Probiotic potential and safety of enterococci strains

Abstract

The aims of this research were to evaluate the safety and probiotic potential of Enterococcus spp. strains and select novel strains for future development of new functional fermented products. Bile salt hydrolase (BSH) activity, capacity of auto-aggregation and co-aggregation, hydrophobicity, tolerance to different pH values and NaCl content, mucin degradation, and antibiotic susceptibility were evaluated. Considering the preliminary probiotic features and safety, the strains were selected for complementary tests: tolerance to gastrointestinal tract (GIT) conditions, adhesion to Caco-2 cells and β-galactosidase activity, and presence of genes encoding virulence factors, antibiotic resistance, and biogenic amines were also performed for the selected strains. Enterococcus faecium SJRP20 and SJRP65 resisted well to the GIT conditions, presented low adhesion property, produced β-galactosidase although they did not present genes implicated in adhesion, aggregation, and colonization. Enterococcus faecium SJRP65 showed fewer genes related to antibiotic resistance and virulence factors and presented good functional properties, with interesting features for future application in dairy products.

Introduction

The knowledge of positive associations established between the intestinal microbiota and the human health has been expanding throughout the last decades. The most common way of delivering health-promoting microorganism to consumers is through fermented dairy products. Moreover, it is a priority for the food industry to search for novel probiotic microorganisms which could be used in innovative physiologically active and nutritive products. Probiotics are living microorganisms which confer a health benefit on the host when administrated in adequate amounts (FAO/WHO 2002).

Among several microorganisms, lactic acid bacteria (LAB) are popular as probiotic candidates due to their generally recognized safe status (GRAS). Bacteria belonging to the genera Lactobacillus and Bifidobacterium are more commonly used in the fermented food production. However, probiotic potential of several other genera of LAB, such as Enterococcus, Aerococcus, Carnobacterium, and others were also explored, due to their technological advantage in the food industry and their health promoting properties (Ashraf and Shah 2014).

Enterococcus is one of the main genera belonging to the LAB group and includes nearly 50 species (Li et al. 2018). They can play an essential technological role in several fermented food products and are part of microbiota of various commercial fermented products, such as dairy, meat, fish, sea food and vegetables, because of their specific metabolite characteristics, their ability to withstand heat stress and other adverse environmental conditions, high salinity and pH levels, and ability to grow in the presence of 40% (w/v) bile salts (Favaro et al. 2015; Penna et al. 2015; Zommiti et al. 2018). Enterococci confer technological functionality in fermented food production, such as contribution to the safety and sensory characteristics by the production of organic acids, antimicrobial compounds, and aromatic volatile compounds. Besides that, they also contribute to the texture and nutritional and beneficial value through the production of bioactive compounds and vitamins (Martín-Platero et al. 2009; Ahmadova et al. 2013; Pingitore et al. 2012; Santos et al. 2015; Amaral et al. 2017) and have attracted great research interest.

However, unlike other LAB genera, Enterococcus has not yet obtained the GRAS status (Hanchi et al. 2018). Besides, Enterococcus species are neither recommended for the Qualified Presumption of Safety (QPS) list from the European Food Safety Authority (EFSA Panel on Biological Hazards 2017). For these reasons, a new potentially probiotic Enterococcus strain should be checked towards its safety before being added in functional food. For a bacterial strain to be considered probiotic, selected microorganism must be safe, non-pathogenic, genetically stable and viable in high population (about 7–9 log CFU mL⁻¹ of product), and lack virulence and antibiotics resistance genes (De Paula et al. 2014). Following, the microorganisms must not only survive the passage through the gastrointestinal tract (GIT) and survive under the acid and bile environment, but also be able to proliferate in the intestine (Vandeplans et al. 2015), present adhesion, and colonization of the intestinal cells, as well as to present therapeutic benefits, besides showing technological and sensory properties in food systems.

Some strains, such as Enterococcus faecium M74 and Enterococcus faecium SF-68 are available as probiotic food supplements, such as Cernivet®; FortiFlora® (NCIMB 10415 Cerbios-Pharma SA, Switzerland); and Symbioflor® (Symbiopharm, Germany), and were considered safe and effective (Zommiti et al. 2018; Hanchi et al. 2018). Additionally, many enterococci strains have been described as useful for many health functions or technological application in food systems, as Enterococcus durans M4-5 (production of short chain fatty acids), Enterococcus faecium M74 and Enterococcus durans KLDS 6.0930 (reduction of cholesterol levels in serum), Enterococcus mundtii ST4SA (production of bacteriocins), Enterococcus faecium LCW 44, and Enterococcus durans 6HL (production of antimicrobial compounds against Gram-positive and Gram-negative bacteria) (Hanchi et al. 2018).

