In vitro growth-inhibitory effect of Brazilian plants extracts against Paenibacillus larvae and toxicity in bees

American foulbrood (AFB) is a serious worldwide spreading disease in bees caused by Paenibacillus larvae. Plants extracts are known to decrease or inhibit the growth of these bacteria. The purpose of this study was to evaluate the antimicrobial activity of Calendula. officinalis, Cariniana domestica, and Nasturtium officinale extracts against the P. larvae and to evaluate the toxicity of the extracts in bees. In vitro activity against P. larvae of the extracts was evaluated by micro dilution method and the minimal inhibitory concentrations (MICs) were also determined. The concentrations used in the toxicity test were established based on the MIC values and by the spraying application method. The P. larvae was susceptible to the evaluated crude extract of C. officinalis and N. officinale. To C. domestica, only the ethyl acetate (EtAc) fraction and n-butanol (BuOH) fractions had activity against P. larvae. Toxicity analysis in bees showed no toxicity for N. officinale crude extract and for C. domestica BuOH fraction during 15 days of treatment, however, some deaths of bees occurred during the first three days of treatment with C. officinalis and C. domestica EtAc fraction. The results with these species were firstly described and showed that N. officinale crude extract and C. domestica BuOH fraction both presented not toxic effects in the concentration tested by the spraying application method, and can be a useful alternative for treatment or prevention of AFB.


INTRODUCTION
American foulbrood (AFB) is a serious worldwide spreading disease of honey bee (Apis mellifera) caused by the spore-forming, Gram-positive bacterium Paenibacillus larvae (Eguaras et al. 2005, Flesar et al. 2010).AFB is highly infectious and can be fatal to colonies.At a minimum it decreases honey productivity of colonies and increases labor and treatment costs to beekeepers to control disease transmission (Hansen andBrødsgaard 1999, Bastos et al. 2008).
The popular approach in treatment of AFB in some countries, such as the US, is to suppress the
clinical phase of the disease with supplemented antibiotics, however, this practice has been shown to lead to bacterial resistance (Eguaras et al. 2005, Flesar et al. 2010) such as tetracycline in which vegetative cells of P. larvae have become increasingly resistant to treatments (Kochansky et al. 2001, Cox et al. 2005).Furthermore, antibiotic residues are found in honey and its use in the colony can also contaminate royal jelly (Bogdanov 2006).This situation calls for an alternative and effective control of the disease with either therapeutics or prophylactic feed additives that do not contribute to the phenomenon of bacterial resistance (Lewis andAusubel 2006, Flesar et al. 2010).
Plant extracts, herbs, spices, essential oils and isolated compounds are known to delay or inhibit the growth of bacteria, yeast and moulds (Hayouni et al. 2008, Boligon et al. 2013).Brazil has in its flora, species with antimicrobial activity such as Calendula officinalis L. known in Brazil as calendula (Efstratiou et al. 2012), Cariniana domestica (Mart.)Miers.known as jequitibá-roxo (Janovik et al. 2009), and Nasturtium officinale commonly called agrião (Antúnez et al. 2008).Studies conducted by Boligon et al. (2013) showed positive results of Scutia buxifolia extract against P. larvae as well as low toxicity to bees.In addition Copaifera officinalis (copaiba) oil in studies by Santos et al. (2012), have also presented similar results.
The aim of the present work was to evaluate the antimicrobial activity of C. officinalis, C. domestica and N. officinale extracts against the etiologic agent of American Foulbroad Disease, the P. larvae.Furthermore, the toxicity of the extracts in honey bees A. Mellifera was also evaluated.

PLANT COLLECTION AND EXTRACTION
The C. officinalis (flowers) and N. officinale produced in a hydroponic system (leaves and branches) used in this study were purchased in supermarkets in the city of Santa Maria-RS, Brazil.The C. domestica (leaves) were collected in Tangará-da-Serra, Mato Grosso do Sul, Brazil.Exsiccate of C. domestica was archived as voucher specimen in the herbarium of the Department of Pharmacology at UFSM (SMDB 11818).The vegetables were cleaned and dried at room temperature.The dried plants were powdered in a knife mill and the powder was macerated with 70% ethanol for a week with daily shake-up.The extract was filtered and concentrated under reduced pressure in rotary evaporator (temperature ± 40°C) in order to eliminate the ethanol, then the aqueous extract was dried, thus obtaining the crude extract.
For the C. domestica, an equal part of aqueous extract was partitioned with solvents of increasing polarity dichloromethane (DCM), ethyl acetate (EtAc) and n-butanol (BuOH) and were also dried to provide each corresponding fraction.

