INFECTION AND IMMUNITY, Nov. 2005, p. 7113–7125 0019-9567/05/$08.00⫹0 doi:10.1128/IAI.73.11.7113–7125.2005 Copyright © 2005, American Society for Microbiology. All Rights Reserved.
Vol. 73, No. 11
Cholesterol-Enriched Membrane Microdomains Are Required for Inducing Host Cell Cytoskeleton Rearrangements in Response to Attaching-Effacing Escherichia coli Jason D. Riff,1,2 John W. Callahan,1,3,4 and Philip M. Sherman1,2,3,5* Research Institute, Hospital for Sick Children,1 and Departments of Laboratory Medicine and Pathobiology,2 Paediatrics,3 and Biochemistry,4 and the Institute of Medical Science,5 University of Toronto, Toronto, Canada Received 26 January 2005/Returned for modification 18 April 2005/Accepted 8 July 2005
The diarrheal pathogens enterohemorrhagic Escherichia coli (EHEC) O157:H7 strain CL56 and enteropathogenic Escherichia coli (EPEC) O127:H6 strain E2348/69 adhere intimately to epithelial cells through attaching-effacing lesions, which are characterized by rearrangements of the host cytoskeleton, intimate adherence, and destruction of microvilli. These cytoskeletal responses require activation of host signal transduction pathways. Lipid rafts are signaling microdomains enriched in sphingolipid and cholesterol in the plasma membrane. The effect of perturbing plasma membrane cholesterol on bacterial intimate adherence was assessed. Infection of both HEp-2 cells and primary skin fibroblasts with strains CL56 and E2348/69 caused characteristic rearrangements of the cytoskeleton at sites of bacterial adhesion. CL56- and E2348/69-induced cytoskeletal rearrangements were inhibited following cholesterol depletion. Addition of exogenous cholesterol to depleted HEp-2 cells restored cholesterol levels and rescued bacterially induced ␣-actinin mobilization. Quantitative bacterial adherence assays showed that EPEC adherence to HEp-2 cells was dramatically reduced following cholesterol depletion, whereas the adherence of EHEC remained high. Cytoskeletal rearrangements on skin fibroblasts obtained from children with Niemann-Pick type C disease were markedly reduced. These findings indicate that host membrane cholesterol contained in lipid rafts is necessary for the cytoskeletal rearrangements following infection with attaching-effacing Escherichia coli. Differences in initial adherence indicate divergent roles for host membrane cholesterol in the pathogenesis of EHEC and EPEC infections. Enterohemorrhagic Escherichia coli (EHEC) serotype O157:H7 frequently causes outbreaks of bloody diarrhea in developed countries, such as occurred in Walkerton, Ontario, Canada, during the summer of 2000, when over 3,000 people were infected and seven deaths occurred (17). Infection with EHEC can be further complicated by the development of the hemolytic uremic syndrome, which is the most common cause of acute renal failure in children (21). EHEC O157:H7 and the related diarrheal pathogen enteropathogenic Escherichia coli (EPEC) serotype O127:H6 both colonize the host intestinal tract by initial binding events followed by the development of intimate adhesion through characteristic attaching and effacing lesions. To achieve the attaching-effacing lesion, these bacteria possess a homologous pathogenicity island termed the locus of enterocyte effacement (6, 34) that encodes a type III secretion system. This secretion system delivers a number of secreted effector proteins into the host cell, including EspE, EspB, EspD, EspF, and Map (5, 22, 23). The eukaryotic plasma membrane is not a homogeneous phospholipid bilayer but contains specialized cholesterol and sphingolipid-rich microdomains, termed lipid rafts (28, 37). Functionally, lipid rafts serve as platforms for protein sorting and membrane trafficking, as well as containing many mole-
cules important for signal transduction events involved in proliferation, apoptosis, cell migration, and adhesion (11). In addition, microorganisms and their secreted products utilize lipid rafts in order to exert their effects on host cells (6, 27, 29, 40). The distinct involvement of lipid rafts in signaling functions (10) led us to hypothesize a role for these cholesterol-enriched microdomains in the formation of E. coli-induced attachingeffacing lesions. In this study, the involvement of host cell lipid rafts in the formation of attaching-effacing lesions was determined using complementary approaches. The results demonstrate that host plasma membrane cholesterol is required for bacterial adherence and attaching-effacing cytoskeleton alterations in response to both EHEC O157:H7 and EPEC O127:H6 infections. MATERIALS AND METHODS Tissue culture and cell lines. HEp-2 human laryngeal epithelial cell line (CCL 23; American Type Culture Collection, Manassas, VA), were cultured at 37°C and 5% CO2 in minimal essential medium supplemented with 10% fetal bovine serum, 1% sodium bicarbonate, 1% Fungizone, and 1% penicillin-streptomycin (all media and supplements from Life Technologies, Grand Island, NY). Patients with Niemann-Pick type C (NPC) disease (Hospital for Sick Children) were defined using the cholesterol esterification assay as previously described (2). Fibroblasts were grown in at 37°C and 5% CO2 in ␣-minimal essential medium (Wisent Inc., Saint-Jean-Baptiste de Rouville, Canada) supplemented with 10% fetal bovine serum (Life Technologies). Bacterial growth and conditions of infection. EHEC serotype O157:H7, strain CL56, and EPEC serotype O127:H6, strain E2348/69, were held on 5% sheep blood agar plates at 4°C. Individual colonies were scraped into Penassay broth (Difco Laboratories, Detroit, Mich.) and grown for 18 h at 37°C before use in experimental infection, as previously described (4). For experimental infection, stationary-phase bacteria were added to tissue culture cells grown in Lab-Tek
* Corresponding author. Mailing address: Room 8409, Hospital for Sick Children, 555 University Avenue, Toronto, Ontario M5G 1X8, Canada. Phone: (416) 813-7734. Fax: (416) 813-6531. E-mail:
[email protected]. 7113
7114
RIFF ET AL.
INFECT. IMMUN.
