ARTICLE IN PRESS

Food Microbiology 25 (2008) 762– 770

Contents lists available at ScienceDirect

Food Microbiology

0740-00

doi:10.1

� Corr

E-m

journal homepage: www.elsevier.com/locate/fm

Inhibition of Bacillus cereus and Bacillus weihenstephanensis in raw vegetablesby application of washing solutions containing enterocin AS-48 alone and incombination with other antimicrobials

Antonio Cobo Molinos a, Hikmate Abriouel a, Rosario Lucas Lopez a, Nabil Ben Omar a, Eva Valdivia b,c,Antonio Galvez a,�

a Area de Microbiologıa, Departamento de Ciencias de la Salud, Facultad de Ciencias Experimentales, Universidad de Jaen, 23071 Jaen, Spainb Departamento de Microbiologıa, Facultad de Ciencias, Universidad de Granada, 18071 Granada, Spainc Instituto de Biotecnologıa, Facultad de Ciencias, Universidad de Granada, 18071 Granada, Spain

a r t i c l e i n f o

Article history:

Received 11 February 2008

Received in revised form

29 April 2008

Accepted 4 May 2008Available online 13 May 2008

Keywords:

Bacillus cereus

Bacillus weihenstephanensis

Bacteriocins

Biopreservation

Vegetable foods

Sprouts

20/$ - see front matter & 2008 Elsevier Ltd. A

016/j.fm.2008.05.001

esponding author. Tel.: +34 953 212160; fax:

ail address: agalvez@ujaen.es (A. Galvez).

a b s t r a c t

Enterocin AS-48 is a broad-spectrum cyclic antimicrobial peptide produced by Enterococcus faecalis. In

the present study, the bacteriocin was tested alone and in combination with other antimicrobials for

decontamination of Bacillus inoculated on alfalfa, soybean sprouts and green asparagus. Washing with

enterocin AS-48 solutions reduced viable cell counts of Bacillus cereus and Bacillus weihenstephanensis by

1.0–1.5 and by 1.5–2.38 log units right after application of treatment, respectively. In both cases, the

bacteriocin was effective in reducing the remaining viable population below detection levels during

further storage of the samples at 6 1C, but failed to prevent regrowth in samples stored at 15 or 22 1C.

Application of washing treatments containing enterocin AS-48 in combination with several other

antimicrobials and sanitizers (cinnamic and hydrocinnamic acids, carvacrol, polyphosphoric acid,

peracetic acid, hexadecylpyridinium chloride and sodium hypochlorite) greatly enhanced the

bactericidal effects. The combinations of AS-48 and sodium hypochlorite, peracetic acid or

hexadecylpyridinium chloride provided the best results. After application of the combined treatments

on alfalfa sprouts contaminated with B. cereus or with B. weihenstephanensis, viable bacilli were not

detected or remained at very low concentrations in the treated samples during a 1-week storage period

at 15 1C. Inhibition of B. cereus by in situ produced bacteriocin was tested by cocultivation with the AS-48

producer strain E. faecalis A-48-32 inoculated on soybean sprouts. Strain A-48-32 was able to grow and

produce bacteriocin on sprouts both at 15 and 22 1C. At 15 1C, growth of B. cereus was completely

inhibited in the cocultures, while a much more limited effect was observed at 22 1C. The results

obtained for washing treatments are very encouraging for the application of enterocin AS-48 in the

decontamination of sprouts. Application of washing treatments containing AS-48 alone can serve to

reduce viable cell counts of bacilli in samples stored under refrigeration, while application of combined

treatments should be recommended to avoid proliferation of the surviving bacilli under temperature-

abuse conditions.

& 2008 Elsevier Ltd. All rights reserved.

1. Introduction

Endospore-forming bacteria of the Bacillus cereus group areoften considered hazardous to food safety because of their capacityto produce enterotoxins, together with their wide distribution innature and frequent contamination of raw materials used in foodproduction. B. cereus is a common soil inhabitant that is oftenpresent in a variety of foods, such as rice, spices, milk and dairy

ll rights reserved.

+34 953 212943.

products, vegetables, meat, cakes and other desserts (Granum,2001). B. cereus produces several toxins (Schoeni and Wong, 2005),including emetic toxin (Agata et al., 1995) and at least four otherenterotoxins: hemolysin BL or Hbl (Beecher and Wong, 1997), non-hemolytic enterotoxin or Nhe (Lindback et al., 2004) and cytotoxinK or CytK (Lund et al., 2000). Bacillus weihenstephanensis is apsychrotolerant bacterium belonging to the B. cereus group(Lechner et al., 1998). Some strains may be highly cytotoxic, andhave been shown to carry at least some of the genetic determinantsof the enterotoxins as well as the emetic toxin described in B. cereus

(Pruß et al., 1999; Stenfors et al., 2002; Thorsen et al., 2006; Baronet al., 2007). B. cereus has been isolated from sprouts and sprouting

ARTICLE IN PRESS

A. Cobo Molinos et al. / Food Microbiology 25 (2008) 762–770 763

seeds (Kim et al., 2004; Pao et al., 2005). In soybean sprouts,endospore-forming bacteria were found in the order of 2 logunits/g,and 53 out of 55 B. cereus isolates were found to produce diarrheicenterotoxins (Kim et al., 2004). B. cereus has also been isolated fromraw vegetables used to prepare refrigerated minimally processedfoods (Valero et al., 2002). The use of sanitizers to reduce theB. cereus levels in fruits and vegetables has been proposed as asafety measure (Beuchat et al., 2004).

