Oral and contact toxicity of the extract obtained with hexane from Achyrocline satureioides on larvae and adult honey bees

Aim of study: To evaluate the antimicrobial activity of the hexanic extract (HE) of Achyrocline satureioides on Paenibacillus larvae - a gram-positive spore-forming bacillus that affects the larval stage and causes American Foulbrood (AFB) - and its oral and contact toxicity on larvae and adult honey bees. Area of study: A. satureioides plants were collected in Santa Monica (32° 05’ 29” S, 64° 36’ 54” W, Córdoba. Argentina). The larvae and adults of Apis mellifera were obtained from the experimental apiary of the University of Córdoba, Spain. Material and methods: P. larvae 9 was previously isolated and identified in the Laboratory of General Microbiology (Dept. of Micro -biology, National University of Río Cuarto, Argentina). The HE was obtained by liquid-liquid extraction. The minimum inhibitory concentration (MIC) of HE was determined by a microdilution method. This concentration and 2 ½ MIC were used for in vitro toxicity tests. Oral toxicity was tested on larvae, feeding them with both concentrations of the HE, while on adult bees the HE was spread to determine contact toxicity. Main results: The HE showed antimicrobial activity, the MIC obtained was 0.4 μg/mL. The HE presented very low toxicity at the MIC and 2 ½ MIC, with survival percentages to be around 95% for adult bees and larvae. Research highlights: The results show that this extract could be used for the development of an alternative product for a safe and effec tive treatment of AFB.


Introduction
The honeybee, Apis mellifera, is a social insect living in a colony or a hive comprising between 40 to 80 thousands individuals. Honeybees are recognized to have an essential role in nature for their efficient abilities as pollinators, considering that they pollinate more than one third of human crops, contributing to the cross-pollination of many fruits and forages (Watson & Stallins, 2016). In addition, they produce honey and other bee products such as propolis, royal jelly and bee wax, which are used by humans for several purposes. Thus, they are very important for mankind survival. Nevertheless, like all living organisms, honeybees can be altered with diseases and pests.
One of the most important pathogen of honeybees is Paenibacillus larvae, a gram positive and spore-forming bacterium, responsible for American Foulbrood (AFB), a highly virulent, contagious and fatal disease of the honeybee larvae (Genersch, 2010). The control of AFB in many countries, like UK and New Zealand, consists in burning the hives suspected to have AFB-affected colonies, leading to several economical losses. In these countries the use of antibiotics is banned or regulated by law; however in other countries the use of antibiotics, such as oxytetracycline hydrochloride or sulfathiazole, is a common strategy for the prevention and treatment of affected colonies, although they are not effective against the infectious spores, hence they only suppress clinical symptoms and mask the disease. As a result of the continuous use of antibiotics, serious problems have been generated, like the appearance of resistant strains of bacteria and the chemical residues that can persist in honey affecting its quality and safety for human consumption (Lodesani & Costa, 2005). Consequently, there is a need to develop new and effective AFB treatments, less prone to generate bacterial resistance, leave residues in honey and that are safe for honeybees.
Non-toxic natural compounds are being studied as an alternative to the use of antibiotics for the control of AFB and to considerably diminish the appearance of resistance in P. larvae, as well as the amount of antibiotic residues in beehive products. Several studies have reported antagonistic properties of natural compounds against P. larvae such as propolis (Antúnez et al., 2008;Bastos et al., 2008), essential oils and other derivated compounds from different plant species (Albo et al., 2003;Gende et al., 2009;Flesar et al., 2010;Sabaté et al., 2012).
Achyrocline satureioides (Lam.) DC, popularly known as "Marcela del campo", belongs to the Asteraceae family and has been used extensively in popular medicine. In recent years, extracts from A. satureioides have been the target of numerous in vitro and in vivo experimental testings confirming the ethnopharmacological uses described for them (Barioni et al., 2013).  tested A. satureioides extracts against the growth of P. larvae. They found that the water remaining after hydro-distillation and the decoctions were good inhibitors of the bacterial growth. In a more recent study (González et al., 2015), they showed that the compounds carrying the biological activity of A. satureioides against the growth of P. larvae were mainly contained in the hexanic extract obtained by liquid-liquid extraction from an aqueous-ethyl alcohol macerate of the aerial parts of the plant. However, the toxicity of the whole hexanic extract (HE), or its individual components, to the honeybee larvae or adult bees has not yet been tested.
Thus, the aims of this study were to evaluate the antimicrobial activity of the HE of A. satureioides on P. larvae and its oral and contact toxicity on larvae and adult honey bees, with a view to develop products based on this extract or its compounds for the control of AFB, in the future.

