INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Crohn's disease (CD) is an inflammatory bowel disease characterized by a chronic transmural, segmental, and typically granulomatous inflammation of the intestine in humans. CD lesions develop from an accumulation of lymphocytes and plasma cells, followed by an influx of macrophages, which transform into epithelioid cells and generate nascent granulomas (for a review see reference 12).

The etiology of CD is still unknown. The general concept is that CD pathogenesis is immunologically mediated and greatly influenced by genetic, environmental, and other endogenous host factors. Infectious agents as a possible environmental cause of CD have been investigated since the disease was first recognized. Some characteristic pathologic elements of CD, including aphthous ulcers of the mucosa, mural abscesses, and macrophage and epithelioid cell granulomas, occur in well-recognized infectious diseases, like shigellosis, salmonellosis, and yersinial enterocolitis, in which invasiveness is an essential virulence factor of the bacteria involved (for a review see reference 40).

We previously showed that Escherichia coli is abnormally predominant (between 50 and 100% of the total number of aerobes and anaerobes) in early and chronic ileal lesions of CD and that a great majority of the E. coli strains isolated from Crohn's ileal mucosa adhere strongly in vitro to intestinal epithelial cells (11). The latter property could enable the bacteria to colonize the intestinal mucosa. Qualitative and quantitative analysis of E. coli strain LF82 isolated from a chronic ileal lesion of a patient with CD revealed that it is a true invasive pathogen (6). It efficiently invades a wide range of human epithelial cell lines in vitro, including HEp-2 cells and the intestinal cells Intestine-407, Caco-2, and HCT-8. In contrast with most invasive bacteria belonging to the Enterobactericeae family, its uptake is dependent upon both functioning host cell actin microfilaments and microtubules. It survives and replicates in the host cell cytoplasm after lysis of the endocytic vacuole. Electron microscope examination of LF82-infected epithelial cells revealed a macropinocytosis-like process of entry, characterized by the elongation of membrane extensions which surrounded the bacteria at the sites of contact between the entering bacteria and the epithelial cells (5). The invasive process was found to be original in that LF82 possessed none of the known genetic invasive determinants described for enteroinvasive, enteropathogenic, or enterotoxigenic E. coli, or Shigella flexneri, i.e., the ipaC plasmid gene encoding the invasin of S. flexneri, and enteroinvasive E. coli (EIEC), the eae gene encoding the intimin of enteropathogenic E. coli (EPEC), and the tia gene encoding a 25-kDa outer membrane protein involved in enterotoxigenic E. coli (ETEC) invasiveness (6, 7). Recent data have revealed a high prevalence of invasive strains associated with the ileal mucosa of patients with CD compared to that for controls, supporting a putative role of E. coli invasiveness in the pathogenesis of CD (7, 29). Characterization of these strains revealed that they shared the same mechanism of entry as LF82 and that they lacked any of the invasive genetic determinants described above. They were thus clustered in a new potentially pathogenic group of invasive E. coli, which we designate AIEC for adherent invasive E. coli (6, 7).

In CD mucosa the numbers of macrophages and antigen-presenting dendritic cells are increased (37). Macrophages residing in the intestine or attracted to the site of inflammation might normally serve as a first line of defense by nonspecifically eliminating microorganisms that have penetrated from the intestinal lumen. Invasive bacteria have developed various strategies to counteract these mechanisms, which, when successful, enable them to survive and multiply and sometimes to destroy macrophages (for reviews see references 1, 13, 16, and 31). The behavior of the microorganisms within macrophages and host responses are key components in the replication and perpetuation of these intestinal pathogens. Elicitation of an inflammatory cascade may either eliminate the invading bacteria or facilitate further bacterial invasion (33, 34, 42).

