Author Correction

Bacteroides fragilis toxin stimulates intestinal epithelial cell shedding and�-secretase-dependent E-cadherin cleavage Shaoguang Wu, Ki-Jong Rhee, Ming Zhang, Augusto Franco and Cynthia L. Sears

Journal of Cell Science 120, 3713 (2007) doi:10.1242/jcs.03493

There was an error published in J. Cell Sci. 120, 1944-1952.

The authors apologise for an error that occurred in Fig. 2 of this article. The last band on each gel was labelled �-catenin butshould have been labelled �-actin (as stated in the legend). This error appeared in both the print and the online versions of thisarticle.

The correctly labelled figure is shown below.

1944 Research Article

IntroductionBacteroides fragilis are common colonic commensals thatcolonize the majority of humans (Moore and Holdeman, 1974).Despite comprising a minority (~0.1%) of the total fecal flora,this organism emerges as the leading anaerobic pathogen inbacteremia and is found in the vast majority, if not all,intraabdominal abscesses (Redondo et al., 1995; Polk andKasper, 1977). Molecular genetic studies indicate that B.fragilis are highly heterogeneous and a subset of strains aretermed enterotoxigenic B. fragilis (ETBF) (Franco et al.,1999). ETBFs are associated with diarrheal disease in children,adults and domestic livestock as well as active inflammatorybowel disease (IBD) (Sears, 2001; Basset et al., 2004;Prindiville et al., 2000). However, ETBFs alsoasymptomatically colonize up to 35% of humans (Basset et al.,2004). To date, the virulence of ETBF strains is ascribed tosecretion of a 20-kDa metalloprotease toxin termed the B.fragilis toxin (BFT) (Sears, 2001).

ETBFs produce BFT as a 44 kDa preproprotein toxin that isprocessed by the organism and secreted as a mature 20 kDaprotein (Franco et al., 2002). BFT induces marked changes inintestinal epithelial cell (IEC) function in vitro and in vivo,which are initiated when BFT binds specifically to a IECreceptor that is not E-cadherin (Wu et al., 2006). Biologicalactivity and specific binding of BFT to IECs is mitigated bymutation of the protease motif (Franco et al., 2005; Wu et al.,

2006). After IEC binding, BFT stimulates multiple host cellchanges including: morphological changes in IECs as well asin polarized renal and pulmonary cell lines (Sears, 2001);reduced barrier function and increased ion transport inintestinal cell monolayers and/or human colonic epithelium(Chambers et al., 1997; Riegler et al., 1999); IEC proliferationdependent on E-cadherin expression and, in part, �-catenin–T-cell-factor (TCF) signaling (Wu et al., 2003); and cytokinesecretion by IECs, most notably interleukin-8 (IL-8) (Wu et al.,2004; Kim et al., 2001; Sanfilippo et al., 2000). Consistent withthese in vitro results, histopathology of animal intestinal tissuetreated with ETBF or BFT reveals disruption of the epitheliallayer with a mixture of acute and chronic inflammatory cellsin the lamina propia (Obiso, Jr et al., 1995; Sears et al., 1995).Human intestinal histopathology from ETBF disease is not yetpublished.

E-cadherin is a 120 kDa glycosylated Type I transmembraneprotein critical to the formation of intercellular adhesionjunctions (zonula adherens in the intestinal epithelium) and tocellular signaling, proliferation and differentiation (Nelson andNusse, 2004). The cytoplasmic domain of E-cadherinassociates with �-catenin, which is in turn tethered to �-cateninand actin (Jou et al., 1995). At the cellular level, BFT inducesrapid cleavage (within 1 minute) of E-cadherin yieldingsequentially 33 kDa and 28 kDa cell-associated E-cadherinfragments in the human colonic carcinoma cell line HT29/C1

Enterotoxigenic Bacteroides fragilis – organisms that live inthe colon – secrete a metalloprotease toxin, B. fragilis toxin.This toxin binds to a specific intestinal epithelial cellreceptor and stimulates cell proliferation, which isdependent, in part, on E-cadherin degradation and �-catenin–T-cell-factor nuclear signaling. �-Secretase (orpresenilin-1) is an intramembrane cleaving protease and isa positive regulator of E-cadherin cleavage and a negativeregulator of �-catenin signaling. Here we examine themechanistic details of toxin-initiated E-cadherin cleavage.B. fragilis toxin stimulated shedding of cell membraneproteins, including the 80 kDa E-cadherin ectodomain.Shedding of this domain required biologically active toxinand was not mediated by MMP-7, ADAM10 or ADAM17.Inhibition of �-secretase blocked toxin-induced proteolysis

of the 33 kDa intracellular E-cadherin domain causing cellmembrane retention of a distinct �-catenin pool withoutdiminishing toxin-induced cell proliferation. Unexpectedly,�-secretase positively regulated basal cell proliferationdependent on the �-catenin–T-cell-factor complex. Weconclude that toxin induces step-wise cleavage of E-cadherin, which is dependent on toxin metalloprotease and�-secretase. Our results suggest that differentiallyregulated �-catenin pools associate with the E-cadherin–�-secretase adherens junction complex; one pool regulated by�-secretase is important to intestinal epithelial cellhomeostasis.

Key words: Bacteroides fragilis, Bacteroides fragilis toxin,Metalloproteases, E-cadherin, ADAM, �-secretase

Summary

Bacteroides fragilis toxin stimulates intestinalepithelial cell shedding and �-secretase-dependentE-cadherin cleavageShaoguang Wu1, Ki-Jong Rhee1, Ming Zhang1, Augusto Franco1 and Cynthia L. Sears1,2,*Divisions of Infectious Diseases1 and Gastroenterology2, Department of Medicine, Johns Hopkins University School of Medicine, 1550 Orleans St,CRB2 Bldg Suite 1M.05, Baltimore, MD 21231, USA*Author for correspondence (e-mail: csears@jhmi.edu)

Accepted 10 April 2007Journal of Cell Science 120, 1944-1952 Published by The Company of Biologists 2007doi:10.1242/jcs.03455

Jour

nal o

f Cel

l Sci

ence

1945E-cadherin cleavage by B. fragilis toxin

(Wu et al., 1998). E-cadherin cleavage by BFT releases �-catenin to the cytoplasm resulting in �-catenin nuclearlocalization and stimulation of �-catenin–TCF-dependentcellular proliferation (Wu et al., 2003).

Recently, basal and agonist-induced (such as bystaurosporine or ionomycin) E-cadherin processing has beenreported to involve sequential steps that include cleavage of theE-cadherin ectodomain by host matrix metalloproteinase 7(MMP-7 or matrilysin) or ADAMs (a disintegrin andmetalloprotease) (Marambaud et al., 2002; Ito et al., 1999;Maretzky et al., 2005; Reiss et al., 2005; McGuire et al., 2003;Noe et al., 2001; Steinhusen et al., 2001) followed byprocessing of the intracellular E-cadherin domain, a 38 kDaprotein (in mouse embryos and A431 human epithelial cells)by �-secretase and caspase-3 (Marambaud et al., 2002).ADAMs, a family of at least 32 transmembrane zinc-dependenthost proteases, are implicated in the ectodomain shedding ofvarious membrane-bound proteins (Blobel, 2005). Presenilin-1 or �-secretase (termed �-secretase hereafter) is a 48 kDatransmembrane protein that associates directly andindependently with both E-cadherin and �-catenin andregulates the stability and function of the cadherins or cateninadhesion complex (Baki et al., 2001; Murayama et al., 1998;Kang et al., 2002). �-Secretase is an aspartyl protease of theintramembrane cleaving protease family (termed iCLiPs) andappears to be a promiscuous enzyme that cleaves various TypeI transmembrane domain proteins after they have undergoneectodomain shedding (Kopan and Ilagan, 2004; Wolfe andKopan, 2004; Chyung et al., 2004). Given these new insightsinto cadherin biology, the goal of this study was to investigatein greater detail the mechanism by which BFT induces E-cadherin cleavage.

