Cytotoxic necrotizing factor-1 (CNF1) is isolated from pathogenic strains of Escherichia coli and catalyzes the activation of Rho GTPases by the deamidation of a glutamine residue. This toxin induces stress fiber formation, cell spreading, and membrane folding and promotes phagocytosis competence in epithelial cells. We show that CNF1 induces morphologic changes in monocytic cells: polarized-like shape in THP-1 cells, lamellipodia, and cell spreading in adherent monocytes. CNF1 also increased filamentous actin (F-actin) content in a time- and dose-dependent manner. In addition, the toxin profoundly reorganized the actin cytoskeleton: redistribution of F-actin in polarized deformations of THP-1 cells and disorganization of microfilament network in monocytes. We also studied the effects of CNF1 on phagocytosis. It markedly impaired the ingestion of unopsonized zymosan involving CR type 3. However, CNF1 had no effect on the uptake of iC3b-coated zymosan or IgG-mediated phagocytosis of SRBC. In addition, CNF1 induced clustering of CR3 and FcγRII (CD32) but selectively impaired the colocalization of CR3 with F-actin. It is likely that CNF1-induced reorganization of actin cytoskeleton down-modulates integrin activation-dependent phagocytosis by preventing the codistribution of CR3 with F-actin. CNF1 may control some features of integrin-dependent phagocytosis in myeloid cells through its action on Rho GTP binding proteins and cytoskeletal organization.

Small GTPases play a role in the control of several transductional pathways and in the regulation of actin cytoskeleton in eukaryotic cells. At least 12 members of the Rho family of small GTPases have been characterized, e.g., RhoA-E, RhoG, Rac1, Rac2, Cdc42, and TC10 (1, 2). Their role in the regulation of actin cytoskeletal organization has been defined in fibroblasts and epithelial cells stimulated with growth factors (3). Indeed, RhoA-C proteins control the assembly of stress fibers and focal adhesions (4). Cdc42 elicits the formation of actin microspikes and filopodia (5). Rac proteins participate in membrane ruffling and the formation of lamellipodia (6). The role of Rho GTPases in the functions of immune cells is not fully understood. The inactivation of Rho by Clostridium botulinum C3 exoenzyme prevents the cytotoxicity (7) and the aggregation of T lymphocytes (8). In addition, C3 exoenzyme inhibits agonist-induced adhesion of α4β1 integrin to VCAM-1 (9). IL-2-driven lymphocyte proliferation requires the integrity of actin cytoskeleton and is under the control of Rho proteins (10). Mice lacking thymic function of the GTPase Rho exhibit severe defects in fetal and adult thymopoiesis, establishing the critical role of Rho in the development of early thymic progenitors (11).

In phagocytes, the role of Rho GTPases has been studied for few cell functions. Rac1 and Rac2 appear to be essential for the activation of NADPH oxidase complex in neutrophils and macrophages (12, 13). C3 exoenzyme-mediated inactivation of Rho reduces neutrophil chemotaxis (9). The abnormality of an effector of Cdc42 is related to the deficiency of neutrophil chemotaxis observed in Wiskott-Aldrich syndrome (14). The spreading of monocyte-derived macrophages correlates with the expression of Cdc42 (15) and is enhanced by the ADP ribosylation of Rho proteins (16). CSF-1 stimulates macrophage spreading apparently via the activation of Cdc42 and Rac GTPases (17), but the role of Rho GTPases in phagocytosis, a major mechanical function of professional phagocytes, has not been determined. Several receptors mediate the binding and internalization of particles, opsonized or not. The best characterized receptors are the different types of receptors for the Fc portion of IgG (FcγR) and CR type 3 (CD11b/CD18), a β2 integrin involved in the recognition of iC3b and several determinants expressed by bacteria and parasites (18). Particle internalization requires actin polymerization and its remodeling at the contact area between phagocytes and targets (19).

In this report we studied the role of Rho GTPases in cytoskeleton organization and the phagocytosis of human monocytes by using a bacterial toxin active on the GTPase Rho. Indeed, cytotoxic necrotizing factor-1 (CNF1),2 isolated from Escherichia coli strains, covalently interacts with Rho, resulting in its activation through the deamidation of a glutamine residue at position 63 (20, 21). CNF1 has been reported to induce membrane ruffling, stress fiber assembly, and phagocytic competence in epithelial cells (20, 22). In human monocytes and the myelomonocytic cell line THP-1, CNF1 induced dramatic morphologic alterations and increased filamentous actin (F-actin) content and reorganization. CNF1 also affected some features of CR3-dependent phagocytosis and impaired the codistribution of CR3 with actin cytoskeleton. We suggest that Rho-mediated modulation of actin cytoskeleton might negatively regulate the phagocytic activity of integrins.

CNF1 from pathogenic E. coli was purified as previously described (22) and filtered on sulfone to retain endotoxins (23). RPMI 1640, FCS, l-glutamine, penicillin, streptomycin, and HBSS without phenol red were purchased from Life Technologies (Eragny, France). All cell preparations and media were checked for the absence of endotoxins by using Limulus amebocyte lysate (Boehringer Ingelheim, Gagny, France). SRBC and the specific rabbit IgG were purchased from BioMérieux (Marcy l’Etoile, France). Bodipy phallacidin and calcein-AM were obtained from Molecular Probes (Eugene, OR). [32P]nicotinamide adenine dinucleotide (NAD; sp. act., 1.11 TBq/mmol) was obtained from New England Nuclear Products (Les Ulis, France). mAb directed against CD11b (IgG1), CD18 (IgG1), CD32 (IgG2a), controls (IgG1 and IgG2a), and secondary Ab were obtained from Immunotech (Marseille, France). mAb 24 was provided by Dr. N. Hogg (London, U.K.). Other reagents were purchased from Sigma (St. Louis, MO).

Blood from healthy adult volunteers was collected in heparinized tubes. PBMC were isolated by Ficoll gradient centrifugation (Nycomed, Oslo, Finland) as previously described (24). Then, they were suspended in RPMI 1640 containing 25 mM HEPES, 10% FCS, 2 mM l-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin (supplemented RPMI 1640). For cytoskeletal determination and phagocytosis assay, 5 × 105 PBMC in a 0.5-ml volume were incubated in Lab-Tek glass chambers (Nunc, Naperville, IL). For superoxide assay, 106 PBMC in a 1-ml volume were cultured in 24-well culture plates (Nunc). After 1 h at 37°C, nonadherent cells were removed by washing. Remaining adherent cells consisted of about 95% monocytes as assessed by morphology, CD14 expression, and latex phagocytosis.