As interactions between bacteria and macrophages dictate the outcome of most of the infectious diseases, the aim of the present work was to analyze the behavior within macrophages of AIEC strains isolated from patients with CD. Three different types of macrophages were used: the universally used J774-A1 murine macrophage-like cell line, which allowed us to compare the behavior of AIEC with other invasive bacteria studied so far, murine peritoneal macrophages, and human monocyte-derived macrophages (HMDM). The survival and replication of the phagocytosed bacteria were studied, as well as the behavior of the infected macrophages, by analyzing the synthesis and release of proinflammatory cytokines and cell integrity.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains and culture conditions. A total of 15 invasive E. coli strains (LF strains) isolated from patients with CD were tested, including the prototype AIEC strain LF82 of serotype O83:H1, which was isolated from a chronic lesion of a patient with CD (6). The strains were recovered from ileal biopsy specimens from patients with chronic lesions who had been operated on for CD (n = 7), from patients with early endoscopic recurrent lesions at 3, 6, or 12 months postsurgery (n = 7), and from a patient without recurrence of CD up to 2 years postsurgery (n = 1) (11).

Cell line and cell culture. The murine macrophage-like cell line J774-A1 (American Type Culture Collection no. TIB67) was maintained in an atmosphere containing 5% CO₂ at 37°C in RPMI 1640 medium (Seromed, Biochrom KG, Berlin, Germany) supplemented with 10% (vol/vol) heat-inactivated fetal calf serum (FCS; Seromed), 1% L-glutamine (Life Technologies), 100,000 of penicillin/liter, 100 mg of streptomycin/liter, and 25 µg of amphotericin B/liter. J774-A1 cells were seeded in 24-well tissue culture plates (Polylabo, Strasbourg, France) at a density of 10⁵ cells per cm² and were grown for 18 h in an atmosphere containing 5% CO₂ at 37°C.

Isolation and culture of murine peritoneal macrophages. Peritoneal cells were isolated by intraperitoneal (i.p.) injection of 5 ml of RPMI 1640 (Seromed) supplemented with 5% (vol/vol) heat-inactivated FCS (Seromed), 1% L-glutamine (Life Technologies), 100,000 U of penicillin/liter, 100 mg of streptomycin/liter, and 25 µg of amphotericin B/liter into OF1 mice (IFFA CREDO, Chatillon sur Chalaronne, France), that had been killed by cervical dislocation. Following i.p. injection, mice were shaken to dislodge peritoneal cells, and the lavage fluids were removed by syringe. The resulting cells, in the above-mentioned medium, were placed in 24-well tissue culture plates (Polylabo) at a density of 5 × 10⁵ cells per well and maintained in an atmosphere containing 5% CO₂ at 37°C. Two hours later, the medium was removed, and the plastic adherent cells were washed three times in 0.5 ml of phosphate-buffered saline (PBS; pH 7.2) and then incubated in 0.5 ml of complete RPMI medium as above at 37°C in a 5% CO₂ incubator. Eighteen hours later, the culture medium was aspirated, and the cells were washed twice with PBS and then incubated with 0.5 ml of the above culture medium lacking antibiotics.

Isolation and culture of HMDM. Isolation of HMDM was performed as described by Libby et al. (27). Briefly, blood was collected from normal donors in heparinized syringes and in additional sterile glass tubes for use as autologous serum. Monocytes were purified by density gradient centrifugation using Isoprep and a differential adherence procedure with 2% gelatin-coated flasks. Nonadherent cells were removed by washes with RPMI, and adherent cells were eluted with RPMI plus 10 mM EDTA in PBS. The eluted cells were centrifuged and resuspended in RPMI containing 15% autologous serum. Monocytes were seeded into 24-well culture plates at a density of 5 × 10⁵ cells per well. The monocytes were maintained in an atmosphere containing 5% CO₂ at 37°C for 5 to 7 days until they differentiated into cells with characteristic macrophage morphology.

