In Vitro Antagonistic Effect of Gut Bacteriota Isolated from Indigenous Honey Bees and Essential Oils against Paenibacillus Larvae

The aim of study was to isolate and identify the gut bacteria of Apis mellifera and to evaluate antagonistic effect of the bacteriota against Paenibacillus larvae, which causes American foulbrood (AFB) in honeybees. The dilution plating method was used for the quantification of selected microbial groups from digestive tract of bees, with an emphasis on the bacteriota of the bees’ intestines. Bacteria were identified using mass spectrometry (MALDI-TOF-MS Biotyper). Overall, five classes, 27 genera and 66 species of bacteria were identified. Genera Lactobacillus (10 species) and Bacillus (8 species) were the most abundant. Gram-negative bacteria were represented with 16 genera, whereas Gram-positive with 10 genera. Delftia acidovorans and Escherichia coli were the most abundant in the digestive tract of honey bee. Resistance to a selection of antimicrobials was assessed for the bacterial isolates from bee gut and confirmed against all antimicrobials included in the study, with the exception of cefepime. Lactobacillus spp., especially L. kunkeei, L. crispatus and L. acidophilus. showed the strongest antimicrobial activity against P. larvae, the causal pathogen of AFB. Antimicrobial activity of essential oils against isolated bacteria and two isolates of P. larvae were assessed. Application of a broad selection of plant essential oils indicated that Thymus vulgaris had the highest antimicrobial activity against P. larvae.


Introduction
The digestive tract of the worker bee is inhabited with a variety of microorganisms diverse in their morphology, physiology and metabolism. The microbiota of digestive tract consists of yeasts (1%), Gram-positive bacteria (29%) and Gram-negative and gram-variable bacteria (70%) [1]. The first research on microbiota of digestive tract of bees had been published in the beginning of the 20th Table 1. Isolated bacteriota of adult worker honeybee guts in in log cfu/g (mean ± SD). 2.52 ± 0.11 a 3.25 ± 0.13 ab 3.57 ± 0.13 abc 3.37 ± 0.14 ac * TCAM-total counts of aerobic microorganisms, TCANM-total counts of anaerobic microorganisms, AG + -anaerobic Gram-positive bacteria, BS-Bacillus spp., LS-Lactobacillus spp., PS-Pseudomonas spp., ES-Enterococcus spp., SS-Staphylococcus spp., CB-coliform bacteria. a,b,c same letters in the raw show statistically significant differences among the groups.

Isolated Bacteria from Bees Gut
A total of five classes of bacteria were obtained from the gut of the honey bee: Actinobacteria, Alphaproteobacteria, Betaproteobacteria, Firmicutes and Gammaproteobacteria. A total of 27 genera were isolated from the honey bee bacteriota: Aeromonas, Arthrobacter, Bacillus, Citrobacter, Delftia, Enterobacter, Enterococcus, Escherichia, Fructobacillus, Hafnia, Klebsiella, Kocuria, Lactobacillus, Lactococcus, Microbacterium, Moraxella, Morganella, Paenibacillus, Pantotea, Proteus, Pseudomonas, Rahnella, Ralstonia, Raoultella, Serratia, Sphingomonas and Staphylococcus. A total of 66 species were isolated from bees, of which the genus Lactobacillus represented by 10 species and the genus Bacillus by eight species were the most numerous (Table 2).  A total of 66 species of bacteria from the digestive tract of bees were isolated, of which 33 were Gram-positive and 33 Gram-negative. Escherichia coli was isolated most frequently from all samples tested, but P. larvae was isolated from only one sample (Table 3).

Antibiotic Resistance of A. mellifera Gut Bacteriota
A total of 5789 isolates were isolated from the digestive tract of 200 bees. Gram-positive and Gram-negative bacteria showed antimicrobial resistance to various classes of antimicrobials (Table 4). Table 4. Antimicrobial resistance of bacteria isolated from bee digestive tracts.