The human myelomonocytic cell line THP-1 was provided by the European Collection of Animal Cell Cultures (Cerdic, Sophia-Antipolis, France). Suspended cells were propagated at an initial density of 5 × 105 cells/ml in supplemented RPMI 1640 by biweekly passages.

Morphologic changes in monocytic cells were studied with scanning electron microscopy (25). Cells were incubated with CNF1 for 24 h and fixed for 30 min in 0.1 M cacodylate buffer (pH 7.2) containing 4% glutaraldehyde. After extensive washings, cells were dehydrated through graded ethanols and critical point-dried in CO2. THP-1 cells and monocytes were examined with a scanning electron microscope (JEOL 35CF, Croissy-sur-Seine, France).

Determination of F-actin.

After treatment with CNF1, monocytic cells were fixed with 3.7% formaldehyde and incubated for 20 min with PBS containing 10 U/ml bodipy phallacidin, a specific fluorescent probe of F-actin, and 100 μg/ml lysophosphatidylcholine, as previously described (25). After being washed in PBS, the cellular content in F-actin was determined. The cell fluorescence was analyzed by an EPICS XL (Coulter, Hialeah, FL) equipped with an argon laser (488 nm excitation and 525 nm fluorescence emission). Linear fluorescence intensities of 10,000 cells were expressed as the mean of arbitrary units ± SD as provided by the data processing software.

Expression of membrane Ags.

Monocytes and THP-1 cells were incubated with anti-CD11b, anti-CD18, anti-CD32 (FcγRII) mAb, or isotypic controls at 1/100 dilution for 30 min at 4°C. After washing, they were incubated with FITC-tagged F(ab′)2 anti-mouse Igs for 30 min at 4°C. After fixation with 1% formaldehyde, cells were analyzed by flow cytometry. Gating was established using forward and side scatters and fluorescence recorded on the log scale.

The cytoskeletal organization of monocytes and THP-1 cells was observed with a Labophot microscope (Nikon, Tokyo, Japan) equipped for epifluorescence. The quantification of cell deformations rich in F-actin was determined by comparing the large and the small axes in fluorescence micrographs of single cells. The intracellular distribution of F-actin was also examined with a laser scanning confocal fluorescence microscope (Leica, DMIRBE, Lyon, France) equipped with a ×100 (NA 1.4) oil-immersion lens. Serial optical sections of images were collected at 1-μm intervals, analyzed using Adobe Photoshop 3.0, and printed with a Mavigraph color video printer (Sony, Tokyo, Japan).

The colocalization between F-actin and membrane receptors was determined with laser scanning confocal fluorescence microscopy. Monocytes and THP-1 cells were incubated first with mAb anti-CD11b, anti-CD18, anti-CD32, or isotypic controls at a 1/100 dilution for 30 min at 4°C and, second, with rhodamine-tagged F(ab′)2 anti-mouse Igs for 30 min at 4°C. After fixation with 1% formaldehyde, cells were incubated with bodipy phallacidin as described above. The specimens were mounted in slowfade solution (Molecular Probes) and examined with the confocal fluorescence microscope equipped with separate filters for each fluorochrome. Monocytes and THP-1 cells were individually labeled with each fluorochrome to evaluate their contribution to fluorescent images under the given confocal conditions. The intensities of bodipy and rhodamine images were adjusted to be roughly equal, then respectively converted into green and red images and merged to synthesize a yellow color.

Rho activation was determined as previously described (21). PBMC and THP-1 cells at 107 cells/assay were incubated with 12.5 nM CNF1 for 18 h in supplemented RPMI 1640. Adherent monocytes and THP-1 cells were scraped; washed twice with 20 mM Tris buffer, pH 7.4, containing 5 mM MgCl2, 10 mM DTT, and protease inhibitors; and harvested in 100 μl of Tris containing MgCl2 and DTT. Then, cells were frozen and thawed at 37°C four times, and spun down (21,500 × g) for 30 min at 4°C. Supernatants were saved, and their protein concentrations were adjusted. Twenty-five microliters of supernatant from control and CNF1-treated cells was added to 2 μg of C3 exoenzyme and [32P]NAD for 1 h at 37°C. RhoA fusion protein (1.5 μM) was also incubated with or without CNF1 and [32P]NAD as controls. Samples were then subjected to 12% SDS-PAGE electrophoresis, and autoradiograms of the gels were performed.

Superoxide production was monitored by measuring the superoxide dismutase (SOD)-inhibitable reduction of ferricytochrome c, as previously described (26). Briefly, THP-1 cells or adherent monocytes were incubated in HBSS containing 120 μM ferricytochrome c and 2 mM sodium azide. The reaction was conducted for 1 h at 37°C in the presence of 50 ng/ml of PMA and was stopped by the addition of 1 mM N-ethylmaleimide. Supernatants were collected, centrifuged, and assayed for absorbance at 550 nm. The generation of superoxide was calculated by subtracting the change absorbance in the presence of SOD (300 U/ml) from that in its absence. The results are expressed as nanomoles of superoxide released by 106 monocytic cells in 1 h with an extinction coefficient of 21,000 M−1/cm.

After CNF1 treatment, THP-1 cells and adherent monocytes were incubated for 1 h at 37°C with unopsonized or opsonized particles. The phagocytosis of unopsonized zymosan is largely mediated by CR3 (24). The opsonization of zymosan particles by complement was conducted as previously described (27). Briefly, zymosan particles were incubated with 50% human serum for 15 min at 37°C. These conditions result in a maximum deposition and conversion of C3 into iC3b. SRBC were opsonized with specific IgG at 1/2000 dilution (IgG-SRBC). One milligram per milliliter of unopsonized zymosan, iC3b-coated zymosan, or 2 × 107 IgG-SRBC were added to cells in RPMI 1640 containing 10% heat-inactivated FCS. Then cells were washed to remove unbound particles and examined microscopically after lysis of opsonized SRBC by distilled water. For the determination of zymosan phagocytosis, cell preparations were stained with Diff Quik (Baxter, Maurepas, France) before microscopic examination. More than 200 cells were counted in each assay. Phagocytosis results are expressed as the product of the percentage of cells having phagocytosed at least one particle and the number of phagocytozed particles per cell.

