Bovine Lactoferrin Decreases Cholera-Toxin-Induced Intestinal Fluid Accumulation in Mice by Ganglioside Interaction

Secretory diarrhea caused by cholera toxin (CT) is initiated by binding of CT’s B subunit (CTB) to GM1-ganglioside on the surface of intestinal cells. Lactoferrin, a breast milk glycoprotein, has shown protective effect against several enteropathogens. The aims of this study were to determine the effect of bovine-lactoferrin (bLF) on CT-induced intestinal fluid accumulation in mice, and the interaction between bLF and CT/CTB with the GM1-ganglioside receptor. Fluid accumulation induced by CT was evaluated in the mouse ileal loop model using 56 BALB/c mice, with and without bLF added before, after or at the same time of CT administration. The effect of bLF in the interaction of CT and CTB with GM1-ganglioside was evaluated by a GM1-enzyme-linked immunosorbent assay. bLF decreased CT-induced fluid accumulation in the ileal loop of mice. The greatest effect was when bLF was added before CT (median, 0.066 vs. 0.166 g/cm, with and without bLF respectively, p<0.01). We conclude that bLF decreases binding of CT and CTB to GM1-ganglioside, suggesting that bLF suppresses CT-induced fluid accumulation by blocking the binding of CTB to GM1-ganglioside. bLF may be effective as adjunctive therapy for treatment of cholera diarrhea.     

Rapid transient expression of cholera toxin B subunit (CTB) in Nicotiana benthamiana

Cholera toxin B subunit (CTB) has the potential to be an effective adjuvant for mucosal vaccines because of its ability to increase antigen uptake and presentation by antigen-presenting cells through GM1-ganglioside binding. CTB has been produced using different recombinant protein expression systems. This study used the geminiviral replicon system to transiently express CTB in Nicotiana benthamiana. The plant-optimized CTB gene was cloned into a geminiviral vector and infiltrated into N. benthamiana leaves. The highest CTB protein level was observed on day 4 with approximately 4 μg/g fresh weight. The Western blot analysis using anti-CTB suggests assembly of CTB into oligomers. Based on the GM1-ELISA results, this CTB transiently expressed in plants showed biological activity for binding the intestinal epithelial cell membrane glycolipid receptor, GM1-glanglioside, which implies its potential as an adjuvant for mucosal vaccines.     

The Effect of the B Subunit of Cholera Toxin on the Action of Nerve Growth Factor on PC12 Cells

Abstract: Exogenous gangliosides, especially ganglioside GM1 (GM1), seem to potentiate the action of nerve growth factor (NGF). We have examined the possible regulation of the NGF signaling pathway in PC12 cells by the B subunit of cholera toxin (CTB), which binds to endogenous GM1 specifically and with a high affinity. CTB treatment (1 μg/ml) enhanced NGF-induced neurite outgrowth from PC12 cells, NGF-induced activation of ribosomal protein S6 kinase, and NGF-induced stimulation of trk phosphorylation. CTB plus NGF also caused a greater inhibition of [³H]-thymidine incorporation into DNA than did NGF alone. These enhancing effects of CTB were blocked by the presence of cytochalasin B in the culture medium but were not affected by the presence of colchicine or by the depletion of Ca²⁺ in the medium. ¹²⁵I-NGF binding experiments revealed that CTB treatment did not affect the specific binding of NGF to the cells. These results strongly suggest that the binding of cell surface GM1 by CTB modulates the pathway of intracellular signaling initiated by NGF and that the association of CTB with a cytoskeletal component is essential for these effects.     

Interactions of ganglioside GM1 with human and fetal calf sera. Formation of ganglioside-serum albumin complexes

The interactions of ganglioside GM1 with human and fetal calf sera were studied, the following main results being obtained: (a) GM1, upon incubation with both sera gave origin to two GM1-protein complexes, which also occurred after interaction of GM1 with the albumin fractions prepared from the same sera. Instead no complex formation occurred using the albumin-free fractions. Therefore GM1 appeared to specifically bind serum albumin and to form GM1-albumin complexes. (b) GM1 binding to serum albumin started at ganglioside concentrations surely micellar (above 10(-6) M), was time and concentration dependent, and resulted in a relevant degree of GM1 complexation (up to 80% of total GM1 in human serum and up to 18% in fetal calf serum). (c) the binding kinetics appeared, in both serum and the correspondent albumin fraction, to be biphasic: in the first phase, occurring till about 2 . 10(-4) M GM1, the ratio between bound and total GM1 increased linearly with increasing GM1 concentration; in the second phase, occurring above 2 . 10(-4) M, the ratio remained practically constant. After these findings it should be expected that GM1, when present in serum containing systems, forms complexes with albumin. This should be approximately considered when studying the effects of exogeneous GM1 in in vivo and in vitro (tissue cultures) systems.     

