|Home | About | Journals | Submit | Contact Us | Français|
The enterovirulent Escherichia coli strains potentially involved in inflammatory bowel diseases include diffusely adherent strains expressing Afa/Dr fimbriae (Afa/Dr DAEC). We have previously observed type 1 pilus-mediated interleukin-8 (IL-8) hyperproduction in infected neutrophils. As pathogen induction of host cell death programs and clearance of apoptotic infected cells are crucial for innate immune system homeostasis and host integrity, we examined modulation of neutrophil cell death by Afa/Dr DAEC. Using the human PLB-985 cell line differentiated into fully mature neutrophils, we found that the wild-type enterovirulent E. coli strain C1845 and the recombinant strain DH5α/pF1845 (expressing the fimbrial adhesin F1845) similarly induced time-dependent phosphatidylserine (PS) externalization, suggesting a major specific role of this virulence factor. Using small interfering RNA (siRNA) decay-accelerating factor (DAF)-transfected PLB-985 cells, we then showed that this PS externalization was triggered in part by glycosylphosphatidylinositol (GPI)-anchored DAF receptor engagement (leading to tyrosine kinase and protein kinase C activation) and that it required cytoskeleton and lipid raft architectural integrity. PS externalization under these conditions was not dependent on caspases, mitochondria, lysosomes, or reactive oxygen or nitrogen species. F1845-mediated PS externalization was sufficient to enable macrophage engulfment of infected differentiated PLB-985 cells. These findings provide new insights into the neutrophil response to Afa/Dr DAEC infection and highlight a new role for F1845 fimbriae. Interestingly, although apoptosis pathways were not engaged, C1845-infected PLB-985 cells displayed enhanced removal by macrophages, a process that may participate in the resolution of Afa/Dr DAEC infection and related inflammation.
Strains of diffusely adhering Escherichia coli expressing Afa/Dr fimbriae (Afa/Dr DAEC) belong to group 6 of the pathogenic E. coli (35). They can cause childhood diarrhea and are responsible for one-third of recurrent urinary tract infections in adults (53). In vitro, enteric wild-type Afa/Dr DAEC strain C1845, which bears F1845 fimbriae, triggers decay-accelerating factor (DAF)-dependent and mitogen-activated protein kinase (MAPK)-dependent interleukin-8 (IL-8) synthesis by polarized colonic epithelial T84 cell monolayers. This leads to transepithelial migration of human polymorphonuclear neutrophils (PMN), which in turn induce epithelial synthesis of the proinflammatory cytokines tumor necrosis factor alpha (TNF-α) and IL-1β (4, 5). These interactions between PMN and apical enterovirulent E. coli colonizing the intestinal brush border were documented only recently. First, Brest et al. obtained evidence that Afa/Dr DAEC could modulate PMN apoptosis and was inefficiently engulfed by PMN (8). Then, using the human myeloid cell line PLB-985 differentiated into fully mature PMN, our group found that Afa/Dr DAEC could activate PMN, triggering an oxidative burst and rapid release of preformed myeloperoxidase and IL-8, followed by IL-1α, TNF-α, and IL-8 synthesis (52); type 1 pili were identified as the promoting bacterial virulence factor, and DAF was identified as the PMN membrane-bound receptor that triggers cell signaling via Erk1/2 and p38 MAPKs, Src tyrosine kinase, and NF-κB.
Proinflammatory responses might contribute to inducing and perpetuating local gut inflammation. Indeed, delayed death and clearance of infected PMN in tissues can cause exaggerated inflammation and prolonged infection (15); in particular, enzymes and reactive oxygen species (ROS) produced by PMN can damage surrounding tissues. Alternatively, a decrease in the PMN life span due to rapid apoptosis can be a contributing factor in severe and recurrent infections (39). PMN become apoptotic and are then recognized, engulfed, and cleared by professional phagocytes, such as tissue macrophages, which prevents them from releasing their toxic contents (22). The detection, recognition, and ingestion of apoptotic cells involve at least three “eat me” molecules, namely, phosphatidylserine (PS), endocytic receptors, and soluble molecules bridging apoptotic PMN and macrophages (33). A nonapoptotic PS externalization mechanism has also been described, which allows PMN engulfment by macrophages in certain conditions (34, 54, 58). Many microbial pathogens have evolved to circumvent PMN attack through six main strategies: activation of survival and stress responses, contact avoidance, phagocytosis prevention, intracellular survival, PMN death induction, and evasion of PMN extracellular traps (32, 60). Pathogen-induced stimulation of host cell death pathways may eliminate key immune cells or be involved in evasion of other host defenses, while, on the other hand, suppression of death pathways may facilitate the proliferation of intracellular pathogens (20, 36).