Bacterial survival and replication in macrophages. Bacterial uptake, survival, and replication were measured by the gentamicin protection assay. The MBC (concentration that reduced the bacterial count by 99.99%) of the antibiotic for all strains included in this study was determined in tissue culture medium, and the drug was used at 15- to 100-fold the MBC (the MBC ranged from 0.5 to 4 µg/ml)

Transmission electron microscopy (TEM). Cross sections of J774-A1 cells were prepared as follows. After infection, cells were fixed with 3% glutaraldehyde in 0.2 M cacodylate buffer at 4°C for 2 h and postfixed in 1% OsO₄ in cacodylate buffer at 4°C for 1 h. After dehydration in a graded series of ethanol, the cultures were embedded in a 2-mm-thick Epon coating in the tissue culture well and polymerized for 3 days at 60°C. Suitable areas were oriented parallel to the cell layer surface on Epon blocks with an Epon mixture. Ultrathin sections were contrasted with uranyl acetate and lead citrate.

LDH activity. J774-A1 cells were seeded in 24-well tissue culture plates at a density of 2 × 10⁵ cells/cm² and grown for 18 h. They were then infected at MOIs of 10 and 100, and processed as described above. Supernatants of the infected macrophages were sampled at 1, 4, 8, and 24 h of gentamicin treatment, centrifuged at 2,500 × g for 3 min at 4°C, and assayed for lactate dehydrogenase (LDH) activity. Enzymatic activity was determined in supernatants by using NADH as a substrate (LDH kit; Boehringer, Mannheim, Germany). Release of LDH was expressed as units per liter of supernatant. The percent cytotoxicity was calculated as (experimental release  spontaneous release)/(total release  spontaneous release) × 100, where spontaneous release is the amount of LDH activity in supernatants of cells incubated in medium alone and total release is the LDH activity measured in macrophage lysates.

Annexin V binding and propidium iodide staining of infected macrophages. J774-A1 cells were seeded in 24-well tissue culture plates at a density of 2 × 10⁵ cells per well. At various time points after infection at an MOI of 100 or after treatment with 5 µM gliotoxin (Sigma), cells and supernatant were harvested and stained with fluorescein isothiocyanate (FITC)-labeled annexin V and propidium iodide (annexin V-FITC apoptosis detection kit; Sigma) according to the supplier's protocol. With this protocol, live cells are not stained by either propidium iodide or annexin V-FITC. Cells which appear early in the apoptotic process are stained with annexin V-FITC alone, and necrotic cells are stained by both propidium iodide and annexin V-FITC. The percent stained cells was scored by analyzing 100 cells of three independent experiments in duplicate using a Zeiss fluorescence microscope.

DNA fragmentation. J774-A1 cells seeded at a density of 3.5 × 10⁶ were infected for 2 h at an MOI of 100, as described above. After a 24-h gentamicin treatment, total genomic DNA was extracted from adherent and floating cells by a standard DNA extraction method (proteinase K and phenol-chloroform), ethanol precipitated, and incubated for 30 min in the presence of 1 µg of RNase (Boehringer)/ml. The DNA solution was analyzed by electrophoresis for 5 h in a 1.5% agarose-0.1% sodium dodecyl sulfate (SDS) gel and stained with ethidium bromide.

Detection of cytokines IL-1 and TNF- in the supernatant of infected macrophages. Macrophages seeded at a density of 2 × 10⁵ cells/cm² were infected at an MOI of 100, and supernatants were collected, centrifuged, and stored at 20°C. For interleukin-1 (IL-1) only, cells were prestimulated 18 h before infection with 1 µg of E. coli O111:B4 lipopolysaccharide (LPS; Sigma)/ml. The amounts of murine cytokines IL-1 and tumor necrosis factor alpha (TNF-) released in the culture supernatant or remaining in cells were determined by enzyme-linked immunosorbent assay (ELISA) (Endogen Inc., Woburn, Mass.). The optical density was determined at a wavelength of 450 nm, and cytokine concentrations in picograms per milliliter were assessed according to the manufacturer's instructions.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Survival and replication of AIEC strains within the murine J774-A1 macrophage-like cell line. In order to examine the resistance of AIEC strains to the intracellular killing by macrophages, murine J774-A1 macrophages were infected with 15 AIEC strains. Unless otherwise stated, a 2-h infection at an MOI of 10 was performed in all the experiments reported. The number of intracellular bacteria surviving the gentamicin exposure was determined after 1, 4, 8, 24, and 48 h of gentamicin treatment. Bacterial uptake was quantified at 1 h postinfection and expressed as CFU per well. Survival at 4, 8, 24, and 48 h postinfection was expressed as mean percentages of the number of bacteria recovered after 1 h of gentamicin treatment, defined as 100% (Table 1).