FIG. 1. EHEC O157:H7-induced cytoskeletal rearrangements are disrupted by cholesterol depletion. Immunofluorescence micrographs (a to c) using an ␣-actinin antibody (left column) and an EHEC O157:H7 antibody (middle column) and merged images (right column). (a) Uninfected cells show no cytoskeletal rearrangements and lack bacterial staining. (b) Cells infected with EHEC O157:H7 showed foci of bacterial adherence at sites of EHEC O157:H7-induced ␣-actinin mobilization (arrows). (c) HEp-2 cells depleted of cholesterol (10 mM MCD; 1 h) prior to EHEC O157:H7 infection did not demonstrate focal cytoskeletal rearrangements at sites of diffusely adherent bacteria. (d) Graphical representation of semiquantification of cytoskeletal rearrangements with increasing concentrations of MCD. The findings show a dose-dependent decrease in the formation of EHEC O157:H7-induced ␣-actinin foci (ANOVA, P ⬍ 0.001). The error bars indicate standard errors.
four-well chamber slides (Nalge Nunc International, Naperville, IL) or 10-cmdiameter tissue culture dishes (Starstedt Inc., Montreal, Canada) at a multiplicity of infection of 100 bacteria to 1 eukaryotic cell, for 3 to 6 h at 37°C in antibioticfree tissue culture medium. The cells were then washed six times with phosphatebuffered saline (PBS) to remove nonadherent bacteria and processed further as described below. Cholesterol perturbation. Methyl--cyclodextrin (MCD; Sigma Chemical Co., St. Louis, MO) was employed to remove cholesterol from the plasma membrane and disrupt the function of lipid rafts in eukaryotic cells (24). Prior to bacterial infection, HEp-2 cells were incubated with 1, 3, or 10 mM MCD in antibiotic-free medium for 1 h at 37°C. The medium was aspirated, and the cells were washed with PBS to remove solubilized cholesterol and remaining MCD. To add cholesterol back into cholesterol-depleted HEp-2 cells, 20, 100, or
200 g/ml soluble cholesterol (cholesterol complexes with MCD; Sigma) in antibiotic-free medium was added to cells for 45 min at 37°C. Following depletion/repletion of cholesterol, the cells were washed with PBS before bacterial infection was continued. In another set of experiments, HEp-2 cells were treated with filipin complex (Sigma) in order to determine the effect of cholesterol sequestration on E. coli infection. HEp-2 cells were incubated with 0.1, 1, and 5 g/ml filipin for 1 h at 37°C prior to infection and during infection with EHEC and EPEC. Thin-layer chromatography. Confluent HEp-2 cells grown in 75-cm2 flasks (approximately 6 ⫻ 107 cells per flask) were either left untreated, depleted with MCD, or replenished with cholesterol-MCD complexes. The cells were then lifted from the flask surface with 0.05% trypsin (Life Technologies), pelleted, and washed twice with PBS. Cells were resuspended in 20-ml glass tubes, and cellular
LIPID RAFTS AND ATTACHING-EFFACING E. COLI ADHERENCE
VOL. 73, 2005
FIG. 1—Continued.
lipids were extracted by incubating them in a 2:1 (vol/vol) chloroform-methanol solution overnight at room temperature with gentle shaking. After being filtered to remove precipitated proteins, the cleared solution was subjected to Folch extraction (9). Briefly, distilled water was added to obtain a solution of chloroform-methanol-water (2:1:0.6 [vol/vol/vol]). The tubes were briefly agitated and then allowed to stand at room temperature overnight for phase separation. The lower organic phase, containing cellular lipids, was then aspirated and dried under nitrogen gas. Samples were resuspended in 0.1 ml 2:1 chloroform-methanol, and a 20-l sample was dotted onto a thin-layer chromatogram plate. Free cholesterol (50 g; Sigma) was employed as the reference standard. A 70:30:1 (vol/vol/vol) hexane-diethyl ether-acetic acid developing solution was used to separate the lipids, and the dried plates were stained with iodine vapor to visualize bands corresponding to cholesterol. Cell viability. The LIVE/DEAD Viability/Cytotoxicity assay (Molecular Probes Inc., Eugene, OR) was used to determine the viability of adherent HEp-2 cells after treatment with MCD. HEp-2 cells were seeded onto 22- by 22-mm glass coverslips (VWR Scientific Inc., Media, PA) in multiwell plates at a concentration of 105 cells/well and allowed to adhere overnight at 37°C in 5% CO2. The cells were then either left untreated or depleted of cholesterol, as described above. Coverslips were washed twice with PBS and, according to the manufacturer’s instructions (Molecular Probes), incubated with 2 ml of 1 M calcein AM and 2 M ethidium homodimer 1 in PBS for 40 min at room temperature. The cells were visualized using a Leitz Dialux 22 microscope (Leica Canada, Willowdale, Ontario, Canada) at ⫻100 magnification. Four random fields per coverslip were counted by direct visualization. Immunofluorescence microscopy. HEp-2 cells and human skin fibroblasts were seeded onto Lab-Tek four-well slides (Nalge Nunc) at an approximate density of 105 cells/well and allowed to adhere overnight in 5% CO2 at 37°C. The cells were then washed twice with PBS and transferred into serum-free medium. After infection for 3 to 4 h at 37°C and 5% CO2 and subsequent removal of nonadherent bacteria, the cells were fixed in 100% methanol for 10 min at 4°C. Labeling of the cell cytoskeleton and bacterial antigens by immunostaining was employed to clearly identify the colocalization of attaching-effacing lesions and bacteria. Cytoskeleton rearrangements were detected using anti-␣-actinin mouse immunoglobulin M (IgM) (Sigma) and fluorescein isothiocyanate-labeled donkey anti-mouse IgM (Jackson Immunoresearch Laboratories Inc., West Grove, PA) as described previously (19). In some experiments, cells were probed for the
7115