Bacteriocins of lactic acid bacteria (LAB) have seldom beenused for decontamination of fresh produce (reviewed by Galvezet al., 2008). Nisin and pediocin individually or in combinationwith antimicrobial agents (sodium lactate, potassium sorbate,phytic acid and citric acid) were tested as possible sanitizertreatments for reducing the population of Listeria monocytogenes

on cabbage, broccoli and mung bean sprouts (Bari et al., 2005).Washing solutions consisting of LAB bacteriocins produced on alettuce extract (nisin Z, coagulin and a nisin:coagulin cocktail)reduced viable cell counts of L. monocytogenes on fresh-cuticeberg lettuce stored in microperforated plastic bags, but didnot prevent further growth of survivors during refrigerationstorage (Allende et al., 2007). Enterocin AS-48 is a cyclic peptidewith a broad antibacterial spectrum (Galvez et al., 1986, 1989;Maqueda et al., 2004). This bacteriocin has been purified on asemi-preparative scale (Abriouel et al., 2003), and is now beingtested as a natural preservative in different food systems (Munoz

0

2

4

0 1 3 5 7Time (days)

Log

CFU

/ml

0

2

4

0 1 3 5 7Time (days)

Log

CFU

/ml

0

2

4

6

0 1 3Time

0

2

4

0 1 3 5 7Time (days)

Log

CFU

/ml

Log

CFU

/ml

Log

CFU

/ml

Log

CFU

/ml

0

2

4

6

0 1 3Time

0

2

4

6

0 1 3Tim

6°C

6°C

6°C

Fig. 1. Effect of immersion treatments with solutions containing enterocin AS-48 on surv

sprouts (D–F) and green asparagus (G–I). After immersion for 5 min at working tempera

(m), samples were incubated at 6 1C (A, D, G), 15 1C (B, E, H) or 22 1C (C, F, I). Data represen

et al., 2004, 2007; Ananou et al., 2005a, b; Grande et al., 2005,2006, 2007a, b; Lucas et al., 2006). Enterocin AS-48 may be aninteresting bacteriocin for decontamination of raw vegetables. Inlettuce juice, added enterocin AS-48 caused a strong inhibition ofStaphylococcus aureus and completely inactivated L. monocyto-

genes and B. cereus (Grande et al., 2005). Enterocin AS-48 wastested previously in washing treatments for decontamination ofseed sprouts against L. monocytogenes (Cobo Molinos et al., 2005).The antilisterial activity of AS-48 was enhanced greatly by severalantimicrobial compounds, reducing the population of viablelisteria below detection limits in the treated sprouts aftertreatment and preventing regrowth of the listeria during storageof the treated samples (Cobo Molinos et al., 2005). Results fromour previous studies indicate that the effectiveness of AS-48treatments greatly depends on the target bacteria and the foodsystem. While added bacteriocin completely inactivated B. cereus

in rice-based foods (Grande et al., 2006), higher bacteriocinconcentrations were required to control mixed populations ofbacilli in purees (Grande et al., 2007b). In washing treatments, thebacteriocin concentration in samples will decrease after treat-ment, limiting the protection of samples during storage. Thepurpose of the present study was to determine the effectiveness ofAS-48 applied in washing treatments against two toxicogenicspecies of the B. cereus group alone and in combination with otherantimicrobials in search for synergistic effects against the bacilli.

0

2

4

6

0 1 3 5 7Time (days)

5 7 (days)

0

2

4

6

8

0 1 3 5 7Time (days)

Log

CFU

/ml

Log

CFU

/ml

Log

CFU

/ml

5 7 (days)

0

2

4

6

8

0 1 3 5 7Time (days)

5 7e (days)

15°C

15°C

15°C 22°C

22°C

22°C

ival and proliferation of B. cereus LWL1 inoculated on soybean sprouts (A–C), alfalfa

tures in solutions containing final bacteriocin concentrations of 0 (O) and 25 mg/ml

t the average of two independent experiments plus standard deviation (error bars).

ARTICLE IN PRESS

0

2

4

6

8

0 1 3 5 7Time (days)

Log

CFU

/ml

0

2

4

6

8

0 1 3 5 7Time (days)

Log

CFU

/ml

0

2

4

0 1 3 5 7Time (days)

Log

CFU

/ml

6°C

15°C

22°C

Fig. 2. Effect of immersion treatments with solutions containing enterocin AS-48

on survival and proliferation of B. weihenstephanensis CECT 5894 inoculated on

soybean sprouts. After immersion for 5 min at working temperatures in solutions

containing final bacteriocin concentrations of 0 (O) and 25 mg/ml (m), samples

were incubated at 6 1C (A), 15 1C (B) or 22 1C (C). Data represent the average of two

independent experiments plus standard deviation (error bars).

A. Cobo Molinos et al. / Food Microbiology 25 (2008) 762–770764

2. Materials and methods

2.1. Bacterial strains and cultivation conditions

Enterococcus faecalis A-48-32 (Martinez Bueno et al., 1990) wasused for the production of enterocin AS-48, and E. faecalis S-47from our collection was used for standard determination ofbacteriocin activity. B. cereus LWL1 is a psychrotrophic andenterotoxigenic strain (Dufrenne et al., 1995), kindly supplied byDr. F.M. van Leusden (Microbiological Laboratory for HealthProtection, Natl. Inst. Publ. Health and Environ., The Netherlands).B. weihenstephanensis CECT 5894 was obtained from the Spanishtype culture collection (CECT). All strains were cultivatedroutinely on brain–heart infusion (BHI, Scharlab, Barcelona, Spain)broth at 37 1C, and stored at 4 1C on BHI-agar slants.

