Lack of Fc Gamma Receptor IIIA Promotes Rather than Suppresses Humoral and Cellular Immune Responses after Mucosal or Parenteral Immunization with Antigen and Adjuvants.

The Fcγ receptor IIIA (FcγRIIIA) has traditionally been known as a positive regulator of immune responses. Consistent with this, mice deficient in FcγRIIIA are protected from various inflammation-associated pathologies including several autoimmune diseases. In contrast to this accepted dogma, we show here that mice lacking FcγRIIIA developed increased rather than reduced both humoral and cellular immune responses to mucosal (sublingual) immunization with ovalbumin (OVA) given together with the strong mucosal adjuvant cholera toxin as well as to parenteral (subcutaneous) immunization with OVA in complete Freund's adjuvant. After either route of immunization, in comparison with concomitantly immunized wild-type mice, FcγRIIIA-/- mice had increased serum anti-OVA IgG (IgG1 but not IgG2) antibody responses as well as augmented cellular responses that included memory B cells and effector T cells. The increments in immune responses in FcγRIIIA-/- mice were similar to those seen in FcγRIIB-/- mice. Furthermore, OVA-pulsed FcγRIIIA-/- DCs, similar to OVA-specific FcγRIIB-/- DCs, had enhanced capacity to activate OVA-specific OT-II T cells, which was even further pronounced when DCs were pulsed with IgG1-complexed OVA. Our data support an inhibitory-regulatory role of FcγRIIIA on vaccine/adjuvant-induced immune responses and demonstrate that lack of FcγRIIIA can promote rather than suppress both humoral and cellular immune responses.

Mucosal delivery routes for optimal immunization: targeting immunity to the right tissues.

The mucosal immune system exhibits a high degree of anatomic compartmentalization related to the migratory patterns of lymphocytes activated at different mucosal sites. The selective localization of mucosal lymphocytes to specific tissues is governed by cellular "homing" and chemokine receptors in conjunction with tissue-specific addressins and epithelial cell-derived chemokines that are differentially expressed in "effector" tissues. The compartmentalization of mucosal immune responses imposes constraints on the selection of vaccine administration route. Traditional routes of mucosal immunization include oral and nasal routes. Other routes for inducing mucosal immunity include the rectal, vaginal, sublingual, and transcutaneous routes. Sublingual administration is a new approach that results in induction of mucosal and systemic T cell and antibody responses with an exceptionally broad dissemination to different mucosae, including the gastrointestinal and respiratory tracts, and the genital mucosa. Here, we discuss how sublingual and different routes of immunization can be used to generate immune responses in the desired mucosal tissue(s).

Interaction of cholera toxin and toxin derivatives with lymphocytes. I. Binding properties and interference with lectin-induced cellular stimulation.