Here, we report that E. coli wild-type strain C1845 and its recombinant counterpart DH5α/pF1845, which harbors a plasmid encoding F1845 fimbriae, similarly induce time-dependent PS externalization on differentiated PLB-985 cells, suggesting a role for the F1845 adhesin. Further investigation showed that PS externalization followed interaction between F1845 fimbriae and PLB-985 cell membrane-bound DAF. F1845-induced DAF-dependent PS externalization involved tyrosine kinase and protein kinase C (PKC) activation and required cytoskeleton and lipid raft integrity. We also found that PS externalization was not related to three of the main apoptosis pathways (caspase activation and the mitochondrial and lysosomal pathways) or to release of reactive oxygen or nitrogen species. Finally, we showed that the nonapoptotic PS externalization enabled macrophage engulfment of infected PLB-985 cells. Together, these results suggest that this PMN response could participate in resolution of Afa/Dr DAEC infection and the related inflammation.
Phorbol myristate acetate (PMA), all-trans retinoic acid (ATRA), 2′,7′-dichlorofluorescin diacetate (DCFH-DA), staurosporine, and DNA-binding Hoechst stain were obtained from Sigma-Aldrich, St. Quentin Fallavier, France. Annexin V and 5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraethylbenzimidazol carbocyanine iodide (JC-1) were obtained from BD Biosciences. Acridine orange (AO) and propidium iodide were obtained from Invitrogen (Cergy-Pontoise, France). N,N-Dimethyl formamid (DMF) was obtained from Carbo Erba (Rodano, Italy). The caspase inhibitor z-Val-Ala-Asp-fluoromethylketone (zVAD-fmk) and the cathepsin inhibitors trans-epoxysuccinyl-l-leucylamindo-3-methylbutane ethyl ester (EST, E-64d) and pepstatin A were obtained from Calbiochem (San Diego, CA). The mitochondrial NADPH oxidase inhibitor rotenone and the NO synthase inhibitor N-nitro-l-arginine methyl ester (L-NAME) were obtained from Sigma. The following signaling inhibitors were obtained from Sigma: the protein kinase C (PKC) inhibitor staurosporine, the tyrosine kinase inhibitor genistein, the phosphatidylinositol 3-kinase (PI3K) inhibitor LY-294,002, the lipid raft-modifying drug methyl-β-cyclodextrin (MβCD), and the actin cytoskeleton inhibitor cytochalasin D. The Src family kinase inhibitor PP2 and the MEK/ERK inhibitor PD 98059 were obtained from Calbiochem, and the p38 MAPK inhibitor SB-203580 was obtained from Cell Signaling Technology (Beverly, MA).
The human myeloid leukemia cell line PLB-985 (wild-type PLB-985) (59) was a generous gift from M. A. Gougerot-Pocidalo and E. Pedruzzi (IB2M Inserm, CHU Bichat, Paris, France). X-CGD cells (30), prepared from wild-type PLB-985 cells after disruption of the X chromosome-linked gp91-phox gene (65), were a generous gift from M. J. Stasia (CHU, Grenoble, France) and Mary Dinauer (Indiana University School of Medicine, Indianapolis, IN). The cells were cultured as previously described (52). To induce granulocytic differentiation, exponentially growing wild-type PLB-985 cells (starting density, 2 × 105 cells/ml) were cultured in RPMI-1640 medium (BioWittaker, Lonza, Verviers, Belgium) supplemented with 2 μM ATRA, 0.5% DMF, and 5% fetal calf serum (FCS) (BioWittaker). X-CGD cells were differentiated in the presence of 0.5% DMF and 10% FCS (40). On day 6, granulocytic differentiation was checked by morphological analysis of cytocentrifuged cells stained with May-Grünwald-Giemsa stain and by measuring CD11b expression at the cell surface and hydrogen peroxide production, as previously described (52). Cell viability was checked before each experiment by measuring trypan blue exclusion and was routinely >95%.
Blood samples were obtained from informed healthy volunteers. As previously described (26), leukocytes were isolated by sedimentation on a separating medium containing 9% dextran T500 (Pharmacia, Uppsala, Sweden) and 38% Radioselectan (Schering, Lys-lez-Lannoy, France). After red cell sedimentation the leukocyte-rich suspension was centrifuged on a Ficoll density gradient (Eurobio, Les Ulis, France). Contaminating erythrocytes were then removed by hypotonic lysis, and leukocytes were resuspended in Hanks balanced salt solution (HBSS) (Ca2+ Mg2+; Life Technologies, San Diego, CA) or phosphate-buffered saline (PBS) (bioMérieux, Marcy l'Etoile, France) depending on the experiment.
We used two wild-type (WT) Afa/Dr DAEC strains, strains C1845 and IH11128, expressing the fimbrial adhesin F1845 and Dr, respectively, an insertional dra mutant of WT strain IH11128 lacking adhesin Dr (WT-IH11128ΔDr; E. coli strain DR14) (24), and recombinant E. coli strain DH5α/pF1845 expressing F1845 fimbriae (6). E. coli laboratory strain DH5α was used as a nonvirulent control. Stock cultures were maintained in 10% glycerol at −80°C. Before experiments, bacteria were transferred to fresh Luria-Bertani (LB) agar (Difco, Invitrogen) and incubated at 37°C for 24 h. For each experiment, bacteria were subcultured in LB broth at 37°C for 18 h with appropriate antibiotics and washed twice with sterile PBS, and the concentration was adjusted to 5 × 108 bacteria/ml.
