and neral (32.5%).
During this investigation M. piperita EO showed high activity against E. coli. These results are in agreement with Goudjil et al. [21], who tested M. piperita EO against some Gram positive and Gram negative bacteria and found the highest antimicrobial activity versus E. coli. Other authors verified the antimicrobial properties of peppermint EO and attributed the activity to the major components menthol and its oxidative compound menthone [22,23].
C. citratus revealed a good activity against E. coli, as also observed by other researchers who related this activity to the main components geranial and neral [24].
EOs from O. basilicum and P. graveolens showed moderate anti-bacterial activity corroborating other studies, which found higher effectiveness against Gram positive than Gram negative bacteria [25]. However, the antimicrobial property of basil EO was reported in other studies, in which a good activity against E. coli related to a significant amount of linalool was evidenced [26].
A. triphylla, C. citratus and L. cubeba EOs appeared to be the most active against A. fumigatus. These compounds would be of interest in poultry breeding, due to their low toxicity [27][28][29][30]. Furthermore, data available on the long lasting persistence of C. citratus EO [31] would suggest its feasibility, when used as disinfectant.
A. triphylla EO showed an interesting antifungal activity. Its composition is characterized by a 24% of sabinene, 36.7% of limonene and 12% of citronellal. The antifungal activity is probably due to sabinene, which appeared effective against A. fumigatus [32], as well as to citronellal [33]. Moreover, the present data are in full agreement with Correa-Royero et al. [34], who reported effectiveness against A. fumigatus at 99.2 µg/mL. A. triphylla was successfully assayed against Candida spp. both as EO [35,36] and as ethanolic extract [37], while it was not effective against Trichoderma viride [36].
In the present study, this EO was active at 0.5% concentration and such dilution appears to be safe for handling, following Tisserand and Young [27], who also recommend a maximum concentration of 0.9% for dermal application, to avoid phototoxic effects.
Lemongrass EO, with its high amount of both citral isomers, appeared to be another promising antifungal phytocomplex. This EO in fact is composed by a mixture of cis and trans-isomers of 3,7-dimethyl-2,6-octadiene-1-al (geranial and neral) with a percentage ranging between 38.4% and 35.2%, respectively.
Our result has been corroborated by Inouye et al. [38], who reported irreversible alterations in an in vitro model on apical growth of A. fumigatus elicited by C. citratus EO. Lemon, lavender and tea tree EOs induced a moderate and reversible inhibition, while cinnamon allowed a partial hyphal regrowth. Lemongrass exerted an antifungal activity against phytopathogenic fungi also at concentration up to 500 ppm [39].
L. cubeba EO showed neral and geranial content very similar to C. citratus, even if their MIC values were quite divergent (1.770 and 0.895, respectively), suggesting a different biological activity of phytocomplex.

Conclusions
This investigation shows promising results in order to use EOs for disinfection purposes in poultry farms and/or in hatcheries environment. The mixture with C. zeylanimcum and S. aromaticum could be a useful alternative treatment against E. coli. Moreover, C. citratus used alone could be employed to fight both E. coli and A. fumigatus, while A. triphylla EO would be used to enhance the control of spores' burden, aiming to reduce the risk of infection for both animals and workers.