Caballero B. (ed.) Encyclopaedia of Food Science, Food Technology and Nutrition. Ten-Volume Set

Подождите немного. Документ загружается.

BACILLUS

Contents

Occurrence

Detection

Food Poisoning

Occurrence

R Leuschner, Central Science Laboratory, York, UK

Systematics and Ecology of

Bacillus

Species

0001 Aerobic endospore-forming bacteria are currently

assigned to four genera in the family Bacillaceae.

Within this family, the genus Bacillus was established

to include the rod-shaped bacteria that grew in the

presence of air, thus distinguishing them from the

strictly anaerobic Clostridium spp. It is possible to

allocate many Bacillus species to one of six taxa that

have distinguishable physiologies and this is generally

consistent with the devision of the genus based on

spore morphologies. The six groups are: B. polymyxa

group (I), B. subtilis group (II), B. brevis group (III),

B. sphaericus group (IV), and thermophiles (V and

VI). In group I all species are facultative anaerobes

and grow strongly in the absence of oxygen. Acid is

produced from a variety of sugars. Endospores are

ellipsoidal. The group II species (B. subtilis group) are

phylogenetically and phenetically consistent. B. sub-

tilis is an appropriate representative of the taxon that

includes many common names. All these bacteria

produce acids from a range of sugars and some,

B. cereus and B. licheniformis, are facultative anaer-

obes. B. licheniformis grows poorly anaerobically

and can use glucose only under anaerobic conditions.

Although B. subtilis is generally regarded as an

aerobe, it can grow and sporulate slowly under strict

anaerobic conditions. Given glucose, with nitrite as a

terminal electron acceptor, it grows strongly anaero-

bically. These bacteria are therefore an intermediate

stage between the true facultative anaerobes of the

group I strains and the strict aerobes in groups III and

IV. This is reflected in their production of acid from

several sugars. All these bacteria produce oval endo-

spores that do not swell the mother cell and are

generally located centrally or subterminally. Group

III represents strict aerobes that generally do not pro-

duce acid from sugars, with the exception of two

species. They produce ellipsoidal spores that swell

the mother cell. In group IV all species produce spher-

ical spores that may swell the mother cell and contain

l-lysine or ornithine in the cell wall. All species

are strictly aerobic, but some have limited ability

to produce acid from sugars. Group V represents

thermophilic species that grow optimally at > 50



Physiologically and morphologically, they are hetero-

geneous, but most produce oval spores that swell the

mother cell. In group VI are thermophilic, acidophilic

species with membraneous o-alicyclic fatty acid.

Occurrence in the Environment and Food

0002Members of the genus Bacillus have a ubiquitous

environmental distribution. The endospore is import-

ant for the dispersion of Bacillus spp. Spores are

readily blown about in dust and air currents and

are prevalent in animal feces. Bacteria related to

B. subtilis are commonly encountered and easily iden-

tified. The soil is the reservoir of these bacteria. From

the soil, they are transferred to various associated

environments, including plants and plant materials,

foods, animals, and marine and fresh-water habitats.

B. cereus, B. licheniformis, B. pumilus,andB. subtilis

are prevalent in soils, particularly low-nutrient soils.

They are also common on straw and cereals, includ-

ing rice and pulses, which they presumably colonize

from wind-blown soil particles and dust. Endospores

of bacilli represent a metabolically inactive survival

form. They are characterized by high resistance

against heat, dryness, irradiation, and other unfavor-

able environmental conditions. The success of the

endospore for the survival of the species depends on

an effective mechanism to resuscitate and enter the

vegetative cell cycle to multiply. It is usually the case

that spores resuscitate under good environmental

conditions that allow germination and cell growth.

Food of various origin and compositions can offer

these conditions and this can lead to spoilage or

even food poisoning due to growth of bacilli in food.

0003 Spores and vegetative cells of Bacillus cereus and

other food-poisoning Bacillus spp. occur widely in

soils (10

–10

CFU g

1

) and may be found in raw

materials and processed foods which were not steril-

ized by heat or irradiation. They constitute a major

portion of the microbial flora of raw milk and easily

contaminate various dairy products, causing spoilage

with their proteolytic, lipolytic, and saccharolytic ac-

tivities. B. cereus was detected and enumerated in

milk, vegetable, and meat-based products (Table 1).

B. cereus is a problem to the dairy industry because it

contaminates the udders of cows while they graze in

the fields and can be introduced into milk. B. licheni-

formis, B. pumilus and B. subtilis were the most

commonly isolated Bacillus species in bakeries and

milk (Tables 2 and 3). The incidence of B. cereus and

B. subtilis in various food products was investigated

in the Netherlands. The total number of products was

229 samples and B. cereus and B. subtilis were isol-

ated at contamination levels between 10

and 10

CFU g

1

or ml in a wide range of food products such

as milk, yeast, flour, pasta products, cocoa, chocolate,

bakery products, meat products, and herbs and

spices. B. cereus was present in 48% and B. subtilis

tbl0001 Table 1 Prevalence of Bacillus cereus in raw and processed

food products

Foodproduct No. of

samples

examined

No. positive

for B. cereus

(%)