ResultsBFT induces E-cadherin ectodomain release andHT29/C1 cell membrane protein shedding Using a polyclonal antibody to the C-terminal 25 amino acidsof the cytoplasmic domain of E-cadherin (E2 antibody), Fig.1A,B demonstrates that, consistent with previous data (Wu etal., 1998), BFT treatment of HT29/C1 cells stimulates rapidcleavage of intact 120 kDa E-cadherin yielding 33 kDa and 28kDa cell-associated fragments of E-cadherin that are degradedover time. To determine whether the extracellular domain ofE-cadherin is shed after BFT treatment, HT29/C1 cells weretreated with BFT for 30 minutes to 3 hours followed byconcentration of the proteins in the culture supernatant by TCAprecipitation. Western blot analysis of the cell culturesupernatant proteins using an antibody to the extracellularregion of E-cadherin (H-108 antibody) revealed the release ofan 80 kDa protein band in the cell supernatants of BFT-treatedcells over time compared with a faint signal in the controlsamples (Fig. 1A,B). Using the H-108 antibody, the 80 kDaband was the only protein band detected in BFT-treated cellsupernatants. To determine whether E-cadherin was the onlycell surface protein released after BFT treatment of HT29/C1cells, total cell membrane protein shedding was examined bybiotinylation of cell membrane proteins prior to BFT treatment.As shown in Fig. 1C, unexpectedly, multiple biotinylatedproteins as well as the E-cadherin ectodomain (80 kDa) werereleased from the HT29/C1 cell surface after 1 hour of BFTtreatment when compared to untreated control cells. These

results demonstrate that BFT induces shedding into the cellsupernatant of a 80 kDa E-cadherin fragment as well as other,as yet unknown, membrane proteins.

To assess whether BFT was responsible for the release ofthe 80 kDa ectodomain of E-cadherin, HT29/C1 cells weretreated with partially purified culture supernatants ofrecombinant B. fragilis strains secreting wild-type BFT orbiologically inactive mutant BFT-H352Y and E-cadherincleavage fragments were assessed in both HT29/C1 cellsupernatants and lysates. Treatment of HT29/C1 cells withbacterial culture supernatants containing wild-type BFTreleased the E-cadherin 80 kDa ectodomain and cleavage ofintact 120 kDa E-cadherin in cell lysates similar to thatobserved in response to purified BFT (Fig. 1D). In contrast,neither release of the 80 kDa E-cadherin ectodomain norcleavage of E-cadherin in HT29/C1 cell lysates was detectedin HT29/C1 cells treated with bacterial culture supernatantscontaining mutant BFT-H352Y (Fig. 1D). These resultsindicate that biologically active BFT is required to stimulateE-cadherin cleavage.

MMP-7 as well as ADAM10, but not ADAM17, mediatesE-cadherin ectodomain shedding in selected epithelial cells(McGuire et al., 2003; Maretzky et al., 2005; Noe et al., 2001).In other cells, ADAM10 and ADAM17 appear to exhibitoverlapping substrate profiles (Blobel, 2005). Becausemutations in the protease domain of BFT ablate binding ofBFT to IECs and hence the biological activity of BFT (Wu etal., 2006; Franco et al., 2005), we tested whether MMP-7and/or ADAM10 or ADAM17 mediate E-cadherin cleavageafter BFT cellular binding. We initially tested the role of MMP-7 by determining whether recombinant active MMP-7 (3.3�g/ml) mimicked the biological activity of BFT by stimulatingE-cadherin cleavage and/or cell shape changes in HT29/C1cells over 24 hours. MMP-7-treated HT29/C1 cells developeda spindle morphology distinct from the cell rounding andseparation observed in BFT-treated cells and no cleavage of E-cadherin was detected at 3 or 24 hours suggesting that MMP-7 does not mimic or mediate BFT-initiated E-cadherincleavage. Using RNA interference to directly deplete HT29/C1cell MMP-7 also did not inhibit BFT-initiated E-cadherincleavage. However, HT29/C1 cells depleted of MMP-7 did notgrow as well as control cells and displayed less E-cadherin(Fig. 2A). We also observed that, although expression ofMMP-7 was induced by BFT after 24 hours, the active formof the protein was not detected by western blot in control orBFT-treated HT29/C1 cells, further suggesting that MMP-7 isnot activated by BFT (data not shown).

RNA interference was also used to test whether ADAM10or ADAM17 mediate BFT-initiated E-cadherin cleavage. Fig.2B,C shows that active ADAM10 and ADAM17 were nearlycompletely depleted without modifying either basal levels ofE-cadherin or BFT-initiated E-cadherin cleavage. Similarresults were obtained with dual ADAM10 and ADAM17depletion by RNA interference (data not shown).

The �-secretase inhibitor L685,458 inhibits theproteolysis of the E-cadherin intracellular domaingenerated by BFT in HT29/C1 cellsOur previous data suggested that the initial generation of the33 kDa C-terminal E-cadherin fragment by BFT was notdependent on cellular ATP, whereas the subsequent

Jour

nal o

f Cel

l Sci

ence

1946

degradation of this 33 kDa E-cadherin fragment wasdependent on cellular ATP (Wu et al., 1998). This suggestedthat degradation of the 33 kDa E-cadherin fragment wasprobably attributable to cellular proteases. Recent data hasidentified E-cadherin as one potential substrate for �-secretase(Marambaud et al., 2002). Thus, we tested whetherdegradation of the C-terminal E-cadherin fragments generatedby BFT treatment of HT29/C1 cells was dependent on �-secretase using the selective �-secretase inhibitor, L-685,458.As shown in Fig. 3A, compared with HT29/C1 cells treatedonly with BFT, pretreatment of HT29/C1 cells with L-685,458, but not the �-secretase inhibitor, AEBSF, inhibitsdegradation of the BFT-generated 33 kDa E-cadherinfragment to a 28 kDa E-cadherin fragment. This result was

Journal of Cell Science 120 (11)

verified using RNA interference to directly deplete HT29/C1cell �-secretase (Fig. 2D).