The interaction of cholera toxin and a number of toxin derivatives, containing different proportions of light and heavy toxin-composing subunits (L and H), with mouse lymphocytes was studied. Experiments with [(125)I]toxin showed that a single cell can rapidly, within minutes, bind up to 40,000 molecules of toxin, the association constant was estimated to 7 +/- 4 x 10(8) liters/mol, and binding was found to be very similar at 37 degrees C and 5 degrees C. Immunofluorescence studies revealed that the toxin attachment is located on the cell surface, and that purified L subunit but not H subunit binds to the cells. A natural cholera toxoid, built up by aggregated L subunits, showed almost identical binding properties as toxin to the cells. Pure G(M1) ganglioside, the proposed membrane receptor structure for toxin, prevented entirely the cellular binding of both toxin and toxoid. Cholera toxin in concentrations down to approximately 5 x 10(-11) mol/liter (corresponding to 10 bound molecules/cell) inhibited thymus cells from being stimulated to DNA synthesis by concanavalin A (con A), and spleen cells from such stimulation by phytohemagglutinin. The G(M1) ganglioside but not a series of other pure structurally related gangliosides and neutral glycosphingolipids neutralized this toxin activity. Toxin derivatives which, in similarity with toxin, possessed H as well as L subunits but in other proportions, were potent inhibitors of con A-induced thymocyte stimulation, whereas the natural toxoid (aggregated L subunits), purified toxin L subunit and purified toxin H subunit were up to 300-fold less active on a weight basis. The capacity of cholera proteins to inhibit con A-induced thymocyte stimulation correlated well with their activity in the rabbit intradermal toxicity assay. The inhibitory action of cholera toxin on con A-induced thymocyte stimulation did not depend on decreased cell viability from the toxin treatment, nor was it caused by a reaction between toxin and con A. [(125)I]con A bound equally well to the cells when toxin was present as when it was absent, which proves that the toxin did not compete for cellular con A receptors. Nor did the toxin seem to disturb the general mobility of membrane receptors or their ability to accumulate in caps. It is concluded that the L type of subunit confers rapid and firm binding of cholera toxin to lymphocyte membranes, probably to G(M1) ganglioside receptors. For biologic activity the additional presence of H subunit is important. One manifestation of toxin action on lymphocytes is inhibition of lectin-induced DNA synthesis; probably this effect relates to the ability of cholera toxin to raise the levels of intracellular cyclic 3'5'-adenosine monophosphate.

Mucosally induced immunological tolerance, regulatory T cells and the adjuvant effect by cholera toxin B subunit.

Induction of peripheral immunological tolerance by mucosal administration of selected antigens (Ags) ('oral tolerance') is an attractive, yet medically little developed, approach to prevent or treat selected autoimmune or allergic disorders. A highly effective way to maximize oral tolerance induction for immunotherapeutic purposes is to administer the relevant Ag together with, and preferably linked to the non-toxic B subunit protein of cholera toxin (CTB). Oral, nasal or sublingual administration of such Ag/CTB conjugates or gene fusion proteins have been found to induce tolerance with superior efficiency compared with administration of Ag alone, including the suppression of various autoimmune disorders and allergies in animal models. In a proof-of-concept clinical trial in patients with Behcet's disease, this was extended with highly promising results to prevent relapse of autoimmune uveitis. Tolerization by mucosal Ag/CTB administration results in a strong increase in Ag-specific regulatory CD4(+) T cells, apparently via two separate pathways: one using B cells as APCs and leading to a strong expansion of Foxp3(+) Treg cells which can both suppress and mediate apoptotic depletion of effector T cells, and one being B cell-independent and associated with development of Foxp3(-) regulatory T cells that express membrane latency-associated peptide and transforming growth factor (TGF-beta) and/or IL-10. The ability of CTB to dramatically increase mucosal Ag uptake and presentation by different APCs through binding to GM1 ganglioside (which makes most B cells effective APCs irrespective of their Ag specificity), together with CTB-mediated stimulation of TGF-beta and IL-10 production and inhibition of IL-6 formation may explain the dramatic potentiation of oral tolerance by mucosal Ags presented with CTB.

Establishment of an adult mouse model for direct evaluation of the efficacy of vaccines against Vibrio cholerae.

We describe here a new animal model that offers the prospect of using conventional adult mice for direct evaluation of the protective potential of new cholera vaccines. Pretreatment of adult mice with oral streptomycin allowed intestinal colonization by streptomycin-resistant Vibrio cholerae strains of either the O1 or the O139 serogroup. Bacteria were recovered in greatest numbers from the cecum and large intestine, but recoveries from all regions of the gut correlated significantly with bacterial excretion in fresh fecal pellets, which thus provides a convenient indicator of the extent and duration of gut colonization. Mice immunized mucosally or systemically with viable or inactivated V. cholerae were shown to be comparatively refractory to colonization after challenge with the immunizing strain. Several variables were examined to optimize the model, the most significant being the size of the challenge inoculum; surprisingly, a smaller challenge dose resulted in more consistent and sustained colonization. Studies with mutant strains unable to produce cholera toxin or toxin-coregulated pili revealed that neither factor contributed significantly to colonization potential. Protection against V. cholerae challenge was shown to be serogroup restricted, and significant inverse correlations were detected between serum and intestinal anti-lipopolysaccharide antibody responses and the levels of excretion of challenge organisms.

