International Journal of Food Microbiology 81 (2003) 73 – 84 www.elsevier.com/locate/ijfoodmicro Applicability of a bacteriocin-producing Enterococcus faecium as a co-culture in Cheddar cheese manufacture M.R. Foulquie Moreno a,b, M.C. Rea b, T.M. Cogan b, L. De Vuyst a,* ´ a Research Group of Industrial Microbiology, Fermentation Technology and Downstream Processing (IMDO), Department of Applied Biological Sciences, Vrije Universiteit Brussel (VUB), Pleinlaan 2, B-1050, Brussels, Belgium b Dairy Products Research Centre, Moorepark, Teagasc, Fermoy, Cork, Ireland Received 25 November 2001; received in revised form 28 March 2002; accepted 25 April 2002Abstract Two strains, Enterococcus faecium RZS C5 and E. faecium DPC 1146, produce listericidal bacteriocins, so-called enterocins.E. faecium RZS C5 was studied during batch fermentation in both a complex medium (MRS) and in milk to understand theinfluence of environmental factors, characteristic for milk and cheese, on both growth and bacteriocin production. Fermentationconditions were chosen in view of the applicability of in situ enterocin production during Cheddar cheese production. Enterocinproduction by E. faecium RZS C5 in MRS started in the early logarithmic growth phase, and enterocin activity decreased duringthe stationary phase. The effect of pH on enterocin production and decrease of activity was as intense as the effect on bacterialgrowth. Higher enterocin production took place at pH 5.5 compared with pH 6.5. The use of lactose instead of glucose increasedthe production of enterocin, and at higher lactose concentration, production increased more and loss of activity decreased. Theproduction in skimmed milk compared to MRS was lower and was detected mainly in the stationary phase. When caseinhydrolysate was added to the milk, enterocin production was higher and started earlier, indicating the importance of an additionalnitrogen source for growth of E. faecium in milk. For co-cultures of E. faecium RZS C5 with the starters used during Cheddarcheese manufacture, no enterocin activity was detected during the milk fermentation. Furthermore, the applicability of E. faeciumRZS C5 and E. faecium DPC 1146 strains was tested in Cheddar cheese manufacture on pilot scale. Enterocin production tookplace from the beginning of the cheese manufacturing and was stable during the whole ripening phase of the cheese. Thisindicates that both an early and late contamination of the milk or cheese can be combated with a stable, in situ enterocinproduction. The use of such a co-culture is an additional safety provision beyond good manufacturing practices.D 2002 Elsevier Science B.V. All rights reserved.Keywords: Enterococcus faecium; Bacteriocin; Enterocin; Cheddar; Cheese1. Introduction consumers’ demand for healthy, safe and fresh food (Smith, 1993). However, chemicals such as potassium In the last decade, the interest in natural preserva- nitrate are still used in cheesemaking to prevent latetives is increasing, which is in accordance with the gas formation by Clostridium tyrobutyricum during cheese ripening. On the other hand, the risk of Listeria * Corresponding author. Tel.: +32-2-6293245; fax: +32-2- spp. contamination of milk and cheese must be mini-6292720. mised to achieve a zero tolerance policy. The use of E-mail address: email@example.com (L. De Vuyst). bacteriocinogenic lactic acid bacterial strains as starter0168-1605/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved.PII: S 0 1 6 8 - 1 6 0 5 ( 0 2 ) 0 0 1 6 7 - 8
74 ´ M.R. Foulquie Moreno et al. / International Journal of Food Microbiology 81 (2003) 73–84or co-cultures is a promising alternative in food fer- pounds (Jensen et al., 1975a; 1975b; 1975c; Trovatellimentation processes like cheese manufacture, both to and Schiesser, 1987; Coppola et al., 1988, 1990;prevent late blowing and to combat Listeria spp. (De Parente et al., 1989; Ledda et al., 1994; Villani andVuyst, 2000). Coppola, 1994; Centeno et al., 1996, 1999). On the Bacteriocins are ribosomally synthesised, extrac- other hand, enterococci are less proteolytic comparedellularly released, antibacterial peptides (De Vuyst to lactococci and lactobacilli (Jensen, et at., 1975b).and Vandamme, 1994). They display a limited inhib- Hence, they may play an important role as co-cultureitory spectrum encompassing Gram-positive bacteria, in milk fermentations for cheese manufacture, provid-in particular closely related strains. Large numbers of ing the cheese with both an organoleptic and safetybacteriocins are produced by all genera of the lactic advantage.acid bacteria (De Vuyst and Vandamme, 1994; Jack et E. faecium RZS C5 is a natural cheese isolate, whichal., 1995). In general, bacteriocins produced by Enter- produces an enterocin with high antilisterial activityococcus faecium and E. faecalis are small, hydro- (Vlaemynck et al., 1994). In this paper, enterocinphobic and thermostable peptides with an interesting production by E. faecium RZS C5 during batch fer-technological potential (Giraffa, 1995). Indeed, an mentation is studied in both a complex medium and inincreasing number of enterococcal bacteriocins milk. The aim of the study is to investigate theagainst foodborne pathogens such as Listeria spp. influence of environmental factors, characteristic ofand Clostridium spp. has been reported during the milk and cheese (Cheddar), on the kinetics of both ´last decade (Ben Embarek et al., 1994; Farıas et al., growth and bacteriocin production. Therefore, fermen-1994; Vlaemynck et al., 1994; Giraffa et al., 1994, tation conditions have been chosen in view of the1995a,b; Maisnier-Patin et al., 1996; Vlaemynck, applicability of in situ enterocin production by E.1996; Bennik et al., 1999; McAuliffe et al., 1999; faecium RZS C5 during Cheddar cheese production.Mendoza et al., 1999). Even activity towards Gram- Furthermore, this