Previous data suggest that �-secretase forms a proteincomplex binding directly and independently (via distinctamino acid residues) to both E-cadherin and �-catenin (Bakiet al., 2001; Murayama et al., 1998; Kang et al., 2002). Wepreviously observed that BFT cleavage of E-cadherin inHT29/C1 cells results in nuclear translocation of a portion ofthe cellular �-catenin with proteasomal degradation of theremaining �-catenin by 3 to 6 hours after BFT treatment ofthe cells (Wu et al., 2003) (our unpublished data). To furtherexamine the effect of �-secretase inhibition on E-cadherincleavage and �-catenin distribution in intact cells, HT29/C1cells were treated with BFT for 1 to 6 hours in the presence

Fig. 1. BFT induces release of the E-cadherin ectodomain and shedding ofIEC membrane proteins. (A) BFTinduces release of the E-cadherinectodomain. Upper panel. Cell lysates ofHT29/C1 cells treated with BFT (5 nM)for 0.5 hour to 3 hours were evaluated bywestern blot using the E2 antibody thatrecognizes the C-terminal domain of E-cadherin. Actin served as an internalcontrol for protein loading. Blot isrepresentative of five experiments. Lowerpanel. TCA-precipitated cell culturesupernatants of HT29/C1 cells treatedwith BFT (5 nM) for 0.5 hour to 3 hourswere evaluated by western blot using theH108 antibody that recognizes theectodomain of E-cadherin. Transferrin,detected by Coomasie Blue staining,served as an internal control for proteinloading. Blot is representative of threeexperiments. (B) Normalized westernblot data demonstrating significant BFT-induced cleavage of intact E-cadherin(120 kDa) on HT29/C1 cells by 0.5 hour(P<0.01) with enhanced detection of a 33kDa cell-associated E-cadherin fragmentat 0.5 hour and release of the 80 kDa E-cadherin ectodomain into cellsupernatants (both P<0.001). The 33 kDaE-cadherin fragment is degraded overtime (see also Fig. 1A). Data are means ±s.d. of three experiments. (C) BFT (5 nM,1 hour) induces shedding of IECmembrane proteins as well as the E-cadherin ectodomain (80 kDa). HT29/C1cell membrane proteins were biotin-labeled and processed as in the Materialsand Methods. Transferrin stained byCoomasie Blue serves as an internalcontrol for protein loading. Blots arerepresentative of four experiments.(D) Cleavage of E-cadherin extracellulardomain requires biologically active BFT.HT29/C1 cells were treated with purifiedBFT (5 nM) or culture supernatants of B. fragilis 9343(pFD340::P-bft) that expresses wild-type BFT or B. fragilis 9343(pFD340::P-bft�H352Y) that expresses mutant biologically inactive BFT. Cell lysates and TCA-precipitated cell culture supernatants were assessed bywestern blot using the E2 (upper lane) and H108 (lower lane) antibodies to the E-cadherin C-terminus or ectodomain, respectively. Lane 1,untreated control; lane 2 purified BFT; lane 3, B. fragilis 9343(pFD340::P-bft); lane 4, B. fragilis 9343(pFD340::P-bft�H352Y); lane 5, brainheart infusion broth alone.

Jour

nal o

f Cel

l Sci

ence

1947E-cadherin cleavage by B. fragilis toxin

or absence of L-685,458 and assessed by immunofluorescentconfocal microscopy (Fig. 3B). In untreated control cells andcells pretreated with L-685,458 for 30 minutes, E-cadherin islocated at the cell membrane and colocalizes with �-catenin

(Fig. 3B, panels a,e). In HT29/C1 cells treated with BFT (5nM), E-cadherin and �-catenin co-association is lost by 1hour, with diffusion of the signals to the cytoplasm over time(panels b-d). By contrast, in HT29/C1 cells pretreated with

Fig. 2. MMP-7, ADAM10 or ADAM17 do notmediate BFT-initiated E-cadherin ectodomainshedding whereas �-secretase mediates, in part,BFT-initiated E-cadherin cleavage. Cell lysates ofcontrol or siRNA-treated [MMP-7 (A), ADAM10(B), ADAM17 (C) or �-secretase (D)ribonucleotide pairs] HT29/C1 cells wereexamined by western blot for each target proteinor E-cadherin (intact 120 kDa or 33 kDadegradation fragment) in the presence or absenceof BFT (5 nM, 2 hours) treatment. �-actin servedas an internal protein loading control. Blots arerepresentative of three experiments.

Fig. 3. Inhibition of �-secretase, but not�-secretase, blocks BFT-inducedcleavage of the C-terminal domain of E-cadherin and partially �-cateninredistribution. (A) Cell lysates ofHT29/C1 cells were analyzed bywestern blot using the E2 antibody tothe C-terminus of E-cadherin. Cellswere treated with BFT alone (30minutes, 5 nM) or with inhibitors [�-secretase inhibitor, AEBSF (0.5 mM) or�-secretase inhibitor, L-685,458 (1.5�M)] for 30 minutes prior to BFTtreatment. Molecular size markersindicate the 120 kDa intact E-cadherinand the 33 and 28 kDa degradationfragments of E-cadherin. �-cateninserves as an internal control for proteinloading. Blot is representative of threeexperiments. (B) HT29/C1 cells weretreated with BFT alone (5 nM) or in thepresence of the �-secretase inhibitor, L-685,458 (1.5 �M), followed byimmunostaining with an anti-�-cateninantibody (CAT-5H10) (green) and theE2 antibody to the C-terminus of E-cadherin (red). Untreated control cellsor cells pretreated for 30 minutes withL-685,458 (Panels a and e,respectively) reveal co-association(yellow) of E-cadherin and �-catenin staining. After BFT treatment, there is diffusion anddissociation of E-cadherin and �-catenin staining at 1 hour (Panel b) with progressivecytoplasmic �-catenin and E-cadherin diffusion over the subsequent 3- and 6-hour timepoints (Panels c,d). In cells pretreated with L-685,458, there is also dissociation of E-cadherin and �-catenin staining at 1 hour but the �-cateninsignal is intense and localized at the cell membrane (Panel f). Over time, partial cytoplasmic diffusion of �-catenin occurs but with retention ofa distinct �-catenin pool on the cell membrane (panels g,h). Representative of three experiments. (C) �-catenin/actin ratios in control HT29/C1cells or cells treated with BFT (5 nM) for 6 hours in the presence or absence of L-685,458 (1.5 �M). L+B6h=L-685,458 + BFT for 6 hours.P<0.05, Control vs BFT, 6 hours. Data are means ± s.d. of five experiments (D) HT29/C1 cells were treated with BFT (5 nM) for 1, 3 or 6hours in the presence or absence of L-685,458 (1.5 �M). Cell lysates were immunoprecipitated using an antibody to �-secretase and thewestern blot analyzed using antibodies to the C-terminus of E-cadherin (E2) (upper panel), �-catenin (CAT-5H10) (middle panel) or �-secretase(lower panel). Intact E-cadherin is 120 kDa. �-secretase inhibition results in stable association of a 33 kDa C-terminal E-cadherin fragment and�-catenin with �-secretase. Western blot is representative of two experiments.

Jour

nal o

f Cel

l Sci

ence

1948

L-685,458, BFT treatment also results in loss of E-cadherinand �-catenin co-association; however, a portion of thecellular �-catenin remains distinctly anchored on the cellmembrane for at least 6 hours after BFT treatment (panels f-h). However, by 6 hours, membrane-associated �-catenin inHT29/C1 cells treated with BFT and L-685,458 is focal ratherthan uniform on the HT29/C1 cell membrane as at earliertime points. By contrast, a portion of the C-terminal E-cadherin fragment appears to lose its membrane staining andbecomes cytoplasmic during the time course of BFTtreatment even in the presence of L-685,458. Additional timecourse analyses indicated that total cellular �-cateninsignificantly declines over 6 hours in BFT-treated cells,consistent with our previously reported data (Wu et al., 2003),whereas �-catenin levels are maintained in HT29/C1 cellstreated with BFT in the presence of L-685,458 (Fig. 3C,D).These data suggest that �-secretase regulates BFT-inducedcytoplasmic release of membrane-associated �-catenin aswell as �-catenin degradation.

