Letters in Applied Microbiology 1998, 26, 161–165

Diarrhoeal enterotoxin production by psychrotrophicBacillus cereus present in reconstituted milk-based infantformulae (MIF)

N.J. Rowan and J.G. AndersonDepartment of Bioscience and Biotechnology, University of Strathclyde, Glasgow, UK

1625/97: received 18 September 1997 and accepted 23 September 1997

N.J. ROWAN AND J.G. ANDERSON. 1998. One hundred reconstituted milk-based infantformulae (MIF) representative of 10 leading brands available in many European EconomicCommunity countries were examined for psychrotrophic Bacillus cereus and for thepresence of diarrhoeal enterotoxin. Of the 38 B. cereus isolates recovered fromMIF, one, four and 16 strains grew at 4, 6 and 8 °C after 15 d. One (2·6%), two (5·3%)and six (15·8%) of the isolates were identified as potential psychrotrophic foodpoisoning strains as they were both enterotoxigenic and exhibited good growth at 4, 6and 8 °C, respectively. Enterotoxin was not detected in MIF in which less than 5·36 log10 cfuof B. cereus ml−1 had grown. While psychrotrophic enterotoxigenic B. cereus strainsoccur occasionally in MIF, brief storage of reconstituted MIF at the recommendedrefrigeration temperature of 4 °C will allow this product to remain safe for consumption.

INTRODUCTION enterotoxin production abilities in refrigerated foods (Inter-national Dairy Federation 1992). There is insufficient evi-

Bacillus cereus, a ubiquitously distributed aerobic spore-dence to establish conclusively whether B. cereus diarrhoeal-

former, tolerates adverse environmental conditions bettertype food intoxication results from consumption of preformed

than most other bacterial pathogens (Rowan 1996). Becker ettoxin (Baker and Griffiths 1995) or from production of

al. (1994) reported that 54% of 261 samples of infant foodenterotoxin by ingested cells or spores in the ileum (Granum

distributed in 17 countries was contaminated with B. cereus,et al. 1993b). However, all of the studies on the stability of

reaching levels of 0·3–600 viable cells g−1. Also, Rowan et al.the diarrhoeal enterotoxins that are cited in the literature have

(1997) demonstrated that hospital prepared infant food maybeen performed in non-food systems, and recent evidence

be contaminated with enterotoxigenic B. cereus at levels abovesuggests that this stability may be significantly greater when

the Association of Dietetic Food Industries of the Europeanthe toxin is preformed in foods such as milk (Baker and

Communities (IDAEC) proposed safety limit of 103 cfu ml−1.Griffiths 1995). The existence is reported here of psychro-

To guard against possible deleterious effects of food-bornetrophic strains of B. cereus capable of growth and enterotoxin

Gram-negative enteric bacteria, hospitals are routinely pas-production in reconstituted milk-based infant formulae.

teurizing many milk-based products (Anderton 1993). Sev-eral researchers have shown, however, that counts of aerobicspore-forming bacteria in these pasteurized foods are littleaffected by this thermal process (Rowan 1996). MATERIALS AND METHODS

Observations that psychrotrophic B. cereus strains havebeen implicated in outbreaks of food-related illness (Turner Strainset al. 1996) and are capable of producing toxins (Christiansson

Bacillus cereus NCTC 11145 and B. licheniformis NCTCet al. 1989 ; Griffiths 1990 ; van Netten et al. 1990 ; Granum10341, obtained from the National Collection of Typed Cul-et al. 1993a), have raised concern about their growth andtures, were used as positive and negative control strains fordiarrhoeal enterotoxin production respectively. CulturesCorrespondence to : Dr N. J. Rowan, Department of Bioscience andwere grown at 30 °C and maintained on Nutrient Agar ; theyBiotechnology, University of Strathclyde, Royal College Building, 204 George

Street, Glasgow G1 1XW, UK. were subcultured every 2 weeks.

© 1998 The Society for Applied Microbiology

162 N.J. ROWAN AND J.G. ANDERSON

Preparation of infant powder samples rhoeagenic B. cereus NCTC 11145 and atoxigenic B.licheniformis NCTC 10341 were used as positive and negative

One hundred milk-based infant formulae products (MIF),controls, respectively. The starting inoculum (reflecting B.

