enterotoxigenic bacteroides fragilis: a rogue among symbiotes · in early studies, oral inoculation...

CLINICAL MICROBIOLOGY REVIEWS, Apr. 2009, p. 349–369 Vol. 22, No. 20893-8512/09/$08.00�0 doi:10.1128/CMR.00053-08Copyright © 2009, American Society for Microbiology. All Rights Reserved.

Enterotoxigenic Bacteroides fragilis: a Rogue among SymbiotesCynthia L. Sears*

Divisions of Infectious Diseases and Gastroenterology, Department of Medicine, Johns Hopkins University School of Medicine,Baltimore, Maryland

INTRODUCTION .......................................................................................................................................................349ETBF INFECTIONS IN ANIMALS .........................................................................................................................349

Naturally Occurring ETBF Infections .................................................................................................................349Experimental ETBF Infections .............................................................................................................................349

ETBF ASSOCIATION WITH HUMAN CLINICAL DISEASE ............................................................................350Epidemiology and Clinical Presentation of ETBF-Associated Diarrheal Disease.........................................350Other ETBF Disease Associations........................................................................................................................351Diagnosis of ETBF Infection.................................................................................................................................354Therapy of ETBF-Associated Diarrheal Disease ................................................................................................356

BFT GENETICS AND PROTEIN STRUCTURE...................................................................................................356GENETICS OF ETBF ................................................................................................................................................359BFT MECHANISM OF ACTION.............................................................................................................................362

In Vitro Studies of Cell Lines, Polarized Epithelial Monolayers, and Human Colon..................................362Molecular Mechanism of Action of BFT .............................................................................................................363

PROPOSED MODEL FOR PATHOGENESIS OF ETBF DISEASE..................................................................364SUMMARY AND FUTURE CHALLENGES ..........................................................................................................366ACKNOWLEDGMENTS ...........................................................................................................................................366REFERENCES ............................................................................................................................................................366

INTRODUCTION

Bacteroides species comprise nearly half of the fecal floracommunity and are host symbionts critical to host nutrition(e.g., Bacteroides thetaiotaomicron) and mucosal and systemicimmunity (e.g., B. fragilis) (25, 35, 54, 55, 73, 98, 110, 127).Among Bacteroides species, B. fragilis strains are opportunisticpathogens, being the leading anaerobic isolates in clinical spec-imens, bloodstream infections, and abdominal abscesses de-spite comprising typically �1 to 2% of the cultured fecal flora(50, 62, 75, 87, 91). In 1984, while investigating the etiology oflamb diarrheal disease, Myers and colleagues provided the firstevidence that certain strains of B. fragilis were epidemiologi-cally associated with diarrheal disease (64). Studies by thesame investigators revealed that both the isolates and theirsterile culture supernatants stimulated intestinal secretion inlamb ligated intestinal loops (64, 69; see reference 105 for areview). The secretory responses in some ligated intestinalloops were so potent that the loops burst, a response reminis-cent of cholera toxin-stimulated secretory responses. The bio-logically active factor was proposed to be a heat-labile, �20-kDa protein toxin, now known to be one of a family of B.fragilis toxins (BFTs) (69, 106). B. fragilis strains eliciting intes-tinal secretion were named enterotoxigenic B. fragilis (ETBF)and their nonsecretory counterparts were termed nontoxigenicB. fragilis (NTBF). This review describes the progress over thesubsequent nearly 25 years in defining the role of ETBF in

human disease, the genetics and mechanism of action of BFT,and insights into the molecular evolution of ETBF strains.

ETBF INFECTIONS IN ANIMALS

Table 1 summarizes the data on ETBF infections in animals.

Naturally Occurring ETBF Infections

ETBF strains were initially identified during an investigationof newborn lamb diarrheic disease (64). Subsequent reportsindicated that calves, foals, and piglets were susceptible toETBF-associated diarrheal illnesses in the field (17, 65, 66, 69,105). In the small number of reports available, the diarrhealillnesses were largely self-limited, with little mortality, exceptpossibly in newborn lambs (64). The limitations of these re-ports include variable assessment for additional enteric patho-gens and a lack of epidemiology to assess ETBF carriage inasymptomatic livestock. Consistent with the native colonichabitat of B. fragilis, exfoliating colitis with an intense neutro-philic mucosal infiltrate was observed in a single piglet exam-ined by histopathology (17).

Experimental ETBF Infections

In early studies, oral inoculation of ETBF into newborn lambs,piglets, or importantly, gnotobiotic (germfree) piglets reproducedthe diarrheal illnesses observed in the field, providing furthersupport for the proposal that select B. fragilis strains stimulatedintestinal secretion and diarrheal disease (23, 64, 65). Histopa-thology from gnotobiotic piglets revealed that lesions were mostsevere in the colon, where crypt hyperplasia and neutrophilicinfiltrates were observed. By scanning electron microscopy, the

* Mailing address: Divisions of Infectious Diseases and Gastroen-terology, Department of Medicine, 1550 Orleans Street, CRB2 Build-ing, Suite 1M.05, Baltimore, MD 21231. Phone: (410) 614-0141. Fax:(410) 955-0740. E-mail: [email protected].

349

on Decem

ber 3, 2020 by guesthttp://cm

r.asm.org/

Dow

nloaded from

https://crossmark.crossref.org/dialog/?doi=10.1128/CMR.00053-08&domain=pdf&date_stamp=2009-04-01

http://cmr.asm.org/

colonic surface epithelium had a cobblestone appearance associ-ated with round, swollen epithelial cells and epithelial cell exfo-liation (23). Similar but more variable lesions were also observedprimarily in the distal half of the small intestine. No extraintesti-nal lesions were noted. Additional studies with infant and 2-week-old rabbits as well as adult rabbits with ligated ceca confirmed theenteropathogenicity of ETBF, but in these disease models bloodydiarrhea and mortality were frequently observed (17, 65–68).However, ETBF virulence was variable in rabbit models (66, 67),consistent with subsequent observations that sterile culture super-natants of ETBF exhibited variable biologic activity on HT29/C1cells (a human colonic epithelial tumor cell line) (13, 30, 82, 116,119). Histopathologic abnormalities in these non-gnotobiotic an-imal models also occurred only in the distal ileum and colon, withdisruption of the epithelial integrity and predominant neutro-philic or mixed neutrophilic and mononuclear cellular infiltratesin the lamina propria; animals colonized with NTBF strains ex-hibited normal colonic histopathology without inflammation bylight microscopy. ETBF adherence and/or invasion of colonocyteswas not observed by light or electron microscopy. Bacteremia hasnot been reported for these animal models (68).

Early studies indicated that mice (suckling and young adult)and hamsters do not exhibit secretory responses to ETBF (68,69). Recently, colonization of gnotobiotic mice with ETBF, butnot NTBF, has been shown to induce acute, sometimes lethal,colitis (71, 92). In contrast, conventional mice colonized withETBF develop rapid-onset, transient diarrhea lasting 3 to 4days. Subsequently, conventional mice colonized with ETBFexhibit persistent, asymptomatic colonization, with ongoinghistopathologic colitis present for as long as 16 months (90, 92)(Fig. 1). Additional studies show that purified BFT, albeit at apharmacologic dose (i.e., 10 �g), stimulates secretion and his-tologic enteritis in mouse ileal loops (42, 44).

In initial studies, ETBF or sterile culture supernatants wereinjected into ligated intestinal (predominantly jejunal) seg-ments of lambs, calves, or rabbits (64, 69). These studies con-firmed that ETBF stimulated intestinal secretion but, moreimportantly, provided the first evidence that ETBF secreted aheat-labile protein toxin responsible for stimulating intestinalsecretion (69). Subsequently, Obiso and colleagues demon-strated that purified BFT (1-�g to 50-�g doses) stimulatesdose-dependent secretion of sodium, chloride, and albumin inboth ileal and colonic ligated loops in rats, rabbits, and lambs(80). Colonic secretory responses exceeded ileal responses inall species. However, species-dependent BFT potency was ob-served, with the greatest fluid responses in the ilea of lambsand the colons of rabbits. The histology of ileal or colonicligated loops inoculated with 20 �g of purified BFT for eachspecies exhibited marked epithelial cell rounding and detach-ment, neutrophilic inflammation, and in some sections, necro-sis and hemorrhage. Time course analyses revealed that histo-logic changes (at 10 h) preceded detection of intestinalsecretion (examined at 18 h). Consistent with the subsequentdetermination that BFT is a zinc-dependent metalloproteasetoxin (see below), metal-chelating agents reduced (�90%) thesecretory and histologic responses to purified BFT in ligatedintestinal loops and zinc reconstituted them (�50%).

ETBF ASSOCIATION WITH HUMANCLINICAL DISEASE

Epidemiology and Clinical Presentation ofETBF-Associated Diarrheal Disease

ETBF strains were first isolated from humans with diarrhealillnesses in an uncontrolled study in 1987 (70) examining 10individuals with diarrhea of unknown origin (from Montana)and 34 infants (2 to 14 months of age; Navajo Area IndianHealth Service, Tuba City, AZ). Overall, ETBF isolates wereidentified in 8 of 44 patients (two adults and six children of �5years of age), with only one infant positive for a second entericorganism, enterotoxigenic Escherichia coli (ETEC). In this ini-

FIG. 1. ETBF induce murine colitis. Colonization of 4-week-oldC57BL/6 mice with ETBF for 2 weeks induces inflammation and hy-perplasia throughout the colon. Panels 1 to 3, mice inoculated withphosphate-buffered saline; panels 4 to 6, mice inoculated with ETBFstrain 86-5443-2-2. In panels 4 and 5, the rounded and detachingcolonic epithelial cells in ETBF-infected mice can be seen. (Repro-duced from reference 90 with permission of the publisher. CopyrightJohn Wiley and Sons Ltd.)

TABLE 1. Animal species susceptible to ETBF infection or BFTbiologic activity

Type of infection or expt and susceptible animal species

Naturally occurring ETBF infectionsa

Newborn lambsb

CalvesFoalsPiglets

Experimental ETBF infections or BFT inoculationsLambs (oral inoculationc and intestinal

loop inoculationd)Calves (intestinal loop inoculationd)Rabbits (oral inoculation, RITARD model, and intestinal

loop inoculationd,e)Rats (intestinal loop inoculationd)Gnotobiotic piglets (oral inoculationf)Mice (oral inoculation and ileal loop inoculationg)

a Presenting with diarrhea.b One diarrheic ewe reported (64).c All oral inoculations are with ETBF strains.d Jejunal, ileal, and/or colonic loops inoculated with ETBF, sterile ETBF

culture supernatants, or purified BFT (64, 69, 80).e Three-day-old (oral inoculation), 2-week-old (direct small bowel inocula-

tion), and adult (oral inoculation with ligated ceca) animals. RITARD, reversibleileal tie adult rabbit diarrhea (17, 65–68, 80).

f One to 2 days of age (23).g Conventional or gnotobiotic mice of more than 3 weeks of age (71, 90, 92).

Inoculation of ileal loops with purified BFT stimulated secretion (42, 44).

350 SEARS CLIN. MICROBIOL. REV.

on Decem


r.asm.org/

Dow

nloaded from

http://cmr.asm.org/

tial study, most patients had self-limited diarrhea, but inter-mittent diarrhea of more than 3 years duration, diarrhea per-sisting 4 weeks, high fever, and frank/occult fecal blood werenoted in ETBF-positive patients (70). The first human studywith control subjects to investigate a putative association be-tween ETBF infection and human diarrheal disease was con-ducted in the pediatric outpatient clinics of the Apache Indianreservation at Whiteriver, AZ, in 1992 (95). The key results ofthis study were that children of �1 year of age did not developETBF-associated diarrhea, whereas ETBF isolates were asso-ciated with diarrheal illnesses in children between the ages of1 and 5 years. ETBF isolates from single families evaluated byrestriction enzyme analysis appeared to be genetically related.

Table 2 summarizes studies with control subjects evaluatingthe role of ETBF as an etiologic agent of diarrheal disease.Table 3 summarizes uncontrolled case series or reports ofhuman ETBF infection. The majority of controlled studiesconducted to date have evaluated ETBF as an agent of acutecommunity-acquired diarrheal disease in children (Table 2).One Swedish study assessed ETBF in adults hospitalized withacute diarrheal disease (128) compared to isolation of ETBFfrom healthy outpatient control adults. Overall, study designshave been variable, and many studies are hampered by smallnumbers of patients and controls, a lack of thorough studies forother enteric pathogens, and a wide range of recovery rates (6to 70%) for B. fragilis.

Of 17 studies with controls (Table 2), 5 did not show anassociation between ETBF and diarrheal disease in differingpatient populations (4, 8, 48, 77, 84). Of these studies, a fol-low-up study of the same patient population by the same au-thors subsequently did show an association of ETBF with di-arrheal disease (59a, 84), one study had a B. fragilis isolationrate of only 6% (77), and three studies were from Brazil (4, 8,48). Excluding the one study (128) focused on adults, all butone (10) of the remaining studies confirmed the original ob-servations of Sack et al. (95) suggesting that, similar to the casewith Clostridium difficile, there is no association between ETBFisolation and diarrheal illnesses in children of �1 year old,whereas diarrheal disease due to ETBF emerges after age 1.The mechanisms accounting for the apparent lack of pathoge-nicity of ETBF in very young children are unknown. However,there is marked variability in the frequencies at which ETBFstrains are associated with diarrhea in different geographiclocales (from 3.5% in Bangladesh to 28% in Italy). Only onelarge study to date has investigated ETBF in adults (defined asindividuals who are �15 years old), identifying ETBF in 27%of 728 patients with diarrhea and 12% of 194 healthy controls(P � 0.01) (128). In this study, 19% of adults of �30 years ofage, 10.6% of adults between the ages of 30 and 60, and 3.7%of adults older than 60 years were asymptomatically colonizedwith ETBF. Among adults between 15 and 30 years of age,ETBF fecal carriage rates were similar for those with andwithout diarrhea, whereas an association of ETBF with diar-rheal disease emerged in older patients. Other studies alsosuggest that asymptomatic colonization with ETBF is common(up to �30%) (7) (Table 2; also see Table 5). Similarly, 9.3%of 237 B. fragilis strains cultured from sewage influent in Boze-man, MT, were ETBF, suggesting moderately high endemiccarriage of ETBF (109). ETBF carriage for up to 16 monthswith fecal ETBF isolates identical by restriction enzyme anal-

ysis has been reported (95). Together, the data suggest thatETBF strains are globally distributed enteric pathogens caus-ing diarrhea in both children and adults. Similar to the case forother enteric pathogens, asymptomatic ETBF colonization iscommon. Geographic diversity in the recovery of entericpathogens is also well known (10a), although the reasons un-derlying the variable isolation rates have not been elucidated.With respect to detection of ETBF in stool, one importantvariable is likely the sensitivity of the diagnostic approachesused (see “Diagnosis of ETBF Infection”).

ETBF clinical illnesses are typically characterized as self-lim-ited with watery diarrhea; if noted, persistent diarrhea (�14 days)has been reported for a minority (0 to 22%) of patients. Table 4summarizes the range of clinical findings reported in studies ofETBF diarrheal illnesses. When assessed, ETBF diarrheal ill-nesses have been clinically indistinguishable from non-ETBF di-arrheal illnesses in the populations studied (10, 95, 128). A recentprospective analysis involving 73 ETBF-infected patients (43 chil-dren under age 15 and 30 adults) in Bangladesh (102) correlatedclinical findings with more detailed stool analyses to evaluate thepathogenesis of ETBF diarrheal illnesses. In this series, althoughillnesses were again self-limited (2 to 11 days), substantial abdom-inal pain (88%), tenesmus (66%), and nocturnal diarrhea (80%)were reported. In contrast, fever (�37.5°C) (7%) and fecal occultblood (8.5%) were infrequent. Leukocytosis was not detected.Stool analyses for leukocytes, fecal lactoferrin, and proinflamma-tory cytokines (tumor necrosis factor alpha and interleukin-8 [IL-8]) indicated that ETBF induced inflammatory diarrhea com-pared to stools from asymptomatic control individuals notinfected with ETBF. Consistent with the inflammatory mucosalresponse to ETBF infection, serum immunoglobulin A (IgA) andIgG antibodies as well as fecal IgA to BFT were detected. In onereport, two to four distinct, sequential episodes of ETBF-associ-ated diarrhea were reported for children in Bangladesh, suggest-ing that acquired immunity to ETBF is incomplete (85).

No studies have been designed to specifically determine ifETBF contributes to persistent or chronic diarrhea or to diar-rheal illnesses in travelers; limited observations are availableon the role of ETBF in diarrhea defined as nosocomial (oc-curring �3 days after hospital admission) (16, 57) or antibioticassociated (39, 86). In one study, ETBF strains were identifiedin stools of 9% of patients with non-C. difficile nosocomialdiarrhea, which was significantly greater than ETBF detectionin outpatient controls (2%; P � 0.04), but hospitalized controlswere not evaluated (16). C. difficile and ETBF were detectedsimultaneously in a small subset of fecal samples from patientswith antibiotic-associated diarrhea (57, 86). In one study ofchildren, there was no association of ETBF with antibiotic-associated diarrhea (39). Well-known extraintestinal or intes-tinal complications of enteric infections, such as reactive ar-thritis or diarrhea-predominant irritable bowel syndrome, havenot yet been reported for ETBF.

Other ETBF Disease Associations

Three recent but small reports raise the specter of adverse,long-term consequences of chronic ETBF colonization (Table5). In two reports examining patients with irritable bowel dis-ease (IBD) (7, 88), active (but not clinically quiescent) IBDwas associated with ETBF infection, although the differences

VOL. 22, 2009 ENTEROTOXIGENIC BACTEROIDES FRAGILIS 351

on Decem


r.asm.org/

Dow

nloaded from

http://cmr.asm.org/

TA

BL

E2.