Expression of β2-integrin (CD11b/CD18), DAF (CD55), CEACAM1 (CD66a), and CEACAM3 (CD66d) was measured after 30 min of incubation at 4°C, using phycoerythrin (PE)-conjugated anti-CD11b (DakoCytomation, Glostrup, Denmark), PE-conjugated anti-CD55 (BD Biosciences, Le Pont de Claix, France), anti-CEACAM1 (R&D Systems, Abingdon, United Kingdom), and anti-CD66d/CD66e (COL-1, Invitrogen) mouse monoclonal antibodies, respectively. After one wash with ice-cold PBS, goat F(ab′)2 anti-mouse IgG-PE (R&D Systems) or goat F(ab′)2 anti-mouse Alexa 488 (Invitrogen) was added to label the anti-CEACAM1 and anti-CEACAM3 antibodies, respectively, and the preparations were incubated for an additional 30 min at 4°C. After one wash with ice-cold PBS, the cells were resuspended in 1% paraformaldehyde-PBS and kept on ice until flow cytometry was performed. Nonspecific binding was determined by using irrelevant antibodies with the same isotypes. We used a FACSCalibur flow cytometer equipped with a 15-mV, 488-nm argon laser (BD Biosciences). The data were analyzed with CellQuest software. Fluorescence was recorded with a constant photomultiplier gain, and the results were expressed as the mean fluorescence intensity (MFI) using a 4-decade logarithmic scale.
Differentiated PLB-985 cells and isolated blood PMN (106 cells/ml of HBSS) were separately preincubated for 15 min with 1.25 μM DCFH-DA in a water bath with gentle shaking at 37°C. The cells were then stimulated for 15 min with 50 ng/ml PMA. H2O2 production was measured with a FACSCalibur device as described above. A stimulation index (SI) was calculated by determining the ratio of the MFI of PMA-treated cells to the MFI of unstimulated cells.
The concentrations of bacteria (C1845, DH5α, and DH5α/pF1845) were adjusted to obtain a final concentration of 5 × 108/ml in a buffer containing 50 mM sodium carbonate, 100 mM NaCl, and 100 μg/ml fluorescein isothiocyanate (FITC) (Sigma-Aldrich), and the preparations were incubated for 30 min at 37°C with protection from light. The labeled bacteria were then washed, and the concentration in RPMI medium supplemented with 2% FCS was adjusted to 5 × 107/ml. Differentiated PLB-985 cells (106 cells/ml) were placed on glass coverslips in 24-well tissue culture plates and allowed to adhere for 1 h at 37°C. The medium was then removed, and freshly prepared labeled bacteria were added. The plates were centrifuged at 140 × g for 5 min to synchronize the infections and incubated at 37°C in humidified air-7% CO2 for 2 h. The coverslips were then washed to remove nonadherent bacteria and fixed with paraformaldehyde. DNA-binding Hoechst stain (2.5 μg/ml) was then added, and the preparations were incubated for 30 min. After washes, the coverslips were mounted in fluorescent mounting medium (DakoCytomation), and epifluorescence was examined with a Leitz Aristoplan microscope (Leica). The number of bacteria per cell was determined for each strain.
To analyze PS externalization, PLB-985 cells (wild type, X-CGD, and DAF small interfering RNA [siRNA] transfected) were resuspended in RPMI-1640 medium supplemented with 2% FCS at a density of 106 cells/ml and challenged with the different E. coli strains (5 × 107 bacteria), PMA (10 ng/ml), or 1 μM staurosporine in low-attachment Costar plates (VWR International, Fontenay Sous/bois, France) at 37°C in the presence of 5% CO2. Bacterial growth was blocked with gentamicin (200 μg/ml; Invitrogen), which was added 1 h after stimulation and kept in the cultures until the end of the incubation period. A 16-h time course was used.
In selected experiments, PLB-985 cells were pretreated for 1 h with the following inhibitors of caspases, cathepsins, and other cell signaling pathways: zVAD-fmk (25 μM), EST (100 μM), pepstatin A (100 μM), LY-294,002 (50 μM), SB-203580 (7.5 μM), PD-98059 (20 μM), PP2 (7 μM), genistein (200 μM), staurosporine (20 nM), MβCD (3 mM), and cytochalasin D (10 μg/ml). The mitochondrial NADPH oxidase inhibitor rotenone (125 μM) and the nitric oxide inhibitor L-NAME (10 mM) were also used. The interaction between the F1845 fimbrial adhesin and its DAF receptor was studied by using the anti-DAF blocking antibody IH4 (20 μg/ml; D. M. Lublin, Washington University, St. Louis, MO).