2.2. Preparation of bacteriocin and chemical preservatives

Enterocin AS-48 was recovered from cultured broths of theproducer strain E. faecalis A-48–32 in CMG-medium by cationexchange chromatography as described elsewhere (Abriouel et al.,2003). Bacteriocin concentrates were filtered through 0.22mmpore size low protein-binding filters (Millex GV; MilliporeCorporation, Belford, MA, USA) under sterile conditions and testedfor bacteriocin activity against the indicator strain E. faecalis S-47by the agar well diffusion method (Abriouel et al., 2003).Immersion solutions were prepared by diluting bacteriocinconcentrates (500 mg/ml) in sterile distilled water or in aqueoussolutions of the chemical preservatives to be tested in the case ofcombined treatments.

Lactic acid, sodium lactate, tri-sodium trimetaphosphate,polyphosphoric acid, hexadecylpyridinium chloride, n-propylp-hydroxybenzoate, and p-hydoxybenzoic acid methyl estherwere purchased from Sigma (Madrid, Spain). Cinnamic acid,hydrocinnamic acid, carvacrol and peracetic acid were from Fluka(Madrid). Sodium hypochlorite was a commercial concentratedbleach (ConejoTM, Henkel Iberica, Barcelona, Spain). The com-mercial solutions or concentrated stock solutions prepared bydissolving the solid compounds in sterile distilled water or inethanol were diluted at least 20-fold in sterile distilled water inorder to prepare the immersion solutions to be used for thewashing treatments. All solutions were prepared fresh before use.

2.3. Assay of enterocin AS-48 in vegetables artificially contaminated

with bacilli

The effect of immersion in solutions containing enterocin AS-48 (25mg/ml, final concentration) on survival and growth ofB. cereus LWL1 inoculated onto fresh alfalfa, soybean sprouts andgreen asparagus as well as B. weihenstephanensis CECT 5894 onsoybean sprouts was investigated at storage temperatures of 6, 15and 22 1C. Cultures of bacilli grown overnight in BHI broth at 37 1Cwere diluted 1:100 in sterile saline solution to a final cell densityof approx. 5.5 log CFU/ml. This Bacillus cell suspension was used toartificially contaminate the vegetables being tested. Fresh greenasparagus (Mary Washington variety, 5–10 mm diameter), soy-bean sprouts (Alleuras Industries, Madrid) and alfalfa sprouts(Productos Fanya, Madrid) were purchased from local super-markets. Asparagus were cut onto 2 cm pieces before treatmentapplication. Samples of the vegetable being tested (2.5 g each)were deposited inside sterile capped 50 ml polypropylene tubes(Sterilin, Stone, UK) and dipped for 5 min in 5 ml sterile distilledwater (negative controls) or in 5 ml of Bacillus cell suspension at 6,15 and 22 1C. Then, they were deposited on a sterile filter paper todrain excess water. The artificially contaminated samples were

dipped for 5 min. at room temperature in 5 ml of sterile distilledwater (controls) or distilled water containing enterocin AS-48 (atfinal concentrations of 5, 12.5 or 25mg/ml). Immersion solutionswere held at room temperature for at least 1 h before use.

ARTICLE IN PRESS

A. Cobo Molinos et al. / Food Microbiology 25 (2008) 762–770 765

Following immersion treatments, excess immersion solution wasremoved as above, and samples were stored in sterile capped50 ml polypropylene test tubes placed in refrigerated storage orincubation chambers (Memmert, Schwabach, Germany) at desiredincubation temperatures (6, 15, 22 1C) for different periods of time.At each step, samples (2.5 g) were mixed with 5 ml of sterile salinesolution (0.85% NaCl) and pummelled for 3 min in a Stomacher 80(Biomaster) before they were serially diluted in sterile salinesolution and spread in triplicate on plates of B. cereus agar(Scharlab, Barcelona) supplemented with egg yolk and polymixinB (Panreac, Barcelona). Plates were incubated at 37 1C for 48 h, andthe number of colonies showing features typical of B. cereus weredetermined in order to calculate viable cell counts.

p-Hydroxibenzoic acid methyl esther + AS-48

p-Hydroxybenzoic acid methyl esther (0.5%)

n-Propyl-p-hydroxybenzoate + AS-48

n-Propyl-p-hydroxybenzoate (0.5%)

Hydrocinnamic acid + AS-48

Hydrocinnamic acid (0.5%)

Cinnamic acid + AS-48

Cinnamic acid (0.3%)

Carvacrol + AS-48

Carvacrol (0.3%)

Polyphosphoric acid + AS-48

Polyphosphoric acid(0.5%)

Polyphosphoric acid + AS-48

Polyphosphoric acid (0.1%)

Peracetic acid + AS-48

Peracetic acid (40 ppm)

Hexadecylpyridinium + AS-48

Hexadecylpyridinium (0.5%)

Trisodium trimetaphosphate + AS-48

Trisodium trimetaphosphate (0.5%)

Sodium hypochlorite + AS-48

Sodium hypochlorite (100 ppm)

Sodiumlactate + AS-48

Sodium lactate (0.5%)

Lactic acid + AS-48

Lactic acid (0.5%)

AS-48

Control

Trea

tmen

t

0

+ *

+

+

+ *

+ *+ *

+ *+ *

+ *

Fig. 3. Effect of enterocin AS-48 (25 mg/ml) in combination with chemical antimicrobial

were determined following application of treatment (open bars) and also after 24 h

statistically significant reduction (Po0.05) compared with the untreated control. *, stat

treatment with the chemical agent alone. Data represent the average of two independ

Combined treatments of enterocin AS-48 and chemical pre-servatives were carried out on Bacillus artificially contaminatedfood samples at room temperature essentially as described above,by using immersion solutions containing enterocin AS-48 (25mg/mlfinal concentration) and/or the corresponding chemical compounds(at the final concentrations indicated in previous heading). Viablecell counts of bacilli were determined as described above followingimmersion treatment (time zero) and after 24 h incubation at 22 1C.Confirmation of B. cereus was done by PCR amplification of the sspE

gene with specific primers sspE1-F (50-GAGAAAGATGAGTAAAAAACAACAA-30) y sspE1-R (50-CATTTGTGCTTTGAATGCTAG) as de-scribed by Kim et al. (2004), followed by detection of the 71-bpamplicon by agarose gel electrophoresis.

