Technological properties and probiotic potential of Enterococcus faecium strains isolated from cow milk



Kolawole Banwo, Department of Microbiology, University of Ibadan, P.M.B 1 Ibadan, Oyo State, Nigeria. E-mail:



To identify enterococci from the fermentation of milk for the production of nono, an African fermented dairy product, to determine the technological properties for suitability as starter cultures and safety as probiotics.

Methods and Results

Enterococcus faecium CM4 and Enterococcus faecium 2CM1 were isolated from raw cow's milk. The strains were phenotypically and genotypically identified. Technological properties, safety investigations, in vitro adherence properties and antimicrobial characteristics were carried out. Strong acidification and tolerance to bile salts were recorded. The strains were bile salts hydrolytic positive and no haemolysis. There was no resistance to clinically relevant antibiotics. The strains exhibited adherence to human collagen type IV, human fibrinogen and fibronectin. The bacteriocins were active against Bacillus cereus DSM 2301, Bacillus subtilis ATCC 6633, Micrococcus luteus and Listeria monocytogenes. Bacteriocins were stable at pH 4–9 and on treatment with lipase, catalase, α-amylase and pepsin, while their activity was lost on treatment with other proteases. The bacteriocins produced were heat stable at 100°C for 10 min. The bacteriocin produced by the strains was identified as enterocin A.


The E. faecium strains in this study exhibited probiotic activity, and the safety investigations indicate their suitability as good candidates for a starter culture fermentation process.

Significance and Impact of the Study

The use of bacteriocin-producing E. faecium strains as starter cultures in fermented foods is beneficial but, however, their safety investigations as probiotics must be greatly emphasized.


Enterococcus species are Gram-positive bacteria, facultative anaerobic cocci that occur singly, in pairs or in chains (Giraffa 2003). They are commensal micro-organisms that colonize in the gastrointestinal tract of humans and animals and are also found in several different food sources, such as meats, milk and cheese. These bacteria are able to survive extreme environments, such as 6·5% NaCl, pH of 9·6, high heat as well as being able to grow and survive under other harsh environmental conditions, like those found in various soils, surface water, raw plants and animal products (Giraffa 2003; Johnston and Jaykus 2004).

Enterococci are utilized in many diverse roles in a multitude of different processes. For instance, they serve as an important contributor in the ripening and flavour enhancement of several types of food such as cheeses and sausages (Giraffa 2003). In dairy products, their major role is in the development of organoleptic characteristics during the development and maturation process (Giraffa 2003; Cocolin et al. 2007). These bacteria have also been utilized as probiotics to improve the microbial balance of the intestine and to treat gastroenteritis in humans and animals (Kayser 2003). Furthermore, enterococci harbour some useful biotechnological and functional properties, such as production of antimicrobials with antilisterial activity (Foulquie-Moreno et al. 2003; Cocolin et al. 2007).

The antimicrobial properties of the strains have been ascribed to the production of bacteriocins (Todorov and Dicks 2006), which are defined as small proteins or peptides with bactericidal or bacteriostatic activity against genetically closely related species (Klaenhammer 1993). There have been published reports on bacteriocin-producing enterococci, mainly among the strains of Enterococcus faecium associated with food ecosystems and dairy products (Cocolin et al. 2007).

There have been numerous reports on bacteriocin-producing bacteria, primarily among strains of lactic acid bacteria (LAB) associated with food systems (Park et al. 2003), but there is little information on Enterococcus species from African fermented food sources (Yousif et al. 2005) especially traditional fermented dairy products (Olasupo et al. 1999; Yousif et al. 2005). Nono is fermented cow milk taken with fura which is made from ground millet. Hence, the name Fura de nono, African fermented milk curd. This is consumed mainly by the northern region of Nigeria and sub-Saharan Africa. The milk is either consumed immediately after milking from the cow or left to the next day. The method employed by the indigenous people in the preservation of the milk is usually back slopping (Eka and Ohaba 1977). Hence, the importance in the development of starter culture fermentation process to ensure consistency in the product and safety.

Bacteriocin production by LAB is believed to confer an ecological advantage over competitors present in the same ecosystem, as Enterococcus species form part of the autochthonous flora of the intestinal tract of humans and animals (Fuller 1989; Javed et al. 2011) and have also been isolated from fermenting plant and dairy products (Villani and Coppola 1994), they show potential for application as probiotics. Their antibiotic susceptibility profile also ensures their safety as probiotics especially the absence of antibiotics and multidrug resistance genes (Yousif et al. 2005). Enterococci from food sources have relatively low virulence, and there are some beneficial activities of some strains; they are also considered as nosocomial pathogens, which cause bacteraemia, endocarditis and other infections (Ben-Omar et al. 2004).

Production of bacteriocins by enterococci maybe an advantage in the use as a probiotic because the bacteriocins add up to antimicrobial effect of other substances such as hydrogen peroxide and organic acids and contribute to the competition in the gastrointestinal tract (Strompfova and Laukova 2007). LAB strains that possess certain functional properties such as tolerance to bile salts, acidification, absence of antibiotic resistance genes, in vitro adhesion capabilities and also production of bacteriocins render them good candidates for starter culture fermentation process and safety as probiotics (Giraffa 2003; Yousif et al. 2005; Strompfova and Laukova 2007; Javed et al. 2011).

In our study, we evaluated the probiotic potentials and the bacteriocin production of two newly isolated Ent. faecium strains from fermenting raw cow's milk in Nigeria. The aim of the study was to investigate the probiotic properties, functionality and safety of the Ent. faecium strains, to employ them as starter cultures for the production of nono African fermented dairy product.

