IntroductionTop

American foulbrood (AFB) is the most severe bacterial disease that affects honey bees, having a nearly cosmopolitan distribution (Genersch, 2010). AFB only kills infected honey bee larvae; however, it eventually leads to the collapse of the entire colony when left untreated. AFB is considered to be very contagious; therefore, it is a notifiable disease in most countries (Djukic et al., 2014). AFB’s causative agent is Paenibacillus larvae, a flagellated gram-positive bacterium, whose main characteristic is the formation of highly resistant endospores. Four P. larvae genoty­pes (Enterobacterial Repetitive Intergenic Consensus, ERIC I-IV) have been described, but only two are commonly detected under field conditions (ERIC I and II) (Genersch & Otten, 2003; Alippi et al., 2004).

In some countries the use of antibiotics, particularly oxytetracycline hydrochloride (OTC) is recommended for the prevention and treatment of infected colo­nies (Hansen & Brødsgaard, 1999; Genersch, 2010). However, in most European countries the use of antibiotics is banned, since chemicals can remain in honey and other bee products affecting their quality for human consumption. Moreover, antibiotic application can affect bee lifespan and can increase the risk of generation of bacterial resistant strains (Martel et al., 2006). To date, the presence of OTC resistant strains has been reported in Argentina, USA, Italy, New Zealand and UK (Alippi et al., 1996; Miyagi et al., 2000).

In this context, the development of alternative and effective methods for the prevention and control of AFB disease is crucial. A promissory strategy is the use of antimicrobial natural bioactive substances. Several studies have been carried out to identify natural mo­lecules with antibacterial activity against P. larvae. Those studies included the evaluation of plant extracts (Flesar et al., 2010; Sabaté et al., 2012; Boligon et al., 2013; Damiani et al., 2014; Chaimanee et al., 2017); essential oils (Alippi et al., 1996; Fuselli et al., 2006; Ansari et al., 2016, Tutun et al., 2018); pure compounds extracted from plants, bacteria or fungus (Lokvam et al., 2000; Fuselli et al., 2006; Flesar et al., 2010; Sabaté et al., 2012); and honey bee by-products, such as propolis (Antúnez et al., 2008; Biliková et al., 2013; Isidorov et al., 2017) and royal jelly (Biliková et al., 2013).

Natural molecules for the control and prevention of AFB are a heterogeneous group that includes a number of compounds which share one common characteristic: they are pure natural substances; either commercial, such as fatty acids or that has been obtained from natural sources, such as fungal strains or plants. Shi­manuki et al. (1992) reported the inhibition of AFB by an ethanolic extract of chalk brood mummies (Apis mellifera larvae infected by Ascosphaera apis fungus). This fact led scientists to seek the identity of the active compound.

On the other hand, Hornitzky (2003) evaluated the antibacterial activity of fifteen fatty acids on European foulbrood, most exhibited activity at the highest tested concentration (250 μg), except for myristoleic and lauric acids, which also showed reduced activity with diffusion disks containing 25 μg of the molecule.

The aim of this study was to evaluate the poten­tial bactericidal activity of natural molecules such as saturated and unsaturated fatty acids, flavonoids, phenolics acid, vitamins and organic acids against P. larvae. Moreover, we investigated whether these mo­lecules that exhibit antimicrobial activity were able to inhibit the pro­teolytic activity of the bacterium.

Material and methodsTop

Chemicals

Mueller-Hinton broth, agar, brain-heart infusion, yeast extract, glucose and peptone were provided by Britania S.A. (Buenos Aires, Argentina); NaCl (99.99%) by Alun (San Martin, Buenos Aires); K2HPO4 (98%) by Cicarelli (San Lorenzo, Santa Fe, Argentina); sodium pyruvate (≥ 99%) by Biopack (Zárate, Buenos Aires); acetonitrile, ethanol and methanol (HPLC grade), glacial acetic acid (ACS reagent), n-butanol (FCC reagent), resazurin sodium salt (forcell culture), lauric acid, palmitic acid, lauric acid monoglyceride, ellagic acid, cinnamic acid, syringic acid, naringe­nin, menadione, alpha-ketoglutaric acid, salicylic acid, phlo­ridzin dihydrate, and chlorogenic acid by Sigma-Aldrich (Saint Louis, MO, USA); dimethyl sulfoxide (analytical grade) by Biopack (Zarate, Buenos Aires); stearic acid by Materia (Mar del Plata, Argentina) and oxytetracycline hydrochloride by Bayer (Buenos Aires). Distilled water was sterilized in autoclave FAC (Buenos Aires). Sterilized polystyrene 96-well culture plates were supplied by Deltalab (Barcelona, Spain).