localization of caveolae with an anti-caveolin-1 antibody (Santa Cruz Biotechnology Inc., Santa Cruz, CA) and rhodamine-conjugated donkey anti-rabbit IgG (Jackson Immunoresearch). Adherent bacteria were visualized with antiEscherichia coli O157:H7 polyclonal goat antibody (Kirkegaard & Perry Laboratories Inc., Gaithersberg, MD) and rhodamine-conjugated donkey anti-goat antibody (Jackson Immunoresearch). Binding of EPEC to cells was detected using rabbit serum obtained from animals immunized with heat-killed EPEC O127:H6 suspended in Freund’s complete adjuvant and rhodamine-conjugated donkey anti-rabbit IgG (Jackson Immunoresearch). Primary and secondary antibodies were both diluted 1:100 in sterile PBS and added separately for 1 h at 37°C or room temperature, respectively. The wells were washed six times with PBS between incubations. Vectashield (Vector Laboratories Inc., Burlingame, CA) mounting medium was then added, and slides were mounted with 22- by 50-mm glass coverslips (VWR) and sealed before being viewed. Samples were viewed under a Leitz Dialux 22 fluorescence microscope (Leica Canada) at ⫻200 and ⫻400 magnifications. Electron microscopy. For scanning electron microscopy, HEp-2 cells were seeded onto 22- by 22-mm glass coverslips (VWR) in multiwell plates at 105 cells/well. The cells were allowed to adhere overnight, washed twice with PBS, and then transferred to antibiotic-free medium. The cells were either left untreated or depleted of cholesterol with 10 mM MCD (1 h at 37°C and 5% CO2) and either left uninfected or infected with EHEC O157:H7 as described above. After infection, the wells were washed six times with PBS and fixed in universal fixative (4% paraformaldehyde, 1% glutaldehyde in 0.1 M phosphate buffer) for at least 1 h. The coverslips were then removed from the multiwell plates and incubated in 2% osmium tetroxide for 1 h at room temperature. The cells were then dehydrated in a graded series of ethanol (50% to 100%), dried through a critical-point dryer, and sputter coated with gold. Samples were viewed with a JSM 820 (Joel USA Corp., Peabody, MA) scanning electron microscope. For transmission electron microscopy, HEp-2 cells were grown in 6-cm-diameter tissue culture dishes (Becton Dickinson and Co., Franklin Lakes, NJ) until confluent. The cells were then left untreated or treated with MCD (1 to 10 mM for 1 h at 37°C) and subsequently infected with EHEC, as described above. After being washed six times in PBS to remove nonadherent bacteria, the monolayers were fixed with 2.5% glutaldehyde in 0.1 M phosphate buffer, pH 7.4, for 10 min. HEp-2 cells were scraped from the tissue culture dishes and pelleted at 600 rpm in the fixative buffer. The HEp-2 cell pellets were next postfixed in 2% aqueous osmium tetroxide for 1 h. Dehydration was then performed in graded acetone, followed by embedding in epoxy resin. Osmium fixation, dehydration, and embedding were conducted in a Pelco Biowave microwave oven (Pelco International, Redding, CA) similar to the procedure described by Giberson et al. (14). One-micrometer-thick sections were stained with toluidine blue, and ultrathin sections were stained with uranyl acetate and lead citrate. Transmission electron microscopy examination was performed under a JEM 1230 (Joel USA Corp., Peabody, MA) transmission electron microscope. Quantification of bacterial adherence. CFU counts were performed to determine the effect of cholesterol depletion on initial bacterial attachment to tissue culture cells. HEp-2 cells were grown in 10-cm-diameter tissue culture dishes (Starstedt) until confluent. The cells were then left untreated or treated with MCD (1 to 10 mM for 1 h at 37°C) and subsequently infected with EHEC or EPEC, as described above. After being washed six times in PBS to remove nonadherent bacteria, HEp-2 cells with adherent bacteria were lysed in distilled water for 5 min at room temperature. Bacteria were then serially diluted in PBS and plated onto McConkey agar, and CFU counts were calculated after overnight growth at 37°C to determine the number of viable bacteria adherent to the tissue culture cells. Semiquantification of bacterially induced cytoskeletal rearrangements. Cytoskeletal-rearrangement events were quantified by visual counting from immunofluorescence photomicrographs. Greater than 100 cells in four random fields containing at least 25 HEp-2 cells stained for ␣-actinin were quantified per well. The results are expressed as the average number of ␣-actinin foci ⫾ standard error per HEp-2 cell in four separate experiments. Statistical analysis. Quantitative and semiquantitative results are expressed as means ⫾ standard errors. Statistical significance was determined by analysis of variance (ANOVA), followed by the Tukey-Kramer multiple-comparison test.
RESULTS HEp-2 cell cholesterol, but not viability, is altered by treatment with MCD. To determine if MCD treatment could affect HEp-2 cell cholesterol levels, lipid extracts of untreated, depleted, and depleted/replenished cells were analyzed by
7116
RIFF ET AL.
INFECT. IMMUN.
FIG. 2. EPEC-induced cytoskeletal rearrangements are disrupted by cholesterol depletion. Immunofluorescence micrographs (a to c) using an ␣-actinin antibody (left column) and EPEC immune serum (middle column) and merged images (right column). (a) Uninfected cells show no cytoskeletal rearrangements and lack bacterial staining. (b) Cells infected with EPEC showed microcolony-type bacterial adherence at sites of EPEC-induced ␣-actinin mobilization (arrows). (c) HEp-2 cells depleted of cholesterol (10 mM MCD; 1 h) prior to EPEC infection show a reduced number of focal cytoskeletal rearrangements at the few sites of bacterial adherence. (d) Graphical representation of semiquantification of cytoskeleton rearrangements with increasing concentrations of MCD. The findings show a dose-dependent decrease in the formation of EPEC-induced ␣-actinin foci (ANOVA, P ⬍ 0.001). The error bars indicate standard errors.
thin-layer chromatography (data not shown). Cholesterol depletion was dose-dependent, since treatment of HEp-2 cells with 1 mM and 3 mM MCD resulted in a slight decrease in cholesterol content, whereas treatment with 10 mM MCD dramatically decreased HEp-2 cholesterol content compared with untreated HEp-2 cells. Addition of soluble cholesterol (20 g/ml, 100 g/ml, and 200 g/ml) increased the HEp-2 cell cholesterol level from the depleted state. The percentage of nonviable cells in treated samples showed only modest increase after treatment with 1 mM MCD (3.8% ⫾ 0.8%), 3 mM MCD (3.3% ⫾ 0.7%), and 10 mM MCD (9.7% ⫾ 1.6%) in comparison with the untreated control (1.6% ⫾ 0.5%).