PBMC and THP-1 cells were incubated with 10 mM calcein-AM for 30 min. Calcein-labeled cells (105 cells/assay) were added to HUVEC monolayers as previously described (28). To determine CR3-dependent adherence, monocytic cells were incubated with HUVEC monolayers in the presence of mAb directed against CD11b. After 1 h of incubation at 37°C and washing, cell-associated fluorescence was measured with a fluorescence multiwell plate reader (Cytofluor, PerSeptive Biosystems, Framingham, MA). Assays were performed in triplicate, and results were expressed as relative mean adherence corresponding to the ratio of fluorescence values before and after washing.

Results are given as the mean ± SEM and compared with Mann-Whitney U test. Differences were considered significant if p < 0.05.

The morphology of control and CNF1-treated THP-1 cells and monocytes was first assessed by scanning electron microscopy. Control THP-1 cells were perfectly spherical, with very few small pseudopodal extensions (Fig. 1,A). The changes in cell shape of CNF1-stimulated THP-1 cells were detectable after 12 h of treatment, but were greatest 24 h after the addition of CNF1 (data not shown). Indeed, after 24 h of treatment with 500 pM CNF1, THP-1 cells showed deformations consisting of a polarized shape (Fig. 1,B). Another cell projection was sometimes found on the opposite side of the cell (Fig. 1,C). Control monocytes were round with discrete deformations caused by their adherence to substrate (Fig. 1,D). The duration of morphologic changes induced by CNF1 in monocytes was the same as that observed in THP-1 cells. After 24 h of treatment, CNF1 increased spreading of monocytes, with a thin circumferential lamellipodium and circular swirls at the top of the cells (Fig. 1,E). Lamellipodia were also associated with major cell deformations consisting of knob-like protuberances (Fig. 1 F).

FIGURE 1.

Electron microscopic study of THP-1 cells and monocytes. THP-1 cells (A–C) and monocytes (D–F) were incubated with 500 pM CNF1 for 24 h. After fixation and dehydration, monocytic cells were examined with a scanning electron microscope. Representative micrographs of control cells (A and D), CNF1-treated THP-1 cells (B and C), and CNF1-treated monocytes (E and F) exhibiting the main shape changes are shown. Bar = 4 μm.

FIGURE 1.

Electron microscopic study of THP-1 cells and monocytes. THP-1 cells (A–C) and monocytes (D–F) were incubated with 500 pM CNF1 for 24 h. After fixation and dehydration, monocytic cells were examined with a scanning electron microscope. Representative micrographs of control cells (A and D), CNF1-treated THP-1 cells (B and C), and CNF1-treated monocytes (E and F) exhibiting the main shape changes are shown. Bar = 4 μm.

Close modal

As morphologic changes probably resulted from cytoskeletal reorganization, we studied F-actin content and its distribution. THP-1 cells were incubated with 500 pM CNF1 for different periods of time. F-actin was labeled with bodipy phallacidin, and the cell content in F-actin was determined by cytofluorographic measurements. The F-actin content of CNF1-treated cells reached a plateau between 8 and 24 h, and declined after 48 h of treatment (Fig. 2,A). After 24 h of treatment, the content of F-actin was significantly higher (p < 0.02) than that in control cells. It is noteworthy that this increase in F-actin was equivalent to that induced by 10−7 M FMLP, a chemoattractant known to elicit actin polymerization in phagocytic cells (data not shown). We also studied the effects of different concentrations of CNF1 on the F-actin content in THP-1 cells treated for 24 h (Fig. 2 B). A progressive increase in the F-actin content was observed for concentrations of CNF1 between 30 pM and 1 nM. Subsequent experiments were performed by incubating cells with CNF1 at 500 pM for 24 h.

FIGURE 2.

Effect of CNF1 on actin polymerization in THP-1 cells. THP-1 cells were incubated with 500 pM CNF1 for different times or with different concentrations of CNF1 for 24 h. Then, they were labeled with bodipy phallacidin. Samples (10,000 cells) were analyzed by flow cytometry. A, The duration (in hours) of the increase in F-actin content was shown as representative histograms. B, The F-actin content as a function of CNF1 concentrations was expressed relative to that in control THP-1 cells. Results represent the mean ± SEM of five experiments.

FIGURE 2.

Effect of CNF1 on actin polymerization in THP-1 cells. THP-1 cells were incubated with 500 pM CNF1 for different times or with different concentrations of CNF1 for 24 h. Then, they were labeled with bodipy phallacidin. Samples (10,000 cells) were analyzed by flow cytometry. A, The duration (in hours) of the increase in F-actin content was shown as representative histograms. B, The F-actin content as a function of CNF1 concentrations was expressed relative to that in control THP-1 cells. Results represent the mean ± SEM of five experiments.

Close modal

The F-actin distribution was investigated by fluorescence microscopy, and the quantification of cells with or without deformations was determined by comparing the large and small axes of individual cells. Without CNF1, the F-actin distribution was homogeneous with sustained staining of THP-1 cell periphery (Fig. 3, A and D). The diameter ratio was 1.05 ± 0.03, thus emphasizing the spherical shape of control cells. After CNF1 treatment, two patterns of F-actin reorganization were found. Using the procedure of diameter comparison, a ratio >1.10 was assigned to polarized cells. A fraction of treated cells (58 ± 7%) remained spherical (diameter ratio, 1.06 ± 0.04) but exhibited numerous peripheral foci of F-actin organized as microspikes (Fig. 3, B and E). The other fraction of THP-1 cells (42 ± 8%) lost the rounded shape and displayed a polarized morphology (diameter ratio, 1.32 ± 0.17) with a large membrane expansion rich in F-actin (Fig. 3, C and F). CNF1 induced more dramatic reorganization of F-actin in adherent monocytes than in suspended THP-1 cells. In control cells, the staining of F-actin was homogeneous with peripheral reinforcement. As monocytes were cultured for 24 h, they already displayed a moderate spreading (Fig. 4, A and C). Monocyte treatment with CNF1 induced an increase in cell size and the formation of large folds and swirls on their surface and at their periphery. The organization of the microfilament network was lost; F-actin was concentrated in several foci, and actin cables were present (Fig. 4, B and D).

