Pathogenicity of many Gram-negative bacteria relies on a type III secretion (T3S) apparatus, which is used for delivery of bacterial effectors into the host cell cytoplasm allowing the bacteria to manipulate host cell cytoskeleton network as well as to interfere with intracellular signaling pathways. In this study, we investigated the potential of the Shigella flexneri T3SA as an in vivo delivery system for biologically active molecules such as cytokines. The anti-inflammatory cytokines IL-10 and IL-1 receptor antagonist (IL-1ra) were genetically fused to the first 30 or 60 residues of the Shigella T3S effector IpaH9.8 or to the first 50 residues of the Yersinia enterocolitica effector YopE and the recombinant fusion proteins were expressed in S. flexneri. YopE50-IL-10, IpaH60-IL-10, and IpaH60-IL-1ra were efficiently secreted via the T3S apparatus of Shigella. Moreover, these recombinant proteins did not impair the invasive ability of the bacteria in vitro. In a murine model, Shigella strains expressing YopE50-IL-10, IpaH60-IL-10, and IpaH60-IL-1ra induced a lower mortality in mice that was associated with reduced inflammation and a restricted localization of bacteria within the lung tissues as compared with wild-type Shigella. Moreover, the level of TNF-α and IL-1β mRNA were reduced in the lungs following infection by IL-10- and IL-1ra-secreting Shigella, respectively. These findings demonstrate that the Shigella T3S apparatus can deliver biologically active cytokines in vivo, thus opening new avenues for the use of attenuated bacteria to deliver proteins for immunomodulation or gene therapy purposes.

Shigellosis is characterized by an acute inflammation of the colonic and rectal mucosa caused by pathogenic Gram-negative bacteria of the genus Shigella. Bacterial invasion of epithelial cells leads to destruction of the intestinal epithelium and accumulation of monocytes and polymorphonuclear leukocytes (PMN)3 at the site of infection (1). The marked inflammation induced at the early stages of infection has paradoxical effects because it leads to further bacterial invasion through rupture of the epithelial cell barrier until complete control of infection occurs (2, 3). Infection with wild-type S. flexneri results in invasion plasmid Ag (Ipa)B-dependent activation of caspase-1 in macrophages and to massive release of mature IL-1β, which plays an important role in the initiation of inflammation (4). TNF-α plays, also, a central role in the development of destructive inflammatory lesions during shigellosis (5). Pathological lesions occurring in the course of this acute infection exhibit similarities to those observed in inflammatory bowel disease, particularly ulcerative colitis and Crohn’s disease, in which proinflammatory cytokines, including IL-1β and TNF-α, have been detected in affected tissues and are thought to be involved in the inflammatory cascade (6).

IL-10 and IL-1 receptor antagonist (IL-1ra) play important roles in the regulation of inflammation. IL-10 knockout mice develop severe colitis (7), while systemic administration of rIL-10 can prevent development of colitis in experimental models (8). In human volunteers, administration of rIL-10 results in inhibition of inflammatory cytokines such as IL-1 and TNF-α (9). In contrast, IL-1ra is a naturally occurring inhibitor of IL-1 (10). IL-1ra is produced by a variety of cells including macrophages, fibroblasts, and keratinocytes (11). IL-1ra binds to IL-1 receptor I on the cell surface but fails to trigger signal transduction, thereby acting as a competitive inhibitor of IL-1 binding to target cells (12). In inflammatory bowel disease, severe cases are associated with IL-1ra decrease leading to an imbalance of IL-1ra and IL-1 (13). The decrease of the IL-1ra:IL-1 ratio was shown to account for acute cases of Crohn’s disease and ulcerative colitis (14, 15). In the rabbit ligated-loop model of Shigella infection, administration of IL-1ra reduces intestinal inflammation (3), supporting the concept that the imbalance between IL-1 and IL-1ra accounts for disease severity (16). Based on these observations, IL-10 and IL-1ra are regarded as potential therapeutic agents against inflammatory diseases (17, 18, 19). However, daily administration of large amounts of these anti-inflammatory proteins seems to be necessary to obtain minimal therapeutic efficacy (18, 20). Gene therapy represents a promising approach to overcome such a drawback (21, 22, 23). However, daily injection of vectors that constitutively secrete anti-inflammatory cytokines in a wide range of colonized tissues might lead to unwanted side-effects, due to pleiotropic functions of these immunoregulatory proteins. The therapeutic potential may also be impaired when using vectors that do not efficiently colonize colonic tissues (23). Therefore, there is a need for an optimal delivery system that targets the colon. This can be achieved by direct injection of drug-secreting vector to the colon via rectal administration, as reported for adenoviral vectors expressing IL-10 (22). Alternatively, one might envision the use of orally administered vectors to deliver high amounts of drug, specifically in the colon.

Modified Shigella strains are an appealing tool for such a purpose, thanks to their tropism for colonic tissue and its type III secretion (T3S) apparatus that delivers effector proteins to the cytoplasmic membrane or the host cell cytoplasm (24, 25). T3S apparatus has been successfully used to deliver immunogenic peptides for vaccine development (26, 27, 28). The aim of this study was to assess the ability of Shigella T3S apparatus to deliver biologically active IL-10 and IL-1ra in vivo. We constructed chimeric rIL-10 and IL-1ra strains and tested their ability to secrete these cytokines and to promote control of mucosal inflammation in murine pulmonary infection model.

The invasive S. flexneri strain M90T and its derivative mxiD mutant (SF401) are described in Ref. 29 .