Log CFU/g1 2 3 4 5 6

+

+

+

+

++ *

+ *++

+ *+ *+

++ *

+

++ *+ *

++ *+ *++

+++

++

+

compounds on the viability of B. cereus LWL1 in alfalfa sprouts. Viable cell counts

(closed bars) of storage at 15 1C. Standard deviation is shown by error bars. +,

istically significant reduction (Po0.05) of the combined treatment compared with

ent experiments plus standard deviation (error bars).

ARTICLE IN PRESS

0

2

4

6

0 1 3 5 7

Log

CFU

/ml

0

2

4

6

0 1 3 5 7Time (days)

Log

CFU

/ml

A. Cobo Molinos et al. / Food Microbiology 25 (2008) 762–770766

2.4. Co-cultivation experiments

Cells from an overnight culture (grown at 37 1C in BHI broth) ofthe enterocin AS-48 producer strain E. faecalis A-48-32 wererecovered by centrifugation (5000g for 10 min at room tempera-ture) and resuspended in sterile saline solution to the initialculture volume. This suspension was added to the bacillicontamination suspensions to obtain a co-culture suspensionwith an approximate enterococci-to-bacilli ratio of 100:1. Vege-table samples (2.5 g each) were immersed in the co-culturesuspension for 5 min at room temperature as above and excesssuspension was removed as above. At desired intervals ofincubation at the test temperature (15 or 22 1C), samples wereplated for viable cell counts of bacilli and enterococci. Forthe bacilli, B. cereus agar was supplemented with ampicillin(80 mg/ml, final concentration) in order to inhibit growth ofenterococci. Bacilli were confirmed by PCR amplification asdescribed above. Viable cell counts of enterococci were alsocarried out, by spreading samples in triplicate on kanamicinesculin azide agar (KAA, Scharlab), followed by 48 h incubation at37 1C. Confirmation of the AS-48 producer strains from KAA plateswas done by testing colonies isolated at random for bacteriocinproduction by the spot on a lawn method with E. faecalis S-47 asthe sensitive indicator strain and A-48-32 strain as the AS-48-resistant indicator strain. Colonies that showed inhibition againststrain S-47 but failed to inhibit strain A-48-32 were considered tobe AS-48 producers.

Bacteriocin activity in samples was determined by testing100 ml of the pummelled vegetable suspension by the agar-welldiffusion method.

2.5. Statistical analyses

The average data from duplicate trials7standard deviationswere determined with Excel programme (Microsoft Corp., USA). Inorder to determine the statistical significance of the data, a t-testwas performed at the 95% confidence interval with StatgraphicsPlus version 5.1 (Statistical Graphics Corp., USA). The significanceof combined treatments was determined by comparison of datafrom the same incubation time.

Time (days)

0

2

4

6

0 1 3 5 7Time (days)

Log

CFU

/ml

Fig. 4. Effect of enterocin AS-48 (25 mg/ml) in combination with the phenolic

compounds carvacrol (A), cinnamic acid (B) and hydrocinnamic acid (C) on the

viability of B. cereus LWL1 in alfalfa sprouts stored at 15 1C for 1 week. Controls (J),

samples treated with AS-48 (m), phenolic compound (’), or AS-48+phenolic

compound (K). Data represent the average of two independent experiments plus

standard deviation (error bars).

3. Results

3.1. Effect of washing treatments containing enterocin AS-48 alone

The effect of washing treatments with solutions containing25 mg/ml AS-48 was studied on B. cereus LWL1 in soybean, alfalfasprouts and green asparagus (Fig. 1) stored at 6, 15 and 22 1C. Bestresults were obtained for samples refrigerated at 6 1C. In all cases,statistically significant (Po0.05) reductions of viable cell counts of1.06, 1.3 and 1.59 log units were obtained after washing treat-ments, and the remaining viable population was reduced belowdetection levels after days 1–3 of storage (Fig. 1A, D and G). Bycontrast, reductions of viable cell counts obtained after treat-ments at 15 or 22 1C were not statistically significant for the threetypes of food tested, and the remaining viable cells multiplied asthe untreated controls at least for the first day of storage. Duringprolonged storage, viable cell counts of treated samples weresignificantly lower than the untreated controls only for alfalfasprouts and asparagus at day 3 and for soybean sprouts after day 3of storage.

Washing treatments with AS-48 (25mg/ml) were also appliedto B. weihenstephanensis CECT 5894 inoculated on soybean sprouts(Fig. 2). The bacteriocin treatment significantly reduced viable cell

counts by 1.5–2.38 log units at all temperatures tested. In thebacteriocin-treated samples, no viable bacilli were detected afterday 1 of storage at 6 1C (Fig. 2A). However, the surviving

ARTICLE IN PRESS

A. Cobo Molinos et al. / Food Microbiology 25 (2008) 762–770 767

population multiplied during storage at higher temperatures,especially at 15 1C and to a less extent also at 22 1C (Fig. 2B and C).