To investigate the protein-protein associations of �-secretase, �-catenin and E-cadherin in HT29/C1 cells, �-secretase was immunoprecipitated from HT29/C1 cell lysatestreated or not with BFT (5 nM) and/or L-685,458 (1.5 �M)and co-precipitated E-cadherin, �-catenin and �-secretase wasexamined by western blot. As shown in Fig. 3D, in untreatedcells or cells pretreated for 30 minutes with L-685,458, bothfull-length E-cadherin (120 kDa) and �-catenin are associatedwith �-secretase. Treatment of HT29/C1 cells with BFT overtime leads to near complete loss of association of �-secretasewith E-cadherin initially and then �-catenin (Fig. 3D, left 4lanes). By contrast, �-secretase remains associated with the 33kDa E-cadherin fragment and �-catenin in cells treated withBFT and L-685,458 (Fig. 3D, four right-hand lanes).Consistent with these results, Western blot analysis of Triton-X-100-soluble and -insoluble cell fractions revealed that aportion of the 33 kDa E-cadherin fragment and �-catenin arepresent in the Triton-X-100-insoluble cell fraction after 3 hoursof treatment with BFT and L-685,458 but is not detectable ineither fraction after BFT treatment alone (data not shown).Together these data suggest that �-secretase regulates BFT-induced E-cadherin cleavage in HT29/C1 cells and that �-secretase co-associates with a portion of the E-cadherincytoplasmic domain and �-catenin. A portion of this complexmay dissociate from the cell membrane over time (Fig. 3B,panel h).

�-secretase regulates basal, but not BFT-induced,cellular proliferation in HT29/C1 cellsWe previously reported that BFT stimulates not only E-cadherin cleavage but also nuclear localization of �-cateninand �-catenin–TCF-dependent cellular proliferation (Wu etal., 2003). To evaluate the impact of �-secretase inhibition onBFT-induced �-catenin–TCF nuclear signaling, HT29/C1cellular proliferation was measured using [3H]thymidineincorporation into parental HT29/C1 and HT29/C1 TCF-dominant negative (HT29/C1::Tcf�N31) cell lines. Because�-catenin nuclear signaling requires co-association with TCF,TCF dominant negative HT29/C1 cells disrupt �-catenin-dependent nuclear signaling (Wu et al., 2003). Time courseanalysis indicated that L-685,458 inhibited the degradation ofthe 33 kDa C-terminal E-cadherin fragment generated by

Journal of Cell Science 120 (11)

BFT treatment of HT29/C1 cells for at least 48 hours and didnot alter cell viability as assessed by trypan blue exclusion(data not shown). Fig. 4 shows that, consistent with previousdata, BFT alone stimulates a significant increase in[3H]thymidine uptake versus control parental HT29/C1 cells(lane 1, P<0.001) that is not diminished when the parentalHT29/C1 cells are treated with BFT in the presence of L-685,458 (lane 3 vs lane 1, P=NS). Unexpectedly, L-685,458alone significantly reduced [3H]thymidine uptake in parentalHT29/C1 cells (lane 2, –34±13%, L-685,458-treated cells vscontrol, P<0.001).

In TCF dominant negative HT29/C1 cells, as previouslyreported (Wu et al., 2003), BFT stimulates a significantincrease in [3H]thymidine incorporation in BFT-treated cellscompared with the control (Fig. 4, lane 4, P<0.001) but the[3H]thymidine uptake stimulated by BFT in HT29/C1 cellslacking functional TCF is ~40% lower than in BFT-treatedparental cells (comparison between lanes 1 and 4, P<0.05).These results indicate that �-catenin–TCF nuclear signalingaccounts for a portion, but not all, BFT-initiated HT29/C1 cellproliferation. Similar to the parental HT29/C1 cells, L-685,458did not inhibit BFT stimulation of [3H]thymidine uptake inTCF dominant negative HT29/C1 cells (lane 6, BFT + L-685,458 vs lane 4, BFT alone, P=NS). By contrast, the abilityof L-685,458 to inhibit [3H]thymidine incorporation in parentalHT29/C1 cells is obliterated in TCF dominant negative

Fig. 4. Effect of �-secretase inhibition on basal and BFT-induced cellproliferation. Cell proliferation in parental HT29/C1 cells and TCFdominant negative HT29/C1 cells (HT29/C1::Tcf�N31). Cell lineswere treated with BFT (5 nM) alone or in the presence of the �-secretase inhibitor, L685,458 (1.5 �M). Results are depicted aschanges in the percentage uptake of [3H]thymidine compared withcontrol cells for each cell line. BFT induces a significant increase inHT29/C1 cell proliferation in both parental and TCF dominantnegative HT29/C1 cell lines (lanes 1 or 4, P<0.001 for both cell linesvs. control) that is unaltered by L685,458 (comparisons lanes 1 vs 3and lanes 4 vs 6, P=NS). By contrast, L685,458 inhibits parentalHT29/C1 cell proliferation (lane 2, P<0.001) that is ablated in TCFdominant negative HT29/C1 cells (lane 5, P<0.01 compared withlane 2) indicating �-secretase regulation of parental HT29/C1 cellproliferation is dependent on �-catenin/TCF nuclear signaling. Dataare means ± s.d. of 3-4 experiments conducted with 3-5 replicatesper experiment.

Jour

nal o

f Cel

l Sci

ence

1949E-cadherin cleavage by B. fragilis toxin

HT29/C1 cells (comparison between lanes 2 and 5, P<0.01).These results suggest that �-secretase does not regulate BFT-induced HT29/C1 cell proliferation but does regulate basalTCF-dependent HT29/C1 cell proliferation.

DiscussionBacteroides fragilis are commensals colonizing the majority ofhumans and are the most frequent Bacteroides speciesidentified at the mucosal surface of the colon where theyprobably contribute to intestinal homeostasis (Moore andHoldeman, 1974; Namavar et al., 1989; Hooper and Gordon,2001; Coyne et al., 2005). Virulence of B. fragilis is ascribed,in part, to the unparalleled diversity of the polysaccharidecapsule (Krinos et al., 2001) and, for ETBF strains, to theproduction of a zinc-dependent metalloprotease toxin termedBFT (Sears, 2001). We previously reported that BFT inducesE-cadherin cleavage on intestinal and other polarizableepithelial cell lines resulting in adherens junction disassembly(Chambers et al., 1997; Wu et al., 1998), which initiates IECproliferation dependent on TCF–�-catenin signaling (Wu et al.,2003).

We have recently demonstrated that BFT binds to a specificHT29/C1 cell membrane receptor; that specific BFT bindingrequires BFT protease activity; and that the BFT receptor isunlikely to be either E-cadherin or one of the describedprotease-activated receptors (PAR1-4) (Wu et al., 2006). Thedata reported here extend our understanding of the mechanismby which BFT acts to induce disruption of the adherensjunction and also suggest that signaling from the adherensjunction in IECs is more complex than previously reported.Our results indicate that the proteolytic activity of BFT isnecessary for shedding of the 80 kDa E-cadherin ectodomainbut do not define whether BFT directly cleaves or indirectlystimulates a host cell protease to cleave the E-cadherinectodomain. Because E-cadherin alone is insufficient to permitspecific IEC binding of BFT, we propose a model of BFTaction (Fig. 5) in which BFT binds to a specific host cellreceptor activating a host cell protease directly or by signal

transduction yielding shedding of the E-cadherin ectodomainas well as other cell membrane proteins. This model in whichthe host cell is proposed to mediate BFT-initiated E-cadherincleavage is consistent with recent data indicating that E-cadherin undergoes stepwise cleavage in response to activationof cellular Ca2+-dependent signal transduction with release ofthe E-cadherin ectodomain (Marambaud et al., 2002; Ito et al.,1999; Steinhusen et al., 2001). Further dissection of the initialsteps in BFT interaction with the cell and the role(s) of the BFTprotease domain and/or cellular proteases in the mechanism ofaction of BFT has been limited because we have not identifieda protease inhibitor that specifically inhibits eukaryoticproteases, but not BFT. Similarly, mutational analyses of BFTsuggest it is a highly conserved protein without, as yet, adefined binding domain distinct from its protease domain(Franco et al., 2005; Sears et al., 2006).