Materials and methods

Identification of bacteriocin-producing strains

Seven isolates from raw milk were characterized phenotypically as Enterococcus species. The strains were grown in De Man, Rogosa, Sharpe agar (Oxoid, Basingstoke, UK) at 37°C for 24–48 h. They were presumptively identified by the following tests: observation of colonial characteristics and cell morphology, Gram staining, catalase, growth at 15 and 45°C, growth in the presence of 6·5% NaCl and at pH 9·6 and fermentation of wide range of sugars. Genetic identification to species level was performed by 16S rRNA sequencing, as described by Brosius et al. (1978) and Kostinek et al. (2005).


The Enterococcus faecium strains were inoculated (1% of an overnight culture) into MRS broth adjusted to pH 6·75 before autoclaving (pH 6·54 after autoclaving) and grown aerobically at 30°C. Acid production was determined by measuring the pH of the culture after 6, 24 and 48 h. The MRS broth medium for all acid production tests was prepared from a single batch that was pH adjusted and then dispensed into tubes of 10 ml each before autoclaving (Kostinek et al. 2005).

Resistance to bile salts

The ability of the strains to grow in the presence of bile was determined according to the method of Walker and Gilliland (1993) as modified by Vinderola and Reinheimer (2003). Each strain was inoculated (2% v/v) into MRS broth with 0·3, 0·5 and 1% (w/v) of bile (Sigma-Aldrich, Saint Louis, MO, USA). Cultures were incubated at 37°C for 24 h; optical density (OD) of the inoculated tube was measured at OD560 nm wavelength and compared with a control culture (without bile salts). This was expressed as the percentage of growth at OD560 nm in the presence of bile salts compared with the control. Each determination was carried out in triplicates.

Bile salts hydrolytic activity

Bile salt hydrolytic (BSH) plates were prepared by adding 0·5% (w/v) sodium glycodeoxycholate (GCDA) and 0·5% (w/v) taurodeoxycholate (TDCA) to MRS agar. The organisms were streaked on the plates and incubated at 37°C for 24–48 h, and the activities were determined by the precipitation of deconjugated bile salts around the colonies, which was recorded as positive (Minelli et al. 2004).

Haemolysis and production of gelatinase

The strains were cultured in MRS broth at 37°C for 12–18 h and then transferred onto blood agar (Difco, Michigan, USA) plates supplemented with 5% defibrinated whole sheep blood as described by Yoon et al. (2008). After 24–48 h, the haemolytic reaction was recorded by the observation of a partial hydrolysis of red blood cells and greening zone (α-haemolysis), clear zone around bacterial growth (β-haemolysis) and no reaction (γ-haemolysis).

To determine the presence of gelatinase activity, the plate assay method on Todd–Hewitt agar containing 30 g of gelatin l−1 was carried out as described by Ben-Omar et al. (2004).

Antibiotic susceptibility testing of the Enterococcus faecium strains

The antibiotic strips for the determination of susceptibility profile were used according to the manufacturer's instructions. The susceptibility was determined towards ten antibiotics, namely amoxycillin (25 μg), ofloxacin (5 μg), streptomycin (10 μg), chloramphenicol (30 μg), cefriazone (30 μg), gentamicin (10 μg), pefloxacin (5 μg), cotrimaxazole (25 μg), ciprofloxacin (10 μg) and erythromycin (5 μg). Tetracycline and vancomycin were tested separately in a dilution of 30 μg ml−1 and 30 μg ml−1 in sterile distilled water, respectively.

This was determined semiquantitatively using the agar overlay diffusion methods of National Committee for Clinical Laboratory Standards (NCCLS 1990). A bacterial suspension was made by picking colonies from MRS agar plates using a sterile loop and making a suspension in sterile MRS broth and left for 18–24 h. With the use of a sterile swab, the suspension was applied on the surface of Mueller–Hinton agar in sterile Petri dishes, and the strips were placed on the surface using a sterile scalpel. For tetracycline and vancomycin, a 3-mm sterile cork borer was used after which 100 μl of the antibiotic solution was dispensed into the holes. The plates were then incubated at 30°C for 48 h after which the readings were taken (Mueller and Hinton 1941; Hummel et al. 2007; Mathara et al. 2008).

In vitro adherence assay

This assay was carried out according to the study by Schillinger et al. (2005) with slight modification. Commercial precoated 96-well human fibronectin (Fn) and human collagen type IV cell culture plates were obtained from Becton-Dickinson (San Jose, CA, USA) and human plasma fibrinogen (Fb) (Sigma-Aldrich). Human plasma fibrinogen was dissolved in 0·1 mol l−1 phosphate-buffered saline (PBS, pH 7·4) at a concentration of 50 μg ml−1 and used to coat 96-well cell culture microtiter plates from Corning (Lowell, MA, USA). The strains were grown in Dulbecco's modified Eagle's medium (DMEM; Invitrogen, Beijing China) supplemented with 2% (v/v) foetal bovine serum (FBS; MDgenics, USA) for 24 h at 37°C. Cells were harvested by centrifugation at 10 000 g for 5 min and the pellets resuspended in DMEM and diluted in 10−4 dilution factor in Tris/borate/EDTA (TBE) buffer pH 8·0, and Syto 62 red (Molecular Probes, Eugene, OR, USA) was added as the fluorescent dye for detection in the flow cytometer. The cell count was determined using a BD FACS cell sorter (BD Biosciences, San Jose, CA, USA). A 1-ml aliquot of the bacterial suspension was adjusted to 1 × 108 cfu ml−1. The wells were filled with 200 μl of extracellular matrices (ECM) protein (Fb, Fn and collagen type IV) and incubated overnight at 4°C. The protein solution was removed and wells washed twice with PBS. Then, FBS (200 μl of a 2% solution in DMEM) was added to the wells to block unoccupied sites and to prevent nonspecific binding of the bacteria. The FBS was removed using a pipette after incubation at 37°C for 1 h. Two hundred microlitres of the initial bacterial suspension in DMEM was added to the ECM-coated cell culture plates. The plates were centrifuged at 2000 g for 120 s and incubated at 37°C for 1 h. The unbound bacteria were removed with a pipette into an Eppendorf tube diluted in TBE buffer and the cell count performed with a BD FACS cell sorter (USA). The number of nonadherent and adherent bacteria was established and expressed as a percentage of the initial bacterial suspension. Each determination was performed in triplicates. Probiotic strain Lactobacillus acidophilus CNRZ 1923 and Lactobacillus rhamnosus GG were used as controls.