View larger version (13K): [in this window] [in a new window]   FIG. 1.   Kinetics of intracellular multiplication of AIEC LF82 in J774-AI macrophages. The numbers of viable LF82 bacteria were determined in attached () and detached () macrophage fractions after different times (hours) of gentamicin treatment. Results are expressed as the number of intracellular bacteria relative to that obtained at 1 h after gentamicin exposure, taken as 100%. Means and standard errors of the means are indicated and correspond to six different experiments.

Survival within murine resident peritoneal macrophages. Assays were performed at an MOI of 10 with AIEC strain LF82, serovar Cholereasuis, and nonpathogenic E. coli K-12 C600. Results presented in Fig. 2 show only the intracellular behaviors of AIEC strain LF82 and serovar Choleraesuis. We were unable to count any intracellular bacteria for E. coli K-12 at the postinfection time points. Moreover, experiments performed at an MOI of 100 revealed that E. coli K-12 C600 was efficiently killed as early as 4 h postinfection (data not shown), confirming the very high bactericidal efficiency of these macrophages. The number of serovar Choleraesuis intracellular bacteria recovered at 4 h represented 34.1% of that recovered at 1 h. This dropped to 1.4% at 8 h postinfection in the attached-macrophage fraction. In the detached fraction, 16.0% of the initially phagocytosed bacteria were counted at 8 h postinfection, reflecting the Salmonella-induced apoptotic process. In contrast, even though the number of AIEC LF82 bacteria decreased relative to the number initially internalized, 52.0% of the bacteria recovered at 1 h postinfection were also recovered at 4 h postinfection. Hence, the number of live bacteria remained constant over a period of 8 h postinfection and decreased to 14.2% at 24 h postinfection. Only very few bacteria were counted in the detached cells over the postinfection time course.

View larger version (12K): [in this window] [in a new window]   FIG. 2.   Survival of AIEC LF82 (A) and S. enterica serovar Choleraesuis (B) in murine peritoneal macrophages. Means and standard errors of the mean are indicated and correspond to four different experiments in duplicate wells. Bacterial CFU per well containing 5 × 105 macrophages (y axis) and the time after addition of gentamicin (x axis) are shown. Numbers of viable bacteria were determined in the attached () and detached () macrophage fractions.

Survival of AIEC strain LF82 within HMDM. In order to confirm the ability of AIEC to survive in human phagocytic cells, experiments were conducted using the AIEC reference strain LF82 with HMDM. The serovar Dublin strain SL2260, which is well characterized for its behavior in HMDM, was included in our experiments for comparative purposes. Detached and attached macrophage fractions were cultured separately to enumerate intracellular bacteria. The results are shown in Fig. 3. Results with serovar Dublin SL2260 were in accordance with already published data (27). The number of CFU recovered after infection with serovar Dublin SL2260 increased slightly in the attached macrophages over 24 h and reached 122% of the number of bacteria recovered after 1 h postinfection, taken as 100%. Bacteria in detached macrophages were observed as early as 4 h postinfection, and their number increased over the postinfection period. The number of bacteria in the attached fraction of AIEC LF82-infected macrophages increased over 4 h postinfection to reach 208% of the initially internalized bacteria and decreased slightly over 48 h postinfection. However, at 48 h, the percent intracellular bacteria still represented 96% of the bacteria recovered after 1 h of gentamicin treatment. In contrast to serovar Dublin-infected macrophages, no bacteria were recovered in the corresponding detached fractions, confirming the lack of cytotoxicity induced by AIEC LF82 infection of J774-A1 macrophages. With the nonpathogenic strain K-12 C600, a decrease in the number of viable bacteria was observed as early as 4 h postinfection and only 5% of the bacteria initially internalized were still alive after 24 h postinfection.