bacteriocin-producing strain is testednegative bacteria like Vibrio cholerae has been shown as well as the bacteriocin-producing strain E. faecium(Simonetta et al., 1997). DPC 1146 in Cheddar cheese manufacture on pilot Studies on in vitro enterocin production through scale.fermentation by E. faecium are scarce, except forenterocin 1146, whose production has been studiedin detail (Parente and Hill, 1992a,b,c; Parente and 2. Materials and methodsRicciardi, 1994; Parente et al., 1997). Enterocin 1146has a rapid bactericidal effect on Listeria monocyto- 2.1. Bacterial strains and mediagenes in buffer systems, broth and milk. The application of several enterocins, either as food E. faecium RZS C5 (deposited as E. faecium FAIRadditive or through in situ production by an appropri- E-171 in the Laboratory of Microbiology (LMG) Gentate starter culture or co-culture during fermentation, Culture Collection, Gent, Belgium) and E. faeciumhas been studied in cheese production (Sulzer and DPC 1146 were used as bacteriocin-producing strains.Busse, 1991; Giraffa, 1995; Giraffa et al., 1995b; E. faecium FAIR E-171 has been shown to be a safe ´˜ ´Joosten et al., 1995; Nunez et al., 1997; Farıas et al., strain, i.e. vancomycin-sensitive and cytolysin-nega-1999), and in other food products (Laukova and ´ tive (unpublished results). The enterocin-sensitive ´Czikkova, 1999; Aymerich et al., 2000; Callewaert et Listeria innocua LMG 13568 strain was used as anal., 2000). It turned out that enterocins are able to indicator for bacteriocin production. The bacteriocin-reduce the Listeria count by 2 to 9 log cycles, depend- producing and the indicator strains were propagated ating on the product and enterocin tested. However, less 37 (or 21) and 30 jC, respectively, in de Man-Rogosa-is known about the kinetics of bacteriocin production Sharpe (MRS) and Brain Heart Infusion (BHI) broth,in food ecosystems. Furthermore, it has been postu- respectively (Oxoid, Basingstoke, United Kingdom).lated that enterococci play an important role in the Lactococcus lactis subsp. cremoris 223 and L. lactisripening of different cheeses, due to their proteolytic subsp. cremoris 227 from Christian Hansen Labora-and lipolytic activities and production of flavour com- tories (Hørsholm, Denmark) were used as commercial
´ M.R. Foulquie Moreno et al. / International Journal of Food Microbiology 81 (2003) 73–84 75cheese starter cultures. Stock cultures of all strains sterilised in situ at 121 jC for 20 min. The energywere kept in the appropriate medium with 25% (v/v) source was sterilised separately, and aseptically addedglycerol, and stored at À80 jC. Strains were propa- to the fermentor. The fermentor was inoculated withgated twice in the appropriate broth before use as 1% (v/v) of a log phase culture of the enterocin-fresh bacteriocin-producing or indicator culture. Cus- producing strain. The fermentor was run at a constanttomised MRS and milk media were used to study the temperature of 21 or 37 jC, without aeration. Slowinfluence of physical and chemical factors on cell agitation (50 rpm) was maintained to keep the fer-growth and bacteriocin production by E. faecium RZS mentation broth homogeneous. The pH value (5.5 orC5. The composition of these media is presented in 6.5) was kept constant by automatic addition of 10 MTable 1. Agar media were prepared by addition of NaOH. The fermentation conditions tested are listed1.5% granulated agar (Oxoid) to the broth medium; in Table 1. At regular intervals, samples were asepti-overlay agar media contained 0.7% granulated agar. cally withdrawn from the fermentor. Cell number (colony forming units or CFU mlÀ1), biomass (grams2.2. Bacteriocin production kinetics in batch fermen- of cell dry mass or CDM lÀ1), and bacteriocin activitytations (activity units or AU mlÀ1) were determined for each sample. Biomass was determined gravimetrically after The kinetics of cell growth and bacteriocin pro- membrane filtration (0.45-Am pore-size filters, typeduction by E. faecium RZS C5 were studied in a 15 L HA; Millipore, Bedford, MA) of a known volume ofBiostatR C fermentor (B. Braun Biotech International, fermentation liquor followed by washing the filterMelsungen, Germany), containing 10 l of medium. with demineralized water and drying it overnight atThe medium was adjusted to pH 5.5 or 6.5, and 105 jC. The maximum specific growth rate (lmax)Table 1Fermentation conditions tested for bacteriocin production by E. faecium RZS C5Fermentation conditions lmax Xmax tstationary Bmax tBmax (hÀ1)a (g lÀ1)b (h)c (AU mlÀ1)d (h)eMRS (glucose, 2.0%), 37 jC, uncontrolled pH 1.0 0.8 7 1200 2–8MRS (glucose, 2.0%), 37 jC, controlled pH 6.5 0.9 2.2 6 2400 3–4MRS (glucose, 2.0%), 37 jC, controlled pH 5.5 0.3 1.1 14 3200 12 – 32MRS (glucose, 2.0%), 21 jC, controlled pH 5.5 0.1 1.2 20 2400 48 – 52MRS (lactose, 2.0%), 37 jC, controlled pH 6.5 0.8 2.3 7 4800 4 – 11MRS (lactose, 2.0%) without peptone and Lab Lemco, 0.8 1.8 9 4800 4–6 casein hydrolysate (1.8%), 37 jC, controlled pH 6.5MRS (lactose, 5.0%) without peptone and Lab Lemco, 0.8 2.5 10 9600 15 – 99 casein hydrolysate (1.8%), 37 jC, controlled pH 6.5MRS (lactose, 2.0%) without peptone and 0.9 2.2 9 6400 7 – 11 Lab Lemco, casein hydrolysate (1.8%), NaCl (2.0%), 37 jC, controlled pH 6.5Skim milk, casein hydrolysate (1.8%), nm nm nm 1200 49 37 jC, controlled pH 6.5Skim milk, 37 jC, controlled pH 6.5 nm nm nm 600 71Skim milk, 37 jC, free pH nm nm nm 200 56Skim milk, Cheddar conditions nm nm nm 150 61Skim milk, Cheddar conditions, starter cultures (L. lactis) nm nm nm 0 nrnm=not measured. nr=not relevant. a lmax=maximum specific growth rate. b Xmax=maximum biomass obtained (g of cell dry mass or CDM lÀ1). c tstationary=time at which was reached the stationary phase. d Bmax=maximum enterocin activity measured. e tBmax=time at which the Bmax was obtained.