representative of 10 leading brands available in many Euro-cereus counts obtained occasionally from dried MIF products

pean Economic Community countries, were purchased and(Rowan 1996)) was achieved by inoculating MIF with a

examined for the presence of B. cereus and for diarrhoealspectrophotometrically (A600) adjusted (Shimadzu uv-120-

enterotoxin. Triplicate 25 g MIF samples were reconstituted02; Shimadzu Corporation, Kyoto, Japan) spore suspension

in 225 ml of sterile water at 56 °C (20·2 °C) by shaking 25prepared in phosphate buffered saline (PBS, 0·01 mol l−1

times through an excursion of 30 cm and were then incubatedsodium phosphate pH 7·2, 0·15 mol l−1 NaCl), where the test

at 30 °C for periods up to and including 24 h.strains and isolates had been harvested from a 72 h culturegrown at 30 °C on NAMS. Before MIF inoculation, the PBScell/spore suspension was tempered in a waterbath (TechneBacteriological analysisTempette Junior TE-8J; Techne Inc., Princetown, USA) at

Milk-based infant formulae samples were spread and spiral 70 °C for 10 min (simulating the water temperature usedplated (Spiral plater model B, Sprial Systems Inc., Don occasionally to reconstitute infant powder in the homeWhitley Scientific, Ltd, Shipley, UK) onto Tryptone Soya (Rowan 1996)) to eliminate vegetative cells. Microbial num-Agar (supplemented with 0·6% yeast extract), Blood Agar bers in the feeding bottles were confirmed by plating (1 : 10)Base (supplemented with 0·7% defibrinated horse blood), dilutions on NAMS plates that were then incubated at 30 °CNutrient Agar (supplemented with 0·5 mg of MnSO4H2O for 48 h.l−1 to aid sporulation) (NAMS), and B. cereus Selective Agar(Oxoid, Basingstoke, UK) and incubated for 72 h at 30 °C.

Detection of diarrhoeal enterotoxin in MIF samplesBacillus cereus isolates were identified and confirmed by estab-lishing certain key morphological and biochemical charac- Milk-based infant formulae seeded with B. cereus wereteristics as described in Rowan et al. (1997). assessed for the presence of diarrhoeal enterotoxin, with the

BCET-RPLA assay kit, after a non-turbid, fat-free extractwas initially obtained via the Filtron stirred cell ultrafiltrationDetection of diarrhoeagenic B. cereus strainssystem fitted with an Omega 300 kDa cut-off polyether-

After overnight cultivation of B. cereus isolates in Brain Heart sulfone membrane (Filtron Technology Corporation;Infusion broth supplemented with 0·25% filter-sterilized glu- Flowgen, Lichfield, UK).cose (BHIG) at 30 °C for 18 h on a rotary shaker (250 revmin−1), duplicate 1 ml samples were examined in diar-

Statistical analysisrhoeagenic properties via the B. cereus enterotoxin—reversepassive latex agglutination assay kit (BCET-RPLA TD950; Fisher’s exact test was used to compare levels of diarrhoealOxoid) as described in Rowan and Anderson (1997). Bacillus enterotoxin (ng ml−1) and/or microbial numbers (cfu ml−1)cereus isolates producing enterotoxin after cultivation in in MIF. All significant differences were reported at the 95%BHIG were designated as potentially enterotoxigenic. levels of confidence (P³ 0·05).

RESULTS AND DISCUSSIONGrowth of B. cereus isolates in MIF atpsychrotrophic temperatures

Thirty-eight diarrhoeal-type B. cereus isolates were recoveredfrom the 100 MIF products analysed, of which only 10Milk-based infant formulae used to study growth profiles as

affected by temperature of incubation were pre-tempered for (26·3%) isolates were designated as potentially entero-toxigenic, producing varying levels of enterotoxin in BHIGat least 24 h at the temperature used in the challenge. Trip-

licate pre-cleaned and sterilized infant feeding bottles con- after 18 h at 30 °C (results not shown). van Netten et al.(1990) reported that 8% of 483 pasteurized milk samplestaining 250 ml of tyndallized Farleys second milk (a milk-

based infant formula containing maltodextrin shown pre- were positive for B. cereus, with approximately half of theisolates capable of growth at 7 °C.viously to support diarrhoeal enterotoxin production (Rowan

and Anderson 1997)), were seeded with approximately 102 B. This study showed that of the 38 B. cereus isolates seededin MIF at low temperatures, one (2·6%), two (5·3%) and sixcereus spores ml−1 and were then incubated at 4, 6, 8 and

30 °C (a control temperature for toxin production) for periods (15·8%) isolates were identified as potential psychrotrophicfood poisoning strains as they were both diarrhoeagenic andup to and including 15 d.