IL-1ra was PCR-amplified from cDNA derived from LPS-stimulated murine macrophage cell lines (RAW) using primers: 5′-CTATTGGTCTTCCTGGAAG-3′ (forward) and 5′-CGCCCTTCTGGGAAAAGA-3′ (reverse) flanked, respectively, by EcoRI and XbaI sites. This IL-1ra fragment lacking the peptide signal (11) was cloned in frame with 30 or 60 amino-terminal of IpaH9.8 (30) cloned in pUC plasmid (a kind gift of Dr. C. Parsot, Pasteur Institute, Paris, France) to create pUC-IpaH30-IL-1ra and pUC-IpaH60-IL-1ra. Site-directed mutagenesis of YopE (31) cloned in pUC plasmid was performed to create NdeI site at position 150 bp. PCR-amplified IL-1ra was cloned between NdeI and HindIII sites to replace NdeI-HindIII YopE fragment and to create pUC-YopE50-IL-1ra. PCR-amplified IL-10 fragment lacking the peptide signal (32) was cloned in frame with 30 or 60 amino-terminal of IpaH9.8 cloned in pUC plasmid to create pUC-IpaH30-IL-10 and pUC-IpaH60-IL-10. PCR-amplified IL-10 was cloned between NdeI and PstI sites in the replacement of NdeI-PstI YopE fragment to create pUC-YopE50-IL-10. After verification of all plasmid constructions by nucleotide sequencing, they were transferred into S. flexneri strain M90T or its derivative mxiD mutant by electroporation.

Bacterial extract and culture supernatant were prepared as previously described (29). Congo red-induced type III secretion was performed in bacteria grown at exponential phase (33). The supernatant was filtered with a 0.45-μm-pore-size filter, and proteins were concentrated through 10 kDa exclusion size filter or precipitated with 10% (vol/vol) trichloroacetic acid on ice. SDS-PAGE and immunoblotting were performed using anti-IpaB and anti-IpaC mAbs, described previously (34, 35), and anti-IL-10 and anti-IL-1ra sera from R&D Systems.

Whole lung tissues were harvested at 72-h postinfection and homogenized in PBS for 20 s at 6000 rpm. After centrifugation at 20,000 g, supernatants were subjected to IL-10-ELISA measurement using standard kit (Duoset; R&D Systems). Experimental conditions used to test TNF-α production by stimulated monocytes were similar to those described in Ref. 5 . Macrophages (J774) were coated in 24-well plates (5 × 105 cells) and allowed to adhere for at least 2 h at 37°C. To answer different multiplicity of infection (MOI), different dilutions of bacteria from a fresh culture adjusted to 108 CFU/ml in 2% FCS-RPMI 1640 were added to adherent macrophages in the presence or absence of 10 μg/ml neutralizing anti-IL-10 mAb (R&D systems) for 4 h at 37°C. After harvesting and centrifuging the culture media, the supernatants were assayed for TNF-α by ELISA (Duoset; R&D system).

As described in Ref. 36 , mice were inoculated intranasally (i.n.) with 2 × 108 bacteria in 20 μl. Bacterial enumeration was performed in the lungs removed from sacrificed mice and ground in 10 ml sterile PBS. Dilutions were plated on Tryptic Soy Broth plates for CFU enumeration. Shigella infections were performed in groups of at least 15 animals, and each data point is the mean of at least 10 infected mice. Experiments in mice compiled with the guidelines of the Université Libre de Bruxelles ethic committee for the human use of laboratory animals.

Total RNA from homogenized lungs was extracted using Tripure solution (Roche Diagnostic Systems) according to the manufacturer’s instructions. Procedures of reverse transcription and amplification of cytokines by real time PCR were described in detail in Ref. 37 . Briefly, after treatment of samples with RNase-free DNase (Promega) to remove contaminating genomic DNA, cDNA was obtained using M-MLV reverse transcriptase (Invitrogen Life Technologies). Real time PCR was conducted on the Lightcycler using DNA Master Hybridization probes (Roche Diagnostic Systems). Cytokine mRNA levels were normalized against β-actin mRNA.

Lung tissue samples were fixed in 4% buffered formalin, dehydrated, embedded in paraffin, and sectioned in 5-μm slices. H&E staining was done as previously described (36). Inflammation index, representing the percentage of inflammatory area recorded in a series of sections covering the whole lung, was calculated as previously described (38). LPS immunostaining was realized with an anti-S. flexneri 5a LPS Ab (IgGC20) conjugated to alcaline phosphatase (39).

Survival analyses were performed using Gehan’s generalized Wilcoxon test. Data were compared using the Mann-Whitney U test.