Determination of bacteriocin production and antimicrobial spectrum

The preparation of the cell-free culture supernatant was carried out, as described by Cocolin et al. (2007). The strains Ent. faecium CM4 and 2CM1 were grown in MRS broth (Oxoid) at 37°C for 18–24 h. Cultures were centrifuged at 13 000 g for 15 min at 4°C (Eppendorf 5810 R, California, USA), and supernatants were collected and adjusted to pH 7 with NaOH and filtered through a 0·2-μm membrane filter (Millipore, Billerica, MA, USA). The antimicrobial spectrum of the cell-free culture supernatant was carried out using agar well diffusion assay, as described by Schillinger and Lucke (1989). The indicator strains were grown in appropriate media such as MRS (Oxoid) for LAB and brain heart infusion agar (Difco) for non-LAB. Wells were punctured in the soft agar with a sterile cork borer (5 mm) after solidification. These wells were inoculated with 100 μl of the cell-free supernatants of the Ent. faecium strains, and the assay plates were incubated at 30–37°C, aerobically for non-LAB and anaerobically for LAB for 18–24 h. At the end of the incubation period, the inhibition zones were detected by a clear region around the well. MRS broth without micro-organisms was used as a control. Each determination was carried out in duplicates.

Effect of enzymes, heat and pH on bacteriocin activity

The stability of the bacteriocins was tested after treatment with catalase, proteinase K, trypsin, pepsin, lysozyme, pronase E, lipase and α-amylase (Sigma-Aldrich). The enzymes were added to 1 ml of neutralized cell-free culture supernatant to a final concentration of 2 mg ml−1 in all cases, and 100 μl was inoculated into assayed plates (Cocolin et al. 2007). The plates were incubated at 37°C for 18–24 h, and the antimicrobial activity spectrum was carried out by measuring the zones of inhibition against indicator strains Bacillus cereus DSM 2301 and Bacillus subtilis ATCC 6633. The untreated cell-free supernatants were used as controls. Each determination was carried out in duplicates.

To determine the effect of pH and heat treatment on the bacteriocin activity, final pH values of the cell-free culture supernatants, ranging from 4·0 to 9·0, were prepared, using 1 mol l−1 NaOH and HCl, and the bacteriocins were treated at different temperatures and time: 121°C for 15 min, 100°C for 10 min and 30 min, 60°C for 10 min and 30 min and 45°C for 10 and 30 min (Cocolin et al. 2007). The plates were incubated at 37°C for 18–24 h, and the antimicrobial spectrum of activity was carried out by measuring the zones of inhibition against indicator strains B. cereus DSM 2301 and B. subtilis ATCC 6633. The untreated cell-free supernatants were used as controls. Each determination was carried out in duplicates.

Partial purification of the bacteriocin

For partial purification of the bacteriocins, the modified method of Callewaert et al. (1999) and Strompfova and Laukova (2007) was employed. The cultures were grown for 18–24 h at 37°C, and the supernatant was obtained by centrifugation at 10 000 g at 4°C for 15 min. The pH was adjusted to pH 7·0 with 1 mol l−1 NaOH and the supernatant cooled in an ice-water bath. Ammonium sulfate was added to the supernatant to a final saturation of 60% and held overnight at 4°C with slow and continuous stirring on a magnetic stirrer. The suspension was centrifuged at 5500 g for 30 min at 4°C (Eppendorf centrifuge 5415 R), and the precipitate (surface pellicle) was resuspended in 50 mmol l−1 sodium phosphate buffer (pH 6·5). The dissolved precipitate was desalted by dialysis for 24 h at 4°C against 50 mmol l−1 sodium phosphate buffer (pH 6·5) with two changes of buffer over a magnetic stirrer. The molecular weight cut-off of the membrane used was 1000 Da. The dialysate was resuspended in sterile Milli Q water and frozen at −20°C until use. The inhibitory activity of the partially purified bacteriocin was quantified using a modified method of Yamamoto et al. (2003) and Prasad et al. (2005). This was performed by the preparation of 10 μl of twofold serial dilution of the partially purified bacteriocin in sterile Milli Q water that was spotted onto the inoculated lawn of the indicator strain (B. cereus DSM 2301) and incubated at 30°C for 18–24 h. The bacteriocin titre was expressed as reciprocal of the highest dilution that exhibited inhibition. The activity was calculated in arbitrary units per ml (AU ml−1) using the formula: 1 AU ml−1 = 2n × (1000 μl per 10 μl−1), where AU ml−1 is the arbitrary unit per ml and n is reciprocal of the highest dilution that gave inhibition.

