Slovak Local Ewe’s Milk Lump Cheese, a Source of Beneficial Enterococcus durans Strain

Slovak ewe’s milk lump cheese is produced from unpasteurized ewe’s milk without any added culture. Because of the traditional processing and shaping by hand into a lump, this cheese was given the traditional specialty guaranteed (TSG) label. Up till now, there have existed only limited detailed studies of individual microbiota and their benefits in ewe’s milk lump cheese. Therefore, this study has been focused on the beneficial properties and safety of Enterococcus durans strains with the aim to contribute to basic dairy microbiology but also for further application potential and strategy. The total enterococcal count in cheeses reached 3.93 CFU/g (log 10) ± 1.98 on average. Based on a MALDI-TOF mass spectrometry evaluation, the strains were allotted to the species E. durans (score, 1.781–2.245). The strains were gelatinase and hemolysis-negative (γ-hemolysis) and were mostly susceptible to commercial antibiotics. Among the strains, E. durans ED26E/7 produced the highest value of lactase enzyme β-galactosidase (10 nmoL). ED26E/7 was absent of virulence factor genes such as Hyl (hyaluronidase), IS 16 element and gelatinase (GelE). To test safety, ED26E/7 did not cause mortality in Balb/c mice. Its partially purified bacteriocin substance showed the highest inhibition activity/bioactivity against Gram-positive indicator bacteria: the principal indicator Enterococcus avium EA5 (102,400 AU/mL), Staphylococcus aureus SA5 and listeriae (25,600 AU/mL). Moreover, 16 staphylococci (out of 22) were inhibited (100 AU/mL), and the growth of 36 (out of 51) enterococcal indicators was as well. After further technological tests, E. durans ED26E/7, with its bacteriocin substance, can be supposed as a promising additive to dairy products.


Introduction
Ewe's milk and cheeses made from it have a high nutritive value, which has led to high demand for those products that is increasing worldwide [1,2]. The indigenous microbiota in raw milk plays an important role in the cheese formation quality [3]. In general, cheese is composed of microbiota originating from the raw ingredients used, the environment and in some types of cheese, the added starter cultures or adjunct cultures. These many sources of microbiota cause considerable variability in the microbiome across cheese varieties [4]. Based on newly developed modern identification techniques, the microbiota of cheeses can be studied in more detail [4][5][6]. Recently, Parente et al. [7] reported a review study associated with the microbiota of dairy milk. Based on meta-analysis results, it was concluded that four phyla, such as Firmicutes, Proteobacteria, Bacteroidetes and Actinobacteria, included the majority of the most abundant and prevalent taxa identified in milk. There were detected psychrotrophs, bacteria associated with teat skin, with gut and also potentially beneficial lactic acid bacteria (LAB) [4][5][6][7]. In general, LAB are considered

Phenotypization and Enzyme Production of E. durans
Phenotypization of E. durans strains was performed based on consensus matrix of tests for identification of Enterococcus spp. as reported by Manero and Blanch [31], also involving the Rapid ID Streptotest (R Inc., Reading, PA, USA), as well as the API-ZYM system (BioMérieux, France). The testing panel involves arginine, esculin, mannitol, sorbitol, raffinose, inulin, galactose, glucose, NAG (N-acetyl-glucosamine), hippurate hydrolysis, etc. Type strain E. durans ATCC 19432 was used as a control.

Hemolysis, Gelatinase Activities and Antibiotic Profile
Hemolysis was detected by streaking the cultures on MRS agar (Difco) supplemented with 5% of defibrinated sheep blood. Plates were incubated at 37 • C for 24-48 h under semi-anaerobic conditions. The presence or absence of clearing zones around the colonies was interpreted as β-hemolysis and negative γ-hemolysis, respectively [32].
Gelatinase is a proteolytic enzyme (extracellular metalloendopeptidase EC 3.4.24.30) that acts on a variety of substrates, such as insulin-beta chain, collagenous material in tissues, the vasoconstrictor endothelin-1, as well as sex-pheromones and their inhibitor peptide [33]. Gelatinase activity was detected with a 3% gelatin medium (Todd-Hewitt agar, Becton and Dickinson, Cockeysville, MD, USA). After the growth of tested strains (48 h at 37 • C), plates were flooded with a 15% solution (HgCl 2 in 20% HCl). The loss of turbidity halos around colonies was then checked at 4 • C [34].

