Antimicrobial and miticide activities of Eucalyptus globulus essential oils obtained from different Argentine regions

The biological activity of Eucalyptus globulus essential oils derived from plant material obtained from different geographic areas was analyzed in in vitro experiments on Paenibacillus larvae, Varroa destructor, and Apis mellifera. The physicochemical properties, composition, antimicrobial, and bioactivity of these essential oils were studied. The bioactivity against P. larvae was analyzed by means of two in vitro techniques (tube dilution and bioautography). Mite and bee lethality were estimated using a complete exposure method test with oils at different concentrations. Essential oils differed in their composition, albeit their similar physicochemical properties. The minimal inhibitory concentrations range for E. globulus essential oil lay between 600-700 μg mL for Mar del Plata and 900-1,200 μg mL for Valle de Conlara essential oils against all P. larvae strains. The bioautography method determined that limonene accounted for the greatest antimicrobial activity with respect to the other compounds. The complete exposure method at 24, 48, and 72 h yielded lower LC50 values for mites exposed to E. globulus essential oils from Conlara. LC50 values for acari after 72 h of treatment with E. globulus from Mar del Plata and Conlara were 47.1 and 11.7 μL capsule, respectively; whilst those for bees was > 20 μL capsule for both oils. The E. globulus essential oils tested in this study featured high efficiency against V. destructor, yet their antimicrobial activity against P. larvae proved to be lower, and innocuous to bees. The present experience promotes the use of active compounds for American fouldbrood and Varroosis management. Additional key words: Apis mellifera, bioactivity, eucalyptus, Paenibacillus larvae, Varroa destructor.


Introduction
American foulbrood (AFB), an infectious disease caused by the spore-forming bacterium Paenibacillus larvae (Genersch et al., 2006), along with the varroasis produced by Varroa destructor (Anderson and Trueman, 2000), an ectoparasitic mite, constitute the two main pathologies affecting brood and adult honeybee (Apis mellifera), thereby posing a genuine threat to apiarian producers.
A common strategy adopted to prevent and treat colonies affected with AFB is the use of antibiotics, particularly oxytetracycline hydrochloride (OTC) (Hansen and Brodsgaard, 1999).On the other hand, other investigations suggest that the early detection of this disease, the withdrawal of affected hives, and the use of hygienic behavior lines, might be suff icient to control AFB making possible to manage apiaries without antibiotics and to preserve honey quality (Basualdo et al., 2008).Efforts to control Varroa have focused on the evaluation of synthetic acaricides such as coumaphos and fluvalinate, among others (Maggi et al., 2008).In both cases, the indiscriminate use of chemicals to control said pathologies has encouraged the emergence of resistance in many countries and led to hive products contamination (Elzen et al., 1999a,b;Mathieu and Faucon, 2000;Thompson et al., 2002;Bogdanov, 2006;Martel et al., 2006;Maggi et al., 2008).Moreover the chemical control of mites and bacterial diseases may not be acceptable in pure honey production, given the contamination issues arising therefrom, as well as from other hive products consumed by the general public (Kevan, 1999).
The current tendency for chemical control is the utilization of natural products as part of an integrated control strategy.This is the case of the neem tree Azadirachta indica (González Gómez et al., 2006) and of essential oils against V. destructor (Imdorf et al., 1999;Eguaras et al., 2005;Ruffinengo et al., 2005;Maggi et al., 2009).Along these lines, plant extracts and essential oils against P. larvae (Alippi et al., 1996;Fuselli et al., 2006;Gende et al., 2007) are now being employed to such an end.The substances contained in natural products can be considered potentially useful to control honeybee diseases (Gende et al., 2008a); and their incorporation into chemical substances as a natural alternative could assist in the reduction of beehive products contamination.
Essential oils are characterized by changes in their chemical composition, which are dependent on the plant developmental state, the part used for extraction, its geographical location, and on the physical and chemical characteristics of the soil and climate in question.Several studies on Eucalyptus spp.cultivated in different regions worldwide have centered their attention on the variability of chemical composition.Different percentages of 1,8-cineole have been reported in E. globulus: 64.5% in Uruguay (Dellacasa et al., 1990), 75% and 77% in Cuba (Montejo Loret de Mole et al., 1985), 86.67% in USA (Nishimura and Calvin, 1979), 58% to 82% in Morocco (Ahmadouch et al., 1985), and 48.7% in South Africa (Thilivahalt Ndou and Von Wandruszka, 1986).Indeed the variation in bioactivity results accounted by different researches can be attributed to the lack of constancy in the chemical composition of essential oils (Imdorf et al., 1999).
In the light of the aforementioned, the aim of this study was to analyze the bioactivity of E. globulus essential oils extracted of plants from different geographic areas against P. larvae and V. destructor, and to ascertain the role that variation in their composition plays in the action against both diseases.
plants located in the Faculty of Exact and Natural Sciences, Universidad Nacional de Mar del Plata (PV 89 and PV 90, respectively).Fresh plant material was dried prior to distillation (20-27°C and ~50% RH); and dried leave oils were obtained by hydrodistillation using a Clevenger-type apparatus (Richard et al., 1992) for 2 h.An average of 100 g of leaves was used in each experiment, and several distillations were performed until the volume required to run all trials was reached.
The oils were dried over anhydrous sodium sulphate and stored in screw-capped dark glass vials at 5°-8°C until further testing.

