Isolation of Enterococcus faecium, characterization of its antimicrobial metabolites and viability in probiotic Minas Frescal cheese

Abstract

The aim of this study was to isolate Enterococcus faecium from raw milk samples, to characterize its antimicrobial metabolites, and to evaluate its viability in a probiotic Minas Frescal cheese. For this, antagonist activity against Listeria monocytogenes, safety aspects and biochemical, genotypic, and probiotic characteristics of the isolates were evaluated. Minas Frescal cheese was manufactured with the isolate that showed the best characteristics in vitro, and its viability in the product was evaluated. It was observed that of the 478 lactic acid bacteria isolates, only isolate E297 presented antagonist activity, genes encoding for enterocin production and absence of virulence factors. Besides that, E297 presented probiotic characteristics in vitro, and maintained its viability (8.09 log CFU mL⁻¹) for 14 days of cold storage, when it was added to cheese. Therefore, isolate E297 can be considered a promising microorganism for the manufacture of probiotic foods, especially Minas Frescal cheese.

Electronic supplementary material

The online version of this article (10.1007/s13197-019-03985-2) contains supplementary material, which is available to authorized users.

Introduction

The demand for new substances that can be used in food preservation and considered natural is a trend that has become a requirement of the consumer market. Accordingly, studies have been carried out in order to isolate new microorganisms (and/or their metabolites) that present biopreservative or probiotic characteristics and that meet this market need (Vera-Pingitore et al. 2016).

Several genera of LAB can produce bacteriocins, and raw milk is an important source of these bacteria. Among LAB, the genus Enterococcus, especially E. faecium, has an important genetic potential to produce several bacteriocins, known as enterocins. Enterocins have attracted scientific and technological interest because they exhibit antimicrobial activity against important foodborne pathogens, such as L. monocytogenes (Zommiti et al. 2018).

Studies have also evaluated the probiotic potential of strains of E. faecium for future food applications. Probiotics are defined as living microorganisms that, when administered in suitable amounts, confer benefits for host health (FAO/WHO 2001). Among their benefits, the following stand out: colonization capacity of the gastrointestinal tract; reduction of cholesterol levels; increased antibody levels; blocking the adhesion of pathogenic bacteria to epithelial tissues; inflammatory and bowel cancer treatment; and reducing symptoms of lactose intolerance (Ilavenil et al. 2015).

However, the use of enterococci in food is controversial due to the pathogenic potential of some isolates, as well as their ability to transfer genes that confer resistance to clinically relevant antimicrobials. The selection and evaluation of safety of an isolate of E. faecium can be accomplished using phenotypic and genotypic markers (Ladero et al. 2012).

Several probiotic species are used in the manufacture of dairy products, since they are considered good vehicles for these microorganisms. Minas Frescal cheese, a typical Brazilian cheese, is one of the most consumed in Brazil. It stands out for its high yield and its nutritional and physicochemical characteristics (Nunes and Caldas 2017), such as low acidity and high water activity, which contribute to the continued viability of a probiotic microorganism during the shelf life of the product and its passage through the gastrointestinal tract (Caggia et al. 2015).

Therefore, as it is important to evaluate each isolate individually to verify its potential for food application, the aim of this study was to isolate E. faecium from raw milk samples, to characterize its antimicrobial metabolites, and to evaluate its viability in a probiotic Minas Frescal cheese.

Materials and methods

Samples

Eighteen samples from a bulk raw milk tank were collected from a dairy industry located in the western region of the state of Santa Catarina, Brazil, during an experimental period of 21 months.

Bacterial strains

The bacterial strains used as control in analyses were Lactococcus lactis subsp. lactis DY13 (bacteriocin producer) and E. faecium FAIR-E178 (carrier of enterocin genes A, B, P and L50A/B); all of these strains were provided by the Microbiology Laboratory of the Department of Veterinary Medicine, Federal University of Viçosa, Minas Gerais, Brazil. In addition, the strain L. monocytogenes Scott A (microorganism indicator) was also used, which belongs to the collection of cultures at the Food Microbiology Laboratory—Department of Science and Agroindustrial Technology, Federal University of Pelotas (UFPel), Brazil.