View larger version (16K): [in this window] [in a new window]   FIG. 3.   Survival of AIEC LF82 (A) compared to those of serovar Dublin SL2260 (B) and nonpathogenic E. coli K-12 C600 (C) in HMDM. Each time point represents the mean of three independent experiments in triplicate wells. Bacterial CFU per well containing 5 × 105 macrophages (y axis) and the time after addition of gentamicin (x axis) are shown. Numbers of viable bacteria were determined in the attached () and detached () macrophage fractions.

AIEC strain LF82 did not induce cell death of J774-A1 infected macrophages. All experiments to analyze the cell integrity of infected macrophages were performed at MOIs of 10 and 100, to check whether a higher number of internalized bacteria was able to induce cell death. The culture supernatants of infected or noninfected macrophages were assayed for the presence of the cytoplasmic enzyme LDH to estimate the membrane integrity of the cells. Samples of the culture medium were tested at 1, 4, 8, and 24 h postinfection (Table 2). S. flexneri M90T and serovar Choleraesuis were included as positive controls, since they are known to induce cell death (30, 42). E. coli K-12 C600 was used as a negative control. The amounts of LDH released were expressed as LDH activity recovered in the supernatant or as a percentage relative to that obtained at 1 h.

View larger version (49K): [in this window] [in a new window]   FIG. 4.   DNA electrophoresis of J774-A1 cells at 24 h postinfection. Lanes: L, ladder of molecular size markers in base pairs (SmartLadder SF; Eurogentec); 1, uninfected cells; 2, cells infected with the AIEC strain LF82; 3, cells infected with S. flexneri M90T. Only M90T-infected cells generated ladders of 180 to 200 bp, characteristic of apoptosis. LF82-infected cells were negative for DNA degradation.

AIEC LF82 induced spacious phagosomes. Morphological data to confirm the survival and replication abilities of AIEC LF82 within J774-A1 cells were obtained with TEM. One hour after infection, LF82 bacteria were observed in small vacuoles (Fig. 5A). LF82-infected cells exhibited larger phagosomes at 8 h, some of which were observed to fuse, leading to the formation of spacious phagosomes (Fig. 5B). At 24 h postinfection, infected macrophages showed a particularly prominent vacuole containing numerous bacteria. None of the internalized bacteria were free in the cytoplasm. The phagosomes occupied most of the cytoplasmic compartment and did not seem to alter the morphology of the nucleus or that of the cytoplasmic membrane (Fig. 5C). There was no morphological evidence of cytotoxicity, confirming the results of LDH release and DNA analysis. Despite this large number of intracellular bacteria, infected macrophages appeared viable even after 48 h postinfection (data not shown). In contrast, after 24 h, macrophages with internalized E. coli K-12 C600 harbored only a few phagosomes and also had small, round vesicular structures in phagosome vacuoles that appeared to represent fragments of organisms (data not shown). Thus, AIEC LF82 can persist within phagocytic cells for a long period in a large vacuole without inducing any apparent cell death or damage to the infected macrophages.

View larger version (94K): [in this window] [in a new window]   FIG. 5.   TEMs of J774-A1 macrophages infected at an MOI of 10 bacteria per cell. At 1 h postinfection, AIEC LF82 bacteria were observed in small vacuoles (A). After 8 h of gentamicin treatment, the size of the vacuoles was increased (B). Note phagosome fusion leading to the formation of a large vacuole observed after 24 h of gentamicin treatment (C). Magnification: ×7,200.