Bacteriocin Activity of E. durans
Ents genes possessing strains were tested for antimicrobial activity using the quantitative agar diffusion method [45]. For this test, concentrates of the strains were prepared as follows: strains (0.1% pre-inoculum) were inoculated in the Brain heart broth (Difco, MD, USA) overnight at 37 • C. Then they were centrifuged for 30 min at 10,000× g (at laboratory temperature). Supernatants were concentrated 10-fold using Concentrator plus (Eppendorf AG, Hamburg, Germany) to achieve a volume of 4 mL. Their activity was tested against the principal (the most susceptible indicator Enterococcus avium EA5, fecal strain of piglet, our laboratory), and against Listeriae from different food products (Table 1) Listeria monocytogenes CCM4699 (Czech Culture Collection, Brno, Czech Republic), L. innocua LMG13568 (University Brussel, Belgium) and Staphylococcus aureus SA5 (our strain isolated from mastitis milk). Bacteriocin activity was expressed in arbitrary units per mL, indicating the highest dilution of bacteriocin that can inhibit the growth of the indicator strain. Based on the results of inhibition activity, ED26E/7 was selected to prepare semi-purified bacteriocin (precipitate). Inhibition activity testing was performed in duplicate.

Partial Purification of Bacteriocin ED26E/7 and Activity Testing
At first, partial purification of the bacteriocin substance followed the protocol according to Mareková et al. [39] for Ent P; the activity of 1600 AU/mL against E. avium EA5 strain was measured. Therefore, partial purification was performed according to the protocol for durancin 28L [22], there was a modification regarding the temperature and time of cultivation (although the durancin gene was not tested). ED26E/7 was cultivated in 500 mL of MRS (Merck, Germany) at 37 • C for 18 h. Then the broth culture was centrifuged (10,000× g) for 30 min (min) at 4 • C. The pH of the supernatant was adjusted to 4.2 and filtrated using a 0.45 µm filter (Millipore Corp. Bedford, MA, USA). The supernatant was precipitated with ammonium sulfate (80% saturation) at 4 • C overnight (18 h). After precipitation and centrifuging (10,000× g) for 30 min at 4 • C, the precipitate was re-suspended in the minimum volume of phosphate buffer (pH 6.5) and inhibition activity was tested against E. avium EA5. Inhibition activity reached 102,400 AU/mL. Moreover, other indicators were used: S. aureus SA5, L. innocua LMG13568, L. monocytogenes (12) from various food products, various staphylococci from ewe's milk lump cheese and raw goat milk, E. faecium, E. faecalis, E. hirae from raw goat milk and meat products, E. thailandicus from beavers, and fecal strains E. mundtii from roe and red deers.

In Vitro Safety Control
To control safety of E. durans ED26E/7, three specific genes for virulence factors were tested. PCR amplification with the primers and conditions used followed the protocols according to Kubašová et al. [46] and Lauková et al. [47]. The genes tested were GelE (gelatinase), hylEfm (hyaluronidase) and Is16 (element IS). The PCR products were separated by means of agarose gel electrophoresis (1.2% w/v, Sigma-Aldrich, Saint Louis, MO, USA) with 1 µL/mL content of ethidium bromide (Sigma-Aldrich) using 0.5× TAE buffer (Merck, Darmstadt, Germany). PCR fragments were visualized with UV light. E. faecium P36 (Dr. Semedo-Lemsaddek, University Lisbon, Portugal) was the positive control. The PCRs were carried out in 25 µL volume with a mixture consisting of 1× reaction buffer, 0.2 mmoL/L of deoxynucleoside triphosphate, 3 mmoL MgCl 2 , 1 µmoL/L of each primer, 1 U of Taq DNA polymerase and 1.5 µL of the DNA template, with the cycling conditions as previously reported by Kubašová et al. [42] and Lauková et al. [43].