3.2. Effect of washing treatments containing enterocin AS-48 in

combination with other antimicrobial compounds

Enterocin AS-48 (25mg/ml) was tested on B. cereus incombination with several other antimicrobial compounds inalfalfa sprouts stored at 15 1C. At this concentration, thebacteriocin alone did not reduce significantly the number ofviable B. cereus after treatment or during the following 24 h ofstorage (Fig. 3).

Many of the antimicrobial compounds tested individuallyreduced the numbers of viable cells significantly (Po0.05)compared with the untreated controls, either at time 0 or after24 h storage or both (Fig. 3). Comparison of viable cell countsobtained for each chemical preservative and the combinationbacteriocin–chemical preservative indicated that the bactericidaleffect of treatments was enhanced significantly by addition of AS-48 (Po0.05) for cinnamic and hydrocinnamic acids, carvacrol,polyphosphoric acid (at 0.1% and 0.5%), peracetic acid, hexade-cylpyridinium chloride and sodium hypochlorite, both at 0 and24 h of storage, and also for trisodium trimetaphosphate at 24 h.Interestingly, some of the combined treatments (with sodiumhypochlorite, peracetic acid, polyphosphoric acid and hydrocin-namic acid) reduced the viable cell counts of B. cereus belowdetection levels.

3.3. Preservation of sprouts after treatment

The efficacy of combinations of selected preservatives andenterocin AS-48 for prolonged inhibition of bacilli were tested in

0

2

4

6

0 1 3 5 7Time (days)

Log

CFU

/ml

0

2

4

6

0 1 3 5 7Time (days)

Log

CFU

/ml

Fig. 5. Effect of enterocin AS-48 (25 mg/ml) in combination with the chemical compound

polyphosphoric acid (D) on the viability of B. cereus LWL1 in alfalfa sprouts stored at 15

(’), or AS-48+chemical compound (K). Data represent the average of two independen

alfalfa sprout samples stored at 15 1C for 1 week. In samplestreated with carvacrol, cinnamic and hydrocinnamic acids incombination with AS-48, viable cell counts of B. cereus LWL1remained significantly lower compared with each individualtreatment (Po0.05) during most part or the whole storage period(Fig. 4). Best results were obtained for the combined treatmentwith hydrocinnamic acid, for which no viable cells were detectedexcept at day 2 (Fig. 4C).

The chemical compounds sodium hypochlorite, hexadecylpyr-idinium chloride, peracetic acid and polyphosphoric acid reducedthe population of B. cereus below detection levels for the whole orat least most of the storage period when tested in combinationwith AS-48 (Fig. 5). However, none of the compounds was able tocompletely eliminate B. cereus when tested without bacteriocin.

Antimicrobial compounds were also tested against B. weihen-

stephanensis CECT 5894 inoculated on alfalfa sprouts stored a15 1C. Treatment with bacteriocin alone reduced viable cell countsbelow detection levels from the beginning, but did not preventfurther proliferation of the surviving bacteria (Fig. 6). By contrast,the combinations of 25 mg/ml AS-48 and hydrocinnamic acid,peracetic acid, sodium hypochlorite as well as hexadecylpyridi-nium chloride reduced the population of B. weihenstephanensis

below detection levels for the whole or at least most of the storageperiod (Fig. 6). In the case of polyphosphoric acid, a higherconcentration of 0.5% was required to achieve such a degree ofinactivation (Fig. 6E and F). Addition of the antimicrobialcompounds alone had much more limited effects in most cases.

3.4. Inhibition of B. cereus in sprouts by cocultivation with a

bacteriocin-producing strain

B. cereus LWL1 was inoculated on soybean sprouts in combina-tion with the bacteriocin-producing strain E. faecalis A-48-32 as a

0

2

4

6

0 1 3 5 7Time (days)

Log

CFU

/ml

0

2

4

6

0 1 3 5 7Time (days)

Log

CFU

/ml

s sodium hypochlorite (A), hexadecylpyridinium chloride (B), peracetic acid (C) and

1C for 1 week. Controls (J), samples treated with AS-48 (m), chemical compound

t experiments plus standard deviation (error bars).

ARTICLE IN PRESS

0

2

4

0 1 3 5 7Time (days)

Log

CFU

/ml

0

2

4

0 1 3 5 7Time (days)

Log

CFU

/ml

0

2

4

0 1 3 5 7Time (days)

Log

CFU

/ml

0

2

4

0 1 3 5 7Time (days)

Log

CFU

/ml

0

2

4

0 1 3 5 7Time (days)

Log

CFU

/ml

0

2

4

0 1 3 5 7Time (days)

Log

CFU

/ml

Fig. 6. Effect of enterocin AS-48 (25 mg/ml) in combination with hydrocinnamic acid (A), peracetic acid (B), sodium hypochlorite (C), hexadecylpyridinium chloride (D), 0.1%

polyphosphoric acid (E) and 0.5% polyphosphoric acid (F) on the viability of B. weihenstephanensis CECT 5894 in alfalfa sprouts stored at 15 1C for 1 week. Controls (J),

samples treated with AS-48 (m), chemical compound (’), or AS-48+chemical compound (K). Data represent the average of two independent experiments plus standard

deviation (error bars).

A. Cobo Molinos et al. / Food Microbiology 25 (2008) 762–770768

protective culture at 15 and 22 1C (Fig. 7). Enterococci multipliedrapidly on sprouts both at 15 and 22 1C. In both cases, bacteriocinactivity could be recovered from the sprouts during days 1–3, butnot at other points of storage. In coculture samples stored at 15 1C,growth of B. cereus was completely inhibited for the whole storageperiod, with viable cell counts being significantly lower thanmonocultures for the first 5 days of storage (Fig. 7A). By contrast,cocultivation with the producing strain at 22 1C only producedsome growth inhibition of B. cereus, during the first 3 days ofstorage (Fig. 7B).