Antimicrobial Activity of Isolated Bee Digestive Tract Bacteriome against P. larvae
The interactions between intestinal bacteria and pathogens of A. mellifera, in particular the action of intestinal bacteria against P. larvae, are an area of great research interest. Research on microbial composition of digestive tract of A. mellifera are perspective from the bee's health point of view. The research on antagonisms of P. larvae may promote the development of bee-friendly compounds, to protect the bees from infection with pathogens.
All microorganisms tested showed antimicrobial activity against P. larvae. The strongest antimicrobial activity was shown by Lactobacillus, whereas the weakest was typical for Enterobacteriaceae (Table 5). Among the species analyzed, L. kunkei, L. crispatus, L. acidophilus were the most active against P. larvae. Klebsiella variicola, Ralstonia picketii, Pantotea agglomerans, Pa. vagans and Serratia liquefaciens were less active against P. larvae isolated from bee intestines. The strongest antimicrobial activity of L. kunkei, L. acidophilus and L. crispatus and the weakest antimicrobial activity of Pa. ananatis and Rahnella aquatilis were found against P. larvae CCM 4483. Table 5. Antimicrobial activity of individual isolates against P. larvae in mm (mean ± SD of three replicates).

Antimicrobial Activity of Essential Oils against P. larvae
The next aim of the work was to determine the antimicrobial activity of essential oils against two strains of P. larvae. The highest antimicrobial activity (

Discussion
The highest counts of the aerobic microorganisms, Bacillus spp., Lactobacillus spp. and coliform bacteria were found in the intestine of winter bees and the lowest in the rectum of summer bees. Similar results of bacterial counts have been reported previously [10][11][12][13]. The microbiome of bees represents not only the microorganisms present in the adult worker bees, but also reflects the hive microbiota. The origin of hive microorganisms are nectar, pollen, dust and other airborne and soilborne environmental contaminants [12][13][14]. The excrement of honey bees and animals could be a source of microbiota during nectar harvesting. A wide variation in bacteria associated with bees have been ascribed to the external environment [15]. The bacteriota of the digestive tract of the Japanese eastern bee (Apis cerana japonica) revealed that Bacillus species could be potential antagonists for biologic control of P. larvae [16].
Non-culture studies of bee microbiome were conducted on the digestive tract or only on the middle and posterior parts of the intestines [17][18][19][20][21][22][23][24][25] and revealed that the pollination-based environmental microbiota and the four nectar-bearing ones are an important source of the beneficiary and potentially beneficiary microorganisms for bees [26][27][28]. Lactobacillus spp. were frequently found in the bee intestines and were considered the most important genus of lactic acid bacteria (LAB) in promoting animal and human health [11,[29][30][31]. Lactobacillus spp. play significant role in feed digestibility in animals and they are important for functioning of gastrointestinal tract and accompanied immunological responses [32][33][34][35][36][37]. In our study, we did not identify species from the Bifidobacterium genus.
Antimicrobial resistance of the bacterial isolates varied in our study, depending on the genus and strain properties. Kačániová et al. [38] found resistance to tigecycline (12.5%) and amikacin (18.2%), gentamicin (9.5%) and chloramphenicol (7.2%) in their bacteriome of honey bees. Administration of antimicrobials triggers changes in the microbiome of humans and livestock, therefore, assessment of the effect of the antimicrobials on bee intestinal microorganisms is important for their health prognosis [23,24,39,40] and a possible explanation of unexpected bee colony deaths [41]. The studies on microbiome diversity and its antimicrobial resistance can provide an overview on nutritional and health problems of honey bees [42].