Stud

ies

ofE

TB

Fin

diar

rhea

ldis

ease

a

Ref

eren

ce(a

utho

r,yr

�ref

eren

ceno

.�)St

udy

desi

gnL

ocat

ion

nE

TB

Fdi

agno

stic

met

hod

Mai

nre

sults

Stre

ngth

s/lim

itatio

nsof

stud

y

Sack

etal

.,19

92(9

5)O

utpa

tient

Apa

che

Indi

anch

ildre

nw

ithan

dw

ithou

tdi

arrh

ea,

1986

–198

8

Whi

teri

ver,

AZ

275

patie

nts

with

diar

rhea

,18

5ag

e-an

dtim

e-m

atch

edco

ntro

ls

Stoo

lcul

ture

,lam

bile

allo

opas

say

Ove

rall,

12%

ofpa

tient

sw

ithdi

arrh

eaan

d6%

ofco

ntro

lsw

ere

ET

BF

posi

tive

(P�

0.03

);E

TB

Fis

olat

ion

rate

sw

ere

high

est

at13

to36

mo

(20

to24

%fo

rch

ildre

nw

ithdi

arrh

eavs

8to

9%fo

rco

ntro

ls);

noas

soci

atio

nbe

twee

nE

TB

Fan

ddi

arrh

eain

child

ren

unde

r1

year

old

B.f

ragi

lisis

olat

ion

rate

of�

70%

;lim

ited

anal

ysis

for

othe

ren

teri

cpa

thog

ens;

inco

mpl

ete

mat

chin

gof

patie

nts

and

cont

rols

Sack

etal

.,19

94(9

4)In

patie

ntB

angl

ades

hich

ildre

nw

ithan

dw

ithou

tac

ute

diar

rhea

,19

91–1

992

Dha

ka,B

angl

ades

h35

8pa

tient

sw

ithdi

arrh

ea,

425

cont

rols

;all

wer

e�

5ye

ars

old

Stoo

lcul

ture

,HT

29/C

1ce

llas

say,

RIT

AR

Dm

odel

Ove

rall,

6.1%

ofpa

tient

sw

ithdi

arrh

eaan

d1.

2%of

cont

rols

wer

eE

TB

Fpo

sitiv

e(P

�0.

001)

;no

asso

ciat

ion

betw

een

ET

BF

and

diar

rhea

inch

ildre

nun

der

1ye

arol

d;fo

rch

ildre

nol

der

than

1ye

ar,

9%w

ithdi

arrh

eaan

d0.

8%of

cont

rols

wer

eE

TB

Fpo

sitiv

e(P

�0.

001)

(diff

eren

cew

asal

sosi

gnifi

cant

whe

nch

ildre

nw

ithm

ultip

lein

fect

ions

wer

eex

clud

ed)

Ext

ensi

vest

udie

sof

ente

ric

path

ogen

s;an

tibio

ticad

min

istr

atio

npr

ior

toad

mis

sion

for

diar

rhea

grou

pw

asa

conf

ound

er;l

owis

olat

ion

rate

sof

B.f

ragi

lis(2

0%)

San

Joaq

uin

etal

.,19

95(1

01)

Urb

anch

ildre

nw

ithan

dw

ithou

tdi

arrh

ea,

1990

–199

1

Okl

ahom

aC

ity,O

K97

8pa

tient

sw

ithdi

arrh

ea,

578

age-

mat

ched

cont

rols

Stoo

lcul

ture

,HT

29/C

1ce

llas

say

B.f

ragi

lisis

olat

edfr

om32

%an

d21

%of

child

ren

with

and

with

out

diar

rhea

,res

pect

ivel

y(P

�0.

001)

;4.

8%of

child

ren

of1

to10

year

sof

age

with

diar

rhea

and

1.5%

ofco

ntro

lsw

ere

ET

BF

posi

tive

(P�

0.02

);no

asso

ciat

ion

betw

een

ET

BF

and

diar

rhea

inch

ildre

nun

der

1ye

arol

d

Lim

ited

anal

ysis

for

othe

ren

teri

cpa

thog

ens;

low

isol

atio

nra

tes

ofB

.fra

gilis

(�35

%)

Pant

osti

etal

.,19

97(8

4)R

ando

mly

sele

cted

diar

rhea

lsto

ols

subm

itted

toa

mic

robi

olog

yla

bfr

omin

patie

nts

and

stoo

lsfr

omhe

alth

yco

ntro

ls,

1994

–199

5

Ital

y23

child

ren

and

106

adul

tsw

ithdi

arrh

ea;2

6he

alth

ych

ildre

nan

d39

heal

thy

adul

ts

Stoo

lcul

ture

,HT

29/C

1ce

llas

say

Sim

ilar

ET

BF

isol

atio

nra

tes

for

heal

thy

child

ren

and

adul

tsan

dch

ildre

nan

dad

ults

with

diar

rhea

(10

to17

%)

Smal

lstu

dy;m

ixed

adul

tpo

pula

tion

with

acut

e,no

soco

mia

l,an

d/or

chro

nic

diar

rhea

;B.f

ragi

lisis

olat

ion

rate

sof

�40

to50

%

Niy

ogie

tal

.,19

97(7

7)In

patie

ntch

ildre

nw

ithan

dw

ithou

tac

ute

diar

rhea

,199

5–19

96

Kol

kata

,Ind

ia22

6pa

tient

sw

ithdi

arrh

ea,

172

cont

rols

Stoo

lcul

ture

,HT

29/C

1ce

llas

say

Ove

rall,

2.6%

ofpa

tient

sw

ithdi

arrh

eaan

d1.

7%of

cont

rols

wer

eE

TB

Fpo

sitiv

e(d

iffer

ence

not

sign

ifica

nt);

noag

eas

soci

atio

nof

ET

BF

with

diar

rhea

was

iden

tified

Lim

ited

anal

ysis

for

othe

ren

teri

cpa

thog

ens;

B.f

ragi

lisw

asis

olat

edfr

omon

ly24

(6%

)ch

ildre

n,bu

t9

(38%

)of

the

isol

ates

wer

eE

TB

FM

enoz

ziet

al.,

1998

(59a

)In

patie

ntch

ildre

nw

ithdi

arrh

eaan

dag

e-m

atch

edco

ntro

lsw

ithou

tdi

arrh

ea(1

996–

1998

)

Parm

a,It

aly

227

patie

nts

with

diar

rhea

,23

7co

ntro

lsSt

oolc

ultu

re,H

T29

/C1

cell

assa

yB

.fra

gilis

was

isol

ated

from

27.7

%of

patie

nts

and

19%

ofco

ntro

ls(P

�0.

05);

ET

BF

was

asso

ciat

edw

ithdi

arrh

eafo

rch

ildre

nof

1to

5ye

ars

ofag

e(3

3.3%

vs16

.4%

�pat

ient

svs

cont

rols

�;P

�0.

01);

noco

rrel

atio

nof

ET

BF

with

antib

iotic

ther

apy

was

foun

d

Low

B.f

ragi

lisis

olat

ion

rate

s;lim

ited

clin

ical

deta

ils

Kat

oet

al.,

1999

(39)

Urb

anho

spita

l-bas

edst

udy

ofch

ildre

nne

gativ

efo

rco

mm

onba

cter

ial

ente

ropa

thog

ens,

1996

Nag

oya,

Japa

n41

6ch

ildre

n(1

37w

ithno

n-an

tibio

tic-a

ssoc

iate

ddi

arrh

ea,1

66w

ithan

tibio

tic-a

ssoc

iate

ddi

arrh

ea,1

13co

ntro

lsw

ithou

tdi

arrh

ea)

Stoo

lcul

ture

,PC

Rfo

rB

.fra

gilis

bft

onba

cter

iali

sola

tes,

HT

29/C

1ce

llas

say

Ove

rall,

ET

BF

was

asso

ciat

edw

ithno

n-an

tibio

tic-a

ssoc

iate

ddi

arrh

eain

child

ren

of1

to14

year

sof

age

(14.

9%vs

4.9%

inco

ntro

ls;P

�0.

05)

but

not

with

antib

iotic

-as

soci

ated

diar

rhea

(ass

esse

don

lyin

child

ren

upto

6ye

ars

ofag

e);n

oas

soci

atio

nbe

twee

nE

TB

Fan

ddi

arrh

eain

child

ren

unde

r1

year

old

Stoo

lcul

ture

spe

rfor

med

onst

ools

froz

enfo

r�

1m

o;B

.fr

agili

sis

olat

edfr

om23

to54

%of

child

ren

of�

1ye

arof

age


on Decem


r.asm.org/

Dow

nloaded from

http://cmr.asm.org/

Alb

ert

etal

.,19

99(2

)H

ospi

tal-b

ased

stud

yof

child

ren

with

acut

edi

arrh

ea;c

omm

unity

child

ren

with

out

diar

rhea

wer

eus

edas

cont

rols

,199

3–19

94

Dha

ka,B

angl

ades

h81

4pa

tient

sw

ithdi

arrh

ea,

814

cont

rols

with

out

diar

rhea

Stoo

lcul

ture

,HT

29/C

1ce

llas

say

Ove

rall,

3.5%

ofpa

tient

sw

ithdi

arrh

eaan

d1.

5%of

cont

rols

wer

eE

TB

Fpo

sitiv

e(P

�0.

01);

ET

BF

was

not

asso

ciat

edw

ithdi

arrh

eain

child

ren

unde

r1

year

old

Ext

ensi

vest

udie

sof

ente

ric

path

ogen

s

Zha

nget

al.,

1999

(128

)A

dult

inpa

tient

s(�

15ye

ars

old)

with

diar

rhea

and

outp

atie

nthe

alth

yco

ntro

ls,1

996–

1997

Hud

ding

e,Sw

eden

(Sto

ckho

lmsu

burb

)72

8pa

tient

sw

ithdi

arrh

ea,

194

cont

rols

IMS-

PCR

,HT

29/C

1ce

llas

say

Ove

rall,

26.8

%of

patie

nts

with

diar

rhea

and

12.4

%of

cont

rols

wer

eE

TB

Fpo

sitiv

e(P

�0.

01);

ET

BF

was

asso

ciat

edw

ithdi

arrh

eain

adul

tsof

�30

year

sof

age

but

not

inth

ose

of15

to30

year

sof

age

Ext

ensi

vest

udie

sof

ente

ric

path

ogen

s;st

ools

wer

efr

ozen

for

anun

spec

ified

time

prio

rto

anal

ysis

for

ET

BF

;19%

ofpa

tient

sw

ithdi

arrh

eaan

d4%

ofco

ntro

lsto

okan

tibio

tics

prio

rto

eval

uatio

nC

acer

eset

al.,

2000

(10)

Out

patie

ntch

ildre

n(�

10ye

ars

old)

with

and

with

out

diar

rhea

,19

98–1

999

Leo

n,N

icar

agua

106

patie

nts

with

diar

rhea

,60

age-

mat

ched

cont

rols

IMS-

PCR

,HT

29/C

1ce

llas

say

Ove

rall,

8.4%

ofpa

tient

sw

ithdi

arrh

eaan

d0%

ofco

ntro

lsw

ere

ET

BF

posi

tive

(P�

0.03

);ch

ildre

nof

�1

year

ofag

ew

ithdi

arrh

eaha

dth

ehi

ghes

tE

TB

Fis

olat

ion

rate

s(1

1.1%

vs4.

6%fo

rch

ildre

nof

�1

year

ofag

e�d

iffer

ence

not

sign

ifica

nt�)

Onl

yen

teri

cpa

rasi

tes

wer

eev

alua

ted

inco

ntro

lchi

ldre

n;st

ools

wer

efr

ozen

for

anun

spec

ified

time

prio

rto

anal

ysis

for

ET

BF

;inc

ompl

ete

mat

chin

gof

patie

nts

and

cont

rols

;ant

ibio

ticad

min

istr

atio

npr

ior

tost

ool

sam

plin

gw

asa

pote

ntia

lco

nfou

nder

Bre

ssan

eet

al.,

2001

(8)

Hos

pita

lized

imm

unod

efici

ent

child

ren

with

orw

ithou

tdi

arrh

ea;h

ealth

yco

ntro

lchi

ldre

nfr

omda

yca

rece

nter

s(d

ates

not

spec

ified

)

Sao

Paul

o,B

razi

l56

imm

unod

efici

ent

child

ren,

74co

ntro

lsSt

oolc

ultu

re,P

CR

for

bft

gene

onba

cter

ial

isol

ates

,HT

29ce

llas

say

Onl

yon

ech

ildw

ithA

IDS

and

nodi

arrh

eaw

asE

TB

Fpo

sitiv

eO

ther

ente

ric

path

ogen

sno

tev

alua

ted;

B.f

ragi

lisis

olat

ion

rate

sof

18to

27%

;all

imm

unod

efici

ent

child

ren

wer

eon

antib

iotic

sw

hen

stud

ied

Ant

unes

etal

.,20

02(4

)U

rban

child

ren

with

and

with

out

diar

rhea

(1-y

ear

stud

y)

Rio

deJa

neir

o,B

razi

l66

child

ren

(0to

5ye

ars

old)

with

diar

rhea

,25

child

ren

with

out

diar

rhea

Stoo

lcul

ture

,PC

Rfo

rbf

tge

nein

bact

eria

lis

olat

es,H

T29

cell

assa

y

ET

BF

was

isol

ated

from

one

child

with

diar

rhea

Clin

ical

deta

ils,i

nclu

ding

age

dist

ribu

tion

ofch

ildre

n,no

tpr

esen

ted;

25.8

%an

d84

%of

diar

rhea

and

cont

rols

tool

s,re

spec

tivel

y,w

ere

cultu

repo

sitiv

efo

rB

.fra

gilis

Krz

yzan

owsk

yan

dA

rila

-Cam

pos,

2003

(48)

Hos

pita

lized

child

ren

with

acut

edi

arrh

ea,h

ealth

yco

ntro

lchi

ldre

nfr

omda

yca

rece

nter

s(d

ates

not

spec

ified

)

Sao

Paul

o,B

razi

l96

patie

nts

with

diar

rhea

,74

cont

rols

Stoo

lcul

ture

,HT

29/C

1ce

llas

say,

PCR

for

bft

gene

onba

cter

ial

isol

ates

2.1%

ofpa

tient

sw

ithdi

arrh

eaan

d0%

ofco

ntro

lsw

ere

ET

BF

posi

tive

Eva

luat

ion

ofot

her

ente

ric

path

ogen

sun

clea

r;B

.fra

gilis

isol

atio

nra

tes

of13

.5%

to24

%

Ngu

yen

etal

.,20

05(1

18)

Hos

pita

l-bas

edst

udy

ofch

ildre

nw

ithdi

arrh

eaan

dhe

alth

yco

ntro

lch

ildre

nfr

omda

yca

reor

heal

thce

nter

,20

01–2

002

Han

oi,V

ietn

am58

7pa

tient

sw

ithdi

arrh

ea,

249

age-

mat

ched

cont

rols

(all

�5

year

sol

d)

IMS-

PCR

(with

only

MA

bC

3,no

tM

Ab

4H8)

Ove

rall,

7.3%

ofpa

tient

sw

ithdi

arrh

eaan

d2.

4%of

cont

rols

wer

eE

TB

Fpo

sitiv

e(P

�0.

01);

noas

soci

atio

nof

ET

BF

with

diar

rhea

inch

ildre

nun

der

1ye

arol

d;fo

rch

ildre

nof

�1

year

ofag

e,8.

9%w

ithdi

arrh

eaan

d2.

9%of

cont

rols

wer

eE

TB

Fpo

sitiv

e(P

�0.

01)

Cam

pylo

bact

ersp

p.no

tso

ught

due

tola

ckof

faci

litie

s;st

ools

wer

efr

ozen

for

anun

spec

ified

time

prio

rto

anal

ysis

for

ET

BF

;B.f

ragi

lisis

olat

ion

rate

sno

tsp

ecifi

ed;i

ncom

plet

em

atch

ing

ofpa

tient

san

dco

ntro

lsPa

thel

aet

al.,

2005

(85)

Lon

gitu

dina

lstu

dyof

birt

hco

hort

,199

3–19

96R

ural

Ban

glad

esh

(Mir

zapu

r)25

2ch

ildre

n(�

2ye

ars

old)

Stoo

lcul

ture

,HT

29/C

1ce

llas

say

ET

BF

isol

ated

from

40/1

97(2

0.3%

)B

.fr

agili

s-po

sitiv

edi

arrh

eals

tool

san

d15

/185

(8.1

%)

B.f

ragi

lis-p

ositi

veno

ndia

rrhe

alst

ools

(P�

0.00

1);

ET

BF

asso

ciat

edw

ithac

ute

diar

rhea

ldis

ease

inch

ildre

nof

�1

year

ofag

e;16

.3%

ofal

lchi

ldre

nw

ere

infe

cted

with

ET

BF

inth

efir

st2

year

sof

life

Fir

stco

mm

unity

-bas

edsu

rvei

llanc

est

udy

ofE

TB

Fin

fect

ion;

exte

nsiv

eev

alua

tion

ofen

teri

cpa

thog

ens;

cultu

rem

etho

dolo

gych

ange

dov

erth

est

udy

peri

od;o

nly

1,08

3/7,

429

(14.

6%)

stoo

lspe

cim

ens

wer

eB

.fra

gilis

cultu

repo

sitiv

eov

ertim

e

Con

tinue

don

follo

win

gpa

ge


on Decem


r.asm.org/

Dow

nloaded from

http://cmr.asm.org/

did not reach statistical significance in one study (7). Of note,in the latter study, ETBF isolates were identified by PCR inmucosal washes of �30% of all individuals examined by en-doscopy (7). In an additional prospective, cross-sectional epi-demiologic report from Turkey (113), fecal ETBF was isolatedfrom 38% of 56 patients with colorectal cancer but only 12% of40 sex- and age-matched concurrent controls (P � 0.01).