Intestinal immune responses in humans. Oral cholera vaccination induces strong intestinal antibody responses and interferon-gamma production and evokes local immunological memory.

We have examined secretory antibody and cell-mediated immune responses to oral cholera vaccine in the human gastrointestinal mucosa. Freshly isolated peripheral blood lymphocytes and intestinal lymphocytes obtained by enzymatic dispersion of duodenal biopsies were assayed for numbers of total and vaccine specific immunoglobulin-secreting cells by enzyme-linked immunospot assay (ELISPOT) techniques; the frequency of cells secreting interferon-gamma (IFN-gamma) was also examined by a new modification of the ELISPOT technique. After booster immunizations with oral cholera vaccine, large numbers of cholera toxin-specific antibody-secreting cells (ASC) appeared in the small intestine. The responses were dominated by IgA ASC. A single immunization, performed 5 mo after the initial vaccinations, gave rise to an ASC response similar to that seen after the first booster immunization, with respect to both magnitude and isotype distribution. Each of the immunizations also evoked an ASC response in blood which was of lower magnitude than that seen in the small intestine, and comprised similar proportions of IgA and IgG ASC. A booster immunization also resulted in increased frequencies of IFN-gamma-secreting cells, but this increase was confined to the duodenal mucosa. This study establishes the feasibility of studying, at the single-cell level, intestinal immune reactivity in humans. Furthermore, it indicates that the small intestinal mucosa is an enriched source of IFN-gamma. It also demonstrates marked differences between intestinal and peripheral blood immune responses after enteric immunization, and confirms the notion that the mucosal immune system in humans displays immunological memory.

Differential expression of tissue-specific adhesion molecules on human circulating antibody-forming cells after systemic, enteric, and nasal immunizations. A molecular basis for the compartmentalization of effector B cell responses.

Expression of the adhesion molecules CD44, L-selectin (CD62L), and integrin alpha 4 beta 7 by antibody-secreting cells (ASC) was examined in human volunteers after oral, rectal, intranasal, or systemic immunization with cholera toxin B subunit. Almost all blood ASC, irrespective of immunization route, isotype (IgG and IgA), and immunogen, expressed CD44. On the other hand, marked differences were observed between systemically and intestinally induced ASC with respect to expression of integrin alpha 4 beta 7 and L-selectin, adhesion molecules conferring tissue specificity for mucosal tissues and peripheral lymph nodes, respectively. Thus, most ASC induced at systemic sites expressed L-selectin, whereas only a smaller proportion of ASC expressed alpha 4 beta 7. In contrast, virtually all IgA- and even IgG-ASC detected after peroral and rectal immunizations expressed alpha 4 beta 7, with only a minor fraction of these cells expressing L-selectin. Circulating ASC induced by intranasal immunization displayed a more promiscuous pattern of adhesion molecules, with a large majority of ASC coexpressing L-selectin and alpha 4 beta 7. These results demonstrate that circulating ASC induced by mucosal and systemic immunization express different sets of adhesion molecules. Furthermore, these findings provide for the first time evidence for differential expression of adhesion molecules on circulating ASC originating from different mucosal sites. Collectively, these results may explain the anatomical division of mucosal and systemic immune responses in humans as well as the compartmentalization of mucosal immune responses initiated in the upper vs. the lower aerodigestive tract.

Cholera as a model for research on mucosal immunity and development of oral vaccines.