Plant material
Plant samples of A. satureioides were collected in Santa Monica, in the southern region of Córdoba Province (32° 05' 29" S, 64° 36' 54" W, Argentina) in the summer of 2016. A voucher specimen was conditioned and deposited at the Herbarium of the Faculty of Agronomy and Veterinary, National University of Río Cuarto, for verification of taxonomic identity as file herbarium 4658.
The aerial parts were dried at room temperature for 3 days; the hydro-alcoholic extracts were obtained by crushing the dried material and macerating it until depletion in a water: ethanol (1:1) mixture. The alcohol was then evaporated under low pressure. The HE was obtained by liquid-liquid extraction. Finally, the hexane was evaporated under low pressure (Gonzalez et al., 2015).

Phytochemical tests
The preliminary phytochemical evaluation of HE was carried out following Sanabria´s (1983) methodology. For this, the presence of alkaloids, flavonoids, tannins, naphthoquinones and terpenoids were determined.

Bacterial strain
P. larvae 9 was previously isolated and identified in the Laboratory of General Microbiology (Dept. of Microbiology, National University of Río Cuarto, Argentina) from a hive located in the southern region of Río Cuarto . The identification was phenotypically corroborated by biochemical characterization and confirmed by mass spectrometry (MALDI-TOF). The maintenance and storage of the strain was performed on MYPGP agar (Mueller-Hinton broth, yeast extract, potassium phosphate, glucose and sodium pyruvate) .

Determination of minimum inhibitory concentration (MIC) of HE of A. satureioides
The antimicrobial activity of the HE against P. larvae was determined by the broth microdilution method 3 Toxicity of HE from A. satureioides on larvae and adult honey bees (Cugnata et al., 2017). Colonies grown for 48 h in a MYP-GP agar plate were suspended in sterile peptone water (peptone 0.1% (w/v) and sodium chloride 0.85 % (w/v)) until reaching a final optical density at 600 nm of 0.1, measured with a UV-VIS spectrophotometer Spectrum.
-Preparation of HE dilutions: 6 mg (0.006 g) of the HE was dissolved in 15% dimethylsulfoxide (DMSO), which was previously determined as the nontoxic solvent concentration for the growth of P. larvae. HE dilutions were prepared in a concentration range from 0.048 to 600 µg/mL.
-Minimum inhibitory concentration (MIC): 100 μL of brain-heart infusion was placed in a 96 well microplate, then 100 μL of HE was added to well-1 and serially dilutions were performed until well-11. Finally, 50 μL of the previously prepared inoculum was added to each well. The microplate was incubated in microaerophilia at 37 °C for 48 h, then 10 μL of 0.01% resazurin were added to each well. It was incubated again during 1½ h under the same conditions and MIC was visually determined by the changing in color from blue (oxidized) to pink (reduced by microbial metabolism). The MIC was established as the last dilution where resazurin was not reduced, that is, it remained blue (Mann & Markham, 1998). DMSO at 15% was used as a negative control.

Determination of the minimum bactericidal concentration (MBC)
To determine the MBC, 100 μL of the MIC dilution and the previous ones were inoculated in MYPGP agar and incubated in microaerophilia at 37°C for 48 h. The MBC was considered as the last dilution without cell growth (Finelgold et al., 1992).

