Susceptibility to Bacteriocins in Biofilm-Forming, Variable Staphylococci Isolated from Local Slovak Ewes’ Milk Lump Cheeses

Seventeen staphylococci isolated from 54 Slovak local lump cheeses made from ewes’ milk were taxonomically allotted to five species and three clusters/groups involving the following species: Staphylococcus aureus (5 strains), Staphylococcus xylosus (3 strains), Staphylococcus equorum (one strain) Staphylococcus succinus (5 strains) and Staphylococcus simulans (3 strains). Five different species were determined. The aim of the study follows two lines: basic research in connection with staphylococci, and further possible application of the bacteriocins. Identified staphylococci were mostly susceptible to antibiotics (10 out of 14 antibiotics). Strains showed γ-hemolysis (meaning they did not form hemolysis) except for S. aureus SAOS1/1 strain, which formed β-hemolysis. S. aureus SAOS1/1 strain was also DNase positive as did S. aureus SAOS5/2 and SAOS51/3. The other staphylococci were DNase negative. S. aureus SAOS1/1 and SAOS51/3 showed biofilm formation on Congo red agar. However, using quantitative plate assay, 12 strains out of 17 showed low-grade biofilm formation (0.1 ≤ A570 < 1), while five strains did not form biofilm (A570 < 0.1). The growth of all strains, including those strains resistant to enterocins, was inhibited by nisin and gallidermin, with high inhibition activity resulting in the inhibition zone in size from 1600 up to 102,400 AU/mL (arbitrary unit per milliliter). This study contributes to microbiota colonization associated with raw ewe’s milk lump cheeses; it also indicates bacteriocin treatment benefit, which can be used in prevention and/or elimination of staphylococci.


Introduction
Milk from farm animals is a good source of nutrients. Consumers increasingly demand traditional/local products, to which those made from ewes' milk belong. Ewes' milk is rich in dry matter, minerals (calcium, zinc, and phosphorus) and B group vitamins. The principal protein in ewes' milk, casein, is tolerated by people who are allergic, for example to cows' milk [1]. Products made from ewes' milk are even more suitable for persons suffering from lactose intolerance.
In ewes' milk lump cheese, lactose is mostly removed into the whey during processing, which means that consumption of such cheese should be safe for lactose-intolerant people. Despite modernization and centralization of the dairy industry, many European cheeses are produced traditionally from raw or heat-treated milk with or without the use of natural undefined starters in farms [2]. This is also the case of local Slovak ewes' milk lump cheese.
It is a traditional Slovak cheese produced instructions (Bruker Daltonics, Billerica, MA, USA). The results were evaluated using the MALDI Biotyper 3.0 (Bruker Daltonics, Billerica, MA, USA) identification database. Taxonomic allocation was evaluated on the basis of highly probable species identification (value score 2300-3000), secure genus identification and/or probable species identification (2000-2299) and probable genus identification . Positive controls were those included in the Bruker Daltonics database. Colonies with the same MALDI-TOF score value were excluded. For subsequent testing, identified strains were maintained on MSA agar or Baird-Parker agar (Becton and Dickinson, Sparks, Maryland, USA) and stored using the Microbank system (Pro-Lab Diagnostic, Richmond, VA, Canada).