Even though a great number of commercial probiotic strains are currently available on the market, the isolation and characterization of novel strains is attractive since each strain shows different modes of action and distinguish benefits to consumer’s health. Considering that gut microbiota plays a pivotal role in host metabolism and that promotion of human health via functional food is in high demand, the aims of this study were to investigate the probiotic potential and safety features of Enterococcus spp. strains isolated from water buffalo mozzarella cheese using in vitro tests for future application in functional fermented products.

Materials and methods

LAB strains

A total of 16 Enterococcus spp. strains from the Culture Collection of Lactic Acid Bacteria of São Paulo State University, UNESP (CCLAB-UNESP, WDCM 1182), previously isolated from the processing of water-buffalo mozzarella cheese and identified by 16S rRNA gene sequencing (Silva et al. 2015; Silva 2015) were used: Enterococcus sp. (SJRP04, SJRP11, SJRP16, SJRP101, SJRP120, SJRP122, and SJRP125); Enterococcus durans (SJRP14, SJRP17, SJRP25, SJRP26, and SJRP68); and Enterococcus faecium (SJRP20, SJRP28, SJRP65, SJRP69). The strains were cultured in MRS broth (Difco, Becton Dickinson Co., Sparks, MD, USA) and stored at − 80 °C. Before their use in the assays, the LAB cultures were sub-cultured at least twice in MRS broth.

Preliminary screening for probiotic potential

Bile salt hydrolase activity

The bile salt hydrolase (BSH) activity was evaluated according to Kumar et al. (2012) with some modifications. The strains were cultivated in MRS broth for 18 h at 37 °C, and 10 μL of each strain were spotted, separately, in MRS agar plates, supplemented with 0.5% (w/v) of sodium salt of taurodeoxycholic acid (TDC; T0875, Sigma-Aldrich, St Louis, MO, USA) or taurocholic acid (TC; T4009, Sigma-Aldrich) and 0.37 g L⁻¹ of CaCl₂ (Vetec, Duque de Caxias, Brazil). Plates were incubated anaerobically at 37 °C for 72 h. The presence of white precipitated bile acid around spots was considered as a positive result. MRS plates without addition of TDC or TC were used as negative control.

Auto-aggregation, co-aggregation assay and cell surface hydrophobicity

The auto-aggregation and co-aggregation was determined according to Todorov et al. (2011) using Eq. 1, wherein OD₀ refers to initial OD and OD₆₀ refers to the OD value measured after 60 min. The ability of the LAB cells surface to adhere to hydrophobic compounds was evaluated according to the methods reported by Doyle and Rosenberg (1995). The percentage of hydrophobicity was determined by Eq. 2, wherein OD₀ refers to initial OD and the OD₃₀ refers to OD value measured after 30 min.

$$ \%\mathrm{aggregation}=\left[\left(\mathrm{O}{\mathrm{D}}_0-\mathrm{O}{\mathrm{D}}_{60}\right)/\mathrm{O}{\mathrm{D}}_0\right]\times 100 $$

(1)

$$ \%\mathrm{hydrophobicity}=\left[\left(\mathrm{O}{\mathrm{D}}_0-\mathrm{O}{\mathrm{D}}_{30}\right)/\mathrm{O}{\mathrm{D}}_0\right]\times 100 $$

(2)

Tolerance to pH and NaCl

The capability of bacterial cells to tolerate different pH values (3.0, 5.0, 7.0, and 9.0) and NaCl concentrations (0, 0.5, 1.0, 3.0, 5.0, and 7.0 (w/v): Synth, Diadema, São Paulo, Brazil) in MRS broth was performed in vitro, according to Todorov et al. (2008). All tests were realized in sterile flat-bottom 96-well micro titer plates (TPP, Trasadingen, Switzerland). Each well was filled with 180 μL of modified MRS broth (pH or NaCl) and 20 μL of LAB culture activated in MRS broth. The optical density was measured to obtain OD_600nm = 0.2. Every hour for 12 h, the OD_600nm was recorded using microplate reader Asys Expert Plus UV (Biochrom Ltd. Cambridge, UK).