The host cell proteases involved in ectodomain cleavage ofE-cadherin have been examined in select cell lines with a roleidentified for MMP-7 in lung epithelium (McGuire et al., 2003)and ADAM10, but not ADAM17, in fibroblast andkeratinocyte cell lines (Maretzky et al., 2005). A similar rolefor ADAM10 in shedding of the N-cadherin ectodomain hasbeen reported (Reiss et al., 2005). Additional data, however,suggest that ADAM10 and ADAM17 (also known as tumornecrosis factor-� converting enzyme or TACE) may, in somecells, modify similar substrates (Blobel, 2005). Thus, wetested, but did not identify any evidence that MMP-7,ADAM10 or ADAM17 mediates BFT-initiated E-cadherincleavage. In particular, siRNA depletion of MMP-7, ADAM10or ADAM17 was nearly complete but did not modify the timecourse of BFT-initiated E-cadherin cleavage (Fig. 2 and datanot shown). In addition, ADAM10 has been proposed toregulate physiological E-cadherin cleavage in selected cells(Maretzky et al., 2005). Thus, we expected, but did notobserve, increased basal intact E-cadherin levels in ADAM10-depleted cells. Our results suggest that, in contrast to other celllines, neither ADAM10 nor ADAM17 are critical to basal orBFT-stimulated E-cadherin cleavage in an IEC model cell line,HT29/C1 cells. Additional studies using, for example, stableinducible knockouts of these enzymes in intestinal epitheliumwill be necessary to definitively evaluate the role of ADAMsin IEC E-cadherin processing.

Interestingly, our data show that BFT induces shedding ofmultiple cell membrane proteins. Although the release ofcellular ligands and receptors by proteases may serve toamplify cell signaling induced by an agonist (Koon et al.,2004), a specific role of the E-cadherin ectodomain as a cellsignaling agonist has not yet been described (Kopan andIlagan, 2004). Nonetheless, it seems likely that one or more ofthe proteins released by BFT may act to modify the cellularresponse. We are presently defining the identity of the BFT-released cell membrane proteins.

After BFT stimulates shedding of the E-cadherinectodomain, processing of the residual E-cadherin cytoplasmicdomain appears to be mediated by �-secretase, similarly topreviously reported data in mouse embryos and A431 cells(Marambaud et al., 2002). Prior reports also indicate thatcleavage of E-cadherin with dissociation of the adherensjunction by BFT, Ca2+ signaling or staurosporine results incytoplasmic redistribution of the nuclear signaling protein, �-catenin (Wu et al., 2003; Marambaud et al., 2002; Ito et al.,

Fig. 5. Proposed model of BFT mechanism of action. BFT binds toand probably cleaves a specific host cell receptor (Wu et al., 2006),potentially activating a host cell protease that either alone orcomplexed with BFT then cleaves the E-cadherin ectodomain andstimulates activation of �-secretase that cleaves the cell-associated E-cadherin remnant. Alternatively, BFT receptor binding may stimulatesignal transduction activating a host cell protease that stimulatescleavage of E-cadherin.

Jour

nal o

f Cel

l Sci

ence

1950

1999). �-secretase has been proposed to inhibit cytoplasmic,but not membrane-associated, �-catenin signaling and thisregulatory role has been reported in embryonic cells as distinctfrom the proteolytic activity of �-secretase (Murayama et al.,1998; Kang et al., 2002). Consistent with this, loss of �-secretase was reported to enhance �-catenin signaling and skintumor development in vivo (Xia et al., 2001). In contrast tothese results, our data suggest that �-secretase activitypositively regulates a membrane- and cytoskeleton-associatedpool of �-catenin that acts in concert with TCF to contributeto basal HT29/C1 cell replication. Consistent with our results,in vivo treatment of rodents with �-secretase inhibitors nearlyhalted intestinal epithelium proliferation, modified IECdifferentiation and altered small intestine morphology(Searfoss et al., 2003; Wong et al., 2004; van Es and Clevers,2005). The impact of in vivo �-secretase inhibition on colonicmorphology was not reported. However, �-secretase inhibitiondid not alter BFT-induced cellular proliferation in parental orTCF dominant negative HT29/C1 cells. Together these resultsreveal, to our knowledge, the first evidence for twodifferentially regulated or heterogeneous pools of �-cateninassociated with the adherens junction of IECs. BFT-inducedcellular proliferation in HT29/C1 cells, although dependent onE-cadherin expression and partially dependent on �-catenin–TCF signaling (Wu et al., 2003), is independent of �-secretase-mediated E-cadherin cleavage whereas basal cellproliferation is dependent on �-secretase-regulated �-catenin–TCF signaling (Fig. 4).

Previous data indicate that �-catenin associates with E-cadherin and that both proteins bind to �-secretase at distinctsites (Murayama et al., 1998; Baki et al., 2001; Kang et al.,2002). Thus, one possibility to explain our results is that BFTtreatment of cells leads to a biochemical modification of aportion of E-cadherin-associated �-catenin contributing to itscytoplasmic and nuclear redistribution whereas unmodified �-catenin remains regulated by �-secretase. Initial studies usingantibodies to mono- or triphosphorylated �-catenin have notidentified differences in these �-catenin species betweenHT29/C1 cells treated with BFT in the presence or absence ofL-685,458 (S.W. and C.L.S., unpublished data). However,additional work is necessary to explore this possibility. Unlikeother �-secretase-regulated pathways, such as proteolysis ofNotch, where the released cytoplasmic domain of the targetprotein acts as a transcriptional regulator, the E-cadherincytoplasmic domain has not yet been identified to act as atranscriptional regulator (Kopan and Ilagan, 2004; Wolfe andKopan, 2004). Our preliminary data suggest BFT does notupregulate expression of Notch-regulated genes (S.W. andC.L.S., unpublished data). Although mutations in �-secretaseaccount for the majority of early-onset familial Alzheimer’sdisease (Brunkan and Goate, 2005), it is unknown whethermutations in �-secretase contribute to IEC signalingdisturbances resulting in intestinal pathology or disease.

Bacterial toxins have long served as tools to investigatenormal and pathological cellular function. BFT sharessequence homology with the eukaryotic MMPs of themetzincin family and has been proposed to be the ancestralorigin of the eukaryotic MMPs (Massova et al., 1998). Furtherstudies to detail the mechanism of action of BFT willcontribute to our understanding of IEC signaling pathways andcell function.