API Staph Profile-Identity Strip System
Identification was also checked with API STAPH strips, which are miniaturized biochemical tests with a specially adapted database. The strips contain dehydrated test substrates in individual microtubes. Each microtube contains a tested medium and isolate inoculated in API STAPH medium added to it. After 18-24 h incubation at 37 • C results are read and interpreted with reference to the information contained in the manual. Control strains are those involved in the test manual or Bergey's Manual of Systematic Bacteriology (Staphylococcus xylosus ATCC 35663; S. simulans ATCC 27848; American Type Culture Collection, Rockville, MA, USA). The tests involved are as follows: acidification due to carbohydrate utilization (glucose, fructose, mannose, maltose, lactose, trehalose, mannitol, xylitol, melibiose, raffinose, xylose, sucrose, α-methyl-D-glucoside, and N-acetyl-glucosamine), reduction of nitrate to nitrite, alkaline phosphatase, acetyl-methyl-carbinol production, arginine dihydrolase, and urease.
To determine nuclease activity, each strain was inoculated onto the surface of DNase agar (Oxoid, USA) and incubated for 24 h at 37 • C. After flooding and acidifying the medium with 1 N HCl, the DNA precipitated out; and the medium became turbid with clear zones around DNase-positive colonies. S. aureus SA4 from a dog (isolated in our laboratory by Dr. Strompfová) was used as positive control.
Hemolysis was detected by streaking the cultures onto BH agar (Difco, Lawrence, KS, USA) supplemented with 5% defibrinated sheep blood (from our stall). Plates were incubated at 37 • C for 24-48 h in an incubator. Presence or absence of clear zones around the colonies was interpreted as αand β-hemolysis, while those testing negative exhibited γ-hemolysis [17].

Virulence Profile (Slime Production)
Slime production or biofilm formation belongs to the virulence factors. This ability in the identified staphylococci was tested using two methods, the qualitative phenotypic method, and the quantitative method. Congo red agar is the medium used in the phenotypic qualitative method [18]. This medium was composed of brain-heart infusion (BHI, Difco; Detroit, Michigan, USA, 37 g/L) enriched with sucrose (36 g/L, Slavus, a.s. Bratislava, Slovakia), pure agar (30 g/L, Becton and Dickinson, Sparks, MA, USA) and Congo red dye (0.8 g/L, Merck, Germany). Plates of the medium were inoculated with the tested strains and incubated at 37 • C for 24 h. A positive result was indicated by black colonies with dry crystalline consistency. Non-slime producers usually remained pink. The color was also checked after 48 and 72 h. Each strain was tested in duplicate.
Biofilm formation measured by means of quantitative biofilm plate assay is a method focusing on the biofilm formation capacity of tested strains. One colony of the tested strain grown on brain-heart agar (BHA) overnight at 37 • C (Difco, Lawrence, KS, USA) was transferred into 5 mL Ringer solution (pH 7.0, 0.75% w/v) to reach McFarland Standard no. 1 corresponding to 1.0 × 10 8 CFU/mL. A volume of 100 µL from that culture was then transferred into 10 mL of brain-heart infusion (BHI). Standardized culture (200 µL) was inoculated into polystyrene microtiter plate wells (Greiner ELISA 12 Well Strips, 350 µL, flat bottom, Frickenhausen GmbH, Germany) and incubated for 24 h at 37 • C. The biofilm formed in the microtiter plate wells was washed twice with 200 µL of deionized water and dried at room temperature for 40 min. The remaining attached bacteria were stained for 30 min at room temperature with 200 µL of 0.1% (m/v) crystal violet in deionized water. The dye solution was aspirated away, and the wells were washed twice with 200 µL of deionized water and dried at room temperature for 30 min. After water removal and drying, the dye bound to the adherent biofilm was extracted with 200 µL 95% ethanol and stirred. A 150 µL aliquot was transferred from each well and placed in a new microplate well for absorbance (A) measurement at 570 nm using an Apollo 11 Absorbance Microplate reader LB 913 (Berthold Technologies, Oak Ridge, TN, USA). Each strain and condition was tested in two independent tests with 12 replicates. A sterile culture medium (BHI) was included in each analysis as negative control. Streptococcus equi subsp. zooepidemicus CCM 7316 was used as positive control (kindly provided by Dr. Eva Styková, University of Veterinary Medicine and Pharmacy in Košice, Slovakia). Biofilm formation was classified as highly positive (A 570 ≥ 1), low-grade positive (0.1 ≤ A 570 < 1) or negative (A 570 < 0.1) according to Chaieb et al. [19] and Slížová et al. [20].