During the past year, the extensive investigational use of a recently developed oral vaccine against cholera has led to significant advances in our understanding of both immunity to cholera and related diarrhoeal diseases, and the mucosal immune response in general after oral immunization. The oral cholera vaccine has been shown to protect, through its cholera toxin B subunit component, against travellers' diarrhoea caused by enterotoxigenic Escherichia coli. The elaboration of sensitive new techniques has allowed detailed clonal analyses of the activation of specific B and T cells and immunologic memory in intestinal mucosa in humans after oral cholera vaccination. These techniques have also been used to demonstrate a transient appearance after immunization of specific gut-derived IgA antibody-producing cells in the circulation and also, a few days later, in a distant mucosal tissue such as the salivary glands.

Detection of a ganglioside antigen associated with small cell lung carcinomas using monoclonal antibodies directed against fucosyl-GM1.

Monoclonal antibodies with an apparent specificity for fucosyl-GM1 (Fuc-GM1) were produced by the immunization of mice with Fuc-GM1 adsorbed to Salmonella minnesota bacteria and fusion of the spleen cells with the myeloma cell line Sp 2/0. The antibodies detected Fuc-GM1 with a unique ceramide composition containing 2-hydroxy fatty acids in 11 of 12 cases of small cell carcinoma of the lung. Trace amounts of Fuc-GM1 were detected in 1 of 11 squamous epithelial cell lung carcinomas. Fuc-GM1 was also detected in 1 of 7 pancreas carcinomas but was not detected in any of the other cancers analyzed. Small amounts of Fuc-GM1 without 2-hydroxy fatty acids were detected in normal adult pancreas, spleen, and brain but could not be detected in normal lung tissue. Fuc-GM1 with 2-hydroxy fatty acids is suggested to be a specific ganglioside associated with small cell lung carcinomas. The monoclonal antibodies directed against Fuc-GM1 may be useful for specific immunodiagnosis of small cell lung carcinomas and might also be useful for specific immunotherapy of these malignant tumors.

Immunohistological detection of fucosyl-GM1 ganglioside in human lung cancer and normal tissues with monoclonal antibodies.

With the aid of a highly specific murine monoclonal antibody, F12, an immunofluorescence method was elaborated that allowed sensitive and specific detection of the ganglioside antigen fucosyl-GM1 (IV2FucII3NeuAcGgOse4Cer) in different types of human lung cancer and normal tissues. Nineteen of 21 cases of small cell lung cancer were positive with the F12 immunofluorescence method as compared to 2 of 10 squamous epithelial cell lung cancers and 1 of 5 large cell lung cancer specimens. Specimens of lung adenocarcinoma (8 cases) and bronchial carcinoid (3 cases) were all negative, as were 2 examined cases of neuroblastoma. No fucosyl-GM1 could be detected in normal lung and bronchus. However, in thymus, spleen, and lamina propria of the small intestine sparsely distributed clusters of small round cells were stained as well as intramural ganglionic cells of the small intestine and islet cells of the pancreas. All other normal tissues tested were negative. Results obtained with immunofluorescence closely agreed with immunochemical determination of fucosyl-GM1 in lipid extracts of tissues. Our findings suggest that fucosyl-GM1 is strongly associated with small cell cancer of the lung and demonstrate that this tumor-associated antigen can be detected with high sensitivity and specificity with an immunofluorescence method based on the use of the F12 monoclonal antibody.

Monoclonal antibodies raised against NeuAc alpha 2-6neolactotetraosylceramide detect carcinoma-associated gangliosides.

Monoclonal antibodies wee obtained by the immunization of mice with 6'LM1 (IV6NeuAc-nLcOse4Cer) adsorbed to Salmonella minnesota. The monoclonal antibodies showed a specificity for gangliosides with a terminal NeuAc alpha 2-6Gal substitution, which was demonstrated in solid-phase binding assay and in liposome inhibition assay. Gangliosides with a NeuAc alpha 2-6Gal substitution were minor components of different normal tissues. However, these gangliosides were enriched in carcinomas of many tissues, and were particularly enriched in most colorectal carcinomas and in lung carcinomas. 6'LM1 is a characteristic ganglioside in fetal intestinal mucosa (meconium). This ganglioside and other gangliosides with a terminal NeuAc alpha 2-6Gal substitution might represent oncofetal antigens expressed in carcinomas owing to an activation of a "fetal' sialyltransferase.