Safety of LAB strains

Mucin degradation

The ability of LAB strains to degrade mucin was evaluated according to Peres et al. (2014) and Zhou et al. (2001). Escherichia coli ATCC 8739 and Salmonella enterica subsp. enterica serovar Typhimurium ATCC 14028 and Lactobacillus rhamnosus GG (ATCC 53103) were used as positive and negative control, respectively. The cultures were cultivated in MRS broth and incubated for 18–24 h. Agar plates were prepared with the basal medium containing 1.5% of agar, 0.5% of porcine mucin (M 1778, Sigma-Aldrich, mucin from porcine stomach type iii, bound sialic acid 0.5–1.5%, partially purified powder), with or without 3% (w/v) of glucose (Sigma-Aldrich). Then, 10 μL of each LAB strain were inoculated by spotting onto the surface of the agar plates and incubated at 37 °C for 48 h anaerobically. Mucin lyse zones (mucin degradation activity) around the colonies were observed.

Susceptibility to antibiotics

The LAB cultures were cultivated in MRS broth, incubated for 18–24 h to obtain 7–8 log CFU mL⁻¹, and spread onto the surface of MRS broth (Difco), supplemented with 1% (w/v) of agar. The antibiotic discs (Oxoid Ltd., Basingstoke, UK), according to the list proposed by the European Food Safety Authority (EFSA 2012), were manually placed on the surface and plates were incubated for 24 h aerobically. The diameters of the inhibition zones surrounding the disks were measured in millimeters and susceptibility of isolates was scored as resistant, moderately susceptible, and susceptible, according to the cutoff values proposed by Cavalieri et al. (2005). The strains Enterococcus faecium SJRP20 and Enterococcus faecium SJRP65, selected on the basis of preliminary screening, were further analyzed to confirm their probiotic potential.

Tolerance to simulated GIT conditions

The tolerance to simulated GIT conditions test was performed by successively exposing the strains to gastric and enteric simulated juices as described by Casarotti et al. (2017). The enterococci strains were grown for 18 h at 37 °C in MRS broth, and 1 mL of each culture (8–9 log CFU mL⁻¹) was distributed into four sterile flasks (two for the gastric phase and two for the enteric phase). The solutions simulating the gastric and enteric juices were prepared according to the method of Bautista-Gallego et al. (2013). The pH values used in gastric and enteric phases were 2.5 and 8.0, respectively. All enzyme solutions were prepared and filter-sterilized using a 0.22-μm membrane filter (Merck Millipore, Cork, Ireland) on the day of analysis. The cells were counted at the beginning (T₀) and the end of the gastric phase (T₁₂₀) and after the enteric phase (T₃₆₀). The cell count was performed by serial dilution and plating in MRS agar (Difco). The plates were incubated at 37 °C for 48 h under anaerobic conditions (Anaerobac, Probac, São Paulo, Brazil).

Adherence to Caco-2 cells

Caco-2 cell line ATCC HTB-37 (Rio de Janeiro Cell Bank, Rio de Janeiro, Brazil) was typically cultured in Dulbecco’s modified Eagle’s medium (DMEM, Sigma-Aldrich) supplemented with 20% (w/v) heat-inactivated fetal bovine serum (Cultilab, Campinas, Brazil), a mix of penicillin (100 UI mL⁻¹) and streptomycin (100 μg mL⁻¹) (Sigma-Aldrich), and 1% non-essential amino acids solution (Sigma-Aldrich) at 37 °C in a 5% CO₂ atmosphere. The adhesion assay was performed as described by Ranadheera et al. (2012) with modifications. In summary, the Caco-2 cells were seeded in a concentration of 10⁵ cells per well into 24-wells tissue culture plates and they were incubated at 37 °C in a 5% CO₂ atmosphere (Thermo Electron, Sunnyvale, CA, USA) until the formation of a monolayer (15–17 days). One day before the adhesion assays, the medium was replaced by the same medium, but without antibiotics. Before the adhesion, the monolayer was washed once with PBS (pH 7.4, 0.1 M) to remove all vestiges of antibiotics. Enterococcus faecium strains (SJRP20 and SJRP65) were cultivated in MRS broth and incubated for 18–24 h to obtain 8–9 log CFU mL⁻¹. Later, an aliquot of 1 mL was transferred to the monolayers of Caco-2 cells and incubated at 37 °C in a 5% CO₂ atmosphere during 2 h. The cells were washed individually at least three times with PBS to remove the non-adherent bacteria. Then, 1 mL of Triton X-100 (0.5% v/v, Sigma-Aldrich) was added for detaching the cells from each well. Following incubation for 5 min at 37 °C, the lysates were serially diluted and plated on MRS agar. Bacterial adhesion (%) was calculated by the rate of adhered bacteria and the total number of added bacteria.

β-galactosidase enzyme activity

The β-galactosidase enzyme activity was determined using paper discs impregnated with o-nitropherol-β-D-galactopyranose (ONPG Disks, Fluka, Buchs, SG, Switzerland), according to the manufacturer’s instructions. Salmonella enterica subsp. enterica serovar Typhimurium ATCC 14028 and Escherichia coli ATCC 25922 were used as negative and positive controls, respectively.