B. c ereus

counts g

1

or ml

1

Rice and oriental foods

Raw rice 13 46 10

–10

Boiled rice 32 38 10

–10

14 93 10

–10

252 10 10

–10

Fried rice 14 86 10

–10

204 24 10

–10

Egyptian rice dishes 172 40 10

–10

Japanese noodles 200 8 10

–10

Soybean curd (tofu) 257 56 10

–10

467 31 10

–10

Sashimi (raw fish) 228 14 10

–10

Milk and dairy products

Raw milk 100 9 10

–10

Pasteurized milk 100 35 10

–10

Milk powder 120 27 10

–10

Icecream 100 48 10

–10

Spices 110 53 10

–10

Meat and meat products

Raw meats 133 2 10

–10

452 6 10

–10

Modified from Kramer JM and Gilbert RJ (1989) Bacillus cereus and other

Bacillus species. In: Doyle MP (ed.) Foodborne Bacterial Pathogens, pp. 21–

70. New York: Marcel Dekker.

tbl0002Table 2 Distribution of 170 bakery Bacillus isolates identified

using API 50CHB

Isolate Bread ingredients

and uncooked

dough (%)

Swabs of

processing

line (%)

Cooked

loaves and

bread

crumbs (%)

B. subtilis 22.9 4.1 5.3

B. amyloliquefaciens 1.2 1.2 0

B. li cheniformi s 18.2 5.9 4.1

B. p umilus 11.2 2.4 2.4

B. circulan s 2.9 0.6 0.6

B. me gaterium 1.8 1.2 0.6

B. p olymyxa 2.9 0 0.6

B. macera ns 2.4 0 0.6

B. c ere u s (including

B. mycoides)

0.6 1.8 0

B. stearothermophilus 0 0 0.6

Unacceptable profile 0.6 2.4 1.2

Reproduced from Thompson JM, Dodd CER and Waites WM (1993)

Spoilage of bread by Bacillus. International Biodeterioration and

Biodegradation 32: 55–66, with permission.

tbl0003Table 3 Mesophilic Bacillus species isolated from milk

samples

Bacillus species No. of sampleswhere a specieswas

isolated (total number of samples: n ¼250)

B. li cheniformi s 131

B. p umilus 156

B. subtilis 127

B. lentus 52

B. pantothenticus 33

B. amyloliquefaciens 25

B. stearothermophilus 19

B. mycoides 22

B. c ere u s 16

B. circulan s 15

B. firmus 10

B. me gaterium 27

B. shaericus 6

B. macera ns 3

B. later o sporus 2

B. p olymyxa 5

Unknown 120

Reproduced from Sutherland AD and Murdoch R (1994) Seasonal

occurrence of psychrotrophic Bacillus species in raw-milk, and studies on

the interactions with mesophilic Bacillus sp.

356

BACILLUS

/Occurrence

in 25% of all samples examined. In spoiled bread

that developed symptoms of ropiness after 2 days’

storage at ambient summer temperatures, identified

species were B. subtilis (70%), B. licheniformis (24%),

B. pumilus (2%), and B. cereus (2%).

Food Spoilage and Poisoning

0004 The occurrence of Bacillus species in raw materials

used for food processing is generally below the infec-

tious dosis required to affect negatively the well-being

of consumers: this is normally above 10

cells g

1

food (Table 4). This indicates that Bacillus cells

or spores must have the opportunity to multiply in the

food environment to enable them to cause food spoil-

age and poisoning.

0005 B. cereus food poisoning is principally associated

with the storage of cooked foods at temperatures and

times that allow growth and result in cell numbers

above 10

CFU g

1

or ml

1

. Rapid cooling of cooked

food below 10



C is generally an effective control

measure. Low-pH foods (< pH 5.0) and dry foods

will not support the growth of B. cereus, although

many dried foods will be contaminated with spores of

this microorganism. Spores are more hydrophobic

than spores of other Bacillus spp., which enables

them to adhere particularly well to many types of

surfaces. This makes them difficult to remove during

cleaning and a difficult target for disinfection.

0006 B. cereus can cause infections and intoxications. In

addition to foodborne diseases, B. cereus causes septi-

cemia, meningitis, and ocular infections. It can cause

two types of food poisoning: the emetic disease, char-

acterized by a short incubation period (1–5 h) display-

ing symptoms such as nausea, vomiting, and stomach

cramps, and the diarrheal disease, characterized by an

incubation period of 8–16 h with symptoms including

abdominal pain, watery diarrhea, and rectal tenes-

mus. Usually both types of food poisoning are rela-

tively mild and last for less than 24 h.

0007 B. subtilis was reported to be the cause of illness in

meat dishes with elements of vegetables, seafood and

rice, bread and pastry products, sandwiches, and

pizza. The infectious dose was > 10

CFU g

1

with

an incubation period of 10 min to 14 h and a duration

of illness of 2–8 h. Symptoms were vomiting (80%),

diarrhea (49%), abdominal pain/cramps (27%), and

nausea, headaches, flushing, and sweating. Bacilli are

capable of causing food spoilage in bread, and this is

known as ropiness. B. subtilis has been described to

be the main cause of the bread spoilage ‘ropiness.’

Ropiness is the most important spoilage of bread after

moldiness and occurs particularly during the summer

when the climatic conditions favor growth of the

bacteria. This involves mainly B. subtilis. Bacillus

counts in white and wholemeal wheat loaves pro-

duced without preservatives or sourdough were con-

sistently 10

CFU g

1

after 2 days of storage at

ambient summer temperatures.

0008B. licheniformis was identified in meat dishes with

elements of vegetables, bread, and pastry products,

and chicken as a cause of food poisoning. The infec-

tious dose was > 10

CFU g

1

, the incubation period

2–14 h and the duration of illness 6–24 h. Symptoms

were vomiting (54%), diarrhea (92%), and abdom-

inal pain/cramps (46%). B. pumilus in meat products,

sandwiches, and canned tomato juice was found to be

responsible for food poisoning. The infectious dose

was > 10

CFU g

1

and the incubation period 15 min

to 11 h. Symptoms were vomiting and diarrhea.