To study the potential use of the S. flexneri T3S apparatus to deliver bioactive anti-inflammatory cytokines, the signal sequences of IL-10 (aa 1–19) and IL-1ra (aa 1–7) were replaced by sequences encoding the secretion signal of T3S effectors. The efficiency of a potential secretion signal that may be contained within amino-terminal of the Shigella effector IpaH9.8 (30) was investigated by using the first 30 and 60 residues of IpaH9.8 to construct IpaH30-IL-10, IpaH30-IL-1ra, IpaH60-IL-10, and IpaH60-IL-1ra. We also used the previously defined secretion domain contained within amino residues derived from Yersinia enterocolitica effector YopE (31) to construct YopE50-IL-10. Full IL-10 and IL-1ra genes lacking T3S domains were used as a control. The pUC plasmids harboring these constructs were introduced into S. flexneri wild-type strain M90T or its derivative SF401, in which the mxiD gene (which encodes an essential T3S apparatus component) was inactivated (29). Expression and secretion of the recombinant proteins were analyzed on whole cell extracts and protein supernatants upon Congo red induction by SDS-PAGE and immunoblotting using anti-IL-10 and anti-IL-1ra Abs (Fig. 1, A and B). IL-10 (20 kDa), YopE50-IL-10 (25 kDa), IpaH30-IL-10 (23 kDa), IpaH60-IL-10 (26 kDa) (Fig. 1,A), IL-1ra (20 kDa), IpaH30-IL-1ra (23 kDa), and IpaH60-IL-1ra (26 kDa) (Fig. 1,B) were expressed in whole cell extracts of corresponding strains. Only YopE50-IL-10, IpaH60-IL-10, and IpaH60-IL-1ra were secreted upon exposure of bacteria to Congo red. In all strains, the secretion of other T3S substrates, such as IpaB and IpaC, was unaffected (Fig. 1, A and B). YopE50-IL-10, IpaH60-IL-10, and IpaH60-IL-1ra were not secreted by derivatives of the mxiD mutant, confirming that their secretion was dependent on the T3S apparatus. These results indicate that the 60 N-terminal residues of IpaH9.8 are sufficient to direct protein secretion via the T3S apparatus, and that the efficiency of secretion is similar to that observed for 50 N-terminal of YopE.

FIGURE 1.

Production and secretion of rIL-10 and IL-1ra proteins. Immunoblotting analysis of whole cell extracts and proteins secreted in response to Congo red induction (supernatants) by derivatives of the wild-type strain (WT) or the mxiD mutant harboring the plasmids (indicated by numbers) using anti-IL-10, anti-IL-1ra, anti-IpaB, and anti-IpaC Abs.

FIGURE 1.

Production and secretion of rIL-10 and IL-1ra proteins. Immunoblotting analysis of whole cell extracts and proteins secreted in response to Congo red induction (supernatants) by derivatives of the wild-type strain (WT) or the mxiD mutant harboring the plasmids (indicated by numbers) using anti-IL-10, anti-IL-1ra, anti-IpaB, and anti-IpaC Abs.

Close modal

Interaction of bacteria with adherent monocytes induces the release of high amounts of the proinflammatory cytokine TNF-α (5). We used this read-out system to test whether IL-10 produced by bacteria inhibits TNF-α production by monocytes in vitro. M12 (YopE50-IL-10), M16 (IpaH60-IL-10), and M90T (wild type) were incubated at various MOI with J774 macrophage cell line for 4 h, and the amount of TNF-α released in the medium was measured by ELISA (Fig. 2). At low MOI, a significant reduction in the amount of secreted TNF-α was observed with the recombinant strains as compared with the wild-type strain M90T. To determine whether reduction in TNF-α production was IL-10-dependent, we performed the test in the presence of neutralizing anti-IL-10 mAb. In these conditions, similar amounts of TNF-α were detected in the supernatant following incubation of macrophages with M12 (YopE50-IL-10), M16 (IpaH60-IL-10), or M90T (data not shown). These results indicate that delivered IL-10 is biologically active, and that recombinant bacteria might potentially be used to down-regulate production of proinflammatory cytokines.

FIGURE 2.

Production of TNF-α by macrophages infected by wild-type (M90T)- or IL-10-secreting Shigella, M12 (YopE50-IL-10), and M16 (IpaH60-IL-10) at increasing MOI. Significant differences (*, p < 0.05) in average levels of TNF-α production between wild-type and recombinant strains were recorded at lower MOI.

FIGURE 2.

Production of TNF-α by macrophages infected by wild-type (M90T)- or IL-10-secreting Shigella, M12 (YopE50-IL-10), and M16 (IpaH60-IL-10) at increasing MOI. Significant differences (*, p < 0.05) in average levels of TNF-α production between wild-type and recombinant strains were recorded at lower MOI.

Close modal

Before testing in vivo the effect of IL-10 and IL-1ra secretion on Shigella pathogenicity, we examined the ability of recombinant strains to enter into and spread between Caco-2 cells in vitro using the plaque assay (40). The strains M12 (YopE50-IL-10), M15 (IpaH60-IL-1ra), and M16 (IpaH60-IL-10) induced the formation of the same number of plaques of similar size as compared with wild-type strain M90T (data not shown), indicating that their invasion and dissemination within epithelial cells were unaffected.

We next analyzed the kinetics of survival and the bacterial load of mice infected i.n. with different bacterial strains. As shown in Fig. 3,A, the survival of mice infected by M12 (YopE50-IL-10), M15 (IpaH60-IL-1ra), and M16 (IpaH60-IL-10) was significantly higher than M90T-infected mice starting from 72-h postinfection. The survival rate of mice infected by M10 (IL-10) and M11 (IL-1ra) was similar to M90T (wild-type)-infected mice (data not shown). When measuring the lung-bacterial load at 72-h postinfection in the different groups of mice (Fig. 3 B), we observed a significantly higher amount of CFU/lungs in mice infected by M12 (YopE50-IL-10), M15 (IpaH60-IL-1ra), and M16 (IpaH60-IL-10) as compared with M90T (wild type), suggesting that the improved mice survival was not due to a lower level of bacterial infection but might be attributed to the degree of inflammation induced during infection.

FIGURE 3.

Improved survival is not associated with low bacterial load in mice infected with IL-10- and IL-1ra-secreting Shigella. A, Percentage of mice survival following i.n. infection with M90T (wild type), M12 (YopE50-IL-10), M15 (IpaH60-IL-1ra), or M16 (IpaH60-IL-10). The data are representative of three independent experiments. B, Number of CFUs in infected lungs from mice sacrificed at 72-h postinfection. Asterisk (*) indicates statistically significant differences (p < 0.05) between recombinant and wild-type strains.