PCR amplification of the enterocin genes

Genes encoding for the bacteriocins were targeted by PCR using the primers and conditions described by De Vuyst et al. (2003) and Cocolin et al. (2007). The specific primers for enterocins A and B were used in the amplification. For enterocin A, the forward primer was made up of EntFa (5′ GGT ACC ACT CAT AGT GGA AA 3′) and reverse primer EntFb (5′ CCC TGG AAT TGC TCC ACC TAA 3′). PCR amplification conditions were as follows: initial denaturation for 5 min at 95°C, 30 cycles of denaturation for 30 s at 95°C, annealing at 58°C for 30 s and extension at 72°C for 30 s and a final elongation step of 5 min at 72°C.

For enterocin B, the forward primer was made up of Ent1 (5′ CAA AAT GTA AAA GAA TTA AGT ACG 3′) and reverse primer Ent2 (5′ AGA GTA TAC ATT TGC TAA CCC 3′). PCR amplification conditions were as follows: initial denaturation of 5 min at 95°C, 30 cycles of denaturation at 95°C for 30 s, annealing at 56°C for 30 s, extension at 72°C for 30 s and final extension at 72°C for 5 min.

The PCR products were subjected to electrophoresis using a 1·8% agarose gel and 1× TBE buffer containing 1 μg ml−1 ethidium bromide (Sambrook et al. 1989).


Identification of the bacteriocin-producing strains

The bacteriocin-producing strains Enterococcus faecium CM4 and Ent. faecium 2CM1 were characterized to be Gram-positive, cocci-shaped, catalase-negative bacteria that did not produce gas from glucose and grew at pH 9·6. Further confirmation was carried out by 16S rRNA sequencing as stated in the Materials and methods. The sequences are deposited in the GenBank database under the following accession numbers JN104687 and JN104688, respectively.


The strain Ent. faecium CM4 reduced the pH to 4·05, while Ent. faecium 2CM1 reduced to pH 3·85 after 48-h incubation. The initial pH of the medium was 6·54. The strains showed strong acidification properties by lowering the pH of the MRS medium to <5.

Bile salts resistance

Enterococcus faecium 2CM1 displayed a growth of 79·13 and 69·1%, while Ent. faecium CM4 exhibited a growth of 74·3 and 61·67% at 0·3 and 1% bile salts concentration, respectively. However, the probiotic strain Lactobacillus acidophilus CNRZ 1923 that was the control had a growth of 77·83 and 70·1% in the presence of bile salts at 0·3 and 1% concentrations, respectively (Fig. 1). It was observed that the Ent. faecium strains in this study generally gave good results as the probiotic strain tested.

Figure 1.

Bile salts tolerance of the bacteriocin-producing Enterococcus faecium strains and reference probiotic strain Lactobacillus acidophilus CNRZ 1923 to 0·3, 0·5 and 1% bile salts concentration. Bars represent the mean and standard deviations of three independent experiments. (image_n/jam12031-gra-0001.png) 0.30%; (image_n/jam12031-gra-0002.png) 0.50% and (image_n/jam12031-gra-0003.png) 1.00%.

Bile salt hydrolytic, haemolytic activity and gelatinase characteristics

In this study, deconjugation activity was observed on sodium glycodeoxycholate (GCDA) and TDCA, there was no haemolytic activity and gelatinase in the plates screening tests for the selected Enterococcus faecium strains (Table 1).

Table 1. Bile salt hydrolase (BSH), Haemolytic and gelatinase activities
  1. GCDA, glycodeoxycholic acid; TDCA, taurodeoxycholic acid; +: positive/precipitation of bile salts deconjugation; −/γ: negative/no haemolysi.

Enterococcus faecium CM4++γ
Ent. faecium 2CM1++γ

Antibiotics susceptibility profile of the strains

Enterococcus faecium CM4 demonstrated high susceptibility to vancomycin but was moderately susceptible to streptomycin, chloramphenicol, gentamicin, pefloxacin, ciprofloxacin and tetracycline. Enterococcus faecium 2CM1 exhibited resistance to pefloxacin, moderate susceptibility to seven antibiotics including streptomycin, gentamicin, tetracycline and vancomycin. There was no resistance to clinically relevant antibiotics in the study (Table 2).

Table 2. Antibiotics susceptibility profile of the Enterococcus faecium strains
  1. +++ (15–20 mm): highly susceptible; ++ (10–14 mm): moderately susceptible; + (1–9 mm): susceptible; −: resistance; Susceptibility: presence of a zone of inhibition (mm); Resistance: absence of a zone of inhibition (mm).

Ent. faecium CM4+ (9)++ (13)++ (12)+ (5)++ (14)++ (13)+ (5)++ (14)++ (14)+++ (18)
Ent. faecium 2CM1+ (5)+ (7)++ (11)++ (14)+ (9)++ (14)++ (13)++ (13)++ (12)++ (13)++ (14)

In vitro adherence assay

The strain Ent. faecium CM4 exhibited the highest value of 38·3%, while Ent. faecium 2CM1, the least value of 37·2% for the adhesion to human fibronectin. For human plasma fibrinogen, Ent. faecium CM4 displayed the value of 29·9% and Ent. faecium 2CM1 value of 28·9%. The strain Ent. faecium CM4 had the value of 32·1% and Ent. faecium 2CM1 exhibited the value of 36·5% for human collagen type IV. The reference strains demonstrated higher values than the selected strains for human plasma fibrinogen and human collagen type IV, while the Ent. faecium strains exhibited higher value for human fibronectin (Fig. 2).

Figure 2.