4. Discussion

B. cereus is the main aerobic mesophilic endospore former ofconcern in the food industry (Schoeni and Wong, 2005). Thespores of B. cereus survive pasteurization processes of 100 1Cduring 2.2–5.4 min and, in addition, vegetative cells from some

strains can grow at low temperatures, of 4–5 1C (Dufrenne et al.,1995; Choma et al., 2000). Seed sprouts can contain spores of B.

cereus coming from diverse sources (vegetal substrate, earth,fertilizers), that can germinate and reach high populationdensities greater than risk levels. The closely related psychroto-lerant species B. weihenstephanensis is now being investigated as itmay also produce food-poisoning toxins. The need to includeadditional hurdles for the control of B. cereus in the food chaincould be solved by using natural antibacterial substances likebacteriocins.

Previous studies have demonstrated that the addition ofenterocin AS-48 presents a good inhibitory activity against thisbacteria in foods of milk origin, boiled rice and vegetable purees(Munoz et al., 2004; Grande et al., 2006, 2007b). Nevertheless, thepotential of this bacteriocin in the control of B. cereus as well as B.

weihenstephanensis in sprouts had still not been investigated. Theresults obtained in the present study are clearly promising for thedecontamination of these two species, mainly for samples stored

ARTICLE IN PRESS

0

2

4

6

8

10

12

0 1 2 3 5 7Time (days)

Log

CFU

/ml

5

10

15

AS

-48

(mm

)

0

2

4

6

8

10

12

0 1 2 3 5 7Time (days)

Log

CFU

/ml

5

10

15A

S-4

8 (m

m)

Fig. 7. Cocultivation of B. cereus LWL1 and the enterocin AS-48 producer strain E.

faecalis A-48-32 on soybean sprouts stored at 15 1C (A) and 22 1C (B). Viable cell

counts of B. cereus LWL in control monocultures (J) and in cocultures (K) are

shown. Growth of strain E. faecalis A-48-32 in cocultures (m) and bacteriocin

production (bars). Data represent the average of two independent experiments

plus standard deviation (error bars).

A. Cobo Molinos et al. / Food Microbiology 25 (2008) 762–770 769

under refrigeration. However, the treatment with bacteriocin didnot provide any protection under abuse temperature conditions of15 or 22 1C. For this reason, in order to increase the temperaturesafety margins it would be necessary to apply combinedtreatments with other antimicrobial agents. Results from thepreliminary screening carried out in the present study were highlysatisfactory for the combined treatments of AS-48 with thecinnamic and hydrocinnamic acids, carvacrol, polyphosphoricacid, peracetic acid, hexadecylpyridinium chloride and sodiumhypochlorite. For most of them, viable cell counts of B. cereus werereduced below the detection limits. In other cases (lactic acid,lactate sodium, n-propyl p-hydroxybenzoate and p-hydoxyben-zoic acid methyl ester), viable cell counts in the bacteriocin-treated samples were significantly lower compared to the non-treated controls, but not with the chemical preservatives alone,indicating that there was no synergism with the bacteriocin.While previous reports have studied the synergistic effects ofantimicrobial compounds with bacteriocins (reviewed by Galvezet al., 2007), this is the first example of application of washingtreatments against endospore-forming bacteria and also forapplication of a bacteriocin against B. weihenstephanensis.

One of the main problems of decontamination treatments isthat the reductions of viable cell counts obtained may not be

sufficient to avoid growth of survivors and sublethally injuredcells during storage. In the present study, those combinations ofAS-48 that were more effective in reducing viable cell countsduring treatment also provided a very good protection againstB. cereus and B. weihenstephanensis during 1 week storage ofsprouts at 15 1C, reducing the concentration of viable cells belowthe detection limits for most of the samples. These results suggestthat such combinations could be used as highly effectivedecontamination treatments against Bacillus in germinated seedsprouts.

Another possible alternative to control undesired microorgan-isms in foods is based on the application of competitive exclusiontechniques, in which the bacteriocin-producing strains are used toinhibit the growth of pathogenic or toxicogenic bacteria. In thissense, the use of producing strains of enterocin AS-48 has showngood results in the control of B. cereus and S. aureus in cheese aswell as of S. aureus and L. monocytogenes in a meat system (Munozet al., 2004, 2007; Ananou et al., 2005a, b). Enterococci are part ofthe normal microbiota of the human gastrointestinal tract, and arealso found in many different types of foods (revised by FoulquieMoreno et al., 2006). Since there is now a concern that enterococcimay be involved in the spread of antibiotic resistance andvirulence traits in foods, the strains used for food applicationsshould be devoid of such traits. In this respect, previous resultshave shown that the genetic traits coding for enterocin AS-48 canbe transferred artificially into suitable strains devoid of virulencetraits and with suitable technological performance (Fernandez etal., 2007). The results obtained in the present study against B.

cereus in soybean sprouts also indicate that the AS-48 producingstrains could serve to control the growth of B. cereus undersuitable temperature conditions that allow production of suffi-cient amounts of bacteriocin without favouring an excessivegrowth of B. cereus. It is noteworthy the good implantation shownby the bacteriocinogenic strain A-48-32 in soybean sprouts,reaching elevated cell densities within the three first days ofstorage of the samples at 15 and 22 1C. However, the application ofthis strain as a protective culture would be limited by the lowbacteriocin production under refrigeration and the lack ofeffectiveness shown at 22 1C. Therefore, addition of ex-situ

produced bacteriocin (either alone or in combination with otherantimicrobials) remains the best choice to control B. cereus and B.

weihenstephanensis on sprouts.