FIGURE 3.

Effect of CNF1 on F-actin distribution in THP-1 cells. THP-1 cells were incubated with 500 pM CNF1 for 24 h and then labeled with bodipy phallacidin. Cells were observed with conventional fluorescence (left panels) and confocal microscopy (right panels). Micrographs of a control cell (A and D), a nonpolarized cell (B andE), and a polarized cell (C andF) after CNF1 treatment are shown.

FIGURE 3.

Effect of CNF1 on F-actin distribution in THP-1 cells. THP-1 cells were incubated with 500 pM CNF1 for 24 h and then labeled with bodipy phallacidin. Cells were observed with conventional fluorescence (left panels) and confocal microscopy (right panels). Micrographs of a control cell (A and D), a nonpolarized cell (B andE), and a polarized cell (C andF) after CNF1 treatment are shown.

Close modal
FIGURE 4.

Effect of CNF1 on F-actin distribution in monocytes. Monocytes were incubated with 500 pM CNF1 for 24 h and then labeled with bodipy phallacidin. Cell fluorescence was observed with a conventional microscope (left panels) or a confocal microscope (right panels). Micrographs of control monocytes (A and C) and CNF1-treated monocytes (B and D) are shown.

FIGURE 4.

Effect of CNF1 on F-actin distribution in monocytes. Monocytes were incubated with 500 pM CNF1 for 24 h and then labeled with bodipy phallacidin. Cell fluorescence was observed with a conventional microscope (left panels) or a confocal microscope (right panels). Micrographs of control monocytes (A and C) and CNF1-treated monocytes (B and D) are shown.

Close modal

It has been reported that CNF1 stimulates Rho activation in several cell types (21). Monocytes and THP-1 cells were incubated with CNF1, and the C3 exoenzyme known as ADP-ribosylate Rho was added to cell extracts. As a control, RhoA fusion protein was incubated in the presence of CNF1 and the C3 exoenzyme. CNF1 caused a clear shift in the apparent m.w. of [32P]ADP-ribosylated RhoA on SDS-PAGE (Fig. 5, compare lanes 5 and 6). In control monocytes (lane 1) and THP-1 cells (lane 3) ADP-ribosylated target of C3 exoenzyme migrated with the same m.w. as the RhoA protein. The treatment of monocytes (lane 2) and THP-1 cells (lane 4) by CNF1 resulted in a change in the mobility of Rho. Thus, CNF1 elicited Rho activation in monocytes and THP-1 cells.

FIGURE 5.

CNF1-stimulated Rho activation. Monocytes (lanes 1 and 2) and THP-1 cells (lanes 3 and 4) were treated (lanes 2 and4) or not treated (lanes 1 and 3) with 12.5 nM CNF1 for 18 h. After freezing/thawing cycles and centrifugation, supernatants were incubated with C3 exoenzyme and [32P]NAD. RhoA fusion protein in the presence (lane 6) or the absence (lane 5) of CNF1 was used as a control. Samples were subjected to 12% SDS-PAGE electrophoresis. The autoradiogram of the mobility shift of Rho represents three experiments.

FIGURE 5.

CNF1-stimulated Rho activation. Monocytes (lanes 1 and 2) and THP-1 cells (lanes 3 and 4) were treated (lanes 2 and4) or not treated (lanes 1 and 3) with 12.5 nM CNF1 for 18 h. After freezing/thawing cycles and centrifugation, supernatants were incubated with C3 exoenzyme and [32P]NAD. RhoA fusion protein in the presence (lane 6) or the absence (lane 5) of CNF1 was used as a control. Samples were subjected to 12% SDS-PAGE electrophoresis. The autoradiogram of the mobility shift of Rho represents three experiments.

Close modal

Some GTPases, such as Rac1 and Rac2, are involved in activation of the NADPH oxidase. The effect of CNF1 on superoxide release was tested by using the SOD-inhibitable reduction of ferricytochrome c. Adherent monocytes (Table I) and THP-1 cells did not release superoxide after 24 h of incubation with CNF1. PMA-stimulated THP-1 cells released low amounts of superoxide (∼2–3 nmol/106 cells), which were not affected by CNF1 treatment (data not shown). PMA at 50 ng/ml stimulated an efficient oxidative response in control monocytes. The CNF1 treatment of monocytes did not modify superoxide production (Table I). Thus, the activation of GTPase Rho is not involved in the activity of NADPH oxidase.

Table I.

Effect of CNF1 on superoxide release by monocytesa

− CNF1+ CNF1
Without PMA 0.4 ± 0.2 0.5 ± 0.3 
With PMA 12.4 ± 2.3 10.8 ± 1.9 
− CNF1+ CNF1
Without PMA 0.4 ± 0.2 0.5 ± 0.3 
With PMA 12.4 ± 2.3 10.8 ± 1.9 
a

Adherent monocytes were treated by CNF1 at 500 pM for 24 h. Then they were stimulated or not by PMA (50 ng/ml) in HBSS containing 120 μM ferricytochrome c for 1 h at 37°C. Superoxide results are expressed as released nanomol/106 cells and represent the mean ± SEM of six experiments.

As phagocytosis requires the reorganization of cytoskeleton, we investigated the effect of CNF1 treatment on FcR- and CR3-dependent phagocytosis using IgG-SRBC and unopsonized zymosan or iC3b-coated zymosan, respectively. CNF1 treatment did not significantly modify IgG-mediated phagocytosis or the uptake of iC3b-coated zymosan in monocytes and THP-1 cells (Table II). In contrast, CNF1 significantly decreased the ingestion of unopsonized zymosan by monocytes (81 ± 8% inhibition; p < 0.01) and THP-1 cells (58 ± 5% inhibition; p < 0.05). These findings suggest that CNF1 specifically affects some features of CR3-mediated functions.

Table II.