MICROORGANISM TESTED
In this study, P. larvae (ATCC 9545) from the collection of the Ministry of Agriculture (LANAGRO/RS) Brazil was used.The microorganism was grown in Mueller-Hinton broth (Difco, Sparks, Maryland, USA) at 37°C for 24h and maintained on slopes of nutrient agar (Difco).MIC was defined as the lowest concentration of compounds that inhibit bacterial growth.This test was performed in triplicate on separate occasions.The 2,3,5-triphenyltetrazolium chloride was used as an indicator of bacterial growth.

TOXICITY ASSAY
The C. officinalis and N. officinale crude extracts, and EtAc and BuOH fractions of C. domestica were dissolved in DMSO.The concentrations used in the toxicity assay were determined based on the MIC values.The spraying application method was performed according to Santos et al. (2012).Petri dishes (150 x 15 mm) padded with absorbent filter paper on the inner bottom and with an extra lid of plastic mesh were used.Fifteen adult worker bees were placed in every modified Petri dish.Then, 1 mL of each concentration (crude extracts or fractions) was individually sprayed on the bees throughout the plastic lid using a hand sprayer.A device with candy and water was placed inside each unit as food for the bees.Fifteen bees in a modified Petri dish sprayed with DMSO were included as negative control, and six bees in a modified Petri dish sprayed with 0.07% Deltamethrin (DTT) (Pirisa-PiretroIndustrial Ltda, Brazil) were included as positive death control.Four replicates for each experimental group were run.Bioassay dishes were placed in incubators at 28 ± 1°C and 60% relative humidity.Mortality of bees was evaluated daily, by visual inspection for a period of 15 days.

STATISTICAL ANALYSIS
Differences in survival after 15 days of observation were assessed by Kaplan-Meier analysis followed by the Logrank test.A p value < 0.05 was considered statistically significant.All statistical analyses were performed with the software package GraphPadPrism 4.00 for Windows (GraphPad Software, San Diego, CA, USA).

P. larvae SUSCEPTIBILITY TEST AND DETERMINATION OF MICS
The P. larva was susceptible to the assessed crude extract of the C. officinalis and N. officinale.For the C. domestica, only the EtAc and BuOH fractions indicated activity against P. larvae.The MICs of these samples ranged from 0.98-30.51mg/mL, and the crude extract and DCM fraction of C. domestica did not show activity against P. larvae (Table I).EtAc fraction (Fig. 1).Bee mortality was evident in treatment with DTT (positive death control group).

