Isolation and identification of bacteriocinogenic strain of Lactobacillus plantarum with potential beneficial properties from donkey milk

Authors

  • A. Murua,

    1. Department of Agriculture and Life Sciences, Faculty of Medicine, University of Maribor, Hoče, Slovenia
    Search for more papers by this author
  • S.D. Todorov,

    Corresponding author
    1. Departamento de Alimentos e Nutrição Experimental, Faculdade de Ciências Farmacêuticas, Laboratório de Microbiologia de Alimentos, Universidade de São Paulo, São Paulo, SP, Brazil
    • Department of Agriculture and Life Sciences, Faculty of Medicine, University of Maribor, Hoče, Slovenia
    Search for more papers by this author
  • A.D.S. Vieira,

    1. Departamento de Alimentos e Nutrição Experimental, Faculdade de Ciências Farmacêuticas, Laboratório de Microbiologia de Alimentos, Universidade de São Paulo, São Paulo, SP, Brazil
    2. EMBRAPA, Caprinos e Ovinos, Sobral, CE, Brazil
    Search for more papers by this author
  • R.C.R. Martinez,

    1. Departamento de Alimentos e Nutrição Experimental, Faculdade de Ciências Farmacêuticas, Laboratório de Microbiologia de Alimentos, Universidade de São Paulo, São Paulo, SP, Brazil
    Search for more papers by this author
  • A. Cencič,

    1. Department of Agriculture and Life Sciences, Faculty of Medicine, University of Maribor, Hoče, Slovenia
    Search for more papers by this author
    • Deceased
  • B.D.G.M. Franco

    1. Departamento de Alimentos e Nutrição Experimental, Faculdade de Ciências Farmacêuticas, Laboratório de Microbiologia de Alimentos, Universidade de São Paulo, São Paulo, SP, Brazil
    Search for more papers by this author

Correspondence

S.D. Todorov, FCF-USP, Av. Prof. Lineu Prestes 580, Bloco 14, 05508-000, São Paulo, SP, Brazil. E-mail: slavi310570@abv.bg

Abstract

Aims

The goal of this study was to isolate and characterize a lactic acid bacteria (LAB) from donkey milk with potential beneficial properties.

Methods and Results

Lactic acid bacteria were isolated from donkey milk and identified based on physiological, biochemical and molecular methods. The isolate that presented highest bacteriocin potential (Lactobacillus plantarum LP08AD) was evaluated for the production of bacteriocin, including stability in the presence of various enzymes, surfactants, salts, pH and temperatures. Bactericidal effect of bacteriocin LP08AD on Listeria monocytogenes, Enterococcus faecium and Lactobacillus curvatus was shown for actively growing and stationary cells. Similar growth and bacteriocin production were observed when strain LP08AD was cultured in MRS broth at 30°C or 37°C. Bacteriocin LP08AD adhered at low levels on the producer cells (200 AU ml−1). The presence of plantaricin W gene on the genomic DNA was recorded based on PCR. Good growth for strain LP08AD was recorded in MRS broth with pH from 5·0 to 9·0 and LP08AD grew well in the absence of oxbile or concentration below 0·8%. Lact. plantarum LP08AD was applied to the small intestinal epithelial polarized monolayers of H4, PSIc1 and CLAB and demonstrated low attachment ability on all cell lines studied, with values with a similar behaviour for cells from human and pig origin.

Conclusions

Bacteriocin-producing Lact. plantarum LP08AD might be useful in the design of novel functional foods with potential probiotic or biopreservation properties.

Significance and Impact of the Study

To the best of our knowledge, this is the first report on detection and characterization of bacteriocinogenic Lact. plantarum from donkey milk. The strain LP08AD shows to have potential beneficial properties, as demonstrated by the use of noncancerogenic cell lines.

Introduction

Lactic acid bacteria (LAB) represent a heterogeneous group of micro-organisms that are naturally present in many foods and in the gastrointestinal and urogenital tract of animals. It has been shown that these LAB can produce antimicrobial compounds, such as bacteriocins or bacteriocin-like inhibitory substances (De Vuyst and Vandamme 1994; Todorov 2008). Bacteriocins of LAB are defined as ribosomally synthesized proteins or protein complexes usually antagonistic to genetically closely related organisms (Nes and Johnsborg 2004). In the last decade, some research papers have demonstrated that certain bacteriocins are also active against certain Gram-negative bacteria, such as Escherichia coli and Salmonella typhimurium (Todorov and Dicks 2005a; Gong et al. 2010) and interestingly against Campylobacter jejuni, the major cause of gastroenteritis worldwide (Stern et al. 2006; Svetoch et al. 2011; Messaoudi et al. 2012). Furthermore, bacteriocins can block the reproduction of some viruses (Wachsman et al. 1999, 2003; Todorov et al. 2005, 2008, 2010). More than 300 different bacteriocins have been described for the genera Lactobacillus, Lactococcus, Leuconostoc, Pediococcus and Enterococcus (De Vuyst and Vandamme 1994; Olasupo 1998; Çetinkaya et al. 2003; Osuntoki et al. 2003). They are generally low-molecular-weight proteins that gain entry into target cells by binding to cell surface receptors and which bactericidal mechanisms vary, including pore formation, degradation of cellular DNA, disruption through specific cleavage of 16S rRNA and inhibition of peptidoglycan synthesis (James et al. 1991; De Vuyst and Vandamme 1994; Heu et al. 2001). A number of bacteriocins have been described for Lactobacillus plantarum, a heterofermentative microaerophilic Gram-positive micro-organism, with rod morphology, which occurs singly or as grouped in short chains (reviewed by Todorov 2009). This species has well-accepted GRAS status, and numerous strains of Lact. plantarum have been isolated from different ecological niches, including meat, fish, fruits, vegetables, milk and cereal products (reviewed by Todorov and Franco 2010). Application of Lact. plantarum as a beneficial LAB/probiotics have been studied extensively in last 20 years (Martinez et al. 2012). Lactobacillus plantarum has been used as a starter culture in various food fermentation processes contributing to the organoleptic properties, flavour and texture. Due to the production of lactic acid and other antimicrobial compounds, Lact. plantarum also contributes to the safety of the final products and then can be used for food biopreservation. Bacteriocinogenic Lact. plantarum have been isolated from fermented meat products, milk, cheese, fermented cucumber, olives, dough, pineapple, grapefruit juice, sorghum beer, molasses, boza, kefir, papaya and Amasi (reviewed by Todorov 2009).

To the best of our knowledge, this is the first report describing the isolation and identification of Lact. plantarum from donkey milk, as well as being a pioneer study on the production of bacteriocin from LAB isolated from this source and evaluation of their probiotic potential.

Materials and methods

Isolation of bacteriocin-producing strains

Donkey milk was collected, from different animals (Equus africanus asinus), in a local farm of the region of Stajerska (Maribor, Slovenia). Collected samples were frozen at −20°C prior to isolation of the LAB. The LAB were isolated by preparing serial dilutions in physiological solution (0·85% NaCl, w/v) and plated onto MRS agar (Difco, BD, Franklin Lakes, NJ, USA). The plates were incubated at 30°C for 2 days. Colonies were isolated randomly based on the differences in the morphology and inoculated into MRS broth (Difco) at 30°C for 24 h for obtaining fresh cultivars. The purity of the cultures was checked by plating on MRS agar and performance of Gram staining and optical microscopy. Pure cultures were stored at −80°C in MRS broth added of 20% (w/v) of glycerol.

