Evidence for the existence of a new genus Chlamydiifrater gen. nov. inside the family Chlamydiaceae with two new species isolated from flamingo (Phoenicopterus roseus): Chlamydiifrater phoenicopteri sp. nov. and Chlamydiifrater volucris sp. nov

The family Chlamydiaceae currently comprises a single genus Chlamydia, with 11 validly published species and seven more taxa. It includes the human pathogens Chlamydia (C.) trachomatis, C. pneumoniae and C. psittaci, a zoonotic agent causing avian chlamydiosis and human psittacosis, as well as other proven or potential pathogens in ruminants, birds, snakes, reptiles and turtles. During routine testing of 15 apparently healthy captive flamingos in a zoo in 2011, an atypical strain of Chlamydiaceae was detected by real-time PCR of cloacal swab samples. Sequence analysis of the 16S rRNA gene revealed high similarity to the uncultured Chlamydiales bacterium clone 122, which previously had been found in gulls. As more samples were collected during annual campaigns of the flamingo ringing program in southern France from 2012 to 2015, Chlamydiaceae-specific DNA was detected by PCR in 30.9% of wild birds. From these samples, three strains were successfully grown in cell culture. Ultrastructural analysis, comparison of 16S and 23S rRNA gene sequences, whole-genome analysis based on de novo hybrid-assembled sequences of the new strains as well as subsequent calculation of taxonomic parameters revealed that the relatedness of the flamingo isolates to established members of the family Chlamydiaceae was sufficiently distant to indicate that the three strains belong to two distinct species within a new genus. Based on these data, we propose the introduction of Chlamydiifrater gen. nov., as a new genus, and Chlamydiifrater phoenicopteri sp. nov. and Chlamydiifrater volucris sp. nov., as two new species of the genus.

Type III secretion, contact-dependent model for the intracellular development of chlamydia

The medically significant genus Chlamydia is a class of obligate intracellular bacterial pathogens that replicate within vacuoles in host eukaryotic cells termed inclusions. Chlamydia's developmental cycle involves two forms; an infectious extracellular form, known as an elementary body (EB), and a non-infectious form, known as the reticulate body (RB), that replicates inside the vacuoles of the host cells. The RB surface is covered in projections that are in intimate contact with the inclusion membrane. Late in the developmental cycle, these reticulate bodies differentiate into the elementary body form. In this paper, we present a hypothesis for the modulation of these developmental events involving the contact-dependent type III secretion (TTS) system. TTS surface projections mediate intimate contact between the RB and the inclusion membrane. Below a certain number of projections, detachment of the RB provides a signal for late differentiation of RB into EB. We use data and develop a mathematical model investigating this hypothesis. If the hypothesis proves to be accurate, then we have shown that increasing the number of inclusions per host cell will increase the number of infectious progeny EB until some optimal number of inclusions. For more inclusions than this optimum, the infectious yield is reduced because of spatial restrictions. We also predict that a reduction in the number of projections on the surface of the RB (and as early as possible during development) will significantly reduce the burst size of infectious EB particles. Many of the results predicted by the model can be tested experimentally and may lead to the identification of potential targets for drug design.

Genome sequence of Chlamydophila caviae (Chlamydia psittaci GPIC): examining the role of niche-specific genes in the evolution of the Chlamydiaceae

The genome of Chlamydophila caviae (formerly Chlamydia psittaci, GPIC isolate) (1 173 390 nt with a plasmid of 7966 nt) was determined, representing the fourth species with a complete genome sequence from the Chlamydiaceae family of obligate intracellular bacterial pathogens. Of 1009 annotated genes, 798 were conserved in all three other completed Chlamydiaceae genomes. The C.caviae genome contains 68 genes that lack orthologs in any other completed chlamydial genomes, including tryptophan and thiamine biosynthesis determinants and a ribose-phosphate pyrophosphokinase, the product of the prsA gene. Notable amongst these was a novel member of the virulence-associated invasin/intimin family (IIF) of Gram-negative bacteria. Intriguingly, two authentic frameshift mutations in the ORF indicate that this gene is not functional. Many of the unique genes are found in the replication termination region (RTR or plasticity zone), an area of frequent symmetrical inversion events around the replication terminus shown to be a hotspot for genome variation in previous genome sequencing studies. In C.caviae, the RTR includes several loci of particular interest including a large toxin gene and evidence of ancestral insertion(s) of a bacteriophage. This toxin gene, not present in Chlamydia pneumoniae, is a member of the YopT effector family of type III-secreted cysteine proteases. One gene cluster (guaBA-add) in the RTR is much more similar to orthologs in Chlamydia muridarum than those in the phylogenetically closest species C.pneumoniae, suggesting the possibility of horizontal transfer of genes between the rodent-associated Chlamydiae. With most genes observed in the other chlamydial genomes represented, C.caviae provides a good model for the Chlamydiaceae and a point of comparison against the human atherosclerosis-associated C.pneumoniae. This crucial addition to the set of completed Chlamydiaceae genome sequences is enabling dissection of the roles played by niche-specific genes in these important bacterial pathogens.