Cell surface phosphatidylserine (PS) expression was measured with PE-conjugated annexin V, as recommended by the manufacturer (BD Biosciences). Briefly, specific annexin V binding was achieved by incubating 105 PLB-985 cells in binding buffer containing a saturating concentration of PE-annexin V and 7-amino-actinomycin (7-AAD) for 15 min in the dark. PE-annexin V and 7-AAD binding to the cells was analyzed by flow cytometry with CellQuest software, as described above.
On the third day of granulocytic differentiation, PLB-985 cells were washed in prewarmed PBS and resuspended (40 × 106 cells/ml) in RPMI-1640 medium containing siRNA strand 145596 (Ambion, Cambridgeshire, United Kingdom) at a concentration of 0.5 μg/ml. The cells were then transferred to a 4-mm-gap electroporation cuvette and were electroporated (300 V, 150 μF) using a Bio-Rad Gene Pulser instrument. Electroporated cells were then cultured for another 3 days in RPMI-1640 medium supplemented with 2 μM ATRA, 0.5% DMF, and 5% FCS for terminal granulocytic differentiation. DAF expression in transfected cells was controlled by Western blotting with a goat polyclonal anti-human DAF antibody (AF 2009; R&D Systems). The cells were then treated with a horseradish peroxidase (HRP)-linked secondary antibody (anti-goat HRP; Jackson Immunoresearch), and chemiluminescence was measured with an ECL Plus kit (Perkin Elmer).
DNA fragmentation in PLB-985 cells was examined by using propidium iodide. Briefly, cells were fixed in 70% (vol/vol) ethanol at −20°C overnight and then centrifuged and resuspended in PBS before they were stained (105 cells) with 50 μg/ml propidium iodide for 15 min. The percentage of cells with a hypodiploid (sub-G1) DNA content was determined by flow cytometry.
Caspase-3 activity was measured by fluorometric detection of cleavage of a 7-amino-4-trifluoromethyl coumarin (AFC)-labeled substrate (DEVD-AFC) specific for caspase-3, as recommended by the manufacturer (BioVision Research Products, Mountain View, CA). When cleaved, AFC emitted yellow-green fluorescence (λmax, 505 nm), which was quantified with a fluorometer (GENios; TECAN SA, Trappes, France).
Poly(ADP-ribose) polymerase (PARP-1) is a nuclear enzyme that catalyzes the transfer of ADP-ribose polymers and other nuclear proteins in response to DNA strand breaks. PARP-1 cleavage by caspase-3 and caspase-7 into 89- and 24-kDa fragments, considered to be a hallmark of apoptosis (10), was analyzed by Western blotting. The membranes were probed overnight with an anti-PARP monoclonal antibody targeting the N-terminal part of PARP-1 (C2-10; Alexis Biochemicals, Enzo Life Sciences) and treated with HRP-linked goat anti-mouse antibody (Beckman Coulter).
The mitochondrial transmembrane potential (ΔΨm) was measured by a flow cytometric method based on the capacity of intact mitochondria to take up and retain the lipophilic cationic fluorescent dye JC-1. This dye leaks into the cytoplasm from mitochondria undergoing the transition from polarized ΔΨm to depolarized ΔΨm (due to apoptosis), resulting in a decrease in fluorescence. PLB-985 cells were loaded with 20 nM JC-1 for 15 min at 37°C and then washed in PBS before fluorescence was measured by flow cytometry. The loss of mitochondrial transmembrane potential was calculated by comparing treated and untreated bacteria and was expressed as the percentage of cells with JC-1 fluorescence intensity lower than the mean for untreated controls.
Acridine orange (AO) is a lysosomotropic weak base (pKa 10.3) that can be retained by proton trapping in its charged form inside the acidic vacuolar compartment. To examine lysosomal membrane stability after bacterial challenge, PLB-985 cells were incubated in RPMI medium supplemented with 5 μg/ml AO for 15 min at 37°C. Cells containing a reduced number of intact AO-accumulating lysosomes (“pale cells”) were detected by flow cytometry as cells with diminished red fluorescence (64).
PS externalization on differentiated PLB-985 cells allows engulfment of these cells by macrophages. To quantify phagocytosis, we used the murine macrophage cell line RAW 264.7 cultured in Dulbecco modified Eagle medium (DMEM) (Gibco, Invitrogen) supplemented with 10% FCS and antibiotics. PLB-985 cells (2 × 106 cells) were pretreated with 10 ng/ml PMA or with 108 bacteria for 6 h in order to induce PS expression, and then they were stained with DNA-binding Hoechst stain (25 μg/ml) for 10 min at 37°C. The cells were then cocultured with 5 × 104 RAW 264.7 cells for 1 h in the presence or absence of annexin V (3 μg/ml), a PS-binding protein. After this coculture, noningested target cells were eliminated by several washes in PBS. A minimum of 100 RAW 264.7 cells per experimental condition were observed with a Nikon ECLIPSE 80i fluorescence microscope using UV illumination. The results were expressed as the percentage of RAW 264.7 cells containing Hoechst stain-labeled PLB-985 target cells.