Effect of CNF1 on particle phagocytosis by monocytic cellsa

− CNF1+ CNF1
MonocytesTHP-1 cellsMonocytesTHP-1 cells
Zymosan 156 ± 31 59 ± 5 30 ± 13** 25 ± 5b 
iC3b-coated zymosan 240 ± 27 91 ± 8 272 ± 54 95 ± 12 
IgG-SRBC 163 ± 28 139 ± 22 110 ± 23 153 ± 32 
− CNF1+ CNF1
MonocytesTHP-1 cellsMonocytesTHP-1 cells
Zymosan 156 ± 31 59 ± 5 30 ± 13** 25 ± 5b 
iC3b-coated zymosan 240 ± 27 91 ± 8 272 ± 54 95 ± 12 
IgG-SRBC 163 ± 28 139 ± 22 110 ± 23 153 ± 32 
a

Monocytes and THP-1 cells were treated by CNF1 at 500 pM for 24 h. Then they were incubated with 1 mg/ml zymosan, iC3b-coated zymosan, or 2 × 107 IgG-SRBC for 1 h at 37°C. After they were washed, phagocytosis was determined by microscopical examination of at least 200 cells. The results are expressed as the product of the percentage of cells having phagocytosed at least one particle and the number of phagocytozed particles per cell. They represent the mean ± SEM of three experiments.

b

p < 0.05; ** p < 0.01.

We also studied the CR3-dependent adherence of monocytes and THP-1 cells to HUVEC monolayers (Table III). The adherence of control cells to HUVEC did not exceed 20% and was partly dependent on CR3, as demonstrated using specific mAb. CNF1 slightly enhanced the adherence of THP-1 cells to HUVEC, but had no effect on the interaction of monocytes with HUVEC. CNF1 did not affect the adherence between monocytes or THP-1 cells and HUVEC under CR3 control.

Table III.

Effect of CNF1 on monocytic cell-HUVEC adherencea

− CNF1+ CNF1
THP-1 cells 21 ± 2 (41 ± 15) 38 ± 7 (34 ± 5) 
Monocytes 18 ± 6 (33 ± 7) 16 ± 5 (35 ± 6) 
− CNF1+ CNF1
THP-1 cells 21 ± 2 (41 ± 15) 38 ± 7 (34 ± 5) 
Monocytes 18 ± 6 (33 ± 7) 16 ± 5 (35 ± 6) 
a

PBMC and THP-1 cells were treated by CNF1 at 500 pM for 24 h, labeled with calcein AM, and then added to HUVEC monolayers. To assess CR3-mediated monocyte adherence, mAb directed to CD11b (5 μg/ml) was added to adherence assay. Cell-associated fluorescence was measured with a fluorescence multiwell plate reader. Each assay was performed in triplicate, and the results of three experiments are expressed as relative mean adherence. The fraction of CR3-mediated adherence, i.e., the ratio of adherence in the presence and in the absence of anti-CD11b mAb, is shown in parentheses.

CNF1-mediated impairment of the phagocytosis of unopsonized zymosan may be related to the modulation of the expression of phagocyte receptors and/or some activation epitopes. The expression of CD11b, CD18, and CD32 (FcγRII) was studied in THP-1 cells and monocytes by flow cytometry. CNF1 treatment did not modify the expression of CD11b, CD18, and CD32 (Fig. 6). We also investigated the expression of the α-subunit epitope of β2 integrins recognized by mAb 24, which is critical for the activation of these integrins. In the absence of divalent cations, no significant binding of mAb 24 to monocytes was found. When 200 μM Mn2+ was added, the binding of mAb 24 to monocytes increased dramatically. CNF1 treatment of monocytes did not affect the Mn2+-induced expression of epitope 24 (data not shown). The modulation of CR3 activity may also be related to its distribution. We studied the effect of CNF1 on the distribution of CR3 and CD32 by confocal microscopy. CD11b and CD32 were detected at the periphery of control monocytes as a fluorescent ring. In CNF1-treated cells, fluorescence staining was concentrated in peripheral patches (Fig. 7). Similar results were obtained for CD18 (data not shown). Thus, CNF1 clearly induced the clustering of CD11b, CD18, and CD32. We then evaluated the relationship between clusterized distribution of the receptors and F-actin. Monocytes were incubated with mAb directed to CD11b or CD32 (in red) and then with bodipy phallacidin (in green). CD11b and CD32 were colocalized with F-actin (in yellow) at the periphery of resting monocytes. In CNF1-treated monocytes, CD32 and F-actin were colocalized as in control cells. In contrast, CNF1 markedly decreased the association of CD11b with F-actin (Fig. 8). Similarly, the colocalization of CD18 with F-actin was altered by CNF1 treatment (data not shown). The CNF1-stimulated impairment of CR3/F-actin colocalization did not result from monocyte adherence. We found that CR3 was colocalized with F-actin in control THP-1 cells as in adherent monocytes, and that again its distribution with F-actin was altered by CNF1 treatment (data not shown). Hence, CNF1 specifically modulated the activation of CR3 and its colocalization with actin cytoskeleton.

FIGURE 6.

Effect of CNF1 on the expression of membrane receptors. Monocytes were treated with 500 pM CNF1 or medium for 24 h. mAb anti-CD11b, anti-CD18, or anti-CD32 was incubated with monocytes at 1/100 dilution; FITC-tagged F(ab′)2 anti-mouse Igs were subsequently added. The expression of receptors was determined by flow cytometry. Representative histograms of three experiments are shown.

FIGURE 6.

Effect of CNF1 on the expression of membrane receptors. Monocytes were treated with 500 pM CNF1 or medium for 24 h. mAb anti-CD11b, anti-CD18, or anti-CD32 was incubated with monocytes at 1/100 dilution; FITC-tagged F(ab′)2 anti-mouse Igs were subsequently added. The expression of receptors was determined by flow cytometry. Representative histograms of three experiments are shown.

Close modal
FIGURE 7.

Effect of CNF1 on the distribution of CD11b and CD32 in monocytes. Adherent monocytes were treated with 500 pM CNF1 or medium for 24 h. mAb anti-CD11b or anti-CD32 at 1/100 dilution were incubated with monocytes; rhodamine-tagged F(ab′)2 anti-mouse Igs were subsequently added. The distribution of receptors was determined by laser scanning confocal fluorescence microscopy. Representative cells are shown.

FIGURE 7.

Effect of CNF1 on the distribution of CD11b and CD32 in monocytes. Adherent monocytes were treated with 500 pM CNF1 or medium for 24 h. mAb anti-CD11b or anti-CD32 at 1/100 dilution were incubated with monocytes; rhodamine-tagged F(ab′)2 anti-mouse Igs were subsequently added. The distribution of receptors was determined by laser scanning confocal fluorescence microscopy. Representative cells are shown.