Biological material

Paenibacillus larvae isolates were obtained from honey combs of bee hives exhibiting clinical symptoms of AFB, from Argentina and Italy. Isolates S1 and S2 were from Bal­­­carce, Buenos Aires province (37° 52’00″S-58°15’00″W), strain S3 from Rio Cuarto, Cordoba province (33°08′00″S-64°21′00″W), strain S4 from Concordia, Entre Rios province (31°

23′32″S-58°01′01″W), strain S5 and S8 from Ne­cochea, Buenos Aires (38°33′44″S-58°44′43″W), strain S6 and S7 from Mar del Plata, Buenos Aires (38°00′00″S-57°33′00″W), strain S9 from Modena, Italy (44°38′45″N-10°55′33″E) and strain S10 from Emilia Reggio, Italy (44°42′00″N-10°38′00″E).

Bacterial isolates were grown and maintained on 2% (w/v) MYPGP agar plates (1% Mueller-Hinton broth (w/v)), 1.5% yeast extract, 0.3% K2HPO4, 0.2% glucose, 0.1% sodium pyruvate and 2% agar (w/v), incubated at 37 °C and 10% (v/v) O2 for 48 h (Dingman & Stahly, 1983; Nordström & Fries, 1995). The bacterial inocula were prepared in sterile peptone water (0.1%, peptone w/v) and 0.85% NaCl (w/v) to a final optical density at 600 nm OD600 = 0.1, using a UV-VIS spectrophotometer Spectrum SP-1103 (Spectrum Instruments Company Ltd., Shanghai, China) (Nordström & Fries, 1995). Brain-heart infusion (3.7%, w/v) was used as a growth medium for the bacterial isolates when performing the broth microdilution assay. P. larvae growth was detected using resazurin sodium salt.

Identification and genotyping of P. larvae

Total DNA of the P. larvae isolates was prepared from overnight cultures of the bacterial isolates grown in J medium (Hornitzky & Nicholls, 1993). DNA was obtained using a commercial genomic DNA purification Kit (Sigma). The DNA concentration and purity were checked using a NanoDrop 2000 spectrometer 169 (Thermo Fisher Scientific Com., Waltham, USA). Bacterial isolates identification was carried out by amplification of a specific 16S rRNA gene fragment of P. larvae using primers PL5 and PL4 (Piccini et al., 2002).

ERIC genotyping of bacterial isolates was per­formed using primers designed by Versalovic et al. (1994) and conditions optimized by Antúnez et al. (2007). The PCR reactions were carried out by triplicate. The PCR reactions were carried out with 1 U Taq DNA polymerase (Invitrogen), 100 ng of DNA, 0.0002 mol/L of each of the four dNTPs, 0.005 mol/L MgCl2 and 0.003 mol/L of each primer in a total volume of 25 μL. The PCR conditions were as follows: a single denaturation step at 94 ºC for 5 min, 40 cycles of denaturation of the DNA template at 94 ºC for 1 min, annealing of primers at 40 ºC for 2 min and extension of PCR products at 65 ºC for 8 min. A final extension step was performed at 65 ºC for 16 min. Amplified products and two DNA markers, GeneRuler DNA Ladders 100 bp and GeneRuler DNA Laders 1 kb Thermo Fisher Scientific Co., were separated in a 0.8% agarose gel and stained with Gel Red (Olerup 187 SSP). The gels were photographed under UV light.

Screening of the antimicrobial activity by agar diffusion method

Screening of the antimicrobial activity of thirteen molecules against four P. larvae isolates was conducted by the agar diffusion technique (Bonev et al., 2008). The natural molecules tested were: saturated and unsaturated fatty acids (lauric acid, palmitic acid, lauric acid monoglyceride, alpha-ketoglutaric acid and stearic acid); phenolics acids (salycilic acid, syringic acid, ellagic acid and chlorogenic acid); flavonoids (naringenin and phloridzin); a vitamin (menadione) and an organic acid (cinamic acid). Each paper disc contained 200 µg of the molecule tested. Oxytetracycline hydrochloride (30 µg) was used as positive control. The natural molecules that showed a strong activity against P. larvae isolates by this method were selected for further analysis.