Cholesterol depletion of HEp-2 cells inhibits cytoskeletal rearrangements induced by both EHEC and EPEC. As shown in Fig. 1a, in the absence of bacteria, the cytoskeletal protein ␣-actinin was distributed throughout the eukaryotic cells in a uniform manner. EHEC O157:H7 infection of HEp-2 cells caused characteristic cytoskeletal rearrangements visible as dense foci of ␣-actinin present directly underneath adherent bacteria (Fig. 1b). At sites of ␣-actinin mobilization, there was colocalization of bacteria and cytoskeletal components, as shown by the presence of both ␣-actinin and bacterial staining in the merged image (Fig. 1b). In contrast, HEp-2 cells depleted of cholesterol with MCD (10 mM; 1 h) showed a
LIPID RAFTS AND ATTACHING-EFFACING E. COLI ADHERENCE
VOL. 73, 2005
FIG. 2—Continued.
reduced number of cytoskeletal-rearrangement events beneath sites of bacterial adhesion (Fig. 1c). The inhibitory effect was quantified by counting the number of dense ␣-actinin foci displayed by infected HEp-2 cells from photomicrographs. As shown in Fig. 1d, MCD inhibited the ability of EHEC O157:H7 to induce host cell cytoskeletal rearrangements in a dose-dependent manner. Similar results were observed following infection of HEp-2 cells with the prototype EPEC strain, E2348/69. Photomicrographs of HEp-2 cells infected with EPEC O127:H6 displayed characteristic microcolony formation and ␣-actinin rearrangements at sites of bacterial adherence (Fig. 2b). Formation of microcolonies and rearrangement of ␣-actinin (Fig. 2c) was reduced in cells depleted of cholesterol by MCD in a dosedependent manner (Fig. 2d). Adhesion of EPEC, but not EHEC, to HEp-2 cells is reduced by cholesterol depletion. Fewer bacteria were adherent to MCD-treated HEp-2 cells following EPEC infection (Fig. 2c) than to untreated cells (Fig. 2c). To confirm this observation, monolayers of HEp-2 cells treated with various doses of MCD (1, 3, and 10 mM; 1 h) were infected with either EHEC or EPEC, and the total number of bacteria bound to the monolayer was assessed by determining the bacterial CFU. As shown in Table 1, the number of EHEC bacteria bound to cholesterol-depleted HEp-2 monolayers remained high (7.64 log10 adherent EHEC bacteria bound to 10 mM MCDtreated monolayers versus 7.87 log10 bound to untreated monolayers). In contrast, EPEC adherence was reduced to approximately 14% of bacterial adherence observed in untreated HEp-2 monolayers (6.46 log10 adherent EPEC bacteria
7117
bound to 10 mM MCD-treated monolayers versus 7.34 log10 bound to untreated monolayers). Replenishment of HEp-2 cells with soluble cholesterol rescues EHEC and EPEC patterns of localized adherence. To confirm that inhibition of ␣-actinin mobilization was attributable to cholesterol removal, cholesterol-depleted HEp-2 cells were replenished with exogenous cholesterol. Replenishing HEp-2 cells with 200 g/ml soluble cholesterol prior to bacterial infection rescued the ability of these bacteria to recruit cytoskeletal components to sites of bacterial adhesion (Fig. 3) to levels comparable to those observed in the positive control. Disruption of caveolae by cholesterol sequestration does not affect attaching-effacing cytoskeleton rearrangements in host cells. The cholesterol-sequestering agent filipin is commonly used to disrupt lipid rafts and caveolae (32, 39, 41), another cholesterol-dependent membrane microdomain involved in protein sorting and membrane trafficking. Unlike MCD, filipin does not remove cholesterol from the plasma membrane but rather binds to cholesterol within the membrane. Treatment of HEp-2 cells with filipin (0.1 to 5 g/ml for 60 min prior to and during infection at 37°C) did not prevent EHEC (Fig. 4a)- or EPEC (Fig. 4b)-induced ␣-actinin recruitment. Furthermore, HEp-2 cells did not show caveolin 1 recruitment to sites of bacterial adhesion and ␣-actinin mobilization (Fig. 4c), suggesting that intact plasma membrane lipid rafts, and not caveolae, are necessary for EHEC- and EPEC-induced cytoskeleton perturbations. Cholesterol depletion prevents effacement of HEp-2 cell surface structures. In addition to cytoskeletal protein rearrangements, bacterial intimate adherence also causes effacement of microvilli on the surfaces of intestinal epithelial cells in vivo (21). The surface of an untreated HEp-2 cell is covered with surface appendages, as visualized by scanning electron microscopy (Fig. 5a). Following EHEC O157:H7 infection, these structures were effaced, revealing a smooth cell surface (Fig. 5b). Cholesterol depletion (10 mM MCD for 1 h at 37°C) alone led to a shortening of appendage length (Fig. 5c). EHEC O157:H7-infected HEp-2 cells that were pretreated with MCD (Fig. 5d) retained these shortened structures, resembling the surface of an uninfected cholesterol-depleted cell. Furthermore, replenishment of HEp-2 cells with 200 g/ml soluble cholesterol restored the surface appendages (Fig. 5e) and rescued the ability of EHEC O157:H7 to efface the surfaces of HEp-2 cells (Fig. 5f). Cholesterol depletion prevents the formation of actin-rich pedestals beneath adherent bacteria. As a complementary approach to further characterize the inhibitory action of host cell
TABLE 1. Adhesion of EHEC and EPEC to HEp-2 cell monolayers after cholesterol depletion No. of adherent bacteriaa MCD (mM)
0 1 3 10 a b c
EHEC
EPEC
7.87 ⫾ 0.03 7.85 ⫾ 0.05 7.79 ⫾ 0.05 7.64 ⫾ 0.04b
7.34 ⫾ 0.06 7.27 ⫾ 0.06 6.75 ⫾ 0.07c 6.46 ⫾ 0.09c
Values are log10 means ⫾ standard error (n ⫽ 3 duplicate samples). P ⬍ 0.01. P ⬍ 0.001.