Essential oil analyses
The oils were analyzed by GC-FID-MS, using a Perkin Elmer Clarus 500 model chromatograph equipped with a single split/splitless injector (split ratio: 1:100) connected, through a flow divisor, to two capillary columns (fused silica, 60 m × 0.25 mm i.d., 0.25 µm film thickness) coated with: a) polyethylenglycol (MW ~ 20,000) (DB-Wax, J&W Scientific) and b) 5% phenyl-95% dimethylpolysiloxane (DB-5, J&W Scientific).The polar column was connected to a FID detector, and the non-polar column to a FID detector and to an MS quadrupole detector (70 eV), using a vent system (MSVent™).Operating conditions were as follows: carrier gas was helium (constant flow: 1.87 mL min -1 ); and oven temperature was programmed at an initial temperature of 90°C, increased at a rate of 3°C min -1 to 225°C, and maintained for 15 min.The injector and FID detectors temperatures were set at 255°C and 275°C, respectively; and that of the transfer line at 180°C.The ion source temperature was of 150°C, and the acquisition mass range of 40-300 m z -1 .Quantitative data were determined from the minor response of both FID area values, and expressed as percentages calculated from the peak-area percentage.Oil components were identif ied by comparing their retention indices to C 6 -C 20 alkanes on both columns, and their mass spectral data to that from electronic libraries (Willey, 2006;Adams, 2007) and to data published in the literature.

Physicochemical properties determinations
Density to 20°C, triplicate of 1.0 mL of essential oil, was weighed, and the average of the values obtai-ned was calculated (Montes, 1981).The refractive index was determined at 20°C ± 0.05°C with an Abbe refractrometer, in compliance with AOAC off icial method 921.08 (AOAC, 1999).

Thin layer chromatography
The thin layer chromatography (TLC) of the essential oils was performed on silica gel plates (0.2 mm Kieselgel 60 F254, Merck).The eucalyptus oils were applied to two TLC plates using an aliquot of 5 µL (using Drummond micro-capillaries), and developed (93:7 toluene/ethyl acetate).On one plate, the separated compounds were sprayed with sulphuric acid in ethanol, and later on with vanillin in ethanol, followed by heating at 110°C (Wagner and Bladt, 1996).The other plates were used to conduct the bioautography assay.