Close modal
FIGURE 8.

Effect of CNF1 on the colocalization of CD11b and CD32 with F-actin. Adherent monocytes were incubated with 500 pM CNF1 or medium for 24 h. mAb anti-CD11b or anti-CD32 at 1/100 dilution were incubated with monocytes; rhodamine-tagged F(ab′)2 anti-mouse Igs were added to cell preparations (membrane Ags are in red). Bodipy phallacidin was then added; F-actin appears in green. The colocalization of CD11b or CD32 with F-actin (in yellow) was determined with laser scanning confocal fluorescence microscopy. Representative cells are shown.

FIGURE 8.

Effect of CNF1 on the colocalization of CD11b and CD32 with F-actin. Adherent monocytes were incubated with 500 pM CNF1 or medium for 24 h. mAb anti-CD11b or anti-CD32 at 1/100 dilution were incubated with monocytes; rhodamine-tagged F(ab′)2 anti-mouse Igs were added to cell preparations (membrane Ags are in red). Bodipy phallacidin was then added; F-actin appears in green. The colocalization of CD11b or CD32 with F-actin (in yellow) was determined with laser scanning confocal fluorescence microscopy. Representative cells are shown.

Close modal

We show that CNF1 causes dramatic morphologic changes and profound reorganization of actin cytoskeleton in human monocytic cells. CNF1-stimulated major shape modifications, increase in F-actin content, and F-actin redistribution were delayed, thus emphasizing the necessity of toxin internalization. Indeed, CNF1, like other bacterial toxins, must undergo receptor-mediated endocytosis to be delivered in the cytosol and to exert its action on Rho GTP binding proteins (29). The microinjection of CNF1 into cells confirmed that the toxin acts on the cytosol (30). In addition, the deamidation of Rho induced in vitro by CNF1 is a relatively slow process (21). CNF1-stimulated changes in cell morphology and F-actin organization differed in suspended THP-1 cells and adherent monocytes. CNF1 elicited a polarized shape with several membrane folds in THP-1 cells, whereas thin circumferential lamellipodia were found in CNF1-treated monocytes. In addition, spreading was increased in CNF1-treated monocytes, in agreement with results for Vero cells or Hep-2 cells (22, 30). In CNF1-treated THP-1 cells, two patterns of F-actin organization were found: a dense peripheral ring consisting of numerous foci of F-actin and a polarized distribution of F-actin. In monocytes, actin cytoskeleton was profoundly disorganized, with cell projections or invaginations rich in F-actin and some intracellular structures evoking actin cables. The differences in CNF1-induced morphologic changes and distribution of F-actin between THP-1 cells and monocytes are probably related to cell adherence, which promotes the immobilization of some cytoskeletal structures (31). The cytoskeletal modifications result from CNF1 treatment, but are not related to the secretion of soluble molecules such as cytokines by monocytes in response to CNF1. The contamination of CNF1 preparations by endotoxins, potent inducers of cytokine release, could be sufficient to elicit the secretion of soluble molecules. CNF1 preparations were filtered on sulfone to remove bacterial endotoxins before use. Furthermore, cytoskeletal changes induced by CNF1 were identical with and without polymyxin B and could not be mimicked by purified LPS from E. coli (data not shown). Cytokines such as TNF-α and chemokines were not secreted by monocytes in response to CNF1 (data not shown), suggesting that CNF1-induced cytoskeleton reorganization did not depend on cytokine release.

The cell shape changes and cytoskeletal reorganization induced by CNF1 directly involve the GTPase Rho but not Cdc42 or Rac proteins. The injection of constitutively active Cdc42 into Bac1 macrophages elicited the formation of long filopodia (17) that were not observed in CNF1-treated monocytes or THP-1 cells (our results). CNF1-stimulated shape changes in adherent monocytes and in THP-1 cells may involve the activation of Rho or Rac proteins. The morphologic modifications after microinjection of constitutively active Rac1 in Bac1 cells (17) evoke some features of actin reorganization in CNF1-treated monocytes. We took advantage of the property that Rac has of activating phagocytic NADPH oxidase (12, 13) to assess the effect of CNF1 on Rac proteins in monocytic cells. Treating monocytes and THP-1 cells only with CNF1 did not induce any oxidative response and did not modify the superoxide generation stimulated by phorbol ester. Hence, CNF1-induced morphologic changes and actin reorganization did not depend on the activation of Rac proteins but required Rho GTPase activity. Three series of data further support this statement. CNF1 decreased the electrophoretic mobility of Rho but not that of Rac (21). CNF1 stimulated a shift in the electrophoretic mobility of ADP-ribosylated Rho in both monocytes and THP-1 cells (our results). Moreover, pretreatment of monocytic cells with C3 exoenzyme before addition of CNF1 prevented morphologic changes (data not shown).

The activation of Rho affected the CR3-dependent phagocytosis of unopsonized zymosan but did not modify the ingestion of IgG-SRBC. These results were not related to changes in receptor expression, since the expression of CR3 and FcγRII (CD32) was similar in control and CNF1-treated cells. There is a growing body of evidence that Rho is important for the clustering of receptors including integrins and FcγR (8, 32). We also found that CNF1 induced the clustering of CR3 and FcγRII. The clustering of integrins may occur through the modulation of actin-myosin tension and cell contractility (33). Indeed, lysophosphatidic acid-mediated activation of Rho resulted in contracted morphology, phosphorylation of myosin light chain, and aggregation of integrins in fibroblasts (21). CNF1 stimulates the relocalization of myosin in stress fibers of epithelial cells; butanedione monoxime, an inhibitor of myosin ATPase and contractility, prevents stress fibers and the relocalization of myosin (30). As butanedione monoxime inhibited morphologic changes in CNF1-treated monocytes (data not shown), contractility may account for the cytoskeletal rearrangements and receptor clustering. Nevertheless, the clustering of CR3 and FcγRII cannot account for the selective down-modulation of CR3-dependent phagocytosis.