Several studies, largely conducted in microbiology laboratories,have examined whether ETBF strains are isolated in excessamong bloodstream isolates (Table 6). In six of eight studies,ETBF isolation from blood exceeded isolation from other ex-traintestinal specimens, but the results reached statistical signifi-cance in only one study (38). Only one study (26) examinedwhether isolation of ETBF from blood was associated with evi-dence of excess clinical morbidity or mortality, and it did not findan association. However, when the studies are combined, 232 of1,325 extraintestinal B. fragilis isolates (17.5%) were identified asETBF, with 86 of 368 blood isolates (23.4%) and 146 of 957 otherextraintestinal isolates (15.3%) being ETBF (P � 0.0005; two-sided chi-square test), suggesting that ETBF may be isolatedmore frequently from bloodstream infections. Additional studieswill be necessary to determine if ETBF strains exhibit increasedvirulence in extraintestinal infections, consistent with their viru-lence in colonic disease.

Diagnosis of ETBF Infection

Table 7 summarizes the approaches to diagnosis of ETBFinfection. Diagnosis of ETBF infection from stool is difficult.Similar to C. difficile diagnosis, detection of the bft gene or itsbiologic activity is required to diagnose ETBF colonization ordisease. Most studies have used one of two approaches, eitherB. fragilis fecal isolation with testing for the bft gene (by PCR)or for BFT (in the HT29/C1 cell assay) or direct examinationfor the bft gene in DNA extracted from stool. The HT29/C1cell assay detects the biologic activity of BFT in culture super-natants of ETBF strains, demonstrating a sensitivity of 89%and a specificity of 100% compared to ETBF strain identifica-tion by the ability to stimulate secretion in the lamb intestinalloop assay (119) (Fig. 2). The HT29/C1 cell assay detects aslittle as 0.5 pM purified BFT; the half-maximal concentrationof BFT detected by the HT29/C1 cell assay is 12.5 pM BFT(97). Analysis of our collection of B. fragilis strains (n � 304strains) indicates that the presence of the bft gene correlateswell with the ability to detect BFT in the HT29/C1 cell assay(29). In rare instances (n � 2 strains), deletions/mutations inthe bft gene have been identified that result in a lack of syn-thesis and/or secretion of biologically active BFT by an ETBFstrain (29).

Available data further suggest that overnight anaerobic cul-tivation of stool in a nutrient broth promotes enhanced recov-ery of B. fragilis and ETBF (102). BFT biologic activity hasbeen detected directly in fecal supernatants from patients withdiarrhea by use of the HT29/C1 cell assay (84). Enzyme im-munoassays (EIAs) have been reported for detection of BFT,and in preliminary data, detection of BFT in stool samples byEIA has been reported (89, 116). A report of a combinedimmunomagnetic separation (to concentrate B. fragilis fromstool) and PCR approach (termed IMS-PCR) to detect ETBFwas encouraging (sensitivity, 50 CFU ETBF/g stool), but this

TA

BL

E2—

Con

tinue

d

Ref

eren

ce(a

utho

r,yr

�ref

eren

ceno

.�)St

udy

desi

gnL

ocat

ion

nE

TB

Fdi

agno

stic

met

hod

Mai

nre

sults

Stre

ngth

s/lim

itatio

nsof

stud

y

Dur

maz

etal

.,20

05(2

4)U

rban

outp

atie

ntch

ildre

nan

dad

ults

with

acut

edi

arrh

ea;h

ospi

taliz

edco

ntro

ls

Mal

atya

,Tur

key

117

child

ren

(0to

16ye

ars

old)

and

104

adul

tsw

ithdi

arrh

ea;1

02ch

ildre

nan

d95

adul

tsw

ithou

tdi

arrh

ea(t

ime

and

age

mat

ched

)

Nes

ted

PCR

for

bft

gene

onst

ool

sam

ples

ET

BF

asso

ciat

edw

ithdi

arrh

eain

child

ren

of1

to5

year

sof

age

(25%

vs9.

5%in

cont

rols

;P�

0.05

);no

asso

ciat

ion

ofE

TB

Fw

ithdi

arrh

eain

patie

nts

of�

1ye

arof

age

or�

5ye

ars

ofag

e

Dia

rrhe

a-as

soci

ated

E.c

oli

clas

ses

not

eval

uate

d;pa

tient

popu

latio

nw

asou

tpat

ient

s,bu

tco

ntro

lpop

ulat

ion

was

hosp

italiz

ed

Coh

enet

al.,

2006

(16)

Cas

e-co

ntro

lser

ies

(dat

esno

tsp

ecifi

ed)

Uni

vers

ityof

Cal

iforn

ia,D

avis

,Sa

cram

ento

,CA

152

patie

nts

with

noso

com

iald

iarr

hea

(with

out

C.d

iffici

le),

85ou

tpat

ient

cont

rols

unde

rgoi

ngsu

rvei

llanc

eco

lono

scop

y

Stoo

lcul

ture

,nes

ted

PCR

for

bft

gene

from

B.f

ragi

lisis

olat

esan

dfe

cal

DN

A

30/1

52(1

9.7%

)an

d28

/85

(32%

)di

arrh

eaan

dco

ntro

lsto

ols,

resp

ectiv

ely,

wer

ecu

lture

posi

tive

for

B.f

ragi

lis;4

patie

nts

wer

eE

TB

Fpo

sitiv

eby

cultu

re,b

ut14

/152

(9.2

%)

patie

nts

wer

epo

sitiv

eby

dire

ctst

oolP

CR

for

bft

sequ

ence

s;2.

3%(n

�2)

ofco

ntro

lsub

ject

sw

ere

posi

tive

for

ET

BF

(P�

0.04

)

Sugg

ests

that

ET

BF

may

cont

ribu

teto

noso

com

ial

diar

rhea

and

that

dire

ctst

ool

PCR

for

bft

gene

impr

oves

diag

nost

icse

nsiti

vity

aA

llst

udie

sin

clud

eda

cont

rolp

opul

atio

nan

dar

elis

ted

inch

rono

logi

calo

rder

.ET

BF

,ent

erot

oxig

enic

B.f

ragi

lis;I

MS-

PCR

,im

mun

omag

netic

sepa

ratio

n-PC

R;R

ITA

RD

,rev

ersi

ble

inte

stin

altie

adul

trab

bitd

iarr

hea;

MA

b,m

onoc

lona

lant

ibod

y;H

T29

cells

,par

enta

lHT

29ce

lllin

e;H

T29

/C1

cells

,clo

ned

HT

29ce

lllin

e.


on Decem


r.asm.org/

Dow

nloaded from

http://cmr.asm.org/

method requires the use of two anti-B. fragilis antibodies forbest sensitivity, and one antibody has been lost (129; A. Wein-traub, Karolinska Institute, personal communication). Furthercommercial development of an EIA to detect fecal BFT or ofIMS-PCR has not occurred. Limited comparative data suggest,not unexpectedly, that direct stool-based PCR approaches aremore sensitive for the diagnosis of ETBF colonization or dis-ease than fecal cultivation of B. fragilis, followed by detectionof bft or BFT (16, 88, 107). Anaerobic stool culture of B. fragilisis affected by delays in processing of samples as well as (likely)

the general difficulty of anaerobic microbiology (95, 107). Us-ing stool culture, detection of ETBF is further affected by thefecal heterogeneity of B. fragilis strains, where both NTBF andETBF may coexist (16), although one report suggested dis-placement of fecal NTBF by ETBF in diarrheal disease (95).

Together, the data suggest that similar to detection ofETEC, testing of multiple fecal B. fragilis colonies for bft/BFTis necessary to enhance the diagnostic sensitivity if fecal cultureis the chosen diagnostic approach. Overall, it remains uncer-tain which method(s), and in particular which PCR approach,

TABLE 3. Miscellaneous clinical reportsa

Reference (author,yr �reference no.�) Study designb Location n ETBF diagnostic

method Main results Limitations

Myers et al., 1987(70)

Case series Montana and NavajoArea IndianHealth Service,AZ

10 patients withdiarrhea inMontana and34 infants (2to 14 monthsold) inArizona

Stool culture, LLIL,and rabbitinoculation

8 patients diagnosed with ETBF,with one infant having aconcurrent infection withETEC; persistent (4 weeks)and chronic (3 years) diarrheaas well as frank/occult fecalblood noted in some patients

No controls includedin the study

Meisel-Mikolajczyket al., 1994 (58)

Case series Warsaw, Poland 120 hospitalizedchildren (2weeks to 3.5years old), 56patients withdiarrhea

Stool culture,HT29/C1 cellassay

16.7% of stools were culturepositive for B. fragilis; twochildren were ETBF positive

Clinical details limited,but majority ofpatients hadpneumonia andantibiotic therapy;low B. fragilisisolation rate

Aucher et al., 1996(5)

Case report Poitiers, France 1 Cerebrospinal fluidculture, HT29/C1cell assay

Meningitis in a neonate withmedullary-colonic fistula, Hxbloody diarrhea; cross-reactionwith Haemophilus influenzaetype 6 latex agglutinationnoted

Single case of ETBFmeningitis; role ofBFT in diseasepathogenesisunknown

Leszczynski et al.,1997 (51)

Case series Warsaw, Poland 120 pregnantwomen

Vaginal culture,PCR for bft gene

6.6% of vaginal cultures werepositive for B. fragilis; oneETBF isolate was identified;no identified clinical sequelae

Single case of vaginalETBF colonization

Meisel-Mikolajczyket al., 1999 (57)

Case series Warsaw, Poland 50 cases ofnosocomialdiarrhea

Stool culture,HT29/C1 cellassay, PCR for bftgene in stool orbacterial isolates

17/50 (34%) stools were culturepositive for B. fragilis; 4/17(23%) patients had ETBF,among whom 3 patients werealso positive for toxigenic C.difficile; DNA from ETBFpositive stools was also PCRpositive for bft gene

No control populationstudied; role ofETBF in nosocomialdiarrhea not defined

Kato et al., 2000(40)

Collection of stoolsfrom individualswithout diarrhea(hospital visitors,employees,students, andnursing homeresidents)

Japan 361 stools(nondiarrheal)

Culture, PCR forbft gene

Overall, B. fragilis was isolatedfrom 38.8% of stools (range,21.1% for children of �1 yearold to 51.3% for 7- to 14-year-olds); 25/140 (17.8%) B.fragilis strains were ETBF(range, 0% for children of �1year old to 8.9% for 1- to 6-year-olds and 17- to 64-year-olds)

Clinical details (e.g.,antibiotic use) notpresented

Martirosian et al.,2001 (52)

Case series Poland 34 appendices Culture, HT29/C1cell assay, PCRfor bft gene

2/34 (5.9%) appendices (fromone adult and one child)carried ETBF; bothappendices were gangrenous

Role of ETBF indisease pathogenesisunknown

Pituch et al., 2003(86)

Case series, 2000–2001

Warsaw, Poland 332 patients withnosocomialdiarrhea (272adults, 60children)

Stool culture,HT29/C1 cellassay, PCR for bftgene on bacterialisolates

50/332 (15%) stools were culturepositive for B. fragilis; 9/50(18%) patients were positivefor ETBF, among whom 5patients were also positive fortoxigenic C. difficile

No control populationstudied; role ofETBF in nosocomialdiarrhea not defined

Antunes et al.,2004 (3)

Case series, 2000–2001

Rio de Janeiro,Brazil

334 hospitalizedchildren withdiarrhea

Stool culture, PCRfor bft gene inbacterial isolates

31/334 (9.3%) stools wereculture positive for B. fragilis;no ETBF identified

Low isolation rate forB. fragilis; no controlpopulation studied;clinical diarrhealdisease data notpresented

a Case reports and series lacking control populations. LLIL, lamb ligated intestinal loop assay.b Dates for studies not specified except in a few instances.


on Decem


r.asm.org/

Dow

nloaded from

http://cmr.asm.org/

will provide the best diagnostic sensitivity and specificity fordetection of ETBF disease or colonization. Table 8 summa-rizes the primers utilized to date for detection of the bft genein B. fragilis isolates or DNA extracted from stool. In someinstances, reported primer use has included multiple base pairmismatches, suggesting the possibility of decreased diagnosticsensitivity and specificity (86). Nested PCR is the most sensi-tive PCR method to detect the bft gene reported to date,detecting 102 to 103 CFU ETBF/g stool (108). The develop-ment of rapid, accurate, and sensitive diagnostic testing forETBF organisms will enhance assessments of the epidemiologyof these bacteria and their disease associations as well as beingan important prerequisite to consideration of therapeutic trialsfor ETBF disease.

Therapy of ETBF-Associated Diarrheal Disease

Because diagnosis of ETBF as an etiology of diarrheal dis-ease is presently limited to the research setting and the diag-nosis is often delayed even if sought, no controlled studies haveevaluated whether antibiotic therapy shortens ETBF-associ-ated diarrheal illnesses. Reports of antibiotic sensitivity testingof clinical ETBF strains are limited to 58 strains from Ban-gladesh and 9 strains from Nicaragua (10, 102). Whereas allstrains were resistant to ampicillin in Nicaragua, 97% of Ban-gladeshi ETBF strains were sensitive to ampicillin; conversely,all strains were sensitive to metronidazole in Nicaragua, but arelatively high proportion (7%) of ETBF strains were metro-nidazole resistant in Bangladesh. Approximately 10% of ETBFstrains were clindamycin resistant in both locales, and 26% ofETBF strains were resistant to tetracycline in Bangladesh (notevaluated in Nicaragua). Because available reports suggest thatETBF disease is self-limited, adequate oral rehydration ther-apy is central to clinical care, consistent with the therapeuticapproach to other diarrheal illnesses.

BFT GENETICS AND PROTEIN STRUCTURE

Three distinct alleles of bft have been identified, namely,bft-1, bft-2, and bft-3 (13, 31, 40, 46). The terminology bft-1,bft-2, and bft-3 was adopted to recognize the temporal se-quence in which the alleles were identified. ETBF strains maypossess two copies of a single bft genotype, but no ETBFstrains have yet been reported to contain mixed bft alleles. Thebft genes are chromosomal, with a G�C content of 39%, andare predicted to encode a 397-residue holotoxin with a calcu-lated molecular mass of ca. 44.5 kDa. The predicted BFTstructure is a preproprotein holotoxin (Fig. 3) (31, 46, 115,121). The initial 18 amino acids of the BFT holotoxin make upa signal peptide (preprotein domain) predicted as important indelivery of the holotoxin to the B. fragilis membrane, where the193-residue proprotein toxin domain is postulated (by analogyto other bacterial zinc-dependent metalloproteases [56]) to beinstrumental in proper protein folding and secretion of mature,biologically active BFT (31). The pre- and proprotein domainsof all BFTs exhibit high sequence homology (ca. 97 to 98%),consistent with conserved functions for these protein domains.Based on N-terminal sequencing of purified BFT, the propro-tein domain is cleaved at an Arg-Ala site (amino acids 211 and212) to release mature BFT from the bacterial cell (31, 115,

TA

BL

E4.

Clin

ical

obse

rvat

ions

ofE

TB

Fdi

arrh

eali

llnes

ses

Clin

ical

char

acte

rist

ic

Val

uefo

rgr

oup

from

indi

cate

dst

udy

refe

renc

ec

Apa

che

Indi

anch

ildre

nan

dad

ults

,AZ

(n�

33)a

(95)

Chi

ldre

nof

�5

yrof

age,

Dha

ka,

Ban

glad

esh

(n�

10)a

(94)

Chi

ldre

n(�

1yr

to�

10yr

),O

klah

oma

City

,O

K(n

�43

)a(9

9)

Adu

lts(�

15yr

),Sw

eden

(n�

91)a

(128

)

Chi

ldre

n(�

2yr

),N

icar

agua

(n�

9)a

(10)

Chi

ldre

n(�

5yr

),V

ietn

am(n

�19

)a(1

18)

Chi

ldre

n(�

2yr

),ru

ral

Ban

glad

esh

(n�

18E

TB

Fep

isod

esin

15ch

ildre

n)a

(85)

Chi

ldre

nan

dad

ults

,Dha

ka,

Ban

glad

esh

(n�

43ch

ildre

nof

�15

yran

d30

adul

ts)a

(102

)

Med

ian

dura

tion

ofdi

arrh

ea(d

ays)

33.

5—

203

—6

3e

Ran

ge(d

ays)

1–14

�1–

�15

——

2–20

—2–

392–

11M

edia

nno

.of

stoo

ls/d

ay6

8—

�5,

29%

;5–1

0,40

%;

�10

,30%

75

5—

Ran

ge3–

163–

15—

—4–

102–

93–

20—

Fev

er(%

)b58

—25

4444

536

7V

omiti

ng(%

)21

—33

4444

426

—A

bdom

inal

pain

(%)

——

—69

——

—88

Blo

ody

stoo

ls(%

)9

1042

14—

522

8f

Deh

ydra

tion

(%)

18—

—37

—68

619

Hos

pita

lizat

ion

(%)

1510

0A

llou

tpat

ient

s49

All

outp

atie

nts

—C

omm

unity

surv

eilla

nce

—O

ther

ente

ric

path

ogen

s(%

)4

547

5345

5655

8Pe

rsis

tent

diar

rhea

(%)d

—10

21—

22—

—0

aN

umbe

rof

patie

nts

with

ET

BF

-ass

ocia

ted

diar

rhea

onw

hich

data

pres

ente

dar

eba

sed.