American foulbrood (AFB) is the most destructive bacterial disease of honey bee larvae [43]. AFB is a contagious infection that begins in an individual bee larva and can cause the collapse of the entire colony because only a few spores of P. larvae are necessary to initiate the disease [44].
The use of antimicrobials, especially oxytetracycline, could protect the bees hives against infection, however, P. larvae resistance to oxytetracycline has been identified in the USA, Argentina and Canada [5,45]. Use of antimicrobials in beekeeping poses a serious risk to human health as their residues may persist in honey and other bee products [46]. Adverse effects of application of antimicrobials on the honey of honey bees [47] and on the beneficial intestinal bacteria [48] have been described.
The biologic control of AFB pathogen is considered an environmentally conscious and bee-friendly perspective. Evans and Armstrong [49,50] found that certain intestinal bacteria of A. mellifera showed antagonistic activity against P. larvae. Eastern Japanese bee (Apis cerana japonica), native to Japan, exhibited resistance against parasitic and microbial pathogens, including mite and AFB pathogen [51]. The antagonistic effect of bacteria may also depend on bacterial communities present or strains properties, including production of antimicrobial substances, e.g., bacteriocins and lysozyme and changes in pH as a result of organic acids production [52]. Bacteria with antagonistic properties enhance control or inhibition of pathogens. Bacillus spp. were found to exhibit bactericidal and fungicidal effects in the host gut as a result of production of various antimicrobial compounds [53,54]. Apis mellifera jemenitica was shown as biologically better adapted to harsh environment with higher productivity [55,56].
Several natural compounds were studied for antagonistic activity against P. larvae in vitro [57][58][59], however, the identified cytotoxic effects on bees had limited their practical application. Alternatives, such as prevention and control methods of the AFB pathogen are an area of great interest. Since the ancient times, the herbal medicine and herbal extracts were applied for treatment of human and animal diseases [60]. Biologically active compounds of honey, propolis, essential oils, agents from spore of bacteria of honey and fungal extract of pollen were tested against AFB pathogen [61][62][63][64][65]. Of these, essential oils showed the strongest antibacterial activity against microorganisms responsible for bee diseases without toxicity on bees in vitro. The main complication in those studies is to obtain the results applicable to beekeeping related to the antimicrobial activity of essential oils and their effect on bees [66,67]. In our study, Thymus vulgaris was the most effective essential oil against both species of P. larvae, whereas the most effective essential oils against P. larvae CCM4483 were those from Pinus silvestris and Abies alba.
Tests of Melaleuca viridiflora and Cymbopogon nardus essential oils against P. larvae have shown an inhibition at 320 mg/L in vitro [68]. Almost all essential oils of Achyrocline satureioides, Chenopodium ambrosioide, Eucalyptus cinerea, Gnaphalium gaudichaudianum, Lippia turbinata, Marrubium vulgare, Minthostachys verticillata, Origanum vulgare, Tagetes minuta and Thymus vulgaris were effective against P. larvae strains. Eucalyptus cinerea and M. verticillata essential oils exhibited 100% efficiency in inhibiting the growth of all P. larvae strains [69]. Essential oils of Schinus molle var. areira L., Acantholippia seriphioides A. Gray, Mintosthachys mollis, Tagetes minuta L. and Lippia turbinata Griseb grown in wild in Argentina shared minimum and maximum MIC and MBC values of 200-250 mg/L and 200-300 mg/L for Andean thyme and 800-1000 mg/L and 850-1100 mg/L. Andean thyme has been shown to be the most effective in vitro against P. larvae and could be a perspective natural alternative to the traditional antimicrobial treatment of AFB pathogen [61].