0009B. stearothermophilus and B. coagulans were

reported to be the cause of flat sour spoilage in evap-

orated milk.

Growth Characteristics and Resistance of

Bacilli

0010Knowledge about the resistance and growth charac-

teristics of Bacillus species is important to enable food

manufacturers and consumers to avoid handling

practice that enable spores to germinate and vegeta-

tive cells to multiply in a food.

0011Psychrotrophic strains of B. cereus were able to

grow and produce diarrheal toxin at temperatures

down to 4



C. Most mesophilic strains are able to

grow in low-acidic foods at temperatures between

15 and 55



C, with an optimum range of 30–40



Growth can generally occur between pH 5.0 and 8.8,

tbl0004 Table 4 Infective or intoxication dosis food poisoning caused by Bacillus species

Bacillus species Intoxication dosis Infective dosis

B. c ere u s (emetic toxin) 12–32 mgkg

1

toxin (Suncus murinus: a small monkey) 10

–10

cells g

1

food

B. c ere u s (diarrheal disease) Usual 10

1

or ml

1

B. subtilis >10

B. li cheniformi s >10

B. p umilus >10

Modified from Granum PE and Baird-Parker TC (2000) Bacillus species. In: Lund BM, Baird-Parker TC and Gould GW (eds) The Microbiological Safety and

Quality of Food, vol. 2, pp. 1029–1039. Maryland: Aspen.

BACILLUS

/Occurrence 357

with an optimum between 6.0 and 7.0. Spores are

variable in their heat resistance, which is normally

moderate, but can be increased by fat components.

The emetic toxin is very heat-resistant and can survive

heating at 126



C for 90 min. The diarrheal toxin is

heat-sensitive and is inactivated at 56



C after 5 min.

Growth of B. cereus was found at water activities of

0.93 depending on the acidulant and humectant.

Other food-poisoning Bacillus spp. have similar

growth and resistance characteristics to B. cereus. B.

licheniformis and B. subtilis were not found to grow

at pH 4–4.2; however, growth in tomato juice at a pH

of 4.4 in the presence of oxygen was reported. Spores

of thermophilic bacilli such as B. stearothermophilus

are far more heat-resistant and are used to test steril-

ization effectiveness. Heat resistance of bacilli spores

is dependent on the temperature during spore forma-

tion and on the pH of the food. Further dependence

was observed on various acidulants (Table 5).

0012 Sterilization will inactivate all Bacillus spores. This

is however also imposing a harsh treatment towards

other valuable food ingredients. The development of

novel and milder food preservation technologies, such

as high-pressure and electromagnetic fields, have

proven to be effective in Bacillus spore inactivation.

Pressure treatments of 400 MPa for 25 min at 30



resulted in a 0.45 log inactivation of B. cereus spores.

Pressure treatments at lower temperatures were less

effective. Spores of B. subtilis proved to be more

resistant to ultraviolet B range (280–330 nm) than

spores of B. cereus. An exposure of 30 min reduced

the viability of spores of B. cereus by 50% and that

of B. subtilis spores by 10%. Levels of resistance to

ultraviolet B of spores from different Bacillus species

appear to be related to the quantity and quality of

small acid-soluble proteins and to activities of DNA

repair systems. The novel, milder nonthermal preser-

vation technologies need further evaluation with par-

ticular emphasis on spore inactivation mechanisms

and synergistic effects by combination of various

techniques in food systems.

See also: Bacillus: Detection; Food Poisoning; Food

Poisoning: Classification; Spoilage: Bacterial Spoilage

Further Reading

Claus D and Berkeley RCW (1986) Genus Bacillus. In:

Sneath PHA (ed.) The Bergey’s Manual of Systematic

Bacteriology, vol. 2, pp. 1105–1139. Baltimore: Wil-

liams & Wilkins.

Granum PE and Baird-Parker TC (2000) Bacillus species.

In: Lund BM, Baird-Parker TC and Gould GW (eds) The

Microbiological Safety and Quality of Food, vol. 2, pp.

1029–1039. Maryland: Aspen.

International Commission on Microbiological Specifica-

tions for Foods (1996) Bacillus cereus. In: Roberts TA,

Baird-Parker AC and Tompkin RB (eds) Microbiological

Characteristics of Food Pathogens, pp. 335–344.

London: Blackie Academic and Professional.

Kramer JM and Gilbert RJ (1989) Bacillus cereus and other

Bacillus species. In: Doyle MP (ed.) Foodborne Bacterial

Pathogens, pp. 21–70. New York: Marcel Dekker.

Priest FG (1993) Systematics and ecology of Bacillus. In:

Sonenshein AL, Hoch JA and Losick R (eds) Bacillus

Subtilis and Other Gram Positive Bacteria, pp. 3–16.

Washington, DC: American Society of Microbiology.

Detection

L A Shelef, Wayne State University, Detroit, MI, USA

Background

0001The genus Bacillus is large, comprising more than 60

species that are mostly saprophytes, widely distrib-

uted in nature, spreading from soil to water, plants,

and animals. The genus shows a great diversity of

strains and species. The organisms are Gram-positive

or Gram-variable spore-forming bacilli, mostly cata-

lase-positive, that may be motile by peritrichous

flagella. Most are mesophiles, but some are psychro-

throphs and thermophiles. Bacillus contains strict

aerobes (e.g., B. megaterium), as well as facultative

anaerobes (e.g., B. cereus, B. licheniformis). The

vegetative cells are rods ranging from 0.5 1.2 mm

to 2.5 10 mm, and the endospores are in the central

or paracentral, subterminal or terminal position. Sur-

vival of the organism results from the resistance of the

spores to adverse conditions.