7118
RIFF ET AL.
INFECT. IMMUN.
FIG. 3. The attaching-effacing phenotype is rescued by the addition of exogenous cholesterol to HEp-2 cells. Immunofluorescence micrographs using an ␣-actinin antibody (left column) and EPEC immune serum (middle column) and merged images (right column). (a) Cells infected with EPEC show the adherence of microcolonies at sites of cytoskeletal rearrangement typical of EPEC-induced attaching-effacing lesions. (b) HEp-2 cells depleted of cholesterol prior to EPEC infection do not possess focal cytoskeletal rearrangements at the few sites of bacterial adherence. (c) HEp-2 cells depleted of cholesterol were replenished and display cytoskeletal rearrangements (arrows), as well as EPEC microcolony adherence patterns comparable to the untreated control. Comparable results were obtained with EHEC O157:H7-infected HEp-2 cells.
cholesterol depletion on EHEC adhesion, transmission electron microscopy was undertaken. Figure 6a shows a crosssectional view of the plasma membrane region of an uninfected and untreated HEp-2 cell. At high magnification, an EHECinfected HEp-2 cell (Fig. 6b) displays distinct actin-rich cups at sites of bacterial intimate adherence and no evidence of surface appendages. In contrast, HEp-2 cells treated with 10 mM MCD for 1 h prior to EHEC infection (Fig. 6d) display surface appendages and no actin-rich cups at sites of bacterial adherence. Addition of exogenous cholesterol (200 g/ml soluble cholesterol; 45 min) restored EHEC-induced actin-rich-
cup formation (Fig. 6f) characteristic of attaching-effacing lesions. Niemann-Pick type C fibroblasts are resistant to attachingeffacing E. coli-induced cytoskeleton rearrangements. Skin fibroblasts from human patients with NPC disease are characterized by a cholesterol-trafficking defect whereby cholesterolenriched microdomains are reduced at the plasma membrane (12). Fibroblasts from NPC patients were analyzed for cholesterol ester (CE) synthesis to confirm their phenotype (2). Two NPC cell lines, 15055 and 16934, showed decreased CE synthesis (1.6 ⫾ 1.3 and 4.1 ⫾ 0.5 nmol/24 h/mg protein, respec-
VOL. 73, 2005
LIPID RAFTS AND ATTACHING-EFFACING E. COLI ADHERENCE
7119
FIG. 4. Bacterially induced cytoskeleton mobilization events are caveola independent. (a and b) Immunofluorescence micrographs stained for ␣-actinin, EHEC, or EPEC and merged images. HEp-2 cells were treated with 1 mg/ml filipin for 1 h prior to infection and during infection at 37°C. Filipin treatment did not prevent EHEC O157:H7-induced (a) or EPEC-induced (b) cytoskeletal rearrangements. (c) Immunofluorescence micrographs stained for ␣-actinin and caveolin 1 and light microscopy photomicrographs show bacterial adherence to HEp-2 cells. EHEC O157:H7-induced cytoskeletal rearrangements (arrows) did not display recruitment of caveolin 1 (middle) to sites of bacterial adhesion (right). Comparable results were seen during EPEC infection.
tively) compared to two control cell lines from non-NPC patients, 4993 and 15215, which showed high CE synthesis (38.4 ⫾ 6.2 and 54.8 ⫾ 2.2 nmol/24 h/mg protein, respectively). Cytoskeletal rearrangements were induced by both EHEC (Fig. 7b) and EPEC (Fig. 7c) on control primary fibroblasts. By contrast, EHEC (Fig. 7e) and EPEC (Fig. 7f) did not induce foci of cytoskeletal components beneath adherent bacteria on NPC fibroblasts.
DISCUSSION In this study, the critical role of host cell plasma membrane cholesterol in bacterial adhesion and attaching-effacing lesion formation in response to the diarrheal pathogens enterohemorrhagic E. coli O157:H7 and enteropathogenic E. coli O127:H6 has been demonstrated for the first time. Treatment of HEp-2 cells with MCD removed cholesterol, did not affect
FIG. 5. EHEC O157:H7-induced HEp-2 cell surface effacement is prevented following cholesterol depletion. Scanning electron photomicrographs of uninfected (left column) and EHEC O157:H7-infected (right column) HEp-2 cells. (a) Untreated HEp-2 cells display appendages covering the surface of the cell, while EHEC O157:H7-infected cells (b) demonstrate a smooth cell surface. HEp-2 cells depleted of cholesterol possess shortened surface structures covering both (c) uninfected and (d) EHEC-infected cells. HEp-2 cells depleted of cholesterol using MCD and then replenished with exogenous cholesterol (e) displayed structures approaching those of untreated cells, which are effaced following EHEC adhesion (f). Approximate magnification, ⫻6,000. 7120
FIG. 6. EHEC O157:H7-induced actin pedestal formation is prevented following cholesterol depletion. Transmission electron photomicrographs of HEp-2 cells (a and b) untreated, (c and d) cholesterol depleted (10 mM MCD; 1 h), and (e and f) replenished with exogenous cholesterol (200 mg/ml; 45 min). HEp-2 cells infected with EHEC O157:H7 (b) possess distinct actin-rich cup-like pedestals (arrows) underlying intimately adherent bacteria. By contrast, dense actin-rich cups are not present at the sites of EHEC adhesion to HEp-2 cells depleted of cholesterol (d). Replenishment of HEp-2 cells previously depleted of cholesterol rescues the ability of EHEC to recruit dense foci of actin (arrows). Bar, 500 nm. 7121
7122
RIFF ET AL.
INFECT. IMMUN.