Acknowledgments

This work was supported by the Spanish Ministry of Education(Research Project AGL2005-07665-C02-02/ALI). We also acknowl-edge the Research Programme of the University of Jaen, and theResearch Plan of the Junta de Andalucıa (Research Group AGR230).

References

Abriouel, H., Valdivia, E., Martınez-Bueno, M., Maqueda, M., Galvez, A., 2003. Asimple method for semi-preparative-scale production and recovery ofenterocin AS-48 derived from Enterococcus faecalis subsp. liquefaciens A-48-32. J. Microbiol. Methods 55, 599–605.

Agata, N., Ohta, M., Mori, M., Isobe, M., 1995. A novel dodecadepsipeptide,cereulide, is an emetic toxin of Bacillus cereus. FEMS Microbiol. Lett. 129, 17–20.

Allende, A., Martınez, B., Selma, V., Gil, M.I., Suarez, J.E., Rodrıguez, A., 2007. Growthand bacteriocin production by lactic acid bacteria in vegetable broth and theireffectiveness at reducing Listeria monocytogenes in vitro and in fresh-cutlettuce. Food Microbiol. 24, 759–766.

Ananou, S., Garriga, M., Hugas, M., Maqueda, M., Martınez-Bueno, M., Galvez, A.,Valdivia, E., 2005a. Control of Listeria monocytogenes in model sausages byenterocin AS-48. Int. J. Food Microbiol. 103, 179–190.

Ananou, S., Maqueda, M., Martınez-Bueno, M., Galvez, A., Valdivia, E., 2005b.Control of Staphylococcus aureus in sausages by enterocin AS-48. Meat Sci. 71,549–576.

ARTICLE IN PRESS

A. Cobo Molinos et al. / Food Microbiology 25 (2008) 762–770770

Bari, M.L., Ukuku, D.O., Kawasaki, T., Inatsu, Y., Isshiki, K., Kawamoto, S., 2005.Combined efficacy of nisin and pediocin with sodium lactate, citric acid, phyticacid, and potassium sorbate and EDTA in reducing the Listeria monocytogenespopulation of inoculated fresh-cut produce. J. Food Prot. 68, 1381–1387.

Baron, F., Cochet, M.F., Grosset, N., Madec, M.N., Briandet, R., Dessaigne, S.,Chevalier, S., Gautier, M., Jan, S., 2007. Isolation and characterization of apsychrotolerant toxin producer, Bacillus weihenstephanensis, in liquid eggproducts. J. Food Prot. 70, 2782–2791.

Beecher, D.J., Wong, A.C.L., 1997. Tripartite hemolysin BL from Bacillus cereus,hemolytic analysis of component interaction and mode for its characteristicparadoxical zone phenomenon. J. Biol. Chem. 272, 233–239.

Beuchat, L.R., Pettigrew, C.A., Tremblay, M.E., Roselle, B.J., Scouten, A.J., 2004.Lethality of chlorine, chlorine dioxide, and a commercial fruit and vegetablesanitizer to vegetative cells and spores of Bacillus cereus and spores of Bacillusthuringiensis. J. Food Prot. 67, 1702–1708.

Choma, C., Clavel, T., Dominguez, H., Razafindramboa, N., Soumille, H., Nguyen-the,C., Schmitt, P., 2000. Effect of temperature on growth characteristics of Bacilluscereus TZ415. Int. J. Food Microbiol. 55, 73–77.

Cobo Molinos, A., Abriouel, H., Ben Omar, N., Valdivia, E., Lucas, R., Maqueda, M.,Martınez Canamero, M., Galvez, A., 2005. Effect of immersion solutionscontaining enterocin AS-48 on Listeria monocytogenes in vegetable foods. Appl.Environ. Microbiol. 71, 7781–7787.

Dufrenne, J., Bijwaard, M., te Giffel, M., Beumer, R., Notermans, S., 1995.Characteristics of some psychrotrophic Bacillus cereus isolates. Int. J. FoodMicrobiol. 27, 175–183.

Fernandez, M., Martınez-Bueno, M., Martın, M.C., Valdivia, E., Maqueda, M., 2007.Heterologous expression of enterocin AS-48 in several strains of lactic acidbacteria. J. Appl. Microbiol. 102, 1350–1361.

Foulquie Moreno, M.R., Sarantinopoulos, P., Tsakalidou, E., De Vuyst, L., 2006. The roleand application of enterococci in food and health. Int. J. Food Microbiol. 106, 1–24.

Galvez, A., Maqueda, M., Valdivia, E., Quesada, A., Montoya, E., 1986. Characteriza-tion and partial purification of a broad spectrum antibiotic AS-48 produced byStreptococcus faecalis. Can. J. Microbiol. 32, 765–771.

Galvez, A., Maqueda, M., Martınez-Bueno, M., Valdivia, E., 1989. Bactericidal andbacteriolytic action of peptide antibiotic AS-48 against gram-positive and gramnegative bacteria and other organisms. Res. Microbiol. 140, 57–68.

Galvez, A., Abriouel, H., Lucas Lopez, R., Ben Omar, N., 2007. Bacteriocin-basedstrategies for food biopreservation. Int. J. Food Microbiol. 120, 51–70.

Galvez, A., Lucas Lopez, R., Abriouel, H., Valdivia, E., Ben Omar, N., 2008.Application of bacteriocins in the control of foodborne pathogenic andspoilage bacteria. Crit. Rev. Biotechnol. 28, 125–152.