In vitro adherence of bacteriocin-producing Enterococcus faecium strains and reference probiotic strains Lactobacillus acidophilus CNRZ 1923 and Lactobacillus rhamnosus GG to immobilized fibronectin, fibrinogen and human collagen type IV. Bars represent the mean and standard deviations of three independent experiments. (image_n/jam12031-gra-0004.png) Human Fibronectin; (image_n/jam12031-gra-0005.png) Human plasma Fibrinogen and (image_n/jam12031-gra-0006.png) Human Collagen Type IV

Detection of bacteriocin produced by Enterococcus faecium CM4 and 2CM1 and the antimicrobial spectrum

The determination of the spectrum of activity of the bacteriocin produced by the Ent. faecium strains considered in this study, using LAB and non-LAB as indicator organisms, is shown in Table 3. Both strains did not show any antibacterial activity towards the representative LAB strains Enterococcus faecium ATCC 6057, Pediococcus acidilactici ATCC 8042, Lactobacillus fermentum ATCC 9338, Lactobacillus plantarum ATCC 14917, Lactobacillus pentosus JCM 1558, Lactobacillus casei CMCC 1.539 and Leuconostoc dextranicum CMCC 181, while activity was observed against Pediococcus pentosaceus ATCC 25745. In the non-LAB strains, activity was observed in Bacillus cereus DSM 2301, Bacillus subtilis ATCC 6633, Micrococcus luteus and Listeria monocytogenes for both strains.

Table 3. Antimicrobial activity of the bacteriocins produced by strains of Enterococcus faecium CM4 and Ent. faecium 2CM1
  1. American type culture collection (ATCC), Rockville, MD, USA; Chinese microbiological culture collection (CMCC), Beijing, China; Deutsche Sammlung von Mikroorganismen, Braunschweig (DSM), Germany; Japanese collection of micro-organisms (JCM), Tokyo, Japan; Zone of inhibition (mm); ND, not detectable.

Lactic acid bacteria (LAB)
 Ent. faecium ATCC 6057NDND
 Lactobacillus casei CMCC 1·539NDND
 Lactobacillus fermentum ATCC 9338NDND
 Lactobacillus plantarum ATCC 14917NDND
 Lactobacillus pentosus JCM 1558NDND
 Leuconostoc dextranicum CMCC 181NDND
 Pediococcus acidilactici ATCC 8042NDND
 Pediococcus pentosaceus ATCC 2574512·5 ± 0·1310·3 ± 0·21
 Bacillus cereus DSM 230116·4 ± 0·1416·3 ± 0·20
 Bacillus subtilis ATCC 663318·3 ± 0·2016·5 ± 0·20
 Candida albicans NDND
 Escherichia coli 0157:H7NDND
 Micrococcus luteus 12·5 ± 0·1210·1 ± 0·20
 Staphylococcus aureus ATCC 27702NDND
 Listeria monocytogenes 13·3 ± 0·1111·5 ± 0·20

Effect of enzymes, pH and temperature on bacteriocin activity

On treatment with enzymes, the cell-free culture supernatant demonstrated that the inhibitory activity of bacteriocin produced by Ent. faecium CM4 and Ent. faecium 2CM1 was completely eliminated by proteinase K and pronase E, while some activity was observed to be retained for Ent. faecium CM4 on treatment with trypsin, as shown in Tables 4 and 5. However, the activity was not affected by treatment with catalase, pepsin, α-amylase, lysozyme and lipase for both strains. The inhibitory activity of both Ent. faecium strains was retained after exposure to a pH range of 4–9. The inhibitory activity of both Ent. faecium strains was relatively stable against heat at 45–60°C (for 10 and 30 min), respectively, and 100°C for 10 min, while its activity was lost after treatment at 100°C for 30 min and 121°C for 15 min.

Table 4. Effect of pH, temperature and enzyme activities on bacteriocin produced by Ent. faecium CM 4 against Bacillus subtilis ATCC 6633 and Bacillus cereus DSM 2301
pHZIBsZIBcTemperature and timeZIBsZIBcEnzymesZIBsZIBc
  1. ZIBS: zone of inhibition for Bacillus subtilis ATCC 6633 (in mm); ZIBC: zone of Inhibition for Bacillus cereus DSM 2301 (in mm); –: no activity.

  2. Data represent the mean ± SD of duplicates.

  3. Each test was carried out at 37°C for 18–24 h.

414·5 ± 0·1214·0 ± 0·2145°C for 10 min15·0 ± 0·1013·0 ± 0·12Pronase E
616·5 ± 0·1115·5 ± 0·1345°C for 30 min15·0 ± 0·1313·0 ± 0·20Trypsin<10·0 ± 0·10
716·0 ± 0·1315·5 ± 0·1060°C for 10 min13·0 ± 0·1011·0 ± 0·12Proteinase K
912·0 ± 0·1411·0 ± 0·1260°C for 30 min13·0 ± 0·2011·0 ± 0·21Pepsin13·0 ± 0·2213·0 ± 0·20
   100°C for 10 min12·0 ± 0·2011·0 ± 0·13α-Amylase17·0 ± 0·1016·0 ± 0·12
   100°C for 30 minLysozyme16·5 ± 0·1514·0 ± 0·12
   121°C for 15 minLipase16·0 ± 0·2015·5 ± 0·15
      Catalase16·0 ± 0·1514·0 ± 0·20
Table 5. Effects of pH, temperature and enzyme activities on bacteriocin produced by Enterococcus faecium 2CM1 against Bacillus subtilis ATCC 6633 and Bacillus cereus DSM 2301
pHZIBsZIBcTemperature and timeZIBsZIBcEnzymesZIBsZIBc
  1. ZIBS: zone of inhibition for Bacillus subtilis ATCC 6633 (in mm); ZIBC: zone of Inhibition for Bacillus cereus DSM 2301 (in mm); –: no activity.