Susceptibility to Bacteriocins
Four enterocins (10 µL from each) characterized in our laboratory were used: enterocin M produced by Enterococcus faecium AL41 (CCM 8558), enterocin A/P produced by Enterococcus faecium EK13 = CCM7419, both of environmental origin [21,22], Ent 412 produced by Enterococcus faecium EF412, horse strain [23] and durancin ED26E/7 produced by food origin strain Enterococcus durans ED26E/7 [24]. Partially purified enterocins (precipitates) were used as previously reported [13,14,[21][22][23]. Nisin and gallidermin used were commercial pure substances. Nisin (product Nisaplin) was supplied by Aplin and Barrett Ltd., Dorset, United Kingdom. A stock solution of Nisaplin, a commercial preparation containing nisin, was prepared according to Lauková et al. [25] and used at a dose 10 µg meaning concentration 10 µg/mL. Gallidermin was supplied by Enzo Life Sci. Corporation, Blvd Farmingdale, NY, USA (MW 2069.4), used at a concentration of 0.5 mg/mL (dose 5 µL). Seventeen staphylococcal strains were tested for bacteriocin susceptibility using the agar diffusion method according to De Vuyst et al. [26]. Brain-heart infusion/broth supplemented with 1.5% agar (BHA, Difco, Lawrence, KS, USA) was used for the bottom layer. For the overlay, 0.7% BHA was used containing 200 µL of an 18-h culture of the indicator strain with absorbance (A 600 ) of up to 0.800. Dilutions of bacteriocins, in doses of 10 µL and 5 µL (decided on the basis of our previous tests) at 1:1 ratio in phosphate buffer (pH 6.5), were aliquoted onto the surface of soft agar for each tested indicator (staphylococcal) strain. The plates were incubated at 37 • C for 18 h. Clear inhibition zones around each dose of bacteriocins were checked and the antimicrobial activity was expressed in arbitrary units per milliliter (AU/mL), which indicates the reciprocal of the highest twofold dilution of bacteriocin demonstrating complete growth inhibition of the indicator strain, and in the case of nisin and gallidermin also as the minimum inhibitory concentration (MIC, µg). The positive control was the principal indicator strain Enterococcus avium EA5 (our isolate from feces of piglets), the growth of which was inhibited by bacteriocin activity of between 25,600-204,800 AU/mL.

Microbial Characterization
The counts of staphylococci in ewe's milk cheeses were quite high. The total count of colonies grown on baird-parker (BP) agar reached 4.14 ± 2.0 log 10 CFU/g on average, while colonies on MSA agar reached 4.02 ± 2.0 log 10 CFU/g on average. BP agar is mostly used for detection of coagulase-positive staphylococci (CoPS), including S. aureus; MSA agar usually serves for determining coagulase-negative staphylococci (CoNS). Šimko and Bartko [27] previously detected staphylococci in ewe's milk in lower counts (up to 3.0, log 10, CFU/mL). After eliminating identical colonies using MALDI-TOF spectrometry score evaluation, 17 identified strains were taxonomically allotted to the following species: S. aureus (5 strains), S. xylosus (3 strains), S. equorum (1 strain) S. succinus (5 strains), and S. simulans (3 strains) ( Table 1). Variability in staphylococcal species was determined, and five different species were detected. S. aureus is a representative of CoPS; the rest of the identified strains species belong among CoNS (Table 1). According to Takashi et al. [5], based on 16S rRNA gene similarity analysis, the staphylococcal species generally fall into 11 clusters. Following this division, staphylococci detected in ewe's milk lump cheeses can be allotted to three clusters/groups involving five species, S. aureus group with five strains of S. aureus; S. simulans group with three strains of S. simulans; and S. saprophyticus group involving five strains of S. succinus, three strains of S. xylosus, and one strain of S. equorum. S. xylosus SXOS7/2 was evaluated with a score associated with highly probable species identification (2300-3000). Eight strains were measured reaching score corresponding with secure genus identification and/or probable species identification, and another eight strains reached score corresponding with probable genus identification (Table 1). Most parameters tested in the commercial phenotypic strips resulted with the parameters of the reference species strains presented in Bergey's Manual of Systematic Bacteriology [28]. Some differences in species strains can appear due to different sources of their isolation.
As mentioned previously, the genus Staphylococcus consists of Gram-positive bacteria belonging to phylum Firmicutes, and the family Staphylococcaceae. So far, this genus includes 58 known species with 25 subspecies. In this study, the most commonly detected were representatives of the S. saprophyticus group (3 species, 9 strains). Kačániová et al. [29] detected Staphylococcus aureus and Staphylococcus pasteuri in bryndza, traditional Liptov sheep cheese. Detection of Staphylococcus equorum and Staphylococcus xylosus in cheeses was mentioned by Even et al. [30]. There is only limited information regarding the microbiota in ewe's milk or cheeses; mostly only detection is reported but representatives have not been studied in detail.