The

brea

kdow

nof

child

ren

and

adul

tsin

refe

renc

e95

isun

clea

r.In

refe

renc

es94

,118

,and

128,

the

data

show

nar

efo

rpa

tient

sw

ithE

TB

Fal

one,

with

outo

ther

ente

ric

path

ogen

s.T

heto

taln

umbe

rsof

ET

BF

-ass

ocia

ted

diar

rhea

case

sre

port

edin

thes

ere

fere

nces

are

22,4

3,an

d19

5,re

spec

tivel

y.R

efer

ence

85re

port

sa

tota

lof4

0E

TB

Fdi

arrh

eal

epis

odes

in33

child

ren.

bW

hen

defin

ed,t

empe

ratu

reof

�38

°C.

c—

,dat

ano

tre

port

ed.

dD

efine

das

�14

days

ofdi

arrh

ea.

eSe

vent

y-ni

nepe

rcen

tof

the

popu

latio

nre

port

edno

ctur

nald

iarr

hea.

fE

ight

perc

ent

ofst

ools

wer

efe

calo

ccul

tbl

ood

posi

tive.


on Decem


r.asm.org/

Dow

nloaded from

http://cmr.asm.org/

121). Consistent with these predictions, ca. 44-kDa and 20-kDaproteins are detected with anti-BFT antisera in lysates ofETBF strains (28, 29).

The predicted mature toxin domain of each bft allelecontains an extended zinc-binding metalloprotease motif,HEXXHXXGXXH, and a perfectly superimposable methio-nine residue close to the metalloprotease motif. These datasuggest that BFTs are members of the matrix metalloproteasesubfamily (matrixins) of the metzincin superfamily of zinc-dependent metalloprotease enzymes (61, 79). Limited homol-ogy between eukaryotic matrix metalloproteases and BFT ledto the hypothesis that BFT may be an ancestor of host matrixmetalloproteases (53). As predicted, the proteolytic activity ofBFT appears to be crucial to its biologic activity (61, 122), andBFT contains 1 g-atom of Zn2� per toxin molecule (61). Fur-thermore, zinc chelation reduces BFT biologic activity by ca.90% (47, 61, 80).

Point mutations to modify each of the conserved aminoacids of the extended metalloprotease motif as well as theconserved downstream methionine also reduce or eliminate

the biologic activities of BFT (28). Although BFT has beenreported to be autoproteolytic (61, 116), BFT metalloproteasepoint mutations do not alter the intracellular processing andsecretion of BFT by B. fragilis, suggesting that other intracel-lular B. fragilis proteases process the holotoxin to mature BFT(28). Additional mutational analysis of the C-terminal regionof BFT indicated that this region is intolerant to modest aminoacid deletions, suggesting that this region is also important forBFT activity (104). Truncation mutations removing only twoC-terminal amino acids reduced BFT biologic activity, andremoval of eight (or more) amino acids obliterated it. BFTmutants lacking eight or more C-terminal amino acids wereexpressed similar to wild-type toxin, but the mutant BFTs wereunstable (104).

The predicted amino acid sequences of the BFT holotoxinproteins reveal highly homologous proteins, with BFT-1 andBFT-2 being 95% similar and 92% identical (31). BFT-3 ismore closely related to BFT-2 (93% and 96% identical toBFT-1 and BFT-2, respectively) (13). The predicted proteindomains of the toxins, however, exhibit differing degrees of

TABLE 5. ETBF association with IBD or colorectal cancer

Reference (author,yr �reference no.�) Study design Location n

ETBFdiagnostic

methodMain results Limitations

Prindiville et al.,2000 (88)

Endoscopy-basedcase series,with controls

University ofCalifornia,Davis,Sacramento,CA

101 patients withIBD or diarrhea,69 controls

Stool culture,HT29 cellassay,nestedPCR forbft geneDNAextractedfrom stool

11/57 (19.3%) patients withactive IBD were positivefor ETBF by stool PCRvs 0/26 patients withinactive IBD (P � 0.01);5/18 (27.8%) patientswith diarrhea and 2/69(2.9%) controls wereETBF positive by stoolPCR (P � 0.0005); only4.8% with IBD and 11%with diarrhea wereETBF positive by cultureand HT29 cell assay

Limited study designdetails, includingsource of controls;stools were frozenfor 2 weeks priorto analysis byculture; B. fragilisisolation rates notreported

Bassett et al.,2004 (7)

Endoscopy-basedcase series,with controls

London,UnitedKingdom

35 patients with IBDand 37 controls;control populationincluded 19patients withculture-negativediarrhea and 18patients withoutdiarrhea, all ofwhom were beinginvestigated byendoscopy forabdominal pain orchanges in bowelhabits

Nested PCRfor bftgene onDNAextractedfromluminalwashingsor biopsies

Of 60 patients with luminalwashings available foranalysis, 10/28 (35.7%)“controls” (6/14 withdiarrhea and 4/14without diarrhea) and8/32 (25%) IBD patientswere ETBF positive; 7/25(28%) patients withactive IBD and 1/7(14.3%) patients withinactive IBD were ETBFpositive (difference notsignificant); 4/32 (12.5%)IBD patients and 6/33(18.2%) controls withcolonic biopsies wereETBF positive

Small study withlimited details oncontrolpopulation;populationdefined ascontrols includedpatients with andwithout diarrhea;colonic biopsiesand luminalwashings werefrozen for anunspecified timeprior to analysis

Ulger et al.,2006 (113)

Prospective,consecutivecase serieswithconcurrentcontrols

Istanbul,Turkey

73 patients withcolorectal cancer;59 healthycontrols (age andgender matched)

Stool culture,PCR forbft gene onfecal B.fragilisisolates

B. fragilis isolated from77% of patients and 68%of controls (differencenot significant); 21/56(38%) patients and 5/40(12%) controls wereETBF positive (P �0.009)

Small cross-sectionalstudy without fullcolon cancer riskfactor analysis


on Decem


r.asm.org/

Dow

nloaded from

http://cmr.asm.org/

TA

BL

E6.

Mic

robi

olog

yla

bora

tory

-bas

edst

udie

sa

Ref

eren

ce(a

utho

r,yr

�ref

eren

ceno

.�)St

udy

desi

gnL

ocat

ion

nE

TB

Fdi

agno

stic

met

hod

Mai

nre

sults

Lim

itatio

ns

Pant

osti

etal

.,19

94(8

2)M

icro

biol

ogy

labo

rato

ry-b

ased

stud

y(d

ates

not

spec

ified

)It

aly

146

B.f

ragi

lisis

olat

es(9

5ex

trai

ntes

tinal

,51

feca

l)H

T29

cell

assa

y11

.5%

ofex

trai

ntes

tinal

and

10%

offe

calB

.fra

gilis

isol

ates

wer

eE

TB

F

No

bloo

dis

olat

esex

amin

ed;

limite

dcl

inic

alde

tails

avai

labl

eK

arun

iaw

ati

etal

.,19

95C

olle

ctio

nof

B.f

ragi

lisis

olat

es,1

987–

1989

Japa

n14

1ex

trai

ntes

tinal

clin

ical

B.

frag

ilis

isol

ates

(25

bloo

dis

olat

es)

HT

29/C

1ce

llas

say

Ove

rall,

23.4

%of

B.f

ragi

lisis

olat

esw

ere

ET

BF

;9/2

5(3

6%)

bloo

dan

d24

/116

(20.

7%)

othe

rex

trai

ntes

tinal

isol

ates

wer

eE

TB

F(d

iffer

ence

not

sign

ifica

nt)

No

clin

ical

deta

ilsav

aila

ble

Mun

dyan

dSe

ars,

1996

(63)

Mic

robi

olog

yla

bora

tory

-bas

edst

udy,

1992

–199

4B

altim

ore,

MD

65ex

trai

ntes

tinal

clin

ical

B.

frag

ilis

isol

ates

(28

bloo

dis

olat

es)

Cul

ture

,HT

29C

1ce

llas

say,

BF

Ten

zym

e-lin

ked

imm

unos

orbe

ntas

say

Ove

rall,

6.2%

ofB

.fra

gilis

isol

ates

wer

eE

TB

F(i

nclu

ding

one

bloo

dis

olat

e)

�33

%of

B.f

ragi

lisis

olat

esfr

omth

est

udie

dtim

epe

riod

are

unav

aila

ble

Kat

oet

al.,

1996

(38)

Cen

tral

mic

robi

olog

yla

bora

tory

-bas

edst

udy,

1987

–198

9

Japa

n18

8un

sele

cted

B.f

ragi

lisis

olat

esfr

omex

trai

ntes

tinal

clin

ical

sam

ples

Cul

ture

,HT

29/C

1ce

llas

say,

PCR

for

bft

gene

Ove

rall,

18.6

%of

B.f

ragi

lisis

olat

esw

ere

ET

BF

;18/

64(2

8.1%

)bl

ood

and

17/1

24(1

3.7%

)ot

her

extr

aint

estin

alis

olat

esw

ere

ET

BF

(P�

0.05

)

No

clin

ical

deta

ilsav

aila

ble

Szok

eet

al.,

1997

(112

)M

icro

biol

ogy

labo

rato

ry-b

ased

stud

y,19

95–1

996

Szeg

ed,H

unga

ry13

4B

.fra

gilis

isol

ates

(40

from

outp

atie

nts

with

diar

rhea

with

out

anot

her

caus

e,20

from

heal

thy

cont

rols

tool

s,74

from

othe

rcl

inic

alsa

mpl

es)

Cul

ture

,HT

29ce

llas

say

40%

,15%

,and

20.3

%of

B.f

ragi

lisis

olat

esfr

omdi

arrh

eals

tool

s,co

ntro

lsto

ols,

and

extr

aint

estin

also

urce

s,re

spec

tivel

y,w

ere

ET

BF

;sto

olfil

trat

esw

ere

also

posi

tive

inH

T29

cell

assa

y

No

bloo

dis

olat

este

sted

;clin

ical

deta

ilsno

tav

aila

ble,

but

extr

aint

estin

alE

TB

Fte

nded

tobe

high

er-t

iter

BF

Tpr

oduc

ers

Chu

nget

al.,

1999

(13)

Mic

robi

olog

yla

bora

tory

-bas

edst

udy,

1995

–199

7Se

oul,

Kor

ea89

extr

aint

estin

alcl

inic

alB

.fr

agili

sis

olat

es(2

2bl

ood

isol

ates

)

HT

29/C

1ce

llas

say,

PCR

and

colo

nybl

othy

brid

izat

ion

for

bft

gene

Ove

rall,

38%

ofB

.fra

gilis

isol

ates

wer

eE

TB

F;1

2/22

(54.

5%)

bloo

dis

olat

esan

d22

/67

(33%

)ot

her

extr

aint

estin

alis

olat

es(P

�0.

07)

wer

eE

TB

F

No

clin

ical

deta

ilsav

aila

ble

Cla

ros

etal

.,20

00(1

4)M

icro

biol

ogy

labo

rato

ry-b

ased

stud

y(d

ates

not

spec

ified

)G

erm

any,

Sout

hern

Cal

iforn

ia93

extr

aint

estin

alcl

inic

alB

.fr

agili

sis

olat

es(1

3bl

ood

isol

ates

)

PCR

for

bft

gene

Ove

rall,

11%

ofB

.fra

gilis

isol

ates

wer

eE

TB

F;3

/13

(23%

)bl

ood

isol

ates

and

7/80

(9%

)ot

her

extr

aint

estin

alis

olat

es(d

iffer

ence

not

sign

ifica

nt)

wer

eE

TB

F

No

clin

ical

deta

ilsav

aila

ble

Fou

lon

etal

.,20

03(2

6)M

icro

biol

ogy

labo

rato

ry-b

ased

stud

yof

bloo

dis

olat

es(1

989–

2002

)an

dot

her

isol

ates

(200

0–20

03)

Bel

gium

166

extr

aint

estin

alcl

inic

alis

olat

es(1

03bl

ood

isol

ates

)PC

Rfo

rbf

tge

neO

vera

ll,24

.1%

ofB

.fra

gilis

isol

ates

wer

eE

TB

F;2

7/10

3(2

6.2%

)bl

ood

and

13/6

3(2

0.6%

)ot

her

extr

aint

estin

alis

olat

esw

ere

ET

BF

(diff

eren

ceno

tsi

gnifi

cant

)

Rev

iew

ofcl

inic

alda

ta,i

nclu

ding

shoc

kan

dm

orta

lity

rate

s,in

dica

ted

nodi

ffere

nces

betw

een

patie

nts

with

ET

BF

orN

TB

Fba

cter

emia

Obu

ch-W

oszc

zaty

nski

etal

.,20

04(8

1)M

icro

biol

ogy

labo

rato

ry-b

ased

stud

y(d

ates

not

spec

ified

)T

heN

ethe

rlan

ds,

Pola

nd21

0B

.fra

gilis

isol

ates

(78

extr

aint

estin

alis

olat

esfr

omT

heN

ethe

rlan

ds;1

32fe

cal

isol

ates

,mai

nly

from

antib

iotic

-ass

ocia

ted

diar

rhea

,fro

mPo

land

)

PCR

for

bft

gene

15%

ofB

.fra

gilis

isol

ates

from

The

Net

herl

ands

and

13%

ofis

olat

esfr

omPo

land

wer

eE

TB

F

No

clin

ical

deta

ilsav

aila

ble

Cla

ros

etal

.,20

06(1

5)M

icro

biol

ogy

labo

rato

ry-b

ased

stud

y(d

ates

not

spec

ified

)U

nite

dSt

ates

(CA

,IL

),G

erm

any

260

extr

aint

estin

alcl

inic

alB

.fr

agili

sis

olat

es(6

3bl

ood

isol

ates

,197

skin

and

soft

tissu

eis

olat

es)

PCR

for

bft

gene

Ove

rall,

12.3

%of

B.f

ragi

lisis

olat

esw

ere

ET

BF

;12/

63(1

9%)

bloo

dis

olat

esan

d20

/197

(10%

)sk

inan

dso

fttis

sue

isol

ates

wer

eE

TB

F(d

iffer

ence

not

sign

ifica

nt)

No

clin

ical

deta

ilsav

aila

ble

Avi

la-C

ampo

set

al.,

2007

(6)

Col

lect

ion

ofB

.fra

gilis

isol

ates

(dat

esno

tsp

ecifi

ed)

Uni

ted

Stat

es(C

A)

323

extr

aint

estin

alcl

inic

alis

olat

es(5

0bl

ood

isol

ates

,6fe

cali

sola

tes)

PCR

for

bft

gene

Ove

rall,

13.6

%of

B.f

ragi

lisis

olat

esw

ere

ET

BF

;4/5

0(8

%)

bloo

dis

olat

es(i

nclu

ding

one

cath

eter

tipis

olat

e)an

d40

/273

(14.

6%)

othe

rex

trai

ntes

tinal

isol

ates

wer

eE

TB

F;n

ofe

cali

sola

tes

wer

eE

TB

F

No

clin

ical

deta

ilsav

aila

ble

aSt

udie

sof

B.f

ragi

lisis

olat

eco

llect

ions

.Ref

eren

ces

for

stud

ies

incl

udin

gB

.fra

gilis

bloo

dis

olat

esar

esh

own

inbo

ld.


on Decem


r.asm.org/

Dow

nloaded from

http://cmr.asm.org/

identity, with only 2 to 5 amino acid changes between BFT-1,BFT-2, and BFT-3 in the preproprotein domains (13), whereasup to 25 amino acid changes occur in the mature toxin proteindomains (13, 31). Consistent with the predicted protein diver-sity in the mature BFT protein, BFT-1 and BFT-2 purify withdistinct biochemical profiles and differ in sodium dodecyl sul-fate-polyacrylamide gel electrophoresis mobility and two-di-mensional gel electrophoresis mobility, consistent with the pre-dicted uniqueness of the BFT-1 and BFT-2 proteins (121).BFT-2 also exhibits modest but consistently greater specificbiologic activity than BFT-1 in vitro, although the importanceof this observation to disease pathogenesis is unknown (121).BFT-1 and BFT-2 are trypsin resistant and stable over a widepH range (i.e., pHs 5 to 10) (97, 115, 121), potentially enablingthese toxins to resist degradation in animal and human guts.BFT-3 exhibits a purification profile and specific activity (bio-logic activity/mg protein) similar to those of BFT-2, consistentwith its greater homology to BFT-2 than BFT-1, but otherdetails on the properties of BFT-3 are not available (13).

Only limited investigations have characterized the epidemi-ology of the specific bft alleles of ETBF, but available dataindicate that all three bft alleles are globally distributed (Table9). Although the distribution of the bft-3 allele is not restrictedto Southeast Asia (6, 21), ETBF strains possessing bft-3 havebeen reported predominantly from Korea, Japan, or Vietnam,suggesting that regional evolution of ETBF may have occurred(13, 40, 76). Overall, the data suggest that the bft-1 allele ismost common among human ETBF strains evaluated to date.