  2. Data represent the mean ± SD of duplicates.

  3. Each test was carried out at 37°C for 18–24 h.

415·0 ± 0·2015·0 ± 0·1245°C for 10 min14·0 ± 0·2014·0 ± 0·10Pronase E
616·0 ± 0·1415·5 ± 0·2145°C for 30 min14·0 ± 0·1012·0 ± 0·15Trypsin
716·5 ± 0·1016·0 ± 0·1060°C for 10 min14·0 ± 0·1212·0 ± 0·12Proteinase K
916·0 ± 0·1216·0 ± 0·1260°C for 30 min11·0 ± 0·1312·0 ± 0·10Pepsin13·5 ± 0·1013·0 ± 0·10
   100°C for 10 min11·0 ± 0·1211·0 ± 0·10α-Amylase16·0 ± 0·1215·5 ± 0·15
   100°C for 30 minLysozyme16·0 ± 0·1516·0 ± 0·12
   121°C for 15 minLipase16·0 ± 0·1016·5 ± 0·10
      Catalase16·0 ± 0·1216·0 ± 0·20

Effect of different treatments on the partially purified bacteriocin

The activity of the control supernatants of the partially purified bacteriocins from the two Ent. faecium strains was determined to be 6400 AU ml−1. The results obtained after treatment with enzymes indicated that the bacteriocin activity was completely lost after treatment with proteinase K, pronase E and trypsin for both strains. There was no effect on the activity of the bacteriocin as observed after treatment with catalase, lipase, lysozyme and α-amylase. The activity was retained on treatment with pH values 4–9 for Ent. faecium 2CM1, but half of the original activity was observed for Ent. faecium CM4 at pH values 8 and 9. The retained activity observed for both strains was not affected by heat treatment at 45–60°C (for 10 and 30 min), respectively, and at 100°C for 10 min but completely lost its activity at 100°C for 30 min and 121°C for 15 min (Table 6).

Table 6. Effect of different treatments on the partially purified bacteriocin produced by Enterococcus faecium CM4 and Ent. faecium 2CM1
TreatmentResidual activity (AU ml−1)
Ent. faecium CM4Ent. faecium 2CM1E
  1. –: no activity.

  2. Bacillus cereus DSM 2301 was used as the indicator strain.

  3. AU ml−1: Arbitrary Unit per ml defined as reciprocal of highest dilution showing inhibition of indicator strain × 100.

  4. Each test was carried out at 37°C for 18–24 h and the residual bacteriocin activity determined.

  5. Data represent test in duplicates (n = 2).

Pronase E
Proteinase K
Temperature (min)
45°C (10)64006400
45°C (30)64006400
60°C (10)64006400
60°C (30)64006400
100°C (10)64006400
100°C (30)
121°C (15)

PCR amplification of the bacteriocin genes

The DNA of Ent. faecium CM4 and 2CM1 with primers for enterocins A and B was amplified using PCR. Enterocins A and B were targeted by the specific primers. The PCR products obtained were for enterocin A, while there was no amplification for enterocin B. The results got after gel electrophoresis are shown in Fig. 3. The result highlighted here describes the presence of enterocin A structural gene in the two Ent. faecium strains studied.

Figure 3.

Polymerase chain reaction (PCR) products obtained from the amplification using the specific primer for enterocins A and B. Lane 1: molecular ladder (50 bp); lanes 2 and 3: enterocin A amplification for Enterococcus faecium CM4 and Ent. faecium 2CM1; lanes 4 and 5: no amplification for enterocin B for Ent. faecium CM4 and Ent. faecium, respectively.


Adequate phenotypic and genotypic characterization was employed in the identification of the Enterococcus faecium strains as reported by Banwo et al. (2012). The ability of the LAB strains to produce higher acid by stronger reduction in pH of the growth medium ascertains them as a candidate for good starter culture fermentation process (Kostinek et al. 2007; Oguntoyinbo 2007), and ability to withstand bile at different concentrations satisfies the selection as probiotic candidates (Vinderola and Reinheimer 2003; Schillinger et al. 2005). Investigation on enterococci of dairy origin states the poor acidifying capacity of these organisms in milk with only a small percentage of the strains showing a pH below 5·0 after 24 h of incubation at 37°C (Andrighetto et al. 2001; Sarantinopoulous et al. 2001). The Ent. faecium strains isolated in our study were good acidifiers with reduction of pH in the range 3·85–4·05 after 48-h incubation. An acidifying potential in skim milk with a pH lowering to about 4·5 after 24-h fermentation has also been reported in Enterococcus faecalis strains isolated from artisan Italian cheese (Giraffa 2003; Foulquie-Moreno et al. 2003). Resistance to bile salts is an important criterion for the selection of probiotic bacteria, which is a prerequisite for the colonization and metabolic activity of the strain in the small intestine of the host (Liong and Shah 2005; Strompfova and Laukova 2007). The mean survival ability of the tested strains to bile was observed to be quite high comparable to the probiotic strain tested.