Bacterial biomass preparation
Bacterial strains of P. larvae, collected from five locations in Buenos Aires province: La Plata, Mar de Cobo, Vivoratá, Mar del Plata and Vidal, were isolated from brood combs of beehives with clinical symptoms of American foulbrood.Isolation was achieved on MYPGP agar (Dingman and Stahly, 1983); and to inhibit Paenibacillus alvei growth, it was supplemented with 9 µg mL -1 of nalidixic acid.Plates were incubated under microaerobic conditions (5-10% of CO 2 ), and strains were identified using biochemical tests (Gordon et al., 1973;Alippi, 1992).Pure strains were maintained on MYPGP agar with 15% v v -1 glycerol until used.
Vegetative cells of P. larvae previously cultivated on MYPGP agar for 48 h at 35 ± 0.5°C were suspended in double distilled sterile water, and the suspension was standardized according to FDA (1998) method.Concentration was adjusted to 0.5 of Mac Farland scale for measuring antimicrobial activity with serial dilution.

Determination of minimal inhibitory concentrations
The minimal inhibitory concentration (MIC) is defined as the lowest concentration of an antimicrobial agent capable of inhibiting the visible growth of a microorganism after incubation (Lennette et al., 1987).
MIC individual determination was directly evaluated by turbidity observation.The oils were mixed in water and emulsified with 8% (v v -1 ) propylene glycol (1,2propanediol, The Merck Index, 1996).For broth microdilution, 100 µL of MYT broth (Gende et al., 2008b) were placed in each of the 96-well microtitre plates and then diluted to obtain serial dilutions.Microbial biomass suspension was added to each serial dilution.Final Serial dilution concentrations ranged between 12.5-2,000 µg mL -1 .Positive and negative controls (with microorganisms and water, respectively) were used.Microtiter plates were incubated at 35 ± 0.5°C for 48 h so as to determine MIC values.Antimicrobial activity was tested by triplicate analyses for oil and strains.The MICs of oxytetracycline were also determined in parallel experiments as a way to control tested microorganisms sensitivity.

Bioautography
This technique was employed to define the active constituents (Iskan et al., 2002).Vegetative cells of P. larvae previously cultivated on MYPGP agar for 48 h at 35 ± 0.5°C were suspended in double distilled sterile water and the suspension was adjusted to 2 of Mac Farland scale.Twenty milliliters of MYPGP medium were poured on TLC plates (not previously revealed) placed in Petri dishes; microbial suspension was placed on the medium and incubated at 35 ± 0.5°C for 48 h under microaerobic conditions.Microbial growth inhibition was determined by measuring the area of the inhibition zones after being revealed with a triphenyl tetrazolium chloride solution to 5% w v -1 in water.Bacteria reduced tetrazolium salt through dehydrogenase activity and produced intensely colored formazan, as reported by Eloff (1998).The inhibition area was observed and classified as 0, -, +, ++, in line with the length area of the inhibition zones in relation to the chromatography spot.Triplicate analyses were performed to determine oil and strains antimicrobial activity by bioautography.

Bioactivity of essential oils against V. destructor and A. mellifera
The bioactivity of E. globulus essential oils against V. destructor was established using a complete exposure method (Ruffinengo et al., 2005).Treatments were assayed in Petri dishes (140 × 20 mm); using new dishes for each experiment.Each essential oil was diluted in ethanol to a desirable concentration, and 1 mL of solution was then applied to the bottom of the Petri dish.Concentrations between 0 and 20 µL per cage of essential oils were used.Ethanol was evaporated from the dishes by exposing it to airflow for 3-5 min.Five newly emerged adult bees (between 0 and 3 days old) and five adhering Varroa mite females obtained from brood cells were placed in each Petri dish.Bees and mites were exposed to essential oils for 72 h.The bees present in Petri dishes were fed on 3 g of candy, and incubated at 30°C and 70% RH during the test.Control treatments consisted of Petri dishes with the evaporated solvent and dishes treated with fluvalinate (technical grade tau-fluvalinate).Five replicates were used for each experimental unit.The number of dead Varroa and dead bees was determined after visual inspection of the dish bottoms after 24, 48, and 72 h, respectively.