Presence of genes encoding adhesion, aggregation and colonization, and genes encoding virulence factors, antibiotic resistance and biogenic amines

The DNA from the selected strains were extracted using QIAgen DNeasy Blood & Tissue Kit (QIAGEN GmbH, Hilden, NRW, Germany), followed by DNA concentration estimation using NanoDrop 2000 UV-Vis spectrophotometer (Thermo Scientific Inc., Waltham, MA, USA). The screening of genes related to adhesion, aggregation, and colonization, and genes encoding virulence factors, resistance to antibiotics, and production of biogenic amines was carried out. PCR were performed, and amplified products were separated by electrophoresis in 0.8 to 2.0% (w/v) agarose gels in 0.5 × TAE buffer. Gels were stained in 0.5 × TAE buffer containing 0.5 μg mL⁻¹ of ethidium bromide (Sigma-Aldrich). All tested genes, encoded factors, and references can be found in the Supplementary Material (Tables S1 and S2).

Statistical analyses

The results were presented as mean ± standard deviation. Results were analyzed using analysis of variance (ANOVA) and the Tukey test, considering the level of significance 5% (P < 0.05), using the software OriginPro® (Northampton, USA), version 8. All tests were repeated on three independent occasions and in triplicate each time (n = 9), except for tolerance to pH and NaCl, tolerance to simulated GIT conditions and adherence to Caco-2 cells, which were repeated on three independent occasions and in duplicate each time (n = 6).

Results

Screening for probiotic potential

Bile salt hydrolase activity

Only Enterococcus sp. SJRP122 and Enterococcus faecium SJRP20 showed positive results on MRS containing the TC and TDC salts (Table 1), suggesting a potential capability to hydrolyze these salts and to reduce blood cholesterol. On the other hand, the growth of all strains of Enterococcus durans was completely inhibited. Enterococcus faecium SJRP65 showed positive results on MRS containing TC while Enterococcus faecium SJRP69 and Enterococcus sp. SJRP16 showed positive results on MRS containing TDC salts (Table 1).

Table 1 Bile salt hydrolase activity of Enterococcus spp. strains in MRS agar modified with salts of taurocholic and taurodeoxycholic acid

Auto-aggregation and co-aggregation properties

For all strains, the auto-aggregation was affected by the incubation temperature (P < 0.05), the values were superior at 4 °C for most of strains (Table 2). Enterococcus sp. SJRP11 and SJRP16, Enterococcus durans SJRP14 and SJRP17, and Enterococcus faecium SJRP20 presented auto-aggregation higher than 56%, while Enterococcus durans SJRP68, Enterococcus faecium SJRP69, and Enterococcus sp. SJRP101 and SJRP122 presented auto-aggregation average lower than 36% at the three evaluated temperatures.

Table 2 Auto-aggregation (%) of Enterococcus spp. strains tested in different temperatures: 4 °C (refrigerated storage temperature), 37 °C (optimal growth temperature), and 42 °C (fermentation temperature)

The tested strains were not able to produce inhibitory substances against pathogens (data not shown), not avoiding nor reducing the development of pathogenic microorganisms when co-aggregated. This way, the best results obtained for co-aggregation were from Enterococcus sp. SJRP101 and Enterococcus durans SJRP68, which presented low co-aggregation capacity with Listeria innocua ATCC 33090 and Listeria monocytogenes ATCC 15313, respectively (Table 3). High co-aggregation values were found for Enterococcus sp. SJRP11 and Enterococcus durans SJRP17 when co-aggregated with Listeria innocua ATCC 33090 and Listeria monocytogenes ATCC 15313, respectively (Table 3).

Table 3 Co-aggregation capacity and hydrophobic potential of Enterococcus spp. strains with indicating microorganisms

Cell surface hydrophobicity

The cell surface hydrophobicity varied among the evaluated strains (P < 0.05). Enterococcus sp. SJRP125 presented 100% of hydrophobic potential, and 60% of the tested strains presented hydrophobicity values higher than 90% (Table 3).

Tolerance to pH and NaCl

The LAB strains resisted well to pH values ranging from 5.0 to 9.0, and there was a slight OD increase in the first 2 h, indicating cell growth. Such a growth was more intense in the following 10 h of analysis (Fig. 1). However, the tested strains were not viable in the pH value 3.0, since none of them presented any OD increase during the 12 h of evaluation (data not shown). The LAB strains grew better in neutral conditions (pH = 7.0), similar to the intestine conditions. The strains Enterococcus sp. SJRP16 and SJRP68 did not grow well at pH 5.0. (Fig. 1).