Binding specificity of monoclonal antibodies to ganglioside, Fuc-GM1.

The binding specificity of thirteen mouse monoclonal antibodies reacting with Fuc-GM1, Fuc alpha 1-2Gal beta 1-3GalNAc beta 1-4(NeuAc alpha 2-3)-Gal beta 1-4Glc beta 1-1Cer, a ganglioside found to be associated with small cell lung carcinoma (O. Nilsson et al. (1984) Glycoconjugate J. 1, 43-49) was studied. The results are based upon radioimmunodetection of their binding to structurally related glycolipids adsorbed to microtiter plates or chromatographed on thin-layer plates. Four of thirteen antibodies reacted only with Fuc-GM1 and both the fucose and the sialic residues were necessary for binding. Optimal binding was obtained when the sialic acid was N-acetylneuraminic acid. When this sialic acid residue was substituted with N-glycoloylneuraminic acid the binding activity was reduced and up to 10-times more Fuc-GM1 was needed for detection. The ceramide composition did not influence the binding. The other nine monoclonal antibodies cross-reacted with glycolipids containing structures closely related to Fuc-GM1 and differed from the specific ones by recognizing a smaller portion of the carbohydrate moiety in Fuc-GM1. These results indicate that anticarbohydrate monoclonal antibodies, recognizing structures involving a large proportion of the sugar in the glycolipid, possess a high specificity and might be useful for detection of tumor-associated ganglioside antigen.

Chemical structure of carcinoma ganglioside antigens defined by monoclonal antibody C-50 and some allied gangliosides of human pancreatic adenocarcinoma.

A hybridoma, C-50, obtained by fusion of mouse myeloma cells with spleen cells from a mouse immunized with cells from the colorectal carcinoma cell line COLO 205, produced antibodies that detected ganglioside antigen in human adenocarcinomas in many organs. The major ganglioside antigen fraction isolated from liver metastases of a pancreatic adenocarcinoma, behaving as a homogenous band on thin-layer chromatography, consisted of three different gangliosides. One of them, A (25%), had the same carbohydrate structure as the ganglioside antigen defined by monoclonal antibody 19-9, NeuAc alpha 2-3Gal beta 1-3(Fuc alpha 1-4)GlcNAc beta 1-3Gal beta 1-4Glc-Cer(Fuc-3'-isoLM1) Magnani, J.L., Nilsson, B., Brockhaus, M., Zopf, D., Steplewski, Z., Koprowski, H. and Ginsburg, V. (1982) J. Biol. Chem. 257, 14365-14369). The major ganglioside, B (60%), was the isomeric hexasaccharide ganglioside (NeuAc alpha 2-3Gal beta 1-4(Fuc alpha 1-3)GlcNAc beta 1-3-Gal beta 1-4Glc-Cer(Fuc-3'-LM1) and the third ganglioside, C, was 6'-LM1, NeuAc alpha 2-6Gal beta 1-4GlcNAc beta 1-3Gal beta 1-4Glc-Cer (15%). Ganglioside B, isolated from human kidney, did not react with the C-50 MAb. Based on this result and on studies of COLO 205 cell induced tumours where the ganglioside antigen fraction only consisted of A, it is suggested that the C-50 MAb defines an antigen determinant present in A.

Oral administration of cholera toxin B-insulin conjugates protects NOD mice from autoimmune diabetes by inducing CD4+ regulatory T-cells.