Journal of Cell Science 120 (11)

Materials and MethodsCell linesHT29/C1 cells are cloned cells derived from a human colon carcinoma (obtainedfrom Dr Daniel Louvard, Institut Pasteur, Paris, France) (Huet et al., 1987;Montrose-Rafizadeh et al., 1991). The HT29/C1::TCF-4�N31 cell line is a derivedpolyclonal stable TCF dominant negative HT29/C1 cell line (Wu et al., 2003). Bothcell lines were grown subconfluently in Dulbecco’s minimum essential medium(DMEM) containing streptomycin (0.1 mg/ml), penicillin (100 U/ml), G418 (0.4mg/ml for HT29/C1::Tcf�N31 cell line only), human transferrin (10 �g/ml, Sigma,St Louis, MO) and 10% fetal bovine serum (HyClone, Logan, Utah). All culturemedia and reagents were purchased from GIBCO BRL Life Technologies(Rockville, MD) unless otherwise stated.

BFT purification and bacterial strainsBFT was purified from culture supernatants of B. fragilis recombinant strainI1345(pFD340::P-bft) as previously described (Wu et al., 2002). In someexperiments, filter-sterilized, concentrated culture supernatants of B. fragilisrecombinant strains 9343(pFD340::P-bft, secretes wild-type BFT) and9343(pFD340::P-bft�H352Y, secretes mutant biologically inactive BFT owing to asingle nucleotide point mutation in the BFT metalloprotease domain) (Franco et al.,2005) were used for the treatment of HT29/C1 cells. Similar quantities of wild-typeand mutated inactive BFT were present in culture supernatants as assessed bywestern blot analysis using anti-BFT antibody (data not shown).

Inhibitors and other treatmentsHT29/C1 cells were washed once with Hank’s balanced salt solution beforetreatment with purified BFT at specified concentrations in DMEM lacking serum.Cells were incubated with the �-secretase inhibitor (L-685,458; Calbiochem, SanDiego, CA) or �-secretase inhibitor [(4-(2-aminoethyl) benzenesulfonylfluoride(AEBSF), Sigma] for 30 minutes before the addition of BFT and then continuouslyduring the experiment unless otherwise described. Active human recombinantmatrilysin (MMP-7; specific activity �3000 U/mg) was obtained from Calbiochem.

AntibodiesE2: polyclonal anti-C-terminal of E-cadherin antibody (provided by James Nelson,Stanford University, Palo Alto, CA); C36: monoclonal anti-C-terminal E-cadherinantibody (BD Biosciences, Palo Alto, CA); H108: polyclonal anti-extracellular E-cadherin domain antibody (Santa Cruz Biotechnology, Santa Cruz, CA); CAT-5H10:monoclonal anti-�-catenin antibody (Zymed Laboratories Inc., San Francisco, CA);polyclonal anti-�-presenilin-1-loop antibody (Calbiochem); monoclonal anti-presenilin-1-loop antibody (Chemicon International, Temecula, CA); AC-40 or AC-15: monoclonal anti-actin antibody (Sigma); monoclonal anti-biotin antibody(Jackson ImmunoResearch Laboratories, West Grove, Pa); polyclonal anti-cytoplasmic ADAM10 or ADAM17 domain antibody (Axxora, San Diego, CA);polyclonal anti-active/latent MMP-7 antibody (Chemicon).

Immunoblot analysis and evaluation of HT29/C1 cellsupernatant proteinsFor western blot analyses, proteins were probed using the primary antibodies asindicated then horseradish peroxidase-coupled secondary antibodies (JacksonImmunoResearch Laboratories) after 10% SDS-PAGE separation and nitrocellulosemembrane transfer (Sambrook et al., 1989). Immunoreactive proteins were detectedusing Super Signal West Pico Chemiluminiscent Substrate (Pierce, Rockford, IL)and quantified by densitometry using AlphaEaseTM (version 5.1; Alpha InnotechCorporation, San Leandro, CA).

To detect released E-cadherin fragments in HT29/C1 cell culture supernatants,proteins in the culture supernatants were precipitated with 10% trichloroacetic acidfollowed by western blot using the H108 antibody. To detect cell membrane proteinsshed from cells treated with BFT, HT29/C1 cell surface proteins were biotinylatedusing EZ-Link Sulfo-NHS-Biotin Reagents (Pierce, Rockford, IL) as per themanufacturer’s instructions. Proteins shed into the culture supernatants wereretrieved using streptavidin beads, eluted with 0.05 M glycine (pH 2.5) and detectedby western blot using an anti-biotin antibody.

Cell proteins were extracted with 1% Triton X-100 in phosphate-buffered salinefor 10 minutes at 4°C. The supernatant (Triton X-100 soluble fraction) was collectedand the residual portion (Triton X-100 insoluble fraction) was solubilized with 1%SDS-lysis buffer (Nathke et al., 1994). The proteins from both fractions wereanalyzed by western blot as described.

RNA interferenceSmall interfering RNA (siRNA) duplex oligoribonucleotides against humanADAM10, ADAM17, MMP-7 or �-secretase were selected and synthesized byInvitrogen (StealthTM RNAi, Carlsbad, CA). The sequences were as follows: (i)ADAM10 sense 5�-GAGGAAAUACCAGAUGACUGGUGUA-3�, antisense 5�-UACACCAGUCAUCUGGUAUUUCCUC-3�, (ii) ADAM 17 sense 5�-GGAA -GCUGACCUGGUUACAACUCAU-3�, antisense 5�-AUGAGUUGUAACCAG -GUCAGCUUCC-3�; (iii) MMP-7 sense 5�-CCCGCGUCAUAGAAAUAAU G -

Jour

nal o

f Cel

l Sci

ence

1951E-cadherin cleavage by B. fragilis toxin

CAGAA-3�, antisense 5�-UUCUGCAUUAUUUCUAUGACGCGGG-3�; (iv) �-secretase sense 5�-GCUCAGGAGAGAAAUGAAACGCUUU-3�, antisense 5�-AAAGCGUUUCAUUUCUCUCCUGAGC-3�. Control siRNA (Cat.No. 12935-300) was purchased from Invitrogen. Two hours after cell plating (5104 cells/wellusing 24-well plate), the siRNA oligoribonucleotides were transfected intoHT29/C1 cells using DharmaFECTTM 4 reagent (Dharmacon, Lafayette, CO)according to the manufacturer’s protocol followed by BFT treatment 3 days later.After washing with cold PBS, cells were lysed with RIPA buffer [Sigma; 50 mMTris-HCl (pH 7.2-8.0), 150 mM sodium chloride, 1.0% NP-40, 0.5-1% sodiumdeoxycholate, 0.1% SDS] supplemented with protease inhibitors (RocheDiagnostics, Indianapolis, IN) followed by western blot analysis.

ImmunoprecipitationHT29/C1 cells were lysed in RIPA buffer and incubated with the desired antibody(1-5 �g) overnight at 4°C. Protein A-Sepharose beads (CL-4B, AmershamBiosciences, Piscataway, NJ) were added and incubated at 25°C for 1 hour. Proteinswere eluted from the beads in 50 �l of 0.1 M citric acid (pH 2.5).

Immunofluorescent stainingHT29/C1 cells grown on eight-well chamber slides were fixed with 4%paraformaldehyde (Sigma) and permeabilized with 1% Triton X-100. Specificproteins were immunostained with antibodies as indicated followed by incubationwith Alexa Fluor 488-labeled or Alexa Fluor 568-labeled secondary antibody(Molecular Probes, Eugene, OR). The signal was observed by single or dual channelconfocal microscopy (Zeiss, LSM410). The images were processed usingMetamorph (Universal Imaging, West Chester, Pa) or Photoshop 8.0.