The study was done in triplicate and the incubation tem- exhibited good growth at 4, 6 and 8 °C, respectively (Table 1).Enterotoxin was detected sooner in MIF incubated at higherperatures varied less than 0·5 °C throughout the study. Diar-

© 1998 The Society for Applied Microbiology, Letters in Applied Microbiology 26, 161–165

TOXIN PRODUCTION BY PSYCHROTROPHIC B. CEREUS IN MIF 163

Table 1 Bacillus cereus isolates supporting diarrhoeal enterotoxin production and/or growth in reconstituted milk-based infant formulae(MIF) after 15 days at ¾8 °C and 2 days at 30 °C—––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––

ng of enterotoxin ml−1* (no. of cells†) in MIF—–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––Incubation temperature (°C)—–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––

Strains Source of strains 4 6 8 30—––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––SU02/95 SMA white cap —‡ — (5·75) — (7·98) — (8·17)SU03/95 Nutrilon plus — (4·76) 13 (6·52) 43 (7·37) 171 (7·48)SU11/95 Nutrilon premium — — 32 (8·08) 213 (8·00)SU27/95 Nutrilon premium — — — (7·46) — (7·78)SU30/95 Milupa milumil — — 96 (7·80) 226 (8·04)SU36/95 Milupa milumil — — — 128 (7·96)SU38/95 SMA gold cap — — — (7·64) — (7·58)SU46/95 Nutrilon plus — — — 213 (8·08)SU48/95 Milupa aptamil — — — (6·45) — (7·76)SU53/95 Nutrilon plus — — — 64 (7·37)SU58/95 Farleys first milk — — — (5·88) — (7·48)SU62/95 Nutrilon premium — — 107 (7·28) 192 (7·56)SU70/95 Milupa aptamil — — — (7·77) — (7·80)SU99/95 Farleys second milk — 3 (6·25) 96 (7·91) 64 (7·84)SU108/96 Nutrilon plus — — — (6·04) — (8·08)SU109/96 SMA white cap — — — 132 (7·67)SU114/96 SMA white cap — — (6·37) — (7·57) — (8·08)SU123/96 Nutrilon premium — — — (7·42) — (7·45)SU125/96 Farleys first milk — — 16 (6·78) 128 (7·85)SU126/96 SMA gold cap — — — (8·17) — (8·08)11145 NCTC — — — 64 (7·37)Bacillus licheniformis NCTC — — — —(10341)—––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––

* Detected by the BCET-RPLA system, where values are averages of three replicate trials.† Measured as log10 cfu ml−1, where counts are averages of three replicate trials.‡ No significant growth (P ³ 0·05) or enterotoxin detected.

temperatures, where tests became positive after 23 d at 4 °C, stiansson et al. 1989 ; Griffiths 1990 ; van Netten et al. 1990 ;Granum et al. 1993a), and convenience meals (Griffiths 1990 ;15 d at 6 °C and 10 d at 8 °C (Table 2). However, enterotoxin

was not detected until B. cereus populations reached 5·36 log10 Beuchat et al. 1997) at or near refrigeration temperatures.These strains are causing increasing concern in the foodcfu ml−1. This appears to be in agreement with the infective

dose proposed by Kramer and Gilbert (1989), where the industry as they have been implicated in a number of foodpoisoning incidents (van Netten et al. 1990). Granum et al.authors stated that at least 105 cfu ml−1 (g−1) were necessary

to cause illness. In an outbreak of B. cereus gastroenteritis (1993a) demonstrated that 6% of 85 strains of psychrotrophicB. cereus (isolated from Norwegian dairy products) werelinked to the consumption of stew, the estimated infective dose

was 104–105 cfu g−1 (Beuchat et al. 1997). enterotoxigenic, while Christiansson et al. (1989) showedthat 75% of 136 B. cereus isolates produced an extracellularWhile typical environmental isolates of B. cereus do not

normally grow at temperatures below 10 °C (Granum et al. verocytotoxin when grown in milk or whipped cream at 8 °C.In conclusion, as infant foods are considered to be a food1993a), other research groups have recently reported the

existence of psychrotrophic strains capable of diarrhoeal class of high risk due to the susceptibility of infants to entericbacterial pathogens, their response to enterotoxins andenterotoxin production in microbiological media (Chri-

stiansson et al. 1989 ; Griffiths 1990), dairy products (Chri- increased mortality (International Dairy Federation 1992),

© 1998 The Society for Applied Microbiology, Letters in Applied Microbiology 26, 161–165

164 N.J. ROWAN AND J.G. ANDERSON

Table 2 Titres of diarrhoealenterotoxin produced inreconstituted MIF inoculated withspores of Bacillus cereus andincubated for up to 15 days at 4, 6and 8 °C

—–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––

ng of enterotoxin ml−1* (no. of cells†) in MIF atincubation time (days)

Incubation —––––––––––––––––––––––––––––––––––––––––––Strains temperature (°C) 3 6 10 15—–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––SU03/95 4 — — (2·65) — (3·52) — (4·76)§

6 — — (3·36) — (4·93) 13 (6·52)8 — (2·86) — (4·22) 5 (5·91) 43 (7·37)

SU11/95 4 — — — —6 — — — —8 — (2·94) — (4·42) 3 (6·26) 32 (8·08)