GENETICS OF ETBF

Virulence genes of some organisms are clustered in uniquechromosomal loci, termed pathogenicity islands, that possessat least two virulence genes and have a G�C content differingfrom that of the host organism chromosome, with the lattersuggesting acquisition of the sequences from a foreign organ-ism (33). Study of our bank of NTBF (n � 191) and ETBF (n �113) strains, using probes derived from restriction enzymemapping of cosmid clones containing the bft gene, revealed fivedistinct patterns of hybridization, three of which predominated(Fig. 4). Strains with these genetic patterns were termed pat-tern I, II, and III B. fragilis strains (29). The chromosomes ofall ETBF strains (pattern I strains) possess a 6-kb DNA regionnot found in NTBF strains (29, 60). This 6-kb region containsthe bft gene and a second putative virulence gene, termed themetalloprotease II gene (mpII). Because this 6-kb DNA regioncontains two potential virulence genes and its G�C content is35%, differing significantly from the predicted ca. 43% G�Ccontent for the B. fragilis chromosome (11, 49), this 6-kb DNAregion, exclusively present in ETBF, was termed the B. fragilispathogenicity island or islet (due to its relatively small size)(BfPAI) (29, 60). Sequence analysis of the BfPAI revealed aca. 700-bp region upstream of bft with five putative B. fragilispromoter consensus sequences; this 700-bp promoter region is

TABLE 7. Approaches to diagnosis of ETBF infection

Diagnostic method Advantages Limitations

Stool culture B. fragilis isolation allowing directconfirmation of bft gene by PCR or ofBFT secretion by the HT29/C1 cell assay

Labor-intensive and expensive; delays diagnosisseveral days; dependent on anaerobicmicrobiology expertise

Stool PCR for bft gene Rapid compared to stool culture for B.fragilis

Requires fecal DNA extraction; sensitivitypotentially limited by fecal inhibitors of PCR;sensitivity enhanced by using overnightenrichment culture; absence of B. fragilisisolation for diagnosis confirmation

HT29/C1 cell assaya Detects BFT biologic activity directly in stoolor in culture supernatants of B. fragilisisolates; excellent correlation with lambintestinal loop assay

Expensive, labor-intensive, requires anaerobicmicrobiology expertise and subjectiveinterpretation of HT29/C1 cell assay unlessBFT-neutralizing antibody available

Enzyme-linked immunosorbentassay for fecal BFT

Potentially rapid diagnostic approach Only limited data support detection of BFT instool

IMS-PCR Potentially time-saving; B. fragilis isolationand PCR for bft gene can be performed inparallel on same sample

Requires noncommercial reagents; variableperformance to date

Intestinal loop assaysb Detects secretion stimulated by ETBF orBFT

Requires prior B. fragilis isolation; expensive andlabor-intensive; a research tool only

RITARD modelc Detects ETBF disease Requires prior B. fragilis isolation; expensive andlabor-intensive; a research tool only

a Parental HT29 cells sometimes substituted.b Usually lamb or rabbit.c RITARD, reversible ileal tie adult rabbit diarrhea.

FIG. 2. Effect of BFT on HT29/C1 cells in vitro. HT29/C1 cells (ahuman colonic carcinoma continuous cell line) exhibit morphologicalchanges, including cell rounding and dissolution of cell clusters, whentreated with BFT (5 nM). (Reprinted from reference 106 with permis-sion from Elsevier.)


on Decem


r.asm.org/

Dow

nloaded from

http://cmr.asm.org/

TABLE 8. Oligonucleotide primers used to identify ETBF strains and bft isoformsd

Primer use and name Primer sequence (5–3)b Primer 5 positionin bft (bp)

No. of mismatchesApproachc Reference(s)

bft-1 bft-2 bft-3

Identification of ETBFstrains

BF1 (forward) GACGGTGTATGTGATTTGTCTGAGAGA

654 3 3 3 PCR 83, 107

BF2 (reverse) ATCCCTAAGATTTTATTATCCCAAGTA

947 0 3 3

BF1 (forward) GACGGTGTATGTGATTTGTCTGAGAGA

654 3 3 3 IMS-PCR 129

BF2 (reverse) ATCCCTAAGATTTTATTATCCCAAGTA

947 0 3 3

GBF-201 GAACCTAAAACGGTATATGT 729a (real start codonat 646)

0 0 0 IMS-PCR 118

GBF-210 GTTGTAGACATCCCACTGGC 1096a (real start codonat 1013)

0 0 0

RS3 (forward outer) TGAAGTTAGTGCCCAGATGCAGG 705 0 1 1 Nested PCR 7, 16, 24, 108RS4 (reverse outer) GCTCAGCGCCCAGTATATGACC 1072 0 5 4RS1 (forward inner) TGCGGCGAACTCGGTTAATGC 729 1 1 1RS2 (reverse inner) AGCTGGGTTGTAGACATCCCA

CTGG1019 0 0 0

P1 (forward) CGCGGAATTCATGTTCTAATGAAGCTGAT

54 0 0 0 PCR 13

P5 (reverse) TGGTCTCGAGATCGCCATCTGCTATTTCC

1191 3 0 0

BFTF (forward) CGCGGCATTATTAGCTGCATGTTCTAATG

36 0 0 0 PCR 30

P4 (reverse) GATACATCAGCTGGGTTGTAGACATCCCA

1027 0 0 0

404, BF3 GAGCCGAAGACGGTGTATGTGATTTGT

646 5 5 5 PCR 86, 113

407, BF4 TGCTCAGCGCCCAGTATATGACCTAGT

1073 0 5 4

GBF-201 GAACCTAAAACGGTATATGT 729a (real start codonat 646)

0 0 0 PCR 39

GBF-210 GTTGTAGACATCCCACTGGC 1096a (real start codonat 1013)

0 0 0

BF5 (forward) GATGCTCCAGTTACAGCTTCCATTG

91 0 1 0 PCR 3

BF6 (reverse) CGCCCAGTATATGACCTAGTTCGTG

1066 0 3 2

Identification of bft isoformsP1 (forward) CGCGGAATTCATGTTCTAATGAA

GCTGAT54 0 0 0 RFLP 13

P5 (reverse) TGGTCTCGAGATCGCCATCTGCTATTTCC

1191 3 0 0

BF5 (forward) GATGCTCCAGTTACAGCTTCCATTG

91 0 1 0 RFLP andPCR

21

BF6 (reverse) CGCCCAGTATATGACCTAGTTCGTG

1066 0 3 2

BFT2R (reverse bft-2) TTGGATCATCCGCATGCCT 1087 5 0 2BFT3R (reverse bft-3) TTGGATCATCCGCATGGTT 1087 5 2 0GBF201 (forward

consensus)GAACCTAAAACGGTATATGT 646 0 0 0 PCR 40

GBF312 (reverse bft-1) CCTCTTTGGCGTCGC 835 0 5 5 118GBF322 (reverse bft-2) CGCTCGGGCAACTAT 820 4 0 3GBF334 (reverse bft-3) TGTCCCAAGTTCCCCAG 931 2 2 0GBF201 (forward

consensus)GAACCTAAAACGGTATATGT 646 0 0 0 PCR 6

BFT-TYPE1 (reverse bft-1) ATTGAACCAGGACATCCCT 960 0 6 5GBF322 (reverse bft-2) CGCTCGGGCAACTAT 820 4 0 3BFT-TYPE3 (reverse bft-3) CGTGTGCCATAACCCCA 931 1 1 0Oligoprobe bft-1 GGCGCTGAGCATACGGATAATT 1063 0 6 6 Hybridization 31Oligoprobe bft-2 GGTGCTAGGCATGCGGATGATC 1063 6 0 2

a The authors labeled the starting nucleotides for their primers as noted. However, the actual starting nucleotides for these primers are nucleotides 646 (for primerlabeled 729) and 1013 (for primer labeled 1096), based on the identified start codons leading to the correct reading frame for the signal peptide of BFT (31).

b The 5 positions in primers P1 and P5 correspond to the nucleotide shown in bold. The underlined sequences in primers P1 and P5 show the restriction sites forEcoRI and XhoI, respectively, created to clone the bft gene.

c RFLP, restriction fragment length polymorphism analysis.d Courtesy of A. Franco-Mora.


on Decem


r.asm.org/

Dow

nloaded from

http://cmr.asm.org/

required for maximal BFT production by ETBF strains, al-though the precise regulatory sequences have not yet beenmapped (30). mpII encodes a protein predicted to be similar insize to BFT and also predicted to be a zinc-dependent, cata-lytically active protein, but with only 56% similarity and 28%identity to the BFT proteins. To date, no in vitro biologicactivity for MPII has been identified, likely due in part to itspoor expression in vitro under growth conditions favoring ex-pression of bft (A. A. Franco, personal communication). Usingan mpII deletion mutant and other recombinant ETBF strains,

the in vitro HT29/C1 cell biologic activity of BFT is not de-pendent on or modified by MPII expression (A. A. Franco andC. L. Sears, unpublished observations). In vivo expression ofmpII or the role of MPII in ETBF disease pathogenesis has notyet been addressed.

The culture supernatants of ETBF strains grown in vitrovary significantly in HT29/C1 cell biologic activity (13, 30, 82,116, 119). Although the mechanism(s) accounting for this vari-ation is incompletely understood, at least two mechanismslikely contribute to the variable biologic activity, including bftcopy number (21) and transcriptional regulation of bft yieldingdifferent amounts of BFT secreted by ETBF strains (30). Incontrast, variation in the biologic activities of purified isoformsof BFT is only modest (13, 121). It is not known if similar

FIG. 3. Schematic of the structure of BFT holotoxin. Each of the BFT isotypes (BFT-1, BFT-2, and BFT-3) consists of three protein domains,i.e., the signal peptide, proprotein, and mature toxin. The holotoxin is cleaved by an as yet unidentified B. fragilis protease at amino acids (AAs)arginine (Arg)211-alanine (Ala)212 prior to release of the mature, �20-kDa BFT protein from the bacterial cells into the colon. H, histidine; G,glycine. (Reprinted from references 103 and 106 with permission from Elsevier.)

TABLE 9. Distribution of bft isotypes among ETBF strains

Reference (author, yr �reference no.�,location) and type of isolate

No. ofETBFstrains

% of strains withallele

bft-1 bft-2 bft-3

Chung et al., 1999 (13), KoreaIntestinal isolates 34 41 22 32

Kato et al., 2000 (40), JapanAsymptomatic control stools 36 72 14 14

Children with diarrheaNon-antibiotic-associated 19 63 26 10.5Antibiotic associated 11 36 54.5 9Cow 16 37.5 50 12.5Extraintestinal isolates 61 69 18 13

Total 143 63 24 13D’Abusco et al., 2000 (21), Italy

Extraintestinal isolates 28 61 36 4Fecal isolatesa 40 70 27.5 2.5

Nguyen et al., 2005 (76), Vietnam 49b 69 18 12Ulger et al., 2006 (114), Turkey 54c 91d 9 0Nakano et al., 2007 (72), Brazil 3e 100 0 0Avila-Campos et al. (6), United States 44f 68 25 7

a Primarily from children (n � 18) and adults (n � 10) with diarrhea; bft-2 wasidentified significantly more often in fecal ETBF isolates from children than inthose from adults.

b Primarily from children with diarrhea; no ETBF strains with the bft-3 genewere identified among the six control children positive for ETBF.

c Thirty-nine fecal ETBF strains (31 from colon cancer patients) and 15 ex-traintestinal ETBF strains.

d Eighty-seven percent, 88%, and 100% of ETBF isolates from colon cancerpatients, control patient feces, and extraintestinal sites, respectively, containedthe bft-1 gene.

e Fecal ETBF was isolated over 8 months from three children, one with AIDSand two immunocompetent children with acute diarrhea.

f Four blood, 25 intra-abdominal, 4 pulmonary, 5 skin and soft tissue, and 6miscellaneous ETBF isolates.

FIG. 4. Schematic of the molecular types of B. fragilis. Pattern I B.fragilis strains are ETBF strains possessing at least one 65-kb conjuga-tive transposon (86 CTn), within which is contained the 6-kb BfPAI.The BfPAI contains two genes, one encoding BFT (bft), demonstratedto be important to ETBF pathogenesis (92), and one encoding metal-loprotease II (mpII), a putative virulence protein. Pattern II B. fragilisstrains lack CTn86 and CTn9343 (or related sequences). Pattern III B.fragilis strains are NTBF strains that possess at least one 65-kb conju-gative transposon (9343 CTn). See the text and Fig. 5 for additionaldetails. (Reprinted from reference 103 with permission from Elsevier.)


on Decem


r.asm.org/

Dow

nloaded from

http://cmr.asm.org/

differences in BFT expression occur in vivo and/or impactclinical disease expression.

Both pattern II and pattern III B. fragilis strains are NTBF(29). Using the B. fragilis genome database, the DNA elementpresent in pattern III NTBF strains and a related elementpresent in all ETBF strains have been identified as members ofa new family of putative conjugative transposons, termedCTn9343 (from NTBF strain NCTC 9343) and CTn86 (fromETBF strain 86-5443-2-2) (27). Pattern II NTBF strains aredistinguished by the absence of either of these CTns. In ananalysis of 191 NTBF strains, 52% of NTBF strains were pat-tern II strains and 43% were pattern III strains. A small num-ber of NTBF strains (5%) displayed other genomic patterns(29). CTn9343 and CTn86 are approximately 64 kb, with openreading frames organized as modules with various G�C con-tents, suggesting regions where the genes may be of Bacteroidesorigin and regions acquired from other genetic sources. Theseputative CTns have very limited sequence similarity to otherdescribed Bacteroides species CTns and are distinct in severalproperties, including their predicted mechanism of transposi-tion, their lack of tetracycline regulation of CTn chromosomalexcision, and the absence of tetQ (27). Although circularizationof the CTns within ETBF and NTBF strains is readily identi-fied, indicating that the CTns can excise from the B. fragilischromosomes, transfer of CTn86 or CTn9343 to other organ-isms has not yet been observed (27; Franco, personal commu-nication). However, it is postulated that interorganism transferof these putative CTns with further acquisition of the BfPAImay be the mechanism by which ETBF strains evolved as adistinct class of B. fragilis strains (27). Additional data suggestthat additional CTns, related to but distinct from CTn86 andCTn9343, are present in both ETBF and NTBF (9) (Fig. 5). Itis unknown how or if these putative CTns modulate B. fragilisvirulence. Similarly, it is unknown what biologic advantage, ifany, CTn86 or the BfPAI confers upon ETBF strains. How-ever, mobilization of a plasmid containing bft plus its promoterregion into pattern III, but not pattern II, NTBF is efficient andyields high-level expression of BFT, indicating the differinggenetic potentials of these two populations of NTBF strainsand also suggesting that CTn9343 or another chromosomallocus unique to pattern III NTBF (and absent in pattern IINTBF) regulates BFT production (30).

By use of a variety of molecular techniques, B. fragilis strainshave also been classified into two phylogenetic divisions (32,34). Division I is characterized by the presence of the cepAgene (encoding a serine--lactamase of class A) and the ab-sence of the cfiA gene (encoding a metallo--lactamase of classB, conferring, for example, imipenem resistance), whereas di-vision II is characterized by the presence of cfiA and the ab-sence of the cepA gene. All ETBF strains, to date, are divisionI B. fragilis, as are ca. 80% of B. fragilis strains isolated inclinical studies (9, 32). Multilocus enzyme electrophoresis andcluster analysis indicate the ETBF strains are nonclonal, con-sistent with the higher recombination rates ascribed to divisionI B. fragilis (32).

Phylogenetic data indicate that B. fragilis strains are diverse,and functional studies and sequences of the genomes of twoNTBF strains identified DNA inversion regulatory mecha-nisms, suggesting that these organisms are highly adaptable,with rapid and dynamic variability in surface molecule expres-

sion patterns (11, 20, 49). B. fragilis can express up to at leasteight distinct capsular polysaccharides, a previously unprece-dented complexity for a single organism (18). It is unknown ifthe enteric pathogenicity of ETBF is modulated by expressionof specific surface characteristics (including capsular polysac-charides), for example, influencing adherence of ETBF to theintestinal mucosa and/or delivery of BFT in vivo.

BFT MECHANISM OF ACTION

In Vitro Studies of Cell Lines, Polarized EpithelialMonolayers, and Human Colon

To date, the in vitro biologic activity of BFT has been re-stricted to continuous epithelial cell lines capable of formingpolarized monolayers (119, 126). Predominantly intestinal celllines (HT29, HT29/C1, Caco-2, T84, SW480, and HCT116)

FIG. 5. Schematic representations of CTn elements found in ETBFand NTBF strains. Both ETBF and NTBF strains may possess a varietyof CTns related to those originally described for ETBF strain 86-5443-2-2 (CTn86) and NTBF strain 9343 (CTn9343) (27). Panels A and Bshow different patterns of CTns present in a collection of 123 ETBFand 73 NTBF strains, respectively. Gray boxes represent the left end ofCTn86, and black boxes represent the left end of CTn9343. (Reprintedfrom reference 9.)


on Decem


r.asm.org/

Dow

nloaded from

http://cmr.asm.org/

have been studied (78, 123, 126; S. Wu and C. L. Sears, un-published observations). More limited studies indicate thatrenal (MDCK) and pulmonary (Calu-3) cell lines that canpolarize in vitro also develop morphological changes and thusexhibit a biologic response to BFT (78; Wu and Sears, unpub-lished observations). Overall, subconfluent cloned HT29/C1cells have been studied in greatest detail due to their exquisiteconcentration-dependent sensitivity to BFT (63, 97). The half-maximal BFT concentration altering HT29/C1 cell morphologyis ca. 12.5 pM, with the onset of activity at 0.5 pM (97). Usingsubconfluent HT29/C1 cells, the hallmarks of the BFT biologicresponse are a rapid onset (within 10 to 15 min) and temper-ature-dependent (maximal activity detected at 37°C) morpho-logical changes in which cell rounding and swelling with loss ofcell-to-cell contact occur (Fig. 2) (61, 63, 97, 119, 122). Thesechanges are reminiscent of the morphological changes in thesurface epithelial cells of intestines infected with ETBF (Fig.1). Although the total F (filamentous)-actin content of cellstreated with BFT is unaltered (22, 47, 96, 97), marked redis-tribution of F-actin occurs, with decreased stress fibers andperipheral F-actin condensation observed in unpolarizedHT29 cells treated with BFT (22, 47). The mechanism(s) re-sponsible for the cell morphology changes stimulated by BFTis unknown, although a broad-spectrum tyrosine kinase inhib-itor was noted to delay the onset of BFT-induced cell mor-phology changes (124). Inhibitors of microtubules and endo-somal or Golgi trafficking do not alter induction of cellmorphology changes by BFT (22, 78, 97).