0002Of the identification schemes proposed, that of

Gordon and co-workers divides the genus Bacillus

into three groups, according to cellular morphology

and physiological properties. Group 1 consists of

tbl0005 Table 5 Decimal reduction (D) values (min) at 95



CofBacillus

cereus spores for various pH and organic acid types

pH Citric acid Lactic acid Acetic acid Malic acid

6.5 1.03 1.09 1.21 0.96

6 0.95 0.85 0.99 0.96

5.5 0.81 0.82 0.82 0.77

5 0.74 0.66 0.67 0.69

4.5 0.64 0.54 0.55 0.62

4 0.62 0.40 ND 0.59

ND, not determined.

Modified from Leguerinel I and Mafart P (2001) Modelling the influence of

pH organic acid types on thermal inactivation of Bacillus cereus spores.

International Journal of Food Microbiology 63: 29–34.

358

BACILLUS

/Detection

Gram-positive species that form ellipsoidal or cylin-

drical spores that do not appreciably distend the spor-

angia. This group is further divided according to the

vegetative cell dimensions and presence of lipid glob-

ules in the protoplasm of the species. The ‘large-

celled’ group 1 Bacillus species have a cell width

greater than 0.9 mm and include B. megaterium,

B. cereus, B. cereus var. mycoides, B. thuringiensis,

and B. anthracis. The ‘small-celled’ species have a cell

diameter < 0.9 mm, and lipid globules are not formed.

They include B. subtilis, B. pumilus, B. licheniformis,

and B. coagulans. Organisms of groups 2 and 3 are

characterized by swollen sporangia and either ellips-

oidal (group 2) or spherical (group 3) spores. Studies

of DNA composition and other studies suggest that

the genus Bacillus consists of some bacterial families

that will necessitate further classification of the genus.

0003 Most Bacillus species are harmless saprophytes

characterized by their proteolytic and saccharolytic

properties and are rarely associated with human

or animal disease. However, several species are

proven or suspect pathogens. Of these, B. cereus is a

recognized food pathogen, and other species, particu-

larly those of the B. subtilis group (B. subtilis,

B. licheniformis, and B. pumilus), have been impli-

cated in foodborne illnesses. B. brevis, also identified

as a causative agent of foodborne disease, has

been split and transferred to the new genus Breviba-

cillus. B. anthracis is the agent of anthrax.

Five Bacillus species, B. thuringiensis, B. popilliae,

B. sphaericus, B. larvae, and B. lentimorbus, are

insect pathogens, the first three being used in com-

mercial insecticides for the control of crop insects.

0004 Of the Bacillus species associated with foods, de-

tection and identification methods have focused on

B. cereus as a recognized food pathogen. Spores and

vegetative forms of B. cereus are frequent inhabitants

of soils, sediments, dust, natural water sources, vege-

tation, and many foods such as grains, dairy prod-

ucts, dried foods, spices, meats, and vegetables.

Examination of various raw and processed foods

shows contamination levels ranging from 10

–10

cells per gram or milliliter. It is estimated

that B. cereus accounts for *2–3% of reported food-

borne outbreaks in the USA. In the UK, a single

outbreak was reported during the period of Octo-

ber–December, 2000. In this outbreak, 30 people

became ill after consuming contaminated soup with

cream in a canteen, and a statistical association was

established between consumption of the food and the

cause.

0005 Detection of B. cereus and its toxins and differenti-

ation between the species within the B. cereus group

are presented in this chapter. Following a brief review

of the properties of the organism and its toxins,

traditional and novel detection of the organism and

its toxins are discussed.

Properties of

B. cereus

0006Cells of B. cereus are large, Gram-positive rods that

are motile by means of peritrichous flagella. They are

typically 1.0–1.2 mm in diameter and 3.0–5.0 mmin

length. Endospores are in either the central or para-

central position without distention of the sporan-

gium. The organism sporulates freely on many

media under aerated conditions. The optimum

growth temperature is 28–35



C, with a minimum of



C and a maximum of 48



C, and the pH for growth

is 4.9–9.3. B. cereus strains are able to utilize glucose,

fructose, and trehalose. Sucrose, salicin, maltose,

mannose, glycerol m-inositol, and lactose are utilized

by some, but not all, members of the species. Pen-

toses, mannitol, and many of the sugar alcohols are

not metabolized. Urease is produced by few strains.

Most of the strains produce acetylmethylcarbinol

(Voges–Proskauer-positive), reduce nitrate to nitrite,

hydrolyze soluble starch, tyrosine, gelatin and casein,

and produce a neutral metalloprotease. B. cereus and

B. thuringiensis produce a broad spectrum of

b-lactamase and are, therefore, resistant to penicillin,

ampicillin, and cephalosporins. Their ability to grow

on agar containing penicillin aids in differentiating

them from B. anthracis. Certain strains have been

reported to synthesize a dehydropeptide reductase

that catalyzes cleavage of the bacteriocins nisin and

subtilin. In contrast to B. megaterium, B. cereus

strains are not sensitive to lysozyme.

0007The organism elaborates a number of toxins with

distinct diarrheal and emetic syndromes. The diar-

rheal syndrome is caused by bacterial growth and

toxin formation in the small intestine, whereas the

emetic syndrome is caused by preformed toxin

resulting from growth of the pathogen in the food.

Of the enterotoxins responsible for the diarrheal syn-

drome, hemolysin, designated BL, comprises three

heat-labile peptides, B (37.8 kDa), L

(38.5 kDa),

and L

(43.2 kDa), and the genes that encode for

these components have been cloned. The L

and L

peptides appear to be responsible for binding to

erythrocytes and the B component for lysing them.