Determination of the minimum inhibitory con­centration of the antimicrobial agent

The antimicrobial activity of selected molecules was evaluated by broth microdilution method (Cugnata et al., 2017) using ten P. larvae isolates. Two parameters were estimated, the minimum inhibitory concentration (MIC), the concentration at which in vitro bacterial growth inhibition is observed; and the minimum non-inhibitory concentration (MNIC), the concentration at which in vitro bacterial growth inhibition is not observed. The repeatability of the method within one day (n=3) and between days (n=3) was determined for each molecule.

Determination of protease activity

Paenibacillus larvae isolates were grown in J Me­dium (Hornitzky & Nicholls, 1993) aerobically (150 rpm) at 37 °C until the stationary growth phase (72 h, OD600 > 1.5). Identical culture volumes (1 mL) were centrifuged for 10 min at 10,000 rpm and the cell-free supernatants were collected. One aliquot was examined on agar plates containing casein to verify the presence of protease activity. The remaining volume (750 µL) was concentrated 4-6x in Centricon filtration units (YM-10) and conserved at 4 °C until analysis.

― Gelatin zymography. The cell free supernatants were electrophoresed on polyacrylamide gels (10%, w/v) containing SDS 1% and copolymerized gelatin (without reducing agents and heating) using a mo­dification of the protocol from Laemmli (1970). After electrophoresis, SDS was removed from the gel by washing in 0.05 mol/L Tris-ClH (pH 7.5), and then incubated in the same buffer in absence or presence of the metalloprotease inhibitor ethylenediamine­tetraa­cetic acid (EDTA) (0.005 mol/L) 37 °C for 3 h. Finally, the gel was stained with Coomasie Brilliant Blue to evidence the bands with protease activity.

― Azocaseinolytic assay. The reaction mix contained 0.05 mol/L Tris-ClH (pH 7.5), 0.5% azocasein (in 0.025 mol/L NaOH) and 0.1 mL cell free supernatants, as a source of enzyme, in a final volume of 0.5 mL. The samples were incubated at 37 °C for different times and the reaction was stopped by addition of cold trichloroacetic acid (TCA) 10% (v/v); the tubes were left on ice for 15 min and centrifuged at 2000 rpm for 10 min. To optimize the assay conditions, TCA soluble products absorbance was measured at 335nm on a UV-visible spectrophotometer (GeneQuant 1300 Spectrophotometer, Classic from GE Healthcare Life Sciences, USA). The reaction was linear for at least 3 h, thus, 2 h was chosen as incubation time. Samples were incubated in absence and presence of EDTA (0.005-0.01mol/L) and phenylmethanesulfonyl fluoride (PMSF) (0.001 mol/L).

Results and discussionTop

In this study, we evaluated the antimicrobial activity of different molecules against P. larvae. Firstly, we generated a collection of ten P. larvae isolates obtained from colonies with clinical signs of AFB from Argentina and Italy. A unique amplicon of 700 bp characteristic of P. larvae was obtained confirming the identity of the bacterial strain isolates S1, S2, S3, S4, S5, S6, S7, S8, S9 and S10. Regarding isolates genotyping using ERIC primers, all P. larvae isolates showed the same DNA pattern, regardless their geographical origin. Based on comparison between the obtained DNA fingerprints, those previously published (Genersch & Otten, 2003; Genersch et al., 2006) and reference strains (Alippi & Aguilar, 1998; Antúnez et al., 2007), the isolates (S1to S10) were classified as ERIC I genotype. This is the most commonly detected genotype and is worldwide distributed (Genersch & Otten, 2003; Alippi et al., 2004; Antúnez et al., 2007).

Then, we evaluated the antimicrobial activity of thirteen molecules by the agar diffusion method using four P. larvae isolates (S1, S2, S9 and S10). Only four of the molecules (menadione, lauric acid, monoglyceride of lauric acid and naringenin) showed inhibition of the bacteria. The inhibitory halo decreased in the following order: menadione, lauric acid, monoglyceride of lauric acid and naringenin (Table 1).