The selected isolates were grown on MRS broth for 24 h at 30°C, and antimicrobial activity was tested by using the agar-spot test method as described by Todorov (2008). Cell-free supernatants of tested LAB obtained by centrifugation (8000 g, 10 min, 4°C) were adjusted to pH 6·0 with 1 mol−1 NaOH in order to prevent the antimicrobial effect of lactic acid and further treated for 10 min at 80°C to deactivate potentiality produced extracellular proteases and hydrogen peroxide.

Antimicrobial activity was expressed as Arbitrary units per millilitre (AU ml−1) and defined as the reciprocal of the highest dilution showing a clear zone of growth inhibition (Todorov 2008) and calculated according to formula AU ml−1 = [(Dn × 1000)/V], where D = the dilution factor; n = number of the highest dilution showing at least 2 mm inhibition zone; V = volume of the spotted cell-free supernatant. The indicator strains used are listed in Table 1.

Table 1. Spectrum of activity of the bacteriocin produced by Lactobacillus plantarum LP08AD
Test micro-organismMediaT (°C)Diameter of inhibition zone (mm)
Enterococcus faecium ET05MRS3011
Ent. faecium ET12MRS300
Ent. faecium ET88MRS300
Ent. faecium ST55MRS3020
Lactobacillus curvatus ET06MRS307
Lact. curvatus ET30MRS3018
Lact. curvatus ET31MRS3010
Lact. curvatus MBSa2MRS300
Lact. curvatus MBSa3MRS300
Lactobacillus delbrueckii ET32MRS300
Lactobacillus fermentum ET35MRS3016
Lactobacillus plantarum MBSa4MRS300
Lactobacillus sakei MBSa1MRS300
Listeria monocytogenes L103BHI3015
L. monocytogenes L104BHI3019
L. monocytogenes L106BHI3020
L. monocytogenes L211BHI3020
L. monocytogenes L302BHI3016
L. monocytogenes L409BHI3016
L. monocytogenes L506BHI3014
L. monocytogenes L603BHI3024
L. monocytogenes L607BHI3022
L. monocytogenes L620BHI3020
L. monocytogenes L703BHI3021
L. monocytogenes L709BHI3021
L. monocytogenes L724BHI3022
Pediococcus acidilactici ET34MRS3010

Identification of bacteriocin-producing strains

The strains isolated from donkey milk were identified according to their physiological and biochemical characteristics, as previously described by Todorov and Franco (2010): Gram staining, growth at different temperatures, pH measurement, assessment of CO2 production from glucose and gluconate, oxidase and catalase reactions. In addition to biochemical tests, biomolecular identification was performed. DNA extracted from the selected LABs was isolated by using Zymo Research DNA isolation kit (Zymo Research, The Epigenetics Company, Irvine, CA, USA). For future identification and confirmation of the results from physiological identification, robust 16S rRNA sequencing was performed (Felske et al. 1997) using the universal primers 8f (5′-CAC GGA TCC AGA CTT TGA TYM TGG CTC AG-3′) and 1512r (5′-GTG AAG CTT ACG GYT AGC TTG TTA CGA CTT), where Y indicates C + T and M indicates A + C. Amplified fragments were cloned into the pGEM®-T Easy Vector system (Promega, Madison, WI, USA) and constructs were used to transform E. coli DH5α. Plasmid DNAs were isolated using a QIAprep Spin miniprep kit (Qiagen®; Valencia, CA, USA) and sequenced using bigdye™ terminator cycle sequencing chemistry (Biosystems, Warrington, UK) on an ABI Genetic Analyzer 3130XL Sequencer (Applied Biosystem, SA, Pty, Ltd., Foster City, CA, USA).

Further differentiation of LAB was performed by random amplification of polymorphic DNA (RAPD-PCR). DNA from the selected LABs was isolated as described before and RAPD-PCR performed with primer OPL-2 (TGG CGG TCA A) (Kit L of the RAPD® lomer kits; Operon Biotechnologies, Cologne, Germany) according to the study described by Todorov (2010).

Characterization of the bacteriocin effect of proteolitic enzymes, chemicals, pH and temperature

Cell-free supernatant of Lact. plantarum LP08AD obtained by centrifugation (8000 g, 10 min, 4°C) was adjusted to pH 6·0 with 1 mol−1 NaOH. Aliquots of 2 ml were incubated for 2 h in the presence of 1·0 mg ml−1 (final concentration) of trypsin (Sigma), pronase (Sigma), proteinase K (Sigma), pepsin (Sigma), papain (Sigma), catalase (Sigma), lipase (Sigma) and α-amylase (Sigma) and then tested for antimicrobial activity using the agar-spot test method against Listeria monocytogenes 703, Enterococcus faecium ET05 and Lactobacillus curvatus ET30 (Todorov 2008).

In another set of experiments, the effect of SDS, Tween 20, Tween 80, urea, Triton X-100, Triton X-114, Na-EDTA and NaCl on bacteriocin activity in cell-free supernatants was determined as described by Todorov (2008). The effect of pH on the bacteriocin activity was determined by adjusting the cell-free supernatant between pH 2·0 and 12·0 with sterile 1 mol−1 HCl or 1 mol−1 NaOH. After 2 h of incubation at 30°C, the samples were readjusted to pH 6·5 and the activity determined as described before (Todorov 2008). The effect of temperature on bacteriocin activity was tested by heating the cell-free supernatants at 30, 37, 45, 60 and 100°C, respectively. Residual bacteriocin activity was tested after 30, 60 and 120 min for all temperatures evaluated, as previously described by Todorov (2008).

In all the above-mentioned studies, uninoculated MRS broth was exposed to the same conditions (digestive enzymes, chemicals, pH and temperature) as were the cell-free supernatants or the semi-purified bacteriocin and used as negative control against L. monocytogenes 703, Ent. faecium ET05 and Lact. curvatus ET30.

Effect of bacteriocin on growth of test micro-organisms

About 20 ml aliquot of bacteriocin-containing filter-sterilized (0·20 μm; Millipore Corporation, Billerica, MA, USA) supernatant (pH 6·0) was added to 100-ml cultures of L. monocytogenes 603, L. monocytogenes 703, Ent. faecium ET05 and Lact. curvatus ET30 at early exponential phase and incubated for 12 h. Optical density readings (at 600 nm) were recorded at 1-h intervals.

Determination of the reduction of viable cells of test micro-organisms in the presence of bacteriocin LP08AD

Early stationary phase (18–h old) cultures of L. monocytogenes 603, L. monocytogenes 703, Ent. faecium ET05 and Lact. curvatus ET30 were harvested (5000 g, 5 min, 4°C), washed twice with sterile physiological solution and resuspended in 10 ml of this solution. Equal volumes of cell suspensions and filter-sterilized (0·20 μm, Minisart®; Sartorius AG, Goettingen, Germany) supernatant from LP08AD (crude preparation of bacteriocin) were mixed, and viable cell numbers were determined before and after incubation for 1 h at 37°C by plating 100 μl aliquots onto MRS agar. Cell suspensions of L. monocytogenes 603, L. monocytogenes 703, Ent. faecium ET05 and Lact. curvatus ET30 without the addition of the bacteriocin were used as negative controls.