Cholera toxin binds to lipid rafts but has a limited specificity for ganglioside GM1

Lipid rafts and the formation of an immunological synapse are crucial for T-cell activation. Binding of cholera toxin B subunit (CTB) to ganglioside GM1 is a marker to identify lipid rafts. Primary human T cells were isolated from healthy donors and were stimulated with superantigen staphylococcus enterotoxin B (SEB) and stained with cholera toxin B-fluorescein isothiocyanate (CTB-FITC). An optimized staining procedure is required to stain lipid rafts exclusively on the cell surface. Unstimulated T cells show a few CTB binding spots on the cell surface. The size and number of CTB-binding lipid rafts are strongly upregulated during T-cell activation in SEB-stimulated CD4(+) T cells. However, our data show that the specificity of CTB for GM1 ganglioside is limited, because the binding capacity is partly resistant to inhibition of ganglioside synthesis and sensitive to trypsin digestion. Our results indicate that the binding of FITC-labeled CTB can be divided into at least three different categories: a specific binding of CTB to ganglioside GM1, a nonspecific binding of CTB probably to glycosylated surface proteins and a nonspecific binding of FITC to the cell surface.     

Cholera toxin entry into pig enterocytes occurs via a lipid raft- and clathrin-dependent mechanism

The small intestinal brush border is composed of lipid raft microdomains, but little is known about their role in the mechanism whereby cholera toxin gains entry into the enterocyte. The present work characterized the binding of cholera toxin B subunit (CTB) to the brush border and its internalization. CTB binding and endocytosis were performed in organ-cultured pig mucosal explants and studied by fluorescence microscopy, immunogold electron microscopy, and biochemical fractionation. By fluorescence microscopy CTB, bound to the microvillar membrane at 4 degrees C, was rapidly internalized after the temperature was raised to 37 degrees C. By immunogold electron microscopy CTB was seen within 5 min at 37 degrees C to induce the formation of numerous clathrin-coated pits and vesicles between adjacent microvilli and to appear in an endosomal subapical compartment. A marked shortening of the microvilli accompanied the toxin internalization whereas no formation of caveolae was observed. CTB was strongly associated with the buoyant, detergent-insoluble fraction of microvillar membranes. Neither CTB's raft association nor uptake via clathrin-coated pits was affected by methyl-beta-cyclodextrin, indicating that membrane cholesterol is not required for toxin binding and entry. The ganglioside GM(1) is known as the receptor for CTB, but surprisingly the toxin also bound to sucrase-isomaltase and coclustered with this glycosidase in apical membrane pits. CTB binds to lipid rafts of the brush border and is internalized by a cholesterol-independent but clathrin-dependent endocytosis. In addition to GM(1), sucrase-isomaltase may act as a receptor for CTB.     

Alteration of ganglioside synthesis by GM3 synthase knockout in murine embryonic fibroblasts

To probe the functions of membrane gangliosides, the availability of ganglioside-depleted cells would be a valuable resource. To attempt to identify a useful genetic model of ganglioside depletion, we assessed ganglioside metabolism in murine GM3 synthase (GM3S)-/- knockout primary embryonic fibroblasts (MEF), because normal fibroblast gangliosides (GM3, GM2, GM1, and GD1a), all downstream products of GM3S, should be absent. We found that heterozygote MEF (GM3S+/-) did have a 36% reduced content of qualitatively normal gangliosides (7.0+/-0.8 nmol LBSA/mg cell protein; control: 11+/-1.6 nmol). However, two unexpected findings characterized the homozygous (GM3-/-) MEF. Despite complete knockout of GM3S, (i) GM3-/- MEF retained substantial ganglioside content (21% of normal or 2.3+/-1.1 nmol) and (ii) these gangliosides were entirely different from those of wild type MEF by HPTLC. Mass spectrometry identified them as GM1b, GalNAc-GM1b, and GD1alpha, containing both N-acetyl and N-glycolylneuraminic acid and diverse ceramide structures. All are products of the 0 pathway of ganglioside synthesis, not normally expressed in fibroblasts. The results suggest that complete, but not partial, inhibition of GM3 synthesis results in robust activation of an alternate pathway that may compensate for the complete absence of the products of GM3S.     

