The JI
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ågren, L.
Right arrow Articles by Lycke, N.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ågren, L.
Right arrow Articles by Lycke, N.
The Journal of Immunology, 2000, 164: 6276-6286.
Copyright © 2000 by The American Association of Immunologists

The ADP-Ribosylating CTA1-DD Adjuvant Enhances T Cell-Dependent and Independent Responses by Direct Action on B Cells Involving Anti-Apoptotic Bcl-2- and Germinal Center-Promoting Effects1

Lena Ågren*, Eva Sverremark{ddagger}, Lena Ekman*, Karin Schön*, Björn Löwenadler{dagger}, Carmen Fernandez{ddagger} and Nils Lycke2,*

* Department of Medical Microbiology and Immunology, University of Göteborg, Göteborg, Sweden; {dagger} Department of Molecular Biology, Astra Hässle AB, Molndal, Sweden; and {ddagger} Department of Immunology, The Wenner-Gren Institute, Arrhenius Laboratories for Natural Sciences, Stockholm University, Stockholm, Sweden


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
We recently developed a novel immunomodulating gene fusion protein, CTA1-DD, that combines the ADP-ribosylating ability of cholera toxin (CT) with a dimer of an Ig-binding fragment, D, of Staphylococcus aureus protein A. The CTA1-DD adjuvant was found to be nontoxic and greatly augmented T cell-dependent responses to soluble protein Ags after systemic as well as mucosal immunizations. Here we show that CTA1-DD does not appear to form immune complexes or bind to soluble Ig following injections, but, rather, it binds directly to B cells of all isotypes, including naive IgD+ cells. No binding was observed to macrophages or dendritic cells. Immunizations in Fc{epsilon}R (common FcR{gamma}-chain)- and Fc{gamma}RII-deficient mice demonstrated that CTA1-DD exerted unaltered enhancing effects, indicating that Fc{gamma}R-expressing cells are not required for the adjuvant function. Whereas CT failed to augment Ab responses to high m.w. dextran B512 in athymic mice, CTA1-DD was highly efficient, demonstrating that T cell-independent responses were also enhanced by this adjuvant. In normal mice both CT and CTA1-DD, but not the enzymatically inactive CTA1-R7K-DD mutant, were efficient enhancers of T cell-dependent as well as T cell-independent responses, and both promoted germinal center formation following immunizations. Although CT augmented apoptosis in Ag receptor-activated B cells, CTA1-DD strongly counteracted apoptosis by inducing Bcl-2 in a dose-dependent manner, a mechanism that was independent of the CD19 coreceptor. However, in the presence of CD40 stimulation, apoptosis was low and unaffected by CT, suggesting that the adjuvant effect of CT is dependent on the presence of activated CD40 ligand-expressing T cells.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Most commonly used adjuvants activate the innate immune system and induce local inflammatory responses. For many of these compounds there is good evidence to suggest that the degree of inflammation directly relates to their adjuvant ability, and that their principal mode of action is to augment Ag processing and presentation (1, 2). At present only aluminum salts can be used clinically. This is despite the fact that many other adjuvants are strikingly more potent, but also more harmful to the host (2). Therefore, the potency of an adjuvant often conflicts with host safety and tolerability. One of the best studied and most effective adjuvants described is cholera toxin (CT)3 (3). The mechanisms for its adjuvant function are not completely understood, although augmenting effects on Ag presentation and direct effects on B cell differentiation explain at least some of the immunomodulating activity (3, 4). However, the strong toxic effects of CT preclude its wider clinical use in most vaccination protocols. It is the ADP-ribosyltransferase activity of the A1 subunit that is responsible for toxicity, and in humans only 5 µg may cause severe symptoms (5). One factor that contributes to its extensive toxicity is the promiscuous binding of CT via its B subunits to the GM1-ganglioside receptor, abundantly distributed on most mammalian cells (6, 7, 8).

We have recently shown that CT can be made nontoxic by targeting of the immunomodulating property of CT and restricting its ability to react with unwanted populations. A novel fusion protein was constructed that combined the enzymatic activity of CTA1 with a dimer of an Ig-binding fragment of Staphylococcus aureus protein A, DD (9, 10). Thus, the fusion protein, CTA1-DD, retained the enzymatically active CTA1 moiety, while the CTB subunits and the ability to bind to the GM1-ganglioside receptor were removed. Extensive investigations revealed that the CTA1-DD fusion protein hosted systemic and mucosal adjuvant function comparable to that of the intact CT molecule, but, in contrast to CT, CTA1-DD was completely nontoxic. The CTA1-DD, therefore, represents a promising breakthrough in vaccine adjuvant construction by demonstrating the feasibility of incorporating potent bacterial immunomodulators in gene fusion proteins, constructs that are both highly effective adjuvants and appear to be nontoxic.

Other groups have taken a different approach to separate toxicity from adjuvanticity in CT and the related Escherichia coli heat-labile toxin (LT), by introducing, by site-directed mutagenesis, single amino acid replacements in the enzymatically active cleft of the A1 subunit (11, 12, 13). These groups are now reporting successful nontoxic constructions with retained adjuvant function where the enzyme is partly or even completely inactive (14, 15, 16, 17). To accommodate these seemingly conflicting results, Guiliani et al. (18) recently suggested that adjuvanticity in the holotoxins may be composed of at least two functions; the enzymatic activity of the A1 subunit and structural or binding properties associated with the AB5 complex.

In agreement, we found that mutations that killed the enzyme in the CTA1-DD fusion protein also inhibited the adjuvant function in vivo, directly supporting the idea that the ADP-ribosyltransferase activity of CT and LT exerts adjuvant function in vivo (19). Furthermore, our fusion protein lost its adjuvant function when we introduced mutations that impaired the binding of CTA1-DD to Ig (19). The latter finding indicated that binding to Ig was important for the adjuvant effect. Whether binding of CTA1-DD to soluble, circulating Ig, membrane-associated Ig on B cells, or both was responsible for the adjuvant effect in vivo was not addressed. However, we have previously argued for a direct effect of CTA1-DD on B cells, but with incomplete experimental data to support that idea. Potentially, a critical element in the adjuvant function could be the binding to soluble Ig, the formation of immune complexes (IC), and the possible interaction with Fc receptors, constitutively expressed on most professional APC. From a strategic as well as a mechanistic point of view it is important to clarify whether the novel adjuvant CTA1-DD acts directly on B cells and to what extent its adjuvanticity is dependent or independent of T cells (9, 19).