Growth dynamics of Lactobacillus plantarum LP08AD, assessment of bacteriocin production and pH changes

An 18-h-old culture of strain LP08AD was inoculated (2%, v/v) into MRS broth (Difco) and dynamics of bacteriocin LP08AD production was assessed at 30 and 37°C during 24 h. Samples were taken every hour and examined for bacterial growth (OD 600 nm), changes in culture medium pH and antimicrobial activity against L. monocytogenes 703, Ent. faecium ET05 and Lact. curvatus ET30. The agar-spot test method was used, and the activity of the bacteriocin expressed as AU ml−1 as described previously. In an independent experiment, the effect of initial medium pH on bacteriocin LP08AD production was also evaluated.

Adsorption to producer cells

The ability of the bacteriocin to adsorb the producer cells was studied as described by Yang et al. (1992). After the growth of Lact. plantarum LP08AD in MRS broth for 18 h at 30°C, the pH of the medium was adjusted to pH 6·0 with 1 mol l−1 NaOH, the cells harvested (10 000 g, 15 min, 4°C) and washed with sterile 0·1 mol l−1 phosphate buffer (pH 6·5). The cells were resuspended in 10 ml 100 mmol l−1 NaCl (pH 2·0), stirred for 1 h at 4°C and then harvested (12 000 g, 15 min, 4°C). The cell-free supernatant was neutralized to pH 7·0 with sterile 1 mol l−1 NaOH and tested for activity using the agar-spot test (Todorov 2008).

Screening for the presence of bacteriocin genes in total DNA isolated from Lactobacillus plantarum LP08AD

Total DNA was isolated as described before and submitted to PCR using different sets of primers to target several plantaricins (plantaricin W, plantaricin S and plantaricin NC8) and pediocin PA-1. PCRs were performed using the GeneAmp® PCR instrument System 9700 (Applied Biosystems, Foster City, CA, USA). The PCR conditions were designed according to Stephens et al. (1998), Holo et al. (2001), Maldonado et al. (2003) and Todorov et al. (2010) and based on the specification of the primers, which are summarized in Table 2. The amplified products were visualized in a 1% or 2% (w/v) agarose gel stained with ethidium bromide solution (0·5 μg ml−1).

Table 2. Primers used for the detection of plantaricin W, plantaricin S, plantaricin NC8 and pediocin PA-1 in Lactobacillus plantarum LP08AD genomic DNA
BacteriocinPrimersTargeted size (bp)
Plantaracin W

planW-F:5′ TCA CAC GAA ATA TTC CA 3′

planW-R:5′ GGC AAG CGT AAG AAA TAA ATG AG 3′

165
Plantaracin S

planS-F:5′ GCC TTA CCA GCG TAA TGC CC 3′

planS-R:5′ CTG GTG ATG CAA TCG TTA GTT T 3′

450
Plantaracin NC8

planNC8-F:5′ GGT CTG CGT ATA AGC ATC GC 3′

planNC8-R:5′ AAA TTG AAC ATA TGG GTG CTT TAA ATT CC 3′

207
Pediocin PA-1

PedPro:5′ CAA GAT CGT TAA CCA GTT T 3′

Pedc1041:5′ CCG TTG TTC CCA TAG TCT AA 3′

1238

Auto- and coaggregation

The strains Ent. faecium ET05, Lact. plantarum ET30, L. monocytogenes 603, L. monocytogenes 703 and Lact. plantarum LP08AD were grown in MRS broth and BHI broth, respectively, for 24 h at 37°C. The cells were harvested by centrifugation (7000 g, 10 min, 20°C), washed, resuspended and diluted in sterile saline water (0·85% NaCl, w/v) to obtain an OD660 nm = 0·3, determined using a spectrophotometer (Ultraspec 2000, Amersham Pharmacia, Uppsala, Sweden). After incubation for 60 min at 37°C, cells were harvested (300 g, 2 min, 20°C), and the OD660 nm of the supernatant was assessed accordingly. Autoaggregation was determined using the following equation (Todorov et al. 2008): percentage autoaggregation = [(OD0–OD60)/OD0] × 100. OD0 refers to the initial OD, and OD60 refers to the OD determined after 60 min.

For the evaluation of coaggregation, strains Ent. faecium ET05, Lact. plantarum ET30, Lact. plantarum LP08AD L. monocytogenes 603 and L. monocytogenes 703 were grown in 10 ml of MRS or BHI broth at 37°C. Cells were harvested after 24 h (7000 g, 10 min, 20°C), washed, resuspended, and diluted in sterile saline water to OD660 nm = 0·3. The degree of coaggregation was determined by OD readings obtained for paired studied culture and coaggregation partners suspensions (ratio of 500 μl and 500 μl of each suspension). Cells were harvested (300 g, 2 min, 20°C), and the OD660 nm of the supernatant was determined. Coaggregation was calculated using the following equation (Todorov et al. 2008): per cent coaggregation = [(ODtot–ODs)/ODtot] × 100. ODtot refers to the initial OD, taken immediately after the relevant strains were paired. ODs refer to the OD of the supernatant obtained after 60 min incubation. Experiments were conducted in triplicates.

Effect of pH and bile on growth of Lactobacillus plantarum LP08AD

Lactobacillus plantarum LP08AD was grown in MRS broth (Difco) adjusted to pH 3·0, 5·0, 7·0, 9·0 and 13·0 by adding 1 mol l−1 HCl or 1 mol l−1 NaOH before autoclaving. If needed, pH was readjusted after sterilization by addition of sterile 1 mol l−1 HCl or 1 mol l−1 NaOH. The strain was also grown in MRS broth containing 0·4, 0·6, 0·8, 1·0, 2·0 and 3·0% (w/v) oxbile (Sigma, St Louis, MO, USA). All tests were conducted in sterile flat-bottom 96-well microtitre plates (TPP, Trasedingen, Switzerland). Each well was filled with 180 μl of the medium and inoculated with 20 μl of cultures of Lact. plantarum LP08AD previously grown in MRS broth (Difco) at 37°C up to OD600 nm = 0·3. Growth of Lact. plantarum LP08AD in the tested conditions was monitored through measurement of OD600 nm every hour for 12 h, using a microtitre plate reader (Versa-max, Sunnyvale, CA, USA). Cultures grown in nonmodified MRS broth served as control. Experiments were performed in triplicates.

Resistance to antibiotics

Lactobacillus plantarum LP08AD was tested for resistance to several antibiotics (Table 3). The antibiotics were supplied by CEFAR, São Paulo (Brazil). Lactobacillus plantarum LP08AD was inoculated into 10 ml MRS broth (Difco) and incubated at 37°C for 18 h and mixed into MRS soft agar (1·0%, w/v) (Difco), to achieve a population of ca. 106 CFU ml−1. After solidification of the agar, each antibiotic (antibiotic disc) was spotted onto the surface of the plates and incubated at 37°C for 24 h. The plates were examined for the presence of inhibition zones around the antibiotic discs. The inhibitory effect of the antibiotics was expressed in millimetres of the inhibition zones.