Grande, M%a.J., Lucas, R., Valdivia, E., Abriouel, H., Maqueda, M., Martınez-

Canamero, M., Ben Omar, N., Galvez, A., 2005. Stability of enterocin AS-48 infruit and vegetable juices. J. Food Prot. 68, 2085–2094.

Grande, M%a.J., Lucas, R., Abriouel, H., Valdivia, E., Ben Omar, N., Maqueda, M.,

Martınez-Bueno, M., Martınez-Canamero, M., Galvez, A., 2006. Inhibition oftoxicogenic Bacillus cereus in rice-based foods by enterocin AS-48. Int. J. FoodMicrobiol. 106, 185–194.

Grande, M%a.J., Lucas, R., Abriouel, H., Valdivia, E., Ben Omar, N., Maqueda, M.,

Martınez-Canamero, M., Galvez, A., 2007a. Treatment of vegetable sauces withenterocin AS-48 alone or in combination with phenolic compounds to inhibitproliferation of Staphylococcus aureus. J. Food Prot. 70, 405–411.

Grande, M%a.J., Abriouel, H., Lucas, R., Valdivia, E., Ben Omar, N., Martınez-

Canamero, M., Galvez, A., 2007b. Efficacy of enterocin AS-48 againstbacilli in ready-to-eat vegetable soups and purees. J. Food Prot. 70,2339–2345.

Granum, P.E., 2001. Bacillus cereus. In: Doyle, M.P., Beuchat, L.R., Montville, T.J.(Eds.), Food Microbiology. Fundamentals and Applications, second ed. ASMPress, Washington, DC, pp. 373–381.

Kim, H.J., Lee, D.S., Paik, H.D., 2004. Characterization of Bacillus cereus isolates fromraw soybean sprouts. J. Food Prot. 67, 1031–1035.

Lechner, S., Mayr, R., Francis, K.P., Pruß, B.M., Kaplan, T., Wießner-Gunkel, E.,Stewart, G.S.A.B., Scherer, S., 1998. Bacillus weihenstephanensis sp. nov. is a newpsychrotolerant species of the Bacillus cereus group. Int. J. Syst. Bacteriol. 48,1373–1382.

Lindback, T., Fagerlund, A., Rodland, M.S., Granum, P.E., 2004. Characterization ofthe Bacillus cereus Nhe enterotoxin. Microbiology 150, 3959–3967.

Lucas, R., Grande, M%a.J., Abriouel, H., Maqueda, M., Ben Omar, N., Valdivia, E.,

Martınez-Canamero, M., Galvez, A., 2006. Application of the broad-spectrumbacteriocin enterocin AS-48 to inhibit Bacillus coagulans in low-pH canned fruitand vegetable foods. Food Chem. Toxicol. 44, 1774–1781.

Lund, T., De Buyser, M.L., Granum, P.E., 2000. A new cytotoxin from Bacillus cereusthat may cause necrotic enteritis. Mol. Microbiol. 38, 254–261.

Maqueda, M., Galvez, A., Martınez Bueno, M., Sanchez-Barrena, M.J., Gonzalez, C.,Albert, A., Rico, M., Valdivia, E., 2004. Peptide AS-48: prototype of a new classof cyclic bacteriocins. Curr. Protein Pept. Sci. 5, 399–416.

Martinez-Bueno, M., Galvez, A., Valdivia, E., Maqueda, M., 1990. A transferableplasmid associated with AS-48 production in Enterococcus faecalis. J. Bacteriol.172, 2817–2818.

Munoz, A., Maqueda, M., Galvez, A., Martınez-Bueno, M., Rodriguez, A., Valdivia, E.,2004. Biocontrol of psychrotrophic enterotoxigenic Bacillus cereus in a non-fathard type cheese by an enterococcal strain-producing enterocin AS-48. J. FoodProt. 67, 1517–1521.

Munoz, A., Ananou, S., Galvez, A., Martınez-Bueno, M., Rodrıguez, A., Maqueda, M.,Valdivia, E., 2007. Inhibition of Staphylococcus aureus in dairy products byenterocin AS-48 produced in situ and ex situ: bactericidal synergism throughheat and AS-48. Int. Dairy J. 17, 760–769.

Pao, S., Khalid, M.F., Kalantari, A., 2005. Sprouting seeds and potential hazardsassociated with enterotoxigenic Bacillus spp. in homegrown sprouts. J. FoodProt. 68, 1648–1653.

Pruß, B.M., Dietrich, R., Nibler, B., Martlbauer, E., Scherer, S., 1999. The hemolyticenterotoxin HBL is broadly distributed among species of the Bacillus cereusgroup. Appl. Environ. Microbiol. 65, 5436–5442.

Schoeni, J.L., Wong, A.C., 2005. Bacillus cereus food poisoning and its toxins. J. FoodProt. 68, 636–648.

Stenfors, L.P., Mayr, R., Scherer, S., Granum, P.E., 2002. Pathogenic potential of fiftyBacillus weihenstephanensis strains. FEMS Microbiol. Lett. 215, 47–51.

Thorsen, L., Hansen, B.M., Nielsen, K.F., Hendriksen, N.B., Phipps, R.K., Budde, B.B.,2006. Characterization of emetic Bacillus weihenstephanensis, a new cereulide-producing bacterium. Appl. Environ. Microbiol. 72, 5118–5121.

Valero, M., Hernandez-Herrero, L.A., Fernandez, P.S., Salmeron, M.C., 2002.Characterization of Bacillus cereus isolates from fresh vegetables andrefrigerated minimally processed foods by biochemical and physiologicaltests. Food Microbiol. 19, 491–499.