We showed previously, using dextran B512 (Dx), that CT acted as an adjuvant for both T cell-independent (TI) and T cell-dependent (TD) responses to Dx in wild-type, but not athymic, mice (20, 21). In the present study we have compared the adjuvant effects of CTA1-DD and CT on TD as well as TI responses in wild-type and athymic mice (3). We asked to what extent the CTA1-DD adjuvant, if able to bind and interact with B cells in vivo, has a stronger enhancing effect on TI responses compared with CT, and, if so, whether we could find evidence for a differential effect on B cell functions that could explain why CTA1-DD, but not CT, is effective in athymic mice.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Animals

C57BL/6 athymic nu/nu mice were obtained from Bomholt Gård (Ry, Denmark), and C57BL/6 wild-type mice were obtained from Charles River (Uppsala, Sweden). Breeding pairs of Fc{epsilon}RI{gamma}-/- (common FcR{gamma} chain) and Fc{gamma}RII-/- mice were obtained from The Jackson Laboratory (Bar Harbor, ME), and CD19-deficient mice were provided by Werner Müller (Köln, Germany) (22, 23, 24). Mice of both sexes were used at the age of 8–12 wk and were maintained under pathogen-free conditions using ventilated microisolator cages and sterile workbenches at the Department of Medical Microbiology and Immunology, Göteborg University, and in the animal facilities at Stockholm University.

Ags and immunizations

Native Dx (TI Ag) with a m.w. of 5–40 x 106 was obtained from INC Pharmaceuticals (Cleveland, OH). A TD protein-Dx conjugate was obtained by conjugating Dx with a m.w. of 103 (3–5 glucose units) to the protein chicken serum albumin (CSA; Sigma, St. Louis, MO). Dx was conjugated to hydrazide-CSA via its terminal aldehyde group using reductive amination (25); it was provided by Dr. Christian Krog-Jensen (Stockholm University, Stockholm, Sweden). DNP-conjugated Dx (DNP-Dx) was prepared according to standard procedures. Keyhole limpet hemocyanin (KLH) was purchased from Calbiochem (San Diego, CA), and CT was obtained from List Biological Laboratories (Campbell, CA). The nontoxic adjuvant CTA1-DD and an enzymatically inactive mutant, CTA1-R7K-DD, were prepared as described in detail in previous publications (9, 19). Biotinylated CTA1-DD was prepared according to standard procedures. 125I radiolabeling of the CTA1-DD fusion protein and OVA was achieved by an oxidative method with chloramine-T (26), provided by Dr. Esbjörn Telemo, Göteborg University (Göteborg, Sweden).

Mice were immunized i.p. with 10 µg of native Dx or 100 µg of DNP-Dx or CSA-Dx in the presence or the absence of optimal doses of the putative adjuvants, 2 µg of CT or 20 µg of CTA1-DD or CTA1-R7K-DD (9). KLH was given i.p. at 5 µg/dose together with the CTA1-DD, CTA1-R7K-DD, or CT in the doses indicated above. Five to 10 mice/group were immunized twice at 10-day intervals, and the mice were bled in the lateral tail vein or by retro-orbital puncture under light ether anesthesia 10 days after the primary immunization and 6–8 days after the secondary immunization.

Detection of specific Abs

Analysis of anti-Dx-, anti-KLH-, or anti-DNP-specific Abs in serum was performed by ELISA as previously described (9, 27). Briefly, 96-well ELISA plates (Costar, Cambridge, MA) were coated with 10 µg/ml Dx T250 (Pharmacia, Uppsala, Sweden), 100 µg/ml of KLH (Calbiochem), or 10 µg/ml of TNP-BSA in PBS at 4°C overnight (29). Following washing and blocking with 0.1% BSA/PBS, sera in serial 2- or 3-fold dilutions were added to corresponding subwells and incubated at 4°C overnight. Bound Abs were detected by isotype-specific alkaline phosphatase-labeled goat anti-mouse IgM or IgG (Southern Biotechnology Associates, Birmingham, AL) or alkaline phosphatase-labeled rabbit anti-mouse total Ig Abs (Dako, Glostrup, Denmark) followed by p-nitro-phenyl phosphatase substrate (Sigma) in diethanolamine buffer. The enzymatic reaction was read at 405 nm using a Thermomax microplate reader (Molecular Devices, Sunnyvale, CA). The anti-KLH and anti-DNP titers were defined as the interpolated value giving rise to an absorbance of 0.4 above background on the linear slope of the titration curve. The mice were analyzed individually, and specific log10 titers were expressed as the mean ± SD of 5–10 mice per group. Dx-specific IgM Ab concentrations were calculated in micrograms per milliliter ± SD from a standard curve generated by incorporating serial dilutions of a mouse anti-Dx IgM mAb of known concentration in the assay (20).

Detection by immunoblotting of free and bound CTA1-DD in serum

C57BL/6 mice were injected i.p. with 500 µg of CTA1-DD, and serum was collected after 1 h. Serum samples in 2 or 4 µl were run on a 12% native Tris-acetate gel (Novex, San Diego, CA) together with purified native CTA1-DD (1 µg) or a sample of CTA1-DD (1 µg) that had been incubated at room temperature for 4 h with normal mouse serum diluted 1/2 as controls. The proteins were electrophoretically transferred to Hybond-C nitrocellulose membranes (Amersham, Aylesbury, U.K.) in 25 mM Tris and 192 mM glycine for 20 min using a Trans-Blot SD SemiDry Transfer Cell (Bio-Rad, Hercules, CA). After blocking with 1% BSA/PBS, the membrane was incubated with rabbit peroxidase-antiperoxidase (PAP) Ab at a 1/1000 dilution (Dako) in blocking solution to detect the presence of DD. Note that DD is a dimer of the synthetic analogue of S. aureus protein A, and the PAP-Ab will bind unspecifically to the DD fragment. The PAP-DD complex was then visualized using diaminobenzidene substrate and H2O2, and the reaction was allowed to proceed for 1 h at room temperature. Serum from uninjected mice gave no detectable bands in the immunoblot.

In vitro B cell cultures

Highly enriched B cells from nu/nu or CD19-deficient mice (after depletion of CD4+ and CD8+ cells with PanT Dynabeads; anti-Thy1.2 (Dynal, Oslo, Norway) according to the manufacturer’s instructions) at 100,000 cells/well in 96-well plates (Nunc, Roskilde, Denmark) were stimulated by anti-IgM at 10 or 2 µg/ml (Jackson ImmunoResearch Laboratories, West Grove, PA) or anti-CD40 at 10 or 2 µg/ml (FGK-45, a gift from Dr. Ton Rolink, Basel Institute for Immunology, Basel, Switzerland) and cultured in Iscove’s complete medium containing 10% FCS and additives for 24 h in the presence or the absence of 5-fold dilutions of CT ranging from 0.1 to 0.016 µg/ml or CTA1-DD in concentrations ranging from 10 to 0.08 µg/ml, as indicated. Following culture, cells were washed in PBS and analyzed for apoptosis or intracellular Bcl-2 expression by FACS, as described in detail below.