Isolation of LAB

The raw milk samples were homogenized and submitted to decimal dilutions in peptone water (0.1% (w/v), Merck). For each dilution, 0.1 mL aliquots were inoculated on the surface of De Man, Rogosa and Sharpe agar (MRS, Merck) and M17 agar (Merck), which were incubated at 25 °C and 35 °C for 72 h and 48 h, respectively, under aerobic and anaerobic conditions. After incubation, 10 typical colonies of LAB (spherical shape of whitish color) were selected. These colonies were submitted to Gram staining and catalase reaction.

Antagonist activity

The antagonist activity was verified by the spot-on-the-lawn technique against L. monocytogenes Scott A (Fleming et al. 1975). As positive control, L. lactis subsp. lactis Dy 13 was used. The presence of an inhibition zone around the spots was considered to indicate an antagonistic effect.

Bacteriocinogenic potential

The isolates that presented antagonist activity were assessed for their bacteriocinogenic potential, according to Fleming et al. (1975). The proteases used were pepsin (porcine stomach mucosa, Sigma^®), chymotrypsin (bovine pancreas, Sigma^®), proteinase K (from Tritirachium album, Sigma^®) and trypsin (bovine pancreas, Sigma^®). The microorganism indicator used was L. monocytogenes. As negative control, ultrapure water (Milli-Q^®) was used. The production of bacteriocin-like substances was confirmed by the sensitivity of the substance produced following the testing of one or more proteases.

Biochemical identification of E. faecium

The isolates that presented bacteriocin-like substances with activity against L. monocytogenes were biochemically identified by Vitek^®2 system (BioMerieux, France), according to the manufacturer’s instructions.

Genotypic identification of E. faecium

DNA extraction

DNA was extracted as described in the manufacturer’s instructions for the Illustra™ bacteria genomic Prep Mini Spin Kit (GE Healthcare, Chalfont St. Giles, UK).

PCR

Genotypic identification of E. faecium isolates was performed by identifying the partial sequencing of the pheS gene using primers and PCR conditions described by Naser et al. (2005).

The reaction mixture was prepared with 25 μL 2X Go Taq Green Master Mix 2X (Promega Corp.^®), 0.5 μL of each primer (10 pmol), 2 μL of DNA (200 ng) and ultrapure water (Integrated DNA Technologies, Iowa, USA) to complete the final volume of 50 μL.

The sequencing of the identified gene was performed at the Molecular and Protein Analysis Unit (Experimental Research Center, HCPA, RS, Brazil). The sequences were compared those deposited in GenBank, using the BLAST algorithm (http://www.ncbi.nlm.nih.gov/BLAST).

Detection of genes encoding for bacteriocin production

The detection of entA, entP, entB and entL50A/B genes was performed by PCR (Özdemir et al. 2011). Each 25 μL reaction included 12.5 μL of 2X Go Taq Green Master Mix, 10 pmol of each primer, 1 μL of DNA, and ultrapure water to complete the final volume. The PCR products were submitted to electrophoresis on 2% (w/v) agarose gel in 0.5 × TBE (Tris borate-EDTA^®) and visualized using the UV light transilluminator L-Pix Image HE (Loccus^®, L-Pix Touch).

Rep-PCR

The isolates confirmed as E. faecium were submitted to Rep-PCR for the identification of genetic profiles using a single primer and PCR conditions as described by Dal Bello et al. (2010). Each 25 μL reaction included 12.5 μL of 2X Go Taq Green Master Mix, 50 pmol of each primer, 2 μL of DNA (200 ng) and ultrapure water to complete the final volume. Electrophoresis was performed as described in “PCR” section. The patterns of the derived bands were analyzed using BioNumerics software (Applied Maths, Inc., Austin, TX, Version 6.0).

Safety aspects

Hemolytic activity

The hemolytic activity was evaluated according to Eaton and Gasson (2001) and classified as α-hemolysis (partial lysis of red blood cells with greenish zones around the colonies) and β-hemolysis (total lysis with the presence of clear zones around the colonies).

Gelatinase, lipase and DNase production

The phenotypic evaluation of gelatinase, lipase and DNase production was performed as described by Barbosa et al. (2010). For both tests, positive results were identified by the formation of clear zones around the colonies.