Although the bioactivity of BFT is not reversible by washingearly on, subconfluent HT29/C1 cells recover a normal appear-ance by light microscopy within 2 to 3 days after BFT treat-ment, indicating that the biologic activity of BFT on HT29/C1cells is reversible over time (97, 119, 122). The majority ofavailable evidence indicates that BFT is a nonlethal and non-cytotoxic protein, namely, BFT stimulates rather than dimin-ishes protein synthesis (12, 47), it does not stimulate cellularlactate dehydrogenase (LDH) or 51Cr release or the cellularuptake of vital dyes (trypan blue and propidium iodide) (12,47, 78, 120), and DNA synthesis continues normally (22). BFTtreatment of polarized colonic epithelial cell (HT29 and T84)monolayers in vitro resulted in delayed (�36 h after BFTtreatment) apoptosis of a minority of treated cells, althoughthe initial response to BFT was in fact shown to be inductionof an antiapoptotic protein (cellular inhibitor of apoptosis pro-tein 2 [cIAP2]) (43; see “Molecular Mechanism of Action ofBFT”). After BFT treatment of T84 cells or human colonbiopsies in vitro, a delayed loss of cell viability and apoptosis ofdetached (and thus dying) epithelial cells were noted (99, 100).Of potential importance to human ETBF disease, heterogene-ity was noted in the rate of onset (2 to 18 h) and severity ofepithelial morphological changes in human colon biopsiestreated with BFT, with no BFT response noted in 20% of theindividuals sampled (99).

BFT stimulates a concentration- and time-dependent in-crease in the permeability of epithelial monolayers (T84,MDCK, HT29, HT29/C1, and Caco-2 cells) (12, 78, 105). No-tably, BFT exhibits polar monolayer bioactivity, increasing per-meability rapidly and at lower toxin concentrations whenplaced on the basolateral membranes rather than the apicalmembranes of epithelial monolayers or human colons studied

in vitro (12, 78, 93, 105, 120; Wu and Sears, unpublishedobservations). In addition, basolateral but not apical BFT canrapidly but transiently increase the short-circuit current (indic-ative of chloride secretion) (12). BFT also stimulates predom-inantly basolateral release of proinflammatory CXC chemo-kines, consistent with a role for BFT in inducing mucosalinflammation (45, 124; see “Molecular Mechanism of Actionof BFT”).

The changes in T84 cell monolayer permeability are accom-panied by cellular morphological changes, with apical BFTcausing only relatively slow (with onset at 6 h), focal changeson the apical epithelial cell membranes. In contrast, basolat-eral BFT more rapidly (by 90 min) modifies the morphology ofevery cell in the monolayer, with cell swelling and unraveling ofthe apical membrane microvilli resulting in a striking domedcellular appearance (12, 47). The disappearance of the mi-crovilli is associated with a loss of F-actin from the apical poleof the cells, with a marked reassembly of the F-actin at thebasolateral pole of the cells (12). Concomitantly, as detailed bytransmission electron microscopy, BFT stimulates structuralchanges and even dissolution of the zonula occludens (tightjunction) and zonula adherens, electron-dense structures thatregulate the permeability of epithelial monolayers, whereasdesmosomes remain intact (12, 105). The redistribution ofF-actin and loss of the zonula occludens and zonula adherensreadily explain the measured increase in monolayer permeabil-ity, although full mechanistic details are not available. BFT-induced colon permeability may further expose the submucosato other bacterial luminal antigens and thus contribute to howETBF fosters colonic inflammation (7, 37, 88, 90, 113). Wellsand colleagues further reported that the basolateral mem-branes of BFT-treated HT29 cells permitted increased associ-ation and invasion of pathogenic enteric bacteria, except forListeria monocytogenes (120). BFT-treated HT29 cells arelikely resistant to L. monocytogenes invasion due to BFT-in-duced loss of cellular E-cadherin (see “Molecular Mechanismof Action of BFT”), one ligand for invasion of this bacterium(59). Lastly, BFT, presumably by modifying mucosal perme-ability, has been reported to act as a mucosal adjuvant, en-hancing the systemic antibody response to an intranasal anti-gen challenge in mice (117).

Molecular Mechanism of Action of BFT

To date, the only cellular protein demonstrated to be rapidlycleaved after treatment of colonic epithelial cells with BFT isthe zonula adherens protein E-cadherin (122, 125). In subcon-fluent HT29/C1 cells, the onset of E-cadherin cleavage is rapid,as it is detectable by 1 min and typically complete within 1 to2 h. BFT induces E-cadherin cleavage via two steps, the initialrelease of the E-cadherin ectodomain, which is dependent onbiologically active BFT, and subsequent processing of the in-tracellular E-cadherin fragment by host cell presenilin-1/�-secretase (a member of the intramembrane cleavage protease[iCLips] family) (125). Only E-cadherin presented on an intact,living cell is cleaved in response to BFT (i.e., in vitro cleavageof E-cadherin cannot be demonstrated), and this E-cadherincleavage requires cellular ATP, suggesting that protein confor-mation and/or other cellular properties contribute to the pro-teolytic event (122). Cellular recovery after BFT treatment


on Decem


r.asm.org/

Dow

nloaded from

http://cmr.asm.org/

correlates with resynthesis of E-cadherin. Cleavage of E-cad-herin also occurs in vivo in a murine model of ETBF diseaseand correlates with the onset of colonic inflammation anddisruption of the epithelial barrier by histopathology (92).

Besides its critical role in intercellular adhesion at thezonula adherens in intestinal tissue, E-cadherin is tethered atits intracellular domain to -catenin, a nuclear signaling pro-tein involved in both normal and dysregulated cellular growth(74). Proteolysis of E-cadherin in response to BFT induces-catenin nuclear localization, upregulation of c-myc (a -cate-nin-regulated oncogene) transcription and translation, and cel-lular proliferation of HT29/C1 cells (123). HT29/C1 cell pro-liferation is stimulated by as little as 0.5 nM BFT (123).Further studies suggest that -catenin signaling accounts foronly ca. 30 to 40% of BFT-induced cellular proliferation, in-dicating that other, as yet unidentified mechanisms contributeto BFT-initiated cell proliferation (123). BFT-treated HT29/C1cells are highly mobile, consistent with observations indicatingthat diminished E-cadherin on tumor cells enhances metastaticpotential (Wu and Sears, unpublished observations). In paral-lel, BFT also stimulates induction of the antiapoptotic proteincIAP2, mediated by p38 mitogen-activated protein kinase (p38MAPK) and cyclooxygenase-2 (COX-2) signaling with produc-tion of prostaglandin E2 (PGE2) (43). BFT-induced colonocyteproliferation and mobility, combined with resistance to apop-tosis, may contribute to the putative oncogenic potential ofBFT and support an initial report of an association betweenETBF colonization and colorectal carcinoma (113). An alter-native hypothesis is that antiapoptotic signaling by BFT in-creases the colonocyte life span, permitting further generationof proinflammatory signaling to the colonic submucosa (43).

Reported in vitro proteolytic substrates of BFT include G(monomeric)-actin, gelatin, azocoll, tropomyosin, collagen IV,human complement C3, and fibrinogen. However, no biologicsignificance of these substrates has yet been identified withregard to the cellular or intestinal mechanism of action of BFT(61, 79, 96). Furthermore, BFT induces shedding of a broadarray of HT29/C1 membrane proteins over time (125). How-ever, the identities of these shed proteins and whether they arecleaved by BFT or specifically due to BFT-induced cell signal-ing are unknown. Of other proteins specifically tested, BFTdoes not cleave the cell surface proteins occludin or claudin 1and 2 (both present in the zonula occludens) and 1-integrin (abasal intestinal epithelial cell protein) (122; Wu and Sears,unpublished observations) or the intracellular proteins -cate-nin, �-catenin, zonula occludens protein-1 (ZO-1), and actin(96, 122). Although these proteins are not BFT substrates,ZO-1, -catenin, and occludin, for example, redistribute incells by 1 hour after treatment with BFT (78, 122).

ETBF induces intestinal inflammation in animals and hu-mans (80, 102, 105). In addition to increased colonic perme-ability that may augment submucosal inflammation throughexposure to luminal bacterial antigens, ETBF likely inducesintestinal inflammation via BFT-stimulated intestinal epithelialcell release of proinflammatory chemokines, including IL-8(also called CXCL8 or [the murine equivalent] macrophageinflammatory protein 2 [MIP-2]), epithelial-neutrophil activat-ing protein 78 (ENA-78), CCL2, monocyte chemotactic pro-tein 1 (MCP-1), and growth-related oncogene alpha (GRO-�;also called CXCL1) (41, 42, 45, 100, 124). The time course of

chemokine release varies, with IL-8 appearing first (at 2 to 4 h)and GRO-� and ENA-78 appearing later (45, 124). Similarly,BFT also stimulates increased synthesis of these chemokinesfrom isolated human colonocytes (45). BFT also induces in-creased expression of the biologically inactive form of trans-forming growth factor beta (TGF-) by intestinal epithelialcells in vitro (100). It is postulated that inactive TGF- issubsequently processed by proteases in the intestinal mucosato active TGF- and contributes to mucosal repair in ETBFdisease. In contrast, no data have yet identified direct effects ofBFT on immune cells or myofibroblasts (92, 100).

NF- B appears to act as a central regulator of chemokineexpression in BFT-stimulated colonic epithelial cells in vitro(41, 45, 124), and NF- B activation in BFT-treated isolatedhuman colonocytes has been demonstrated (41). BFT inducesan unusual discrete supranuclear localization of NF- B, a find-ing also previously reported in response to IL-1 (36, 124). Theregulation of NF- B activation stimulated by BFT is complex,involving receptor and nonreceptor tyrosine kinases, MAPKs(p38, extracellular signal-related kinase, and c-Jun N-terminalkinase), Ras, and AP-1 (42, 124). NF- B is also reported tomediate BFT-induced colonic epithelial cell expression (in-cluding in human colonocytes) of COX-2 but not COX-1,resulting in increased levels of cellular PGE2 and cyclic AMP(cAMP).

Initial experiments implicate BFT-initiated NF- B activa-tion in colonic epithelial cells as the orchestrator of inflamma-tion and secretion in ETBF disease. First, inhibition of NF- Bactivation diminished BFT-induced chemokine release andpolymorphonuclear leukocyte transepithelial migration in co-lonic epithelial monolayers in vitro, suggesting that NF- Bsignaling was directly linked to polymorphonuclear leukocytemucosal influx in ETBF disease (41). Second, both p38 MAPKand COX-2 inhibition significantly decreased secretion inmouse ileal loops treated with BFT, implicating NF- B acti-vation by BFT as a key cellular coordinator of secretion inETBF disease. Inhibition of p38 MAPK activation also re-versed BFT-induced ileal inflammation in the mouse (42). Theeffect of COX-2 inhibition on murine ileal histology was notreported, although COX-2 inhibition did not diminish MIP-2production, suggesting at least some differential regulation ofsecretion and inflammation in response to ETBF infection(44).

PROPOSED MODEL FOR PATHOGENESIS OFETBF DISEASE

Figure 6 proposes a model for the pathogenesis of ETBFdisease. Available data suggest that among Bacteroides species,B. fragilis seeks a mucosal niche aided by its decoration withfucosylated molecules mimicking host proteins (19, 73). Al-though ETBF adherence in vivo or in vitro has not yet beenexamined, the pathogenesis of ETBF intestinal disease is ex-pected to be initiated by adherence of the organism to thecolonic mucosa, with local delivery of BFT. The histology ofETBF disease in animals has not identified adherent organ-isms, but available results are limited by formalin fixation,which can interfere with detection of mucosal adherence in thecolon (111). Given the in vitro potency of BFT in cell assays(97) and the limited quantities of BFT secreted into cultures in


on Decem


r.asm.org/

Dow

nloaded from

http://cmr.asm.org/

vitro (121), it seems probable that small amounts of BFTdelivered by adherent ETBF cells to the colonic epitheliummay be sufficient to modify colonic epithelial cell structure andfunction. Consistent with this postulate, recent data indicatethat bft expression is necessary and sufficient to induce colitis inmurine models (92). No quantitative data have yet correlatedETBF colonization or disease with levels of BFT expressed by

ETBF strains. It is predicted that BFT promotes disease bybinding to apical membrane receptors on colonic epithelialcells (CECs), initiating a burst of complex signal transductionresulting in rapid E-cadherin cleavage (122, 125, 126). E-cad-herin cleavage releases -catenin associated with the cytoplas-mic domain of E-cadherin. -Catenin nuclear translocationalong with activation of tyrosine kinases, MAPKs, and NF- B

FIG. 6. Model of ETBF colitis pathogenesis. ETBF colonizes the colon, where BFT is released and attaches to a specific colonic epithelial cell(CEC) receptor, triggering complex (and incompletely understood) CEC signal transduction involving -catenin, tyrosine kinases (TK), mitogen-activated protein kinases (MAPK), and NF- B. CEC signal transduction results in the cleavage of E-cadherin as well as in new CEC proteinsynthesis, with increased expression of c-Myc, cyclooxygenase-2 (COX-2), and chemokines/cytokines, including IL-8 and TGF-. E-cadherincleavage initiates decreased barrier function of the colonic mucosa, with the potential for increased exposure of the mucosal immune system toantigens of ETBF as well as the colonic flora fostering an inflammatory mucosal response. c-Myc expression stimulates CEC proliferation, at leastin part. Release of chemokines/cytokines by CECs into the submucosa enhances mucosal inflammation in response to ETBF colonization. Theprecise contributions of different mucosal immune cells to the inflammatory response to ETBF colon colonization are unknown, but data suggestthat both polymorphonuclear leukocytes and lymphocytes are important (41; S. Wu and C. L. Sears, submitted for publication). DC, dendritic cell;M�, macrophage; PMNs, polymorphonuclear leukocytes.


on Decem


r.asm.org/

Dow

nloaded from

http://cmr.asm.org/

results in nuclear signaling with new gene transcription (41, 42,100, 124). E-cadherin cleavage, along with cellular F-actin re-arrangement, further promotes colonic permeability (93) andaccess of innate mucosal immune cells to luminal bacterialantigens. This likely promotes mucosal inflammatory and se-cretory responses that are augmented by BFT-induced CECchemokine expression and PGE2 synthesis. Induction of c-Mycsynthesis further induces CEC proliferation (123). No datahave yet identified a role for BFT in pathogenesis beyond itsdirect CEC actions. The precise timing and order of the sig-naling cascades initiated by BFT remain to be deciphered, asdo the details of how or which specific signal transductionmechanisms contribute to the cell morphology changes, E-cadherin cleavage, new gene expression, and CEC prolifera-tion stimulated by BFT. Together, the data support the con-cept that ETBF strains, through secretion of BFT, areproinflammatory and oncogenic bacteria, at least in some hosts(99). Limited clinical observations are consistent with this con-cept (7, 88, 102, 113).

SUMMARY AND FUTURE CHALLENGES

ETBF organisms are common human colonic symbioteswhose potential to cause human disease is incompletely under-stood. In murine models, ETBF induces acute, self-limited,symptomatic colitis that transitions to long-term carriagewhere the murine host constrains but does not eliminate ETBFcolitis (92). Well-designed human investigations are needed toassess whether this paradigm occurs in humans and to deter-mine the impact of long-term carriage of ETBF on colonicstructure, function, and disease incidence. Based on availableclinical and experimental data, conditions where ETBF maycontribute to disease include IBD, colorectal cancer, and po-tentially other enigmatic conditions where colonic inflamma-tion may be pathogenic, such as necrotizing enterocolitis inneonates, postinfectious irritable bowel syndrome, antibiotic-associated diarrhea, and even perhaps malnutrition in childrenin the developing world. Possible mechanisms regulating anddefining the outcome of the host-ETBF interaction includedifferences in ETBF strain virulence, genetically determinedhost differences in adhesion, immune, or inflammatory re-sponses to ETBF, and/or modulation of ETBF virulence by theother host colonic flora. Studies to detect ETBF in patientpopulations should employ both sensitive molecular microbi-ologic fecal diagnosis and serology to detect anti-BFT antibod-ies that develop, at least for patients with acute ETBF diar-rheal disease (102). Whether anti-BFT antibodies are useful tofurther evaluate the epidemiology of ETBF human infection orcolonization is as yet unknown. There is surging interest in theimpact of symbiont bacteria, particularly the colonic flora, onnormal host physiology, immunology, and disease. AmongBacteroides species, the molecularly diverse B. fragilis strainsare distinguished by being critical symbionts and importanthuman pathogens. Future studies of the colonic flora, includ-ing ETBF, hold promise for illuminating the mechanisms gov-erning the essential yet sometimes pathogenic relationship be-tween flora and host.

ACKNOWLEDGMENTS

I thank R. Bradley Sack, Shaoguang Wu, and Augusto Franco-Morafor reviewing the manuscript and providing their insights. I also thankA. Franco-Mora for providing Table 8 in the manuscript and all thecontributors over time to studies from my laboratory.

I have no conflicts of interest to report.This work was supported by grants RO1 DK45496 and RO1

DK080817 and by a Senior Investigator Award from the Crohn’s andColitis Foundation.

REFERENCES

1. Reference deleted.2. Albert, M. J., A. S. Faruque, S. M. Faruque, R. B. Sack, and D. Ma-

halanabis. 1999. Case-control study of enteropathogens associated withchildhood diarrhea in Dhaka, Bangladesh. J. Clin. Microbiol. 37:3458–3464.

3. Antunes, E. N., E. O. Ferreira, L. S. Falcao, G. R. Paula, K. E. Avelar, D. E.Barroso, J. P. Leite, M. C. Ferreira, and R. M. Domingues. 2004. Non-toxigenic pattern II and III Bacteroides fragilis strains: coexistence in thesame host. Res. Microbiol. 155:522–524.