76 ´ M.R. Foulquie Moreno et al. / International Journal of Food Microbiology 81 (2003) 73–84was calculated as the derivative of the linear regres- temperature for cheesemaking was 32 jC. Rennet wassion obtained for the logarithm of the biomass as a added (Chr. Hansen Laboratories’ Chymax Plus, 0.18function of time. The bacteriocin activity was quanti- ml lÀ1 diluted in 400 ml of sterile water) 30 min afterfied through a twofold serial dilution by the agar spot the starter addition, followed by stirring of the milktechnique (see below), using L. innocua LMG 13568 for 3 min. The rennet was allowed to set for 40 minas indicator strain. For reproducibility reasons, four before cutting, and the curd was allowed to rest for 10fermentations (those with glucose) were carried out min after cutting. The curd was then stirred andtwice. cooked to a maximum of 38.5 jC (maximum scald) at a rate of 1 jC every 5 min. Growth of the starter2.3. Bacteriocin production during fermentation si- was monitored by pH. At pH 6.18, the whey wasmulating Cheddar cheese production drained. The curd was stacked and allowed to Ched- dar until the pH reached a value of approximately To prepare a preculture, 0.1% (v/v) of a fresh E. 5.35, at which time it was milled and salted at a rate offaecium RZS C5 culture was inoculated in separate 2.7% NaCl (m/m). After salting, the curd was kept atlots of 10% heat-treated (30 min at 90 jC) reconsti- room temperature for 15 min; the curd was thentuted skimmed milk powder (RSM), which were packed in two moulds of 23 kg and pressed overnightincubated at 21 jC for approximately 16 h. The at 412 kPa. Finally, the cheese was cut in two pieces,fermentation broth was inoculated with 0.75% (v/v) vacuum-packed, and ripened at 8 jC.of the previous fermented milk. The initial pH of the Microbiological analyses of milk and cheese sam-fermentation broth was 6.52. The initial temperature ples were performed as follows. Raw milk and pas-was 32 jC, and it was kept constant for 1.5 h. The teurised milk of each vat were analysed for aerobictemperature was then increased at a rate of 1 jC every plate count (Milk Plate Count Agar, MPCA, Merck,5 min until the temperature reached 38.5 jC. The Darmstadt, Germany; 3 days at 30 jC) and coliformsincubation continued until the pH reached a value of (Violet Red Bile Agar, VRBA, Merck; 24 h at 30 jC).6.15. Then the temperature was decreased to 33 jC After addition of the starter and bacteriocin producerover the next 10 min (1 jC/2 min), and maintained strains, all milks were analysed for coliformsconstant until a pH of 5.2 was achieved. After this (VRBA), enterococci (Kanamycin Aesculin Azidetime, the pH was kept constant with 10 M NaOH. agar, KAA, Merck; 24 h at 37 jC), and lactobacilli (LactoBacillus Selective agar, LBS agar, Beckton,2.4. Cheddar cheese production on pilot scale Dickenson and Company, Cockeysville, USA; 5 days at 30 jC). Curd samples were also analysed at max- Two Cheddar cheese trials were undertaken. The imum scald, pressing and day 1 for starters (MRS andfirst trial consisted of two vats, one control and one L-M17 (M17 with lactose) agar, Difco Laboratories,with the enterocin producer E. faecium RZS C5, and Detroit, MI, USA; 3 d at 30 jC), enterococci (KAA),the second trial consisted of four vats, one control, one lactobacilli (LBS) and coliforms (VRBA). Cheese waswith the enterocin producer E. faecium RZS C5 and sampled after 3 days, 1, 2 and 4 weeks, and 3, 6, 9 andtwo with the enterocin producer E. faecium DPC 12 months for starters (MRS and L-M17), enterococci1146. The starter cultures L. lactis subsp. cremoris (KAA) and lactobacilli (LBS).223 and L. lactis subsp. cremoris 227 were inoculatedinto heat-treated (90 jC for 30 min) reconstituted 2.5. Quantitative determination of bacteriocin titres in(10%, m/v) skim milk, and incubated at 21 jC for liquid samples16 h. For cheesemaking, the pregrown cultures wereinoculated at 0.75% (m/v) into 450 l of pasteurised Bacteriocin activity in the liquid was determined(72 jC for 15 s) milk, which had been cooled to 32 by an adaptation of the critical dilution method,jC. E. faecium RZS C5 and DPC 1146 were subcul- currently used for the assay of bacteriocins, astured in 500 ml of MRS broth, grown overnight at 37 reported by De Vuyst et al. (1996b). Briefly, serialjC, and then inoculated (0.02%, v/v) into the cheese twofold dilutions of cell-free culture supernatant con-milk at the same time as the starter. The initial milk taining bacteriocin were spotted (10 Al) on agar plates