AIEC LF82 induced secretion by macrophages of TNF- but no IL-1 release. To explore the possible mechanisms in an inflammatory response induced by AIEC LF82, we examined its ability to induce (IL-1) release and TNF- secretion by ELISA.

View this table: [in this window] [in a new window]   TABLE 4.   Production of proinflammatory cytokines IL-1 and TNF- by J774-A1 cells after AIEC LF82 infection

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

The search for infectious agents as a possible cause of CD has mainly focused on intracellular pathogens which have evolved to resist phagocytosis and often persist for long periods within macrophages leading to chronic antigenic stimulation and T-cell and macrophage activation. This results in the formation of granulomas surrounding the microbes. Granulomatous inflammation is not only a histologic hallmark of infections with some intracellular bacteria and fungi but also a characteristic feature of CD, with granulomas found in the bowel wall and in regional lymph nodes (12). Moreover the presence of specific E. coli antigens was previously found in macrophages distributed within the lamina propria, in granulomas, and in the germinal centers of mesenteric lymph nodes in patients with CD (9, 28). We previously reported that the ileal mucosa of patients with CD is abnormally colonized by pathogenic E. coli strains termed AIEC, which have the ability to adhere to and invade epithelial intestinal cultured cells (6, 11). The aim of the present work was to determine the intramacrophagic fate of AIEC strains isolated from patients with CD and the behavior of the infected macrophages in order to evaluate the ability of such invasive bacteria to participate in granuloma formation.

We showed in the present study that all the AIEC strains that we isolated from patients with CD were able to survive and to replicate extensively within J774-A1 phagocytic cells. After 48 h postinfection, the number of intracellular AIEC bacteria increased 2.2- to 74.2-fold over the number at initial infection, depending on the AIEC strains studied. Moreover, the AIEC strain LF82 was able to persist efficiently until 5 days postinfection within J774-A1 macrophages. No detached cells, no LDH release, no annexin V-FITC binding, and no sign of DNA fragmentation or degradation were observed with J774-A1 infected cells even after 24 h postinfection, indicating that there was no cell death. The behavior of the AIEC strains within macrophages is different from that of other invasive bacteria, since most of them induce cell death of the infected macrophages (for a review see reference 31). Through cell death, infected macrophages attempt to both prevent the spread of intracellular pathogens and elicit a potent antibacterial immune response via the release of proinflammatory cytokines.

Comparative studies of intracellular survival of AIEC LF82 within two cell models indicated that this strain had a survival advantage in HMBM in comparison with murine peritoneal macrophages. However, in comparison to published data concerning the intracellular resistance of various Salmonella serovars or Shigella within murine peritoneal macrophages at early postinfection times before induction of macrophage cell death, the fraction of AIEC LF82 resistant to bactericidal killing was similar or higher, depending on the strain (2, 8, 19, 20, 22, 34). In the same way, the survival of intracellular AIEC LF82 within HMDM was similar to or even higher than those previously reported for human-pathogenic Salmonella serovars or Shigella (2, 15, 22, 27, 35, 36, 38). However, a great difference was observed between Salmonella- or Shigella- and AIEC LF82-infected macrophages. No detached cells corresponding to dead cells were observed with AIEC LF82-infected cells, even after 24 or 48 h postinfection.