Comparative analysis of Chlamydia bacteriophages reveals variation localized to a putative receptor binding domain

Three recently discovered ssDNA Chlamydia-infecting microviruses, phiCPG1, phiAR39, and Chp2, were compared with the previously characterized phage from avian C. psittaci, Chp1. Although the four bacteriophages share an identical arrangement of their five main genes, Chpl has diverged significantly in its nucleotide and protein sequences from the other three, which form a closely related group. The VP1 major viral capsid proteins of phiCPG1 and phiAR39 (from guinea pig-infecting C. psittaci and C. pneumoniae, respectively) are almost identical. However, VP1 of ovine C. psittaci phage Chp2 shows a high rate of nucleotide sequence change localized to a region encoding the "IN5" loop of the protein, thought to be a potential receptor-binding site. Phylogenetic analysis suggests that the ORF4 replication initiation protein is evolving faster than the other phage proteins. phiCPG1, phiAR39, and Chp2 are closely related to an ORF4 homolog inserted in the C. pneumoniae chromosome. This sequence analysis opens the way toward understanding the host-range and evolutionary history of these phages.

Phage infection of the obligate intracellular bacterium, Chlamydia psittaci strain guinea pig inclusion conjunctivitis

The infectious cycle of phiCPG1, a bacteriophage that infects the obligate intracellular pathogen, Chlamydia psittaci strain Guinea Pig Inclusion Conjunctivitis, was observed using transmission electron microscopy of phage-hyperinfected, Chlamydia-infected HeLa cells. Phage attachment to extracellular, metabolically dormant, infectious elementary bodies and cointernalisation are demonstrated. Following entry, phage infection takes place as soon as elementary bodies differentiate into metabolically active reticulate bodies. Phage-infected bacteria follow an altered developmental path whereby cell division is inhibited, producing abnormally large reticulate bodies, termed maxi-reticulate bodies, which do not mature to elementary bodies. These forms eventually lyse late in the chlamydial developmental cycle, releasing abundant phage progeny in the inclusion and, upon lysis of the inclusion membrane, into the cytosol of the host cell. Structural integrity of the hyperinfected HeLa cell is markedly compromised at late stages. Released phage particles attach avidly to the outer leaflet of the outer membranes of lysed and unlysed Chlamydiae at different stages of development, suggesting the presence of specific phage receptors in the outer membrane uniformly during the chlamydial developmental cycle. A mechanism for phage infection is proposed, whereby phage gains access to replicating chlamydiae by attaching to the infectious elementary body, subsequently subverting the chlamydial developmental cycle to its own replicative needs. The implications of phage infection in the context of chlamydial infection and disease are discussed.

Type III secretion genes identify a putative virulence locus of Chlamydia

Four genes of Chlamydia psittaci strain guinea pig inclusion conjunctivitis (GPIC), whose predicted products are highly homologous to structural and regulatory components of a contact-dependent or type III secretion apparatus, were isolated. Related to genes present in several animal and plant bacterial pathogens, these genes may represent a section of a previously undetected chromosomal virulence locus analogous to several recently described virulence-associated type III secretion loci. The existence of contact-dependent secretion in Chlamydia strongly suggests that these bacteria use pathogenic mechanisms that are similar to those of other intracellular bacterial pathogens. Unlike other intracellular bacteria, however, chlamydiae are metabolically inactive extracellularly and only become capable of global protein synthesis several hours after infection. This implies that chlamydial contact-dependent secretion is only active from within, uniquely after the bacteria have been internalized by eukaryotic cells. The possible role(s) of this pathway in chlamydial pathogenesis are discussed.