VOL. 73, 2005
LIPID RAFTS AND ATTACHING-EFFACING E. COLI ADHERENCE
host cell viability, and yet prevented bacterially induced rearrangement of the host cytoskeleton. This effect was dose dependent and reversible by reintroducing cholesterol into depleted HEp-2 cells. Cholesterol depletion of HEp-2 cells did not affect the initial adhesion of EHEC to HEp-2 cells, whereas EPEC adherence was reduced. This indicates that membrane cholesterol may differentially regulate the binding of EHEC and EPEC to host cell surfaces. While removal of cholesterol by MCD showed inhibitory effects, filipin did not prevent EHEC O157:H7-induced cytoskeleton rearrangements, nor was caveolin 1 recruited to these sites of bacterial adhesion. This demonstrates that the caveolar subclass of lipid rafts is not involved in attaching-effacing pathogenesis. Although lipid raft-associated components are reported to localize to sites of EPEC binding (16, 42), this is the first report demonstrating that removal of cholesterol, a required structural element of lipid rafts (41), from host cells inhibits attachingeffacing lesion formation. The necessity for cholesterol was confirmed using a naturally occurring mutant human cell with decreased lipid raft cholesterol (12). Taken together, these results indicate that intact sphingolipid/cholesterol-enriched microdomains are required for attaching-effacing lesion formation in response to both EHEC and EPEC infections. In the present study, while the number of EHEC- and EPEC-induced ␣-actinin foci was reduced following cholesterol depletion of host cells, EPEC, but not EHEC, adhesion to HEp-2 cells was decreased. These findings indicate that while only intimate adherence of EHEC O157:H7 is regulated by cholesterol, cholesterol may function in the regulation of both initial binding and intimate adhesion of EPEC. Precisely how cholesterol mediates adhesion of EPEC is not known. While EPEC does not directly bind to cholesterol (1), manipulation of the cholesterol concentration in cells could indirectly affect lipid raft localization of an EPEC receptor. Such speculation is not without precedent, since studies of the oxytocin receptor have shown that binding of oxytocin is dependent on membrane interactions of the receptor with cholesterol in lipid rafts (15). Alternatively, cholesterol depletion could result in the efflux of other lipids, such as phosphatidylethanolamine, a lipid that has been shown to be a receptor of EPEC adherence (1) which is released after cholesterol depletion (33). On the other hand, EPEC forms characteristic microcolonies mediated by interbacterial binding of the bundle-forming pilus (21, 26). A decrease in EPEC microcolony formation on cholesterol-depleted HEp-2 cells could be responsible for the reduced adherence observed. The cholesterol depletion agent MCD has been used in other signal transduction-dependent systems to study the ef-
7123
fects of cholesterol removal on lipid raft function, including epidermal growth factor receptor (18) and T-cell receptor signaling (31). It has recently been appreciated that lipid raft function is important for microbial pathogens, including human immunodeficiency virus infection of CD4⫹ T cells (38) and Mycobacterium tuberculosis infection of macrophages (13), as well as bacterial toxins, such as cholera toxin (29) and vacuolating cytotoxin A (34). MCD-mediated cholesterol efflux has been described as highly specific compared to other cyclodextrins (24), but it may also induce efflux of other membrane lipids, including glycosphingolipids, as well as induce cytotoxicity (33). Since previous studies have demonstrated variable effects on the viability of cells after depletion of cholesterol by MCD, the viability of MCD-treated HEp-2 cells was compared to that of untreated controls. As cholesterol depletion of epithelial cells was reversible by addition of exogenous cholesterol without inducing cytotoxicity, it is reasonable to conclude that the inhibitory effects on host cytoskeleton recruitment are specifically due to the removal of cholesterol. Signal transduction responses and remodeling of the cytoskeleton in host epithelial cells are both required for formation of the attaching-effacing lesion. Dense foci of cytoskeletal proteins, including F-actin, ␣-actinin, talin, and ezrin, aggregate immediately beneath adherent EHEC and EPEC (16). Activation of signaling pathways is involved in reorganization of the cytoskeleton, including phosphatidylinositol 3⬘ kinase and phospholipase C-gamma (19). Membrane-associated EHEC and EPEC effectors, such as EspE/Tir and EspB, may associate with lipid rafts, either directly or indirectly, to exert local effects at sites of bacterial adhesion. Pathogenic bacteria commonly inject type III secreted proteins into host cells to usurp or disrupt signaling pathways to induce cytoskeletondependent internalization of bacteria (3) or prevent phagocytosis by macrophages (20). Many of the signaling molecules affected, including small G proteins, such as Rac and Cdc42, and phosphatidylinositides, are lipid raft dependent (8). Salmonella enterica type III effectors, PipB and PipB2, are enriched in detergent-resistant microdomains (25). As a complementary approach to depletion of cholesterol with MCD, primary human skin fibroblasts from patients with NPC disease were infected with EHEC and EPEC and assessed for attaching-effacing lesion formation. NPC is a rare, heritable, and fatal neurodegenerative disorder affecting humans (35). Mutations of NPC1 or NPC2 lead to a dysfunction in intracellular cholesterol trafficking, resulting in high internal cholesterol levels retained within lysosomal compartments and decreased cholesterol in the trans-Golgi network (30). Al-
FIG. 7. Cytoskeletal rearrangements formed in response to EHEC and EPEC infections are reduced on Niemann-Pick type C primary human skin fibroblasts compared with wild-type fibroblasts. Immunofluorescence micrographs using an ␣-actinin antibody (left column) and bacterial antiserum (middle column) and merged images (right column). Results using fibroblasts from an unaffected individual (cell line 4993) are shown in panels a to c. Results using fibroblasts derived from a patient with NPC (cell line 16934) are presented in panels d to f. (a) Uninfected cells show no cytoskeletal rearrangements and lack staining with primary antibodies against bacteria. (b) Cells infected with EHEC show bacterial adherence at sites of cytoskeletal rearrangement typical of EHEC-induced attaching-effacing lesions (arrows). (c) Cells infected with EPEC also demonstrated microcolonies adherent at sites of cytoskeletal rearrangement characteristic of EPEC-induced attaching-effacing lesions. (d) Uninfected NPC cells show no cytoskeletal rearrangements and lack bacterial staining. NPC fibroblasts infected with EHEC (e) or EPEC (f) do not possess foci of cytoskeleton proteins at sites of bacterial adhesion. Comparable results were observed using two additional skin fibroblast cell lines (wild-type 15215 and NPC cell line 15055).
7124
RIFF ET AL.