Rho activation affected only some features of CR3 functions. Since CNF1 did not modify the monocyte-HUVEC adherence dependent on CR3 or the uptake of iC3b-coated zymosan, it is probable that the activation of Rho modifies the functional state of CR3. Indeed, lymphocyte aggregation is inhibited by C3 exoenzyme through avidity change of LFA-1 (34). Similarly, C3 exoenzyme blocks agonist-induced neutrophil β2 integrin adherence to fibrinogen (9). Thus, one potential mechanism of the CR3 impairment may be the prevention of an active conformation. Divalent cations induce the expression of an activation epitope recognized by mAb 24 on the α subunit of β2 integrins (35). This mAb inhibits LFA-1-dependent, Ag-specific T cell proliferation and CR3-mediated neutrophil chemotaxis to FMLP (36). We found that the treatment of monocytes with CNF1 did not affect Mn2+-induced expression of epitope 24. It has been demonstrated that some lectin sites on CR3 are able to bind zymosan polysaccharides and β-glucan. The fact that these sites are located outside the CD11b I domain that contains the binding sites for iC3b, ICAM-1, and fibrinogen (37) means that CNF1 selectively affects the CR3 functional state, probably through lectin sites.

As the interaction of integrins, including β2 integrins, with the cytoskeleton modulates their activity (38, 39), Rho-mediated cytoskeletal reorganization may affect the activity of CR3. The cytoskeletal distribution of CR3 was clearly distinct from that of FcγRII. First, the association of FcγRII with F-actin in resting monocytes probably resulted from their adhesion to substrate, since FcγRII was not localized with F-actin in suspended THP-1 cells. Second, the activation of the GTPase Rho did not affect the cytoskeletal distribution of FcγRII in monocytes and THP-1 cells. Thus, Rho-mediated F-actin reorganization did not impair FcγRII distribution and FcγRII-dependent phagocytosis. This may be related to the inability of cytoskeletal inhibitors to affect FcR diffusion in adherent macrophages (40). We show here that CNF1 decreased the colocalization of CR3 with F-actin. Since CNF1 treatment is also associated with the impairment of unopsonized zymosan phagocytosis, we suggest that the distribution of CR3 with actin cytoskeleton is necessary for zymosan phagocytosis.

In conclusion, CNF1 dramatically affects actin cytoskeleton in monocytes and specifically impairs some features of CR3-dependent phagocytosis. It also acts on both the activation of CR3 and its codistribution with actin cytoskeleton. The GTPase Rho differentially controls CR3 and FcγR pathways of phagocytosis. Although the activation of Rho creates the driving strength for pseudopodal formation and the ingestion process in epithelial cells, it appears to restrict ingestion in professional phagocytes. By limiting integrin-mediated uptake, CNF1 may prevent the entry of several other pathogens and maintain a favorable ecologic niche for pathogenic strains of E. coli.

We thank the Centre de Microscopie Electronique (Faculté de Médecine, Marseille, France) for scanning electron microscopy.

2

Abbreviations used in this paper: CNF1, cytotoxic necrotizing factor-1; F-actin, filamentous actin; NAD, nicotinamide adenine dinucleotide; SOD, superoxide dismutase.