DISCUSSION
The search for the discovery of drugs, as well as the determination of the prudent use of antimicrobial agents, are the basis to solve the global problem of microbial resistance; and the resources for the discovery of drugs are the natural products (Santos et al. 2012).I) exhibited an antibacterial effect against P. larvae, with respective MICs of 12.76 mg/mL and 30.51mg/mL.Therefore these two species were not fractionated.The C. domestica crude extract did not show such activity, for this reason, its aqueous extract was fractionated using solvents of increasing polarity, with the objective of concentrating molecules with similar polarity in the same fractions and, consequently, promote better activities, as observed in Table I, in which low MIC values were found for AcEt and BuOH fractions of C. domestica.
One of the most usual problems in the use of medicinal plants is the amount of extract used.In this case the MIC values were between 0.98 and 30.51 mg/mL, however this is not a serious problem, since the herbal products are relatively safer than synthetic drugs (Abubakar 2010).The MIC values obtained for the crude plant extracts were high when compared with the MIC values of 0.01 -10 μg/mL, frequently recorded for conventional antibiotics.The differences can be due to the fact that synthetic antibiotics are in a pure form, plants extracts don't act the same way as synthetic antibiotics, that's why were found high MIC values (Abubakar 2010).
To do the spraying on A. mellifera extracts or fractions that have shown activity against P. larvae were used in exactly the same concentration, in order to verify, by the percentage of survival, the possible toxic effects (Fig. 1).
N. officinale crude extract and C. domestica BuOH fraction showed 100% survival in all periods analyzed, demonstrating no toxic effects at the concentrations tested.
The C. officinalis crude extract and C. domestica EtAc fraction showed, respectively, 66.68% and 26.64 % of survival in the concentration tested at the end of the experiment.
In a similar study performed by Janovik et al. (2009), the C. domestica showed that in the BuOH fractions had 486.22 mg/g of polyphenols and 15.26mg/g of flavonoids, and in the EtAc fraction had 510 mg/g of polyphenols and 39.92 mg/g of flavonoids.In this same study the EtAc fraction also presented: gallic acid (2.5 mg/g), caffeic acid 12.82 (mg/g), chlorogenic acid (15.27 mg/g), rutin (11.45), and kanferol (0.85 mg/g).
A study performed by Ducat et al. (2011) showed that this species contains, 92.35 mg/g of phenolic acid and 13.55 mg/g of flavonoids in a 70% ethanolic extract, and also showed an antimicrobial activity for Staphylococcus aureus and Klebsiella pneumoniae.Other study by Efstratios et al. (2012) showed antifungal and antibacterial activity against Bacillus subtilis, Escherichia coli, Enterococcus faecalis, Pseudomonas aeruginosa and Klebsiella pneumoniae.
It has been proposed that the antimicrobial activity could be due to synergism between the various components.It was observed that none of the single component showed a higher activity than the total extract (Bonvehi et al. 1994, Flesar et al. 2010, Boligon et al. 2013).
Similarly, the antibacterial activity of the propolis extract could be related with its chemical composition, which includes phenolic compounds (flavonoids and aromatic acids), terpenes and essential oils among others (Sforcin 2007, Antúnez et al. 2008).
A very similar study of Boligon et al. ( 2013) evaluated the antimicrobial susceptibility of the P. larvae to the crude extract, dichloromethane , EtAc and BuOH fractions of the S. buxifolia, where the MICs were 50, 1.56, 6.25, 25 mg/mL, respectively.The antimicrobial activity was related to the chemical composition of the S. buxifolia which is rich in steroids, triterpenes, phenolic compounds, flavonoids and alkaloids (Boligon et al. 2013).In the same study, in the toxicity assay by spraying, S. buxifolia showed not be toxic for bees during 15 days.
However, the same did not occur with C. officinalis, this species showed an acute toxicity and growth inhibition in milkweed bug, Oncopeltus fasciatus Dallas.Also, Calendula micrantha officinalis showed molluscicidal activity (Abd-El-Megeed 1999).These important activities can explain the mortality of bees with C. offinalis crude extract in this study.Janovik et al. (2009) isolated lupeol, β-amyrin, β-sitosterol and stigmasterol fraction of C. domestica EtAc and also the same compounds of Inula japonica, which showed acaricidal effect in vitro against carmine spider mite, Tetranychus cinnabarinus (Boisduval) in a study performed by Duan et al. (2011), this may explain the reason for the higher mortality in bees, even containing a large amount of polyphenols in this fraction.
Similarly, in a study conducted by Santos et al. ( 2012) with the sprinkling of Carapa guaianensis (andiroba) oil, only 20% of bee survival was observed after 10 days, the insecticidal effect of andiroba oil was associated with this result.
DETERMINATION OF THE MINIMUM INHIBITORYCONCENTRATIONThe minimum inhibitory concentrations (MICs) of C. officinalis, N. officinale and C. domestica crude extracts and C. domestica fractions were determined by microdilution techniques in Mueller-Hinton broth (Difco) for P. larvae(CLSI 2008).The assay was carried out in 96-well microtitre plates.Each sample was mixed with an inoculum prepared in the same medium at a density adjusted per tube to 0.5 of the McFarland scale (1.5 x 10 8 CFU/mL) and diluted 1:10 for the broth microdilution procedure.Microtitre trays were incubated at 37°C and the MICs were recorded after 24h of incubation.The 1043 GROWTH-INHIBITORY PLANTS AGAINST P. larvae

1044
MARIANA PIANA et al.The crude extracts of three Brazilian species were investigated: N. officinale, C. officinalis and C. domestica.The determination of the MIC through the method of broth micro dilution showed that only the crude extract of N. officinale and C. officinalis (Table

TABLE I MICs of C. officinalis, N. officinale and C. domestica on P. larvae.
TOXICITY ASSAY IN BEESToxicity analysis in bees, evaluated by the spraying application method of C. officinalis, N. officinale crude extracts, and EtAc, BuOH fractions of C. domestica, showed no toxicity for N. officinale crude extract and C. domestica BuOH fraction during 15 days of treatment, however, some deaths of bees occurred during the first three days of treatment with C. officinalis and C. domestica