Fig. 1

Growth of L. fermentum ( SJRP42, L. delbrueckii subsp. bulgaricus ( SJRP49), L. helveticus ( SJRP56, SJRP191), E. durans ( SJRP14, SJRP17, SJRP25, SJRP26, SJRP68), E. faecium ( SJRP20, SJRP28, SJRP65, SJRP69) and Enterococcus sp. ( SJRP04, SJRP11, SJRP16, SJRP101, SJRP120, SJRP122, SJRP125) strains in MRS broth at pH values 5, 7 and 9 (a, b and c, respectively), shown as OD (600 nm) measurements. The results are represented as an average of three readings.

Most of the strains tolerate the different concentrations of salt, since they presented high viabilities at temperatures of 30 °C and 37 °C (Fig. 2). Nonetheless, the tolerance was restricted at 5 °C (data not shown). Most of the tested strains did not have significant values of OD, ranging between 0.3 and 0.4 at 5 °C in all tested concentrations. The strains Enterococcus durans SJRP26 and SJRP68; Enterococcus faecium SJRP69; and Enterococcus sp. SJRP101, SJRP120, SJRP122, and SJRP125 presented OD values higher than 0.3 in the absence of NaCl after 12 h (data not shown); nevertheless, only Enterococcus faecium SJRP69 and Enterococcus sp. SJRP101, SJRP122, and SJRP125 maintained these values in the presence of 0.5% (w/v) NaCl (Fig. 2). None of the strains grew in the presence of 3% NaCl at this temperature (data not shown).

Fig. 2

Growth of L. fermentum ( SJRP42), L. delbrueckii subsp. bulgaricus ( SJRP49), L. helveticus ( SJRP56, SJRP191), E. durans ( SJRP14, SJRP17, SJRP25, SJRP26, SJRP68), E. faecium ( SJRP20, SJRP28, SJRP65, SJRP69) and Enterococcus sp. ( SJRP04, SJRP11, SJRP16, SJRP101, SJRP120, SJRP122, SJRP125) strains in MRS broth with 0.5, 1.0 and 3.0% NaCl at 30 °C (a, b and c, respectively) and in MRS broth with 0.5, 1.0 and 3.0% NaCl at 37 °C (d, e and f, respectively), shown as OD (600 nm) measurements. The results are represented as an average of three readings.

At 30 °C, the LAB strains presented growth in the absence and in concentrations until 3.0% NaCl, while in concentrations ≥ 5% NaCl, significant OD increase was not observed over the time for all tested strains. Enterococcus durans SJRP26 and SJRP68 and Enterococcus sp. SJRP101 and SJRP120 presented OD values higher than 0.3 with 5% NaCl over the 12 h of analysis, and only Enterococcus durans SJRP68 grew in 7% NaCl (data not shown).

The LAB strains grew in the absence and in concentrations until 3% NaCl at 37 °C; however, the OD increase over the time was lower than at 30 °C (Fig. 2). None of the strains tolerate concentrations of 5% and 7% NaCl at this temperature (data not shown).

Safety of LAB strains

Mucin degradation

No mucinolysis activity was detected in cultures on medium containing glucose (3%), while in medium where porcine mucin was the only energy source, Escherichia coli ATCC 8739 and Salmonella enterica subsp. enterica serovar Typhimurium ATCC 14028 exhibited mucinolytic activity. In comparison, no mucinolytic zone appeared around the colonies of the tested strains and in the negative control.

Susceptibility to antibiotics

All the 16 strains presented resistance to kanamycin and vancomycin; 93.75% of them were resistant to gentamycin, and the percentages of strains resistant to streptomycin and clindamycin were also considerable (62.5% and 50%, respectively). All strains were resistant to at least two class of antibiotics, and most of the strains (93.75%) were classified as multi-drug resistant (MDR; resistant to three or more classes of antibiotics) (Table 4).

Table 4 Profile of antibiotics susceptibility of Enterococcus spp. strains. The concentrations of antibiotics are expressed in micrograms per disc

Among the Enterococcus sp., Enterococcus sp. SJRP120 was the most sensible to antibiotics. It presented resistance to vancomycin and kanamycin, like all Enterococcus sp., was moderate sensibility to gentamycin and streptomycin and sensible to the other antibiotics (Table 4). Among the tested antibiotics, the ampicillin resulted in higher inhibition halos and higher percentage of sensible strains (43.75%) and moderate sensibility strains (56.25%). Enterococcus sp. SJRP11 and SJRP 125, Enterococcus durans SJRP14, and Enterococcus faecium SJRP28 and SJRP69 were resistant to erythromycin (Table 4).