Binding of fluorescently labeled cholera toxin subunit B to glycolipids in the human submandibular gland and inhibition of binding by periodate oxidation and by galactose

FITC-labeled cholera toxin subunit B (CTB) stained the surfaces of cells of mucous acini in the submandibular gland. CTB, also called choleragenoid, binds to the GM1 glycolipid in the cell membrane. The binding in most acini was inhibited by periodic acid oxidation of the sections, while some acini remained unaffected even after increased oxidation. Staining with the subunit was also reduced significantly by adding galactose to the incubation medium. Binding of CTB to cell surfaces apparently requires intact sialic groups on most, but not all, cell surfaces. Oxidation of the sialic acid residues may influence the structure of the sialylated GM1 molecules on the cell surface in different ways. It is possible that both the sialic acid residue and the terminal galactose are oxidized. Alternatively, the sialic acid may be resistant to acid hydrolysis in gangliosides in which the sialic acid is attached to the internal galactose residue linked to GalNAc, as in the GM1 glycolipid. Inhibition of the GM1 receptor binding to cholera toxin has potential for protection of humans against cholera. Galactose and agents that modify sialic acid inhibit the accessibility of the toxin to the GM1 carbohydrate receptor. Human milk contains high levels of sialic acid glycoconjugates that may provide defense mechanisms.     

Effect of immunization with a recombinant cholera toxin B subunit/somatostatin fusion protein on immune response and growth hormone levels in mice

Somatostatin (SS) is a hormone that inhibits growth hormone secretion. Cholera toxin B subunit (CTB) is a widely used adjuvant to improve the immunogenicity of co-administrated antigen. To block the growth hormone-inhibiting effect of SS, a fusion gene of CTB and SS was constructed and expressed in Escherichia coli. The purified CTB/SS fusion protein polymerized into a biologically active pentamer required for CTB binding to the GM1 ganglioside receptor. Immunization with the CTB/SS protein induced specific immunity against CTB and SS in mice. The serum growth hormone of the CTB/SS-treated mice increased by 29?% (P?<?0.05) compared with the control. The results indicated that the CTB/SS fusion protein was effective in inducing immune response against SS as well as elevating the growth hormone level.     

Very efficient extracellular production of cholera toxin B subunit using Bacillus brevis

Abstract We have constructed a very efficient synthesis and secretion system for cholera toxin B subunit (CTB) of Vibrio cholerae 569B using Bacillus brevis . The constructed expression-secretion vector has the multiple promoters and the signal peptide coding region of the mwp gene, a structural gene for one of the major cell wall proteins of B. brevis strain 47, directly followed by the gene encoding the mature CTB. A large amount of mature CTB (1.4 g per liter of culture) was secreted into the medium. It had the same amino terminal amino acid sequence as that of authentic CTB and was fully active in GM1 ganglioside binding assay.     

Structural studies of receptor binding by cholera toxin mutants.

The wide range of receptor binding affinities reported to result from mutations at residue Gly 33 of the cholera toxin B-pentamer (CTB) has been most puzzling. For instance, introduction of an aspartate at this position abolishes receptor binding, whereas substitution by arginine retains receptor affinity despite the larger side chain. We now report the structure determination and 2.3-A refinement of the CTB mutant Gly 33-->Arg complexed with the GM1 oligosaccharide, as well as the 2.2-A refinement of a Gly 33-->Asp mutant of the closely related Escherichia coli heat-labile enterotoxin B-pentamer (LTB). Two of the five receptor binding sites in the Gly 33-->Arg CTB mutant are occupied by bound GM1 oligosaccharide; two other sites are involved in a reciprocal toxin:toxin interaction; one site is unoccupied. We further report a higher resolution (2.0 A) determination and refinement of the wild-type CTB:GM1 oligosaccharide complex in which all five oligosaccharides are seen to be bound in essentially identical conformations. Saccharide conformation and binding interactions are very similar in both the CTB wild-type and Gly 33-->Arg mutant complexes. The protein conformation observed for the binding-deficient Gly 33-->Asp mutant of LTB does not differ substantially from that seen in the toxin:saccharide complexes. The critical nature of the side chain of residue 33 is apparently due to a limited range of subtle rearrangements available to both the toxin and the saccharide to accommodate receptor binding. The intermolecular interactions seen in the CTB (Gly 33-->Arg) complex with oligosaccharide suggest that the affinity of this mutant for the receptor is close to the self-affinity corresponding to the toxin:toxin binding interaction that has now been observed in crystal structures of three CTB mutants.     