Antimicrobial resistance

The susceptibility of the isolates to antimicrobials was evaluated by the agar diffusion method, as described by the Clinical and Laboratory Standards Institute (CLSI 2017). The antimicrobials (Laborclin^®, Brazil) tested were ampicillin (10 μg), penicillin G (10 μg), vancomycin (30 μg) and tetracycline (30 μg). These antimicrobials were chosen based on the literature (Renye et al. 2009; Marco et al. 2018; Bagci et al. 2019). The CLSI table for Gram-positive microorganisms was used to classify the isolates of LAB as resistant (R), intermediate (I), and susceptible (S) (CLSI 2017).

Detection of genes encoding virulence factors and antimicrobial resistance

The isolates were submitted to the evaluation of the presence of the virulence genes responsible for the production of gelatinase, hyaluronidase, aggregation substance, surface protein, cytolysin and collagen adhesion, as well as for antimicrobial resistance genes such as erythromycin (macrolide), tetracycline (tetracycline), gentamicin (aminoglycoside), terramycin (tetracycline), amikacin (aminoglycoside), kanamycin (aminoglycoside) and vancomycin (glycopeptide). The primers and the PCR conditions are shown in Table 1. Isolates that did not present genes encoding virulence factors and antimicrobial resistance were tested for survival under simulated gastric and intestinal conditions.

Table 1.

Primers and PCR conditions used to identify genes encoding for virulence factors and antimicrobial resistance production

Genes	Sequence (5′–3′)	Tm (°C)	References
gelE	TATGACAATGCTTTTTGGGAT AGATGCACCCGAAATAATATA	47	Vankerckhoven et al. (2004)
hyl	ACAGAAGAGCTGCAGGAAATG GACTGACGTCCAAGTTTCCAA	53	Vankerckhoven et al. (2004)
asa1	GCACGCTATTACGAACTATGA TAAGAAAGAACATCACCACGA	50	Vankerckhoven et al. (2004)
esp	AGATTTCATCTTTGATTCTTG AATTGATTCTTTAGCATCTGG	47	Vankerckhoven et al. (2004)
cylA	ACTCGGGGATTGATAGGC GCTGCTAAAGCTGCGCTT	52	Vankerckhoven et al. (2004)
ace	GAATTGAGCAAAAGTTCAATCG GTCTGTCTTTTCACTTGTTTC	48	Martín-Platero et al. (2009)
erma	TCTAAAAAGCATGTAAAAGAA CTTCGATAGTTTATTAATATTAG	52	Zou et al. (2011)
ermB	GAAAAGTACTCAACCAAATAA GTAACGGTACTTAAATTGTTTA	52	Rizzotti et al. (2005)
ermC	TCAAAACATAATATAGATAAAG CAAATATTGTTTAAATCGTCAAT	52	Rizzotti et al. (2005)
tetK	TTAGGTGAAGGGTTAGGTCC GCAAACTCATTCCAGAAGCA	55	Aarestrup et al. (2000)
tetL	CATTTGGTCTTATTGGATCG ATTACACTTCCGATTTCGG	50	Aarestrup et al. (2000)
tetM	GTTAAATAGTGTTCTTGGAG CTAAGATATGGCTCTAACAA	52	Aarestrup et al. (2000)
tetO	GATGGCATACAGGCACAGAC CAATATCACCAGAGCAGGCT	52	Aarestrup et al. (2000)
tetS	TGGAACGCCAGAGAGGTATTACATAGACAAGCCGTTGACC	52	Aarestrup et al. (2000)
aac(6′)-Ie-aph(2′′)-Ia	CCAAGAGCAATAAGGGCA TACACTATCATAACCACTACCG	60	Van de Klundert and Vliegenthart (1993)
aph(3′)-IIIa	GCCGATGTGGATTGCGAA AAGCTTGATCCCCAGTAAGTCA	60	Van de Klundert and Vliegenthart (1993)
ant(4′)-Ia	GGAAGCAGAGTTCAGCCAT GTGCCTGCATATTCAAACAGC	58	Ounissi and Courvalin (1987)
vanA	TCTGCAATAGAGATAGCCG CGGAGTAGCTATCCCAGCATT	60	Martín-Platero et al. (2009)
vanB	GCTCCGCAGCCTGCATGGACA ACGATGCCGCCATCCTCCTGC	62	Martín-Platero et al. (2009)
vanC1	GGTATCAAGGAAACCTC CTTCCGCCATCATAGCT	58	Biavasco et al. (2007)
vanC2	CTCCTACGATTCTCTCTT GCGAGCAAGACCTTTAAG	58	Dutka-Malen et al. (1995)
vanC1(2)	GATGGCWGTATCCAAGGA GTGATCGTGGCGCTG	60	Patel et al. (1997)
vanC1/C2	CTCCTACGATTCTCTTG CGAGCAAGACCTTTAAG	58	Biavasco et al. (2007)