Cereolysin (56 kDa), a cytolytic protein responsible

for hemolysis, consists of phospholipases that hydro-

lyze phosphatidylinositol, some of which are metal-

loenzymes requiring divalent cations for activity, and

sphingomyelinase, also a metalloenzyme that has

hemolysin-like activity, but shows no phospholipase

activity. The nonhemolytic enterotoxin is heat-labile,

comprising three proteins (39, 45, and 105 kDa).

Cereulide is a putative emetic toxin. It has been

BACILLUS

/Detection 359

isolated and identified as a small (1.2 kDa), heat-

stable dodecadepsipeptide, structurally related to the

ionophore valinomycin, producing a vacuole re-

sponse in HEp-2 cells. Psychrophilic strains of B.

cereus capable of producing toxins at refrigerator

temperatures have been reported.

Isolation and Identification of

B. cereus

Standard Agar Media

0008 The characteristics of B. cereus and of the culturally

similar B. mycoides, B. thuringiensis,andB. anthra-

cis are summarized in Table 1. Most procedures for

the isolation, identification, and enumeration of

B. cereus involve direct plating techniques. Media

developed for the identification of the organism pro-

vide conditions for the organism to exhibit fermenta-

tion properties, hemolysin production, lecithinase

activity, and morphological characteristics. Selective

agents to inhibit competitive microorganisms, such as

antibiotics (most often 10 mg of polymyxin per milli-

liter), are normally incorporated into the media.

0009 MYP agar Mannitol–egg yolk-polymyxin agar

(MYP) was the first selective medium, introduced in

1967 by Mossel and coworkers. It relies on lecithin

hydrolysis (the egg-yolk reaction), negative mannitol

reaction for differentiation of B. cereus from other

commonly occurring Bacillus species, and polymyxin

B. The agar plate is surface-inoculated, and typical

colonies of B. cereus after 20–24 h of incubation at

30–32



C have a dry, flat, ‘ground-glass’ appearance,

translucent to creamy white with a violet–red back-

ground surrounded by a readily visible zone of egg-

yolk precipitate (Figure 1). Less commonly, colonies

may be large (5 to over 10 mm in diameter), amorph-

ous, spreading, and highly irregular. MYP remains to

date a popular choice for the isolation of B. cereus

and is available commercially. Plates can be stored for

up to 7 days at 4



0010KG agar In 1971, Kim and Goepfert developed the

Kim–Goepfert (KG) medium with a similar sensitivity

and selectivity to MYP. It contains no carbohydrates

and low levels of peptone to promote free spore for-

mation within 24 h of incubation at 37



C. Surface

colonies of B. cereus after overnight incubation at



C are flat, dry, and of a rough appearance, similar

to the appearance of the colonies on MYP agar. The

composition of KG agar allows direct confirmation

of lecithinase-producing organisms by microscopic

examination, differentiation of B. cereus from

B. thuringiensis based on production of parasporal

crystals of d-endotoxin by the latter species during

sporogenesis, and the use of fluorescent-labeled anti-

bodies to spore-coat antigens as a rapid and specific

serological test for confirmation of identity. Bacillus

spp. that may produce lecithinase, such as B. poly-

myxa and B. laterosporus, are unable to do so under

the nutritionally poor KG medium.

0011PEMBA medium Polymyxin–pyruvate–egg yolk–

mannitol–bromothymol blue agar (PEMBA), de-

veloped by Holbrook and Anderson in the early

1980s, also relies on the mannitol-negative, lecithi-

nase-positive properties of the organism. The medium

contains pyruvate to reduce the tendency of B. cereus

to form rhizoid colonies, improve the egg yolk reac-

tion, and enhance sporulation. After 18–24 h of incu-

bation at 37



C, B. cereus strains form flat, crenate to

rhizoid, turquoise–peacock blue colonies, 2–5mm in

diameter, with a ‘ground-glass’ surface appearance.

After a further 24 h of incubation at ambient tem-

perature, all colonies turn peacock blue. B. cereus is

differentiated microscopically from other mannitol-

negative species by microscopic examination of the

stained cells for the presence of lipid granules and

spores in the cytoplasm.

0012PEMPA medium Polymyxin–Pyruvate–egg yolk–

mannitol–bromocresol purple agar (PEMPA), a

modified agar, gives comparable results to PEMBA

after shorter incubation times (18–24 h).

tbl0001 Table 1 Characteristics of Bacillus cereus and other members

of the B. cereus group

Reaction B.

cereus

mycoides

thuringiensis

anthracis

Egg-yolk reaction þ

þþ (þ)

Acid from mannitol   

Catalase þþ þ þ

Gram reaction þþ þ þ

Motility +



+ 

Hemolysis

(sheep RBC)

þ (þ)(þ) 

Rhizoid growth þ  

Production of

toxin crystals

 þ 

Anaerobic

utilization

of glucose

þþ þ þ

Nitrate reduction þþ þ þ

VP reaction þþ þ þ

Tyrosine

utilization

þ (þ) þ (þ)

Resistance

to lysozyme

þþ þ þ

þ, 90–100% of strains are positive.

(þ), usually weakly positive.

+, 50–90% of strains are positive.

, <10% of strains are positive.