4. Antunes, E. N., E. Ferreira, D. Vallim, G. Paula, L. Seldin, A. Sabra, M.Gerreira, and R. Domingues. 2002. Pattern III non-toxigenic Bacteroidesfragilis (NTBF) strains in Brazil. Anaerobe 8:17–22.

5. Aucher, P., J. P. Saunier, G. Grollier, M. Sebald, and J. L. Fauchere. 1996.Meningitis due to enterotoxigenic Bacteroides fragilis. Eur. J. Clin. Micro-biol. Infect. Dis. 15:820–823.

6. Avila-Campos, M. J., C. Liu, Y. Song, M. C. Rowlinson, and S. M. Finegold.2007. Determination of bft gene subtypes in Bacteroides fragilis clinicalisolates. J. Clin. Microbiol. 45:1336–1338.

7. Basset, C., J. Holton, A. Bazeos, D. Vaira, and S. Bloom. 2004. Are Heli-cobacter species and enterotoxigenic Bacteroides fragilis involved in inflam-matory bowel disease? Dig. Dis. Sci. 49:1425–1432.

8. Bressane, M., L. Durigon, and M. Avila-Campos. 2001. Prevalence of theBacteroides fragilis group and enterotoxigenic Bacteroides fragilis in immu-nodeficient children. Anaerobe 7:277–281.

9. Buckwold, S. L., N. B. Shoemaker, C. L. Sears, and A. A. Franco. 2007.Identification and characterization of conjugative transposons CTn86 andCTn9343 in Bacteroides fragilis strains. Appl. Environ. Microbiol. 73:53–63.

10. Caceres, M., G. Zhang, A. Weintraub, and C.-E. Nord. 2000. Prevalenceand antimicrobial susceptibility of enterotoxigenic Bacteroides fragilis inchildren with diarrhoea in Nicaragua. Anaerobe 6:143–148.

10a.Centers for Disease Control and Prevention. 2008. Preliminary FoodNetdata on the incidence of infection with pathogens transmitted commonlythrough food—10 states, 2007. MMWR Morb. Mortal. Wkly. Rep. 57:366–370.

11. Cerdeno-Tarraga, A. M., S. Patrick, L. C. Crossman, G. Blakely, V. Abratt,N. Lennard, I. Poxton, B. Duerden, B. Harris, M. A. Quail, A. Barron, L.Clark, C. Corton, J. Doggett, M. T. Holden, N. Larke, A. Line, A. Lord, H.Norbertczak, D. Ormond, C. Price, E. Rabbinowitsch, J. Woodward, B.Barrell, and J. Parkhill. 2005. Extensive DNA inversions in the B. fragilisgenome control variable gene expression. Science 307:1463–1465.

12. Chambers, F. G., S. S. Koshy, R. F. Saidi, D. P. Clark, R. D. Moore, andC. L. Sears. 1997. Bacteroides fragilis toxin exhibits polar activity on mono-layers of human intestinal epithelial cells (T84 cells) in vitro. Infect. Immun.65:3561–3570.

13. Chung, G. T., A. A. Franco, S. Wu, G. E. Rhie, R. Cheng, H. B. Oh, and C. L.Sears. 1999. Identification of a third metalloprotease toxin gene in extraint-estinal isolates of Bacteroides fragilis. Infect. Immun. 67:4945–4949.

14. Claros, M. C., Z. C. Claros, Y. J. Tang, S. H. Cohen, J. Silva, Jr., E. J.Goldstein, and A. C. Rodloff. 2000. Occurrence of Bacteroides fragilis en-terotoxin gene-carrying strains in Germany and the United States. J. Clin.Microbiol. 38:1996–1997.

15. Claros, M. C., Z. Claros, D. Hecht, D. Citron, E. Goldstein, J. Silva, Jr., Y.Tang-Feldman, and A. Rodloff. 2006. Characterization of the Bacteroidesfragilis pathogenicity island in human blood culture isolates. Anaerobe12:17–22.

16. Cohen, S. H., R. Shetab, Y. J. Tang-Feldman, P. Sarma, J. Silva, Jr., andT. P. Prindiville. 2006. Prevalence of enterotoxigenic Bacteroides fragilis inhospital-acquired diarrhea. Diagn. Microbiol. Infect. Dis. 55:251–254.

17. Collins, J. H., M. E. Bergeland, L. L. Myers, and D. S. Shoop. 1989.Exfoliating colitis associated with enterotoxigenic Bacteroides fragilis in apiglet. J. Vet. Diagn. Investig. 1:349–351.

18. Comstock, L. E., and D. L. Kasper. 2006. Bacterial glycans: key mediatorsof diverse host immune responses. Cell 126:847–850.

19. Coyne, M. J., B. Reinap, M. M. Lee, and L. E. Comstock. 2005. Humansymbionts use a host-like pathway for surface fucosylation. Science 307:1778–1781.

20. Coyne, M. J., K. G. Weinacht, C. M. Krinos, and L. E. Comstock. 2003. Mpirecombinase globally modulates the surface architecture of a human com-mensal bacterium. Proc. Natl. Acad. Sci. USA 100:10446–10451.

21. d’Abusco, A. S., M. Del Grosso, S. Censini, A. Covacci, and A. Pantosti.


on Decem


r.asm.org/

Dow

nloaded from

http://cmr.asm.org/

2000. The alleles of the bft gene are distributed differently among entero-toxigenic Bacteroides fragilis strains from human sources and can be presentin double copies. J. Clin. Microbiol. 38:607–612.

22. Donelli, G., A. Fabbri, and C. Fiorentini. 1996. Bacteroides fragilis entero-toxin induces cytoskeletal changes and surface blebbing in HT-29 cells.Infect. Immun. 64:113–119.

23. Duimstra, J. R., L. L. Myers, J. E. Collins, D. A. Benfield, D. S. Shoop, andW. C. Bradbury. 1991. Enterovirulence of enterotoxigenic Bacteroides fra-gilis in gnotobiotic pigs. Vet. Pathol. 28:514–518.

24. Durmaz, B., M. Dalgalar, and R. Durmaz. 2005. Prevalence of enterotoxi-genic Bacteroides fragilis in patients with diarrhea: a controlled study.Anaerobe 11:318–321.

25. Eckburg, P. B., E. M. Bik, C. N. Bernstein, E. Purdom, L. Dethlefsen, M.Sargent, S. R. Gill, K. E. Nelson, and D. A. Relman. 2005. Diversity of thehuman intestinal microbial flora. Science 308:1635–1638.

26. Foulon, I., D. Pierard, G. Muyldermans, K. Vandoorslaer, O. Soetens, P.Rosseel, and S. Lauwers. 2003. Prevalence of fragilysin gene in Bacteroidesfragilis isolates from blood and other extraintestinal samples. J. Clin. Mi-crobiol. 41:4428–4430.

27. Franco, A. A. 2004. The Bacteroides fragilis pathogenicity island is containedin a putative novel conjugative transposon. J. Bacteriol. 186:6077–6092.

28. Franco, A. A., S. Buckwold, J. W. Shin, M. Ascon, and C. L. Sears. 2005.Mutation of the zinc-binding metalloprotease motif affects Bacteroides fra-gilis toxin activity without affecting propeptide processing. Infect. Immun.73:5273–5277.

29. Franco, A. A., R. K. Cheng, G. T. Chung, S. Wu, H. B. Oh, and C. L. Sears.1999. Molecular evolution of the pathogenicity island of enterotoxigenicBacteroides fragilis strains. J. Bacteriol. 181:6623–6633.

30. Franco, A. A., R. K. Cheng, A. Goodman, and C. L. Sears. 2002. Modulationof bft expression by the Bacteroides fragilis pathogenicity island and itsflanking region. Mol. Microbiol. 45:1067–1077.

31. Franco, A. A., L. M. Mundy, M. Trucksis, S. Wu, J. B. Kaper, and C. L.Sears. 1997. Cloning and characterization of the Bacteroides fragilis metal-loprotease toxin gene. Infect. Immun. 65:1007–1013.

32. Gutacker, M., C. Valsangiacomo, and J. C. Piffaretti. 2000. Identification oftwo genetic groups in Bacteroides fragilis by multilocus enzyme electro-phoresis: distribution of antibiotic resistance (cfiA, cepA) and enterotoxin(bft) encoding genes. Microbiology 146:1241–1254.

33. Hacker, J., G. Blum-Oehler, I. Muhldorfer, and H. Tschape. 1997. Patho-genicity islands of virulent bacteria: structure, function and impact onmicrobial evolution. Mol. Microbiol. 23:1089–1097.

34. Holton, J. 2008. Enterotoxigenic Bacteroides fragilis. Curr. Infect. Dis. Rep.10:99–104.

35. Hooper, L. V., and J. I. Gordon. 2001. Commensal host-bacterial relation-ships in the gut. Science 292:1115–1118.

36. Jobin, C., S. Haskill, L. Mayer, A. Panja, and R. B. Sartor. 1997. Evidencefor altered regulation of I kappa B alpha degradation in human colonicepithelial cells. J. Immunol. 158:226–234.

37. Karin, M., and F. R. Greten. 2005. NF-kappaB: linking inflammation andimmunity to cancer development and progression. Nat. Rev. Immunol.5:749–759.

38. Kato, N., H. Kato, K. Watanabe, and K. Ueno. 1996. Association of ente-rotoxigenic Bacteroides fragilis with bacteremia. Clin. Infect. Dis. 23(Suppl.1):S83–S86.

39. Kato, N., C. Liu, H. Kato, K. Watanabe, H. Nakamura, N. Iwai, and K.Ueno. 1999. Prevalence of enterotoxigenic Bacteroides fragilis in childrenwith diarrhea in Japan. J. Clin. Microbiol. 37:801–803.

40. Kato, N., C. X. Liu, H. Kato, K. Watanabe, Y. Tanaka, T. Yamamoto, K.Suzuki, and K. Ueno. 2000. A new subtype of the metalloprotease toxingene and the incidence of the three bft subtypes among Bacteroides fragilisisolates in Japan. FEMS Microbiol. Lett. 182:171–176.

41. Kim, J. M., S. J. Cho, Y. K. Oh, H. Y. Jung, Y. J. Kim, and N. Kim. 2002.Nuclear factor-kappa B activation pathway in intestinal epithelial cells is amajor regulator of chemokine gene expression and neutrophil migrationinduced by Bacteroides fragilis enterotoxin. Clin. Exp. Immunol. 130:59–66.

42. Kim, J. M., H. Y. Jung, J. Y. Lee, J. Youn, C. H. Lee, and K. H. Kim. 2005.Mitogen-activated protein kinase and activator protein-1 dependent signalsare essential for Bacteroides fragilis enterotoxin-induced enteritis. Eur.J. Immunol. 35:2648–2657.

43. Kim, J. M., J. Y. Lee, and Y. J. Kim. 2008. Inhibition of apoptosis inBacteroides fragilis enterotoxin-stimulated intestinal epithelial cells throughthe induction of c-IAP-2. Eur. J. Immunol. 38:2190–2199.

44. Kim, J. M., J. Y. Lee, Y. M. Yoon, Y. K. Oh, J. S. Kang, Y. J. Kim, and K. H.Kim. 2006. Bacteroides fragilis enterotoxin induces cyclooxygenase-2 andfluid secretion in intestinal epithelial cells through NF-kappaB activation.Eur. J. Immunol. 36:2446–2456.

45. Kim, J. M., Y. K. Oh, Y. J. Kim, H. B. Oh, and Y. J. Cho. 2001. Polarizedsecretion of CXC chemokines by human intestinal epithelial cells in re-sponse to Bacteroides fragilis enterotoxin: NF-kappa B plays a major role inthe regulation of IL-8 expression. Clin. Exp. Immunol. 123:421–427.

46. Kling, J. J., R. L. Wright, J. S. Moncrief, and T. D. Wilkens. 1997. Cloning

and characterization of the gene for the metalloprotease enterotoxin ofBacteroides fragilis. FEMS Microbiol. Lett. 146:279–284.

47. Koshy, S. S., M. H. Montrose, and C. L. Sears. 1996. Human intestinalepithelial cells swell and demonstrate actin rearrangement in response tothe metalloprotease toxin of Bacteroides fragilis. Infect. Immun. 64:5022–5028.

48. Krzyzanowsky, F., and M. Avila-Campos. 2003. Detection of non-entero-toxigenic and enterotoxigenic Bacteroides fragilis in stool samples fromchildren in Sao Paulo, Brazil. Rev. Inst. Med. Trop. Sao Paulo 45:225–227.

49. Kuwahara, T., A. Yamashita, H. Hirakawa, H. Nakayama, H. Toh, N.Okada, S. Kuhara, M. Hattori, T. Hayashi, and Y. Ohnishi. 2004. Genomicanalysis of Bacteroides fragilis reveals extensive DNA inversions regulatingcell surface adaptation. Proc. Natl. Acad. Sci. USA 101:14919–14924.

50. Lassmann, B., D. R. Gustafson, C. M. Wood, and J. E. Rosenblatt. 2007.Reemergence of anaerobic bacteremia. Clin. Infect. Dis. 44:895–900.

51. Leszczynski, P., A. van Belkum, H. Pituch, H. Verbrugh, and F. Meisel-Mikolajczyk. 1997. Vaginal carriage of enterotoxigenic Bacteroides fragilisin pregnant women. J. Clin. Microbiol. 35:2899–2903.

52. Martirosian, G., M. Bulanda, B. Wojcik-Stojek, P. Obuch-Woszczatynski,G. Rouyan, P. Heczko, and F. Meisel-Mikolajczyk. 2001. Acute appendici-tis: the role of enterotoxigenic strains of Bacteroides fragilis and Clostridiumdifficile. Med. Sci. Monit. 7:382–386.

53. Massova, I., L. P. Kotra, R. Fridman, and S. Mobashery. 1998. Matrixmetalloproteinases: structures, evolution, and diversification. FASEB J.12:1075–1095.

54. Mazmanian, S. K., C. H. Liu, A. O. Tzianabos, and D. L. Kasper. 2005. Animmunomodulatory molecule of symbiotic bacteria directs maturation ofthe host immune system. Cell 122:107–118.

55. Mazmanian, S. K., J. L. Round, and D. L. Kasper. 2008. A microbialsymbiosis factor prevents intestinal inflammatory disease. Nature 453:620–625.

56. McIver, K. S., E. Kessler, J. C. Olson, and D. E. Ohman. 1995. The elastasepropeptide functions as an intramolecular chaperone required for elastaseactivity and secretion in Pseudomonas aeruginosa. Mol. Microbiol. 18:877–889.

57. Meisel-Mikolajczyk, F., P. Leszczynski, A. van Belkum, H. Pituch, P.Obuch-Woszczatynski, and G. Rouyan. 1999. Enterotoxin-producing Bac-teroides fragilis (ETBF) strains in stool samples submitted for testing ofClostridium difficile and its toxins. Anaerobe 5:217–219.

58. Meisel-Mikolajczyk, F., M. Sebald, E. Torbicka, K. Rafalowska, and U.Zielinska. 1994. Isolation of enterotoxigenic Bacteroides fragilis strains inPoland. Acta Microbiol. Pol. 43:389–392.

59. Mengaud, J., H. Ohayon, P. Gounon, R. Mege, and P. Cossart. 1996.E-cadherin is the receptor for internalin, a surface protein required forentry of L. monocytogenes into epithelial cells. Cell 84:923–932.

59a.Menozzi, M. G., et al. October 1998. 2nd World Congress on AnaerobicBacteria and Infections, Nice, France, abstr. 5.007.

60. Moncrief, J. S., A. J. Duncan, R. L. Wright, L. A. Barroso, and T. D.Wilkins. 1998. Molecular characterization of the fragilysin pathogenicityislet of enterotoxigenic Bacteroides fragilis. Infect. Immun. 66:1735–1739.

61. Moncrief, J. S., R. Obiso, L. A. Barroso, J. J. Kling, R. L. Wright, R. L. VanTassell, D. M. Lyerly, and T. D. Wilkins. 1995. The enterotoxin of Bacte-roides fragilis is a metalloprotease. Infect. Immun. 63:175–181.

62. Moore, W. E. C., and L. V. Holdeman. 1974. Human fecal flora: the normalflora of 20 Japanese-Hawaiians. Appl. Microbiol. 27:961–979.

63. Mundy, L. M., and C. L. Sears. 1996. Detection of toxin production byBacteroides fragilis: assay development and screening of extraintestinal clin-ical isolates. Clin. Infect. Dis. 23:269–276.

64. Myers, L. L., B. D. Firehammer, D. S. Shoop, and M. M. Border. 1984.Bacteroides fragilis: a possible cause of acute diarrheal disease in newbornlambs. Infect. Immun. 44:241–244.

65. Myers, L. L., and D. S. Shoop. 1987. Association of enterotoxigenic Bacte-roides fragilis with diarrheal disease in young pigs. Am. J. Vet. Res. 48:774–775.

66. Myers, L. L., D. S. Shoop, and T. D. Byars. 1987. Diarrhea associated withenterotoxigenic Bacteroides fragilis in foals. Am. J. Vet. Res. 48:1565–1567.

67. Myers, L. L., D. S. Shoop, and J. E. Collins. 1990. Rabbit model to evaluateenterovirulence of Bacteroides fragilis. J. Clin. Microbiol. 28:1658–1660.

68. Myers, L. L., D. S. Shoop, J. E. Collins, and W. C. Bradbury. 1989. Diar-rheal disease caused by enterotoxigenic Bacteroides fragilis in infant rabbits.J. Clin. Microbiol. 27:2025–2030.

69. Myers, L. L., D. S. Shoop, B. D. Firehammer, and M. M. Border. 1985.Association of enterotoxigenic Bacteroides fragilis with diarrheal disease incalves. J. Infect. Dis. 152:1344–1347.