FACS analysis

The ability of native CTA1-DD to bind to freshly isolated B cells ex vivo or after injection in vivo was assessed by FACS. Two hours before analysis mice were injected i.v. with PBS or 20 µg of CTA1-DD fusion protein. Splenic B cells from PBS injected mice were isolated and incubated with unconjugated or biotin-conjugated CTA1-DD at 4°C on ice for 30 min. Thereafter, the appropriate PE-conjugated streptavidin (PharMingen, San Diego, CA) or FITC-conjugated chicken anti-DD FITC (Immunsystem, Uppsala, Sweden) Abs were added to the cells together with FITC- or PE-conjugated anti-mouse CD19, CD4, IgD, IgM (PharMingen), IgG (Southern Biotechnology), or isotype controls (PharMingen) for these Abs directed against unrelated human lymphocyte markers. The cells were allowed to incubate at 4°C on ice for an additional 30 min. After careful washing in buffer the cells were analyzed for binding of the CTA1-DD fusion protein, and live gates were set on lymphocytes by forward and side scatter in combination with specific gates on subsets of cells identified in FL-1 or FL-2 as FITC- or PE-labeled CD19+, CD4+, IgD+, IgM+, or IgG+ cells and analyzed for mean fluorescence intensity using a FACScan (Becton Dickinson, San Jose, CA) (9). The anti-DD FITC Ab (Immunsystem) did not label B cells that had not been exposed to CTA1-DD, nor did the isotype control Abs unspecifically bind to cells that already had bound CTA1-DD.

Cultured B cells were analyzed for apoptosis by FACS (Becton Dickinson) using the In Situ Cell Death Detection Kit/TUNEL technique (Roche, Indianapolis, IN) as described by the manufacturer. For these investigations, cells were fixed in 4% paraformaldehyde for 30 min at room temperature, washed, and permeabilized (0.1% Triton X-100) for 2 min on ice. After washing, the cells were colabeled with anti-CD19-PE Ab (PharMingen) and incubated with the TUNEL reaction mixture for 1 h at 37°C and analyzed for the level of FITC-dUTP in CD19+-gated cells. Similarly, cultured B cells were analyzed for intracellular expression of Bcl-2 using specific hamster anti-bcl-2 Ab (PharMingen) followed by FITC-conjugated anti-hamster Ab and anti-CD19-PE Ab (PharMingen). For Bcl-2 detection cells were incubated in buffer containing 0.1% saponin according to the manufacturer’s instructions.

Immunohistochemistry

Wild-type mice were given a single i.p injection of DNP-Dx or native Dx with or without CT, CTA1-DD, or CTA1-R7K-DD mutant as indicated. Spleens were removed 10–12 days after the primary immunization or 7 days after the booster immunization and immediately frozen in liquid nitrogen and stored at -70°C. Spleens were embedded in Tissue-Tek OCT compound (Miles, Elkhart, IN) and frozen sections (5 µm thick) were prepared on microslides using a Zeiss cryostat (Zeiss, Cambridge, U.K.) and frozen at -70°C. Cryostat sections were fixed in 50% acetone for 30 s followed by 100% acetone for 5 min and then air dried. Subsequently, slides were rinsed with PBS and blocked with normal horse serum (1/20) for 15 min. Cryosections were double-labeled with Texas Red-conjugated anti-IgM (Southern Biotechnology Associates) and FITC-labeled peanut (Arachis hypogaea) hemagglutinin (PNA-FITC; Sigma) to detect germinal center (GC) formation. For detection of Ag-specific GC we incubated the sections with FITC-conjugated Dx to allow for enumeration of Dx-specific GC by counting visual fields as previously described (20).

Analysis of the distribution and binding of CTA1-DD to splenic cells following an i.v. injection of 20 µg of biotin-conjugated CTA1-DD were performed by double labeling with Abs specific for B cells, macrophages, dendritic cells, or CD3+ T cells. Frozen sections from two time points, at 2 and 6 h after injection, were fixed as described above and incubated with HRP-conjugated avidin-biotin complex (Dako) followed by AEC substrate (Sigma). For detection of CTA1-DD binding in vivo to different cellular subsets at 6 h postinjection we used streptavidin-Texas Red conjugate (Sera-Lab, Crawley, U.K.) to visualize the fusion protein and FITC-conjugated anti-B220, anti-CD11b (Mac-1), or rat anti-mouse CD3 molecular complex (PharMingen; anti-CD3 has catalogue no. 28004D) to visualize specific cellular subsets. For detection of biotin-conjugated CTA1-DD and dendritic cells, sections were incubated with alkaline phosphatase-conjugated avidin-biotin complex (ABC complex; Dako) followed by 5-bromo-4-chloro-3-indoyl-phosphate substrate (Dako) and thereafter biotin-conjugated anti-CD11c Abs (N418; Serotec, Oxford, U.K.), followed by HRP-conjugated avidin-biotin complex (ABC complex; Dako) and 3-amino-9-ethylcarbazole (AEC) substrate (Sigma). The sections were evaluated and photographed using DAS Mikroscop, Leica DMLD (Leica Mikroscope Systems, Welzar, Germany).

Tissue distribution of [125I]CTA1-DD

Evaluation of the tissue distribution of CTA1-DD and the tendency to form IC were investigated after an i.v. injection of 200 µl of [125I]CTA1-DD or a similarly labeled control soluble protein [125I]OVA, both corresponding to a total radioactivity of 160 x 106 cpm. After 15, 60, 120, and 240 min mice were killed, and serum, PBL, spleen, kidney, and liver were removed, weighed, and assayed for 125I radioactivity by gamma counting (1282 Compu{gamma}, LKB, Wallac Oy, Turku, Finland). The distribution of labeled protein was calculated as the percent radioactivity in a specific organ per total radioactivity in all removed organs including serum: organ-specific cpm - background cpm/sum of cpm from all tissues - background cpm x 100. Organs from different animals did not differ more than 5% in weight. The results are given as a percentage and are the mean ± SD of each group and three independent experiments.

Statistical analysis

We used the Student’s t test for independent samples for analysis of significance.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
The CTA1-DD adjuvant binds to B cells of many isotypes, including naive IgD+ cells.