Restoration of peripheral tolerance to target autoantigens during autoimmune diseases has met with several limitations because of the limited efficacy of this approach in an already immune host. To optimize the induction of tolerance, we have shown that feeding insulin conjugated to cholera toxin B-subunit (CTB), a potent mucosal adjuvant, reduced by 5,000 the amounts of antigen necessary for delaying diabetes onset in NOD mice. To analyze these protective mechanisms, we have performed cotransfer experiments using splenocytes from young females fed once with 10 microg of CTB-insulin, mixed with diabetogenic T-cells, and intravenously injected into irradiated syngeneic male recipients. We demonstrated that the delayed onset of diabetes relied on CD4+ T-cells. We studied the cytokine production from plate-bound anti-CD3-stimulated cells. Higher interleukin (IL)-4 amounts were observed in both splenocytes and pancreatic lymph node (PLN) cell cultures from CTB-insulin-fed mice as soon as 4 h after the feeding. An increase in the levels of transforming growth factor-beta was seen after 24 h only in the mesenteric lymph nodes (MLN). In both of these organs, a reduction of gamma-interferon (IFN-gamma) production occurred after CTB-insulin treatment, at 24 h in the PLN and at 7 days in the MLN. Reverse transcription-polymerase chain reaction analysis indicated an increase in the level of IL-4 and a reduction in IFN-gamma transcripts in the PLN of mice treated orally with CTB-insulin and of the recipients of regulatory T-cells. Using different strains of congenic NOD mice at the Thy1 locus, we showed that protection was associated with the accumulation of T-cells from CTB-insulin-fed mice in the lymph nodes from draining sites containing functional islets, i.e., the PLN in normal mice and the renal lymph nodes after a syngeneic islet graft under the kidney capsule of streptozotocin-treated mice. Taken together, our results clearly indicate that oral administration of CTB-insulin conjugates in NOD mice produced a shift from a T-helper type 1 to a type 2 profile with the induction of antigen-specific regulatory CD4+ T-cells in the vicinity of the mucosal barrier and close to the inflamed islets.

Intestinal antibody responses to oral vaccination in HIV-infected individuals.

OBJECTIVE: To examine whether and at what stage mucosal immune responsiveness is impaired during HIV-1 infection. DESIGN: Intestinal and systemic antibody-secreting cell (ASC) responses were examined in eight HIV-1-infected volunteers and 10 seronegative control subjects after oral cholera and parenteral tetanus vaccinations. METHODS: ASC numbers were determined before and after booster vaccinations by the enzyme-linked immunospot (ELISPOT) technique. This technique was performed on cell suspensions obtained from enzymatically dispersed duodenal biopsies and from peripheral blood. RESULTS: Oral cholera vaccination evoked ASC responses in the intestinal mucosa of six out of eight HIV-1-infected volunteers, including patients with advanced disease and very low levels of circulating CD4+ T cells. The intestinal cholera ASC responses in HIV-infected volunteers were comparable to those in uninfected controls with regard to both magnitude and distribution of antibody classes. Most HIV-infected volunteers with only moderately reduced CD4+ T-cell counts also responded with vaccine-specific ASC in the blood, whereas none of the patients with < 200 x 10(6)/l CD4+ T cells per litre blood had detectable circulating ASC. CONCLUSION: These findings indicate that mucosal humoral immune responsiveness to a T-cell dependent antigen is maintained in HIV-infected individuals, despite concomitant systemic humoral hyporesponsiveness.

Virus-specific antibody production and polyclonal B-cell activation in the intestinal mucosa of HIV-infected individuals.