[3H]thymidine incorporation experimentsHT29/C1 or HT29/C1::TCF-4�N31 cells in a 96-well plate treated with or withoutBFT in the presence or absence of L-685,458 were exposed to [3H]thymidine (1�Ci/ml) for 3 hours prior to the designated end of the experiment 48 hours afterinitiation of BFT treatment as previously described (Wu et al., 2003).

Statistical analysisData are expressed as means ± s.d. GraphPad Instat Version 3.05 was used forStudent’s t-test or ANOVA determinations to examine for significant statisticaldifferences considered to be P values less than 0.05.

The authors thank Dwight Derr for assistance with tissue cultureand the laboratory of Anirban Maitra for assessment of Notch-regulated gene expression. Supported by grant RO1 DK 45496 (toC.L.S.) and R24 DK 64388 (PI: Mark Donowitz).

ReferencesBaki, L., Marambaud, P., Efthimiopoulos, S., Georgakopoulos, A., Wen, P., Cui, W.,

Shioi, J., Koo, E., Ozawa, M., Friedrich, V. L., Jr et al. (2001). Presenilin-1 bindscytoplasmic epithelial cadherin, inhibits cadherin/p120 association, and regulatesstability and function of the cadherin/catenin adhesion complex. Proc. Natl. Acad. Sci.USA 98, 2381-2386.

Basset, C., Holton, J., Bazeos, A., Vaira, D. and Bloom, S. (2004). Are Helicobacterspecies and enterotoxigenic Bacteroides fragilis involved in inflammatory boweldisease? Dig. Dis. Sci. 49, 1425-1432.

Blobel, C. P. (2005). ADAMs: key components in EGFR signalling and development.Nat. Rev. Mol. Cell Biol. 6, 32-43.

Brunkan, A. L. and Goate, A. M. (2005). Presenilin function and gamma-secretaseactivity. J. Neurochem. 93, 769-792.

Chambers, F. G., Koshy, S. S., Saidi, R. F., Clark, D. P., Moore, R. D. and Sears, C.L. (1997). Bacteroides fragilis toxin exhibits polar activity on monolayers of humanintestinal epithelial cells (T84 Cells) in vitro. Infect. Immun. 65, 3561-3570.

Chyung, J. H., Raper, D. M. and Selkoe, D. J. (2005). gamma-secretase exists on theplasma membrane as an intact complex that accepts substrates and effectsintramembrane cleavage. J. Biol. Chem. 280, 4383-4392.

Coyne, M. J., Reinap, B., Lee, M. M. and Comstock, L. E. (2005). Human symbiontsuse a host-like pathway for surface fucosylation. Science 307, 1778-1781.

Franco, A. A., Cheng, R. K., Chung, G. T., Wu, S., Oh, H. B. and Sears, C. L. (1999).Molecular evolution of the pathogenicity island of enterotoxigenic Bacteroides fragilisstrains. J. Bacteriol. 181, 6623-6633.

Franco, A. A., Cheng, R. K., Goodman, A. and Sears, C. L. (2002). Modulation of bftexpression by the Bacteroides fragilis pathogenicity island and its flanking region. Mol.Microbiol. 45, 1067-1077.

Franco, A. A., Buckwold, S., Shin, J. W., Ascon, M. and Sears, C. L. (2005). Mutationof the zinc-binding metalloprotease motif affects Bacteroides fragilis toxin activitywithout affecting propeptide processing. Infect. Immun. 73, 5273-5277.

Hooper, L. V. and Gordon, J. I. (2001). Commensal host-bacterial relationships in thegut. Science 292, 1115-1118.

Huet, C., Sahuquillo-Merino, C., Coudrier, E. and Louvard, D. (1987). Absorptiveand mucus-secreting subclones isolated from a multipotent intestinal cell line (HT29)provide new models for cell polarity and terminal differentiation. J. Cell Biol. 105,345-357.

Ito, K., Okamoto, I., Araki, N., Kawano, Y., Nakao, M., Fujiyama, S., Tomita, K.,Mimori, T. and Saya, H. (1999). Calcium influx triggers the sequential proteolysis ofextracellular and cytoplasmic domains of E-cadherin, leading to loss of beta-cateninfrom cell-cell contacts. Oncogene 18, 7080-7090.

Jou, T., Stewart, D. B., Stappert, J., Nelson, W. J. and Marrs, J. A. (1995). Geneticand biochemical dissection of protein linkages in the cadherin-catenin complex. Proc.Natl. Acad. Sci. USA 92, 5067-5071.

Kang, D. E., Soriano, S., Xia, X., Eberhart, C. G., De Strooper, B., Zheng, H. andKoo, E. H. (2002). Presenilin couples the paired phosphorylation of beta-cateninindependent of axin: implications for beta-catenin activation in tumorigenesis. Cell110, 751-762.

Kim, J. M., Oh, Y. K., Kim, Y. J., Oh, H. B. and Cho, Y. J. (2001). Polarized secretionof CXC chemokines by human intestinal epithelial cells in response to Bacteroidesfragilis enterotoxin: NF-kappa B plays a major role in the regulation of IL-8 expression.Clin. Exp. Immunol. 123, 421-427.

Koon, H. W., Zhao, D., Na, X., Moyer, M. P. and Pothoulakis, C. (2004).Metalloproteinases and transforming growth factor-alpha mediate substance P-inducedmitogen-activated protein kinase activation and proliferation in human colonocytes. J.Biol. Chem. 279, 45519-45527.

Kopan, R. and Ilagan, M. X. (2004). Gamma-secretase: proteasome of the membrane?Nat. Rev. Mol. Cell Biol. 5, 499-504.

Krinos, C. M., Coyne, M. J., Weinacht, K. G., Tzianabos, A. O., Kasper, D. L. andComstock, L. E. (2001). Extensive surface diversity of a commensal microorganismby multiple DNA inversions. Nature 414, 555-558.

Marambaud, P., Shioi, J., Serban, G., Georgakopoulos, A., Sarner, S., Nagy, V., Baki,L., Wen, P., Efthimiopoulos, S., Shao, Z. et al. (2002). A presenilin-1/gamma-secretase cleavage releases the E-cadherin intracellular domain and regulatesdisassembly of adherens junctions. EMBO J. 21, 1948-1956.

Maretzky, T., Reiss, K., Ludwig, A., Buchholz, J., Scholz, F., Proksch, E., De Strooper,B., Hartmann, D. and Saftig, P. (2005). ADAM10 mediates E-cadherin shedding andregulates epithelial cell-cell adhesion, migration, and beta-catenin translocation. Proc.Natl. Acad. Sci. USA 102, 9182-9187.

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

McGuire, J. K., Li, Q. and Parks, W. C. (2003). Matrilysin (matrix metalloproteinase-7) mediates E-cadherin ectodomain shedding in injured lung epithelium. Am. J. Pathol.162, 1831-1843.

Montrose-Rafizadeh, C., Guggino, W. B. and Montrose, M. H. (1991). Cellulardifferentiation regulates expression of Cl- transport and cystic fibrosis transmembraneconductance regulator mRNA in human intestinal cells. J. Biol. Chem. 266, 4495-4499.