INFECT. IMMUN.
though the total plasma membrane cholesterol contents in npc1⫺/⫺ cells are similar to those in wild-type cells, the cholesterol contents of lipid rafts isolated from npc1⫺/⫺ cells are markedly reduced (12). Thus, npc1⫺/⫺ cells are an effective tool to study lipid raft-dependent processes. Both EHEC and EPEC were attenuated in their ability to recruit foci of ␣-actinin to sites of bacterial adhesion on npc1⫺/⫺ fibroblasts in comparison with control cells. This observation also indicates that intact plasma membrane cholesterol-enriched microdomains are required for EHEC and EPEC intimate attachment. In summary, this study has shown that perturbation of plasma membrane cholesterol in host cells reduces EHEC- and EPEC-induced cytoskeletal alterations. Since bacterial adherence and attaching-effacing lesion formation are important for eliciting human disease, this study highlights the role of cholesterol-enriched microdomains in the pathobiology of disease. Developing a more precise understanding of the molecular mechanisms underlying EHEC and EPEC infections should aid in the development of novel intervention strategies. ACKNOWLEDGMENTS We thank Clifford Lingwood and Anita Nutikka for assistance with thin-layer chromatography procedures, Nancy Cracknell for assistance with primary fibroblast culturing procedures, and Yew Meng Heng and Julia Huang for assistance with electron microscopy. We also thank Danny Aguilar for assistance with preparation of the figures and Peter Ceponis for critical review of the manuscript. This work was supported by an operating grant from the Canadian Institute for Health Research. J.D.R. is supported through a studentship by the Ontario Student Opportunity Trust Fund-Hospital for Sick Children Foundation Student Scholarship Program and a University of Toronto Fellowship. P.M.S. is the recipient of a Canada Research Chair in Gastrointestinal Disease. REFERENCES 1. Barnett-Foster, D., D. Philpott, M. Abul-Milh, M. Huesca, P. M. Sherman, and C. A. Lingwood. 1999. Phosphatidylethanolamine recognition promotes enteropathogenic E. coli and enterohemorrhagic E. coli host cell attachment. Microb. Pathog. 27:289–301. 2. Bowler, L. M., R. Shankaran, I. Das, and J. W. Callahan. 1990. Cholesterol esterification and Niemann-Pick disease: an approach to identifying the defect in fibroblast. J. Neurosci. Res. 27:505–511. 3. Brumell, J. H., and S. Grinstein. 2003. Role of lipid-mediated signal transduction in bacterial internalization. Cell Microbiol. 5:287–297. 4. Ceponis, P. J., D. M. McKay, J. C. Ching, P. Pereira, and P. M. Sherman. 2003. Enterohemorrhagic Escherichia coli O157:H7 disrupts Stat1-mediated gamma interferon signal transduction in epithelial cells. Infect. Immun. 71:1396–1404. 5. DeVinney, R., M. Stein, D. Reinscheid, A. Abe, S. Ruschkowski, and B. B. Finlay. 1999. Enterohemorrhagic Escherichia coli O157:H7 produces Tir, which is translocated to the host cell membrane but is not tyrosine phosphorylated. Infect. Immun. 67:2389–2398. 6. Duncan, M. J., J. S. Shin, and S. N. Abraham. 2002. Microbial entry through caveolae: variations on a theme. Cell Microbiol. 4:783–791. 7. Elliott, S. J., L. A. Wainwright, T. K. McDaniel, K. G. Jarvis, Y. K. Deng, L. C. Lai, B. P. McNamara, M. S. Donnenberg, and J. B. Kaper. 1998. The complete sequence of the locus of enterocyte effacement (LEE) from enteropathogenic Escherichia coli E2348/69. Mol. Microbiol. 28:1–4. 8. Fielding, C. J., and P. E. Fielding. 2004. Membrane cholesterol and the regulation of signal transduction. Biochem. Soc. Trans. 32:65–69. 9. Folch, J., M. Lees, and G. H. Sloane Stanley. 1957. A simple method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem. 226:497–509. 10. Foster, L. J., C. L. De Hoog, and M. Mann. 2003. Unbiased quantitative proteomics of lipid rafts reveals high specificity for signaling factors. Proc. Natl. Acad. Sci. USA 100:5813–5818. 11. Galbiati, F., B. Razani, and M. P. Lisanti. 2001. Emerging themes in lipid rafts and caveolae. Cell 106:403–411. 12. Garver, W. S., K. Krishnan, J. R. Gallagos, M. Michikawa, G. A. Francis, and R. A. Heiderenreich. 2002. Niemann-Pick C1 protein regulates cholesterol transport to the trans-Golgi network and plasma membrane caveolae. J. Lipid Res. 43:579–589.