FIGURE 3.

Improved survival is not associated with low bacterial load in mice infected with IL-10- and IL-1ra-secreting Shigella. A, Percentage of mice survival following i.n. infection with M90T (wild type), M12 (YopE50-IL-10), M15 (IpaH60-IL-1ra), or M16 (IpaH60-IL-10). The data are representative of three independent experiments. B, Number of CFUs in infected lungs from mice sacrificed at 72-h postinfection. Asterisk (*) indicates statistically significant differences (p < 0.05) between recombinant and wild-type strains.

Close modal

i.n. administration of wild-type strain M90T in mice induces a strong inflammatory response within the lung parenchyma, characterized by a massive infiltration of PMN within alveoli and bronchi, mimicking the inflammation induced in the intestinal tract with Shigella spp (36). Before testing whether the difference observed in survival was due to limited tissue damage as a result of the control of inflammatory cytokines by IL-10- and IL-1ra-secreting strains, we examined the amount of IL-10 and IL-1ra mRNA produced within infected lung tissues (Fig. 4). As expected, a high amount of IL-10 mRNA was detected in mice infected with M12 (YopE50-IL-10) and M16 (IpaH60-IL-10) as compared with mice infected with other strains (p < 0.05). Moreover, the protein level of IL-10 was quantified by ELISA in lung tissues derived from mice infected with M12 (YopE50-IL-10) (n = 16). Significant levels of IL-10 (1.292 ± 0.654 μg/ml) were detected in supernatants of homogenized lung tissues as compared with M90T (0.536 ± 0.219 μg/ml) (p < 0.05). In the same way, IL-1ra transcripts were expressed at high levels in the lung tissue of mice infected with M15 (IpaH60-IL-1ra) compared with other recombinant Shigella. These data indicate that Shigella is able to deliver recombinant cytokines to the site of infection. To assess the in vivo biological activity of delivered IL-10 and IL-1ra, we performed the quantification of major proinflammatory cytokine transcripts produced within infected lungs 72-h postinfection (Fig. 4). Interestingly, a sharp decrease of TNF-α transcripts was found in mice infected with strains secreting IL-10 (M12 and M16) compared with those infected with other strains (p < 0.05). In lungs infected by the strain secreting IL-1ra (M15), TNF-α transcripts were lower but did not differ significantly from those obtained with M90T (wild type) or M11 (IL-1ra). Notably, the amount of IL-1β mRNA present in lungs infected with M15 (IpaH60-IL-1ra) decreased significantly (p < 0.05) as compared with those infected with other bacteria. In contrast, no significant differences were seen in IL-1β transcript in mice infected with M12 (YopE50-IL-10) or M16 (IpaH60-IL-10). These results indicate that strains delivering IL-10 and IL-1ra down-regulate proinflammatory cytokines TNF-α and IL-1β, respectively. The evaluation of IL-6 transcripts in mice infected by M12 (YopE50-IL-10), M15 (IpaH60-IL-1ra), or M16 (IpaH60-IL-10) did not show a significant difference compared with mice infected by M10 (IL-10), M11 (IL-1ra), or M90T (wild type).

FIGURE 4.

Real time PCR quantification of cytokine transcripts in infected lungs. RNA was extracted at 72-h postinfection from lungs derived from mice inoculated with M90T (wild type), M10 (IL-10), M11 (IL-1ra), M12 (YopE50-IL10), M15 (IpaH60-IL-1ra), or M16 (IpaH60-IL-10) or not infected (C) and used as template for real time PCR experiments. Quantification of cytokines was recorded as the mRNA level relative to that of β-actin. Differences of levels of cytokine transcripts were illustrated by arbitrary units. Asterisk (*) indicates statistically significant differences (p < 0.05).

FIGURE 4.

Real time PCR quantification of cytokine transcripts in infected lungs. RNA was extracted at 72-h postinfection from lungs derived from mice inoculated with M90T (wild type), M10 (IL-10), M11 (IL-1ra), M12 (YopE50-IL10), M15 (IpaH60-IL-1ra), or M16 (IpaH60-IL-10) or not infected (C) and used as template for real time PCR experiments. Quantification of cytokines was recorded as the mRNA level relative to that of β-actin. Differences of levels of cytokine transcripts were illustrated by arbitrary units. Asterisk (*) indicates statistically significant differences (p < 0.05).

Close modal

To further evaluate the extent of inflammation, we examined the histopathological changes in infected lung tissues (Fig. 5,A). As expected, an acute bronchopneumonia was observed in lungs from animals infected with wild-type M90T. Lungs of mice infected by M10 (IL-10) or M11 (IL-1ra) strains exhibited an inflammation comparable to that induced by wild-type strain M90T (not shown). In contrast, a marked reduction of the degree of inflammation and tissue damages was observed in mice infected by M12 (YopE50-IL-10), M15 (IpaH60-IL-1ra), and M16 (IpaH60-IL-10) strains. The inflammation index representing the percentage of area with symptoms of inflammation was recorded in a series of sections covering the whole pulmonary tissues. As shown in Fig. 5 B, the inflammation index was lower in lungs of mice infected by M12 (YopE50-IL-10), M15 (IpaH60-IL-1ra), and M16 (IpaH60-IL-10) strains as compared with that measured in mice infected by M90T (wild type), M10 (IL-10), and M11 (IL-1ra). Overall, these data demonstrate that Shigella delivering IL-10 and IL-1ra by the T3S apparatus is able to control the inflammation during experimental infection through the control of local production of inflammatory cytokines.