In Vivo Safety Control
For the in vivo safety control of E. durans ED26E/7 strain, pathogen-free eight-weekold male Balb/c mice (VELAZ Prague, Czech Republic) were used. Their weight was around 18-20 g. Mice maintenance conditions were the same as previously reported by Vargová et al. [48]. Mice were kept under a 12-h light/dark regimen at a temperature of 22-24 • C with a humidity of 56%. They were on a commercial diet, and water was available without restriction. Mice were divided randomly into 2 groups: Control (n = 10) and Group 1 (n = 10). To differ the ED26E/7 strain from other enterococci, its rifampicin-resistant Foods 2021, 10, 3091 6 of 13 variant was prepared [49]. ED26E/7 was administered per os daily at a dose of 10 9 CFU/mL in a total dose 100 µL. Counts of ED26E/7, as well as other enterococci, were enumerated after standard microbiological dilution of feces and jejunum (jejunum was homogenized using Masticator, Spain) and plated on BHI agar enriched with rifampicin (100 µg), and M-Enterococcus agar. Counts were expressed in CFU/g (log10) ± SD. Sampling occurred at the start of the experiment (n = 20) and on days 7 and 30.

Identification of E. durans Strains
The total enterococcal count in the screened cheeses reached 3.93 CFU/g (log 10) ± 1.98 on average. Based on MALDI-TOF mass spectrometry evaluation, strains were allotted to the species E. durans with evaluation scores from 1.781 up to 2.245 (Figure 1). Among the nine allotted colonies, five strains (ED24E/9, ED25E/6, ED26E/1, ED26E/7 and ED7E/9) were selected for the next testing. Four identical colonies were excluded. E. durans ED26E/7 reached a score of 2.125, ED7E/9 had a score of 1.781, ED26E/1 possessed a score of 1.921, ED25E/6 was scored with a value of 2.245 and ED24E/9 had a score of 2.154. To support taxonomical allotment, phenotyping was provided, and the results were compared with the type strain E. durans ATCC 19432. In these five strains, the N-acetylglucosamine test was positive, as well as the fermentation of galactose, D-glucose and lactose, while mannitol, inulin and sorbitol were not fermented (they were negative), as well as raffinose. The Voges-Proskauer test was positive; the hydrolysis of esculin showed a dubious reaction and hippurate hydrolysis as well.
22-24 °C with a humidity of 56%. They were on a commercial diet, and water was av ble without restriction. Mice were divided randomly into 2 groups: Control (n = 10) Group 1 (n = 10). To differ the ED26E/7 strain from other enterococci, its rifampicin sistant variant was prepared [49]. ED26E/7 was administered per os daily at a dose o CFU/mL in a total dose 100 µL. Counts of ED26E/7, as well as other enterococci, w enumerated after standard microbiological dilution of feces and jejunum (jejunum homogenized using Masticator, Spain) and plated on BHI agar enriched with rifamp (100 µg), and M-Enterococcus agar. Counts were expressed in CFU/g (log10) ± SD. S pling occurred at the start of the experiment (n = 20) and on days 7 and 30.