Apis mellifera material
The larvae and adults of A. mellifera were obtained from the experimental apiary managed by the beekeeping research group of the University of Córdoba, Spain. The collection was carried out from February to June 2018 from hives free of signs or symptoms of disease. To obtain the sample of bees, insects were chosen from the outer and middle frames of the hive, in order to have a representative sample of the population of worker bees. The larvae were obtained from a central breeding pattern; larvae were free of symptoms of disease and with an age of 1 to 2 days of development.

Toxicity test of hexanic extract (HE) on adult bees
The toxicity of HE was assessed at the MIC concentration (0.4 μg/mL) and at 2 ½ times MIC (1 μg/mL).
The toxicity of higher concentrations of the HE (1000 and 2000 μg/mL) was also assessed. Petri dishes (150 × 15 mm) were padded with filter paper on the inner bottom. Immediately after arriving from the field, the bees were slept with carbon dioxide gas for no more than 3 minutes, and then transferred to Petri dishes. Fifteen adult worker bees were placed in each Petri dish and 1 mL of each concentration was sprayed using a manual sprayer. They were fed with candy and water using a device placed inside each unit; fifteen bees in a Petri dish were sprayed with DMSO at 15% as the negative control and 0.07% deltamethrin (Deltavex Pour-On, Spain) was used as the positive death control. In addition, a survival check was performed where the bees were sprayed with distilled water. Petri plates were placed in incubators at 28 ºC ± 1°C and 60% relative humidity. The mortality of bees was assessed daily by visual inspection for 72 h (De Almeida et al., 2015;Piana et al., 2015). Four replicates for each experimental group were run.

Oral toxicity test of hexanic extract (HE) on larvae
First stage larvae were transferred from the hive combs to sterile 48 well culture plates. Twenty larvae were placed per plate, which were fed with a nutritious preparation containing 50% royal jelly, 12% D-glucose, 12% D-fructose, 2% yeast extract and 24% distilled water. The larvae were placed on 5 μL of food previously placed at the bottom of the wells, and viability was checked by movement. All process was carried out in a laminar flow cabin, and the tools and materials used were previously sterilized under UV light for 15 minutes (Schmehl et al., 2016).
First, the toxicity of the solvent (DMSO) on the larvae was determined in order to know its non-toxic concentration. Therefore, five concentrations of DMSO were evaluated: 50%, 25%, 15%, 1.2% and 0.008%. The HE was used at the concentrations of 0.4 μg/mL and 1 μg/mL, equivalent to the MIC and 2½ times the MIC, respectively. In addition, it was decided to test higher concentrations of HE to determine at which concentration toxic effects on larvae were first observed; therefore the toxicity of HE at 48, 120, 480 and 960 μg/mL was assessed. Honeybee larvae were fed daily with the different diets (controls and HE treatments) for a period of 3 days.
DMSO (1.2%) and distilled water were used as negative control and 0.07% deltamethrin was used as a positive control of death. Each plate was incubated in dark at 34°C and 95% humidity throughout the experiment. Larval mortality was recorded by visual inspection during the period of 24, 48 and 72 h. The dead larvae were removed daily and the diet was replaced every day (Zhu et al., 2014;De Almeida et al., 2015). The test was performed four times.

Analysis of results
The data obtained were analyzed under a completely randomized experimental design with four replications per treatment using the IBM SPSS Statistics statistical program. To observe the variation in the survival rate of individuals in each treatment throughout the test, survival graphs were performed using the GraphPad Prism 6 program; in addition, the data were analyzed by the variance analysis method (ANOVA) of a factor.

Phytochemical tests
The presence of several types of secondary metabolites was found in the HE of A. satureioides when qualitative identification tests were performed. Flavonoids, tannins, naphthoquinones and steroids were detected ( Table 1).

Determination of the minimum inhibitory concentration (MIC) of HE of A. satureioides
The MIC of the HE against P. larvae using the broth microdilution technique was 0.4 μg/mL. No bactericidal effects were observed. The MIC and 2 ½ times the MIC were used for the toxicity tests on larvae and bees (0.4 and 1 μg/mL).