Selection of Enterococcus spp. strains for additional tests

Enterococcus faecium SJRP20 and SJRP65 presented the best collected features to be used as probiotic, because they demonstrate production of the BSH activity, high capacity of auto-aggregation and hydrophobicity, low co-aggregation with pathogens, tolerance to different pH and NaCl, and low resistance to antibiotics, satisfying in an efficient way the desirable features in the selection process. Additional tests were realized to these two strains to complement their probiotic potential evaluation.

Enterococcus faecium SJRP20 and SJRP65 survived until 360 min of analysis (Table 5), which is equivalent to the passage of the bacteria through the GIT in human beings; the strains survived well during 2 h at pH 2.5 and besides, there was a significant reduction of 2 log CFU mL⁻¹ when compared the initial and enteric phases in both strains, the viability was higher than 8 log CFU mL⁻¹ after 360 min (Table 5).

Table 5 Viability of Enterococcus faecium SJRP20 and SJRP65 after simulated GI tract conditions (log CFU mL⁻¹)

Both strains presented low adherence to Caco-2 cells; Enterococcus faecium SJRP20 presented 5.70% ± 0.90 of adhesion capacity, while Enterococcus faecium SJRP65 presented 8.42% ± 0.81. Additionally, both strains demonstrated the capacity to produce β-galactosidase enzyme, which is a beneficial attribute for the probiotic potential LAB.

Enterococcus faecium SJRP20 and SJRP65 did not present any genes implicated in adhesion, aggregation and colonization (Table S1). Regarding antibiotics resistance, Enterococcus faecium SJRP20 presented the genes vanC1, bcr(B), erm(B), erm(C), and ant(4′)-Ia, which encode for vancomycin, bacitracin, erythromycin, and aminoglycoside resistance (Table S2). Enterococcus faecium SJRP65 showed less genes related to antibiotic resistance, only ant(4′)-Ia and ant(3′)-III-a genes were detected. In this study, although only two genes encoding aminoglycoside resistance were detected, both strains showed phenotypical resistance to aminoglycosides (resistant to gentamycin and kanamycin), and moderately sensitive to streptomycin (Table 4).

Regarding the virulence factors, Enterococcus faecium SJRP20 and SJRP65 presented efaA gene (endocarditis antigen), and ccf gene (sex pheromones) was a feature of Enterococcus faecium SJRP20. None of the genes related to biogenic amine production (hdc1, hdc2, tdc, and odc) were detected in the tested strains.

Discussion

In the present study, Enterococcus spp. strains isolated from the processing of water-buffalo mozzarella cheese were tested for their probiotic properties and safety. Only five strains showed some BSH activity, which is an important feature, emphasizing its functional qualities as a potential probiotic strain. BSH was efficient in the hydrolysis of bile salts of glycoconjugates and tauroconjugates. The advantages of BSH enzyme-producing strains are detoxification of bile salts; ensuring a longer retention of strains in GIT; improving nutrient absorption; altering membranes (increasing resistance to bile, intestinal defensins and lysozyme); and prevention of some types of cancer (Liong and Shah 2006; Kumar et al. 2012).

Regarding the aggregation characteristics, the isolates showed auto-aggregation varying from 23.49 to 73.25%, depending on the temperature. The capacity of adherence to the intestinal mucous surface plays an essential role in the definition of a probiotic culture. The intestine colonization by probiotic strains can generate beneficial biological responses: they can influence the immune system, increase the competition with pathogens to the epithelial receptor cells of the intestine, decrease the presence of undesirable intestinal microorganisms’ due to the production of antimicrobial compounds or other factors (Adams 2010), and prevent their elimination from the GIT by peristalsis. When the adhesion occurs, and the bacterium begins to colonize the GIT, it avoids the pathogens attachment through the block of the interaction with specific cell receptors or inhibiting its fixation by enteric interactions (Kos et al. 2003).

The co-aggregation of Enterococcus spp. strains with LAB varied from 38.97 to 72.37% and the co-aggregation with Listeria sp. varied from 24.24 to 67.59%. A high co-aggregation of tested strains with Listeria sp. is undesirable, once it can facilitate the permanence of pathogens in the GIT. Besides the fact that LAB produce several organic acids, which has a certain antimicrobial effect, Listeria spp. are acid-tolerant, and these changes in pH cannot interfere their viability. Similar results are available in the literature (Pingitore et al. 2012; Todorov et al. 2014). These low values of co-aggregation with Liseria monocytogenes may avoid biofilm formation and elimination of the pathogenic agent from the gut.

