Antagonistic Effect of Pseudomonas sp . CMI-1 on Foodborne Pathogenic Listeria monocytogenes

Foodborne pathogenic bacteria can cause serious illnesses via their growth or toxin production, thus leading to signifi cant health problems of consumers. Listeria monocytogenes is frequently transferred by consumption of contaminated food and beverages, and is an important causative agent of foodborne diseases; therefore, inhibition or elimination of this pathogenic bacterium is an essential task for food producers. L. monocytogenes can be found in raw and processed foods that are contaminated during and/or aft er processing.


Introduction
Foodborne pathogenic bacteria can cause serious illnesses via their growth or toxin production, thus leading to signifi cant health problems of consumers.Listeria monocytogenes is frequently transferred by consumption of contaminated food and beverages, and is an important causative agent of foodborne diseases; therefore, inhibition or elimination of this pathogenic bacterium is an essential task for food producers.L. monocytogenes can be found in raw and processed foods that are contaminated during and/or aft er processing.
Although the Listeria genus comprises fi ft een species, L. monocytogenes is the causative agent of human listerio-sis almost exclusively.The number of foodborne cases caused by L. monocytogenes in 2012 in the European Union was 1642 including 13 in Hungary (1).
Various microorganisms are able to inhibit pathogenic microorganisms by overgrowing them or producing antibiotic metabolites (2).The native microbiota of food have the ability to inhibit the contaminating foodborne pathogens, hence these microorganisms can prevent the growth of pathogenic bacteria by diff erent control mechanisms (predation, competitive exclusion, production of antimicrobial metabolites or quorum sensing) (3,4).Therefore, the application of specifi c bacteria isolated from foods or raw materials for the inhibition of pathogens can ISSN 1330-9862 preliminary communication doi: 10.17113/ft b.53.02.15.3731 be promising in the food industry from a safety perspective.These microorganisms have the advantage of being part of the food natural microbiota, thus they can easily colonise the food and inhibit the pathogens when present in appropriate numbers (3).Inhibition of several foodborne pathogens can also be achieved by the application of bacteriophages (5), and in the last few years virulent bacteriophages have been eff ectively used for the inhibition of L. monocytogenes (6,7).
The spoilage process of milk, fresh meat and other protein-rich raw materials represents a characteristic interaction of diff erent microorganisms when competition of saprophytic and pathogenic bacteria for the available nutrients results in the succession of populations comprising diff erent species.Diff erent antagonistic bacteria have already been isolated from natural or spoiling microbiota of foods or raw materials (3,8,9).Among the inhibitory bacteria several have been identifi ed as species of Pseudomonas, and they exhibited pathogen inhibitory features (9)(10)(11).
The objective of this research is to determine the inhibitory potential of bacteria isolated from food and raw materials of animal origin against foodborne pathogenic L. monocytogenes.Aft er performing the screening tests we focused mainly on a Pseudomonas isolate, as this bacterium showed signifi cant inhibitory activity against not only L. monocytogenes, but other non-monocytogenes Listeria species.

Bacterial strains
Bacteria tested for inhibitory eff ect had been previously isolated from diff erent processed foods or raw food materials of animal origin: among 94 isolates, 61 originated from chilled poultry meat, 9 from chilled fi sh, 14 from milk and 10 from liquid egg.They were maintained on peptone glucose yeast extract (PGY) agar slants (containing in g/L: peptone 5, glucose 1, yeast extract 2.5 and bacteriological agar 15) at 5 °C.Diff erent strains of L. monocytogenes, L. innocua and L. ivanovii ssp.ivanovii (Table 1) were used for testing the antagonistic eff ect of the bacterial isolates.

Screening of bacteria for antilisterial activity
Ninety-four bacterial isolates were screened for the ability to inhibit L. monocytogenes CCM 4699 using a spot method as described below.L. monocytogenes was cultured on tryptone soya (TS) agar plate at 37 °C for 18 h, and a cellular suspension was prepared in sterile distilled water.The absorbance of the suspension was adjusted to 0.5 at 600 nm, which corresponds to approx.10 9 CFU/mL.A tenfold serial dilution was made and 1 mL of the dilutions in the range of 10 -2 -10 -6 was massively inoculated onto TS and PGY agar plates.Aft er drying the plates, 10 μL of cell suspensions (containing approx.10 6 cells of the overnight cultures) were dropped onto the agar surface.The plates were incubated at 5, 10, 20, 25, 30 and 37 °C for 6 days.Growth inhibition was detected by formation of clearing zones around macrocolonies of the tested iso-lates.For determination of the optimal antagonistic/ pathogenic cell ratio, the selected inhibitory strains were applied in the range of 10 1 -10 7 cells/mL during co-culturing on TS and PGY agar plates.
All the bacterial isolates exhibiting inhibitory eff ect on L. monocytogenes CCM 4699 were further tested against fi ve other L. monocytogenes and two non-monocytogenes Listeria strains (Table 1) by the spot method.