Results are expressed below as means ± standard errors of the means. Bacterium-treated and untreated samples were compared by using analysis of variance (ANOVA), followed by multiple comparisons of means with Fisher's least significance procedure. The effects of the different bacteria were compared by using the Wilcoxon paired test. P values of <0.05 were considered significant.
PLB-985 myeloid cells were differentiated into terminally mature neutrophils by culture with 0.5% DMF and 2 μM ATRA for 6 days. As shown in Table Table1,1, β2-integrin expression and H2O2 production in response to PMA were not significantly different for PLB-985 cells and PMN freshly isolated from blood, confirming that there was terminal granulocytic differentiation of PLB-985 cells under our experimental conditions.
The expression of selected Afa/Dr DAEC fimbria receptors at the PLB-985 cell surface was also quantified by flow cytometry (Table (Table1);1); cell surface DAF or CEACAM1 expression was similar on PLB-985 cells and blood PMN, while CEACAM3 expression was lower on PLB-985 cells.
The binding capacity of the bacteria was then evaluated. We found that the DH5α control strain and the C1845 wild-type strain bound similarly to PLB-985 cells (10 ± 5 and 11 ± 4 bacteria/cell, respectively) and that DH5α/pF1845 was the most potent strain (22 ± 6 bacteria/cell), probably because of the greater expression of F1845 fimbriae at the surface of the recombinant strain.
These results indicate that differentiated PLB-985 cells are suitable for studying neutrophil functional responses following infectious challenge with Afa/Dr DAEC.
To determine whether E. coli wild-type strain C1845 induces PLB-985 cell apoptosis, PS externalization was measured by using PE-conjugated annexin V and 7-AAD to counterstain late apoptotic or necrotic cells. Cells that stained positive for PE-annexin V and negative for 7-AAD were considered cells that were undergoing apoptosis, while cells that were positive for both markers were either in the end stage of apoptosis or undergoing necrosis. Cells were challenged for 2 to 16 h with wild-type strain C1845, E. coli strain DH5α, or E. coli recombinant strain DH5α/pF1845 expressing F1845 fimbriae. As shown in Fig. Fig.1A,1A, a small number of annexin V-positive, 7-AAD-negative PLB-985 cells were found as early as 2 h after challenge with the different bacteria. In the presence of DH5α, the number of cells remained the same during the 16-h study period. In contrast, cells challenged with wild-type strain C1845 or recombinant strain DH5α/pF1845 exhibited parallel marker kinetics; the maximum increase in the number of annexinV-positive, 7AAD-negative PLB-985 cells occurred at 8 h (Fig. (Fig.1B),1B), and this was followed by a decrease (Fig. (Fig.1A).1A). Interestingly, no significant difference was observed between wild-type strain C1845 and the DH5α/pF1845 recombinant strain, indicating that F1845 fimbriae have a role in PS externalization. Moreover, wild-type strain IH11128, another member of the Afa/Dr DAEC family, and an isogenic IH11128 mutant lacking the Dr adhesin gave results similar to those obtained with the DH5α control strain, further supporting the hypothesis that F1845 fimbriae have a role in PS externalization (Fig. (Fig.1C).1C). Similar results were obtained with blood PMN, validating our cell model (data not shown).
To assess the role of DAF in F1845-induced PS externalization on infected PLB-985 cells, we used a small-RNA-based gene-silencing strategy. As shown in Fig. Fig.22 A, a 60% decrease in DAF expression in differentiated PLB-985 cells was observed by means of flow cytometry after transfection with DAF siRNA. Western blotting confirmed that PLB-985 cells transfected with DAF siRNA showed reduced DAF protein expression compared with nontransfected cells (Fig. (Fig.2B).2B). In addition, DH5α- and DH5α/pF845-infected PLB-985 cells were treated with the anti-DAF blocking antibody IH4, which recognizes the short consensus repeat 3 domain of DAF. As shown in Fig. Fig.2C,2C, DH5α/pF1845-induced PS externalization was significantly reduced, but not abolished, in DAF siRNA-transfected PLB-985 cells and in IH4-treated PLB-985 cells. These results indicate that DAF is engaged during F1845-induced PS externalization.