Table 3. Profile of antimicrobial resistance of Lactobacillus plantarum LP08AD determined against several antibiotic agents (inhibition zones are expressed in millimetres)
Antibiotic (μg per disc)Inhibition zone (mm)
Amicacin (30)0
Amoxicilin + clavulanic acid (30)30
Ampicillin (10)31
Ampillicin (20)31
Bacitracin (10)10
Cefaclor (30)20
Cefazolin (30)23
Cefepime (30)14
Cefotaxim (30)15
Cefoxitin (30)10
Ceftazidim (30)0
Ceftriaxon (30)20
Ceftriofur (30)20
Cefuroxim (30)20
Ciprofloxacin (5)0
Claritromycin (15)26
Clindamycin (2)15
Cotrimazin (25)27
Cotrimixazol (25)18
Doxycicline (30)15
Erythromicin (15)15
Eurofloxacyn (5)10
Florphenicol (30)0
Furazolidon (10)20
Gentamycin (10)0
Impenem (10)32
Kanamycin (30)0
Linezolid (30)10
Minocycline (30)25
Moxifloxacin (5)0
Nalidixic acid (30)0
Neomycin (30)11
Nitrofurantoin (300)21
Ofloxacin (5)26
Oxacillin (1)12
Penicillin G (10)25
Rifampicin (30)25
Rifampicin (5)18
Spectinomycin (10)8
Streptomycin (10)0
Streptomycin (300)11
Sulfonamide (300)0
Tetracycline (30)23
Tilmicosin (15)20
Tobramycin (10)0
Vancomycin (30)0

Evaluation of interaction of Lactobacillus plantraum LP08AD with a cell lines

Bacterial cultures, including strains belonging to different species of Lactobacillus spp. and Ent. faecium, Lactobacillus casei and Carnobacterium divergens were used as positive and negative controls, respectively. The strains were maintained at −80°C in MRS broth (Merck, KGaA, Darmstadt, Germany) added of 20% (v/v) glycerol. Propagation, prior to tests performed with cell lines, was later carried out in MRS broth (Merck) for 24 h at 37°C and under anaerobic conditions by the use of Anaerogen (Oxoid Ltd, Hampshire, UK).

The small intestinal epithelial cell lines H4, PSIc1 and CLAB were grown in advanced Dulbecco's Modified Eagle's Medium (DMEM) (Sigma-Aldrich, Grand Island, NY, USA), supplemented with 5% foetal calf serum (Lonza, Basel, Switzerland), l-glutamine (2 mmol l−1; Sigma), penicillin (100 units ml−1, Sigma) and streptomycin (1 mg ml−1; Fluka, Buchs, Switzerland). Cell lines were routinely grown in 25-cm2 culture flasks (Corning, New York, NY, USA) at 37°C in a humidified atmosphere of 5% CO2 and 95% air, until confluent monolayers were obtained. Culture medium was changed routinely. To obtain polarized monolayers, cells were seeded on Transwell filter inserts (0·4-μm pore size, 12 mm; Corning) placed into 12-well plates (22·1 mm; Corning) to reach a density of 1 × 105 cells cm−2. Functional polarity was established when transepithelial resistance (TER) between the apical and basolateral surface of the PSIc1 and H4 monolayers was >1000 Ω cm−2. Before inoculating the bacterial strains, the monolayers cells were washed twice with 100 μl DMEM without phenol red added of supplements.

Adhesion ability of potential probiotic strains was assayed on polarized PSIc1, H4 and CLAB cells. When cell monolayers reached functional polarity, bacterial strains were resuspended in DMEM without supplements and added to the apical compartment of individual cell monolayers at a concentration of 1 × 107 CFU ml−1 and further incubated for 90 min in a humidified atmosphere of 5% CO2 and 95% air, as previously described (Ivec et al. 2007; Maragkoudakis et al. 2010). Following incubation, the medium supernatant with the nonattached bacteria was removed, and the intestinal cell lines were washed twice with nonsupplemented DMEM. Aliquots of 100 μl of trypsin solution (Botić et al. 2007) were then added to detach adhered bacteria, which were subsequently enumerated by the standard serial dilution method on MRS agar plates at 30 or 37°C for 48 h under microaerobic conditions (Anaerocult A).

Results

Isolation and identification of LAB from donkey milk

The average count of LAB obtained in the donkey milk samples collected from local farm of the region of Stajerska (Maribor, Slovenia) was estimated to be 7·2 × 103 CFU g−1. However, the aim of this work was to select a potential beneficial bacteriocinogenic LAB, and further tests have been performed in this direction. From randomly selected 30 colonies, only seven presented rod morphology, catalase negative and Gram positive. When tested for antibacterial activity, the seven isolates presented inhibitory effect against L. monocytogenes 603, L. monocytogenes 703, Ent. faecium ET05 and Lact. curvatus ET30. These seven isolates were successfully identified as Lact. plantarum according to physiological and biochemical tests. When submitted to API50CHL, all isolates were positively identified as Lact. plantarum with levels of identification between 99·8 and 100%. Further identification of these isolates by 16S rRNA gene sequencing indicated 99% homology to Lactobacillus arizonensis. Based on the similarity of the biochemical features, morphology, RAPD-PCR and spectrum of activity, all isolates were considered as a replicate of the same strain (data not shown). In this work, only the isolate designated as LP08AD, presenting the largest inhibition zones against L. monocytogenes 603, L. monocytogenes 703, Ent. faecium ET05 and Lact. curvatus ET30 was further studied.

Characterization of bacteriocin LP08AD

Bacteriocin LP08AD presented a large spectrum of activity, inhibiting the growth of LAB, food spoilage bacteria and foodborne pathogens, including Ent. faecium, Lact. curvatus, Lactobacillus fermentum, Pediococcus acidilactici and L. monocytogenes (Table 1). In addition, no bacteriocinogenic activity was detected against more than 50 Lactobacillus spp. from the culture collection of University of Sao Paulo, Faculty of Pharmaceutical Sciences, Department of Food Science. It is important to emphasize the bioactivity of LP08AD bacteriocin against L. monocytogenes, a foodborne pathogen of increasing importance worldwide.

Similar growth rates of Lact. plantarum LP08AD and production of its bacteriocin (102 400 AU ml−1 and 51 200 AU ml−1, respectively) were recorded, when the strain was cultivated either at 30 or 37°C in MRS broth for 24 h. However, very low growth rates were observed, when the micro-organism was grown at 15°C (data not shown), and under this condition, a reduced level of bacteriocin activity was detected (200 AU ml−1). It is important to underline that Lact. plantarum LP08AD produced higher levels of bacteriocin against L. monocytogenes in comparison with Lact. curvatus and Ent. faecium (Table 4 and Fig. 1).