Determination of the antagonistic eff ect of the selected bacterial isolates in culture broth
Co-culturing studies using TS and PGY broths were carried out in static and shake fl ask cultures.Overnight cells of L. monocytogenes CCM 4699 and the selected bacterial strains were suspended in sterile distilled water and the absorbance was adjusted to 0.5 at 600 nm.Aft er preparing tenfold serial dilutions, the culture media were inoculated with the cells of pathogenic and antagonistic bacteria in volume ratios of 1:1, 1:10, 1:100 and 1:1000, respectively.The fl asks were incubated for 2 days at 20 °C under shaking (either at 140 or 180 rpm) and static conditions.Samples were taken aft er 24 and 48 h of incubation, and cell counts were determined by the spread plate method using PGY and Listeria selective COMPASS ® agar (Biokar Diagnostics, Beauvais, France) plates.

Antagonistic eff ect of cell-free supernatants on the growth of L. monocytogenes
Isolates showing signifi cant antagonistic activity against L. monocytogenes were further studied for the production of extracellular inhibitory substances using the agar diff usion technique, as well as microplate cultures.Overnight cultures of the selected isolates were produced at 20, 25 and 30 °C, and aft er separating the cells from the culture medium by centrifugation (10 976×g, 15 min), the supernatants were removed and fi ltered through 0.2-μm pore size membrane fi lters.For the agar diff usion technique, the L. monocytogenes CCM 4699 was inoculated on the surface of TS and PGY agar plates at a fi nal count of about 10 4 CFU/mL and allowed to dry at room temperature.Wells (7 mm) were cut in the inoculated agar using a sterile metal cork borer, and fi lled with 100 μL of the supernatant.The plates were left at 5-10 °C for 2 h to allow diff usion of the compounds from the tested supernatants into the agar media, and then incubated for 24-48 h at 20 °C.Absence or presence of any inhibitory zone was recorded.
Eff ect of the inhibitory substances on 10 6 of L. monocytogenes CCM 4699 cells was tested by measuring the growth in microplate cultures using a Multiskan Ascent (Thermo Electron Corporation, Thermo Fisher Scientifi c, Waltham, MA, USA) instrument.Wells of the plates were fi lled with 300 μL of liquid consisting of 75 μL of fourfold strength culture broth, 75 μL of L. monocytogenes cell suspension and cell-free supernatants of the test strain in four diff erent (150, 100, 50 and 25 μL) volumes.Final volumes were adjusted to 300 μL by adding distilled water.The amounts of the cell-free supernatants corresponded to 1:2, 1:3, 1:6 and 1:12 dilutions.Inoculated microplates were incubated at 20 °C and the absorbance values at 595 nm were recorded automatically every 30 min during 24 h of cultivation.Growth curves were generated from the absorbance values vs. time data.

Protease and heat treatments of the cell-free supernatants
Inhibitory eff ect of the cell-free supernatants prepared from 1-, 3-and 6-day-old cultures of the tested bacteria incubated at 20 °C was determined aft er: (i) protease treatment by one of the following (in μg/mL): proteinase K (Sigma-Aldrich, St. Louis, MO, USA) 200, protease from Streptomyces griseus (Sigma) 200, trypsin from bovine pancreas (Sigma) 100 and α-chymotrypsin from bovine pancreas (Sigma) 100, at 37 °C for 90 min, and (ii) heat treatment at 95 °C for 5 and 30 min, or at 121 °C for 15 min.The growth of L. monocytogenes in the presence of the cell--free supernatants was determined by Multiskan Ascent (Thermo Electron Corporation) instrument as described above.

Detection of siderophore production
Bacterial strains were cultivated in PGY and TS broths and their growth was monitored by photometric determination of cell density at 600 nm.Cell-free supernatants were prepared from the cultures aft er 1, 3 and 6 days of incubation and siderophore production was determined by measuring the absorbance at 405 nm as described by Manninen and Matt ila-Sandholm (12).