1
Nobes, C. D., A. Hall.
1995
. Rho, Rac and Cdc42 GTPases regulate the assembly of multimolecular focal complexes associated with actin stress fibers, lamellipodia, and filopodia.
Cell
81
:
53
2
Aktories, K..
1997
. Bacterial toxins that target Rho proteins.
J. Clin. Invest.
99
:
827
3
Hall, A..
1994
. Small GTP-binding proteins and the regulation of the actin cytoskeleton.
Annu. Rev. Cell Biol.
10
:
31
4
Ridley, A. J., A. Hall.
1992
. The small GTP-binding protein Rho regulates the assembly of focal adhesions and actin stress fibers in response to growth factors.
Cell
70
:
389
5
Kozma, R., S. Ahmed, A. Best, L. Lim.
1995
. The Ras-related protein Cdc42Hs and bradykinin promote formation of peripheral actin microspikes and filopodia in Swiss 3T3 fibroblasts.
Mol. Cell. Biol.
15
:
1942
6
Ridley, A. J., H. F. Paterson, C. L. Johnston, D. Diekmann, A. Hall.
1992
. The small GTP-binding protein Rac regulates growth factor-induced membrane ruffling.
Cell
70
:
401
7
Chavrier, P., J. P. Gorvel, J. Bertoglio.
1993
. An immunologist’s look of the Rho and Rab GTP-binding proteins.
Immunol. Today
14
:
440
8
Machesky, L. M., A. Hall.
1996
. Rho: a connection between membrane receptor signalling and the cytoskeleton.
Trends Cell Biol.
6
:
304
9
Laudanna, C., J. J. Campbell, E. C. Butcher.
1996
. Role of Rho in chemoattractant-activated leukocyte adhesion through integrins.
Science
271
:
981
10
Gomez, J., A. Garcia, L. R. Borlado, P. Bonay, C. Martinez-A, A. Silva, M. Fresno, A. C. Carrerra, C. Eicher-Streiber, A. Rebollo.
1997
. IL-2 signaling controls actin organization through Rho-like protein family, phosphatidyl 3-kinase, and protein kinase C-ζ.
J. Immunol.
158
:
1516
11
Galandrini, R., S. W. Henning, D. A. Cantrell.
1997
. Different functions of the GTPase Rho in prothymocytes and late pre-T cells.
Immunity
7
:
163
12
Abo, A., E. Pick, A. Hall, N. Totty, C. G. Teahan, A. W. Segal.
1991
. Activation of the NADPH oxidase involves the small GTP-binding protein p21rac1.
Nature
353
:
668
13
Knaus, U. G., P. G. Heyworth, T. Evans, J. T. Curnutte, G. M. Bokoch.
1991
. Regulation of phagocyte oxygen radical production by the GTP-binding protein Rac 2.
Science
254
:
1512
14
Symons, M., J. M. Derry, B. Karlak, S. Jiang, V. Lemahieu, F. McCormick, U. Francke, A. Abo.
1996
. Wiskott-Aldrich syndrome protein, a novel effector for the GTPase CDC42Hs, is implicated in actin polymerization.
Cell
84
:
723
15
Aepfelbacher, M., M. F. Vauti, P. C. Weber, J. A. Glomset.
1994
. Spreading of differentiating human monocytes is associated with a major increase in membrane-bound CDC42.
Proc. Natl. Acad. Sci. USA
91
:
4263
16
Aepfelbacher, M., M. Essler, E. Huber, A. Czech, P. C. Weber.
1996
. Rho is a negative regulator of human monocyte spreading.
J. Immunol.
157
:
5070
17
Allen, W. E., G. E. Jones, J. W. Pollard, A. J. Ridley.
1997
. Rho, Rac and Cdc42 regulate actin organization and cell adhesion in macrophages.
J. Cell Sci.
110
:
707
18
Greenberg, S..
1995
. Signal transduction of phagocytosis.
Trends Cell Biol.
5
:
93
19
Stossel, T. P..
1992
. The mechanical responses of white blood cells. J. I. Gallin, and I. M. Goldstein, and R. Snyderman, eds.
Inflammation: Basic Principles and Clinical Correlates
459
Raven Press, New York.
20
Fiorentini, C., G. Donelli, P. Matarrese, A. Fabbri, S. Paradisi, P. Boquet.
1995
. Escherichia coli cytotoxic necrotizing factor 1: evidence for induction of actin assembly by constitutive activation of the p21 Rho GTPase.
Infect. Immun.
63
:
3936
21
Flatau, G., E. Lemichez, M. Gauthier, P. Chardin, S. Paris, C. Fiorentini, P. Boquet.
1997
. Toxin-induced activation of the G protein p21 Rho by deamidation of glutamine.
Nature
387
:
729
22
Falzano, L., C. Fiorentini, G. Donelli, E. Michel, C. Kocks, P. Cossart, L. Cabanié, E. Oswald, P. Boquet.
1993
. Induction of phagocytic behaviour in human epithelial cells by Escherichia coli cytotoxic necrotizing factor type 1.
Mol. Microbiol.
9
:
1247
23
Capo, C., Y. Zaffran, P. Houpikian, D. Raoult, J. L. Mege.
1996
. Production of interleukin-10 and transforming growth factor β by peripheral blood mononuclear cells in Q fever endocarditis.
Infect. Immun.
64
:
4143
24
Sanguedolce, M. V., C. Capo, P. Bongrand, J. L. Mege.
1992
. Zymosan-stimulated tumor necrosis factor-α production by human monocytes: down-modulation by phorbol ester.
J. Immunol.
148
:
2229
25
Lepidi, H., Y. Zaffran, J. L. Ansaldi, J. L. Mege, C. Capo.
1995
. Morphological polarization of human polymorphonuclear leucocytes in response to three different chemoattractants: an effector response independent of calcium rise and tyrosine kinases.
J. Cell Sci.
108
:
1771
26
Sanguedolce, M. V., C. Capo, M. Bouhamdan, P. Bongrand, C. K. Huang, J. L. Mege.
1993
. Zymosan-induced tyrosine phosphorylations in human monocytes: role of protein kinase C.
J. Immunol.
151
:
405
27
Newman, S. L., L. K. Mikus.
1985
. Deposition of C3b and iC3b onto particulate activators of the human complement system: quantitation with monoclonal antibodies to human C3.
J. Exp. Med.
161
:
1414
28
Dignat-George, F., N. Teysseire, M. Mutin, N. Bardin, G. Lesaule, D. Raoult, J. Sampol.
1997
. Rickettsia conorii infection enhances vascular cell adhesion molecule-1- and intercellular adhesion molecule-1-dependent mononuclear cell adherence to endothelial cells.
J. Infect. Dis.
175
:
1142
29
Sandvig, K., E. Dubinina, O. Garred, K. Prydz, J. V. Kozlov, S. H. Hansen, B. Van Deurs.
1992
. Protein toxins: mode of action and cell entry.
Biochem. Soc. Trans.
20
:
724
30
Fiorentini, C., A. Fabbri, G. Flatau, G. Donelli, P. Matarrese, E. Lemichez, L. Falzano, P. Boquet.
1997
. Escherichia coli cytotoxic necrotizing factor 1 (CNF1), a toxin that activates the Rho GTPase.
J. Biol. Chem.
272
:
19532
31
Cassimeris, L., H. McNeill, S. H. Zigmond.
1990
. Chemoattractant-stimulated polymorphonuclear leukocytes contain two populations of actin filaments that differ in their spatial distributions and relative stabilities.
J. Cell Biol.
110
:
1067
32
Hackam, D. J., O. D. Rotstein, A. Schreiber, W. J. Zhang, S. Grinstein.
1997
. Rho is required for the initiation of calcium signaling and phagocytosis by Fcγ receptors in macrophages.
J. Exp. Med.
186
:
955
33
Burridge, K., M. Chrzanowska-Wodnicka.
1996
. Focal adhesions, contractility and signaling.
Annu. Rev. Cell. Dev. Biol.
12
:
463
34
Tominaga, T., K. Sugie, M. Hirata, N. Morii, J. Fukata, A. Uchida, H. Imura, S. Narumiya.
1993
. Inhibition of PMA-induced, LFA-1-dependent lymphocyte aggregation by ADP-ribosylation of the small molecular weight GTP-binding protein, Rho.
J. Cell Biol.
120
:
1529
35
Dransfield, I., C. Cabanas, A. Craig, N. Hogg.
1992
. Divalent cation regulation of the function of the leukocyte integrin LFA-1.
J. Cell Biol.
116
:
219
36
Dransfield, I., C. Cabanas, J. Barrett, N. Hogg.
1992
. Interaction of leukocyte integrins with ligand is necessary but not sufficient for function.
J. Cell Biol.
116
:
1527
37
Thorton, B. P., V. Vetvicka, M. Pitman, R. C. Goldman, G. D. Ross.
1996
. Analysis of the sugar specificity and molecular location of the β-glucan-binding lectin site of complement receptor type 3 (CD11b/CD18).
J. Immunol.
156
:
1235
38
Mould, A. P..
1996
. Getting integrins into shape: recent insights into how integrin activity is regulated by conformational changes.
J. Cell Sci.
109
:
2613
39
Elemer, G. S., T. S. Edgington.
1994
. Microfilament reorganization is associated with functional activation of αMβ2 on monocytic cells.
J. Biol. Chem.
269
:
3159
40
Michl, J., M. M. Pieczonka, J. C. Unkeless, G. I. Bell, S. C. Silverstein.
1983
. Fc receptor modulation in mononuclear phagocytes maintained on immobilized immune complexes occurs by diffusion of receptor molecules.
J. Exp. Med.
157
:
2121