´ M.R. Foulquie Moreno et al. / International Journal of Food Microbiology 81 (2003) 73–84 77containing overlayers of fresh cultures of a sensitive ´ method (Lopez-Lara et al., 1991). Also, uniformstrain. These overlayer cultures were prepared by cheese pieces were cut using a cork borer (numberpropagating fresh cultures to an optical density at 12). Agar plates were prepared with the indicator600 nm of 0.45, and adding 100 Al of the cell strain L. innocua LMG 13568, and before the agarsuspension to 3.5 ml of overlay agar. Overlaid agar solidified, the cheese pieces were placed on the agar.plates were incubated for 18 h at the appropriate After incubation at 37 jC for 12 h, the plates weretemperature. The bacteriocin activity was defined as checked for zones of inhibition.the reciprocal of the highest dilution that demonstra-ted complete inhibition of the indicator lawn, and wasexpressed in activity units (AU) per millilitre of 3. Resultsculture medium. 3.1. Enterocin production in modified MRS broth and2.6. Bacteriocin detection in cheese samples skimmed milk Bacteriocin activity was measured in frozen cheese The influence of temperature and pH on thesamples of maximum scald, pressing, day 1 and 3, kinetics of growth and bacteriocin production by E.week 1 and 2, and month 1, 3, 6, 9 and 12. Different faecium RZS C5 was studied during different batchprotocols were used for the bacteriocin extraction and fermentations. The results are summarised in Table 1.detection from cheese. Trisodium citrate (2.0%, m/v) During batch fermentation at uncontrolled pH and at aor tetrasodium EDTA (2.0%, m/v) at dilutions of 1:1, controlled pH of 6.5, both at 37 jC, lmax was ap-1:2, 1:3, 1:4, 1:5 and 1:10 (m/v) were tested as proximately the same. In both cases, bacteriocin pro-extraction liquors. These sample dilutions were trea- duction started in the early logarithmic growth phaseted in a stomacher (Seward Medical, London, UK) for (Fig. 1a and 1b). At controlled pH, the biomass in-15 min. The resulting suspension was either centri- crease was more than twice the value at uncontrolledfuged 5 min or heat treated for 10 min at 80 jC. The pH, and the bacteriocin activity also increased con-extracted bacteriocin was tested by the well-diffusion siderably (Table 1). In both fermentations, the max-Fig. 1. Biomass (n; CDM, g lÀ1), bacteriocin activity (Â; AU mlÀ1), number of cells (E; CFU mlÀ1), and pH (continuous line) during batchfermentation of E. faecium RZS C5 at 37 jC in MRS containing (a) 2.0% glucose at uncontrolled pH; (b) 2.0% glucose at constant pH 6.5; (c)2.0% lactose and 1.8% casein hydrolysate (instead of Lab Lemco and peptone) at constant pH 6.5; (d) 5.0% lactose and 1.8% casein hydrolysate(instead of Lab Lemco and peptone) at constant pH 6.5.
78 ´ M.R. Foulquie Moreno et al. / International Journal of Food Microbiology 81 (2003) 73–84Fig. 2. Bacteriocin activity (Â; AU mlÀ1), number of cells (E; E. faecium RZS C5, cfu mlÀ1—n; starters, cfu mlÀ1), pH (continuous line), andtemperature (discontinuous line; jC) during batch fermentation of (a and c) E. faecium RZS C5, and (b and d) E. faecium RZS C5 and startercultures, in skim milk with the pH and temperature profiles of Cheddar cheese production.imum activity was obtained early in the exponential pH value and at 21 jC, a similar amount of biomassgrowth phase (after about 3 h of fermentation). How- was produced, and after 48 h of fermentation theever, a higher decrease of bacteriocin activity was maximum bacteriocin activity (2400 AU mlÀ1) wasobserved during the stationary phase at controlled pH, detected, followed by a considerable decrease at thecompared with uncontrolled pH (Fig. 1a and b). At 37 end of the fermentation (data not shown).jC and at a lower controlled pH of 5.5, lmax was Two other fermentations at controlled pH 6.5 andlower, while the biomass amount was approximately 37 jC were performed to mimic milk fermentationsthe same as in the former fermentation. The maximum (Table 1). One was carried out with lactose (2.0%, m/v)bacteriocin activity (3200 AU mlÀ1) was obtained instead of glucose as the energy source (data notafter 12 h of fermentation, and was constant until the shown), and the other was performed with lactoseend of the fermentation (data not shown). At the same and casein hydrolysate instead of meat extract (LabFig. 3. pH profile during Cheddar cheese production of Trial 2: x, control; n, vat-2: E. faecium DPC 1146; E, vat-3: E. faecium RZS C5; Â,vat-4: E. faecium DPC 1146.