Contrary to CT, the novel adjuvant, CTA1-DD failed to cause inflammation when injected into the footpad of a mouse, documenting its nontoxic nature (10). In the absence of overt inflammation and because adjuvanticity of most compounds is positively correlated to an ability to activate innate immunity, we predicted that a prerequisite for an immunoenhancing effect of CTA1-DD would require that the molecule can bind and directly affect APCs in vivo. Especially binding to naive B cells could be important for adjuvanticity, because CTA1-DD was designed to bind to Ig (9). Splenic B cells were isolated and double labeled with isotype-specific Abs and biotin-conjugated CTA1-DD or CTA1-DD followed by FITC-conjugated chicken anti-DD. We found that CTA1-DD bound to the membrane of IgM+, IgG+ as well as IgD+ cells, demonstrating that CTA1-DD could bind both naive and Ag-experienced B cells (Fig. 1Go). Isotype control Abs or anti-CD4 Abs did not label cells that bound CTA1-DD, indicating that the fusion protein did not unspecifically bind to cells other than Ig-positive cells, nor did it bind to Abs bound to non-Ig cells, such as CD4+ T cells (Fig. 1Go). In addition we observed no binding of CTA1-DD to macrophages (Mac-1+) or dendritic cells (CD11c+) (9). Following i.v. injection with biotin-labeled CTA1-DD we isolated IgD+ cells that clearly had bound CTA1-DD, confirming the binding also to naive B cells in vivo (Fig. 1Go). Moreover, immunohistochemical analysis of frozen sections of the spleen revealed that biotin-labeled CTA1-DD accumulated 2 h after injection in the periphery of the B cell follicles (Fig. 2Go, upper left). At 6 h postinjection CTA1-DD had concentrated to the central regions of the B cell follicle (Fig. 2Go, upper right). CTA1-DD was not found in the T cell areas (Fig. 2Go, lower right) or associated with macrophages (Fig. 2Go, middle right) or dendritic cells (Fig. 2Go, middle left) after colabeling of the sections with Abs specific for these subsets.



View larger version (33K):
[in this window]
[in a new window]
  		FIGURE 1. CTA1-DD adjuvant specifically binds naive and secondary B cells ex vivo and in vivo. To investigate whether the adjuvant bound directly to B cells ex vivo or in vivo, we performed the following analyses by FACS. Splenic B cells from C57BL/6 mice were isolated and double labeled with CTA1-DD followed by anti-IgG, anti-IgM, anti-IgD, or CD4-specific Abs. Isotype and specificity controls were also included. A live gate was set on lymphocytes by forward and side scatters and on the labeled population by gating in FL-1 or FL-2. The labeled cells were then analyzed for the degree of binding of CTA1-DD (filled profiles) using the direct method with biotinylated CTA1-DD followed by streptavidin-PE or indirectly by FITC-conjugated anti-DD Abs, as indicated. The FACS profiles show binding of CTA1-DD to IgG+ cells but no binding to CD4+ cells, which were simultaneously analyzed (upper left). The IgM+ cells bound CTA1-DD, but cells incubated with isotype control Abs did not label with CTA1-DD (upper right). The naive IgD+ cells also bound CTA1-DD, whereas IgD+ cells incubated without CTA1-DD but with anti-DD FITC (negative control) showed poor fluorescence (lower left). Splenic IgD+ cells were isolated 2 h after an i.v. injection with 20 µg of biotinylated CTA1-DD or PBS and analyzed for bound CTA1-DD. All experiments included isotype control Abs for all subsets tested as well as anti-CD3-, anti-CD11c-, or anti-Mac-1-specific Abs, and on no occasion did we see binding of CTA1-DD to these controls. These are representative histograms of 10 experiments.

 


View larger version (106K):
[in this window]
[in a new window]
  		FIGURE 2. Following i.v. injection CTA1-DD accumulates in the B cell follicle in the spleen, but appears not to be associated with T cells, macrophages, or dendritic cells. Biotin-conjugated CTA1-DD was injected i.v. in C57BL/6 mice. At 2 and 6 h following injection mice were sacrificed, and the spleens were snap-frozen and cryosectioned. Tissue sections were then incubated with avidin-HRP or avidin-Texas Red, as indicated, with or without colabeling with subset-specific biotin- or FITC-conjugated Abs against Mac-1, CD11c, CD3, or B220. The panels show CTA1-DD (dark brown) localized to the periphery of the B cell follicle at 2 h (left) and more centrally at 6 h (upper panels). Dendritic cells labeled with CD11c (red) did not colocalize with the CTA1-DD fusion protein (purple; middle left panel). Macrophages labeled with Mac-1 (green) did not colocalize with CTA1-DD (red; middle right panel). B cells labeled with anti-B220 colocalized with CTA1-DD (orange/red; bottom left panel). CTA1-DD (red) was not found in CD3+ T cell areas (green; bottom right panel). Two mice were analyzed at a time, and the experiment was repeated with the same result on three separate occasions.

 
The adjuvant effect of CTA1-DD is independent of FcR expression

Because CTA1-DD has a strong propensity to bind to Ig and may form IC in vivo we could not exclude that an important mechanism for adjuvanticity could be through interactions with FcR-carrying cells (28). To investigate the degree of IC formation after i.v. injection we radioactively labeled CTA1-DD with 125I and followed the accumulation of radioactivity to various tissues following i.v. injection in wild-type mice. As illustrated in Table IGo, we found that radioactivity did not accumulate in liver or kidney as would have been expected if significant IC formation had occurred (29, 30). In fact, in three independent experiments the distribution of radioactivity associated with CTA1-DD largely followed a similar excretion pattern as that observed for 125I-labeled OVA (Table IGo). However, confirming the immunohistochemical finding we observed a significant (p < 0.05) accumulation of radioactivity to the spleen over time with 125I-labeled CTA1-DD, but not with OVA.


View this table:
[in this window]
[in a new window]
  		Table I. Kinetic analysis of the distribution of [125I]CTA1-DD following i.v. injection1

 
Although IC did not appear to be formed after injection, the CTA1-DD could still bind to circulating Ig and potentially interact with FcR-carrying cells such as macrophages and dendritic cells. Therefore, we undertook studies in Fc{epsilon}R-deficient (common FcR{gamma}-chain) and Fc{gamma}RII-deficient mice to compare the immunoenhancing effect of CTA1-DD with that observed in wild-type mice (22, 23). Following i.v. immunizations with KLH we observed no reduction of the immunoenhancing effect of CTA1-DD adjuvant in Fc{epsilon}R- or Fc{gamma}RII-deficient mice compared with that seen in wild-type mice (Fig. 3Go). The enhancing effect was also similar to that observed with CT in these mice (Fig. 3Go). We consistently found that anti-KLH titers were enhanced to an equal degree in wild-type and FcR-deficient mice compared with controls immunized with KLH alone (p < 0.05). Of note, these mice have previously been found to completely lack Fc{gamma}R-dependent functions, such as phagocytosis and ADCC reactions (23). Here we observed uninhibited adjuvant function of CTA1-DD in Fc{gamma}R-deficient mice. As expected, Fc{gamma}RII-deficient mice exhibited increased anti-KLH Ab responses compared with wild-type mice, but also in these mice the adjuvanticity of CTA1-DD was intact and comparable to that of CT (22). Thus, these results clearly demonstrated that the adjuvant function of CTA1-DD did not rely on Fc{gamma}R-expressing cells. Of note, control mice immunized with CT or CTA1-DD alone demonstrated no KLH-specific titers, indicating that the adjuvant effect was Ag-specific and -dependent (not shown).