Moore, W. E. C. and Holdeman, L. V. (1974). Human fecal flora: the normal flora of20 Japanese-Hawaiians. Appl. Microbiol. 27, 961-979.

Murayama, M., Tanaka, S., Palacino, J., Murayama, O., Honda, T., Sun, X.,Yasutake, K., Nihonmatsu, N., Wolozin, B. and Takashima, A. (1998). Directassociation of presenilin-1 with beta-catenin. FEBS Lett. 433, 73-77.

Namavar, F., Theunissen, E. B., Verweij-Van Vught, A. M., Peerbooms, P. G., Bal,M., Hoitsma, H. F. and MacLaren, D. M. (1989). Epidemiology of the Bacteroidesfragilis group in the colonic flora in 10 patients with colonic cancer. J. Med. Microbiol.29, 171-176.

Nathke, I. S., Hinck, L., Swedlow, J. R., Papkoff, J. and Nelson, W. J. (1994). Defininginteractions and distributions of cadherin and catenin complexes in polarized epithelialcells. J. Cell Biol. 125, 1341-1352.

Nelson, W. J. and Nusse, R. (2004). Convergence of Wnt, beta-catenin, and cadherinpathways. Science 303, 1483-1487.

Noe, V., Fingleton, B., Jacobs, K., Crawford, H. C., Vermeulen, S., Steelant, W.,Bruyneel, E., Matrisian, L. M. and Mareel, M. (2001). Release of an invasionpromoter E-cadherin fragment by matrilysin and stromelysin-1. J. Cell Sci. 114, 111-118.

Obiso, R. J., Jr, Lyerly, D. M., Van Tassell, R. L. and Wilkins, T. D. (1995). Proteolyticactivity of the Bacteroides fragilis enterotoxin causes fluid secretion and intestinaldamage in vivo. Infect. Immun. 63, 3820-3826.

Polk, B. F. and Kasper, D. L. (1977). Bacteroides fragilis subspecies in clinical isolates.Ann. Intern. Med. 86, 569-571.

Prindiville, T. P., Sheikh, R. A., Cohen, S. H., Tang, Y. J., Cantrell, M. C. and Silva,J., Jr (2000). Bacteroides fragilis enterotoxin gene sequences in patients withinflammatory bowel disease. Emerging Infect. Dis. 6, 171-174.

Redondo, M. C., Arbo, M. D., Grindlinger, J. and Snydman, D. R. (1995). Attributablemortality of bacteremia associated with the Bacteroides fragilis group. Clin. Infect. Dis.20, 1492-1496.

Reiss, K., Maretzky, T., Ludwig, A., Tousseyn, T., De Strooper, B., Hartmann, D. andSaftig, P. (2005). ADAM10 cleavage of N-cadherin and regulation of cell-cell adhesionand beta-catenin nuclear signalling. EMBO J. 24, 742-752.

Riegler, M., Lotz, M., Sears, C., Pothoulakis, C., Castagliuolo, I., Wang, C. C., Sedivy,R., Sogukoglu, T., Cosentini, E., Bischof, G. et al. (1999). Bacteroides fragilis toxin2 damages human colonic mucosa in vitro. Gut 44, 504-510.

Sambrook, J., Fritsch, E. F. and Maniatis, T. (1989). Molecular Cloning: A LaboratoryManual. Cold Spring Harbor: Cold Spring Harbor Laboratory Press.

Sanfilippo, L., Li, C. K., Seth, R., Balwin, T. J., Menozzi, M. G. and Mahida, Y. R.(2000). Bacteroides fragilis enterotoxin induces the expression of IL-8 andtransforming growth factor-beta (TGF-beta) by human colonic epithelial cells. Clin.Exp. Immunol. 119, 456-463.

Searfoss, G. H., Jordan, W. H., Calligaro, D. O., Galbreath, E. J., Schirtzinger, L.M., Berridge, B. R., Gao, H., Higgins, M. A., May, P. C. and Ryan, T. P. (2003).

Jour

nal o

f Cel

l Sci

ence

Adipsin, a biomarker of gastrointestinal toxicity mediated by a functional gamma-secretase inhibitor. J. Biol. Chem. 278, 46107-46116.

Sears, C. L. (2001). The toxins of Bacteroides fragilis. Toxicon 39, 1737-1746.Sears, C. L., Myers, L. L., Lazenby, A. and Van Tassell, R. L. (1995). Enterotoxigenic

Bacteroides fragilis. Clin. Infect. Dis. 20 Suppl. 2, S142-S148.Sears, C. L., Buckwold, S. L., Shin, J. W. and Franco, A. A. (2006). The C-terminal

region of Bacteroides fragilis toxin is essential to its biological activity. Infect. Immun.74, 5595-5601.

Steinhusen, U., Weiske, J., Badock, V., Tauber, R., Bommert, K. and Huber, O.(2001). Cleavage and shedding of E-cadherin after induction of apoptosis. J. Biol.Chem. 276, 4972-4980.

van Es, J. H. and Clevers, H. (2005). Notch and Wnt inhibitors as potential new drugsfor intestinal neoplastic disease. Trends Mol. Med. 11, 496-502.

Wolfe, M. S. and Kopan, R. (2004). Intramembrane proteolysis: theme and variations.Science 305, 1119-1123.

Wong, G. T., Manfra, D., Poulet, F. M., Zhang, Q., Josien, H., Bara, T., Engstrom,L., Pinzon-Ortiz, M., Fine, J. S., Lee, H. J. et al. (2004). Chronic treatment with thegamma-secretase inhibitor LY-411,575 inhibits beta-amyloid peptide production andalters lymphopoiesis and intestinal cell differentiation. J. Biol. Chem. 279, 12876-12882.

Wu, S., Lim, K.-C., Huang, J., Saidi, R. F. and Sears, C. L. (1998). Bacteroides fragilisenterotoxin cleaves the zonula adherens protein, E-cadherin. Proc. Natl. Acad. Sci. USA95, 14979-14984.

Wu, S., Dreyfus, L. A., Tzianabos, A. O., Hayashi, C. and Sears, C. L. (2002).Diversity of the metalloprotease toxin produced by enterotoxigenic Bacteroidesfragilis. Infect. Immun. 70, 2463-2471.

Wu, S., Morin, P. J., Maouyo, D. and Sears, C. L. (2003). Bacteroides fragilisenterotoxin induces c-Myc expression and cellular proliferation. Gastroenterology 124,392-400.

Wu, S., Powell, J., Mathioudakis, N., Kane, S., Fernandez, E. and Sears, C. L. (2004).Bacteroides fragilis enterotoxin induces intestinal epithelial cell secretion ofinterleukin-8 through mitogen-activated protein kinases and a tyrosine kinase-regulatednuclear factor-kappaB pathway. Infect. Immun. 72, 5832-5839.

Wu, S., Shin, J., Zhang, G., Cohen, M., Franco, A. and Sears, C. L. (2006). TheBacteroides fragilis toxin binds to a specific intestinal epithelial cell receptor. Infect.Immun. 74, 5382-5390.

Xia, X., Qian, S., Soriano, S., Wu, Y., Fletcher, A. M., Wang, X. J., Koo, E. H., Wu,X. and Zheng, H. (2001). Loss of presenilin 1 is associated with enhanced beta-catenin signaling and skin tumorigenesis. Proc. Natl. Acad. Sci. USA 98, 10863-10868.

Journal of Cell Science 120 (11)1952

Jour

nal o

f Cel

l Sci

ence