´ M.R. Foulquie Moreno et al. / International Journal of Food Microbiology 81 (2003) 73–84 79Fig. 4. Numbers of enterococci during cheese production. x, control-vat, Trial 1; n, control-vat, Trial 2; E, E. faecium RZS C5, Trial 1; Â, E.faecium DPC 1146, Trial 2; B, E. faecium RZS C5, Trial 2; , E. faecium DPC 1146, Trial 2.Lemco) and peptone as complex nitrogen source (Fig.1c). In both fermentations, the maximum enterocinactivity (4800 AU mlÀ1) was seen after 4 h of fermen-tation, and in both cases the activity decreased duringthe stationary phase. The maximum biomass was 2.3and 1.8 g CDM lÀ1, respectively, indicating theimportance of a complex nitrogen source for cellsynthesis. To mimic cheese manufacture much better, salt wasadded and the lactose concentration was increased intwo independent fermentations (Table 1). When 2.0%NaCl was added, a higher decrease in the enterocinactivity was observed as compared with the samemedium without salt. When 5.0% lactose was fer-mented (Fig. 1d), a higher maximum enterocin acti-vity (9600 AU ml À1 ) was seen after 15 h offermentation compared to the same medium with2.0% lactose, and it remained constant until the endof the fermentation. For the fermentations carried out in skimmed milk,bacteriocin activity was higher (1200 vs. 600 AUmlÀ1) and started earlier (after 49 and 71 h, respec-tively), when casein hydrolysate was added to themilk. Maximum bacteriocin production was observedin the late stationary phase, and no decrease of thebacteriocin activity was seen. At uncontrolled pH,very low activity was detected (200 AU mlÀ1).3.2. Bacteriocin production in fermentations simulat-ing Cheddar cheese production conditions Fig. 5. Bacteriocin activity of cheese (Trial 1) visualised on plates containing L. innocua LMG 13568 as indicator strain: (a) following Two fermentations in 10% skimmed milk were per- the well-diffusion method; (b) placing a piece of cheese on the agarformed with the temperature and pH profile of Ched- (A, control-vat; B, bacteriocin producer E. faecium RZS C5).
80 ´ M.R. Foulquie Moreno et al. / International Journal of Food Microbiology 81 (2003) 73–84dar cheese production. During the first fermentation, These last two fermentations were repeated and theonly the bacteriocin producer strain E. faecium RZS same results were obtained.C5 was added to the milk (Fig. 2a and c), while in thesecond fermentation, the bacteriocin producer strain 3.3. Cheddar cheese production on pilot scaleand the commercial starter cultures for Cheddar cheeseproduction were included (Fig. 2b and d). When the The follow-up of the cheese production on pilottemperature was modified during the fermentation scale is represented in Figs. 3 –6. The pH profile offollowing the temperature profile of Cheddar cheese cheese production was the same for all vats in bothproduction, no significant differences were detected in Cheddar cheese production trials. The results for thethe enterocin production, which reached its maximum second trial are shown in Fig. 3. In the first trial,(150 AU mlÀ1) after 38 h of incubation. However, no 4Â101 CFU/g of enterococci were found in milk ofenterocin activity could be measured when the bacter- the control vat after pasteurisation, which increased toiocin producer strain E. faecium RZS C5 was grown as 5Â103 CFU/g during cheese manufacture. This num-co-culture with the commercial starter cultures L. lactis ber remained constant during the ripening. No enter-subsp. cremoris 223 and L. lactis subsp. cremoris 227. ococci were found in milk of the control vat of theFig. 6. Bacteriocin activity of cheese (Trial 2) visualised on plates containing L. innocua LMG 13568 as indicator strain. A, control vat; B,enterocin producer E. faecium DPC 1146; C, bacteriocin producer E. faecium RZS C5; C, bacteriocin producer E. faecium DPC 1146.
´ M.R. Foulquie Moreno et al. / International Journal of Food Microbiology 81 (2003) 73–84 81second trial, however 4Â101 CFU/g of enterococci produced, and losses were more pronounced than atwere counted after 14 days. The counts of enterococci constant pH 5.5. The specific bacteriocin activity wasin both trials of the vats inoculated with the enterocin higher at pH 5.5 than at pH 6.5, indicating thatproducers E. faecium RZS C5 or E. faecium DPC bacteriocin production is enhanced at pH 5.5. These1146 were approximately the same, increasing from results are in accordance with those obtained previ-105 to 107 CFU/g during cheese manufacture (Fig. 4), ously for enterocin 1146 by Parente and Ricciardiafter which they remained constant during ripening (1994), where bacteriocin activity displayed a de-(data not shown). The counts of the starters were the crease in the early stationary phase at pH valuessame in the control vat as in the vats inoculated with higher than 4.5. The decrease of activity was ascribedenterococci, remaining constant during the ripening to the adsorption of the bacteriocin molecules on the(108 CFU/g). The number of lactobacilli increased cell surface of producer cells and depends on the pHduring ripening (data not shown). of the cell environment, being more pronounced at Enterocin production was determined by two meth- higher pH (Yang et al., 1992; Parente and Ricciardi,ods, either bacteriocin extraction of cheese samples 1994; De Vuyst et al., 1996a; Leroy and De Vuyst,with 2.0% tetrasodium EDTA or 2.0% trisodium 1999). However, with decreasing temperature, thecitrate followed by a well-diffusion method, or plac- adsorption increased. Thus, one can assume that theing a piece of cheese on agar plates containing L. enterocin activity measured in the liquid is lower thaninnocua LMG 13568 as indicator strain. However, the the total amount really produced. Therefore, this so-indicator strain was sensitive to tetrasodium EDTA, called bioavailable bacteriocin activity will be mostwhich was not the case with trisodium citrate. Bacter- probably the determining factor for in situ antibacte-iocin extraction of cheese samples with trisodium rial action. The use of lactose instead of glucosecitrate, followed by heating or centrifugation of the increased the production of enterocin, and at highersamples revealed clear inhibition zones for a 1:4 lactose concentrations, enterocin production increaseddilution, indicating a good extraction of the heat- more and adsorption decreased. At present, it isstable bacteriocin. No inhibition zones were observed difficult to explain this phenomenon. The enterocinin the product from the control vat (Fig. 5a). When the production in skimmed milk was lower and took placecheese samples were placed on agar plates containing mainly in the stationary phase, a phenomenon thatthe indicator strain, clear inhibition zones were seen could be very interesting in view of