The cell surface hydrophobicity of Enterococcus spp. strains varied from 58.54 to 100.0%, and according to Taheri et al. (2009), when bacteria show high aggregation, they also have high cell surface hydrophobicity. Other strains isolated from the same sample (water-buffalo mozzarella cheese) also presented high levels of hydrophobicity (Jeronimo-Ceneviva et al. 2014). The hydrophobicity of the bacterial surface can be associated to the growth of bacteria in hydrophobic substrates, to the biofilm formation and to the adhesion, aggregation, and cell flocculation to host cells. Thus, the strains with high levels of cell hydrophobicity can have an easier access to the soluble materials and organic matter linked to the intestinal mucous surface, facilitating the food digestion (Sánchez-Ortiz et al. 2015).

Regarding the tolerance to varied pH and concentration of NaCl, most enterococci strains resisted well to pH values ranging from 5.0 to 9.0 and tolerated the different concentrations of salt. This finding agrees with the Todorov et al. (2011) study; however, in contrast to that observed by Acurcio et al. (2014). The results of tolerance to NaCl are not surprising, since it is well-documented that Enterococcus spp. have tolerance to the presence of NaCl up to 6.5% (Devriese et al. 2006). This tolerance is an important feature for its use in dairy products, especially in cheese (de Paula et al. 2014).

All the tested LAB strains were unable to degrade the mucin, which is a crucial safety feature for probiotic strains. The mucin is a layer of mucous that coats the surface of the GIT of animals and plays an important role in the barrier system of the mucous. Whatever damage or disturbance of this mucin layer will compromise the defense function of the host (Zhou et al. 2001).

The other safety assessment of probiotics concerns about antibiotic resistance, especially because starter cultures used in fermented foods can be potential reservoirs of resistant genes; therefore, the risk of transfer of these genes to other bacteria is increased (Hummel et al. 2007). Then, a strain resistant to antibiotics must be carefully evaluated before to be applied as a commercial probiotic strain. Thus, it is always necessary to analyze the presence of antibiotic resistance genes in probiotic cultures. In this study, the LAB strains presented variable resistance to antibiotics. Some strains were classified as multi-drug resistant and Enterococcus sp. SJRP120 was the most sensible to antibiotics. Resistance to erythromycin can be a matter of concern because macrolides are common substitutes for individuals with penicillin allergy.

Among the tested strains, Enterococcus faecium SJRP20 and SJRP65 were distinguished for their higher probiotic potential and low resistance to antibiotics and were submitted to additional tests. According to the results from the assay evaluating simulated GIT conditions, both strains tolerated well (> 8 log CFU mL⁻¹) to subsequent exposure to acidic pH and bile salt. Studies conducted by Jeronimo-Ceneviva et al. (2014) using LAB strains isolated from the same source demonstrated lower cell viability under GIT, from 2.79 log CFU mL⁻¹ for Leuconostoc mesenteroides subsp. mesenteroides SJRP58 to 4.08 log CFU mL⁻¹ for Leuconostoc citreum SJRP44. When probiotic bacteria are ingested, sufficient numbers of metabolically active bacteria must overcome the GIT barrier and transitorily persist in the GIT to exert their beneficial effects.

The adherence to Caco-2 cells was low, which is similar to that reported by Todorov et al. (2011) for Enterococcus faecium (ET 05, ET 12, and ET 88) and slightly lower the reference Lactobacillus rhamnosus GG (11.3%), using the same method. High ability to adhere to mucosal surfaces in the intestine plays an important feature because the strain may be able to activate the genes encoding antimicrobial compounds, such as bacteriocins, which may act against pathogens present in the GIT (De Paula et al. 2014). In contrast, the low capacity of adhesion to intestinal cells can result in reduced time for the excretion of probiotics in feces (Casarotti et al. 2014).

Enterococcus faecium SJRP20 and SJRP65 were shown to be β-galactosidase positive, which is a common feature of the genera Enterococcus (Todorov et al. 2014). The ability of producing β-galactosidase has a great importance for the consumer health, and for this reason, the dairy industry has invested in searches by β-galactosidase enzyme producer strains to help consumers with problems of intolerance to lactose. Additionally, studies have demonstrated that people with hypolactasia can consume fermented milk containing β-galactosidase enzyme producer strains, because it hydrolyzes the lactose and helps the absorption of the monosaccharides in the intestine, and consequently relieves the symptoms of the intolerance (Gheytanchi et al. 2010).