SU30/95 4 — — — —6 — — — —8 — (2·60) — (3·90) — (5·65) 96 (7·80)

SU62/95 4 — — — —6 — — — —8 — — (3·45) — (5·45) 107 (7·28)

SU99/95 4 — — — —6 — — (3·08) — (4·60) 3 (6·25)8 — (3·00) — (4·50) 13 (6·35) 96 (7·91)

SU125/96 4 — — — —6 — — — —8 — — (3·18) — (5·17) 16 (6·78)

NCTC 11145 4 — — — —6 — — — —8 — — — —

—–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––

* Detected by the BCET-RPLA system, where values are averages of three replicatetrials.† Measured as log10 cfu ml−1, where counts are averages of three replicate trials.‡ No significant difference (P ³ 0·05) or enterotoxin produced.§ 16 ng of enterotoxin ml−1 detected after 23 days at 4 °C.

Beuchat, L.R., Clavero, M.R.S. and Jaquette, C.B. (1997) Effectsimproper cold storage of infant formulations may pose aof nisin and temperature on survival, growth and enterotoxinthreat to consumer safety as the products may be con-production characteristics of psychrotrophic Bacillus cereus intaminated with psychrotrophic B. cereus capable of bothbeef gravy. Applied and Environmental Microbiology 63, 1953–growth and enterotoxin production in these products.1958.

Christiansson, A., Naidu, A.S., Nilsson, I., Waldstrom, T. andPetterson, H.-E. (1989) Toxin production by Bacillus cereus dairyACKNOWLEDGEMENTSisolates in milk at low temperatures. Applied and Environmental

The authors thank the Scottish Office, Home and Health Microbiology 55, 2595–2600.Department (K/MRS/50/C2205) for funding this research. Granum, P.E., Brynestad, S. and Kramer, J.M. (1993a) Analysis of

enterotoxin production by Bacillus cereus from dairy isolates, foodpoisoning incidents and non-gastrointestinal infections. Inter-REFERENCESnational Journal of Food Microbiology 17, 269–279.

Granum, P.E., Brynestad, S., O’Sullivan, K. and Nissen, H. (1993b)Anderton, A. (1993) Bacterial contamination of enteral feeds andEnterotoxin from Bacillus cereus : production and biochemicalfeeding systems. Clinical Nutrition 12, 97–113.characterisation. Netherlands Milk Dairy Journal 47, 63–70.Baker, J.M. and Griffiths, M.W. (1995) Evidence for increased

Griffiths, M.N. (1990) Toxin production by Bacillus spp. presentthermostability of Bacillus cereus enterotoxin in milk. Journal ofin milk. Journal of Food Protection 53, 790–792.Food Protection 58, 443–445.

International Dairy Federation (1992) Bacillus cereus in milk andBecker, H., Schaller, G., von Wiese, W. and Terplan, G. (1994)milk products. In Bulletin No. 275. International Dairy FederationBacillus cereus in infant foods and dried milk products. Inter-

national Journal of Food Microbiology 23, 1–15. pp. 1–48. Brussels, Belgium.

© 1998 The Society for Applied Microbiology, Letters in Applied Microbiology 26, 161–165

TOXIN PRODUCTION BY PSYCHROTROPHIC B. CEREUS IN MIF 165

Kramer, J.M. and Gilbert, R.J. (1989) Bacillus cereus and other Rowan, N.J., Anderson, J.G. and Anderton, A. (1997) The bac-teriological quality of hospital-prepared infant feeds. Journal ofBacillus species. In Foodborne Bacterial Pathogens ed. Doyle, M.P.

pp. 22–70. New York : Marcel Dekker. Hospital Infection 35, 259–267.Turner, B.E., Foegeding, P.M., Larick, D.K. and Murphy, A.H.Rowan, N.J. (1996) Studies on the growth, survival, interaction and

detection of potentially pathogenic Listeria and Bacillus species (1996) Control of Bacillus cereus spores and spoilage microflora insous vide chicken breast. Journal of Food Science 61, 217–219,in infant milk formulae. Ph.D. thesis, University of Strathclyde,

Glasgow, UK. 234.van Netten, P., van de Moosdijk, A., van Hoensel, P., Mossel,Rowan, N.J. and Anderson, J.G. (1997) Maltodextrin stimulates

growth of Bacillus cereus and synthesis of diarrhoeal enterotoxin D.A.A. and Perales, I. (1990) Psychrotrophic strains of Bacilluscereus producing enterotoxin. Journal of Applied Bacteriology 69,in infant milk formula. Applied and Environmental Microbiology

63, 1182–1184. 73–79.

© 1998 The Society for Applied Microbiology, Letters in Applied Microbiology 26, 161–165