the applicabilityaround the cheese pieces with the bacteriocin-produc- of enterococci in cheese production. When caseining strains (Figs. 5b and 6). The antilisterial activity hydrolysate was added to the milk, bacteriocin pro-was detected from the maximum scald to after 12 and duction was higher and started earlier in the growth6 months for the first and second trial, respectively, phase, indicating the importance of an additionalindicating production of the enterocin during fermen- nitrogen source for growth of E. faecium in milk.tation and its persistence upon further cheese manu- This can be explained by the low proteolytic activityfacturing and ripening. of enterococci. For co-cultures of E. faecium RZS C5 and the starters, no enterocin activity was detected during the milk fermentation, nevertheless, the acid-4. Discussion ifying activity of the starters did not inhibit the growth of E. faecium RZS C5 significantly. However, it may E. faecium RZS C5 and E. faecium DPC 1146 be that the amount of cells (1 log less) resulted in aproduce enterocins with anti-Listeria activity. Enter- nondetectable amount of enterocin produced (seeocin production by E. faecium RZS C5 in a complex below). This effect was also observed by Giraffa etmedium (MRS) started in the early logarithmic al. (1994, 1995a), who found that the final levels ofgrowth phase, and bacteriocin activity decreased dur- bacteriocin produced in milk by E. faecium 7C5 in co-ing the stationary phase. The effect of pH on bacter- culture with starters were lower than in pure cultures.iocin production and apparent degradation was as One can assume either reduced growth, degradation ofintense as the effect on bacterial growth. At constant the enterocin by proteases derived from the starterpH 6.5 growth was higher, but less enterocin was cultures, or loss of bacteriocin adsorbed to coagulating
82 ´ M.R. Foulquie Moreno et al. / International Journal of Food Microbiology 81 (2003) 73–84milk proteins with the curd. In contrast, when enter- production and stability during cheese manufacturingococci were used as co-cultures during the cheese in the case of E. faecium RZS C5 and E. faecium DPCproduction, a low but clear inhibition was detected in 1146, because of their high bacteriocin production insamples taken when the maximum scald was reached laboratory media. Second, during laboratory fermen-(38.5 jC). Cheese samples placed on agar plates tations, enterocin production took place in the earlycontaining the indicator strain also showed the pres- exponential growth phase in a complex mediumence of enterocin in the cheese. Therefore, it is (MRS) and in the late stationary phase in milk. Itassumed that the bacteriocin is produced during could very well be that the bioavailable enterocincheese manufacturing and that it is stable during the activity level in milk is only measured when a certainwhole ripening phase of the cheese. Giraffa et al. amount is already bound to milk particles like pro- ´˜(1995b) and Nunez et al. (1997) observed that enter- teins. During the cheese processing, enterocin produc-ocin 7C5 and enterocin 4 are stable during the ripen- tion took place in the early logarithmic phase as well,ing of Taleggio and Manchego cheese, respectively. and it was carried over into the cheese. This indicatesThis is in contrast with other studies where a degra- that both an early and late contamination of the milkdation of the bacteriocin by peptidases present in the or cheese can be combated with a stable, in situcheese was seen (Sulzer and Busse, 1991; Nettles and enterocin production. The use of such a co-culture isBarefoot, 1993). Sulzer and Busse (1991) found that an additional safety provision beyond good manufac-different enterocins are not stable during the ripening turing practices in cheesemaking.of Camembert because of the proteases produced bythe cheese mould. In the latter case, Listeria is onlysuppressed when the contamination occurs in the early Acknowledgements ´stage of ripening. Farıas et al. (1999) could not detectenterocin CRL35 during ripening, although suppres- This work was supported by the FAIR Programmesion of Listeria took place during the whole period of of the European Commission (grants CT97-3078 andripening. Since in Cheddar manufacture, only limited ´ ´ CT97-5013). Marıa Remedios Foulquie Moreno was aautolysis of starters occurs, and since lysis is then a recipient of a Marie Curie Fellowship from the Com-gradual process during ripening, the peptidase poten- mission of the European Communities (grant FAIR-tial in Cheddar is initially low (these being intra- CT97-5013).cellular enzymes). The stability of enterocins in oursystem is therefore as expected and enterocin activity Referencescould be detected from the start of the cheesemakingto the ripening period. Another point to take in Aymerich, T., Garriga, M., Ylla, J., Vallier, J., Monfort, J.M., Hugas,account is to detect their presence in the media. M., 2000. Application of enterocins as biopreservatives againstEnterocins are hydrophobic and heat-stable, extraction Listeria innocua in meat products. Journal of Food Protectionmethods are very tedious, cumbersome and not 63, 721 – 726. Ben Embarek, P.K., Jeppesen, V.F., Huss, H.H., 1994. Antibacterialalways successful. However, our results show that, potential of Enterococcus faecium strains isolates from sous-provided the bioavailable activity is high enough vide cooked fish fillets. Food Microbiology 11, 525 – 536.(which will be necessary to see an in situ effect), it Bennik, M.H.J., van Overbeek, W., Smid, E.J., Gorris, L.G.M.,is possible to detect enterocin activity in cheese by 1999. Biopreservation in modified atmosphere stored mungbeanplacing carefully cut cheese pieces on plates contain- sprouts: the use of vegetable-associated bacteriocinogenic lactic acid bacteria to control the growth of Listeria monocytogenes.ing an appropriate indicator strain. This method is Letters in Applied Microbiology 28, 226 – 232.easy to perform, time-saving, and gives a qualitative Callewaert, R., Hugas, M., De Vuyst, L., 2000. Competitivenessreflection of the real situation. and bacteriocin production of Enterococci in the production of To conclude, this study of in vitro batch fermenta- Spanish-style dry fermented sausages. International Journal oftions of the enterocin producer E. faecium RZS C5 Food Microbiology 57, 33 – 42. ´ ´ Centeno, J.A., Menendez, S., Rodrıguez-Otero, J.L., 1996. Mainwill contribute to a better understanding of what could microbial flora present as natural starters in Cebreiro raw cow’shappen in cheese processing concerning in situ bac- milk cheese. International Journal of Food Microbiology 33,teriocin production. First, one could expect enterocin 307 – 313 (Northwest Spain).