Regarding the presence of antibiotic resistance genes, the results observed for Enterococcus faecium SJRP20 and SJRP65 were similar to the reported by Gaglio et al. (2016). Besides the presence of vancomycin resistance gene vanC1 in Enterococcus faecium SJRP20, the other vanA, vanB, vanC2, and vanC3 genes investigated by PCR, resulted in negative amplification for any of the tested genes. However, both strains Enterococcus faecium SJRP20 and SJRP65 presented the resistance phenotype to vancomycin and only Enterococcus faecium SJRP20 was positive for vanC1 gene, indicating that this characteristic could be related to other resistance types, such as vanD, vanE, or vanG (Courvalin 2006). In addition, vanA and vanB are generally located in plasmid DNA, while vanC, vanD, vanE, and vanG genes are located on bacterial chromosomes (Martín-Platero et al. 2009). Thus, a major concern with the spread of antibiotic resistance through horizontal gene transfer is associated with the vanA and vanB genes, which are absent in both strains.

The gene bcr(B) was found in Enterococcus faecium SJRP20; in Enterococcus faecalis, bcr (ABC) genes were associated with high-level bacitracin resistance and clustering of genes located in a transferable plasmid (Matos et al. 2009), which may be a reason for accepting food applications.

The erm(B) and erm(C) genes, which encode for erythromycin resistance was positive for genotypic character in Enterococcus faecium SJRP20; however, the gene is silent because it was not phenotypically expressed—the erythromycin resistance was not observed (Table 4). The presence of the erm(B) gene was also reported by other authors. In the study of Guerrero-Ramos et al. (2016), the erm(B) gene was the most prevalent in strains resistant to erythromycin and was observed in 17 of the 49 isolates.

The presence of ant (4′)-Ia and ant(3′)-III-a genes in Enterococcus faecium SJRP65, which are associated with aminoglycoside resistance, is due to inactivation by cellular aminoglycoside modifying enzymes. This gene had also been detected in clinical isolates of Staphylococcus aureus (Sattari 2009). In relation to the presence of the ccf gene, although the ccf code determinant for a production of sex pheromone, Sánchez-Valenzuela et al. (2010) do not consider the presence of the ccf gene as related to the safety of strains isolated from swordfish (Xiphias gladius) due to the low incidence (4.2% among isolates). The presence of the ccf gene, as well as the presence of gelatinase, enterococcal surface protein (esp), enterococcal surface adhesion (ace), aggregation substance (agg), cytolysin operon (cylA), surface adhesin gene (efaA), and biofilm formation has been associated with enterococcal virulence factors (Chajęcka-Wierzchowska and Zadernowska 2016); however, Sauer et al. (2009) also observed the ccf gene in the control group of enterococci susceptible to vancomycin from clinical isolates. The endocarditis antigen (efaA) was also detected in enterococci isolated from chorizo, a small fermented sausage (Martin et al. 2005) and in Enterococcus durans, Enterococcus faecalis, Enterococcus faecium, Enterococcus hirae, Lactococcus lactis and Lactococcus lactis subsp. lactis isolated from raw goat’s milk (Perin et al. 2014). Although sex pheromones (ccf, cob, cpd) are not considered virulence factors, their production in enterococci may favor the dissemination of virulence determinants and promote acquisition of antibiotic resistance and other linked traits from other enterococci and lead to increased virulence (Eaton and Gasson 2001).

The presence of antibiotic resistant genes and virulence factors is a concern because they increase the ability of microorganisms for the dissemination of antimicrobial resistance (Martin et al. 2005) and to cause diseases and contribute to enhance infection risks. Enterococci are often associated with nosocomial infections and cause human diseases, such as bacteremia, endocarditis, or urinary tract infections (Hwanhlem et al. 2013). In addition, virulence factors are found as bacteria act as opportunistic pathogens (Ahmadova et al. 2013).

None of the genes related to biogenic amine production were detected in the tested strains. These results are interesting because enterococci are one of the main groups of LAB responsible for the production of biological amines in fermented dairy products (Jiménez et al. 2013). On the other hand, the presence of the genes was reported for Lactobacillus sp. (Jeronimo-Ceneviva et al. 2014; Casarotti et al. 2017) and Enterococcus sp. (Perin et al. 2014). Biogenic amines are formed as a result of bacterial decarboxylation of amino acids and may be toxic to consumers (Jiménez et al. 2013); for this reason, it is necessary to investigate whether the strains that are candidates to be applied in foods possess genes of biogenic amines.

Conclusion

Considering all the screening results, Enterococcus faecium SJRP65 was considered safe and showed the best probiotic potential among the strains evaluated. Further in vivo tests must be realized to guarantee its safety and probiotic effects prior its use in food products.