Strain LF82 does not induce any cell death of the infected macrophages. To our knowledge, this is the first report describing the ability of E. coli strains to replicate extensively within macrophages without inducing cell membrane damage or cell death within a 24-h period. Indeed, the ability to induce macrophage cell death is a common feature of most of the diarrheagenic E. coli strains. Members of all the diarrheagenic pathovars of human E. coli studied so far (enteroaggregative E. coli, enterohemorrhagic E. coli, EIEC, EPEC, ETEC, and diffusely adhering E. coli) were cytotoxic to J774-A1 after 24 h of infection at an MOI of 10 or 100 (24, 25). Except for the ETEC strain H10407, for which the cell death induced has not been elucidated, all the different E. coli pathovars induce cell death via an apoptotic mechanism (25). Infection of HMDM and J774-A1 murine macrophages with several enteroaggregative and cytodetaching E. coli strains demonstrated that these strains were not able to replicate intracellularly and induced macrophage cell death accompanied by LDH and IL-1 release into the culture supernatant (14). Moreover, assays performed with a few invasive E. coli strains that we isolated from stools of healthy controls indicated that these strains were not resistant to macrophage killing within the J774-A1 cells (data not shown). Thus, invasive E. coli strains isolated from CD lesions, which possess none of the known genetic invasive determinants of E. coli, Shigella, or Salmonella, share a virulence factor(s) which confers on the bacteria the ability to evade the bactericidal pathways to which intracellular bacteria are normally exposed within macrophages.

The mechanism by which AIEC strains resist bacterial killing by macrophages is still under investigation. Pathogens that survive within host phagocytes have various mechanisms of survival: (i) by avoiding phagocytosis (13, 17), (ii) by inhibiting fusion of bacteria-containing phagosomes with lysosomes and endosomes (3), (iii) by remodeling their phagosome (10), (iv) by moving out of the phagosome (21, 23), or (v) by resisting the antimicrobial environment of the mature phagolysosome (4).

LF82 bacterial replication does not require bacterial escape into the cytoplasmic compartment, in contrast to its behavior within intestinal epithelial cells (6). Within J774-A1 macrophages, the bacteria induced the formation of a single spacious vacuole by fusion of initial phagosomes. Such large vacuoles containing numerous bacteria have also been observed with J774-A1 cells infected with Salmonella or Yersinia enterocolitica clinical isolates (2, 3, 18). The formation of this specific compartment induced by AIEC LF82 within the infected macrophages is probably the key to its ability to resist macrophage killing. Spacious-phagosome formation may promote LF82 survival by dilution of toxic lysosomal compounds or attenuation of antimicrobial factors, including decreased phagosomal acidification.

It is noteworthy that AIEC strains are able to invade intestinal epithelial cells by a macropinocytosis-like process requiring actin polymerization and recruitment of microtubules and by inducing host cell membrane elongations at the sites of contact with epithelial cells. Type 1 pili were shown to play an essential role in promoting AIEC LF82 internalization within intestinal epithelial cells (5). We thought that they were also involved in stimulating the entry of bacteria into macrophages, and possibly subsequent resistance. Preliminary experiments already performed showed that all type 1 pilus-negative mutants, and the AIEC LF82 wild-type strain in the presence of 2% D-mannose, were able to survive and replicate intracellularly as efficiently as the wild-type LF82. The mechanism of entry, which may involve a specific phagosome formation with an unusual endocytic trafficking as described for Salmonella-infected macrophages (3, 30, 34), is currently under investigation.

We addressed the question of whether LF82-infected macrophages released IL-1 or TNF-, both capable of mediating an inflammatory process. No IL-1 was released even after 24 h post-AIEC LF82 infection. This result was not surprising, since it has been shown that this proinflammatory cytokine was only released in the culture medium during apoptosis of S. flexneri-infected macrophages (41). Compared to LPS-stimulated macrophages, a high level of TNF- was secreted after 24 h of infection with AIEC LF82. As TNF- is transcribed and translated de novo after macrophage stimulation, the synthesis of this cytokine demonstrated that macrophages were still active even with numerous intracellular bacteria. Moreover, proinflammatory TNF- secretion reflects an activation of infected macrophages. We can speculate that such AIEC-infected macrophages are continuously activated by the sustained presence of intracellular bacteria resistant for a long period to the bacterial killing.

In conclusion, AIEC strains, which are able to invade intestinal epithelial cells, are also able to trigger uptake into and survival within macrophages without inducing host cell death. These properties could allow the bacteria to translocate across the human intestinal barrier, to move to deep tissues, to continuously activate macrophages, and to potentially induce the formation of granulomas, one of the hallmarks of CD lesions.