Probiotic characteristics in vitro

Survival in simulated gastric and intestinal juice

The ability of the isolate to tolerate simulated gastric (pH 2.0 with 0.1 N HCl, supplemented with 1 mg.mL⁻¹ of pepsin—Sigma^®) and intestinal (pH 8.0 with 0.2% of bile salts, 0.1 N of NaOH and 1 mg mL⁻¹ of pancreatin—Sigma^®) juice was evaluated according to the protocol of Caggia et al. (2015). The results were evaluated by viable cell count at the initial time and after 1 h (gastric juice) and 4 h (intestinal juice) of incubation at 37 °C in Petri dishes with MRS agar.

Survival in simulated gastric and intestinal juice simultaneously

The ability of the isolate to tolerate simulated gastric and intestinal juice simultaneously (gastrointestinal tract) was evaluated according to the protocol of Caggia et al. (2015).

The isolate was inoculated in 10 mL of simulated gastric juice, prepared as described in the previous section. The suspensions were centrifuged at 7000 × g for 10 min, and the pellets were resuspended in 10 mL of simulated intestinal juice and incubated for 5 h under 150 rpm shaking. The viable cell counts were determined as described in the previous section. The survival rate of isolates was calculated using the equation below:

$$\text{Survival rate}\left(\%\right):\left(log\text{CFU}\text{N}1/log\text{CFU}\text{N}0\right)\times 100$$

*N1—total number of viable cells exposed to simulated gastric and intestinal juice simultaneously, N0—number of initial viable cells, before gastrointestinal exposure.

Manufacture of Minas Frescal cheese with isolate E297

The cheeses were prepared as described by Back et al. (2013). For the manufacture of Minas Frescal cheese, 2.0 L of raw milk, obtained in the western region of Santa Catarina, was submitted to slow pasteurisation (65 °C/30 min). After cooling to 35 °C, 0.25 g L⁻¹ of calcium chloride (Rica Nata^®), 9 log CFU mL⁻¹ of E297 isolate (12 h of incubation at 37 °C—log phase) (Supplementary material 1) and 0.1% of commercial coagulant (HA-LA^®) were added. After 40 min of coagulation, the clot was cut into cubes of approximately 5 cm³, followed by shaking at 30 rpm for 15 min to promote serum removal.

The salting was performed on the surface of the solid mass in the proportion of 15 g kg⁻¹ of salt (NaCl). The cheese pieces were stored in plastic containers with a cooling cap at 7 °C with humidity and temperature control for 14 days.

Viability of E297 isolate in cheese

The viability of E297 isolate in Minas Frescal cheese was evaluated at the initial time and after 3, 7 and 14 days of storage at 7 °C. To carry out the viable cell count of E297, 25 g of the sample was homogenized in 225 mL of 0.1% peptone water (Acumedia^®). Aliquots of 1 mL were diluted in 9 mL of peptone water, inoculated on MRS agar and incubated at 37 °C for 48 h.

To evaluate the survival ability of isolate E297 in cheese under simulated gastric juice conditions, 25 g of the sample was homogenized in 225 mL of 0.1% peptone water; the pH of the sample was adjusted to 2.0 (with HCl 0.1 N—Synth^®) and 0.05 mL of pepsin (25 mg mL⁻¹—Sigma^®) added per g of sample. It was kept under constant shaking at 130 rpm at 37 °C for 90 min.