Identifi cation and characterisation of the antagonistic bacteria
Antagonistic isolates were identifi ed at species level by direct sequencing of the amplifi ed rpoB gene or the 16S rDNA PCR products, generated by the LAPS-LAPS27 and the 27f-1492r primer pairs, respectively (13,14), and alignment of the generated sequences with those deposited in the GenBank was done by the application of the ClustalW program (UCD Conway Institute of Biomolecular and Biomedical Research, Dublin, Ireland).
The bacterial isolate with the highest inhibitory eff ect was further characterised by inoculating it onto PGY agar plates and incubating at 5, 10, 15, 20, 25, 30, 37 and 45 °C for 3 days to determine its optimal growth temperature.The isolate was also inoculated onto Cetrimid agar (Merck, Darmstadt, Germany), Pseudomonas Agar F (corres ponds to King B) and Pseudomonas Agar P (King A) (Merck, Darmstadt, Germany), and aft er incubating at 20 and 25 °C, the growth and pigmentation were recorded, while its fl uorescence was detected under UV light at 365 nm.

Screening for the antagonistic eff ect of bacterial isolates against diff erent Listeria species
Altogether 94 bacterial isolates derived from diff erent processed foods or raw materials of animal origin (chilled poultry meat or fi sh, milk and liquid egg) were tested for the inhibitory eff ect against L. monocytogenes CCM 4699 strain.For screening the inhibitory eff ect of the isolates, co-culturing investigations on PGY and TS agar plates were performed as described in Materials and Methods.Using diff erent counts of L. monocytogenes CCM 4699, it was observed that the biggest clearing zones (highest inhibition) were formed when the number of the pathogen was 10 4 CFU/mL.Ten out of the 94 isolates were able to suppress the growth of L. monocytogenes CCM 4699, which originated from chilled stored poultry meat (four isolates), milk (three isolates), chilled fi sh (two isolates) and liquid egg (1 isolate) (Table 2) (15).The most effi cient growth inhibition was detected at 20 °C, while the isolates belonged to either the psychrotrophic or mesophilic group of microorganisms according to their optimal growth temperatures.
Testing the antagonistic interactions of the ten inhibitory isolates with eight Listeria strains belonging to L. monocytogenes and non-monocytogenes Listeria species (Table 1) indicated signifi cant diff erences concerning either the sensitivity or the inhibitory eff ects of the investigated strains, as shown in Fig. 1.Pseudomonas CMI-1 isolate showed the best inhibitory eff ect as it inhibited all the tested Listeria strains.Sensitivity of the L. innocua CCM 4030 T strain was very similar to the majority of the investigated L. monocytogenes strains.On the other hand, L. ivanovii ssp.ivanovii CCM 5884 was the most sensitive, as all of the tested antagonistic strains were able to inhibit it, while L. monocytogenes P1 proved to be the most resistant.L. ivanovii is a ruminal pathogenic bacterium, although it has recently been found that it is a newly emerging opportunistic human pathogenic bacterium (16), therefore it can be considered as a feed-and food-contaminating target of biocontrol bacteria.

Identifi cation and characterisation of the most effi cient antagonistic isolates
The ten selected antagonistic bacteria with the best inhibitory eff ect were identifi ed by sequencing either the rpoB or the 16S rRNA genes, which led to the conclusion that all the strains belonged to the genus Pseudomonas (Table 2).Five of them were identifi ed as P. fl uorescens (TM--131, TM-161, TS-17, TC-4 and TE-8), which had very weak inhibition against most of the tested Listeria strains.Four isolates with considerable and uniform spectral inhibitory activity belonged to P. aeruginosa (TMI-7, TMI-9, TF-14 and TF-25).It is worth mentioning that one L. monocytogenes strain (P1) was resistant against all of the four P. aeruginosa isolates, however, because of the opportunistic human pathogenic nature of this species, these isolates were excluded from further studies.Pseudomonas CMI-1 strain, which had the best inhibitory eff ect on the growth of Listeria strains, could not be identifi ed with high certainty at species level.As similarity value between the aligned rpoB DNA sequence of this isolate and P. fredericksbergensis or P. antarctica strains deposited in the Data Bank was only 96 %, more detailed genotypic and phenotypic characterisation is required for the exact identifi cation of the isolate.This is supported by the fact that, based on our investigations, this strain is diff erent from both of the most closely sequence-related species in several characteristics.For example, the CMI-1 isolate produces fl uorescent pigment(s) on King B agar property typical of neither P. antarctica (17) nor P. fredericksbergensis (18).