OBJECTIVE: To examine possible changes in mucosal B-cell activation status. DESIGN: To examine the frequency and isotype distribution of total and HIV-specific antibody-secreting cells (ASC) in the intestinal mucosa of HIV-infected individuals. METHODS: Mucosal lymphocytes were obtained by enzymatic treatment of duodenal pinch biopsies and the numbers of ASC were assayed with the enzyme-linked immunospot technique. RESULTS: High numbers of HIV-specific ASC were found in the intestine of all HIV-infected individuals despite low levels of HIV-specific blood ASC. All HIV-infected individuals had large numbers of intestinal immunoglobulin (Ig) A-ASC against the HIV envelope glycoprotein gp160. Eight out of nine patients also had HIV gp160-specific intestinal IgG-ASC. These HIV-specific ASC were detected irrespective of disease stage, route of infection, or levels of circulating CD4+ T cells. HIV-specific ASC were found in peripheral blood from patients with CD4+ T cells > or = 100 x 10(6)/l blood, but in none of three patients with low CD4+ T-cell counts. The frequencies of virus-specific ASC in the blood were on average 100-fold lower than that observed within the intestinal mucosa. Mucosal polyclonal B-cell activation was evident in HIV-infected individuals, as documented by significantly elevated numbers of Ig-secreting cells (ISC) in all three major Ig classes; on average, seven-, five- and 20-fold numbers of IgA, IgG and IgM-ISC compared with healthy controls. Furthermore, substantial numbers of ASC reacting with unrelated antigens such as dog albumin and keyhole limpet haemocyanin were detected in HIV-infected patients. Interestingly, patients with CD4+ T cells < 100 x 10(6)/l blood displayed large numbers of HIV-specific intestinal ASC even though total numbers of ISC, including ASC reactive to unrelated antigens, were decreased. CONCLUSIONS: The large numbers of virus-specific ASC found in the intestine of HIV-infected individuals may be a consequence of local replication of HIV-1 resulting in a continuous antigen stimulation. The persistence of strong intestinal anti-HIV responses even at late stages of disease suggest that the mucosal B-cell responses are functionally intact throughout the disease. Furthermore, these results suggest that there is no correlation between HIV-specific ASC numbers and polyclonal B-cell activation. These observations indicate that intestinal B-cell activation is profoundly disregulated in HIV-infected individuals.

Characterization of an internal permissive site in the cholera toxin B-subunit and insertion of epitopes from human immunodeficiency virus-1, hepatitis B virus and enterotoxigenic Escherichia coli.

We previously described the construction of novel hybrid proteins based on the B-subunit of cholera toxin (CTB) [Bäckström et al., Gene 149 (1994) 211-217], in which a neutralizing B-cell epitope from the third variable (V3) loop in the envelope glycoprotein gp120 from human immunodeficiency virus type 1 (HIV-1) was inserted within a surface-exposed region between amino acids (aa) 55 and 64. The resulting protein retained properties of native CTB and could induce strong anti-CTB antibody (Ab) responses, but the inserted gp120 epitope was only modestly immunogenic. In this study, the potential use of this internal permissive site in CTB for the insertion of heterologous epitopes has been further investigated. Six additional plasmids were constructed encoding HIV::CTB hybrid proteins with ten to fourteen aa from the V3 loop of gp120 genetically inserted at different positions between aa 52 and 65, with deletions of different CTB aa. Plasmids encoding proteins with peptides inserted between aa 53 and 64 in CTB gave rise to stable proteins which reacted with CTB-specific monoclonal antibodies (mAb) and bound to GM1 gangliosides (GM1), indicating that insertions between these positions do not drastically alter the conformation or the receptor-binding properties of native CTB. Plasmids were also constructed encoding CTB hybrid proteins which had either an 11-aa peptide from hepatitis B virus (HBV) pre-S(2) or one of two peptides related to the heat-stable toxin (STa) of enterotoxigenic Escherichia coli inserted between aa 55 and 64 of CTB. This resulted in the production of HBV::CTB or ST::CTB hybrid proteins and illustrated that the internal permissive site can be used for insertion of peptides of varying aa composition. The reactivity of the inserted epitopes with epitope-specific mAb in GM1-ELISA and immunoblots varied greatly between hybrid proteins and depended on the position in CTB and the aa composition of the inserted peptides. Despite these differences, all the HIV::CTB, ST::CTB and HBV::CTB hybrid proteins could induce low, but significant, levels of serum Ab in mice against gp120, STa or pre-S(2), in addition to strong serum Ab responses against CTB. The Ab response against the internally inserted gp120 peptide was similar to that against the same peptide fused to the N-terminus of CTB, indicating that internally placed peptides had similar immunogenicity to the same peptides added terminally.