´ M.R. Foulquie Moreno et al. / International Journal of Food Microbiology 81 (2003) 73–84 83 ´ ´Centeno, J.A., Menendez, S., Hermina, Ma., Rodrıguez-Otero, J.L., tivity and lactic acid development. Journal of Milk and Food 1999. Effects of the addition of Enterococcus faecalis in Ce- Technology 38, 3 – 7. breiro cheese manufacture. International Journal of Food Micro- Jensen, J.P., Reinbold, G.W., Washam, C.J., Vedamuthu, E.R., 1975c. biology 48, 97 – 111. Role of enterococci in Cheddar cheese: organoleptic considera-Coppola, S., Parente, E., Dumontet, S., La Peccerella, A., 1988. The tions. Journal of Milk and Food Technology 38, 142 – 145. microflora of natural whey cultures utilized as starters in the ´˜ Joosten, H.M.L.J., Gaya, P., Nunez, M., 1995. Isolation of tyrosine manufacture of Mozzarella cheese from water-buffalo milk. Lait decarboxylase-less mutants of a bacteriocin-producing Entero- 68, 295 – 310. coccus faecalis strain and their application in cheese. Journal ofCoppola, S., Villani, F., Coppola, R., Parente, E., 1990. Comparison Food Protection 58, 1222 – 1226. of different starter systems for water-buffalo Mozzaarella cheese ´ Laukova, A., Czikkova, S., 1999. The use of enterocin CCM 4231 manufacture. Lait 70, 411 – 423. in soy milk to control the growth of Listeria monocytogenes andDe Vuyst, L., 2000. Technology aspects related to the application of Staphylococcus aureus. Journal of Applied Microbiology 87, functional starter cultures. Food Technology and Biotechnology 182 – 186. 38, 105 – 112. Ledda, A., Scintu, M.F., Pirisi, A., Sanna, S., Mannu, L., 1994.De Vuyst, L., Vandamme, E.J., 1994. Bacteriocins of lactic acid Technological characterization of lactococci and enterococci bacteria. In: De Vuyst, L., Vandamme, E.J. (Eds.), Microbiol- for the manufacture of Fiore Sardo sheep cheese. Scienza e ogy, Genetics and Applications Blackie Academic & Professio- Tecnica Lattiero-Casearia 45, 443 – 456. nal, London, pp. 1 – 11. Leroy, F., De Vuyst, L., 1999. Temperature and pH conditions that ´De Vuyst, L., Callewaert, R., Crabbe, K., 1996a. Primary metab- prevail during fermentation of sausages are optimal for produc- olite kinetics of bacteriocin biosynthesis by Lactobacillus amy- tion of the antilisterial bacteriocin sakacin K. Applied and En- lovorus and evidence for stimulation of bacteriocin production vironmental Microbiology 65, 974 – 981. under unfavourable growth conditions. Microbiology 142, ´ ´ ´ Lopez-Lara, I., Galvez, A., Martınez-Bueno, M., Maqueda, M., 817 – 827. Valdivia, E., 1991. Purification, characterization, and biologicalDe Vuyst, L., Callewaert, R., Pot, B., 1996b. Characterization of the effects of a second bacteriocin from Enterococcus faecalis ssp. antagonistic activity of Lactobacillus amylovorus DCE 471 and liquefaciens S-48 and its mutant strain B-48-28. Canadian Jour- large scale isolation of its bacteriocin amylovorin L471. System- nal of Microbiology 37, 769 – 774. atic and Applied Microbiology 19, 9 – 20. Maisnier-Patin, S., Forni, E., Richard, J., 1996. Purification, partial ´Farıas, M.E., de Ruiz Holgado, A.A.P., Sesma, F., 1994. Bacteriocin characterisation and mode of action of enterococcin EFS2, an production by lactic acid bacteria isolated from regional antilisterial bacteriocin produced by a strain of Enterococcus cheeses: inhibition of foodborne pathogens. Journal of Food faecalis isolated from a cheese. International Journal of Food Protection 57, 1013 – 1015. Microbiology 30, 255 – 270. ´ ˜Farıas, M.E., Nunez de Kairuz, M., Sesma, F., Palacios, J., de Ruiz McAuliffe, O., Hill, C., Ross, R.P., 1999. Inhibition of Listeria Holgado, A.P., Oliver, G., 1999. Inhibition of Listeria monocy- monocytogenes in cottage cheese manufactured with a lacticin togenes by the bacteriocin enterocin CRL35 during goat cheese 3147-producing starter culture. Journal of Applied Microbio- making. Milchwissenschaft 54, 30 – 32. logy 86, 251 – 256.Giraffa, G., 1995. Enterococcal bacteriocins: their potential as anti- ´ ´ Mendoza, F., Maqueda, M., Galvez, A., Martınez-Bueno, M., Val- Listeria factors in dairy technology. Food Microbiology 12, divia, E., 1999. Antilisterial activity of peptide AS-48 and study 291 – 299. of changes induced in the cell envelope properties of an