360

BACILLUS

/Detection

0013 Blood agar On 5% horse blood agar, after 24 h at

35–37



C, B. cereus colonies are 3–8 mm in diameter,

flat to low convex, gray–white to gray–green, with a

matt ‘ground glass’ surface. Growth is also associated

with a strong ‘mouse-like’ odor. Hemolytic reactions

on the medium vary from partial lysis of erythrocytes

to complete b-hemolysis. Small colonies, 2–3mm in

diameter, form during anaerobic incubation. The ir-

regular, b-hemolytic colonies resemble those of Clos-

tridium perfringens and require further identification.

0014 Of the plating media, MYP agar is the official

medium in the USA (Association of Official Analyt-

ical Chemists), Canada, France, and several other

countries. PEMBA is widely used in the UK. These

plating media do not differentiate between B. cereus

and B. thuringiensis organisms (Table 1). In general,

the numbers of B. cereus recovered from foods are

sufficiently high for direct plating, and preenrichment

is unnecessary. However, preenrichment may be re-

quired when low viable counts or damaged cells are

present. The most probable number (MPN) tech-

nique, using trypticase soy-polymyxin B broth and

incubation for 48 h at 30



C, is recommended for

estimation of B. cereus numbers in foods expected to

contain less than 10 cells per gram. Tubes showing

growth are streaked on to MYP agar. On egg-yolk-

containing agar plates, zones of egg-yolk precipita-

tion from individual colonies tend to coalesce if high

levels of the organisms are present (Figure 1), and

enumeration of colonies may become difficult. Fur-

ther dilutions of samples to three to 30 colonies are

therefore recommended in quantitative analyses.

Chromogenic Agar Media

0015 Chromogenic substrates that utilize specific enzym-

atic activities are increasingly used for rapid detection

and identification of microorganisms. Their incorpor-

ation into selective agar media facilitates identifica-

tion of colonies, improves the accuracy of the test and

reduces the need for isolation of pure cultures and

confirmation. The enzyme inositol-specific phospho-

lipase C (PI-PLC) was first reported in culture filtrates

of B. cereus in 1965, and a similar enzyme was later

identified in B. thuringiensis. On a chromogenic, se-

lective, and differential plating medium that detects

the enzyme phosphatidylinositol-specific phospholi-

pase C (PI-PLC) in B. cereus and B. thuringiensis

(BCM

Bacillus cereus and Bacillus thuringiensis,

Biosynth, Staad, Switzerland), these organisms form

large, 2–7-mm, turquoise, flat colonies with or with-

out turquoise halos on the medium. Other Bacillus

spp. form white, 1–2-mm colonies or are inhibited;

some Listeria strains form colonies less than 1 mm in

diameter, white or turquoise on the medium. Com-

parison with MYP agar has shown a significantly

higher inhibition of Gram-positive non-B. cereus/B.

thuringiensis organisms that interfere with enumer-

ation of small numbers of B. cereus in heavily con-

taminated samples. Typical colonies of B. cereus on

the chromogenic agar plate are shown in Figure 1.

The shelf stability of the chromogenic agar plates (3

months at 2–5



C) is superior to that of MYP plates.

As with other procedures, the chromogenic medium

does not differentiate between B. cereus and B. thur-

ingiensis organisms.

Confirmatory Procedures

0016The specific confirmation tests that distinguish

B. cereus from the other members of the B. cereus

group include motility, toxin crystal formation, rhi-

zoid growth, and hemolysin activity (Table 1).

fig0001 Figure 1 Typical colonies of B. ce r e u s on plates of MYP agar (left) and BCM

agar (right).

BACILLUS

/Detection 361

Motility is determined by stabbing the center of a

tube of semisolid medium and incubating for 18 h at



C. The motile organisms (most B. cereus and

B. thuringiensis) diffuse out from the stab, forming

an opaque growth pattern. The only established dif-

ference between B. cereus and B. thuringiensis strains

is the presence of genes encoding for the insecticidal

toxins, usually present on plasmids. The ability of

B. thuringiensis to form toxin crystals is determined

by inoculating nutrient agar slants with the

overnight culture and incubation for 2–3 days at

room temperature. Toxin crystals appear on a smear

of a microscope slide stained with 0.5% basic fuchsin

as dark-staining diamond-shaped objects after their

release from the sporangium upon lysis. The test is

inconclusive if the release is not observed. Rhizoid

growth, characteristic of B. mycoides, is observed by

the production of hair- or root-like structures project-

ing from an agar plate inoculated with the organism

and incubated for 48–72 h. Rhizoid growth may be

lost by the organisms. All three species show strong

hemolytic activity on trypticase soy–sheep blood agar

plates inoculated with an overnight culture and incu-

bated at 35



C for 24 h.

0017 A rapid confirmatory staining technique for the

identification of B. cereus developed by Holbrook

and Anderson uses colonies grown on PEMBA

medium and a combined spore and intracellular

lipid stain. Only the B. cereus group bacilli produce

the intracellular lipid globules.

0018 The US Food and Drug Administration confirms

presumptive B. cereus colonies from MYP agar plates

by anaerobic production of acid from glucose, reduc-

tion of nitrate to nitrite, production of acetylmethyl-

carbinol, decomposition of l-tyrosine, and growth in

the presence of 0.001% (wt/vol) lysozyme. Add-

itional tests for rhizoid growth, hemolysis of sheep

erythrocytes, motility, and production of parasporal

d-endotoxin crystals enable differentiation of

B. mycoides, B. thuringiensis,andB. anthracis. The

Central Public Health Laboratory, London, UK con-

firms presumptive B. cereus by testing motility, hem-

olysis, rhizoid growth, susceptibility to g-phage

(B. cereus-negative; B. anthracis-positive), and fer-

mentation of ammonium salt-based glucose, but not

of mannitol, arabinose, or xylose, after 5 days of

incubation at 36



Serological Tests

0019 The serological relationships within the genus Bacil-

lus are not as well defined as for other genera, and the

development of serological procedures for use as

diagnostic aids has met with limited success. Studies

with spore antigens have shown cross-reactions with

other Bacillus species. The multiplicity of spore sur-

face antigens in a single strain of B. cereus has also

been demonstrated. Some of the problems in the use

of spore antigens stem from contamination of their

preparations with vegetative cell debris and forma-

tion of antivegetative cell immunoglobulins, germin-

ation of spores during serum production, which also

results in the formation of antivegetative cell im-

munoglobulins, and in-vitro spore germination

during immunoassay.