Inhibition of L. monocytogenes by Pseudomonas sp. CMI-1 in co-culture experiments
The optimum cell count of Pseudomonas sp.CMI-1 for the inhibition of L. monocytogenes L4 and CCM 4699 strains was determined by co-culturing the CMI-1 and the Listeria strains on TS and PGY agar plates.As illustrated in Fig. 2, maximum inhibition against L. monocytogenes was achieved when cell count of the CMI-1 strain was three orders of magnitude higher than that of the Listeria strains.The minimum inhibition could be detected when the diff erence was only one order of magnitude or less.However, minimum inhibitory cell count depended on the Listeria strain tested.
Interaction of the Pseudomonas sp.CMI-1 and L. monocytogenes CCM 4699 strains was also studied in liquid cultures by co-culturing these strains in PGY and TS broths in diff erent ratios as described in the section Materials and Methods.Results shown in Fig. 3 indicate that increasing the ratio of CMI-1 slightly inhibits the growth of L. monocytogenes aft er two days of incubation.Decrease in the count of the pathogen was achieved when the cell count of Pseudomonas sp. was ten to one hundred times higher than that of L. monocytogenes.In these cases, Listeria cell number decreased by 1 or 2 orders of magnitude, while when number of the interacting bacteria was al-  1) by the screened antagonistic bacterial isolates (Table 2) most the same, the growth rate of the pathogen was not aff ected.Signifi cant diff erences between interactions of L. monocytogenes with Pseudomonas sp.CMI-1 using PGY and TS media for culturing were not observed.It can be concluded that co-culturing of Pseudomonas sp.CMI-1 and L. monocytogenes CCM 4699 in liquid broth did not result in signifi cant growth inhibition of the pathogen.Despite the fact that the cell number of L. monocytogenes decreased in the presence of the Pseudomonas CMI-1 strain, it was not able to outcompete L. monocytogenes.

Eff ect of cell-free supernatant on the growth of L. monocytogenes
When the eff ect of the cell-free supernatant of Pseudomonas CMI-1 strain was tested on L. monocytogenes CCM 4699 by the agar well diff usion test, no formation of an inhibition zone was detected.Therefore, a more sensitive growth inhibition test was used that monitored the growth of the pathogen in the presence of cell-free supernatant.In microplate culturing experiments, where cells of the pathogen coped directly with the potential extracellular inhibitory compound(s) of the antagonist, decrease in the growth of L. monocytogenes was detected, depending on the volume of the supernatant generated aft er 24 h of incubation.Applying the highest volume (1:2 dilution) of the supernatant, growth of L. monocytogenes was inhibited by approx.40 % (Fig. 4).Moreover, diff erences in the inhibition were also observed if supernatants obtained from 1-, 3-and 6-day-old antagonistic cultures were applied in diff erent ratios; 6-day-old supernatants had the strongest inhibition when applied in the highest ratio (data are not shown).