Toxicity test of the HE of A. satureioides by contact on adult bees
The toxicity of different concentrations of HE on bees was assessed. In Fig. 1, the survival rates of bees in the different treatments were observed at 24, 48 and 72 h of exposure. During the test, the percentages remained around 95% for all the treatments: negative control (99.6% ± 0.8%), DMSO (99.0% ± 1.0%) and treatment with EH at 0.4 (99.6% ± 0.8) and 1 µg/mL (97.9% ± 2.7), meaning that there was no acute toxicity in bees when they were in contact with HE at MIC and 2½ times MIC concentrations. In the positive control (deltamethrin), it was observed a lethal effect on bee survival at 24 h of exposure.
The HE was also tested at higher concentrations than MIC (1000 and 2000 µg/mL) to analyze what would happen with an HE overdose. At 1000 µg/mL the survival rate slightly decreased to 91.0% ± 3.8% at the end of the test, and at the concentration of 2000 µg/mL a greater toxic effect by contact in bees was observed since the survival rate decreased to 66.6% ± 6.6. The negative control showed statistically significant differences (p<0.05) with the positive control, EH at 1000 µg/mL and EH at 2000 µg/mL. Also, the treatment with EH at 2000 µg/mL showed statistically significant difference with the negative control, DMSO, positive control, EH at 0.4 µg/mL and 1 µg/mL. The NOAEL (No Observed Adverse Effect Level) in bees was 1 µg/mL (p=0.05).

Oral toxicity test of the HE of A. satureioides on larvae
In order to perform the larval oral toxicity test, it was necessary to determine the previous concentration of DMSO without any toxic effect on it. The maximum non-toxic concentration found was 1.2% in water, so the HE dilutions were prepared at this DMSO concentration, thus avoiding toxic effects produced by the solvent.
First, the toxicity of HE at concentrations of MIC and 2 ½ times MIC was assessed. The results showed that the survival rate of larvae remained close to 95% until 72 h (Fig. 2); therefore, HE at these concentrations had no toxic effect on larvae compared to positive control. The statistical analysis of covariance showed that the negative control only presented a statistically significant difference with the positive control (p<0.05). This indicated that the HE of A. satureioides did not cause toxic effects on larvae at concentrations of 0.4 and 1 µg/mL.
When higher HE concentrations were used, the results indicated that the survival rate of the larvae was similar for all the treatments: control, DMSO 1.2%, DMSO 0.6%, and HE (480 µg/mL, 120 µg/mL and 48 µg/mL) (Fig. 3). Therefore, the larvae were not affected by HE at these concentrations. On the other hand, the survival rate of the larvae began to decrease in the positive control and in the treatment with HE at 960 µg/mL, after 24 h of exposure till 72 h.
The statistical analysis showed significant differences in the negative control with respect to positive control and the treatment with HE at 960 µg/mL, indicating that HE of A. satureioides caused toxic effects on larvae at the