0020Forty-two flagellar (H) and 13 somatic (O) anti-

gens have been identified. A serological typing

scheme for B. cereus strains based on flagellar anti-

gens was developed by Gilbert and coworkers and is

used routinely by the Food Hygiene Laboratory, Cen-

tral Public Health Laboratory, Colindale, London,

UK for investigations of food-poisoning outbreaks

and for serotyping strains from nongastrointestinal

infections. It is composed of 42 H antigens raised

against prototype strains of diverse origin. Twenty-

three of the 42 H serotypes are linked to illness, and

serotype H1 is dominant and contains both emetic

and diarrheal strains. A system similar to that used in

the UK was also developed by the Tokyo Metropol-

itan Research laboratory of Public Health in Japan.

0021B. thuringiensis strains have been classified on the

basis of their flagellar antigens. A serological typing

scheme for this species has been developed, and 14

serotypes have been recognized. There is evidence

that certain strains of B. cereus and B. thuringiensis

share common antigens.

Rapid Identification Methods

0022The term rapid methods refers to a variety of tests,

which can shorten the analysis time and reduce labor

(i.e., by automation). Some eliminate the need for

confirmation, whereas others require enrichment

prior to testing, which limits the assay speed but

provides more reliable identification of the target

organism. All require pure culture isolates of the bac-

teria. Rapid methods applicable to B. cereus are dis-

cussed below.

0023Miniaturized biochemical kits to identify Bacillus

species have been developed to produce standardized

test media that improve reproducibility, have a long

shelf-life, are rapid and convenient to use, economize

the use of media, and are inexpensive. The API 20E

and 50 CHB strips (bioMe

rieux, Marcy l’Etoile,

France; Hazelwood, MO, USA), are ready-to-use

microtube systems that contain dehydrated media

for standard biochemical tests. The API 50 CHB for

the identification of the genus Bacillus tests the fer-

mentation of 49 carbohydrates, and the phenotypic

profile obtained from these tests is used to identify the

362

BACILLUS

/Detection

strain. The API 20E strips are used for the identifica-

tion in the Vitek automated identification system,

which eliminates the need for manual identification.

Incubation for 24–48hat30or37



C (mesophiles) is

required before reading the results. Other manual and

automated systems for identification of Bacillus

based on biochemical properties are available. Micro-

Log

microbial carbon-source utilization (Biolog

Inc., Hayward, CA) generates an identity profile

based on the ability of the organism to oxidize

different carbon sources based on changes in the

redox potential following respiration of the viable

bacterial cells. Oxidation is determined colorimetri-

cally with a tetrazolium-based compound, and the

color change data measured automatically by a spec-

trophotometer are fed into a computer. Bacteria are

identified by comparison with a reference data base.

The BBL Crystal

Gram-positive Identification

System (Beckton Dickinson, Sparks, MD) is another

example of miniaturized identification method for

Gram-positive aerobes employing modified conven-

tional, fluorogenic, and chromogenic substrates.

0024 Another approach for the identification of the or-

ganism is based on its fatty acid profile. Analysis of

the cellular fatty acid composition is carried out by

gas chromatography following extraction of pure cul-

tures and injection into a gas chromatograph (Micro-

bial Identification System, MIS, Microbial-ID,

Newark, DE). The cellular fatty acid profile is ana-

lyzed by computer and compared with database

values to provide the best-match identification of

the organism.

0025 Other diagnostic schemes include polyacrylamide

gel electrophoresis analysis, pyrolysis mass spectrom-

etry, and Fourier-transform infrared spectroscopy.

Classification and identification based on differences

in rRNA are available commercially for food patho-

gens including B. cereus. Ribotyping (RiboPrinter

Microbial Characterization System, DuPont Quali-

con, Wilmington, DE) generates precise fingerprints

that can be used to classify isolates at the species level

and trace them to the contaminated source. Steps of

the identification, from cell lysis to data capture and

database comparisons, are automated. All of the

above identification methods require large updated

databases. DNA sequence data for the design of spe-

cific primers for B. cereus are being investigated. The

design of primers within the coupled sequences for

phospholipase C and sphingomyelinase of B. cereus

has been reported, and PCR tests with these primers

are positive only with isolates of the B. cereus group.

Moreover, positive results have also been observed

with an egg-yolk-negative B. cereus isolate, suggest-

ing that the primers could be specific for atypical

B. cereus.

0026Phage typing of B. cereus strains is another epi-

demiological tool in outbreaks of B. cereus food

poisoning. Phages are used in routine test dilution,

and drops are deposited on bacterial lawns on nutri-

ent agar plates that are incubated overnight at 32



and examined for lysis. A typing scheme using a set

of 12 phages selected on the basis of host range,

stability, and reproducibility of reaction showed that

most isolates of the pathogen were typeable. A good

correlation (80–100%) was observed between phage

types of strains isolated from specimens of suspected

foods and from symptomatic patients. Most B. thur-

ingiensis strains were also typeable by the same set of

phages.