Samples of Bees
A total of 200 samples of Apis mellifera carnica workers were examined. Samples of bees were taken from hives from the eastern Slovakia in the Košice area (48.7164 • N, 21.2611 • E). Bees were sampled in winter and summer, with samples from the digestive tract (intestines and rectum). examined separately. Workers of honey bees were anesthetized on ice and washed in 86% ethanol before dissection. The head or thorax of a honeybee was fixed and the entire intestine was removed by pulling the stinger using sterile dissecting forceps. The intestines and rectum were separated and collected into sterile, separate microcentrifuge tubes.
The basic dilution (10 −2 ) was obtained by homogenizing 0.1 g of the digestive tract contents of five bees and 9.9 mL of peptone saline (0.89%). Selection for groups of microorganisms followed as shown in Table 7. All agars were purchased from Oxoid (Basingstoke, United Kingdom).

Identification of Bacteria
Identification of bacteriota was performed using MALDI-TOF-MS Biotyper (Bruker Daltonics, Bremen, Germany). All the preparatory stages for the samples were carried out according to the MALDI-TOF-MS Biotyper manufacturer's recommendations. Bacterial colonies were transferred into 300 µL of distilled water and 900 µL of ethanol in Eppendorf tubes, which were centrifuged for 2 min at 14,000 rpm. The supernatant was removed, and centrifugation was repeated for the pellet, which was subsequently allowed to dry. Ten microliters of 70% formic acid and 10 µL of acetonitrile were added to the dried pellet. Tubes were centrifuged for 2 min at 14,000 rpm and 1 µL of the supernatant was applied for identification with the MALDI-TOF. Matrix, α-cyano-4-hydroxycinnamic acid in a volume of 1 µL, was added to that 1 µL of supernatant and allowed to dry. The analysis was performed with a Microflex LT (Bruker Daltonics, Bremen, Germany) instrument and Flex Control 3.4 software and Biotyper Realtime Classification 3.1 with BC specific software. Confidence scores of ≥2.0 and ≥1.7 were the criteria for successful identification at the levels of species and genus, respectively [70].
For Lactobacillus spp. strains, the MICs (µg/mL) of AMP, MER, IMI and CHL were evaluated using the commercial E-test ® (Oxoid, Basingstoke, UK). The concentrations of antimicrobials ranged from 0.016 to 256 µg/mL. Bacterial cultures in exponential growth phase were adjusted to a suitable turbidity (10 6 to 10 7 CFU/mL) and used for inoculation of iso-sensitized agar (90% w/v, Oxoid, UK) supplemented with main Rogosa agar (MRS) or TPY agar (10% w/v) (Oxoid, Basingstoke, UK). E-test strips were placed on the surface of the inoculated agar and incubated at 37 • C for 24 h microaerophilically. The MIC test result was interpreted as the point at which the ellipse intersected the E-test strip as described in the E-test technical guide.

Antimicrobial Activity of Bacterial Suspensions against P. larvae
Bacterial strains after 24 h of incubation on MRS and tryptone soya agar (TSA) medium were centrifuged at 5500× g for 10 min at 4 • C and 0.1 mL of the supernatant was used for detection of activity against P. larvae. A suspension (0.1 mL, 10 5 CFU/mL) was plated on Mueller-Hinton agar. Filter paper discs (6 mm diameter) were impregnated with 15 µL of supernatant from each bacteria and placed on the P. larvae-inoculated agar. The agars were incubated initially at 4 • C for 2 h and then at 37 • C for 16 h. All tests were performed in triplicate. Filter discs impregnated with 10 µL of distilled water were used as a negative control and antibiotics (amikacin, 10 µg and gentamicin, 10 µg) were used as a positive control [71]. Two P. larvae isolates were tested in this study: one isolate was from bee hive and second isolate was purchased (P. larvae CCM 4483) from the Czech collection of microorganisms (Brno, Czech Republic).

Statistical Analyses
All measurements were made in triplicate. Statistical processing of data of the bacterial counts was performed using Microsoft Excel ® software. Bacterial counts and measurements of inhibition zones were expressed as the means and standard deviation (SD). Student's t-test was used for calculation of significance of variability in distribution of bacteria among seasons as well as among different parts of bee gut for individual groups of analysed microorganisms. Significance of the results was considered at the following thresholds: p ≤ 0.05, p ≤ 0.01, p ≤ 0.001.

Conclusions
Understanding of bacteriome inhabiting the intestine of bees has a potential to help beekeepers and promote bee health. Apis mellifera is the most important pollinator insect in means of global food security. Our studies on characterization and functional role of the bee's intestinal microbiota reveal the unique properties of A. mellifera bacteriota. EU prohibited antibiotics in beekeeping practice and P. larvae after antibiotics treatments can develop resistance. Natural antimicrobials as probiotic bacteria and essential oils can play the biggest role in control of bee pathogens.
The antimicrobials may cause an alteration in bee gut microbiota so the studies of beneficiary intestinal bacteria, which may increase colony resistance to various bee's pathogens, is a promising alternative to bee's antimicrobial treatment. Essential oils showed the inhibitory effect on P. larvae isolated from bees, so the application of essential oils may be expanded in beekeeping. Therefore, the present results on the antimicrobial activity of bee-beneficial bacteria and essential oils from plants can help increase the beekeepers' awareness of these possibilities and possibly reduce bee colony mortality on a global scale.