To evaluate the survival ability of isolate E297 in cheese submitted to simulated intestinal juice, the pH of gastric juice was adjusted to 5.0 (with NaOH 0.1 N—Synth^®), and 0.25 mL bile salts (12 g L⁻¹ in 0.1 mol L⁻¹ of NaHCO₃) and pancreatin (2.0 g L⁻¹—Sigma^®) were added. The mixture was kept under constant shaking at 45 rpm at 37 °C for 20 min. The pH was adjusted to 6.5 and it was kept under constant shaking at 45 rpm for 90 min.

For both analyses, serial decimal dilutions and counts of viable cells were performed in Petri dishes with MRS agar at 37 °C for 48 h.

Statistical analysis

The data were submitted to analysis of variance followed by the Tukey test (p < 0.05), using ASSISTAT software 7.6 beta (2011).

Results and discussion

Isolation of LAB and antagonist activity

In the present study, 478 isolates of LAB were obtained from raw milk. Of these isolates, 307 (64%) showed antagonist activity against L. monocytogenes, of which 28 (5.8%) showed bacteriocinogenic potential, evaluated through the sensitivity of the antimicrobial substance produced for one or more proteases tested.

Among the 28 isolates that showed bacteriocinogenic potential, most were sensitive to more than one protease tested (Table 2), suggesting that several bacteriocin-like substances are produced by the same isolate (Arauz et al. 2009). However, according to these authors, evaluating only the sensitivity to proteases cannot determine the particular compounds being produced by the isolate, as each bacteriocin-like substance may be sensitive to one or more protease.

Table 2.

Sensitivity to proteases, biochemical identification by Vitek^® 2, genotypic confirmation of Enterococcus faecium and identification of the enterocin genes A, B, P and L50A/B by PCR

Isolates	Sensivity to proteases				Biochemical identification by Vitek®2	Confirmation of the E. Faecium	Enterocin genes
	Proteinase K	Pepsin	α-chymotrypsin	Trypsin
5	+	–	+	+	Enterococcus faecium	+	P, L50A/B
13	+	+	−	+	Enterococcus faecium	+	A, P, L50A/B
16	+	+	+	−	Enterococcus faecium	−	*
19	−	−	+	+	Enterococcus faecium	+	L50A/B
30	+	+	−	−	Enterococcus faecium	+	A, P, L50A/B
43	+	−	+	−	Enterococcus faecium	+	A e P
57	−	−	+	−	Enterococcus faecium	+	A, B, P, L50A/B
69	+	−	−	+	Enterococcus faecium	−	*
74	+	−	−	+	NI	*	*
89	+	+	+	+	Enterococcus faecium	+	A, L50A/B
104	+	−	+	+	Enterococcus faecium	+	A, P, L50A/B
233	+	+	+	+	Enterococcus faecium	−	*
234	+	+	+	+	NI	*	*
242	−	−	+	−	NI	*	*
276	+	−	+	−	Enterococcus faecium	+	A, P, L50A/B
297	+	−	−	−	Enterococcus faecium	+	P, L50A/B
315	+	−	+	+	NI	*	*
323	+	−	−	+	Leuconostoc pseudomesenteroides	*	*
336	−	−	+	+	Enterococcus faecium	+	P, L50A/B
340	+	−	+	+	Enterococcus faecium	+	P, L50A/B
341	+	+	+	+	Enterococcus faecium	+	P, L50A/B
342	+	+	+	+	Enterococcus faecium	+	P, L50A/B
356	+	+	+	+	Enterococcus faecium	+	A, L50A/B
428	−	+	+	+	Enterococcus faecium	+	A, L50A/B
433	+	+	+	+	Enterococcus gallinarum	−	*
434	+	+	−	+	Enterococcus faecium	−	*
453	+	+	+	+	Leuconostoc pseudomesenteroides	*	*
464	+	+	+	+	Leuconostoc pseudomesenteroides	*	*

NI not identified, * not reviewed

Biochemical and genotypic identification of E. faecium

Among the 28 LAB isolates with bacteriocinogenic potential, 20 (74%) were identified biochemically, using the Vitek^®2 system, as E. faecium (Table 2), and 16 isolates were confirmed by sequencing the pheS gene, which encodes the α-subunit of phenylalanyl-tRNA synthetase (Zhou et al. 2016), at the species level. The sequences obtained were compared with the GenBank database and identified as E. faecium.