Identification of E. durans Strains
The total enterococcal count in the screened cheeses reached 3.93 CFU/g (log 1 1.98 on average. Based on MALDI-TOF mass spectrometry evaluation, strains were a ted to the species E. durans with evaluation scores from 1.781 up to 2.245 (Figur Among the nine allotted colonies, five strains (ED24E/9, ED25E/6, ED26E/1, ED26E/7 ED7E/9) were selected for the next testing. Four identical colonies were excluded. E rans ED26E/7 reached a score of 2.125, ED7E/9 had a score of 1.781, ED26E/1 possess score of 1.921, ED25E/6 was scored with a value of 2.245 and ED24E/9 had a score of 2 To support taxonomical allotment, phenotyping was provided, and the results were c pared with the type strain E. durans ATCC 19432. In these five strains, the N-acetyl cosamine test was positive, as well as the fermentation of galactose, D-glucose and lact while mannitol, inulin and sorbitol were not fermented (they were negative), as we raffinose. The Voges-Proskauer test was positive; the hydrolysis of esculin showed a bious reaction and hippurate hydrolysis as well. E. durans strains produced the enzyme alkaline phosphatase at a slight amoun nmoL, similar to lipase, leucine arylamidase, valine arylamidase, cystin arylamidas chymotrypsin, trypsin and acid phosphatase. The enzyme β-glucuronidase was not duced (except for in ED24E/9), and α-mannosidase and α-fucosidase were produced nmoL or not produced. However, beneficial hydrolase, β-galactosidase was produce ED26E/7 strain at the amount 10 nmoL; the other strains either produced 5 nmoL or not produce this enzyme. Esterase, esterase lipase and naphtol-AS-Bi-phosphohydro E. durans strains produced the enzyme alkaline phosphatase at a slight amount, 5 nmoL, similar to lipase, leucine arylamidase, valine arylamidase, cystin arylamidase, α-chymotrypsin, trypsin and acid phosphatase. The enzyme β-glucuronidase was not produced (except for in ED24E/9), and α-mannosidase and α-fucosidase were produced in 5 nmoL or not produced. However, beneficial hydrolase, β-galactosidase was produced by ED26E/7 strain at the amount 10 nmoL; the other strains either produced 5 nmoL or did not produce this enzyme. Esterase, esterase lipase and naphtol-AS-Bi-phosphohydrolase were also produced in amounts of 5 or 10 nmoL. The amount of β-galactosidase was the highest in the ED26E/7 strain (10 nmoL).

In Vitro and In Vivo Safety Control (Virulence Factor)
ED26E/7 did not possess genes for hyl and IS 16 elements. It also did not possess the GelE gene. ED26E/7 survived well in feces of Balb/c mice; at day 30, it reached almost 10 5 CFU/g (log 10) (4.61 ± 0.14 CFU/g). It does not cause the mortality of mice. Table 2. (a) Inhibition activity of durancin substance produced by ED26E/7 strain against the principal strains (EA5, SA5) and Listeria spp. in arbitrary unit per mL (AU/mL) ± SD. (b) Inhibition activity of the durancin substance produced by ED26E/7 against enterococci in arbitrary unit per mL (AU/mL) ± SD. (c) Inhibition activity of the durancin substance produced by ED26E/7 against enterococci in arbitrary unit per mL (AU/mL) ± SD. (d) Inhibition activity of the durancin substance produced by ED26E/7 against various staphylococci from ewe's milk lump cheeses and raw goat milk in arbitrary unit per mL (AU/mL) ± SD.