CTB/CS3融合蛋白与GM1结合能力和免疫原性的分析

免疫霍乱毒素B亚单位（CTB）或肠毒素大肠杆菌（ETEC）定居因子CS3可使人体对ETEC的侵染有保护作用.为探索研制ETEC双组分亚单位疫苗的可行性,利用大肠杆菌诱导表达系统表达了CTB与CS3的融合蛋白（CTB/CS3）.蛋白质印迹结果表明,诱导表达的29 ku蛋白具有CTB和CS3蛋白双重抗原性.经Ni-NTA亲和层析纯化获得重组蛋白CTB/CS3,复性的重组蛋白可以部分形成五聚体并保留了与神经节苷脂GM1的结合能力.动物实验表明,融合蛋白CTB/CS3具有CTB和CS3蛋白的双重免疫原性,同时,CTB的免疫载体作用提高了CS3的免疫强度.     

Novel carbohydrate binding site recognizing blood group A and B determinants in a hybrid of cholera toxin and Escherichia coli heat-labile enterotoxin B-subunits

The B-subunits of cholera toxin (CTB) and Escherichia coli heat-labile enterotoxin (LTB) are structurally and functionally related. However, the carbohydrate binding specificities of the two proteins differ. While both CTB and LTB bind to the GM1 ganglioside, LTB also binds to N-acetyllactosamine-terminated glycoconjugates. The structural basis of the differences in carbohydrate recognition has been investigated by a systematic exchange of amino acids between LTB and CTB. Thereby, a CTB/LTB hybrid with a gain-of-function mutation resulting in recognition of blood group A and B determinants was obtained. Glycosphingolipid binding assays showed a specific binding of this hybrid B-subunit, but not CTB or LTB, to slowly migrating non-acid glycosphingolipids of human and animal small intestinal epithelium. A binding-active glycosphingolipid isolated from cat intestinal epithelium was characterized by mass spectrometry and proton NMR as GalNAcalpha3(Fucalpha2)Galbeta4(Fucalpha3)Glc NAcbeta3Galbeta4Glc NAcbeta3Galbeta4Glcbeta1Cer. Comparison with reference glycosphingolipids showed that the minimum binding epitope recognized by the CTB/LTB hybrid was Galalpha3(Fucalpha2)Galbeta4(Fucalpha3)GlcNAc beta. The blood group A and B determinants bind to a novel carbohydrate binding site located at the top of the B-subunit interfaces, distinct from the GM1 binding site, as found by docking and molecular dynamics simulations.     

Neurite outgrowth in dorsal root neuronal hybrid clones modulated by ganglioside GM1 and disintegrins.

Subclones of F11 neuronal hybrid cells (neuroblastoma x dorsal root ganglion neurons) have segregated differing and/or overlapping neuritogenic mechanisms on three substrata--plasma fibronectin (pFN) with its multiple receptor activities, cholera toxin B subunit (CTB) for binding to ganglioside GM1, and platelet factor-4 (PF4) for binding to heparan sulfate proteoglycans. In this study, specific cell surface receptor activities for the three substrata were tested for their modulation during neuritogenesis by several experimental paradigms, using F11 subclones representative of three differentiation classes (neuritogenic on pFN only, on CTB only, or on all three substrata). When cycloheximide was included in the medium to inhibit protein synthesis during the active period, neurite formation increased significantly for all subclones on all three substrata, virtually eliminating substratum selectivity for differentiation mediated by cell surface integrin, ganglioside GM1, or heparan sulfate proteoglycans. Therefore, one or more labile proteins (referred to as disintegrins) must modulate functions of matrix receptors (e.g., integrins) mediating neurite formation. To verify whether cycloheximide-induced neuritogenesis was also regulated by integrin interaction with cell surface GM1, two approaches were used. When (Arg-Gly-Asp-Ser)-containing peptide A was added to the medium, it completely inhibited cycloheximide-induced neuritogenesis on all three substrata of all subclones, indicating stringent requirement for cell surface integrin function in these mechanisms. In contrast, when CTB or a monoclonal anti-GM1 antibody was also added to the medium, cycloheximide-induced neuritogenesis was amplified further on pFN and sensitivity to peptide A inhibition was abolished. Therefore, in some contexts ganglioside GM1 must complex with integrin receptors at the cell surface to modulate their function. These results also indicate that (a) cycloheximide treatment leads to loss of substratum selectivity in neuritogenesis, (b) this negative regulation of neurite outgrowth is affected by integrin receptor association with labile regulatory proteins (disintegrins) as well as with GM1, and (c) complexing of GM1 by multivalent GM1-binding proteins shifts neuritogenesis from an RGDS-dependent integrin mechanism to an RGDS-independent receptor mechanism.     