Heparin-mediated inhibition of Chlamydia psittaci adherence to HeLa cells

The adherence of human strains of Chlamydia trachomatis has been recently shown to be inhibitable by heparin and heparitinase, leading to the proposal that Chlamydia binding to host cells may be mediated by a glycosaminoglycan (GAG)-dependent mechanism. We here describe the adherence of the guinea-pig pathogen, Chlamydia psittaci GPIC, to HeLa cells, which was measured by cytofluorometry with chlamydiae whose DNA was fluorescently labelled. Adherence could be inhibited by heat or trypsin pretreatment of the bacteria, and binding was much faster at 37 degrees C (reaching a plateau within 1 h) than 4 degrees C. Little binding remained when host cells were pre-fixed with paraformaldehyde, suggesting that host cell receptor mobility may be required for effective adherence. Visualization by confocal microscopy confirmed that the bacteria were at or near the host cell surface during the entire time-course of these experiments. Adherence increased as a function of pH between pH 6 and pH 8.0-8.5. Both adherence and infection of HeLa cells could be inhibited with heparin when the adherence step was performed at 4 degrees C, but only infection was inhibited when the adherence step was performed at 37 degrees C, even though heparitinase could block adherence at either 4 degrees C or 37 degrees C. Even at 4 degrees C, heparin-mediated inhibition was significantly lower at pH 8 than pH 7.4, suggesting that GAG-independent mechanisms may play a role in the higher adherence observed at basic pH. These results therefore demonstrate that a GAG-dependent adherence step may be operative in C. psittaci, and raise the possibility that other adherence mechanisms may also contribute to binding by this chlamydial strain. Furthermore, they suggest that there may not be a strict correlation between C. psittaci adherence and the ability to cause productive infections.

Homologs of Escherichia coli recJ, gltX and of a putative 'early' gene of avian Chlamydia psittaci are located upstream of the 'late' omp2 locus of Chlamydia psittaci strain guinea pig inclusion conjunctivitis

The nucleotide sequence of nearly 6 kb of genomic DNA located immediately upstream of the omp3-omp2 operon of Chlamydia psittaci strain GPIC was obtained, revealing four significant open reading frames (ORFs), named ORF1, ORF2, ORF4 and ORF5. Searches for homologous sequences in the GenBank/EMBL databases have revealed that: (a) the open-ended ORF1 putatively encodes an homolog of RecJ of Escherichia coli, thought to be required for RecBCD-independent and conjugational recombination, and for UV repair; (b) the predicted translation product of ORF4 is highly homologous to the putative product of EUO, a previously described ORF of avian C. psittaci strain 6BC which is preferentially transcribed early during the life cycle; and (c) ORF5 putatively encodes an homolog of bacterial glutamyl-tRNA synthetases. This analysis establishes the genetic linkage of late (omp3-omp2) and of a proposed early (EUO) genes in Chlamydia.

Sequence analysis of the omp2 region of Chlamydia psittaci strain GPIC: structural and functional implications

The nucleotide sequence of a 3.1-kb genomic DNA fragment carrying the omp3, omp2 and srp gene homologs from Chlamydia psittaci strain GPIC was determined. A comparative analysis of the GPIC sequence with other chlamydial omp2-linked sequences reveals highly conserved omp3 and omp2 upstream sequences across species, suggesting a unified mechanism of transcription regulation. In contrast, the omp2-srp intergenic segment, which encompasses hypothetical srp transcriptional initiation sites, is relatively less conserved in length and in sequence. Examination of the predicted translation products reveals a high degree of homology within Omp3 and Omp2 across species, with the notable exception of the N-terminal fifth of Omp2. Although the latter segment displays relatively high interspecies sequence variation, it includes a smaller segment, whose high positive charge density is conserved across species, suggesting a conserved structure/function. In contrast to Omp2 and Omp3, a comparative analysis of the predicted amino acid (aa) sequence of the srp product reveals high homology within species, but relatively little across species. A 38-aa segment near the C-terminus of Srp, whose sequence is 64% identical between C. psittaci GPIC and C. trachomatis, is partially truncated in C. psittaci 6BC.