Table 4. Antimicrobial activity (AU ml−1) of bacteriocin LP08AD against Listeria monocytogenes 703, Lactobacillus curvatus ET30 and Enterococcus faecium ET05
Time (h)Cultivation of Lactobacillus plantarum LP08AD at 30°CCultivation of Lactobacillus plantarum LP08AD at 37°C
Bacteriocin activity (AU ml−1) againstBacteriocin activity (AU ml−1) against
L. monocytogenes 703Lact. curvatus ET30Ent. faecium ET05L. monocytogenes 703Lact. curvatus ET30Ent. faecium ET05
032002000320000
325 60064000640016000
651 2006400051 20012 800200
951 2006400051 20012 800200
15102 40012 800051 20012 800400
18102 4006400051 20025 600400
21102 4003200051 20025 600400
Figure 1.

Growth of Lactobacillus plantarum LP08AD (expressed as OD600 nm and indicated by ♦ symbol) in MRS broth adjusted to pH 6·5 under incubation at 30°C (a) and 37°C (b). Standard deviation recorded from three repeats was <5% and then is not indicated.

Treatment of the cell-free supernatant with proteolytic enzymes resulted in a complete inactivation of antimicrobial activity (Table 5). Treatment with catalase did not affect the activity against the target strains (Table 5) clearly excluding the involvement of H2O2 in the antagonism process. Moreover, treatment with α-amylase and lipase did not affect the antimicrobial activity needier (Table 5).

Table 5. Stability of bacteriocin produced by Lactobacillus plantarum LP08AD after treatment with enzymes, detergents, urea, NaCl, heat and pH
TreatmentTest micro-organisms
Listeria monocytogenes 703Enterococcus faecium ET05Lactobacillus curvatus ET30
  1. The symbol ‘+’ means that LP08AD bacteriocin was stable against a given treatment, whereas the symbol ‘−’ means it was not.

Enzymes, 0·1 mg ml−1
Trypsin, pronase, pepsin, proteinase K, papain
Catalase+++
Lipase+++
α-amylase+++
Detergents, salts (1%, m/v)
SDS, Tween 20, Tween 80, urea, Na-EDTA and NaCl+++
Triton X-100, Triton X-114
pH
2·0–3·0+++
4·0–8·0+++
10·0+++
12·0
Temperature (for 120 min)
30°C, 37°C, 45°C, 60°C+++
100°C+++
121°C for 20 min+

The activity of bacteriocin LP08AD was not affected by 1% SDS, Tween 20, Tween 80, urea, EDTA and NaCl (Table 5). Only the exposition of the free cell supernatant containing bacteriocin LP08AD to 1% Triton X-100 and Triton X-114 resulted in the reduction of the bacteriocin activity (Table 5). In the present study, bacteriocin LP08AD was not affected by the pH, since it remained active after incubation for 2 h at pH ranging from 2·0 to 10·0 (Table 5). Bacteriocin LP08AD remained stable after 2 h at 30, 37, 45, 60 and 100°C (Table 5). However, when it was exposed to 121°C for 20 min at pH 6·0, bacteriocin LP08AD activity was recorded only against Ent. faecium ET05, but not against L. monocytogenes 703 and Lact. curvatus ET30.

Effect of bacteriocin LP08AD on test micro-organisms

Addition of cell-free supernatant of a 24-h-old culture of Lact. plantarum LP08AD to a 3-h-old culture of L. monocytogenes 603, L. monocytogenes 703, Lact. curvatus ET30, or Ent. faecium ET05 (early exponential phase) repressed cell growth of the indicator strains over 9 h (Fig. 2). When the supernatant was added to a 7-h-old culture, a similar inhibition was observed (data not shown). After treatment with bacteriocin LP08AD, no viable cells of L. monocytogenes 603, L. monocytogenes 703, Lact. curvatus ET30 and Ent. faecium ET05 were detected after 8 or 10 h, underlining the bactericidal mode of action of this bacteriocin. By challenge of L. monocytogenes 603, L. monocytogenes 703, Lact. curvatus ET30 and Ent. faecium ET05 (108–109 CFU ml−1), at stationary phase, with bacteriocin LP08AD, a complete loss of the micro-organisms cellular viability was recorded. After 1-h contact time, no viable cells of L. monocytogenes 603, L. monocytogenes 703, Lact. curvatus ET30 and Ent. faecium ET05 were detected (data not shown), whereas the counts were not modified when these micro-organisms were incubated with the control samples.

Figure 2.

Effect of bacteriocin LP08AD on growth (expressed as OD600 nm readings) of (a) Listeria monocytogenes 703, (b) L. monocytogenes 603, (c) Enterococcus faecium ET05 and (d) Lactobacillus curvatus ET30. (image) L. monocytogenes 703; (image) L. monocytogeness 703 in presence of bacteriocin LP08AD; (image) L. monocytogenes 603; (image) L. monocytogenes 603 in presence of bacteriocin LP08AD; (image) Ent. faecium ET05; (image) Ent. faecium ET05 in presence of bacteriocin LP08AD; (image) Lact. curvatus ET30 and (image) Lact. curvatus ET30 in presence of bacteriocin LP08AD.

Adsorption of bacteriocin to producer cells

Treatment of Lact. plantarum LP08AD with 100 mmol l−1 NaCl (pH 2·0) for 1 h resulted in the adsorption of the bacteriocin to these cells, since the antimicrobial activity under this condition was determined as 200 AU ml−1.

Growth dynamics of Lactobacillus plantarum LP08AD and bacteriocin production

Similar growth and bacteriocin production were observed, when Lact. plantarum LP08AD was cultured for 24 h in MRS broth at 30 or 37°C. Under both temperatures, the activity against L. monocytogenes 703 was higher (51 200 AU ml−1) as recorded after 6 h of growth. However, when Lact. plantarum LP08AD was cultivated at 30°C, the levels of bacteriocin production reached 102 400 AU ml−1 after 15 h of incubation (Fig. 1, Table 4).

Detection of bacteriocin genes

Based on the performed PCRs targeting plantaricin W, plantaricin S, plantaricin NC8 and pediocin PA-1 genes in the total DNA of Lact. plantarum LP08AD, positive results were generated only with primers for plantaricin W (data not shown). Based on these findings, most probably, Lact. plantarum LP08AD is a plantaricin W producer. However, further purification and amino acid sequence of the expressed bacteriocin is required to confirm the expression of plantaricin W gene.

Aggregation studies

Autoaggregation of Lact. plantarum LP08AD was 36·65%, slightly higher than figures recorded for L. monocytogenes 603 (33·96%) and L. monocytogenes 703 (34·24%), but lower compared with autoaggregation ability determined for Ent. faecium ET05 (51·28%) and Lact. curvatus ET30 (57·06%) (Fig. 3). Coaggregation results of Lact. plantarum LP08AD with the pathogens L. monocytogenes 603, L. monocytogenes 703 and opportunistic pathogen Ent. faecium ET05 were determined as 33·55, 38·65 and 45·96%, respectively; in fact, these values are lower than the one found for Lact. curvatus ET30, a nonpathogen micro-organism (49·77%) (Fig. 3).

Figure 3.

Auto-aggregation and co-aggregation figures recorded for strains Lactobacillus plantarum LP08AD (LP08AD), Listeria monocytogenes 703 (L703), L. monocytogenes 603 (L603), Enterococcus faecium ET05 (ET05) and Lactobacillus curvatus ET30 (ET30).