Discussion
Enterococcal counts in ewe's milk cheese were similar as, e.g., in fermented meat products [50], but a lower count (1.82 CFU/mL, log10) was detected in raw goat milk [5]. Lihui Du et al. [24] isolated anti-listerial bacteriocin, the durancin GL producer of which E. durans 41D was isolated from Hispanic style cheese. In this study, three out of five identified strains showed identification scores corresponding to secure genus identification/probable species identification (ED26E/7, ED25E/6, ED 24E/9); the scores of the strains ED7E/9 and ED26E/1 corresponded with probable genus identification. However, phenotypic biochemical testing supported characteristics for the species E. durans [31].
To evaluate strains as safe, no antibiotic resistance and/or virulence factor genes can be present, or their phenotype has to be assessed, as well as no or slight production of damaging enzymes (disease markers). The production of intestinal disorder enzymes, such as β-glucuronidase, α-chymotrypsin or β-glucosidase, was slight in the E. durans tested. N-acetyl-β-glucosaminidase was negative or only 5 nmoL; that one enzyme is required to proliferate in vivo [51]. The tested strains seem to be safe regarding enzyme evaluation. In addition, ED26E/7 produced beneficial lactase (β-galactosidase) at the value of 10 nmoL; this enzyme is used in the dairy industry for the production of lactosefree milk to be consumed by lactose-intolerant people [52]. It is known that bacterial β-glucuronidase can play a role in colon cancer; therefore, a zero (0) value for this enzyme measured in E. durans strains is also promising from a safety aspect. Enterococci considered for industrial applications should be antibiotic-resistant or have virulence factor genes absent [53]. In general, the production of gelatinase increased the pathogenicity in the animal model [54]. Furthermore, gelatinase can cleave fibrin, which was suggested to have important implications in the virulence of E. faecalis, as the secreted protease can damage host tissue and thus allow bacterial migration and spread infection. However, ED26E/7 was hemolysis and gelatinase-negative. The bacteriocin-producing species strain E. durans was also isolated from important Egyptian cheese OSY-W [55]. Pieniz et al. [56] described E. durans LAB18s isolated from Brazilian soft cheese with the antioxidant activity of its culture supernatant and intracellular extract. Further beneficial/desirable traits may include the production of antioxidants or the expression of pathogens antagonism [55]. Dvorožňáková et al. [57] reported the highest stimulative effect on phagocytosis induced by E. durans ED26E/7 strain in mice infected with Trichinella spiralis. Moreover, the respiratory burst of blood polymorphonuclear cells was stimulated, which can contribute to a decreased larval migration and destruction of muscle larvae, and therefore, reduced parasite burden in the host. The application of ED26E/7 caused a significant decrease in the number of muscle larvae of T. spiralis and showed the highest inhibition effect on female fecundity (94%) [58]. Moreover, the respected increase in macrophage's metabolic activity induced by ED26E/7 in the intestinal phase of trichinellosis augmented the hostparasite defense (damage and killing newborn larvae with reactive oxygen species from macrophages [48]). The implication of enterococci in food safety was previously reported by Franz et al. [13]. However, the principal condition is that the strain for use has to be safe [59]. In our previous studies, the anti-listerial effect of enterocin CCM4231 was reported in Saint-Paulin cheese or traditional Slovak dairy product "Bryndza" [60,61]. In Saint-Paulin cheese experimentally infected with L. monocytogenes Ohio, its initial count of 6.7 (log 10) CFU/g was decreased to 1.9 CFU/g in one week after enterocin addition. In "Bryndza", bacterial reduction (expressed in order of magnitude) was noted after Enterocin CCM4231 application.
It can be summarized that bioactive E. durans ED26E/7 was selected from ewe's milk lump cheese. This gelatinase and hemolysis-negative (γ-hemolysis) strain, most susceptible to commercial antibiotics producing 10 nmoL lactase enzyme β-galactosidase, represents a promising additive for dairy products. ED26E/7 did not possess virulence factor genes, such as hyl (hyaluronidase), the IS16 element or GelE gene. Its partially purified bacteriocin substance inhibited the growth of Gram-positive indicator bacteria; the principal indicator E. avium EA5, S. aureus SA5 and the other 16 out of 22 staphylococci, listeriae, and also 36 out of 51 enterococcal indicators were inhibited. In the experimental Balb/c mice, ED26E/7 did not cause mortality.

Conclusions
It can be concluded that E. durans ED26E/7 represents a promising beneficial strain for dairy products manufacturing as a natural, bacteriocin-producing adjunct, which can prolong the stability of the products (cheese) by influencing microbiota without negative influence on cheese quality, and with respect for the One Health concept strategy. This study also contributes new alternative approaches for safety assessment in food hygiene/technology and based on previous results, also in feed or food-producing animal production system.