Discussion
The Asteraceae family is particularly rich in phenolic compounds (Emerenciano et al., 2001;Heiden et al., 2007;Katinas et al., 2007) and A. satureioides, a member of this family, has been reported as one of the species with higher flavonoid content like others such as Ginkgo biloba and Epilobium parviflorum (Arredondo et al., 2004). In this work, a qualitative assay of the HE of this plant was analyzed and the results showed that flavonoids were present as it was reported by other authors who studied the chemical composition of A. satureioides extracts and considered them as an interesting source of flavonoids, regardless of whether the extraction was performed with polar or nonpolar solvents (De Souza et al., 2007;Fachinetto et al., 2007;Santin et al., 2010). In addition, the presence of other phenolic compounds such as tannins was observed in the HE of this plant species, which coincided with a study carried out by Casero et al. (2015) where they isolated and identified from the HE of A. satureioides four new compounds of phenolic nature derived from phloroglucinol which also presented antimicrobial activity against gram positive and gram negative bacteria.
The test for naphthoquinones and/or anthraquinones indicated the possible presence of these compounds in the HE of A. satureioides, which has been reported in other species belonging to the Asteraceae family, such as Senecio subpanduratus (Gratti et al., 2014). The steroid and/ or triterpenoid test (CCF Libermann-Burchard) indicated the presence of these metabolites in the HE of A. satureioides, which is consistent with a study conducted by Gillij et al. (2008) where they found monoterpenoids and sesquiterpenes.
The antibacterial activity of the HE of A. satureioides was evaluated against the strain P. larvae 9 and the effective action of this extract was probed. The HE inhibited the growth of this pathogenic bacteria at concentrations lower than or equal to conventional antibiotics such as oxytetracycline, whose reported value was 0.5 to 1 μg/mL (Sabaté et al., 2012). According to Alippi et al. (2007), P. larvae can be considered susceptible to antibiotics when MIC < 4 µg/mL, intermediate for MIC values between 4 and 8 µg/mL and resistant when MIC >16 µg/mL. Therefore, the low concentration of HE necessary to inhibit the growth of this microorganism indicated the high sensitivity of strain 9 to HE.
The biological activity of HE was compared with other investigations carried out with A. satureioides; e.g. González et al. (2015) reported mean MIC values for the HE of 60 μg/mL, with the lowest value of 16 μg/mL. Likewise, Sabaté et al. (2012) found a MIC range from 16 to 125 μg/mL for the HE of the same plant, resembling more to Gonzalez et al.´s (2015) study than to this study, where a better effect was observed with a minor MIC. These differences could be attributed to the place of collection of the plant material and/or the date of collection, since it is well known that the qualitative and quantitative composition and, therefore, the biological activities of a plant extract, could vary according to the different geographic, climatic factors, altitude, harvest time and growth stage. Consequently, the chemical composition is not homogeneous and different chemotypes could be originated (Ferraro et al., 2008). This was evidenced in a research conducted by Cezarotto et al. (2011) with the essential oil of the aerial parts of A. satureioides where they found a significant variation in chemical composition and antibacterial activity when the plant was collected during the same year but at different times.
Comparing the results obtained in this study with other research works of the antibacterial activity against P. larvae of different extracts of other plant species, it could be observed that the HE of A. satureioides presented better activity than the methanolic-dichloromethane extracts of Humulus lupulus L (MIC=2 μg/mL), Myrtus communnis L. (MIC=2-8 μg/mL) and Eucalyptus gunnii Hook.f. (MIC=16-32 μg/mL) (Flesar et al., 2010). The crude extracts of Brazilian species exhibited MIC values of 12.76 mg/mL for Nasturtium officinale and 30.51 mg/mL for Copaifera officinalis (Piana et al., 2015); the species Chromolaena odorata, also belonging to the Asteraceae family, presented a MIC range between 16 and 64 μg/mL (Chaimanee et al., 2017) and the ethyl acetate fraction of the species Buddleja thyrsoides Lam. showed MIC results of 1.68-3.36 mg/mL (Boligon et al., 2013).
The antimicrobial activity of essential oils of different plant species against P. larvae was reported by other authors; Cinnamomum zeylanicum oil showed MIC values between 38 and 50 μg/mL, Lavandula officinalis between 350 and 400 μg/mL, Mentha piperita between 650 and 700 μg/mL, Citrus paradise MIC of 385 μg/mL, Citrus sinensis MIC of 840 μg/mL, Citrus nobilis MIC of 815 μg/ mL, Citrus limon MIC of 764 μg/mL (Fuselli et al., 2008), Mentha spp. MIC >600 μg/mL (Gende et al., 2014) and Pimenta dioica (L.) Merr. MIC of 78 μg/mL (Ansari et al., 2015). A MIC value of 142 μg/mL has also been reported when glycerol monolaurate nanocapsules were evaluated (Lopes et al., 2016). Therefore, it could be observed that the HE of A. satureioides in this study demonstrated significantly lower MIC values than essential oils, which are usually described as the most biologically active plants compounds.