View larger version (23K):
[in this window]
[in a new window]
  		FIGURE 3. The adjuvant effect of CTA1-DD does not require Fc{gamma}R-expressing cells. To rule out that CTA1-DD acted through Fc{gamma}R-expressing cells we compared the immune responses to KLH with or without adjuvant in wild-type, Fc{epsilon}R-deficient (devoid of the common FcR{gamma}-chain), and Fc{gamma}RII-deficient (deficient in the down-regulatory Fc{gamma}RIIb1 on B cells) mice. The animals were injected twice i.p with KLH together with PBS ({square}), CTA1-DD (20 µg/dose; {blacksquare}), or CT (2 µg/dose; ). Serum was collected from individual mice 8 days following the booster immunization and analyzed by ELISA for the presence of anti-KLH Abs. The values represent total Ig anti-KLH log10 titers ± SD from 10 mice in each group. Three identical experiments gave similar results. The enhancement of specific serum anti-KLH titers was significant (*, p < 0.05) in all three mouse strains, but no statistical difference could be found between the CT and CTA1-DD adjuvant groups. The unaltered and significant enhancing effect of CTA1-DD in Fc{epsilon}R-/- mice proves that the adjuvant effect was not dependent on Fc{gamma}R-expressing dendritic cells or macrophages.

 
To finally exclude that injected CTA1-DD preferentially bound to soluble Ig we undertook a careful analysis of serum at 1 h after an i.p. injection of the fusion protein. The serum was diluted in buffer and analyzed for the distribution of CTA1-DD of different molecular sizes using a nonreducing, native gel. After immunoblotting we found that most of the detectable material had not associated with Ig or other proteins that formed larger size molecules (Fig. 4Go). A distinct band corresponding to 37 kDa, equivalent in size to the purified CTA1-DD fusion protein, was clearly detectable and little was trapped as IC or large size protein complexes (Fig. 4Go).



View larger version (81K):
[in this window]
[in a new window]
  		FIGURE 4. Injected CTA1-DD circulates in serum in soluble nonassociated form. Mice injected i.p with CTA1-DD fusion protein were analyzed for the presence of CTA1-DD in free or complexed form. Serum, collected 1h after injection, was run on a 12% native Novex gel and transferred to a nitrocellulose membrane. Immunoblotting was performed using the PAP-Ab system to detect DD in its native form (37-kDa CTA1-DD) or as part of larger molecules, indicating IC formation or Ig association. Lane 1, Purified native CTA1-DD (1 µg); lane 2, ex vivo control with CTA1-DD (1 µg) incubated for 4 h with normal mouse serum at 1/1; lane 3, 2 µl of serum collected at 1 h following injection i.p with 500 µg of CTA1-DD; lane 4, 4 µl of the same serum from CTA1-DD-injected mice. The molecular mass markers (low range molecular weight standards) are shown in kilodaltons. This is a representative experiment of two with identical results. Three mice were analyzed in each experiment. Serum from uninjected mice failed to show bands detectable by PAP immunoblotting.

 
Taken together, we found that a majority of the i.p injected CTA1-DD did not form IC or associate with Ig or other serum proteins, it was circulated and excreted similar to an unrelated soluble protein (OVA) with the exception that it was also found to accumulate in B cell follicles of the spleen over time. Based on these results it appeared that CTA1-DD could bind and act directly on naive B cells and, thus, also might be effective in augmenting TI responses.

ADP-ribosyltransferase-dependent effects of CTA1-DD-adjuvant on Ab responses and germinal center reactions to both TD and TI-2 Ags

Next, we investigated whether CTA1-DD could augment TI-type responses. The hapten DNP was conjugated to native Dx to obtain a TI-2 type system (31, 32). Dx is a simple, linear molecule of high m.w. (5–40 x 106 m.w.) consisting of glucose subunits. Like many carbohydrates it is a weak immunogen and stimulates no immunological memory (32, 33, 34). Following a priming immunization i.v. with DNP-Dx with or without admixed adjuvant, serum anti-DNP responses were determined by ELISA. We found that serum anti-DNP responses were more strongly augmented in the presence of CTA1-DD than in the presence of CT (Fig. 5Go). Furthermore, when the CTA1-DD, was replaced by an enzymatically inactive mutant, CTA1-R7K-DD, no enhancement was observed, suggesting that the augmenting effect of CTA1-DD was critically dependent on its ADP-ribosyltransferase activity. This result was consistent with our previous findings with TD responses and the CTA1-R7K-DD mutant (19). Furthermore, serum taken from experiments reported in Fig. 3Go were all negative for specific titers against DNP, indicating that the adjuvant effect was Ag specific and dependent (not shown).



View larger version (11K):
[in this window]
[in a new window]
  		FIGURE 5. Enzymatically active CTA1-DD adjuvant enhances TI responses. Native Dx was labeled with DNP and used to immunize wild-type mice in the presence or the absence of CT (2 µg), CTA1-DD (20 µg), or the enzymatically inactive CTA1-R7K-DD (20 µg) mutant. Ten days following a primary immunization anti-DNP serum Ab levels were determined by ELISA. The total Ig anti-DNP log10 titers ± SD from two experiments with 10 mice/group in each experiment are given, and the CT and CTA1-DD groups were significantly enhanced (*, p < 0.05) compared with PBS or the inactive mutant CTA1R7K-DD. Control mice injected with CTA1-DD or CT alone demonstrated no anti-DNP titers.

 
The formation of GC in response to Ag is generally associated with TD-type responses (35). However, a few TI Ags, such as Dx, can under certain circumstances promote the formation of specific GC reactions (36). The effect of the adjuvant molecules, CTA1-DD and CT, on GC formation was, therefore, investigated. Spleens were taken out at 10–12 days following the immunizations, cryosectioned and labeled with the FITC-conjugated marker, PNA, which detects B cells in the GC (37). Interestingly, both adjuvants, CT as well as CTA1-DD, promoted strong GC formation in wild-type mice (Fig. 6Go). By contrast, DNP-Dx alone or given together with the enzymatically inactive CTA1-R7K-DD mutant largely failed to stimulate GC reactions (Fig. 6Go). Unimmunized mice exhibited no GC reactions in their spleens. Thus, the CTA1-DD mimicked CT in its ability to stimulate GC formation in normal mice following immunizations with the TI Ag, Dx. The effect on GC formation appeared to result from an ADP-ribosyltransferase-dependent reaction, because both CT and CTA1-DD, but not the enzymatically inactive CTA1-R7K-DD mutant, stimulated GC development (Fig. 6Go).