Detection of genes encoding for bacteriocin production and Rep-PCR

All E. faecium isolates had at least one of the structural genes encoding for the enterocins A, B, P or L50A/B (Table 2). The presence of these enterocin genes among the 16 isolates of E. faecium was 93.7% for the gene encoding for enterocin L50A/B (entL50A/B), 68.7% for enterocin P (entP), 56.25% for enterocin A (entA) and 6.25% for enterocin B (entB).

Most of the isolates (93.7%) carried more than one enterocin gene. This finding can be attributed to the ability of these microorganisms to disseminate and receive genetic material, both among bacteria of the same genus and among other bacteria. The presence of several bacteriocin genes does not mean that all bacteriocins will be expressed at the same time; besides, the expression of one bacteriocin may regulate the expression of another (Perez et al. 2012).

It is interesting that only isolate E57 showed the entA/B/P/L50AB profile. However, Özdemir et al. (2011) observed this same genetic profile in 14.53% and 14% of the isolates, respectively. Only isolate E57 presented the gene encoding for entB. It is noteworthy that this isolate also presented the gene encoding for entA, which corroborates the findings of other authors, who reported that the isolates that present entB gene also carried the entA gene. According to De Vuyst et al. (2003), the enterocin B gene lacks a transport protein for translocation and therefore requires the transporter proteins encoded in the operons of other enterocins, such as enterocin A.

Figure 1 shows the similarity among the 16 E. faecium isolates evaluated in this study and isolates included in GenBank, which ranged from 90 to 99%. It is noteworthy that significant variability was present among the isolates according to the Rep-PCR analysis, which grouped the 16 isolates into four distinct genetic profiles due to the presence of enterocin genes, demonstrating the diversity among these isolates and confirming the results observed for sensitivity to various proteases. However, no relationship was observed between the profiles and the sensitivity to enzymes or between enterocin genes.

Fig. 1.

Dendrogram of the Rep-PCR of 16 isolates of Enterococcus faecium, including the molecular profiles, presence of genes encoding for bacteriocin production and identification of the species using partial sequencing of the pheS gene

A similar result was reported by Dal Bello et al. (2010) using the same primers with Rep-PCR to assess the genetic profile of enterococci isolates from cheeses and fermented meat products. These authors also did not find any relationship between the Rep-PCR profile and the presence of entA and entP genes.

Four isolates (E297, E336, E342 and E356), belonging to each genetic profile demonstrated by Rep-PCR, were randomly selected to continue the tests (Fig. 1).

Safety aspects

For isolates to be used in food matrices, it is necessary that they do not present virulence factors and/or antimicrobial resistance, in order to ensure the quality and safety of the product to the consumer. This assessment depends on the isolate, because these characteristics are specific for each isolate, even among a group of bacteria that is Generally Recognized as Safe (GRAS) (Eaton and Gasson 2001). In the present study, none of the isolates of E. faecium showed production of gelatinase, lipase, DNase and hemolysin, corroborating studies developed by Angmo et al. (2016).

In the present study, all isolates were susceptible to ampicillin (10 μg), penicillin G (10 μg), vancomycin (30 μg) and tetracycline (30 μg). Similar results were described by Prichula et al. (2013), where 100% of the 23 isolates of E. faecium from buffalo milk were sensitive to ampicillin, tetracycline and vancomycin.

Among the virulence factors that increase the ability of a microorganism to cause disease, it is important to highlight cytolysin (cylA), aggregation substance (asa1), gelatinase (gelE), surface protein (esp), collagen adhesin (ace) and hyaluronidase (hyl). According to Zommiti et al. (2018), the ace gene encodes for production of collagen adhesin. The protein encoded this gene is able to adhere to type I and IV collagens, and to laminin, and to bind the bacterium to the cell matrix, contributing to the pathogenicity in endorcaditis. According to Billstrom et al. (2008), the esp gene is responsible for the coding of proteins necessary for the initial adhesion and biofilm formation of E. faecium isolates. The main function of gelatinase is the supply of nutrients to bacteria from the degradation of the host tissue (Fisher and Phillips 2009). The hyl gene encodes a hyaluronidase homologous protein that assists in the dissemination of the microorganism in host tissues (Rice et al. 2009).