Cooperativity of ganglioside-dependent with protein-dependent substratum adhesion and neurite extension of human neuroblastoma cells

The potential involvement of gangliosides in the adherence and neurite extension of human neuroblastoma cells (Platt and La-N1) was investigated on tissue culture substrata coated with the ganglioside GM1-binding protein, cholera toxin B (CTB) subunit, for comparison with similar processes on plasma fibronectin (pFN)-coated substrata. Cells attached with reduced efficiency on CTB substrata as compared with pFN substrata and required a much longer time to form neurite processes for a small percentage of cells on CTB. The specificity of these processes for GM1 binding was tested in a variety of ways. Supplementation of the cells with exogenous GM1, but not GD1a, identified a larger population of cells adherent on CTB (comparable to pFN-adherent cells) and dramatically increased the proportion of cells capable of forming neurites without reducing the time requirement. In ultrastructural studies using the scanning electron microscope (SEM) and immunofluorescence (IF) analyses to discriminate microtubule distributions, neurites of GM1-supplemented cells on CTB were virtually identical with pFN-adherent neurites, whereas unsupplemented cells on CTB generated processes with fine-structural differences. Treatment of cells during the GM1 supplementation period with cycloheximide completely abolished the ability of cells to generate neurites on CTB and decreased the adhesive capacity of cells as well; a similar treatment of cells had no adverse effect on adherence or neurite extension on pFN. The importance of one or more proteins in GM1-dependent processes was further confirmed by demonstrating the trypsin sensitivity of a cell surface component(s) required to achieve maximal attachment on CTB; in contrast, adherence and neurite extension on pFN were much more resistant to this treatment process. Therefore, these experiments demonstrate (a) that certain cell surface gangliosides are capable of mediating adherence and neurite outgrowth of human neuroblastoma cells on a suitable ganglioside-binding substratum; (b) this ganglioside dependence is cooperative with one or more cell surface proteins which can now be analysed. These results are discussed in light of the identification in ref. [16] (Exp cell res 169 (1987) 311) of a second ‘cell-binding’ domain on the pFN molecule competent for adherence and neurite extension of these neuroblastoma cells, as well as the potential role of pFN binding to a complex ganglioside on the surface of these neural tumor cells in these processes.     

Visualization of GM1 with cholera toxin B in live epididymal versus ejaculated bull, mouse, and human spermatozoa

The organization of membrane subdomains in mammalian sperm has recently generated controversy, with several reports describing widely differing localization patterns for the ganglioside GM1. Using the pentameric B subunit of cholera toxin (CTB), we found GM1 to be restricted to the plasma membrane overlying the acrosome in the heads of live murine sperm. Interestingly, CTB had minimal binding to live bovine and human sperm. To investigate whether this difference in GM1 localization was because of species differences or differences between collection from the epididymis (mouse) or an ejaculate (bull, human), we examined epididymal bovine and human sperm. We found that GM1 localized to the plasma membrane overlying the acrosome in sperm from these species. To determine whether some component of seminal plasma was interfering with the ability of CTB to access GM1, we incubated epididymal mouse sperm with fluid from murine seminal vesicles and epididymal bull sperm with bovine seminal plasma. This treatment largely abolished the ability of the CTB to bind to GM1, producing a fluorescence pattern similar to that reported for the human. The most abundant seminal plasma protein, PDC-109, was not responsible for this loss. As demonstration that the seminal plasma was not removing GM1, sperm exposed to seminal plasma were fixed before CTB addition, and again displayed fluorescence over the acrosome. These observations reconcile inconsistencies reported for the localization of GM1 in sperm of different species, and provide evidence for the segregation of GM1 to a stable subdomain in the plasma membrane overlying the acrosome.     