Characterisation of metabolites responsible for the antagonistic eff ect of Pseudomonas sp. CMI-1
As cell-free supernatants of Pseudomonas sp.CMI-1 cultivated for 6 days resulted in total growth inhibition of L. monocytogenes CCM 4699, we investigated whether the responsible extracellular metabolites could be digested by proteases of diff erent origin or if they could be inactivated by shorter or longer heat treatments at diff erent temperatures (5 or 30 min at 95 °C, and 15 min at 121 °C).Fig. 5 shows that digestion of the supernatants by four diff er-  ent proteases (protease from Streptomyces griseus, proteinase K, trypsin and α-chymotrypsin) did not decrease the inhibitory activity, indicating that a protein is not responsible for growth inhibition or resistance against the applied protease treatments.In the case of short (5 min) heat treatment at 95 °C, no change was observed in the inhibition; however, treatment of the supernatants at this temperature for 30 min and at increased temperature (121 °C) for 15 min resulted in considerable but not complete elimination of the inhibitory activity (Fig. 6).A possible explanation could be that the antilisterial compound is relatively heat stable or more than one compound is responsible for the inhibition.
It has been frequently reported that siderophore production plays an important role in the antagonistic eff ect of Pseudomonas species (19), therefore, we checked for the presence of siderophore(s) in the supernatants by detecting the absorbance in the range of 400-410 nm.A peak could always be detected at 405-410 nm in the antilisterial supernatants of 6-day-old static and shaken cultures, indicating the presence of one or more siderophores.Activity of the antilisterial substance of Pseudomonas sp.CMI-1 decreased aft er the addition of 0.1 % FeCl 3 , as the peak detected at 405 nm fl att ened out.Moreover, if the strain was cultured in the presence of 1 mM FeSO 4 , production of an A 405 nm compound could not be detected (results not shown).These results indicate that one of the metabolites responsible for the antilisterial activity of Pseudomonas sp.CMI-1 can be a heat-stable compound with an absorbance maximum typical of iron-chelating substances of fl uorescent Pseudomonas strains (20,21).Chromopeptides, such as pyoverdine, are specifi c iron-binding compounds of fl uorescent pseudomonads (22), seeking to deprive pathogens of iron, thus suppressing their growth (23).Chromogenic siderophores are peptides, but in our case digestion with diff erent proteases had no eff ect on the activity; therefore, further analyses for determining the characteristics of the inhibitory compound(s) produced by Pseudomonas sp.CMI-1 are necessary.
Formation of siderophores is also aff ected by the oxygen concentration.Sabra et al. (24) observed that high concentrations of oxygen in the culture broth of P. aeruginosa PAO1 caused an oxidative stress to the cells which led to the reduction of the growth rate, production of exopolysaccharides (mainly alginate), and enhanced release of diff erent proteins.However, greater pyocyanin formation was detected under microaerophilic conditions.This is in accordance with our observations in the case of Pseudomonas sp.CMI-1, which produced higher levels of the siderophore(s) in static cultures (at low oxygen concentration) than in shake fl ask cultures (Fig. 7).Formation of iron chelators was also infl uenced by the concentration of iron (25) and the presence of diff erent organic carbon sources in the culture medium (26).Production of pyoverdine and/or pyochelin increased signifi cantly under iron limitation, while the highest concentration of siderophores was observed in an iron-free standard succinate medium containing succinate as the sole carbon source (25,26).
In our study, we compared the growth of cells and siderophore production in TS and PGY culture media.As it is shown in Fig. 7, TS broth supported the growth of cells bett er than PGY broth both in aerated and static cultures; however, the specifi c siderophore production, calculated as the ratio of siderophore content and cell density, was higher in PGY.This could be explained by the fact that TS broth is a more complex medium than PGY; therefore, more nutrients are available for the cells during their growth.In PGY broth the amount of accessible nutrients is lower, which can contribute to the production of stress--related substances.Moreover, the digested soya bean component contains low levels of iron, and in its presence the production of important virulence factors, such as siderophores, is partially inhibited.
When cells in PGY static culture were stressed by two rate-limiting factors (i.e.oxygen and iron availability), they responded by considerably increased siderophore production.This supports the observations of Rachid and Ahmed (26) who found that siderophore biosynthesis of Pseudomonas isolates was aff ected by diff erent environparameters, and iron limitation was the most signifi cant.

Conclusions
Raw food materials of animal origin can harbour bacteria that are able to inhibit the growth of foodborne pathogenic bacteria, like Listeria monocytogenes.The Pseudomonas sp.CMI-1 strain originating from milk was selected as the most promising antagonistic bacterium, which was able to inhibit all the tested pathogenic strains of the genus Listeria.Antagonistic eff ect of this strain was demonstrated in contact inhibition tests against L. monocytogenes; however, because direct application of the an- tagonistic cells for biocontrol purposes in food might lead to a spoilage-like eff ect, the extracellular nature of the growth inhibitory metabolites has been investigated too.
Cell-free supernatants generated under diff erent growth conditions also inhibited L. monocytogenes, and siderophores were detected as a group of potentially inhibitory compounds.However, heat-and protease-resistant compounds might also be responsible for growth decline of foodborne pathogenic L. monocytogenes.The identifi ed and well characterised inhibitory compound(s) can be applied in food production to combat L. monocytogenes and in the agrochemical segment for the inhibition of pathogenic bacteria like L. monocytogenes.This bacterium is an ubiquitous microorganism which can be found not only in food and raw materials but also on/in plants.Earlier studies have demonstrated that pathogens may colonise the internal tissues of plants, and consumption of such contaminated vegetables can cause health hazard.

Table 1 .
List of Listeria strains used for the analysis of the antagonistic eff ect of bacterial isolates

Table 2 .
Identifi cation of the antagonistic bacteria isolated from diff erent raw food materials of animal origin Fig. 1.Inhibition of diff erent L. monocytogenes, L. innocua and L. ivanovii strains (Table