Insertion of a HIV-1-neutralizing epitope in a surface-exposed internal region of the cholera toxin B-subunit.

The non-toxic B-subunit of cholera toxin (CTB) is a powerful immunogen and has been investigated as a carrier for foreign peptide epitopes, with peptides genetically fused to either the N- or C terminus of CTB. In the present study, we have constructed a plasmid encoding a novel intrachain CTB fusion protein with a peptide epitope inserted into an internal region of CTB: eight amino acids (aa) in CTB (56-63) were substituted with a 10-aa peptide from the third variable (V3) loop of the HIV-1 envelope protein gp120. The resulting chimeric protein retained important functional characteristics of the native CTB including pentamerization and GM1 ganglioside receptor binding. The internal hybrid protein was also shown to be resistant to proteolytic degradation during production in Vibrio cholerae, whereas a terminal hybrid protein, where the same gp120-epitope was fused to the N terminus of CTB, was rapidly cleaved during culture. The inserted epitope, which is known to give rise to HIV-1 neutralizing Ab, could be detected with a V3 loop-specific monoclonal Ab when the chimeric protein was analyzed in ELISA and immunoblot, indicating that the epitope inserted at this site is presented on the surface of the protein. Consistent with these observations, immunization of mice with the CTB::HIV hybrid protein elicited a high titered serum Ab response to the CTB moiety and also, in some but not all animals, a detectable response to the inserted gp120 epitope.

Synthesis in Vibrio cholerae and secretion of hepatitis B virus antigens fused to Escherichia coli heat-labile enterotoxin subunit B.

A simple and effective electroporation method for the transformation of Vibrio cholerae with nonmobilizable plasmids is described. Expression plasmids directing the synthesis of fusion proteins with the subunit B of Escherichia coli heat-labile enterotoxin B (LT-B) were transformed into nontoxinogenic V. cholerae vaccine strains. A protein consisting of two overlapping immunodominant antibody-binding sites of the hepatitis B virus (HBV) middle surface antigen fused to the C terminus of full-length LT-B was secreted into the supernatant of V. cholerae cultures, whereas two other LT-B/HBV fusion proteins were mostly retained within the cells or rapidly degraded in the culture supernatant. While the secretion of fusion proteins with cholera toxin subunit B (CT-B) from V. cholerae has been described, this is to our knowledge the first report describing extracellular secretion of defined foreign epitopes fused to LT-B in V. cholerae. The fusion of guest epitopes to LT-B or CT-B and secretion in V. cholerae could be an interesting system to rapidly produce pure fusion proteins for immunisation, functional studies or diagnostic procedures. An LT-B/pre-S2 fusion protein purified from the supernatant of recombinant V. cholerae induced serum IgG antibodies against LT-B and against the HBV middle surface antigen in mice after parenteral and oral immunisation.

Immunological characterization of a rotavirus-neutralizing epitope fused to the cholera toxin B subunit.

A highly conserved neutralizing epitope from the surface protein VP4 (amino acids 296-313) of human rotaviruses was genetically fused to the B subunit of cholera toxin (CTB). Synthetic oligodeoxyribonucleotides encoding the VP4 peptide were inserted between the 3' end of the DNA that codes for the leader peptide, and the 5' end of the gene encoding mature CTB. The hybrid protein synthesized in Escherichia coli was found to maintain the ability of CTB to pentamerize, and to adhere to its cell receptor, the GM1 ganglioside. The chimera was efficiently recognized by a monoclonal antibody (mAb) directed at CTB and by a virus-neutralizing mAb against the VP4 peptide. The hybrid polypeptide was shown to induce high titers of serum antibodies (Ab) against CTB and the synthetic VP4 peptide following subcutaneous immunization; paradoxically, however, the Ab obtained did not recognize the virus by an enzyme-linked immunosorbent assay method, nor had detectable neutralizing activity. Potential implications of these results for future design and evaluation of fusion proteins as immunogens are discussed.