No haemolytic activity was observed for the Ent. faecium strains. Haemolysin plays an important role in enterococcal virulence, as it may increase the chance of the infection (Morandi et al. 2006). Gelatinase activity was not detected in the isolates by plate screening assay. This agrees with the results of Mannu et al. (2003) and Franz et al. (2001). Gel genes may be silent, and phenotypes may be negative, even in the presence of a Gel gene (Franz et al. 2001; Yoon et al. 2008). Enterococcus faecium has been reported to be free of virulence determinants, but there have been notable exceptions (Yoon et al. 2008). The gelE gene has been shown to be present more frequently in clinical isolates than in noninfectious strains (Yoon et al. 2008). Eaton and Gasson (2001) reported that Ent. faecium strains isolated from dairy products and ham could test positive for one or more virulence determinants. This determinant is found at similar frequencies in the Ent. faecalis and Ent. faecium strains and appears in all culture types, including isolated food, starter and medical cultures (Yoon et al. 2008).

Enterococci are well known to be intrinsically resistant to cephalosporins, β-lactams, sulfonamides and low levels of clindamycin and aminoglycosides (Yousif et al. 2005). Multiple resistant strains of enterococci had emerged in the last decade, which showed resistance to tetracycline, chloramphenicol and vancomycin (Yousif et al. 2005). High levels of antibiotics resistance are related to a combination of nonprescription antibiotic usage and the circulation of resistant isolates in the environment with limited sanitation facilities (Al-Jabouri and Al-Meshhadani 1985). According to the findings of Mathur and Singh (2005), there was a possible baseline for the reference probiotic strain Ent. faecium 68, used as a probiotic for man as well as for animal and as silage inoculant. This strain isolated in the pre-antibiotic era is susceptible to erythromycin (15 μg), streptomycin/penicillin (streptopen 35 μg), gentamicin (10 μg), chloramphenicol (30 μg) and tetracycline (30 μg). It also possesses intrinsic resistance to kanamycin (30 μg), streptomycin (10 μg) and oxacillin (5 μg) (Mathur and Singh 2005).

The Enterococcus strains were moderately sensitive to streptomycin, gentamicin and ciprofloxacin. Enterococcus faecium CM4 showed resistance to erythromycin (5 μg), which agreed with the intrinsic resistance characteristics of reference probiotic strain of Ent. faecium 68. However, it was observed that Ent. faecium CM4 showed high susceptibility to vancomycin, while Ent. faecium 2CM1 moderately susceptible to this antibiotic. In another study, enterococci isolated from Portuguese dairy products such as milk and cheese were screened for gentamicin resistance (Lopes et al. 2003). Although enterococci are generally regarded as being intrinsically resistant to low levels of gentamicin, a high-level gentamicin resistance was detected in many dairy isolates (Giraffa 2003; Hummel et al. 2007).

Lactic acid bacteria are reported to bind to ECM such as human collagen, fibronectin and human fibrinogen, which may be exposed if the epithelial layer is injured (Kapczynski et al. 2000; Lorca et al. 2002; Schillinger et al. 2005). The ECM proteins binding to enterococci could be involved in tissue colonization by enterococci as shown for other Gram-positive pathogens (Wadstrom et al. 1987; Kostrzynska et al. 1992). Our results showed the ability of enterococci to bind to fibronectin at 30% in contradiction with the findings of Zareba et al. (1997) who reported that enterococci are nonbinders of relatively low binding capacities of 5% to 7%. The strongest binding capacities were observed in the human fibronectin, but binding to human collagen type IV and fibrinogen was lower for the strains tested. As reported by Morelli (2000), adherent strains of probiotic bacteria are more likely to stay longer in the intestinal tract and therefore may have a better opportunity to exhibit metabolic and immunomodulatory effects than the nonadhering counterparts. The enterococci exhibited sufficient binding rate for probiotics. Moreover, as reported by Strompfova and Laukova (2007), probiotic strains that may not be able to show in vitro and/or in vivo adhesion characteristics can still show positive effects in the hosts. In addition, the rate of equality of in vitro and in vivo capabilities is also questionable.

Numerous strains of enterococci associated with food systems, mainly Ent. faecium and Ent. faecalis, are capable of producing a variety of bacteriocin called enterocin with broad spectrum activity (Franz et al. 1996; Ennahar et al. 2001; Giraffa 2003). Enterocins usually belong to class II bacteriocins, which are small, heat stable and nonlantibiotics (Giraffa 2003). In this study, two Enterococcus strains isolated from cow's milk show antimicrobial activity against Gram-positive bacteria, including Bacillus cereus DSM 2301, Bacillus subtilis ATCC 6633, Micrococcus luteus, Listeria monocytogenes and Pediococcus pentosaceus ATCC 25745. The inhibitory activity of the bacteriocin produced by the Ent. faecium strains comprises Gram-positive bacteria including List. monocytogenes but showed limited or no activity against Gram-negative indicators. This observation is in agreement with previous workers on enterocins (Cocolin et al. 2007; Javed et al. 2011). Enterococcal bacteriocins that show antilisterial activity are of great importance in food and dairy industries (Rodriguez et al. 2000; De Vuyst et al. 2003). Olasupo et al. (1999) isolated a bacteriocin-producing Ent. faecium NA01 strain from wara, a fermented skimmed cow milk product from West Africa. Enterocins characterization and screening from Ent. faecium of different origins has been an important area of research, to enhance the safety of this strain as starter culture for food products (Foulquie-Moreno et al. 2003; Cocolin et al. 2007).