FIGURE 5.

Strains secreting IL-10 and IL-1ra induce a lower inflammatory reaction and exhibit reduced LPS staining. A, H&E staining of tissue sections of lungs from mice infected with M90T (a and b), M12 (YopE50-IL-10) (e and f), M16 (IpaH60-IL-10) (i and j), and M15 (IpaH60-IL-1ra) (m and n). B, Inflammatory indexes were evaluated by analysis of serial sections of lungs from mice infected with M90T, M12, M15, M16, M10 (IL-10), or M11 (IL-1ra) at 72-h postinfection. Significant differences (p < 0.05) are indicated by an asterisk (*). C, LPS staining (arrow head) of tissue sections of lungs from mice infected with M90T (c and d), M12 (YopE50-IL10) (g and h), M16 (IpaH60-IL-10) (k and l), and M15 (IpaH60-IL-1ra) (o and p).

FIGURE 5.

Strains secreting IL-10 and IL-1ra induce a lower inflammatory reaction and exhibit reduced LPS staining. A, H&E staining of tissue sections of lungs from mice infected with M90T (a and b), M12 (YopE50-IL-10) (e and f), M16 (IpaH60-IL-10) (i and j), and M15 (IpaH60-IL-1ra) (m and n). B, Inflammatory indexes were evaluated by analysis of serial sections of lungs from mice infected with M90T, M12, M15, M16, M10 (IL-10), or M11 (IL-1ra) at 72-h postinfection. Significant differences (p < 0.05) are indicated by an asterisk (*). C, LPS staining (arrow head) of tissue sections of lungs from mice infected with M90T (c and d), M12 (YopE50-IL10) (g and h), M16 (IpaH60-IL-10) (k and l), and M15 (IpaH60-IL-1ra) (o and p).

Close modal

The localization of bacteria was examined in infected lung tissues by specific anti-LPS Ab. As shown in Fig. 5 C, the immunostaining appeared diffused in the bronchial epithelium, the luminal purulent exudate, and alveolar infiltrate and was associated with tissue destruction in lung tissues of mice infected by wild-type M90T. Similar results were observed with M10 (IL-10) and M11 (IL-1ra) (data not shown). Notably, in the lungs of mice infected by M12 (YopE50-IL-10), M15 (IpaH60-IL-1ra), and M16 (IpaH60-IL-10) strains, LPS staining was mainly restricted to the surface of the bronchial epithelium, indicating that bacteria did not invade massively the pulmonary tissues.

In the present study, we report the in vivo delivery of bioactive cytokines by the S. flexneri T3S apparatus. This secretion machinery has been so far investigated in Salmonella for delivery of antigenic peptides to elicit specific immune response for vaccine purposes (26, 27, 28). This is the first report describing the potential use of the T3S apparatus for in vivo delivery of host anti-inflammatory mediators that could find a broad range of application purposes.

The replacement of the N-terminal domain of IL-10 or IL-1ra with 60 N-terminal residues of IpaH9.8 or 50 N-terminal residues of YopE results in an efficient secretion of the hybrid proteins via the T3S apparatus of Shigella upon induction with Congo red. This indicates that N-terminal residues of IpaH9.8 can be used as a signal sequence to allow secretion of foreign proteins by Shigella. Furthermore, our data are consistent with the finding that N-terminal residues from YopE of Yersinia can be successfully recognized by a heterologous type III secretion system (25, 27).

i.n. inoculation of S. flexneri in mice results in an acute bronchopneumonia with alveolitis, characterized by a massive influx of PMN. The invasive and inflammatory properties of S. flexneri in this murine lung model mimic those observed in the infected human intestinal mucosa (36). The in vivo bioactivity of delivered IL-10 and IL-1ra was assessed by investigating the virulence profile of recombinant strains and the inflammation-related parameters in the murine pulmonary model of infection. Shigella strains secreting IL-10 or IL-1ra exhibit a reduced inflammation as compared with the wild-type strain. In the case of IL-10 expressing Shigella, the control of inflammation was correlated with the down-regulation of TNF-α detected in pulmonary tissues. This is consistent with our previous studies with the rabbit ligated-loop model of infection, in which we observed that neutralization of TNF-α by specific Ab decreased tissue injury (5). Indeed, elevated levels of TNF-α were associated with increased severity of inflammation in shigellosis (5, 41, 42, 43). Notably, there was no significant difference in IL-1β mRNA synthesis after infection with wild-type or IL-10-expressing strains. It is possible that the control of IL-1β by delivered IL-10 may occur earlier during infection. Indeed, IL-1β is released massively during initial phases of shigellosis, following apoptosis of resident tissue macrophages, and plays a key role in the early triggering of inflammation (3, 4). In contrast to Shigella strains secreting IL-10, we found that IL-1β was down-regulated with Shigella strains secreting IL-1ra. This is likely due to the ability of IL-1ra produced in situ to compete with IL-1 for binding to its receptor, thereby suppressing the auto and paracrine effect of IL-1β (10, 11). This is in agreement with the finding that blocking IL-1 with IL-1ra decreases inflammation in the rabbit ileal loop model of shigellosis (3, 16). Therefore, the anti-inflammatory properties of delivered IL-10 and IL-1ra are mediated by the control of the major proinflammatory cytokines TNF-α and IL-1β. Given the in vivo bioactivity of delivered IL-10 and IL-1ra, it is likely that these cytokines were released from bacteria and bind to their respective receptors on the surface of the target cells (10, 32).