Low-efficiency (macro-)pinocytic internalization of non-pathogenic Escherichia coli into HEp-2 cells

HEp-2 cells internalize non-pathogenic Escherichia coli bacteria by a low-efficiency internalization mechanism which is upregulated in Pho-derepressed strains (as shown by Sinai and Bavoil in 1993), and is independent of microfilament integrity but requires functional microtubules. Here, we further characterize the microtubule requirement of this pathway using various effectors of microtubule integrity and function. Furthermore, we show that internalization is enhanced upon treatment with monodansylcadaverine, a specific inhibitor of receptor mediated endocytosis, and is insensitive to brefeldin A, which promotes the microtubule-dependent reorganization of the endosome. An assay system is also described to directly evaluate the contribution of pinocytosis to this pathway based on the ability of the bacteria to cointernalize and consequently colocalize with the fluid-phase marker, Texas-red-conjugated dextran (TRD). Using this assay, Hoescht-stained bacteria were observed in TRD-containing vesicles in numbers that are consistent with their observed internalization rate. Overall, these data are strongly supportive of the existence of a low-efficiency macropinocytic mechanism of entry for these non-pathogenic bacteria. Moreover, the observed requirements for host tyrosine kinase and protein kinase C activities suggest that it is inducible.

Interaction of outer envelope proteins of Chlamydia psittaci GPIC with the HeLa cell surface

The chlamydial life cycle involves the intimate interaction of components of the infectious elementary body (EB) surface with receptors on the susceptible eukaryotic cell plasma membrane. We have developed an in vitro ligand binding assay system for the identification and characterization of detergent-extracted EB envelope proteins capable of binding to glutaraldehyde-fixed HeLa cell surfaces. With this assay, the developmentally regulated cysteine-rich envelope protein Omp2 of Chlamydia psittaci strain guinea pig inclusion conjunctivitis was shown to bind specifically to HeLa cells. HeLa cells bound Omp2 selectively over other cell wall-associated proteins, including the major outer membrane protein, and the binding of Omp2 was abolished under conditions which alter its conformation. Furthermore, trypsin treatment, which reduces EB adherence, resulted in the proteolytic removal of a small terminal peptide of Omp2 at the EB surface and inactivated Omp2 in the ligand binding assay, while having a negligible effect on the major outer membrane protein. Collectively, our results suggest that Omp2 possesses the capacity to engage in a specific interaction with the host eukaryotic cell. We speculate that, since Omp2 is present only in the infectious EB form, the observed in vitro interaction may be representative of a determining step of the chlamydial pathogenic process.

Systemic immunization with Hsp60 alters the development of chlamydial ocular disease

PURPOSE

To determine whether immunization with recombinant Hsp60 would exacerbate ocular pathology on challenge with viable chlamydial elementary bodies.

METHODS

Guinea pigs were immunized either subcutaneously with recombinant Hsp60 or both subcutaneously with recombinant Hsp60 and ocularly with attenuated Salmonella typhimurium expressing the guinea pig inclusion conjunctivitis (GPIC) Hsp60 antigen. All animals were challenged in the conjunctiva with the agent of GPIC, and the degree of gross ocular pathology was determined. Immunoglobulin G (IgG) and immunoglobulin A (IgA) antibody titers to Hsp60 were measured in ocular secretions as a measure of the degree of immunization.

RESULTS

In primary and challenge GPIC infection, the degree of gross ocular pathology was lower in the immunized group. The presence of high IgA and IgG antibody titers to Hsp60 in tears suggested that the response may have been modified by the presence of blocking antibodies that either may have removed the antigen quickly or prevented interaction with sensitized T cells. In contrast to subcutaneous immunization, the combined immunization regimen, consisting of subcutaneous recombinant Hsp60 followed by ocular inoculation of the attenuated Salmonella, resulted in no difference in gross pathology after reinfection of guinea pigs with GPIC.

CONCLUSIONS

These data indicated that the immunization with Hsp60 did not produce exacerbated disease on challenge with viable organisms; however, the data suggested that the route of administration, form of antigen, or both may be critical in the disease process.