In this study, of the four E. faecium isolates evaluated, isolates E336, E342 and E356 presented the ace gene, isolate E342 carried the hyl gene and isolate E356 presented the ace, esp and gelE genes. No isolate presented cylA and asa1 gene coding for the production of virulence factors of Enterococcus spp. Only isolate E297 showed no gene coding for the virulence factors evaluated (Table 3).

Table 3.

Identification of genes encoding virulence factors and antimicrobial resistance

Gene	Isolates
	E297	E336	E342	E356
gelE	−	−	−	+
hyL	−	−	+	−
asa1	−	−	−	−
esp	−	−	−	+
cylA	−	−	−	−
ace	−	+	+	+
ernA ernB	− −	− −	− −	− −
ernC	−	−	−	−
tetK	−	−	−	−
tetL	−	−	−	−
tetM	−	−	−	−
tetO	−	+	−	+
tetS	−	−	−	−
Aac(6′)-Ie-aph(2′’)-Ia	−	+	+	+
Aph(3′)-IIIa	−	−	−	−
ant(4′)-Ia	−	−	−	−
vanA	−	−	−	−
vanB	−	−	−	−
vanC1	−	−	−	−
vanC2	−	−	+	−
vanC1 (2)	−	−	−	+
vanC1/C2	−	−	−	−

Antimicrobial resistance is one of the major safety concerns in strains of Enterococcus spp. (Zommiti et al. 2018). Therefore, the evaluation of the resistance to antimicrobials of clinical use is necessary in order to ensure safety in the application of Enterococcus isolates in foods, such as in probiotic products.

The four isolates evaluated presented a varied profile of antimicrobial resistance genes (Table 3). It was observed that 75% presented resistance genes coding for gentamicin, terramycin, amikacin and kanamycin, 50% for tetracycline and vancomycin, and none presented resistance genes to streptomycin, kanamycin and erythromycin. Only isolate E297 did not present a gene coding for the antimicrobial resistance evaluated. Different results were reported by Zommiti et al. (2018), in studies which did not identify the presence of genes coding for resistance to gentamicin and vancomycin by E. faecium isolates.

Considering that of the four isolates evaluated, only isolate E297 did not present genes that encode virulence factors or antimicrobial resistance, this isolate was selected to assess its probiotic potential.

Probiotic characteristics of isolate E297 in vitro

The ability to survive the passage through the simulated gastrointestinal tract is an important feature for probiotic candidate microorganisms, because they must be able to resist the gastric juice secreted by the stomach and the presence of bile in the small intestine (Caggia et al. 2015).

Bile tolerance is an important characteristic used for the selection of probiotic isolates. The resistance of an isolate to bile salts can be attributed to their ability to produce bile salt hydrolase (BSH), which is an enzyme responsible for bile acid deconjugation, reducing the toxic effects of bile salts on bacterial cells and increasing its survival in the intestine (Mojgani et al. 2015; Oh and Jung 2015). When the probiotic bacteria reach the gastrointestinal tract, they must be able to survive hostile conditions, such as low pH values and high bile salts concentrations.

The viable cell counts of isolate E297 after the simulation of gastric, intestinal and gastrointestinal conditions in vitro can be observed in Table 4. It was noted that isolate E297 presented 7.30 log CFU mL⁻¹ of viable cell count after the passage through the simulated gastrointestinal tract, corresponding to 88% survival rate. According to Ricciardi et al. (2014), probiotics with a survival rate in the gastrointestinal tract that is > 60% are considered potentially capable of reaching the colon.

Table 4.