AS-48-Giraffa, G., Neviani, E., Torri Tarelli, G., 1994. Antilisterial activity adapted strain of Listeria monocytogenes. Applied and Environ- by enterococci in a model predicting the temperature evolution mental Microbiology 65, 618 – 625. of Taleggio, an Italian soft cheese. Journal of Dairy Science 77, Nettles, C.G., Barefoot, S.F., 1993. Biochemical and genetical char- 1176 – 1182. acteristics of bacteriocins of food-assiciated lactic acid bacteria.Giraffa, G., Carminati, D., Torri Tarelli, G., 1995a. Inhibition of Journal of Food Protection 56, 338 – 358. Listeria innocua in milk by bacteriocin-producing Enterococcus ´˜ ´ ´ Nunez, M., Rodrıguez, J.L., Garcıa, E., Gaya, P., Medina, M., 1997. faecium 7C5. Journal of Food Protection 58, 621 – 623. Inhibition of Listeria monocytogenes by enterocin 4 during theGiraffa, G., Picchioni, N., Neviani, E., Carminati, D., 1995b. Pro- manufacture and ripening of Manchego cheese. Journal of Ap- duction and stability of an Enterococcus faecium bacteriocin plied Microbiology 83, 671 – 677. during Taleggio cheesemaking and ripening. Food Microbiology Parente, E., Hill, C., 1992a. A comparison of factors affecting the 12, 301 – 307. production of two bacteriocins from lactic acid bacteria. JournalJack, R.W., Tagg, J.R., Ray, B., 1995. Bacteriocins of Gram-po- of Applied Bacteriology 73, 290 – 298. sitive bacteria. Microbiology Reviews 59, 171 – 200. Parente, E., Hill, C., 1992b. Characterization of enterocin 1146, aJensen, J.P., Reinbold, G.W., Washam, C.J., Vedamuthu, E.R., bacteriocin from Enterococcus faecium inhibitory to Listeria 1975a. Role of enterococci in Cheddar cheese: free fatty acid monocytogenes. Journal of Food Protection 55, 497 – 502. appearance and citric acid utilization. Journal of Milk and Food Parente, E., Hill, C., 1992c. Inhibition of Listeria in buffer, broth, Technology 38, 78 – 83. and milk by enterocin 1146, a bacteriocin produced by Enter-Jensen, J.P., Reinbold, G.W., Washam, C.J., Vedamuthu, E.R., ococcus faecium. Journal of Food Protection 55, 503 – 508. 1975b. Role of enterococci in Cheddar cheese: proteolytic ac- Parente, E., Ricciardi, A., 1994. Influence of pH on the production
84 ´ M.R. Foulquie Moreno et al. / International Journal of Food Microbiology 81 (2003) 73–84 of enterocin 1146 during batch fermentation. Letters in Applied of enterococci in hard cheese made from raw cow and sheep Microbiology 19, 12 – 15. milk. Milchwissenschaft 42, 717 – 718.Parente, E., Villani, F., Coppola, R., Coppola, S., 1989. A multiple Villani, F., Coppola, S., 1994. Selection of enterococcal strains for strain starter for water-buffalo Mozzarella cheese manufacture. water-buffalo Mozzarella cheese manufacture. Annual Reviews Lait 69, 271 – 279. of Microbiology and Enzymology 44, 97 – 105.Parente, E., Brienza, C., Ricciardi, A., Addario, G., 1997. Growth Vlaemynck, G., 1996. Inhibition of Listeria monocytogenes by and bacteriocin production by Enterococcus faecium DPC 1146 the application of bacteriocins: study of some parameters af- in batch and continuous culture. Journal of Industrial Microbio- fecting production of enterocin RZS C5 and pediocin RZS logy & Biotechnology 18, 62 – 67. C8. Mededelingen Faculteit Landbouwkundige en Toegepaste ´Simonetta, A.C., Moragues de Velasco, L.G., Frison, L.N., 1997. Biologische Wetenschappen Universiteit Gent 61 (4a), 1605 – Antibacterial activity of enterococci strains against Vibrio chol- 1611. erae. Letters in Applied Microbiology 24, 139 – 143. Vlaemynck, G., Herman, L., Coudijzer, K., 1994. Isolation andSmith, J., 1993. In: Smith, J. (Ed.), Technology of Reduced-Additive characterization of two bacteriocins produced by Enterococcus Foods Blackie Academic & Professional, London, pp. 123 – 138. faecium strains inhibitory to Listeria monocytogenes. Interna-Sulzer, G., Busse, M., 1991. Grown inhibition of Listeria spp. on tional Journal of Food Microbiology 24, 211 – 225. Camembert cheese by bacteria producing inhibitory substances. Yang, R., Jonhson, M., Ray, B., 1992. Novel method to extract large International Journal of Food Microbiology 14, 287 – 296. amounts of bacteriocins from lactic acid bacteria. Applied andTrovatelli, L.D., Schiesser, A., 1987. Identification and significance Environmental Microbiology 58, 3355 – 3359.