Effect of initial pH and the presence of oxbile on growth of Lactobacillus plantarum LP08AD

Good growth of Lact. plantarum LP08AD was recorded in MRS broth (Difco), when the initial pH was adjusted to 5·0, 6·0, 7·0 or 9·0 (Fig. 4). Lactobacillus plantarum LP08AD grew well in the absence of oxbile and when the concentration of the compound was below 0·8% (Fig. 4).

Figure 4.

Growth of Lactobacillus plantarum LP08AD (expressed as OD600 nm) in MRS broth (Difco) supplemented with different concentrations of ox-bile (a) and in MRS broth (Difco) adjusted to different pH values (b). Every data showed represents an average of three optical density readings. (a) (image) No oxbile; (image) oxbile 0·40%; (image) oxbile 0·60%; (image) oxbile 0·80%; (image) oxbile 1·00%; (image) oxbile 2·00%; (image) oxbile 3·00%; (b) (image) pH 6·5; (image) pH 3·0; (image) pH 5·0; (image) pH 7·0; (image) pH 9·0 and (image) pH 13·0.

Susceptibility of Lactobacillus plantarum LP08AD to antibiotics

Except for amicacin (30 μg per disc), ceftazidim (30 μg per disc), ciprofloxacin (5 μg per disc), florphenicol (30 μg per disc), gentamycin (10 μg per disc), kanamycin (30 μg per disc), moxifloxacin (5 μg per disc), nalidixic acid (30 μg per disc), streptomycin (10 μg per disc), sulfonamide (300 μg per disc), tobramycin (10 μg per disc) and vancomycin (30 μg per disc), all the antibiotics tested in this study inhibited the growth of Lact. plantarum LP08AD to some extent (Table 3).

Evaluation of interaction of Lactobacillus plantraum LP08AD with a cell lines

Lactobacillus plantarum LP08AD was applied to the small intestinal epithelial polarized monolayers of H4, PSIc1 and CLAB. The application of the selected putative probiotic strain (LP08AD) on the examined cell lines did not lead to any detrimental effects on the cell line polarized monolayer integrity and viability as compared to healthy monolayers. In this work, Lact. plantarum LP08AD demonstrated very low attachment ability on all cell lines studied, with values estimated as below 1% showing a similar behaviour for cells from human and pig origin.

Discussion

Seven isolates from donkey milk samples presenting antimicrobial activity were successfully identified as Lact. plantarum according to physiological and biochemical tests. Based on 16S rRNA gene sequencing, these isolates were indicated 99% homology to Lact. arizonensis. According to Kostinek et al. (2005), Lact. arizonensis should be considered as a heterotypic synonym of Lact. plantarum, since no sufficient evidences exist to distinguish it as a new species. In the present work, based on the similarity of the biochemical features, morphology, RAPD-PCR and spectrum of activity, all isolates were considered as a replicate of the same strain, and one of the isolates, designated as LP08AD, was selected for further studies. Total microbiota of donkey milk collected in this study is being subjected to further analysis by biomolecular and biochemical approaches. Bacteriocin LP08AD was able to inhibiting the growth of many food spoilage bacteria and foodborne pathogens, including Ent. faecium, Lact. curvatus, Lact. fermentum, Ped. acidilactici and L. monocytogenes (Table 1). Most of the bacteriocins described for Lact. plantarum are active only against a narrow range of genera and species (De Vuyst and Vandamme 1994).

Similar growth rates of Lact. plantarum LP08AD and production of its bacteriocin were recorded, when the strain was cultivated either at 30°C or at 37°C in MRS broth for 24 h, compared with a very low bacteriocin production when the micro-organism was grown at 15°C. However, it is important to underline that Lact. plantarum LP08AD produced higher levels of bacteriocin against L. monocytogenes in comparison with Lact. curvatus and Ent. faecium (Table 4 and Fig. 1). This phenomenon may be explained by the presence of specific receptors on the surface of L. monocytogenes, but not in Lact. curvatus nor in Ent. faecium, important for the interaction of bacteriocin and target cells. This high activity against L. monocytogenes may have a practical application in the control of the foodborne pathogen and prevention of listeriosis in fermented milk products. It is important to clarify that fermented milk products very frequently contain different Lactobacillus spp. as a starter culture, and in this case, low activity levels against Lactobacillus spp. is considered a positive characteristic for this bacteriocin.

Bacteriocin LP08AD was deactivated in the presence of proteolytic enzymes. Similar results were observed when cell-free supernatant and semi-purified bacteriocin produced by Lact. plantarum ST16PA were tested (Todorov et al. 2011). However, as been expected, treatment with catalase did not affect the activity against the target strains, clearly excluding the involvement of H2O2 in the antagonism process. Similar results have been reported for other bacteriocins of Lact. plantarum (De Vuyst and Vandamme 1994; Kelly et al. 1996; Todorov et al. 2004). Moreover, treatment with α-amylase and lipase did not affect the antimicrobial activity, suggesting that bacteriocin LP08AD does not belong to the controversial group IV (Klaenhammer 1988) bacteriocins, which contain carbohydrates or lipids in the active molecule structure.

The activity of bacteriocin LP08AD was not affected by 1% SDS, Tween 20, Tween 80, urea, EDTA and NaCl. Only the exposition of the cell-free supernatant-containing bacteriocin LP08AD to 1% Triton X-100 and Triton X-114 resulted in the reduction in the bacteriocin activity. The sensitivity to detergents, NaCl and urea seems to be bacteriocin-dependent. Bioactivity of plantaricin C19 was not affected by the presence of SDS or Triton X-100 (Atrih et al. 2001) contrarily to bacteriocins produced by Ent. faecium ST311LD (Todorov and Dicks 2005c), plantaricin 423 (Verellen et al. 1998), pediocin PA-1/AcH (Biswas et al. 1991), lactacin B (Barefoot and Klaenhammer 1984) and lactocin 705 (Vignolo et al. 1995). However, bacteriocin ST16PA was sensitive to Triton X-114 but not affected by presence of Triton X-100 (Todorov et al. 2011).

In the present study, bacteriocin LP08AD was not affected by the pH, since it remained active after incubation for 2 h at pH ranging from 2·0 to 10·0. Similar observations have been previously reported for pediocin PA-1/AcH (Bhunia et al. 1988). However, at pH 12, bacteriocin LP08AD lost its activity that might be ascribed to proteolytic degradation, protein aggregation or instability of protein at this extreme pH (Parente and Riccardi 1994; Parente et al. 1994; De Vuyst et al. 1996; Aasen et al. 2000).

Bacteriocin LP08AD remained stable after 2 h at various temperatures between 30°C and 100°C. However, when it was exposed to 121°C for 20 min at pH 6·0, bacteriocin LP08AD activity was recorded only against Ent. faecium ET05, but not against L. monocytogenes 703 and Lact. curvatus ET30. Similar results were reported for pediocin PA-1/AcH, which was resistant to heat treatment at 80°C for 60 min and 100°C for 10 min, but not to 121°C (Ray et al. 1992). Heat resistance of pediocin PA-1/AcH was pH-dependent; at pH 6·0, 84% of the activity was lost after heating at 121°C for 15 min, and no activity was registered after the same heat treatment at pH 7·0 and 8·0. At pH 4·0, only 11% of the activity was lost. Similar results were recorded for other plantaricins (Todorov et al. 1999; Todorov and Dicks 2005a,b,c).