We have previously described involvement of several signaling pathways, as well as the F-actin cytoskeleton and lipid rafts, in Afa/Dr fimbria-induced intestinal epithelial cell injury (28, 50). Treatment of DH5α/pF1845-infected PLB-985 cells with cytochalasin D (an agent that disrupts the microfilament network of the actin cytoskeleton) or the lipid raft-disrupting agent methyl-β-cyclodextrin significantly reduced the percentage of cells expressing PS (Fig. (Fig.3).3). A small but significant decrease was also observed with the DH5α strain, suggesting that this observation might have been related to nonspecific bacterial adherence. As we previously showed that wild-type strain C1845 triggers kinase-dependent IL-8 synthesis by PLB-985 cells, we pretreated PLB-985 cells with various kinase inhibitors in order to determine the role of kinases in F1845-induced PS externalization (Fig. (Fig.3).3). In response to both bacteria, PS externalization was reduced by the tyrosine kinase inhibitor genistein, suggesting that an additional common virulence factor may be involved. The protein kinase C inhibitor staurosporine strongly inhibited only F1845-induced PS externalization. In contrast, inhibitors of Erk1/2 and p38 MAPKs, PI3K, and Src kinase had no effect (data not shown).
We first analyzed apoptotic and necrotic processes in PLB-985 cells challenged with the different E. coli strains in a 16-h kinetic study. Flow cytometric analysis of DNA fragmentation was used to identify sub-G1 cells. Although the PLB-985 cells exposed to staurosporine contained a high percentage of sub-G1 cells as early as 8 h, the percentage remained low (<5%) throughout the kinetic study, and the results were similar with C1845-, DH5α/pF1845-, and DH5α-infected PLB-985 cells (data not shown). Moreover, 7-AAD staining showed that few PLB-985 cells were necrotic, whatever the bacterial challenge (data not shown).
This result was unexpected, given the cell death observed elsewhere for isolated human PMN infected with an F1845-positive E. coli strain (8), which prompted us to examine apoptotic pathways potentially activated in infected PLB-985 cells.
No cleavage of DEVD, a specific substrate of caspase-3-like proteases, was induced by wild-type strain C1845, recombinant strain DH5α/pF1845, or nonpathogenic strain DH5α (Fig. (Fig.4A).4A). In addition, no cleavage of poly(ADP-ribose) polymerase (PARP-1) into 89- and 24-kDa fragments was observed in infected PLB-985 cells (Fig. (Fig.4B).4B). Finally, PS externalization on C1845-, DH5α/pF1845-, and DH5α-infected PLB-985 cells was not affected by the pancaspase inhibitor zVAD (data not shown), confirming that caspases are not activated in our model.
Stimuli originating from various subcellular compartments can converge on mitochondria and induce mitochondrial outer membrane permeabilization (38). We therefore investigated the possible role of mitochondria in F1845-induced, caspase-independent PS externalization on PLB-985 cells. Changes in the mitochondrial membrane potential (ΔΨm) of infected PLB-985 cells were measured by using the cationic lipophilic fluorochrome JC-1. As shown in Fig. Fig.4C,4C, no significant change in ΔΨm was seen in C1845-, DH5α/pF1845-, or DH5α-challenged PLB-985 cells compared with controls.
Lysosomes can also be involved in caspase-independent apoptosis, depending on the degree of lysosomal permeability and on the quantitative release of proteolytic enzymes, such as cathepsins (27). We used the fluorophore acridine orange (AO) to monitor the loss of lysosomal acidity in infected PLB-985 cells. As shown in Fig. Fig.4D,4D, there was no change in lysosomal membrane stability in PLB-985 cells challenged with strain C1845, DH5α/pF1845, or DH5α. Likewise, PLB-985 cells pretreated with the specific cysteine-cathepsin inhibitor EST (targeting cathepsins B and L) or the aspartic cathepsin inhibitor pepstatin A (targeting cathepsin D) showed no change in PS externalization when they were challenged with bacteria (data not shown).
Similar results were obtained with blood PMN (data not shown). Together, these findings suggest that F1845-induced PS externalization on differentiated PLB-985 cells is not related to cell death mechanisms.
In response to bacteria, neutrophils can produce reactive oxygen species (ROS) and reactive nitrogen species (RNS) through activation of NADPH oxidase (NOX2) and nitric oxide synthase (NOS), respectively (61). We therefore examined whether ROS and/or RNS production by infected PLB-985 cells contributed to F1845-induced PS externalization. As shown in Fig. Fig.5A,5A, both E. coli wild-type strain C1845 and E. coli recombinant strain DH5α/pF1845 induced PS externalization on X-CGD PLB-985 cells lacking NOX2 with the same intensity as that on wild-type PLB-985 cells (Fig. (Fig.1C).1C). PS externalization was therefore not affected by NOX2 inactivation. Moreover, no significant change in PS externalization occurred in the presence of L-NAME, a constitutive NOS inhibitor (Fig. (Fig.5B).5B). Finally, as mitochondria are an alternative source of ROS (2), we studied the possible role of mitochondrial ROS production in PS externalization by using rotenone, a mitochondrial NOX inhibitor. As shown in Fig. Fig.5B,5B, PS externalization was not affected by rotenone, whatever the bacterial challenge. Together, these results suggest that F1845-induced PS externalization on PLB-985 cells is not related to ROS or NOS production.