View larger version (73K):
[in this window]
[in a new window]
  		FIGURE 6. CTA1-DD and CT greatly augment germinal center formation following immunization. Wild-type mice were given a single i.p injection of DNP-Dx with or without CT (2 µg), CTA1-DD (20 µg), or the enzymatically inactive CTA1-R7K-DD (20 µg). On days 12–13 following the injection cryosections of the spleens were prepared and double labeled with Texas Red-conjugated anti-IgM (red) and FITC-labeled PNA to detect GC (green). Light level micrographs consistently demonstrated the presence of large numbers of GC in mice treated with CT or CTA1-DD, while CTA1-R7K-DD or PBS failed to support GC development. IgM+ cells in GC are double labeled (yellow), while IgM+ cells outside the GC are red. This is one representative experiment of six performed on different occasions.

 
The augmenting effect on TI Ab levels in serum seemed to correlate well with a strong stimulating ability on GC formation in the spleen. Therefore, we asked whether such an effect of CTA1-DD and CT was restricted to TI responses or could also be found after immunizations with TD Ags. Immunogenicity of Dx decreases with decreasing molecular mass, and below 105 kDa Dx is no longer immunogenic. However, after conjugation of low m.w. Dx to protein, a TD-type response can be stimulated. We conjugated CSA to low m.w. Dx, to obtain a highly TD Ag, CSA-Dx (20). As illustrated in Table IIGo, both CTA1-DD and CT significantly promoted the development of GC reactions to CSA-Dx. Again, this appeared to be dependent on the ADP-ribosyltransferase activity, because the mutant CTA1-R7K-DD failed to show this effect (Table IIGo). Moreover, FITC-conjugated Dx bound to the GC, indicating that the GC reactions were Dx specific and a result of immunizations using enzymatically active adjuvant molecules (Table IIGo). Control mice injected with CT alone demonstrated no Dx-specific GC (not shown). These findings suggested that the adjuvant effect of CTA1-DD and CT involved an augmenting effect on the GC reaction.


View this table:
[in this window]
[in a new window]
  		Table II. Comparison of CT and CTA1-DD as adjuvants for the TD form of Dx, CSA-Dx

 
The B cell-targeted CTA1-DD adjuvant, but not CT, enhances TI-2 responses in athymic mice

GC reactions are thought to require the participation of CD4+ T cells (38). In support of this, athymic nude mice, devoid of TCR-{alpha}ß+ CD4+ T cells, largely fail to exhibit significant GC reactions following immunizations (39, 40). However, athymic mice respond well to immunizations with TI Ags, and in the next experiment we investigated whether CTA1-DD still acted as an adjuvant when GC reactions are impaired. Athymic, nu/nu mice on the C57BL/6 background were immunized with native Dx (TI Ag) in the presence or the absence of CT or CTA1-DD adjuvants. CTA1-DD strongly augmented anti-Dx serum Ab levels (p < 0.01) following immunizations with native Dx, whereas, by contrast, CT failed to augment Dx responses in nu/nu mice (p < 0.05), as evidenced in three separate experiments (Fig. 7Go). A mean 5-fold increase in the IgM titers against Dx was seen in CTA1-DD adjuvant-treated mice, whereas CT-treated mice exhibited titers not significantly different from nonadjuvant PBS-treated mice (Fig. 7Go). Cryosections demonstrated increased numbers of Dx-specific cells in the spleens of CTA1-DD-treated mice, whereas the frequency of such cells was unaltered in CT-treated compared with that in control mice (not shown). The majority of spleens from nu/nu mice were devoid of GC reactions, indicating that CTA1-DD’s adjuvant effect was independent of the development of GC and the presence of CD4+ T cells. By contrast, CT appeared to require the presence of CD4+ T cells for an adjuvant effect on Dx-specific responses.



View larger version (7K):
[in this window]
[in a new window]
  		FIGURE 7. CTA1-DD, but not CT, is an effective adjuvant in athymic nude mice. Athymic C57BL/6 nu/nu mice were primed and boosted i.p. with 10 µg of native Dx with or without CT (2 µg) or CTA1-DD (20 µg). The mice were bled at 7 days after the second immunization, serum samples were analyzed for specific anti-Dx IgM Abs by ELISA, and the results were expressed as micrograms per ml ± SD, as calculated from a standard curve generated with an anti-Dx IgM mAb of known concentration. This is one representative experiment of three performed with five to seven mice per group in each experiment. The enhancing effect of CTA1-DD was significant (*, p < 0.05) in all three experiments, while CT failed to augment the response to Dx. The mean enhancing effect of CTA1-DD was 4.9 ± 0.3 in the three experiments.

 
CTA1-DD and CT differentially affect apoptosis and Bcl-2 levels in activated B cells

Having found that CTA1-DD, but not CT, could enhance TI responses in nu/nu mice in the absence of significant GC reactions we asked whether CTA1-DD and CT affected B cell expansion and survival differently. To this end we enriched splenic B cells from athymic mice and stimulated the cells with anti-IgM (TI-type activation) or anti-CD40 (TD-type activation) in culture in the presence or the absence of CTA1-DD or CT. Using the TUNEL technique we found that CTA1-DD effectively prevented apoptosis in anti-IgM-activated B cells, whereas CT, by contrast, increased apoptosis in such cultures (Fig. 8Go). The effect was dose dependent, and <20% of the cells were labeled in the high concentration (10 µg/ml) of CTA1-DD, while the effect was reduced to 33, 42, and 52%, respectively, as 5-fold dilutions were tested (mean values of two experiments; not shown). We observed no similar effect of CT in any concentration tested, rather the proapoptotic effect was reduced from almost 90% labeled cells at 2 µg/ml to 50% at the lowest concentration tested (0.016 µg/ml). The effect of a lower concentration of anti-IgM (2 µg/ml) did not influence the anti-apoptotic pattern of CTA1-DD, nor did it change the pro-apoptotic pattern of CT in these cultures. Moreover, in the anti-CD40-treated B cells, neither CT nor CTA1-DD affected the level of apoptosis (Fig. 8Go). These results demonstrated that CTA1-DD counteracted apoptosis in B cells activated by TI-type Ags. CT, on the other hand, increased apoptosis after anti-IgM activation, but had no negative effect on B cell survival in anti-CD40-driven TD-type cultures. Furthermore, the anti-apoptotic effect of CTA1-DD in TI cultures was associated with an increased production of Bcl-2 in B cells (Fig. 9Go). CT did not affect intracellular Bcl-2 levels in activated or resting B cells. Because Bcl-2 induction was recently shown to depend on the CD19 coreceptor (41), we also undertook these experiments with B cells from CD19-deficient mice. However we observed no difference between B cells from the CD19-deficient or wild-type mice with regard to the Bcl-2-inducing ability by CTA1-DD, indicating that the effect was not mediated through the CD19 coreceptor (not shown). The Bcl-2-inducing effect was dose dependent, and it correlated well with the reduction in TUNEL+ cells mentioned above. By contrast, CT had no effect on Bcl-2 levels in any concentration, nor did it have an effect with anti-IgM or anti-CD40 in any other combination that we tested (not shown).