Characterization of natural killer cells and their precursors in the murine bone marrow

We have fractionated murine bone marrow cells according to their density on bovine serum albumin (BSA) gradient and studied (a) the NK activity against YAC-1 targets, (b) the proportion of asialo GM1+ lymphocytes, (c) and the presence of large granular lymphocytes (LGL) in the different fractions (A, B, C, D). The NK activity was found mainly in the C fraction, but the proportion of asialo GM1+ cells was the same in every fraction. No LGLs were found in the bone marrow. Cells from the various fractions were also transplanted into irradiated recipients. Seven days later the highest NK activity was found in the spleens of mice injected with cells from the A + B fractions indicating that the immediate precursors for NK cells reside in the low density fractions of the BSA gradient. Mice transplanted with C or D fractions needed longer time to develop normal NK levels. The treatment of bone marrow cells before transplantation with anti-asialo GM1+ complement did not inhibit the development of NK activity, so it can be concluded that the precursor for NK is asialo GM1-.     

Differential activities of plant polyphenols on the binding and internalization of cholera toxin in vero cells

Plant polyphenols, RG-tannin, and applephenon had been reported to inhibit cholera toxin (CT) ADP-ribosyltransferase activity and CT-induced fluid accumulation in mouse ileal loops. A high molecular weight fraction of hop bract extract (HBT) also inhibited CT ADP-ribosyltransferase activity. We report here the effect of those polyphenols on the binding and entry of CT into Vero cells. Binding of CT to Vero cells or to ganglioside GM1, a CT receptor, was inhibited in a concentration-dependent manner by HBT and applephenon but not RG-tannin. These observations were confirmed by fluorescence microscopy using Cy3-labeled CT. Following toxin binding to cells, applephenon, HBT, and RG-tannin suppressed its internalization. HBT or applephenon precipitated CT, CTA, and CTB from solution, creating aggregates larger than 250 kDa. In contrast, RG-tannin precipitated CT poorly; it formed complexes with CT, CTA, or CTB, which were demonstrated with sucrose density gradient centrifugation and molecular weight exclusion filters. In agreement, CTA blocked the inhibition of CT internalization by RG-tannin. These data suggest that some plant polyphenols, similar to applephenon and HBT, bind CT, forming large aggregates in solution or, perhaps, on the cell surface and thereby suppress CT binding and internalization. In contrast, RG-tannin binding to CT did not interfere with its binding to Vero cells or GM1, but it did inhibit internalization.     

Surface display of the cholera toxin B subunit on Staphylococcus xylosus and Staphylococcus carnosus.

The heterologous surface expression of the cholera toxin B subunit (CTB) from Vibro cholerae in two staphylococcal species, Staphylococcus xylosus and Staphylococcus carnosus, has been investigated. The gene encoding native CTB (103 amino acids) was introduced into gene constructs encoding chimeric receptors designed to be translocated and anchored on the outer cell surface of the staphylococci. Since functionality of CTB is correlated with its ability to form pentamers and the capacity of the pentameric CTB to bind the GM1 ganglioside, both the surface accessibility and the functionality of the surface-displayed CTB receptors were evaluated. It could be concluded that the chimeric receptors were targeted to the cell wall of the staphylococci, since they could be released by lysostaphin treatment and, after subsequent affinity purification, identified as full-length products by immunoblotting. Surface accessibility of the chimeric receptors was demonstrated by a colorimetric assay and by immunofluorescence staining with a CTB-reactive rabbit antiserum. Pentamerization was investigated by using a monoclonal antibody described to be specific for pentameric CTB, and the functionality of the receptors was tested in a binding assay with digoxigenin-labelled GM1. It was concluded that functional CTB was present on both types of staphylococci, and for S. carnosus, the reactivity to the pentamer-specific monoclonal antibody and in the GM1 binding assay was indeed significant. The implications of the results for the design of live bacterial vaccine delivery systems intended for administration by the mucosal route are discussed.