Viable cell counts (log CFU mL⁻¹) of Enterococcus faecium E297 during simulation of gastric, intestinal and gastrointestinal conditions in vitro

Conditions	E297	Survival rate (%)
Gastric	4.50b ± 0.23	54
Intestinal	7.00a ± 0.36	84
Gastrointestinal	7.30a ± 0.24	88

Mean ± SD accompanied by the same superscripts letter in the column do not differ by Tukey’s test (p ≤ 0.05)

The initial concentration of isolate E297 used in the gastrointestinal simulation was 8.25 log CFU mL⁻¹. A reduction of 3.75 log CFU mL⁻¹ was observed in the initial viable cell counts in the simulated gastric juice. This reduction can be explained due to the low pH of the gastric juice, a similar condition to that found in the stomach, which is able to destroy a number of microorganisms (Mojgani et al. 2015). In the intestinal simulation, it was observed that E297 presented an increase of 2.50 CFU mL⁻¹ in the viable cell counts, which can be explained by the increase in pH creating favourable conditions for the growth of the isolate (Oh and Jung 2015).

Due to its suitable characteristics for a probiotic microorganism, isolate E297 was applied in a Minas Frescal cheese, evaluating its viability during cold storage and under simulated gastrointestinal conditions protected by a food matrix.

Viability of isolate E297 in Minas Frescal cheese

Maintaining the viability of a probiotic microorganism in food during storage period is essential. The viable cell counts of E297 were 6.58 and 8.09 log CFU mL⁻¹ at the initial time and at 14 days of storage at 7 °C, respectively (Table 5).

Table 5.

Viability (log CFU mL⁻¹) of Enterococcus faecium E297 in probiotic Minas Frescal cheese during refrigerated storage at 4 °C

	Time (days)
	0	3	7	14
E297	6.58bB ± 0.33	7.55aA ± 0.16	7.41aA ± 0.06	8.09aA ± 0.14
E297 in simulated gastric juice	6.46bB ± 0.13	6.30bB ± 0.21	6.63bB ± 0.13	7.61aB ± 0.27
E297 in simulated intestinal juice	8.53aA ± 0.09	7.68aA ± 0.31	8.02aA ± 0.44	8.13aA ± 0.41

Mean ± SD accompanied by the same lower case and upper case superscripts letter in the line and column do not differ by the Tukey’s test (p ≤ 0.05), respectively

According to the FAO/WHO (2002), counts above 6 log CFU g⁻¹ are considered necessary to promote probiotic effect. It was observed that E297 presented viable cell counts above the recommended limit and multiplied during the cold storage period of 14 days, increasing 1.51 log CFU mL⁻¹ regarding its initial count. This increase suggests that E297 adapted well to the food matrix, probably because it had been isolated from raw milk, the raw material used in the manufacture of cheese.

Regarding the difference in viable cell counts between simulation of the gastric and intestinal juice (Table 5), an increase of 2.07 and 0.52 log CFU g⁻¹ was observed at initial time and at 14 days of storage at 7 °C, respectively. This increase in the bacterial counts can be explained by the increase in pH, which favors the development of bacteria and the recovery of cells that were injured in the previous stages. According to Gomes et al. (2009), cheese has a buffering capacity that contributes to the protective effect on cell viability and survival after exposure to stress conditions.

At the end of the cold storage period (14 days), counts higher than 6 log CFU g⁻¹ were obtained in all conditions tested. This is an important result, given that probiotic microorganisms must reach the intestine in high concentration and remain viable (Ricciardi et al. 2014). It can be suggested that the food acted as a buffer, protecting the microorganism, increasing the pH value and maintaining cell viability. Thus, Minas Frescal cheese can be considered a promising food vehicle for isolate E297, as tested.

Conclusion

Isolate E297 from raw milk presented genes for bacteriocin production, did not carry genes for the virulence factors or antimicrobial resistance evaluated, and it showed probiotic characteristics in vitro. When this isolate was applied in Minas Frescal cheese it remained viable for 14 days in cold storage and increased its cell counts after passage through the simulated gastrointestinal tract, protected by the food matrix. Thus, isolate E297 is a promising candidate for use as a probiotic microorganism, and Minas Frescal cheese is a good vehicle for the incorporation of this probiotic isolate.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Compliance with ethical standards

Ethical standards

The article complies with all points of ethical standards.