Macrophage recognition and disposal of neutrophils are important steps in the resolution of inflammation. To determine whether F1845-induced PS externalization is sufficient for PLB-985 cell engulfment by macrophages, PLB-985 cells were cocultured with RAW 264.7 murine macrophages for 1 h. As shown in Fig. Fig.6,6, the percentage of phagocytosis-positive RAW 264.7 cells was significantly higher with PMA-treated and C1845-treated PLB-985 cells than with untreated cells and DH5α-challenged cells. Moreover, annexin V, a PS-binding protein, significantly reduced PLB-985 phagocytosis by RAW 264.7 cells, providing further evidence that PS externalization is sufficient for macrophage disposal of C1845-challenged PLB-985 cells.
Neutrophils are the first cells to be recruited to sites of bacterial infection and provide the first line of defense against invading microorganisms. Pathogen-induced death program activation in neutrophils and clearance of apoptotic infected cells are both crucial for innate immune system homeostasis and host integrity (36). As different bacteria have different effects on neutrophil cell death mechanisms, we examined the effects of enterovirulent Afa/Dr diffusely adherent E. coli (DAEC) on neutrophil apoptosis and on the clearance of apoptotic neutrophils. We chose the PLB-985 cell line, which differentiates into fully mature neutrophils, as a model system. We first showed that PLB-985 cells express several receptors involved in the interaction with Afa/Dr DAEC, including DAF (48), CEACAM1 (3), and CEACAM3. We found that the wild-type enterovirulent strain E. coli C1845 and the recombinant strain E. coli DH5α/pF1845, which expresses the fimbrial adhesin F1845, induced similar time-dependent PS expression on differentiated PLB-985 cells, suggesting that this virulence factor has a major role in PS externalization. In particular, PS externalization was triggered by DAF engagement (leading to tyrosine kinase and PKC activation) and required architectural integrity of the cytoskeleton and lipid rafts. No involvement of necrosis or apoptosis was found; in particular, PS externalization was not dependent on caspases, mitochondria, lysosomes, or reactive oxygen or nitrogen species. Moreover, nonapoptotic F1845-mediated PS externalization was sufficient for infected PLB-985 cells to be engulfed by macrophages.
Bacterial pathogens are armed with an array of virulence determinants that interact with key components of host cell death pathways or interfere with the regulation of transcription factors required for cell survival (36, 37, 41, 63). Phagocytosis-induced cell death (PICD) is another mechanism by which neutrophils that contain pathogens, particularly E coli, are cleared (62). Whether cell death is induced or delayed by E. coli depends on the strain and/or the number of bacteria. For instance, adherent invasive E. coli isolated from patients with Crohn's disease survives and replicates within phagocytes but does not induce cell death (23). Conversely, enteroaggregative E. coli and uropathogenic E. coli induce phagocyte cell death, notably via type 1 fimbriae and lipopolysaccharide (LPS) (7, 19). The most widespread and best-characterized change on the surface of dying cells is the loss of phospholipid asymmetry and the translocation of PS to the outer leaflet of the lipid bilayer, which occurs very early in the apoptotic process (45, 51). This “eat me” signal binds to bridging molecules, such as MFG-E8, Gas6, C1q, and β2GPI, thereby enhancing the capacity of apoptotic cells to be engulfed by macrophages (1, 14).
Afa/Dr DAEC harboring the fimbrial adhesin F1845 recognizes membrane-associated glycosylphosphatidylinositol (GPI)-anchored proteins, such as DAF (53). As we found here that DAF was expressed on the surface of differentiated PLB-985 cells and as GPI-anchored protein engagement can induce PS externalization (54), we studied the role of DAF in our model by establishing a DAF siRNA-transfected PLB-985 cell line. PS externalization in response to F1845 fimbriae was indeed reduced in the transfected cells, as well as in anti-DAF antibody (IH4)-treated wild-type cells. This suggested that other receptors, such as CEACAM1, might also participate in F1845-induced PS externalization. We then analyzed the molecular mechanism of PS externalization in PLB-985 cells through GPI-anchored DAF. Smrz et al. obtained evidence that GPI-anchored protein engagement in mastocytes could be sensed and transduced by signaling molecules located in lipid rafts, leading to nonapoptotic PS externalization (54). Moreover, we found previously that infection of intestinal Caco-2 cells by E. coli wild-type strain C1845 and by E. coli recombinant strain DH5α/pF1845 provokes drastic F-actin rearrangement following phosphotyrosine clustering and activation of a cascade of signaling molecules, including protein tyrosine kinase (PTK), phospholipase C γ (PLCγ), PKC, and PI3K (29, 53). In the present study we found that most of these pathways were also involved in DAF-mediated PS externalization on PLB-985 cells, the most specific signaling pathway involving PKC activation. The latter finding is interesting, as PS expression on the outer leaflet is dependent on PKCδ-induced scramblase activation in both activated and apoptotic cells (21, 55). Our results partially elucidate the signaling pathways involved in the response to Afa/Dr DAEC, showing a role for lipid rafts, the cytoskeleton, PTK, and PKC. As IL-8 production is dependent mainly on Src family kinase and MAPK activation (52), PS externalization does not appear to be driven by these two signaling pathways.