Addition of cell-free supernatant of Lact. plantarum LP08AD to a early exponential phase cultures of L. monocytogenes 603, L. monocytogenes 703, Lact. curvatus ET30 or Ent. faecium ET05 repressed cell growth of the indicator strains over tested period of 9 h. Similar results were recorded with 7-h-old culture of test micro-organisms. After treatment with bacteriocin LP08AD, no viable cells of L. monocytogenes 603, L. monocytogenes 703, Lact. curvatus ET30 and Ent. faecium ET05 were detected after 8 or 10 h, underlining the bactericidal mode of action of this bacteriocin.

By the challenge of stationary phase cells of L. monocytogenes 603, L. monocytogenes 703, Lact. curvatus ET30 and Ent. faecium ET05 (108–109 CFU ml−1) in nongrowing conditions with bacteriocin LP08AD, a complete loss of the micro-organisms cellular viability (death) was recorded. Similar results were obtained, when lower concentrations of those tested micro-organisms were used. Other bacteriocins, such as the one produced by Lact. plantarum ST16 Pa (a strain isolated from papaya) or Ped. acidilactici HA-6111-2 (a strain obtained from a fermented sausage traditionally produced in Portugal) presented a similar behaviour (Albano et al. 2007; Todorov et al. 2011).

Treatment of Lact. plantarum LP08AD with 100 mmol l−1 NaCl (pH 2·0) for 1 h resulted in the adsorption of the bacteriocin to these cells, since the antimicrobial activity under this condition was determined as 200 AU ml−1. Similar results were reported for bacteriocin ST16Pa (Todorov et al. 2011). In contrast, no bacteriocin adsorption was reported for plantaricin ST31 (Todorov et al. 1999), bacteriocins ST194BZ and ST414BZ (Todorov and Dicks 2006), bacteriocins ST26MS and ST28MS (Todorov and Dicks 2005a) or bozacin B14 to the producer strains (Ivanova et al. 2000).

Similar growth and bacteriocin production were observed, when Lact. plantarum LP08AD was cultured for 24 h in MRS broth at 30 or 37°C. However, when Lact. plantarum LP08AD was cultivated at 30°C, the levels of bacteriocin production reached 102 400 AU ml−1 after 15 h of incubation. This is in agreement with results obtained for plantaricin ST31 (Todorov et al. 1999), bacteriocins ST26MS and ST28MS (Todorov and Dicks 2005a) and mundticin ST4SA (Todorov and Dicks 2009). Based on these results and taking into consideration a potential application of this strain as a potential probiotic, all further experiments were conducted at 37°C.

Several studies have shown that bacteriocin production is dependent on the biomass. Todorov and Dicks (2005b) reported that optimal levels of plantaricin ST194BZ, produced by Lact. plantarum ST194BZ, were obtained in growth media that supported high biomass production, such as MRS. For Lact. plantarum ST16Pa, a decrease in the activity of ST16Pa bacteriocin at the end of the monitored period was observed and could be explained by the degradation of the bacteriocin by extracellular proteolytic enzymes (Todorov et al. 2011). Similar reductions have also been observed for bacteriocins produced by Enterococcus mundtii ST4SA (Todorov and Dicks 2009) and Ped. acidilactici NRRL B5627 (Anastasiadou et al. 2008). From a metabolic point of view, this trend would be characteristic of primary metabolite production.

Based on the performed PCRs targeting plantaricin W, plantaricin S, plantaricin NC8 and pediocin PA-1 genes in the total DNA of Lact. plantarum LP08AD, positive results were generated only with primers for plantaricin W. Based on these findings, most probably Lact. plantarum LP08AD is a plantaricin W producer. However, further purification and amino acid sequence of the expressed bacteriocin is required to confirm the expression of plantaricin W gene or to claim the production of new bacteriocin(s) by Lact. plantarum LP08AD.

It is important to point out that the application of different approaches for the characterization of a bacteriocin is critical. Very frequently authors report on new bacteriocin identification merely based in the determination of the presence of gene(s) for bacteriocin(s) production (De Kwaadsteniet et al. 2006, 2008). Albano et al. (2007) verified the presence of pediocin PA-1 genes in two different strains of Ped. acidilactici isolated from ‘Alheira’. The evidences for the expression of this bacteriocin were not reported in both cited works (De Kwaadsteniet et al. 2006; De Kwaadsteniet et al. 2008; Albano et al. 2007) and only may be presumed that were expressed. The purification of the bacteriocin followed by mass spectrometry and amino acid sequencing (even only partial sequence or amino acid composition) is a proof that the genes are actually being expressed (Zendo et al. 2008). It was shown that some bacteriocin-producer strains may carry several genes for bacteriocin production and depending on growth conditions they express one or another gene (Poeta et al. 2007). Reports based only on the detection or identification of bacteriocin genes need to be considered with high scepticism, and results should be accepted only when additional research and sufficient evidence of bacteriocin production and expression is made available.

Autoaggregation of Lact. plantarum LP08AD was 36·65%. Coaggregation results of Lact. plantarum LP08AD with the pathogens L. monocytogenes 603, L. monocytogenes 703 and opportunistic pathogen Ent. faecium ET05 were determined as 33·55, 38·65 and 45·96%, respectively; in fact, these values are lower than the one found for Lact. curvatus ET30, a nonpathogen micro-organism (49·77%). The low levels of autoaggregation and coaggregation with pathogens may play an important role in preventing the formation of biofilms and, in this way, may help eliminating pathogens in the gastrointestinal tract.

Good growth of Lact. plantarum LP08AD was recorded in MRS broth (Difco) when the initial pH was set at 5·0, 6·0, 7·0 or 9·0 (Fig. 4). Similar results were reported for Lactococcus lactis HV219 (Todorov et al. 2007). In another study, growth of several strains of Lact. plantarum, Lactobacillus rhamnosus, Lactobacillus pentosus and Lactobacillus paracasei was suppressed at pH 3·0 and 4·0, while variable results were recorded in medium with an initial pH of 11·0 and 13·0, with poor growth recorded for strains Lact. paracasei ST242BZ and ST284BZ, Lact. rhamnosus ST462BZ, Lact. plantarum ST664BZ and Lact. pentosus ST712BZ at pH 13·0 (Todorov et al. 2008).

Lactobacillus plantarum LP08AD grew well in the absence of oxbile and when the concentration of the compound was below 0·8% (Fig. 4). The bacteriocinogenic Lactococcus lactis HV219 strain was less tolerant to bile salts, since its growth was inhibited by testing concentrations above 0·3% (Todorov et al. 2007). Other studies have also reported similar effects of oxbile, and pH for different strains of Lactobacillus spp., such as Lact. plantarum 423, Lactobacillus salivarius 241 and Lact. curvatus DF38 (Brink et al. 2006). Haller et al. (2001) reported variable results for the growth of different strains of Lact. plantarum when exposed to HCl (pH 2·0) and bile salts, and the figures were as high as 10% for Lact. plantarum cells, but <0·001% for Lactobacillus sakei and Lact. paracasei. Although these assays cannot predict patterns of behaviour in the human body, the results are valuable in selecting Lactobacillus spp. for probiotic applications, as resistance to low pH and high concentrations of bile salts is important for growth and survival of bacteria in the intestinal tract (Havenaar et al. 1992; Carvalho et al. 2009).