Antimicrobial Phenotype Testing, Determination of Nuclease, and Hemolysis Activity
The identified staphylococci were mostly susceptible to antibiotics. They were susceptible to 10 out of 14 antibiotics, namely penicillin, erythromycin, tobramycin, streptomycin, ampicillin, cefoxitin, chloramphenicol, azithromycin, vancomycin, and rifampicin. Resistance was only rarely found ( Table 2). The tested strains were each resistant only to one or two antibiotics out of the 14 used, except for S. xylosus SXOS2/3, which was resistant to three antibiotics-clindamycin, oxacillin, and lincomycin. S. xylosus SXOS7/2 was resistant to neomycin. Neomycin resistance was also found in S. aureus SAOS2/1. S. sciuri Sci OS18/1 was resistant to lincomycin (Table 2). However, the most important finding was that all strains except for strains SXOS2/3 and SAOS2/1 were susceptible to cefoxitin and oxacillin, in which resistance was associated with methicillin [31], meaning that only two strains tested could be methicillin-resistant; however, presence of mec gene was not tested there. Susceptibility to antibiotics is one of the properties indicating their non-pathogenic character.

Biofilm Testing as Virulence Profile
Using the Congo red agar method to evaluate biofilm, we found only two strains which were positive (S. aureus SAOS1/1 and SAOS51/3). The rest of the strains did not form biofilm on Congo red agar (Table 3). However, using quantitative plate assay, 12 strains out of 17 showed low-grade biofilm formation (0.1 ≤ A 570 < 1, Table 3). Five strains did not form biofilm at all (A 570 < 0.1). Those two strains of S. aureus with positive biofilm formation on Congo red agar were also found to be positive using plate assay (Table 3).
Biofilm formation is assumed to be a factor of pathogenicity especially in pathogenic/spoilage bacteria; biofilm growth protects bacteria against host defenses and the action of antimicrobial agents; therefore, biofilm can be a source of persistent infection [32]. However, the staphylococci in this study were only low-grade biofilm-positive (Table 3). S. aureus are often detected as biofilm-positive [33]. In this study, four S. aureus strains out of five showed biofilm formation (Table 3). However, biofilm formation was also found in the strains of S. simulans, S. succinus, and S. xylosus. It seems that biofilm ability was not species-dependent.
S. aureus is one of the causative agents of mastitis in dairy herds. It could be speculated that the S. aureus detected in our ewe's milk lump cheeses may have this origin [34], even though our cheeses were made from milk sampled from healthy animals. The major sources of contaminated milk and milk products are usually considered to be improperly cleaned and sanitized farm and factory equipment. It is speculated that contaminants in milk samples may show biofilm formation [6], as shown in the staphylococci from our tested cheeses. Dairy products are very susceptible to contamination by biofilms, and it is challenging to eliminate such microbiota. How can this be done? One possibility is indicated in this study: by using bacteriocins.