Statistical analysis
To determine the bioactivity of E. globulus essential oils (obtained from different geographic areas) against V. destructor and A. mellifera, statistical analyses were conducted in accordance with the specific software for LC 50 values calculation, using 95% confidence interval (USEPA, 1986) and EPA software (vers.1.5) as proposed by Lindberg et al. (2000).Mortality values were adjusted as a function of natural mortality in agreement with Abbott (1925).
Comparisons between LC 50 value pairs of the different E. globulus oils in V. destructor were carried out each time (24, 48, and 72 h) by means of LC 50 greater LC 50 lower -1 quotient; and statically significant differences were detected when the statistical value was higher than the corresponding critical value set forth by APHA (1992).

Results
Oils obtained by distillation with Clevenger apparatus from materials from different geographic areas yielded the same chemical composition, except for aromadendrene, trans-pinocarveol and alpha-terpineol; and featured similar physicochemical properties.Table 1 lists the essential oil composition of the E. globulus samples collected from Mar del Plata and Conlara.As Biological activity of E. globulus essential oils against honey bees diseases a whole, 11 components were fully characterized by CG/ MS analyses, eucalyptol (1,8-cineole) and alphapinene being the major compounds.Physicochemical properties values were as follows: density of 0.9232 g mL -1 at 20°C and refractive index of 1.4922 for E. globulus from Mar del Plata; and density of 0.9232 g mL -1 at 20°C and refractive index of 1.4672 for essential oil from Conlara.
Table 2 lists MIC values for the five strains of P. larvae in the presence of both essential oils.MIC values for E. globulus from Mar del Plata ranged from 600 to 700 µg mL -1 , while those from Conlara oil from 900 to 1,200 µg mL -1 .As it can be noticed, the higher percentage of 1,8-cineole is not connected to the higher antimicrobial activity values.
On silica gel thin-layer chromatograms, eucalyptus essential oils were separated into 6 and 4 spots for Mar del Plata and Conlara, respectively.Both samples were characterized by one main zone based on 1,8-cineole (Rf 0.35-0.4).In addition, only two of the bands could be identified by co-chromatography using pure standards, alpha-pinene and limonene with Rfs 0.95; 0.73, respectively.
The bioautography method was applied to strains of P. larvae.This bioassay allowed to evaluate the minimum concentrations of the compound responsible for the biological activity observed, at first sight, as growth inhibition areas.The clearest inhibition zones in this experiment corresponded to limonene if compared to the other compounds identified by TLC technique.Table 3 displays the results obtained for the main components in antimicrobial activity tests against P. larvae strains from five different geographic areas.
The percentage of mortality recorded for V. destructor and A. mellifera at 24, 48, and 72 h is shown in   of exposition (p < 0.001).Table 5 exhibits the estimated LC 50 values for V. destructor obtained at each time interval for each treatment along with their comparison.EPA software version 1.5 was employed to estimate the specific values of LC 50 based on the acari mortality percentage obtained.On the other hand, with regard to A. mellifera, the estimated LC 50 values reached at each time interval for each treatment exceeded 20 µL per cage, owing to the low bee mortality (0, 0.4, and 8%) for all concentrations assayed.For mites and bees, LC 50 of fluvalinate decreased with time, and significant differences were identified regarding treatments in each time (p < 0.001).