Interestingly, a significant reduction in the mortality was observed in mice infected with IL-10- and IL-1ra-secreting Shigella. Survival was correlated with reduced inflammation and damage to the pulmonary tissues but not with the level of infection. However, specific LPS staining indicated that the distribution of bacteria in mice infected with Shigella secreting IL-10 or IL-1ra largely restricted to the surface of bronchial epithelium and did not invade, massively, the lung parenchyma. This cannot be ascribed to a defect in the ability of the recombinant bacteria to enter epithelial cells or to spread between cells, because recombinant strains were shown to be able to invade and spread between cells as efficiently as the wild type in vitro. The data obtained with IL-1ra-secreting Shigella are in line with a previous work (38) showing that in contrast to wild-type mice, the i.n. inoculation of IL-1β-deficient mice with M90T strain resulted in the localization of bacteria at restricted infectious foci within lung tissues. This limited distribution of bacteria would contribute to an efficient containment of the infection leading mice to progress toward resolution of the infection and survival.

It is well established that the acute inflammation triggered during Shigella infection is necessary to the development of harmful dysenteric symptoms, characterized by a massive destruction of infected tissues due to the recruitment of inflammatory cells such as PMN. Paradoxically, inflammation is also necessary for the effective eradication of the pathogen. Therefore, the control of inflammation is a critical point, in the sense that it should avoid the over-response of the host while preserving an effective immune response against the pathogen. Our data indicate that a balance between inductive and inhibitory signals of inflammation brought about by using Shigella delivering IL-1ra and IL-10 by the T3S apparatus seems to be maintained in infected tissues. Future investigations will concern the monitoring of the immune cells involved in this equilibrium during experimental infection.

In conclusion, the T3S system can be considered as a powerful tool for localized delivery of bioactive molecules. Unlike the constitutive secretion that might be undesirable for a number of pleiotropic cytokines such as IL-10, especially when vectors naturally colonize widespread organ tissues, the T3S apparatus-mediated active secretion upon contact with host cells is highly efficient for optimizing drug delivery. It is, however, worth noting that a shut-off mechanism should be considered to prevent potential overdosing with cytokines. This can be explored by using attenuated strains that not only are defective in virulence genes but also in genes encoding critical enzymes involved in biosynthetic pathways of aromatic amino acids or nucleotides (44, 45). Such auxotrophic mutants will be useful for the biological containment of cytokine delivery, because the replication and survival of bacteria would be dependent on metabolites present in limited amounts within host cells. The current work can obviously be considered proof of concept. More will be needed to validate the possible clinical value of such strains. In intestinal inflammatory diseases, gene therapy is an interesting tool to overcome a repetitive administration of high amounts of anti-inflammatory cytokines (21, 22, 23). The use of this bacterial delivery system in human therapeutical strategy awaits further studies on the biological role of type III secreted anti-inflammatory cytokines within intestinal tissue in murine experimental model. The refractory state of conventional mice model to Shigella due to the loss of IL-8 gene (46) can be bypassed in the future by using transgenic mice expressing human IL-8 gene. This study opens new approaches for the potential use of the T3S apparatus of highly attenuated Shigella targeting the colon for immunomodulation or gene therapy purposes related to intestinal disorders.

We are grateful to Dr. C. Parsot for providing pUC-IpaH9.8 plasmid and for the critical reading of the manuscript. We thank Dr. C. Tang for careful reading of this manuscript and Dr. L. Bernardini for helpful discussion. We also thank Drs. R. Tournebize, M. Tanguy, and A. Thuizat for help and V. Vercrysse for technical assistance.

The authors have no financial conflict of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was supported by grant from the European Union (QLK2-C1999-00938), the Actions de Recherche concertées (convention 98/03-224) de la Communauté Française de Belgique, and by Fonds national de la Recherche Scientifique Médicale, convention 3.4623.06. P.S. is a Howard Hughes Medical Institute Scholar.

3

Abbreviations used in this paper: PMN, polymorphonuclear leukocyte(s); T3S, type III secretion; IL-1ra, IL-1 receptor antagonist; MOI, multiplicity of infection; i.n., intranasal(ly); Ipa, invasion plasmid Ag.