MIC of menadione and lauric acid was evaluated by the broth microdilution method, using ten P. larvae isolates. Table 2 shows the results of the MIC values obtained, which ranged from 0.78 to 3.125 µg/mL for menadione, and from 25 to 50 µg/mL for lauric acid.

Feldlaufer et al. (1993) had already found that lauric and tridecanoic acids (isolated from mycelia and spores of Ascosphaera apis) were the most active saturated fatty acids against P. larvae (2.5 µg per disc), while palmitoleic and linoleic acids were the most active among the unsaturated ones. These authors demonstrated that the introduction of a double bond or multiple double bonds seemed to be necessary to maintain the antibiotic activity once the chain length of the fatty acid exceeds fourteen carbons. Furthermore, Hornitzky (2003) showed that fifteen fatty acids proved to have antibacterial activity against European foulbrood. The majority of the tested acids exhibited activity at the highest amount tested (250 μg), except for myristoleic and lauric acids, which showed activity even with diffusion disks containing 25 μg of the molecule. In our study, lauric acid also showed a good antimicrobial activity; however, stearic and palmitic acids did not present any antimicrobial activity.

Flesar et al. (2010) analyzed a total of 26 natural compounds of different chemical classes and 19 crude extracts by the broth microdilution method. The MICs, ranging from 2 to 256 µg/mL, showed that 13 compounds exhibited an antimicrobial effect against P. larvae. Out of all plant-derived products tested, the greatest antibacterial activity against P. larvae was observed for sanguinarine with a MIC = 4 µg/mL. This molecule presents the most compact spatial conformation of all the analyzed molecules. In this sense, the molecular structure of menadione is similar to the sanguinarine molecule (both present aromatic rings); this spatial conformation seems to confer the best activity against microorganisms among the molecules studied in the present work, and may explain the antibacterial activity against P. larvae of menadione.

According to Michielin et al. (2009), plant materials can be classified as antimicrobial agents based on the MIC values of its extracts. In this sense, Duarte et al. (2007) and Wang et al. (2008) classified the extracts as: strong inhibitors for MIC value below 500 µg/mL; moderate inhibitors for MIC between 600 and 1500 µg/mL; weak inhibitors for MIC above 1600 µg/mL. In consonance with this classification and the MIC values obtained for menadione and lauric acid in the present study, these molecules are strong inhibitors of P. larvae.

In order to deep-in in the study of the antimicrobial activity of those molecules, we investigated its po­tential inhibitory effect on the proteolytic activity of P. larvae. Previous reports showed that P. larvae secrets metalloproteases which are considered virulence factors of this microorganism (Dancer & Chantawannakul, 1997; Antúnez et al., 2009). Then, we first confirmed the capacity of the P. larvae isolates obtain in this study to produce these proteins. Protease activity was maximal in the stationary growth phase (60-70 h). Then, the proteolytic profile was analyzed by gelatin zymography. All the isolates presented proteolytic activity, showing the same pattern with at least three gelatinolytic bands ranging from 20 to 38 kDa (Fig. 1). The intensity of these bands was reduced by incubation with EDTA, indicating the presence of metalloprotease/s, in agree­ment with Antúnez et al. (2009). This result was con­firmed with the azocaseinolytic assay, which showed 80% inhibition of the protease activity in presence of EDTA and no inhibition with the serine protease inhibitor PMSF. To explore the potential inhibitory effect of natural molecules on the metalloproteases secreted by P. larvae, we assayed the extracellular azocaseinolytic activity of two isolates in presence of menadione (0.15, 1.5 and 3 µg/mL) and lauric acid (50 µg/mL). However, no inhibition of the protease activity was observed under the concentrations tested.

Figure 1. Extracellular proteolytic profile of P. larvae isolates. Cultures of ten P. larvae isolates were grown in J medium until the stationary phase (OD600 = 1.5-2) and the cell-free supernatants were concentrated and examined by gelatin zymography.

In conclusion, the present study shows that the natural molecules menadione and lauric acid exhibited in vitro bactericidal activity against Paenibacillus larvae, at concentrations feasible to be applied in the field. The use of non-toxic compounds represents a natural alternative to synthetic antibiotics for the control of AFB, reducing considerably the controversies of antibiotic residues and bacterial strain resistance. Further research will focus on the efficacy of these compounds and delivery on the field condition and the methods of dosage (sugar solution, candy or aerosol), as well as on their mechanisms of action against P. larvae.