The better effect of HE on P. larvae could be due to a combination of the activities of the different compounds that constitute it, since it has been observed that single compounds did not show better activity than the whole extract (Flesar et al., 2010;Boligon et al., 2013;Piana et al., 2015). These observations could be indicating the existence of synergistic interactions between the metabolites present in the HE of this plant that enhance the effect for the control of pathogenic bacteria. Joray et al. (2010), evaluating the interactions between the metabolites of A. satureioides in relation to their antibacterial activity, found the existence of a combined action between the metabolites of the plant.
The results obtained in this study in relation to antimicrobial activity demonstrate that A. satureioides HE had good antimicrobial properties, and taking into consideration that when performing the preliminary phytochemical analysis, metabolites such as flavonoids, tannins, steroids, triterpenoids and naphthoquinones were identified in abundance, it could be said that they were directly responsible for the activity on P. larvae, since several investigations frequently reported the antibacterial activity of these secondary metabolites (Flesar et al., 2010;Piana et al., 2015;Chaimanee et al., 2017). It was described that for flavonoids, the bacteriostatic effect was attributed to the presence of phenolic hydroxyls in their structure, which easily penetrate through the bacterial cell membrane, combine and precipitate protoplasmic proteins, denaturing them and acting as protoplasmic poisons (Coloma, 2009).
In vitro confirmation of toxicity or side effects on the larvae and bees health is a prerequisite for assessing the practical usefulness of extracts with biological activity. For example, an interesting "biological" activity of Sysygium aromaticum against Varroa destructor has been reported, but the oil produced high bee mortality, even at low concentrations (Maggi et al., 2010). As regards essential oils with good effects for the control of P. larvae, it has been reported that andiroba oil is toxic for bees (Santos et al., 2012); similarly, nanoemulsions of the same oil presented a significant toxic effect for the larvae reporting a mortality of 26%. On the other hand, nanoemulsion with Copaiba oil did not present toxic effects on larvae or bees (De Almeida et al., 2015). A research work conducted with glycerolmonolaurate showed high toxic effect on bees with a survival rate of 55%, but the nanoencapsulation of the same decreased the toxic effect, with a survival rate of 95% at the end of the test (Joray et al., 2010). Chaimanee et al. (2017) found that the extract of Piper ribesioides and Piper sarmentosum were highly toxic to bees at 48 h; on the contrary, the crude extracts of Allium sativum, Cymbopogon citratus, Senna alata and Thunbergia laurifolia did not cause toxic effects in adult bees, even at high concentrations. A study of ethyl acetate, butanolic and dichloromethane fractions and the crude extract of the species Buddleja thyrsoides Lam. showed a clear mortality of bees only in the treatment with the dichloromethane fraction (Boligon et al., 2013). The investigation of the toxic effect of the crude extract of Calendula officinalis and the ethyl acetate fraction of Cariniana domestica showed, respectively, 66.68% and 26.64% survival in bees at the end of the experiment. On the other hand, no toxic effects were found to be caused by the crude extract of Nasturtium officinale and the n-butanol fraction of C. domestica (Piana et al., 2015), as well as the crude extract and fractions of Scutia buxifolia (Boligon et al., 2013). Our results showed that the HE of A. satureioides did not cause the death of larvae and bees at the same concentration that exhibited antimicrobial activity against P. larvae; even at higher concentrations of MIC no bee/larvae mortality was observed until 72 h. Considering that HE did not have toxic effects on larvae or bees at concentrations of MIC and 2½ times MIC, the HE could be used at a concentration of 0.4 µg/mL for the control of the pathogen in hives. Furthermore, it could be used up to a concentration of 480 µg/mL, since such concentrations would not affect the health of Apis mellifera.
In summary, the results presented in this research demonstrated the antibacterial activity of the HE of A. satureioides on P. larvae at relatively low concentrations. The toxicity analysis of the HE of A. satureioides on larvae and adult bee of Apis mellifera showed no toxicity at concentrations of MIC (0.4 µg/mL) and 2 ½ times MIC (1 µg/mL) until 72 h of the test. Therefore, since HE is safe for larvae and bees at the concentration that inhibits the growth of the pathogen, it could be used in the future as a biopesticide for the control of AFB in hives.