Deviation of immune response to Chlamydia psittaci outer membrane protein in lipopolysaccharide-hyporesponsive mice

The outcome of infection is determined by both the quantity and the quality of an induced immune response. In particular, it has been demonstrated for selected pathogens that induction of TH1 or TH2 type helper T-cell subsets determines whether an immune response gives rise to protective immunity or disease-associated immunopathology. The nature of the antigen and the type of antigen-presenting cells recruited in the induction of a response are critical factors that influence the quality of the immune response. Of particular interest in this respect is the immune response to bacterial particles and the impact of cell wall-associated lipopolysaccharide (LPS) on that response. Nonspecific activation of macrophages and B lymphocytes by LPS could skew the phenotype of activated antigen-presenting cells and selectively alter the immunoglobulin isotypes and helper T-cell subsets that are induced following infection. In an initial attempt to detect immune deviation associated with LPS stimulation, we have compared the immunoglobulin isotypes of antibodies specific for the cysteine-rich outer membrane protein Omp2 induced in normal and LPS-hyporesponsive mice following immunization with Chlamydia psittaci strain guinea pig inclusion conjunctivitis whole elementary bodies. We report that there is a dramatic shift of Omp2-specific antibody from predominantly immunoglobulin G2a (IgG2a) isotype in LPS-hyporesponsive mice to high levels of IgG1 isotype in LPS-responder strains. The dependence of the IgG1 isotype shift on the LPS responder status is linked to the structure of the antigen and its natural processing pathway since LPS-hyporesponsive mice are not, in general, deficient in IgG1 antibody production. In particular, the antibody response to purified recombinant Omp2 is predominantly of the IgG1 isotype even in LPS-hyporesponsive mice.

Dissociation of immune determinants of outer membrane proteins of Chlamydia psittaci strain guinea pig inclusion conjunctivitis

Chlamydia trachomatis is an important human pathogen. Research to develop a Chlamydia vaccine has focused on the major outer membrane protein (MOMP). Determinants of this protein elicit serovar-specific neutralizing antibodies which are thought to play a critical role in protective immunity. MOMP-specific antibody responses are highly variable in the polymorphic population. Genetic factors which might influence the MOMP-specific immune response are consequently of particular interest. The C. psittaci strain guinea pig inclusion conjunctivitis (GPIC) is a natural pathogen of the guinea pig that causes both ocular and genital tract infections that closely resemble those caused by C. trachomatis in humans. As such, it provides an excellent model for disease. In this report, we explore the influence of major histocompatibility complex-linked genes on the MOMP-specific antibody response in mice immunized with either whole GPIC elementary bodies or recombinant GPIC MOMP. Our results indicate that the MOMP-specific antibody response is major histocompatibility complex linked such that mice of the H-2d haplotype are high responders while mice of the H-2k haplotype are low responders. We demonstrate that MOMP-specific B cells are present in H-2k strains which are, however, deficient in MOMP-specific helper T cells. Although immunization of low-MOMP-responder strains with whole chlamydial elementary bodies induces high levels of immunoglobulin G antibody specific for Omp2, the cysteine-rich outer membrane protein, MOMP-specific B cells are unable to receive help from Omp2-specific T cells. The failure of intermolecular help from Omp2-specific T cells and related observations raise important issues regarding the processing and presentation of chlamydial antigens and the design of optimal subunit vaccines.

Hyper-invasive mutants define a novel Pho-regulated invasion pathway in Escherichia coli

We have isolated two transposon insertion mutations of the pst-phoU operon which result in the constitutive expression of the phoA gene product, alkaline phosphatase. The two mutations also render Escherichia coli invasive towards cultured HEp-2 cells and define a novel Pho-regulated invasion pathway. The presence of the large 'invasion' plasmid derived from an entero-invasive E. coli (EIEC) clinical isolate in these mutants leads to enhanced invasiveness toward cultured HEp-2 cells, a phenomenon referred to as the 'hyper-invasive' phenotype. Transduction of a pst-phoU insertion mutation into clinical isolates of EIEC and Shigella flexneri results in constitutive PhoA expression and coupled hyper-invasiveness in the former but not the latter. We speculate that the Pho-regulated invasion pathway described here, while silent in bacteria grown in standard laboratory rich media, may become functional in the host when invasive bacteria encounter nutrient starvation and/or other related stress conditions.