Discussion
Among the 11 compounds identified by means of CG/MS analyses eucalyptol and alpha pinene were the main components, according with Retamar (1982).With regard to the role that environmental factors play on plants secondary metabolites, expectations lie in detecting variations in these compounds in the oils of the same species obtained from diverse geographic regions.These composition differences with respect to the geographic distribution of the plant material can explain the distinct composition of the two eucalyptus essential oils analyzed.
In general, the E. globulus oil shows a lesser inhibitory effect on P. larvae growth when compared to the essential oils of other plants, such as lemon grass (Cymbopogon citratus) with MIC values ranging from 50 to 100 mg L -1 , thyme (Thymus vulgaris) with values ranging from 100 to 150 mg L -1 (Alippi et al., 2001), and cinnamon (Cinnamomum zeylanicum) with MIC values between 25 and 100 mg L -1 (Floris et al., 1996;Gende et al., 2008a).The antimicrobial activity data obtained in this work mirror those reported by Alippi et al. (1996) for the same vegetal species against P. larvae.The poor antimicrobial activity of E. globulus could be ascribed to the chemical composition rich in terpenic compounds, which is consistent with previous studies reporting that these chemical structures are the least bioactive of the essential oil components (Lattaoui and Tantaoui-Elaraki, 1994;Cosentino et al., 1999).
The results achieved by bioautography suggest that limonene was the main compound responsible for the inhibitory activity against the P. larvae strains analyzed.These results are in line with those obtained by Fuselli et al. (2008), in which the antimicrobial activity of citrus oils against P. larvae strains is attributed to the volatile specific component, i.e., limonene.The major antimicrobial activity of limonene in relation to that of the other two major components of these essential oils could be explained by the delocalized system of electrons resulting from the presence of two double bonds that allow proton exchange, thereby rendering the substances more active against microorganisms Different letters indicate significant differences between each other within each time (p < 0.001).(Ben Arfa et al., 2006).An increase in activity depends on the type of alkyl substituent incorporated into a nonphenolic ring structure, an alkenyl substituent (1-methylethenyl) results in an increased antibacterial activity (Dorman and Deans, 2000).More often than not, the inhibitory action of aroma compounds is associated with hydrophobicity, which is directly correlated to the log P (partitioning behavior of the lipophilic compounds in octanol/water) as well as to their partition in the cytoplasmic microbial membranes (Lanciotti et al., 2003).The most hydrophobic compounds are generally considered to be the most toxic (Sikkema et al., 1995;Weber and de Bont, 1996).In this particular case, limonene featured a log P value of 4.83 in comparison with 1,8-cineole with 2.74 (Onken and Berger, 1999), which would also explain why limonene is more effective against P. larvae than 1,8-cineole is.
If a comparison of both essential oils tested in this study is drawn, E. globulus from Conlara yielded high efficiency against V. destructor, but exhibited lesser antimicrobial activity against P. larvae, while they were well tolerated by worker honeybees, according to the complete exposure method.This effect may be due to the greater 1,8-cineole content.The results for V. destructor were congruent with the data obtained by Imdorf et al. (2006) and Ruffinengo et al. (2007), who stated that 1,8-cineole led to high mite mortality.Yet, with respect to adult bee workers mortality, results differed from those accounted by these authors.The differences noticed could be explained by the application forms adopted, being, in this study, that of complete exposure and of pulverization in the others.
Previous researches have evaluated the activity of essential oils against V. destructor, P. larvae, and A. mellifera on a separate basis; paying no attention to the composition variation associated with the geographic origin.This work centers its efforts on the relevance that knowing about oils composition has, in view of the fact that the activity against both pathogens varies depending on the geographic distribution of the plant material.With the demand for naturally occurring bioactive (both acaricid and antibacterial) agents and a never ending search for natural biologically-active products, it is hence of interest to develop a study to identify the components of essential oils with both activities.
The present experience, in which eucalyptus oil antimicrobial and miticide properties were tested in vitro, promotes the use of its active compounds for American fouldbrood and Varroosis management in honey bee colonies.

Table 1 .
Chemical composition of E. globulus essential oils

Table 2 .
Antimicrobial activity expressed as minimal inhibitory concentration (MIC, µg mL -1 ) of E. globulus essential oils against P. larvae strains P.

Table 4 .
Biological activity of E. globulus essential oils against honey bees diseases 647 Mite and bee mean mortality (%) after 24, 48 and 72 h for different concentrations of essential oils of E. globulus from two different geographic locations (CL: Conlara.MdP: Mar del Plata)

Table 5 .
LC 50 (µL Petri dish -1 ) estimated for V. destructor and A. mellifera for essential oils and fluvalinate from Conlara and Mar del Plata