13. Gatfield, J., and J. Pieters. 2000. Essential role for cholesterol in entry of mycobacteria into macrophages. Science 288:1647–1650. 14. Giberson, R. T., R. L. Austin, J. Charlesworth, G. Adamson, and A. H. Guillermo. 2003. Microwave and digital imagin technology reduce turnaround times for diagnostic electron microscopy. Ultrastruct. Pathol. 27:1–10. 15. Gimpl, G., V. Wiegand, K. Burger, and F. Fahrenholz. 2002. Cholesterol and steroid hormones: modulators of oxytocin receptor function. Prog. Brain Res. 139:43–55. 16. Goosney, D. L., R. DeVinney, R. A. Pfuetzner, E. A. Frey, N. C. Strynadka, and B. B. Finlay. 2000. Enteropathogenic E. coli translocated intimin receptor, Tir, interacts directly with alpha-actinin. Curr. Biol. 10:735–738. 17. Hrudey, S. E., P. Payment, P. M. Huck, R. W. Gillham, and E. J. Hrudey. 2003. A fatal waterborne disease epidemic in Walkerton, Ontario: comparison with other waterborne outbreaks in the developed world. Water Sci. Technol. 47:7–14. 18. Hur, E. M., Y. S. Park, B. D. Lee, I. H. Jang, H. S. Kim, T. D. Kim, P. G. Suh, S. H. Ryu, and K. T. Kim. 2004. Sensitization of epidermal growth factorinduced signaling by bradykinin is mediated by c-Src. Implications for a role of lipid microdomains. J. Biol. Chem. 279:5852–5860. 19. Johnson-Henry, K., J. L. Wallace, N. S. Basappa, R. Soni, G. K. Wu, and P. M. Sherman. 2001. Inhibition of attaching and effacing lesion formation following enteropathogenic Escherichia coli and Shiga toxin-producing E. coli infection. Infect. Immun. 69:7152–7158. 20. Juris, S. J., F. Shao, and J. E. Dixon. 2002. Yersinia effectors target mammalian signalling pathways. Cell Microbiol. 4:201–211. 21. Kaper, J. B., J. P. Nataro, and H. L. Mobley. 2004. Pathogenic Escherichia coli. Nat. Rev. Microbiol. 2:123–140. 22. Kenny, B. 2002. Mechanisms of action of EPEC type III effector molecules. Int. J. Med. Microbiol. 291:469–477. 23. Kenny, B., R. DeVinney, M. Stein, D. J. Reinscheid, E. A. Frey, and B. B. Finlay. 1997. Enteropathogenic E. coli (EPEC) transfers its receptor for intimate adherence into mammalian cells. Cell 91:511–520. 24. Kilsdonk, E. P., P. G. Yancey, G. W. Stoudt, F. W. Bangerter, W. J. Johnson, M. C. Phillips, and G. H. Rothblat. 1995. Cellular cholesterol efflux mediated by cyclodextrins. J. Biol. Chem. 270:17250–17256. 25. Knodler, L. A., B. A. Vallance, M. Hensel, D. Jackel, B. B. Finlay, and O. Steele-Mortimer. 2003. Salmonella type III effectors PipB and PipB2 are targeted to detergent-resistant microdomains on internal host cell membranes. Mol. Microbiol. 49:685–704. 26. Knutton, S., R. K. Shaw, R. P. Anantha, M. S. Donnenberg, and A. A. Zorgani. 1999. The type IV bundle-forming pilus of enteropathogenic Escherichia coli undergoes dramatic alterations in structure associated with bacterial adherence, aggregation and dispersal. Mol. Microbiol. 33:499–509. 27. Lafont, F., and F. G. van der Goot. 2005. Bacterial invasion via lipid rafts. Cell Microbiol. 7:613–620. 28. Laude, A. J., and I. A. Prior. 2004. Plasma membrane microdomains: organization, function and trafficking. Mol. Membr. Biol. 21:193–205. 29. Lencer, W. I. 2001. Microbes and microbial toxins: paradigms for microbialmucosal toxins. V. Cholera: invasion of the intestinal epithelial barrier by a stably folded protein toxin. Am. J. Physiol. Gastrointest. Liver Physiol. 280: G781–G786. 30. Liscum, L., and S. L. Sturley. 2004. Intracellular trafficking of Niemann-Pick C proteins 1 and 2: obligate components of subcellular lipid transport. Biochim. Biophys. Acta 1685:22–27. 31. Nagafuku, M., K. Kabayama, D. Oka, A. Kato, S. Tani-Ichi, Y. Shimada, Y. Ohno-Iwashita, S. Yamasaki, T. Saito, K. Iwabuchi, T. Hamaoka, J. Inokuchi, and A. Kosugi. 2003. Reduction of glycosphingolipid levels in lipid rafts affects the expression state and function of glycosylphosphatidylinositolanchored proteins but does not impair signal transduction via the T cell receptor. J. Biol. Chem. 278:51920–51927. 32. Orlandi, P. A., and P. H. Fishman. 1998. Filipin-dependent inhibition of cholera toxin: evidence for toxin internalization and activation through caveolae-like domains. J. Cell Biol. 141:905–915. 33. Ottico, E., A. Prinetti, S. Prioni, C. Giannotta, L. Basso, V. Chigorno, and S. Sonnino. 2003. Dynamics of membrane lipid domains in neuronal cells differentiated in culture. J. Lipid Res. 44:2142–2151. 34. Patel, H. K., D. C. Willhite, R. M. Patel, D. Ye, C. L. Williams, E. M. Torres, K. B. Marty, R. A. MacDonald, and S. R. Blanke. 2002. Plasma membrane cholesterol modulates cellular vacuolation induced by the Helicobacter pylori vacuolating cytotoxin. Infect. Immun. 70:4112–4123. 35. Patterson, M. C. 2003. A riddle wrapped in a mystery: understanding Niemann-Pick disease, type C. Neurologist 9:301–310. 36. Perna, N. T., G. F. Mayhew, G. Posfai, S. Elliott, M. S. Donnenberg, J. B. Kaper, and F. R. Blattner. 1998. Molecular evolution of a pathogenicity island from enterohemorrhagic Escherichia coli O157:H7. Infect. Immun. 66:3810–3817. 37. Pike, L. J. 2004. Lipid rafts: heterogeneity on the high seas. Biochem. J. 378:281–292. 38. Popik, W., T. M. Alce, and W. C. Au. 2002. Human immunodeficiency virus type 1 uses lipid raft-colocalized CD4 and chemokine receptors for productive entry into CD4⫹ T cells. J. Virol. 76:4709–4722.
VOL. 73, 2005
LIPID RAFTS AND ATTACHING-EFFACING E. COLI ADHERENCE
39. Schnitzer, J. E., P. Oh, E. Pinney, and J. Allard. 1994. Filipin-sensitive caveolae-mediated transport in endothelium: reduced transcytosis, scavenger endocytosis, and capillary permeability of select macromolecules. J. Cell Biol. 127:1217–1232. 40. Simons, K., and R. Ehehalt. 2002. Cholesterol, lipid rafts and disease. J. Clin. Investig. 110:597–603.
Editor: V. J. DiRita
7125
41. Simons, K., and D. Toomre. 2000. Lipid rafts and signal transduction. Nat. Rev. Mol. Cell Biol. 1:31–39. 42. Zobiack, N., U. Rescher, S. Laarmann, S. Michgehl, M. A. Schmidt, and V. Gerke. 2002. Cell-surface attachment of pedestal-forming enteropathogenic E. coli induces a clustering of raft components and a recruitment of annexin 2. J. Cell Sci. 115:91–98.