Treatment with Bacteriocins
Only three out of five S. aureus strains tested were resistant to the enterocins used (Table 3). S. aureus SAOS2/1 was susceptible to all enterocins with inhibition activity ranging from 800 up to 12,800 AU/mL, and while growth of S. aureus SAOS51/3 was inhibited by treatment of enterocin (ent) M, A/P, and 412, this was associated with lower inhibition activity (100 AU/mL). S. aureus strains often show resistance against enterocins under in vitro conditions; however, they can be inhibited under in vivo conditions [24]. In vivo, e.g., fecal CoPS isolated from broiler rabbits were significantly decreased (p < 0.05) after application of durancin ED26E/7 and ent M as well. Among the species S. xylosus, growth of two strains was inhibited using all enterocins (ents), but SXOS13/4 was inhibited only after treatment with ent A/P and ent 412 (100 AU/mL). Regarding the S. succinus, three strains were inhibited with all ents; SciOS17/4 and SciOS8/1 were resistant to ent M only. The situation was similar with S. simulans strain SmiOS18/4 ( Table 3). The growth of SQOS54 was inhibited. In general, most strains were resistant to ent M but susceptible to other enterocins. The enterocins used in this study are classified as thermo-stable, small peptides with broad antimicrobial effect (involving Gram-negative bacteria) belonging in enterocins classification group II [8,[21][22][23][24], which act as pore-forming substances in the cell membrane. Their inhibition activity was reported previously under both in vitro as well as in vivo conditions [14]. In horses, administration of ent M led to mathematical reduction of coliforms, campylobacters (p < 0.05), and significant reduction of Clostridium spp. (p < 0.001). Ents 412, Ent A/P (EK13), Ent M were also effective against various staphylococci isolated from free-living trout [35]. Lauková et al. reported inhibition of Salmonella enterica serovar Dusseldorf by ent A in gnotobiotic Japanese quails. There stronger therapeutic than prophylactic effect of ent A (produced by E. faecium EK13 = CCM7419) was observed [36]. Significant decrease in CoPS as well as phagocytic activity stimulation was also noted in broiler rabbits after durancin ED26E/7 administration [24]. Nisin and gallidermin are two lantibiotic bacteriocins which are known to inhibit predominantly Gram-positive cocci. The growth of all strains tested, also including those resistant to enterocins, was inhibited with high inhibition activity by both nisin and gallidermin, with a measure inhibition zone ranging from 1600 up to 102,400 AU/mL; MIC was in the range from 0.0625 µg up to 0.005 µg (Table 4). Nisin is a peptide composed of 34 amino acid residues. Gallidermin belongs in the group of polycyclic proteinaceous bacteriocins. They both contain unusual amino acid residues such as lanthionine, β-methyllanthionine or α, β-didehydroamino acids, which are capable of building intramolecular thioether bridges, and so this group of bacteriocins is called lantibiotics [10,37]. The primary site of nisin action against cells is the cytoplasmic membrane; nisin disrupts the membrane by forming pores [37]. The mode of action of gallidermin is similar, as it integrates into the plasma membrane, forming pores and inhibiting cell membrane synthesis [38]. Similarly, as in this study, variable fecal staphylococcal strains of different species isolated from ostriches and pheasants, and producing biogenic amines, were inhibited by gallidermin, reaching inhibition zones in the range from 3200 up to 25,600 AU/mL [39]. All results from previous studies and the present one underline the effective use of enterocins and lantibiotic bacteriocins against contaminating bacteria, representing an opportunity which could be used in further practice. Their beneficial potential against clinical multiresistant staphylococci has already been indicated using aureocin A53 and epidermin [40]. Of course, additional studies are required.

Conclusions
Seventeen different staphylococci were identified in local Slovak ewe's milk lump cheeses. They belong in five species and three clusters/groups. These strains were mostly susceptible to antimicrobials and they showed low-grade biofilm ability. In addition, they were susceptible to lantibiotic bacteriocins, nisin, and gallidermin but their growth was mostly inhibited using enterocins. This is a contribution to the knowledge on the microbiota in lump cheeses made from ewes' milk, and also an indication of how to use bacteriocins for their prevention and/or elimination.
Author Contributions: A.L. was the author of a scientific idea; she treated samples to isolate microbiota, tested biofilm and susceptibility to bacteriocins, summarized results, wrote the manuscript, and designed the project; M.P.S. and V.F. tested the other properties; and M.K., M.T., and E.D. provided samples. All authors approved the final manuscript and agree with its submission. All authors have read and agreed to the published version of the manuscript.