Except for amicacin, ceftazidim, ciprofloxacin, florphenicol, gentamycin, kanamycin, moxifloxacin, nalidixic acid, streptomycin, sulfonamide, tobramycin and vancomycin, all the antibiotics tested in this study inhibited the growth of Lact. plantarum LP08AD to some extent. However, it is important highlight that normally nalidixic acid is an antibiotic active against Gram-negative bacteria. Resistance of LAB to antibiotics is a controversial subject. It is important to keep in mind that antibiotic-resistant probiotic LAB may be responsible for horizontal transfer of resistance genes to other bacteria present in the human GIT (Dicks et al. 2009). Resistance may be inherent to a bacterial genus or species but may also be acquired through exchange of genetic material, mutations or incorporation of new genes (Ammor et al. 2007). Teuber (1999) and Salyers et al. (2004) suggested that starter cultures and probiotics may serve as vectors in the transfer of antibiotic-resistant genes. Such transfers had been documented in other bacterial groups by Levy and Marshall (2004) and Salyers et al. (2004). On the other hand, probiotics needs to be resistant to certain antibiotics in order to be facilitating their application as a combined treatment with antibiotics.

To investigate in vitro interactions between probiotic bacteria and the intestinal mucosa, several cell lines and different culturing models have been used elsewhere (Botić et al. 2007; Yang et al. 2007; Ewaschuk et al. 2008; Zoumpopoulou et al. 2009). The human colon tumorigenic cell lines CaCo2, T84 and HT-29 have been widely utilized for bacterial adhesion assays and mechanistic studies but are not an appropriate model for pathogen–host interactions, because they do not derive from the small intestine, have a tumorigenic phenotype distinct from that of normal gut epithelia, express modified surface glycoconjugates and are much more glycosylated than normal cells. These latter considerations are becoming important due to the increasing significance of cell culture-based assays that mimic the in vivo environment (Klingberg et al. 2005; Tremblay and Slutsky 2007). These problems on accuracy can be overtaken by the use of H4, PSIc1 and CLAB cell lines; among them, the first one was obtained from small intestinal tissue of a human foetus aged 20–22 weeks of gestation (Sanderson and Walker 1995) and further subcloned. Moreover, these cell lines are not of tumour origin and are, therefore, a better in vitro model for studying mechanisms of gut interactions than human tumorigenic cell lines (e.g. CaCo2). The cell line H4 is a noncancerogenic cell line, positive for epithelial markers cytokeratins (CK), alkaline phosphatase and laminin. Cells polarize when growing on microporous inserts, develop microvilli and TER up to 1000 ohms. H4 cell line obtained from pig is the closest one to humans in terms of genome, organ, development, anatomy, physiology and metabolism of the intestinal tract (Brown and Timmermans 2004), disease progression (Lunney 2007) and also for intestine–microbe interactions (Pipenbaher et al. 2009; Maragkoudakis et al. 2010).

PSI cell line was recovered from the dissected small intestine of an adult pig by using limiting dilution technique. PSI cells are positive for epithelial markers CK 5-18 and weakly positive for alkaline phosphatase and PAS staining, but negative for CK-19 and all markers characteristic for myeloid cells, myofybroblasts and mesenchyme. PSI cells form tightly packed epithelial barrier when grown on microporous membranes with or without collagen. PSI clones give very high TER (up to 7000 Ω) and exhibit high potential (up to 40 mV), characteristic for a highly differentiated epithelial character and high transmembrane transport. Upon treatment and activation, cells usually respond by elevated mitochondrial dehydrogenases activity and weakly for reactive oxygen species (ROS) and nitric oxide (NO) production. These cells are classified as cryptic, since they are not fully differentiated from intestinal epithelial cells (Cencic and Langerholc 2010).

On other hand, CLAB cells are enterocytes obtained from the dissected small intestine of an adult pig. CLAB cells are positive for epithelial markers such as CK 5-18 and CK-19 (a specific marker for porcine enterocytes) and for alkaline phosphatase, whereas essentially negative for CK-18 (a marker for pig M-cells) and actin (marker for myofibroblasts). Some of the cells are positive for desmin (a marker for cells derived from mesenchyme). These cells are strongly positive for PAS staining, indicating the presence of mucins at the surface of the cells, which consists of one of the most significant features of differentiated enterocytes and goblet cells. Although classified as epithelial, they do not display TER or transepithelial potential (TEP). CLAB cell line is weakly positive for nonspecific esterases, amido-black staining and acid-phosphatase but essentially negative for markers indicating cells of myeloid origin (Cencic and Langerholc 2010). Upon treatment and activation, CLAB cells respond significantly by increasing the secretion of NO, ROS and mitochondrial dehydrogenases activity (Nissesn et al. 2009). The established small intestinal cell lines were validated and proven to be excellent and reliable models for the evaluation of safety and probiotic activitiy (Trapecar et al. 2011).

Lactobacillus plantarum LP08AD was applied to the small intestinal epithelial polarized monolayers of H4, PSIc1 and CLAB. The application of the selected putative probiotic strain (LP08AD) on the examined cell lines did not lead to any detrimental effects on the cell line polarized monolayer integrity and viability as compared to healthy monolayers.

In the present work, Lact. plantarum LP08AD demonstrated very low attachment ability on all cell lines studied, with values estimated as below 1% showing a similar behaviour for cells from human and pig origin. It is important to take in consideration the fact that adhesion properties of the potential probiotic bacteria are very depended on the future application of the LAB. As it has been shown previously (Nissesn et al. 2009; Cencic and Langerholc 2010; Trapecar et al. 2011), it is of outmost importance to use polarized small intestinal monolayers to determine the attachment ability of putatitve probiotic strains, as in simple monolayers grown on plastic, the epithelial cells do not exert the intestinal characteristics; therefore, the attachment sites of the basal and basolateral membranes may be available to the bacteria. Such false-positive results can mislead and bring to wrong conclusions, since the attachment sites may not be available on the apical membrane of fully differentiated small intestinal epithelia.

Conclusions

To the best of our knowledge, this is the first report on detection of Lact. plantarum from donkey milk. Based on the results obtained, including bacteriocin LP08AD activity and mode of action, we can conclude that bacteriocin-producing Lact. plantarum LP08AD might be useful in the design of novel functional foods with potential probiotic or biopreservation properties.

Acknowledgements

This research project was funded by a grant obtained from CNPq (Conselho Nacional de Desenvolvimento Científico e Tecnológico) (Process # 490098/2009-6) - Brasília, DF, Brazil. S. D. Todorov acknowledge to CNPq and FAPESP for providing his visiting researcher grant for the performance of research work at Faculdade de Ciências Farmacêuticas – Universidade de São Paulo (São Paulo, SP, Brazil). The authors are grateful to Prof. Maria Teresa Destro (FCF-USP) for providing L. monocytogenes strains and to Prof. Elisabetta Tome (Universidad Central de Venezuela, Caracas, Venezuela) for L. curvatus and Ent. faecium used in this study.

Ancillary