View larger version (23K):
[in this window]
[in a new window]
  		FIGURE 8. CTA1-DD, but not intact CT, prevents apoptosis in Ag-receptor-activated B cells. The effect of Ig-receptor cross-linking on apoptosis in B cell cultures in the presence or the absence of CT or CTA1-DD was investigated. Highly enriched splenic B cells from nu/nu mice were stimulated by two concentrations of anti-IgM or anti-CD40 (FGK-45; 10 and 2 µg/ml) (dotted lines) and cultured for 24 h together with different concentrations of CTA1-DD or CT, ranging from 0.08 to 10 and from 0.016 to 2 µg/ml, respectively (solid lines). Cells were gated by live gate set on CD19+ lymphocytes (FL-2) and were analyzed using the TUNEL technique for the distribution of cells undergoing apoptosis (FL-1). Here we depict the results from cultures with CT at 0.1 µg/ml and CTA1-DD at 10 µg/ml. We consistently found that CTA1-DD, in a dose-dependent manner, prevented apoptosis in anti-IgM (10 µg/ml)-stimulated cells. By contrast, CT dramatically aggravated apoptosis in these cultures. At no concentration did we find an anti-apoptotic effect of CT. On the contrary, in the presence of anti-CD40 (10 µg/ml), CT as well as CTA1-DD did not affect the level of apoptotic cells in culture. This is one representative experiment of three with similar results. Cell cultures stimulated with 2 µg/ml of anti-IgM or CD40 Abs showed the same patterns, albeit at a lower activation level, as the higher concentrations depicted in the figure.

 


View larger version (23K):
[in this window]
[in a new window]
  		FIGURE 9. The CTA1-DD adjuvant prevents apoptosis by strongly augmenting intracellular Bcl-2 levels in Ag-receptor-activated B cells. We correlated the anti-apoptotic effect of CTA1-DD to the presence of intracellular Bcl-2 protein in anti-IgM- or anti-CD40-activated B cells. Highly enriched B cells from nu/nu mice were stimulated by anti-IgM or anti-CD40 (10 µg/ml) for 24 h in the presence or the absence of CT (0.1 µg/ml) or CTA1-DD (10 µg/ml). The cells were labeled with anti-CD19 PE and anti-Bcl-2 FITC, gated for CD19+ cells (FL-2), and analyzed for Bcl-2 levels in FL-1 by FACS. Similar to the TUNEL analysis, the effect of CTA1-DD on Bcl-2 expression was dose dependent. CT failed to affect Bcl-2 levels in cultured cells at all concentrations tested (5-fold dilutions from 0.016 to 2 µg/ml). This is one representative experiment of four performed with similar results.

 
Taken together these results suggest that CTA1-DD directly interacts with activated B cells to prevent apoptosis, and this effect may allow for enhanced Ab responses to TI Ags in vivo in the absence of T cells (nu/nu mice). CTA1-DD added to the culture at 5 h or later after the onset of anti-IgM stimulation had a significantly lesser anti-apoptotic effect than when added at the onset of culture (not shown). By contrast, CT aggravated apoptosis in anti-IgM-activated B cells and failed to support TI-type responses in the absence of T cells (nu/nu mice). However, in the presence of T cells, i.e., with a possibility for CD40-CD40L interactions, both CT and CTA1-DD promote B cell expansion and GC formation.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
The present study clearly illustrates that targeted immunomodulation using bacterial ADP-ribosylating enterotoxins in novel gene fusion proteins may achieve effects on immunocytes not seen with the native CT or LT holotoxins. Here we show that the CTA1-DD adjuvant, contrary to the holotoxin, was effective at augmenting TI responses in athymic nu/nu mice, i.e., in the absence of T cell help. The CTA1-DD adjuvant acted directly on the B cell. By contrast, it was evident that adjuvanticity of CT required the presence of T cells, because CT enhanced TI responses in wild-type mice, but failed to do so in athymic mice. We believe that the reason for this difference can be found in the differential effects on apoptosis of activated B cells that were observed. Whereas CTA1-DD induced intracellular Bcl-2 and prevented apoptosis, CT strikingly augmented apoptosis and had no effect on Bcl-2 levels in Ag receptor-activated B cells. In wild-type mice B cells under the influence of CT adjuvant have access to T cells expressing CD40L, and our experiments using anti-CD40 stimulation of B cells in vitro support an essential role for the CD40-CD40L interaction for the adjuvant function of CT. In fact, CT fails to augment B cell responses in CD40-deficient mice (N. Lycke, unpublished observation). Thus, it appears that TI responses can be augmented by CT in vivo provided that T cells are present that express CD40L, which counteracts apoptosis in Ag-activated B cells. This conclusion agrees well with our earlier observations (20, 31, 42) and also helps explain conflicting information on CT’s suppressive effect on B cell responses in vitro and the enhancing effects that are observed in vivo (43, 44).

Here and in our previous work we have demonstrated that the strategy of separating adjuvanticity from toxicity in CT by constructing a targeted fusion protein, CTA1-DD, has been highly successful (9, 10, 19, 45). However, whether CT and CTA1-DD act through the same or different mechanisms remains to be shown. Extending our information from a recent study, we report that the ADP-ribosyltransferase activity of CTA1-DD is critical for the adjuvant function, not only for TD responses (19), but also for TI responses. We observed strong adjuvant effects of CTA1-DD in athymic mice, suggesting a direct immunoenhancing effect of CTA1-DD on the B cells, while the enzymatically inactive mutant, CTA1-R7K-DD, failed to augment TI responses. Thus, the enzymatic activity is essential for both TD and TI responses, indicating that the adjuvant mechanism of CTA1-DD differs from that of the newly described, enzymatically inactive mutant holotoxins, such as CTS61F (14, 46), which appear to rely on the structural or binding properties of the AB5 complex (14, 15, 16, 18). Although a common denominator for CT and CTA1-DD is the ADP-ribosyltransferase activity, we still lack information about which cells are affected by the enzyme (3). It is possible that both adjuvants act on the same cells, or they could act on completely different cells.

Previous studies from our laboratory have described direct effects of CT on IgM+ B cell differentiation in vitro, showing enhanced isotype switching to downstream isotypes, especially in the presence of IL-4 and T cell help (42, 44). The results of the present study on TI responses in athymic mice further underscore a requirement for T cell help in this process. By contrast, CTA1-DD promoted B cell responses in the absence of T cell interactions. Our data suggest that the CTA1-DD adjuvant achieved this effect after direct binding to B cells, as evidenced by in vitro and in vivo detection of CTA1-DD binding using FACS and immunohistochemical analysis. The fact that CTA1-DD could be detected on B cells in the spleen at 2 h following i.v. injection also strongly argues against the idea that the fusion protein predominantly binds to soluble Ig, forming IC or circulating CTA1-DD-Ig fusions. No binding to dendritic cells or macrophages was observed (9), arguing against the idea that CTA1-DD must interact with these professional APCs to exert an adjuvant effect. Also, in support of such an interpretation we observed an excretion pattern of [125I]CTA1-DD closely resembling that of the similarly labeled OVA, indicating that CTA1-DD did not form circulating IC. It would have been expected that radioactively labeled material would have been cleared from the circulation and accumulated in the kidney and liver over time had IC formation been prominent (29, 30). As final proof that CTA1-DD can circulate without binding to soluble Ig, we detected intact CTA1-DD, of the expected molecular size (37 kDa), by immunoblotting serum taken 1 h following an i.p. injection.