70. Myers, L. L., D. S. Shoop, L. L. Stackhouse, F. S. Newman, R. J. Flaherty,G. W. Letson, and R. B. Sack. 1987. Isolation of enterotoxigenic Bacteroidesfragilis from humans with diarrhea. J. Clin. Microbiol. 25:2330–2333.

71. Nakano, V., D. A. Gomes, R. M. Arantes, J. R. Nicoli, and M. J. Avila-Campos. 2006. Evaluation of the pathogenicity of the Bacteroides fragilistoxin gene subtypes in gnotobiotic mice. Curr. Microbiol. 53:113–117.

72. Nakano, V., T. A. Gomes, M. A. Vieira, R. C. Ferreira, and M. J. Avila-


on Decem


r.asm.org/

Dow

nloaded from

http://cmr.asm.org/

Campos. 2007. bft gene subtyping in enterotoxigenic Bacteroides fragilisisolated from children with acute diarrhea. Anaerobe 13:1–5.

73. Namavar, F., E. B. Theunissen, A. M. Verweij-Van Vught, P. G. Peerbooms,M. Bal, H. F. Hoitsma, and D. M. MacLaren. 1989. Epidemiology of theBacteroides fragilis group in the colonic flora in 10 patients with coloniccancer. J. Med. Microbiol. 29:171–176.

74. Nelson, W. J., and R. Nusse. 2004. Convergence of Wnt, beta-catenin, andcadherin pathways. Science 303:1483–1487.

75. Nguyen, M. H., V. L. Yu, A. J. Morris, L. McDermott, M. W. Wagener, L.Harrell, and D. R. Snydman. 2000. Antimicrobial resistance and clinicaloutcome of Bacteroides bacteremia: findings of a multicenter prospectiveobservational trial. Clin. Infect. Dis. 30:870–876.

76. Nguyen, T. V., P. Le Van, C. Le Huy, K. N. Gia, and A. Weintraub. 2005.Detection and characterization of diarrheagenic Escherichia coli fromyoung children in Hanoi, Vietnam. J. Clin. Microbiol. 43:755–760.

77. Niyogi, S. K., P. Dutta, U. Mitra, and D. K. Pal. 1997. Association of entero-toxigenic Bacteroides fragilis with childhood diarrhoea. Indian J. Med. Res.105:167–169.

78. Obiso, R. J., Jr., A. O. Azghani, and T. D. Wilkins. 1997. The Bacteroidesfragilis toxin fragilysin disrupts the paracellular barrier of epithelial cells.Infect. Immun. 65:1431–1439.

79. Obiso, R. J., Jr., D. Bevan, and T. D. Wilkins. 1997. Molecular modelingand analysis of fragilysin, the Bacteroides fragilis toxin. Clin. Infect. Dis.25:S153–S155.

80. Obiso, R. J., Jr., D. M. Lyerly, R. L. Van Tassell, and T. D. Wilkins. 1995.Proteolytic activity of the Bacteroides fragilis enterotoxin causes fluid secre-tion and intestinal damage in vivo. Infect. Immun. 63:3820–3826.

81. Obuch-Woszczatynski, P., R. G. Wintermans, A. van Belkum, H. Endtz, H.Pituch, D. Kreft, F. Meisel-Mikolajczyk, and M. Luczak. 2004. Enterotoxi-genic Bacteroides fragilis (ETBF) strains isolated in The Netherlands andPoland are genetically diverse. Acta Microbiol. Pol. 53:35–39.

82. Pantosti, A., M. Cerquetti, R. Colangeli, and F. D’Ambrosio. 1994. Detec-tion of intestinal and extra-intestinal strains of enterotoxigenic Bacteroidesfragilis by the HT-29 cytotoxicity assay. J. Med. Microbiol. 41:191–196.

83. Pantosti, A., M. Malpeli, M. Wilks, M. G. Menozzi, and F. D’Ambrosio.1997. Detection of enterotoxigenic Bacteroides fragilis by PCR. J. Clin.Microbiol. 35:2482–2486.

84. Pantosti, A., M. G. Menozzi, A. Frate, L. Sanfilippo, F. D’Ambrosio, and M.Malpeli. 1997. Detection of enterotoxigenic Bacteroides fragilis and its toxin instool samples from adults and children in Italy. Clin. Infect. Dis. 24:12–16.

85. Pathela, P., K. Z. Hasan, E. Roy, K. Alam, F. Huq, A. K. Siddique, and R. B.Sack. 2005. Enterotoxigenic Bacteroides fragilis-associated diarrhea in chil-dren 0–2 years of age in rural Bangladesh. J. Infect. Dis. 191:1245–1252.

86. Pituch, H., P. Obuch-Woszczatynski, F. Luczak, and F. Meisel-Mikolajczyk.2003. Clostridium difficile and enterotoxigenic Bacteroides fragilis strains iso-lated from patients with antibiotic associated diarrhoea. Anaerobe 9:161–163.

87. Polk, F. B., and D. L. Kasper. 1996. Bacteroides fragilis subspecies in clinicalisolates. Ann. Intern. Med. 86:569–571.

88. Prindiville, T. P., R. A. Sheikh, S. H. Cohen, Y. J. Tang, M. C. Cantrell, andJ. Silva, Jr. 2000. Bacteroides fragilis enterotoxin gene sequences in patientswith inflammatory bowel disease. Emerg. Infect. Dis. 6:171–174.

89. Qadri, F., M. G. Mohi, A. Chowdhury, K. Alam, T. Azim, C. L. Sears, R. B.Sack, and M. J. Albert. 1996. Monoclonal antibodies to the enterotoxin ofBacteroides fragilis: production, characterization, and immunodiagnostic ap-plication. Clin. Diagn. Lab. Immunol. 3:608–610.

90. Rabizadeh, S., K. J. Rhee, S. Wu, D. Huso, C. M. Gan, J. E. Golub, X. Wu,M. Zhang, and C. L. Sears. 2007. Enterotoxigenic Bacteroides fragilis: apotential instigator of colitis. Inflamm. Bowel Dis. 13:1475–1483.

91. Redondo, M. C., M. D. Arbo, J. Grindlinger, and D. R. Snydman. 1995.Attributable mortality of bacteremia associated with the Bacteroides fragilisgroup. Clin. Infect. Dis. 20:1492–1496.

92. Rhee, K.-J., S. Wu, X. Wu, D. L. Huso, B. Karim, A. A. Franco, S. Rabi-zadeh, J. Golub, L. E. Mathews, J. Shin, R. B. Sartor, D. Golenbock, A.Hamad, C. M. Gan, F. Housseau, and C. L. Sears. 2009. Induction ofpersistent colitis by a human commensal, enterotoxigenic Bacteroides fragi-lis, in wild-type C57BL/6 mice. Infect. Immun. 77:1708–1718.

93. Riegler, M., M. Lotz, C. Sears, C. Pothoulakis, I. Castagliuolo, C. C. Wang,R. Sedivy, T. Sogukoglu, E. Cosentini, G. Bischof, W. Feil, B. Teleky, G.Hamilton, J. T. LaMont, and E. Wenzl. 1999. Bacteroides fragilis toxin 2damages human colonic mucosa in vitro. Gut 44:504–510.

94. Sack, R. B., M. J. Albert, K. Alam, P. K. B. Neogi, and M. S. Akbar. 1994.Isolation of enterotoxigenic Bacteroides fragilis from Bangladeshi childrenwith diarrhea: a controlled study. J. Clin. Microbiol. 32:960–963.

95. Sack, R. B., L. L. Myers, J. Almeido-Hill, D. S. Shoop, W. C. Bradbury, R.Reid, and M. Santosham. 1992. Enterotoxigenic Bacteroides fragilis: epide-miologic studies of its role as a human diarrhoeal pathogen. J. DiarrhoealDis. Res. 10:4–9.

96. Saidi, R. F., K. Jaeger, M. H. Montrose, S. Wu, and C. L. Sears. 1997.Bacteroides fragilis toxin alters the actin cytoskeleton of HT29/C1 cells invivo qualitatively but not quantitatively. Cell Motil. Cytoskelet. 37:159–165.

97. Saidi, R. F., and C. L. Sears. 1996. Bacteroides fragilis toxin rapidly intox-

icates human intestinal epithelial cells (HT29/C1) in vitro. Infect. Immun.64:5029–5034.

98. Salyers, A. A. 1984. Bacteroides of the human lower intestinal tract. Annu.Rev. Microbiol. 38:293–313.

99. Sanfilippo, L., T. J. Baldwin, M. G. Menozzi, S. P. Borriello, and Y. R.Mahida. 1998. Heterogeneity in responses by primary adult human colonicepithelial cells to purified enterotoxin of Bacteroides fragilis. Gut 43:651–655.

100. Sanfilippo, L., C. K. Li, R. Seth, T. J. Balwin, M. G. Menozzi, and Y. R.Mahida. 2000. Bacteroides fragilis enterotoxin induces the expression ofIL-8 and transforming growth factor-beta (TGF-beta) by human colonicepithelial cells. Clin. Exp. Immunol. 119:456–463.

101. San Joaquin, V. H., J. C. Griffis, C. Lee, and C. L. Sears. 1995. Associationof Bacteroides fragilis with childhood diarrhea. Scand. J. Infect. Dis. 27:211–215.

102. Sears, C. L., S. Islam, A. Saha, M. Arjumand, N. H. Alam, A. S. G. Faruque,M. A. Salam, J. Shin, D. Hecht, A. Weintraub, R. B. Sack, and F. Qadri.2008. Enterotoxigenic Bacteroides fragilis infection is associated with inflam-matory diarrhea. Clin. Infect. Dis. 47:797–803.

103. Sears, C. L. 2001. The toxins of Bacteroides fragilis. Toxicon 39:1737–1746.104. Sears, C. L., S. L. Buckwold, J. W. Shin, and A. A. Franco. 2006. The

C-terminal region of Bacteroides fragilis toxin is essential to its biologicalactivity. Infect. Immun. 74:5595–5601.

105. Sears, C. L., L. L. Myers, A. Lazenby, and R. L. Van Tassell. 1995. Ente-rotoxigenic Bacteroides fragilis. Clin. Infect. Dis. 20(Suppl. 2):S142–S148.

106. Sears, C., A. Franco, and S. Wu. 2005. Bacteroides fragilis toxins, p. 535–546.In J. Alouf and M. Popoff (ed.), The comprehensive sourcebook of bacterialprotein toxins. Academic Press, Oxford, England.

107. Sharma, N., and R. Chaudhry. 2006. Rapid detection of enterotoxigenicBacteroides fragilis in diarrhoeal faecal samples. Indian J. Med. Res. 124:575–582.

108. Shetab, R., S. H. Cohen, T. Prindiville, Y. J. Tang, M. Cantrell, D. Rah-mani, and J. Silva, Jr. 1998. Detection of Bacteroides fragilis enterotoxingene by PCR. J. Clin. Microbiol. 36:1729–1732.

109. Shoop, D. S., L. L. Myers, and J. B. LeFever. 1990. Enumeration of ente-rotoxigenic Bacteroides fragilis in municipal sewage. Appl. Environ. Micro-biol. 56:2243–2244.

110. Sonnenburg, J. L., J. Xu, D. D. Leip, C. H. Chen, B. P. Westover, J.Weatherford, J. D. Buhler, and J. I. Gordon. 2005. Glycan foraging in vivoby an intestine-adapted bacterial symbiont. Science 307:1955–1959.

111. Swidsinski, A., J. Weber, V. Loening-Baucke, L. P. Hale, and H. Lochs.2005. Spatial organization and composition of the mucosal flora in patientswith inflammatory bowel disease. J. Clin. Microbiol. 43:3380–3389.

112. Szoke, I., E. Dosa, and E. Nagy. 2006. Enterotoxigenic Bacteroides fragilis inHungary. Anaerobe 3:87–89.

113. Toprak, N. U., A. Yagci, B. M. Gulluoglu, M. L. Akin, P. Demirkalem, T.Celenk, and G. Soyletir. 2006. A possible role of Bacteroides fragilis entero-toxin in the aetiology of colorectal cancer. Clin. Microbiol. Infect. 12:782–786.

114. Ulger, T. N., D. Rajendram, A. Yagci, S. Gharbia, H. N. Shah, B. M.Gulluoglu, L. M. Akin, P. Demirkalem, T. Celenk, and G. Soyletir. 2006.The distribution of the bft alleles among enterotoxigenic Bacteroides fragilisstrains from stool specimens and extraintestinal sites. Anaerobe 12:71–74.

115. Van Tassell, R. L., D. M. Lyerly, and T. D. Wilkins. 1992. Purification andcharacterization of an enterotoxin from Bacteroides fragilis. Infect. Immun.60:1343–1350.

116. Van Tassell, R. L., D. M. Lyerly, and T. D. Wilkins. 1994. Characterizationof enterotoxigenic Bacteroides fragilis by a toxin-specific enzyme-linked im-munosorbent assay. Clin. Diagn. Lab. Immunol. 1:578–584.

117. Vines, R. R., S. S. Perdue, J. S. Moncrief, D. R. Sentz, L. A. Barroso, R. L.Wright, and T. D. Wilkins. 2000. Fragilysin, the enterotoxin from Bacte-roides fragilis, enhances the serum antibody response to antigen co-admin-istered by the intranasal route. Vaccine 19:655–660.

118. Vu, N. T., P. Le Van, C. Le Huy, and A. Weintraub. 2005. Diarrhea causedby enterotoxigenic Bacteroides fragilis in children less than 5 years of age inHanoi, Vietnam. Anaerobe 11:109–114.

119. Weikel, C. S., F. D. Grieco, J. Reuben, L. L. Myers, and R. B. Sack. 1992.Human colonic epithelial cells, HT29/C1, treated with crude Bacteroidesfragilis enterotoxin dramatically alter their morphology. Infect. Immun.60:321–327.

120. Wells, C. L., E. M. A. Van De Westerlo, R. P. Jechorek, B. A. Feltis, T. D.Wilkins, and S. L. Erlandsen. 1996. Bacteroides fragilis enterotoxin modu-lates epithelial permeability and bacterial internalization by HT-29 entero-cytes. Gastroenterology 110:1429–1437.

121. Wu, S., L. A. Dreyfus, A. O. Tzianabos, C. Hayashi, and C. L. Sears. 2002.Diversity of the metalloprotease toxin produced by enterotoxigenic Bacte-roides fragilis. Infect. Immun. 70:2463–2471.

122. Wu, S., K.-C. Lim, J. Huang, R. F. Saidi, and C. L. Sears. 1998. Bacteroidesfragilis enterotoxin cleaves the zonula adherens protein, E-cadherin. Proc.Natl. Acad. Sci. USA 95:14979–14984.

123. Wu, S., P. J. Morin, D. Maouyo, and C. L. Sears. 2003. Bacteroides fragilis


on Decem


r.asm.org/

Dow

nloaded from

http://cmr.asm.org/

enterotoxin induces c-Myc expression and cellular proliferation. Gastroen-terology 124:392–400.

124. Wu, S., J. Powell, N. Mathioudakis, S. Kane, E. Fernandez, and C. L. Sears.2004. Bacteroides fragilis enterotoxin induces intestinal epithelial cell secre-tion of interleukin-8 through mitogen-activated protein kinases and a ty-rosine kinase-regulated nuclear factor-kappaB pathway. Infect. Immun.72:5832–5839.

125. Wu, S., K. J. Rhee, M. Zhang, A. Franco, and C. L. Sears. 2007. Bacteroidesfragilis toxin stimulates intestinal epithelial cell shedding and �-secretase-dependent E-cadherin cleavage. J. Cell Sci. 120:1944–1952.

126. Wu, S., J. Shin, G. Zhang, M. Cohen, A. Franco, and C. L. Sears. 2006. The

Bacteroides fragilis toxin binds to a specific intestinal epithelial cell receptor.Infect. Immun. 74:5382–5390.

127. Xu, J., M. K. Bjursell, J. Himrod, S. Deng, L. K. Carmichael, H. C. Chiang,L. V. Hooper, and J. I. Gordon. 2003. A genomic view of the human-Bacteroides thetaiotaomicron symbiosis. Science 299:2074–2076.

128. Zhang, G., B. Svenungsson, A. Karnell, and A. Weintraub. 1999. Prevalenceof enterotoxigenic Bacteroides fragilis in adult patients with diarrhea andhealthy controls. Clin. Infect. Dis. 29:590–594.

129. Zhang, G., and A. Weintraub. 1998. Rapid and sensitive assay for de-tection of enterotoxigenic Bacteroides fragilis. J. Clin. Microbiol. 36:3545–3548.

Cynthia L. Sears (M.D.) received her med-ical degree from Thomas Jefferson MedicalCollege, followed by training in InternalMedicine at The New York Hospital in NewYork City. She trained in Infectious Dis-eases at the Memorial Sloan Kettering Can-cer Institute and the University of Virginia,where she was a member of the infectiousdiseases faculty until 1988. Subsequently,Dr. Sears joined the faculty at Johns Hop-kins University School of Medicine, whereshe is now a Professor of Medicine in the Divisions of InfectiousDiseases and Gastroenterology, Department of Medicine. As a resultof international experiences, she developed a clinical interest in food-borne and enteric infections. She has conducted laboratory as well asclinical research on enterotoxigenic Bacteroides fragilis (ETBF) overthe past 15 years. Her present work focuses on studies to define howETBF induces colitis and the link between chronic colonic inflamma-tion induced by bacteria and colonic tumorigenesis.


on Decem


r.asm.org/

Dow

nloaded from

http://cmr.asm.org/

enterotoxigenic bacteroides fragilis: a rogue among symbiotes · in early studies, oral inoculation...

Documents