Partial protection against genital reinfection by immunization of guinea-pigs with isolated outer-membrane proteins of the chlamydial agent of guinea-pig inclusion conjunctivitis

Because partial protection against reinfection is induced by experimental infection in the guinea-pig model of genital chlamydial infection, we sought to induce immunity by immunization. Female guinea-pigs were immunized subcutaneously with the major outer-membrane protein (MOMP) and the 61 kDa cysteine-rich outer-membrane protein (61 kDa) of the agent of guinea-pig inclusion conjunctivitis (GPIC) eluted from SDS-polyacrylamide gels (SDS-MOMP, SDS-61 kDa). Post-immunization sera and secretions contained antibodies to the SDS-purified proteins at high titre as measured by immunoblotting, whereas enzyme immunoassays (EIA) using whole elementary bodies as antigen showed significantly lower titres (P < 0.001). Likewise, blastogenic responses of peripheral mononuclear cells to GPIC elementary bodies were weak. Animals immunized with SDS-MOMP and SDS-61 kDa were fully susceptible to intravaginal challenge, as were control animals immunized with buffer without protein. Another group of animals were immunized with material prepared by extraction of chlamydial outer-membrane complexes with octyl beta-D-glucopyranoside (OGP) and dithiothreitol, which consisted largely of MOMP (OGP-MOMP). In contrast to the SDS-MOMP group, sera and secretions in the OGP-MOMP group showed high titres in EIA, and high titre antibodies to MOMP by immunoblot; however, most animals also had antibodies to 61 kDa, 72 kDa and ca. 84 kDa outer-membrane proteins. OGP-MOMP animals were partially protected against genital challenge as evidenced by low inclusion scores compared to control animals, although duration of infection measured by culture isolation was similar to controls.(ABSTRACT TRUNCATED AT 250 WORDS)

Expression and translocation of the chlamydial major outer membrane protein in Escherichia coli

The entire gene encoding the major outer membrane protein (MOMP) from Chlamydia psittaci strain GPIC has been cloned and expressed in Escherichia coli. A tightly regulated T7 promoter is used to control expression of the protein in Escherichia coli. Upon induction of expression, the precursor (pre-MOMP) is synthesized in the cell. This is followed by the appearance of a lower molecular weight protein that comigrates with mature MOMP from chlamydial elementary bodies by both one-dimensional sodium dodecyl sulfate polyacrylamide gel electrophoresis and two-dimensional gel electrophoresis. When E. coli cells expressing MOMP are converted to spheroplasts and subjected to protease treatment, MOMP is quantitatively degraded while cytoplasmic pre-MOMP is protected from degradation. Whole cells subjected to the same protease treatment show no degradation of MOMP. Furthermore, MOMP is not detected in surface-labeling experiments using several MOMP-specific antibodies. These data indicate that pre-MOMP is translocated to the periplasmic space and processed but is not surface exposed in E. coli. Expression of MOMP in this system causes a significant reduction in cell viability. In addition, coexpression in E. coli of MOMP or a MOMP-PhoA fusion with various chaperone proteins does not alter the level of MOMP translocation.

Characterization of virulence genes of enteroinvasive Escherichia coli by TnphoA mutagenesis: identification of invX, a gene required for entry into HEp-2 cells

While enteroinvasive Escherichia coli (EIEC) and shigellae are genotypically nearly identical, a difference has been reported in the infective dose to humans: EIEC is 10,000-fold less infectious than shigellae. A possible basis for this difference lies in the inherent invasiveness of these bacteria toward epithelial cells. Thus, despite the high degree of homology between the invasion plasmids of EIEC and shigellae, substantial differences in genetic organization and/or sequence may exist. We have undertaken a systematic genetic analysis of the EIEC plasmid pSF204, using transposon mutagenesis. Congo red-negative TnphoA insertion mutants (Pcr- PhoA-) and TnphoA fusion mutants (PhoA+) were isolated and screened for the ability to invade cultured HEp-2 cells. Most invasion-negative (Inv-) mutations mapped to a 30-kb segment of the invasion plasmid, including homologs of the Shigella flexneri ipa, mxi, and spa genes. Inv- PhoA+ fusions in the EIEC ipaC, mxiG, mxiJ, mxiM, and mxiD homologs and in a proposed new gene, named invX, located downstream of the spa region were identified and characterized. This analysis indicates the presence of the ipaC, mxiG, mxiJ, mxiM, mxiD, and invX gene products in the EIEC cell envelope and demonstrates a strict requirement for these genetic loci in invasion. Overall, our results suggest a high degree of genetic, structural, and functional homology between the EIEC and S. flexneri large invasion plasmids.