Detection of

B. cereus

Toxins

0027In-vivo and in-vitro methods for the detection of

B. cereus diarrheal and emetic toxins are summarized

in Table 2. In-vivo methods for the detection of

B. cereus diarrheal enterotoxins include the rabbit

or guinea-pig ileal loop test, dermonecrotic test in

guinea-pig, mouse lethality test, vascular permeabil-

ity test, and Rhesus monkey feeding trials. In the

ileal loop assay, the lower portion of the intestine of

the animal is surgically exposed to allow ligation,

usually of the ileal regions, thereby testing different

samples in a single animal. The sample is injected

into a ligated section, and fluid accumulation is

monitored with time and compared with control

sections. Animal feeding trials (Rhesus monkey and

Suncus marinus) and lethality tests (mouse and

Suncus marinus) are used to detect the presence of

the emetic toxin. Animals (a set of six) are observed

for about 5 h after samples have been introduced

tbl0002Table 2 Detection methods of B. c ereus toxins

In-vivo methods In-vitro methods

Diarrheal syndrome

Rabbit or guinea-pig

ligated ileal loop

Gel-diffusion assays

Guinea-pig

dermonecrotic test

Cell cytotoxicity

(hemolysin BL)

Mouse lethality test

BCET-RPLA kit (L

of BL)

Vascular permeability test

BDE-VIA kit

(nonhemolytic toxin)

Rhesus monkey feeding trials

Emetic syndrome

Feeding trials

(Rhesus monkey and

Suncus marinus)

Cytotoxicity (proliferation of

HEp-2 cell line)

Lethality tests (mouse and

Suncus marinus)

See text for details.

BACILLUS

/Detection 363

using a stomach tube. An emetic response in two

of the six animals is considered positive for the

toxin.

0028 The disadvantages of bioassays are the time to

complete the assay, the need for special animal facil-

ities, the cost, as well as the concern about the use

of experimental animals in research. Alternatives to

the in-vivo tests are in-vitro assay methods that

include diffusion techniques, enzyme-linked immu-

nosorbent assays (ELISA), and cell cytotoxicity

tests. The gel diffusion assay detects hemolysin BL,

producing a discontinuous hemolysis pattern on

blood agar plates. The microslide immunodiffusion

assay and the fluorescent immunodot assay detect

enterotoxin by line of identity and fluorescence,

respectively.

0029 Two immunoassay kits are available commercially

for the detection of the B. cereus diarrheal toxins.

These include the B. cereus enterotoxin-reversed pas-

sive latex agglutination (BCET-RPLA) assay (Oxoid,

Unipath, Basingstoke, UK), and the Bacillus diarrheal

enterotoxin visual immunoassay (BDE-VIA) (TECRA

Diagnostics, Batley, UK). The RPLA procedure uses

latex particles to amplify the antibody: antigen reac-

tion. The antibody detects the L

component of

hemolysin BL, and can provide a semiquantitative

measure of the enterotoxin in foods. In the TECRA

immunoassay, a sandwich ELISA is used, in which

the antibody is absorbed on the solid phase, and the

enterotoxin is added. The colorimetric reaction

detects the 45-kDa and 105-kDa proteins of the non-

hemolytic enterotoxin and other proteins. The pres-

ence of preformed toxin is detected in 4 h, and

production of the toxin in samples containing

enterotoxigenic Bacillus spp. is detected after over-

night enrichment. Since these two kits detect compon-

ents in the different enterotoxins, isolates should

be tested by both methods. Many B. cereus strains

have been shown to react with both the Oxoid and

the TECRA detection kits, suggesting that they are

able to produce both enterotoxin complexes.

0030 B. thuringiensis, B. subtilis, B. licheniformis,and

B. pumilus strains have been involved in foodborne

illnesses. Although the nature of their toxins is not

well defined, reaction with antibodies to B. cereus

enterotoxins in both the Oxoid and the TECRA de-

tection kits has been demonstrated for some of them.

0031 Cell cytotoxicity techniques have been used for

both the diarrheal toxin and the emetic toxin. Cell

lines used include HeLa, HEp-2, Vero, and others,

and cellular responses in the presence of the toxin

range from morphological changes (subjective), to

more specific, e.g., metabolic status of the cells and

detection of lactate dehydrogenase release from dam-

aged cells.

Taxonomic Relationship of

B. cereus

B. thuringiensis

, and

B. anthracis

0032The lower end of the GC range of the genus Bacillus

(32–38%) is occupied by B. cereus and the closely

related species B. anthracis, B. mycoides, and B. thur-

ingiensis. It was suggested in 1952 that there are no

consistent phenotypic properties that differentiate

these species and that these three species be desig-

nated varieties of B. cereus. However, this has not

been accepted for B. anthracis and B. thuringiensis

because of their pathogenic qualities. The high hom-

ology between DNA from B. anthracis, B. thuringien-

sis, and B. cereus suggests that organisms in these

taxa should have the same name. Recently, research-

ers used multilocus enzyme electrophoresis and se-

quence analysis of nine chromosomal genes to

provide further evidence that B. anthracis should

be considered a lineage of B. cereus. Evidence of

the close taxonomic relationship of B. cereus, B. thur-

ingiensis,andB. anthracis was indirectly obtained

from serological studies that showed extensive

cross-agglutination between the spores of the three

species.

0033The only established difference between B. cereus

and B. thuringiensis strains is the presence of genes

encoding for the insecticidal toxins, usually present

on plasmids. None of the detection methods discussed

above distinguishes between the two species.

See also: Bacillus: Occurrence; Food Poisoning