PS externalization has been implicated in neutrophil viability in several models (8, 36, 63). Surprisingly, we found no change in the viability of F1845-positive E. coli-infected PLB-985 cells, based on DNA fragmentation status; moreover, 7-AAD staining showed that the percentage of necrotic cells was very low. These results are in keeping with those of recent studies suggesting that PS translocation may involve nonapoptotic pathways in lymphocytes (13), mastocytes (54), and blood neutrophils in response to formyl-methionyl-leucyl-phenylalanine (fMLP) (21) and galectin (57, 58). However, our results for PLB-985 cells are different from those of Brest et al. (8), who reported that both C1845 infection and DH5α/pF1845 infection induced blood neutrophil apoptosis. These different results indicate the importance of the experimental conditions, including the culture medium, bacterial inoculum, and human cell model.
Different pathways of PS externalization in neutrophils have been reported to be either dependent on or independent of caspase activation (16, 34). Although caspase-3 activation has been observed in some models of bacterium-induced neutrophil apoptosis (8, 37, 44), this was not the case in our model, as shown using three different approaches. Certain stimuli that trigger the intrinsic apoptotic pathway may disrupt the mitochondrial transmembrane potential and may or may not induce caspase activation (25). In particular, mitochondria can release mediators, such as ROS, the apoptosis-inducing factor (AIF) that interacts with DNA (9), and the serine protease HtrA2/Omi, that can mediate caspase-independent apoptosis (47). We found no evidence for a role of mitochondria in F1845-induced PS externalization. Stimuli that cause lysosomal membrane permeabilization in turn induce the release of cathepsin proteases (42). In our model, F1845-bearing bacteria did not modify lysosome membrane stability, and PS externalization was not prevented by cathepsin B, D, or L inhibitors.
In a previous study we found that Afa/Dr DAEC, particularly strains bearing F1845 fimbriae, were able to activate the neutrophil oxidative burst (52). As ROS have sometimes been found to play a role in neutrophil PS externalization (16, 43) and as patients with chronic granulomatous disease (CGD) exhibit deficient PS expression in response to stimuli (18), we examined the X-CGD PLB-985 cell line, which lacks functional NOX2 (12). We obtained similar results for PS externalization in response to E. coli wild-type strain C1845 and E. coli recombinant strain DH5α/pF1845 in both wild-type and deficient PLB-985 cells. In keeping with these results, we found that rotenone, a mitochondrial ROS inhibitor, had no effect and that cNOS-generated NO intermediates were not involved in F1845-induced PS externalization. Thus, ROS and NOS are not critical for F1845-induced PS translocation in our model.
In conclusion, our findings provide new insights into the neutrophil response to Afa/Dr DAEC infection. After our initial demonstration that type 1 pili on Afa/Dr DAEC can drive IL-8 hyperproduction by PLB-985 cells (52), we identified a major role for the F1845 fimbria-DAF interaction in PS externalization. Interestingly, despite the lack of engagement of irreversible apoptosis pathways, wild-type strain C1845-infected PLB-985 cells displayed increased sensitivity to engulfment by macrophages. PS externalization is a common recognition signal for macrophages (11, 34). Externalized PS interacts with macrophages either directly through specific receptors, such as stabilin-2 and Tim-4, or indirectly via bridging molecules, allowing neutrophil removal and thus resolution of inflammation (36, 46, 49). The advantage for the pathogen of F1845-induced PS externalization, which promotes bacterial engulfment by macrophages, remains to be determined. Alternatively, this phenomenon may be a host defense mechanism that allows macrophage clearance of Afa/Dr DAEC-infected PMN and thus resolution of inflammation. Interestingly, we observed that Afa/Dr DAEC-infected PLB-985 cells produce macrophage chemotactic peptide 1 (MCP-1) (data not shown), a factor that can attract circulating monocytes and resident macrophages and thereby participate in neutrophil-mediated monocyte recruitment and activation (56).
We are grateful to Mary Dinauer (Indiana University School of Medicine, Indianapolis, IN) and Marie-José Stasia (Centre Diagnostic et Recherche Granulomatose Septique, CHU, Grenoble, France) for the generous gift of the X-CGD PLB-985 cell line. We also thank Imad Kansau for helpful discussions concerning E. coli, Mojgan Djavaheri-Mergny for her expert assistance with evaluation of apoptosis, and Saadia Kerdine and Affef Tolbi for their help with transfection studies.
Editor: S. M. Payne
Published ahead of print on 19 April 2010.