1
Sansonetti, P. J..
2001
. Rupture, invasion, and inflammatory destruction of the intestinal barrier by Shigella, making sense of prokaryote-eukaryote cross-talks.
FEMS Microbiol. Rev.
25
:
3
-14.
2
Perdomo, J. J., J. M. Cavaillon, M. Huerre, H. Ohayon, P. Gounon, P. J. Sansonetti.
1994
. Acute inflammation causes epithelial invasion and mucosal destruction in experimental shigellosis.
J. Exp. Med.
180
:
1307
-1319.
3
Sansonetti, P. J., J. Arondel, J. M. Cavaillon, M. Huerre.
1995
. Role of IL-1 in the pathogenesis of experimental shigellosis.
J. Clin. Invest.
96
:
884
-892.
4
Zychlinsky, A., P. J. Sansonetti.
1997
. Apoptosis as a proinflammatory event: what can we learn from bacteria-induced cell death?.
Trends Microbiol.
5
:
201
-204.
5
D’hauteville, H., S. Khan, D. J. Maskell, A. Kussak, A. Weintraub, J. Mathison, R. J. Ulevitch, N. Wuscher, C. Parsot, P. J. Sansonetti.
2002
. Two msbB genes encoding maximal acylation of lipid A are required for invasive Shigella flexneri to mediate inflammatory rupture and destruction of the intestinal epithelium.
J. Immunol.
168
:
5240
-5251.
6
Shanahan, F..
2002
. Crohn’s disease.
Lancet
359
:
62
-69.
7
Kuhn, R., J. Lohler, D. Rennick, K. Rajewsky, W. Muller.
1993
. Interleukin-10-deficient mice develop chronic enterocolitis.
Cell
75
:
263
-274.
8
Herfarth, H. H., S. P. Mohanty, H. C. Rath, S. Tonkonogy, R. B. Sartor.
1996
. Interleukin-10 suppresses experimental chronic, granulomatous inflammation induced by bacterial cell wall polymers.
Gut
39
:
836
-845.
9
Chernoff, A. E., E. V. Granowitz, L. Shapiro, E. Vannier, G. Lonnemann, J. B. Angel, J. S. Kennedy, A. R. Rabson, S. M. Wolff, C. A. Dinarello.
1995
. A randomized, controlled trial of IL-10 in humans: inhibition of inflammatory cytokine production and immune responses.
J. Immunol.
54
:
5492
-5499.
10
Arend, W. P..
1993
. IL-1 receptor antagonist.
Adv. Immunol.
54
:
167
-227.
11
Arend, W. P., F. G. Joslin, R. J. Massoni.
1985
. Effects of immune complexes on production by human monocytes of IL-1 or an IL-1 inhibitor.
J. Immunol.
134
:
3668
-3875.
12
Dripps, D. J., B. J. Brandhuber, R. C. Thompson, S. P. Eisenberg.
1991
. Interleukin-1 receptor antagonist binds to the 80-kDa IL-1 receptor but does not initiate IL-1 signal transduction.
J. Biol. Chem.
266
:
10331
-10336.
13
Casini-Raggi, V., L. Kram, Y. J. T. Chong, C. Fiocchi, T. T. Pizarro, F. Comineli.
1995
. Mucosal imbalance of IL-1 and IL-1 receptor antagonist in inflammatory bowel disease.
J. Immunol.
154
:
2434
-2440.
14
Andus, T., R. Daig, D. Vogel.
1997
. Imbalance of the interleukin-1 system in colonic mucosa: association with intestinal inflammation and interleukin-1 receptor antagonist genotype 2.
Gut
41
:
651
-657.
15
Heresbach, D., M. Alizadeh, A. Dabadie.
1997
. Significance of interleukin-1 β and interleukin-1 receptor antagonist genetic polymorphism in inflammatory bowel diseases.
Am. J. Gastroenterol.
92
:
1164
-1169.
16
Arondel, J., M. Singer, A. Matsukawa, A. Zychlinsky, P. J. Sansonetti.
1999
. Increased IL-1 and imbalance between IL-1 and IL-1 receptor antagonist during acute inflammation in experimental shigellosis.
Infect. Immun.
67
:
6056
-6066.
17
Pretolani, M., M. Goldman.
1997
. IL-10: a potential therapy for allergic inflammation?.
Immunol. Today
18
:
277
-280.
18
Schreiber, S., R. N. Fedorak, O. H. Nielsen, J. Wild, C. N. Williams, S. Nikolaus, M. Jacyna, B. A. Lashner, A. Gangl, P. Rutqeerts, et al
2000
. Safety and efficacy of recombinant human interleukin 10 in chronic active Crohn’s disease: Crohn’s Disease IL-10 Cooperative Study Group.
Gastroenterology
119
:
1461
-1472.
19
Gabay, C., W. P. Arend.
1998
. Treatment of rheumatoid arthritis with IL-1 inhibitors.
Springer Semin. Immunopathol.
20
:
229
-246.
20
Fedorak, R. N., A. Gangl, C. O. Elson, P. Rutqeerts, S. Schreiber, J. Wild, S. B. Hanauer, A. Kilian, M. Cohard, A. Lebeaut, B. Feagan.
2000
. Recombinant human interleukin 10 in the treatment of patients with mild to moderately active Crohn’s disease: the Interleukin 10 Inflammatory Bowel Disease Cooperative Study Group.
Gastroenterology
119
:
1473
-1482.
21
Steidler, L., W. Hans, L. Schotte, S. Neirynck, F. Obermeier, W. Falk, W. Fiers, E. Remaut.
2000
. Treatment of murine colitis by Lactococcus lactis secreting interleukin-10.
Science
289
:
1352
-1355.
22
Lindsay, J. O., C. J. Ciesielski, T. Scheinin, F. M. Brennan, H. J. Hodgson.
2003
. Local delivery of adenoviral vectors encoding murine interleukin-10 induces colonic IL-10 production and is therapeutic for murine colitis.
Gut
52
:
981
-987.
23
Ricci, A., G. Macchia, P. Ruggiero, T. Magi, P. Bossù, L. Xu, D. Medaglini, A. Tagliabue, L. Hammartsröm, G. Pozzi.
2003
. In vivo mucosal delivery of bioactive human interleukin-1 receptor antagonist produced by Streptococcus gordonii.
BMC Biotechnol.
3
:
1472
-1483.
24
Cossart, P., P. J. Sansonetti.
2004