The effect of pH on enterocin production and apparent degradation was as intense as the effect on bacterial growth (Foulquie-Moreno et al. 2003). This effect was similar in our study for Ent. faecium CM4, which expressed a reduction in activity at pH 8 and 9 and still had a very good activity at pH 4–7; however, Ent. faecium 2CM1 exhibited good activity. This agrees with the work of Parente and Ricciardi (1999), who reported the bacteriocin activity for Enterocin 1146, which displayed a decrease in the early stationary phase at pH values higher than 4·5. The decrease of activity was ascribed to the adsorption of bacteriocin molecules on the cell surface of producer cells; and this depends on the pH of cell environment being more pronounced at higher pH (Parente and Ricciardi 1999; Leroy and De Vuyst 2003; Vizoso-Pinto et al. 2006). The bacteriocins produced lost its activity after exposure to heat at 100°C for 30 min but still had full activity at 100°C for 10 min. This indicates sensitivity to heat, which was also reported by Du Toit et al. (2000) and Cocolin et al. (2007) who observed that Enterocin 1170 was sensitive to heat as its activity decreased at 80 and 100°C after 30 min, and the activity was retained after 100 and 60°C for 10 min for Ent. faecium M241 and M249, respectively. pH values affecting the activity of enterocins produced are in accordance with those reported by Du Toit et al. (2000). The activity of Enterococcus over a wide range of pH may be advantageous when produced in the gastrointestinal tract, where pH levels are known to vary from pH 3 in the stomach to >pH 7·0 in the large intestine as reported by Du Toit et al. (2000) and Strompfova and Laukova (2007).

The effect of enzyme was investigated on the bacteriocin produced; it was observed that the activity of the enterocin was completely lost after the action of proteinase K and pronase E and trypsin for Ent. faecium 2CM1. This indicates that the bacteriocin produced is proteinaceous in nature. The ability to retain some activity after treatment with trypsin in Ent. faecium CM4 and pepsin for both strains agrees with the work of Cocolin et al. (2007), in which some activity was retained for strains Ent. faecium M241 and M249 after treatment with pepsin. This is in agreement with report of other works on enterocin production by Ent. faecium strains (Cocolin et al. 2007; Strompfova and Laukova 2007). However, the enterocins produced in our study retained its full activity on treatment with α-amylase, lysozyme, lipase and catalase and reduced activity for pepsin. The neutralized culture supernatants were not inactivated by lysozyme and lipase, which suggest that the Enterococcus in this study do not require a lipid or carbohydrate moiety for activity as also reported by Du Toit et al. (2000).

The genetic determinants of the bacteriocin genes for the bacteriocin-producing organisms were targeted for enterocins. The primer specific for enterocins A and B were used as described by De Vuyst et al. (2003) and Cocolin et al. (2007). The PCR product of the expected size for the enterocin produced by the bacteriocin genes was observed to be about 140 bp for enterocin A. This is in agreement with the size observed also by Cocolin et al. (2007). After amplification, it was confirmed that the bacteriocin-producing Enterococcus species (Ent. faecium CM 4 and Ent. faecium 2 CM1) had the presence of enterocin A but not enterocin B. In the works of Cocolin et al. (2007), there was the presence of both enterocins A and B in Ent. faecium M 241 and Ent. faecium M 249. The structural gene of enterocin A is widely distributed among Ent. faecium strains, whereas that of enterocin B always occurs in the presence of enterocin A as reported by De Vuyst et al. (2003). This implies that there are no transport genes for enterocin B production as reported by Franz et al. (1999) in De Vuyst et al. (2003). The authors suggested that these genes may be lost due to mutations, which could mean that the structural gene of enterocin B was lost by several strains, resulting in the observation that the structural gene of enterocin A occurs solely in some strains (De Vuyst et al. 2003). The bacteriocin genes were observed to be chromosomal because there was no plasmid isolated from the Ent. faecium strains (results not shown). This indicates that the genes for bacteriocin activities are located on the chromosomal DNA. This is in agreement with the work of De Vuyst et al. (2003); they further stated that chromosomal localization of enterocin genes can be an intrinsic characteristic, while plasmid localization is a result of acquired antibacterial potential through conjugal gene transfer.

The application and importance of bacteriocin-producing Ent. faecium strains is of immense usefulness in food and dairy industries (Giraffa 2003; Cocolin et al. 2007). The use of these strains as starter cultures, cocultures or protective cultures is becoming an important aspect that can lead to the production of safer food (Knorr 1998; Giraffa 2003; Cocolin et al. 2007). The use as a coculture was addressed by Cocolin et al. (2007) who cocultured the bacteriocin-producing Ent. faecium M 241 and Ent. faecium M 249 in skimmed milk with Lact. monocytogenes. Both enterococci strains studied by the authors showed a notable inhibition towards List. monocytogenes with delay in growth of the pathogen (Cocolin et al. 2007). In this study, the enterocins produced by the strains of Enterococcus faecium showed considerable inhibition against List. monocytogenes. It has also been reported that enterocin-producing strains of Ent. faecium is of great potential in dairy technology (Giraffa 2003). Therefore, enterocins or the producers show a potential for a diary application as biopreservatives or protective cultures as reported by Cocolin et al. (2007).

In conclusion, the Ent. faecium strains exhibited a good probiotic activity because of their tolerance to low pH, BSH activity, safety and in vitro adherence properties and the enterocins produced displayed inhibitory activities against food pathogens and spoilage micro-organisms. The use of bacteriocin-producing Enterococcus strains as starter cultures, protective agents in fermented foods and as probiotics is strain specific as reported by De Vuyst et al. (2003). The authors also reiterated that it is safer to use Ent. faecium strains than Ent. faecalis strains. The bacteriocin produced by the strains demonstrates their ability to be included in a starter culture for food and dairy fermentation.


The award of a postgraduate fellowship by CAS-TWAS to Kolawole Banwo for the research period at Institute of Microbiology, Chinese Academy of Sciences, Beijing, China, is gratefully acknowledged. The authors appreciate Dr. Jin Zhong for kindly providing some of the reference strains used in this study.