. Bacterial invasion: the paradigms of enteroinvasive pathogens.
Science
304
:
242
-248.
25
Cornelis, G. R..
2006
. The type III secretion injectisome.
Nat. Rev. Microbiol.
4
:
811
-825.
26
Rüssmann, H., H. Shams, F. Poblete, Y. Fu, J. E. Galan, R. O. Donis.
1998
. Delivery of epitopes by the Salmonella type III secretion system for vaccine development.
Science
281
:
565
-568.
27
Rüssmann, H., E. I. Igwe, J. Sauer, W. D. Hardt, A. Bubert, G. Geginat.
2001
. Protection against murine listeriosis by oral vaccination with recombinant Samonella expressing hybrid Yersinia type III proteins.
J. Immunol.
167
:
357
-365.
28
Nishikawa, H., E. Sato, G. Briones, L. M. Chen, M. Matsuo, Y. Nagata, G. Ritter, E. Jäger, H. Nomura, S. Kondo, et al
2006
. In vivo antigen delivery by a Salmonella typhimurium type III secretion system for therapeutic cancer vaccines.
J. Clin. Invest.
116
:
1946
-1954.
29
Allaoui, A., P. J. Sansonetti, C. Parsot.
1993
. MxiD, an outer membrane protein necessary for the secretion of the Shigella flexneri lpa invasins.
Mol. Microbiol.
1
:
59
-68.
30
Demers, B., P. J. Sansonetti, C. Parsot.
1998
. Induction of type III secretion in Shigella flexneri is associated with differential control of transcription of genes encoding secreted proteins.
EMBO J.
17
:
2894
-2903.
31
Sory, M. P., A. Boland, I. Lambermont, G. Cornelis.
1995
. Identification of the YopE and YopH domains required for secretion and internalization into the cytosol of macrophages, using the cyaA gene fusion approach.
Proc. Natl. Acad. Sci. USA
92
:
1998
-2002.
32
Moore, K. W., A. O’Garra, R. de Waal Malefyt, P. Vieira, R. T. Mosmann.
1993
. Interleukin-10.
Annu. Rev. Immunol.
11
:
165
-190.
33
Bahrani, F. K., P. J. Sansonetti, C. Parsot.
1997
. Secretion of Ipa proteins by Shigella flexneri: inducer molecules and kinetics of activation.
Infect. Immun.
65
:
4005
-4010.
34
Barzu, S., F. Nato, S. Rouyre, J. C. Mazié, P. J. Sansonetti, A. Phalipon.
1993
. Characterization of B-cell epitopes on IpaB, an invasion-associated antigen of Shigella flexneri: identification of an immunodominant domain recognized during natural infection.
Infect. Immun.
61
:
3825
-3831.
35
Phalipon, A., J. Arondel, F. Nato, S. Rouyre, J. C. Mazié, P. J. Sansonetti.
1992
. Identification and characterization of B-cell epitopes of IpaC, an invasion-associated protein of Shigella flexneri.
Infect. Immun.
60
:
1919
-1926.
36
Phalipon, A., M. Kaufmann, P. Michetti, J. M. Cavaillon, M. Huerre, P. Sansonetti, P. J. Krehenbuhl.
1995
. Monoclonal IgA antibody directed against serotype-specific epitope of Shigella flexneri lipopolysaccharide protects against murine experimental shigellosis.
J. Exp. Med.
182
:
769
-778.
37
Stordeur, P., L. F. Poulin, L. Craciun, L. Zhou, L. Schadené, A. de Lavareille, S. Goriely, M. Goldman.
2002
. Cytokine mRNA quantification by real-time PCR.
J. Immunol. Methods
259
:
55
-64.
38
Sansonetti, P. J., A. Phalipon, J. Arondel, K. Thirumlai, S. Banerjee, S. Akira, K. Takeda, A. Zychlinsky.
2000
. Caspase-1 activation of IL-1β and IL-18 are essential for Shigella flexneri-induced inflammation.
Immunity
12
:
581
-590.
39
Phalipon, A., C. Costachel, C. Grandjean, A. Thuizat, C. Guerreiro, M. Tanguy, F. Nato, B. Vulliez-Le Normand, F. Bélot, K. Wright, et al
2006
. Characterization of functional oligosaccharide mimics of the Shigella flexneri serotype 2a O-antigen: implications for the development of a chemically defined glycoconjugate vaccine.
J. Immunol.
176
:
1686
-1694.
40
Oaks, E. V., M. E. Wingfield, S. B. Formal.
1985
. Plaque formation by virulent Shigella flexneri.
Infect. Immun.
48
:
124
-128.
41
Van de Verg, L. L., C. P. Mallett, H. H. Collins, T. Larsen, C. Hammack, T. L. Hale.
1995
. Antibody and cytokine responses in a mouse pulmonary model of Shigella flexneri serotype 2a infection.
Infect. Immun.
63
:
1947
-1954.
42
Raqib, P., A. A. Linderbeg, B. Wretlind, P. K. Bardahan, U. Anderson, J. Andersson.
1995
. Persistence of local cytokine production in shigellosis in acute and convalescent stages.
Infect. Immun.
63
:
289
-296.
43
Cersini, A., M. C. Martino, I. Martini, G. Rossi, M. L. Bernardini.
2003
. Analysis of virulence and inflammatory potential of Shigella flexneri purine biosynthesis mutants.
Infect. Immun.
71
:
7002
-7013.
44
Cersini, A., A. M. Salvia, M. L. Bernardini.
1989
. Intracellular multiplication and virulence of Shigella flexneri auxotrophic mutants.
Infect. Immun.
66
:
549
-557.
45
Coster, T. S., C. W. Hoge, L. L. VanDeVerg, A. B. Hartman, E. V. Oaks, M. M. Venkatesan, D. Cohen, G. Robin, A. Fontaine-Thompson, P. J. Sansonetti, T. L. Hale.
1999
. Vaccination against Shigellosis with Attenuated Shigella flexneri 2a Strain SC602.
Infect. Immun.
67
:
3437
-3443.
46
Singer, M., P. J. Sansonetti.
2004
. IL-8 is a key chemokine regulating neutrophil recruitment in a new mouse model of Shigella-induced colitis.
J. Immunol.
173
:
4197
-4206.