Moreover, to exclude the possibility that CTA1-DD, after forming IC or CTA1-DD-Ig fusions, exerted immunoenhancing effects through uptake via an Fc{gamma}R-dependent mechanism, we compared the augmenting ability in Fc{epsilon}R-deficient (common FcR{gamma}-chain) or Fc{gamma}RII-deficient mice with that observed in wild-type mice (22, 23). The Fc{epsilon}R-/- mice do not express the common FcR{gamma}-chain and, hence, are devoid of the Fc{gamma}RI, Fc{gamma}RIII, and Fc{epsilon}R (23). The common FcR{gamma}-chain has been found to be critically involved in phagocytosis by macrophages, Ab-dependent cell-mediated cytotoxicity, Ab-dependent inflammation, degranulation of mast cells and basophils, and triggering of peripheral and systemic anaphylaxis (45). By contrast, the Fc{gamma}RII acts primarily as a negative regulator of B cells and of IC-triggered activation of macrophages (22). We found unaltered enhancing effects of CTA1-DD on Ab responses following immunizations in Fc{gamma}R-deficient mice. This suggests that Fc{gamma}R-expressing cells are dispensable for the augmenting effect of CTA1-DD and supports the idea that Fc{gamma}R uptake of CTA1-DD-Ig fusions or IC is not required for the adjuvant function. In agreement with previous studies we observed stronger responses in Fc{gamma}RII-deficient mice, confirming the inhibitory role of this receptor in B cell responses (22). Taken together, the present study has convincingly demonstrated that CTA1-DD 1) has intact adjuvanticity in the absence of Fc{gamma}Rs, 2) does not primarily bind to soluble Ig in serum, 3) directly binds to B cells and affects survival/apoptosis, and 4) augments TI responses in athymic mice. Thus, these results all favor the interpretation that CTA1-DD exerts its immunoenhancing effect through direct effects on B cells. Importantly, we demonstrated that CTA1-DD has access to both naive, i.e., IgD+, and Ag-experienced, B cells in vivo, indicating that CTA1-DD can bind to a wide range of B cells in different stages of differentiation.

It has previously been reported that fusions of Ag to ZZ dimers, which are closely related to the DD dimers and are also derived from S. aureus protein A, have increased immunogenicity compared with Ag given alone (47, 48). Whether the enhanced immunogenicity was associated with IC formation and uptake via FcR-expressing cells or direct binding to B cells was not investigated. Because Leonetti et al. demonstrated that the immunogenicity of Ag-ZZ fusions could be further improved by noncovalently complexing these with Abs that target cell surface molecules such as MHC II and IgG, it could be argued that IC formation and FcR uptake are less likely to contribute to the strong immunogenicity of Ag-ZZ (47, 49, 50). By inference, this observation also supports our conclusion that CTA1-DD does not primarily form IC and that the adjuvant function is independent of FcR-expressing cells.

The mechanisms responsible for the adjuvant effect of CT are gradually getting less elusive. Nonetheless, we lack detailed knowledge about which of the described effects are critical for the adjuvant function (3). The use of targeted CTA1-DD may provide useful insights into the adjuvant mechanisms of the bacterial enterotoxins. We found that the holotoxin and CTA1-DD affected apoptosis and Bcl-2 induction in activated B cells differently. In the presence of Ag-receptor activation, CTA1-DD, but not CT, induced strong Bcl-2 expression. At variance with a recent study, this effect on Bcl-2-induction in B cells was independent of the CD19 coreceptor (41). Although the mechanism for the anti-apoptotic/Bcl-2 inducing effect of CTA1-DD remains to be investigated, it is possible that membrane interactions other than CD19 or the Ag-receptor may be involved. It has been reported that CD2-CD48 interactions up-regulate Bcl-2 and rescue B cells from Ag-receptor-induced apoptosis (51). The CD2-CD48 interaction also prevented cAMP-mediated apoptosis by reducing intracellular cAMP levels. However, this effect was insufficient against CT-induced apoptosis, which required CD40-CD40L interactions to be counteracted, a mechanism that appeared to be independent of intracellular cAMP levels (51). This agrees well with our findings in the present study, in which the apoptotic effect of CT was completely absent in the presence of CD40 stimulation. It also nicely explains our lack of enhancement of TI responses in athymic mice, which are devoid of CD40L-expressing CD4+ T cells. In addition, this information argues that the anti-apoptotic effect of CTA1-DD is independent of cAMP. In fact, in a recent study we demonstrated that CTA1-DD indeed ADP-ribosylated intact B cells, but we failed to detect changes in intracellular cAMP, suggesting that it did not affect adenylate cyclase (19). Whether CTA1-DD can replace CD40L binding and prevent CT-induced apoptosis has as yet not been investigated, but such studies may help explain at which level CTA1-DD prevents apoptosis. Of note, Bcl-2-induction was much weaker in the presence of CD40 compared with Ag-receptor stimulation, suggesting that CTA1-DD required Ag-receptor activation or at least that pathway to have an anti-apoptotic effect.

A question that arises is whether the anti-apoptotic effect is important for the adjuvant function in normal individuals. Both CT and CTA1-DD promoted the GC reaction and expansion of B cells in wild-type mice. In this regard they shared the same effect, but did CTA1-DD accomplish this by preventing apoptosis rather than by promoting expansion of B cells, and did CT act primarily by allowing CD40L-mediated expansion of B cells in the GC? At present, we have no exclusive answer to these questions. Previous studies have shown that CT dramatically affects B cell proliferation in vitro. In fact, most in vitro studies have demonstrated impaired proliferation to mitogens in the presence of biologically relevant concentrations of CT (52, 53, 54, 55, 56). Thus, CT does not appear to directly affect B cell survival and proliferation in a positive fashion; rather, the adjuvant effect is secondarily to the recruitment of activated T cells. As mentioned above, CD40-CD40L interactions could prevent apoptosis induced by CT/cAMP (51). Also, other T cell-derived factors can prevent the negative effects of CT on B cells. For example, the antiproliferative effect of CT can be overcome by IL-4, indicating that there are membrane activation pathways that are insensitive to CT treatment and cAMP increases (53, 54, 55, 56). These and other reports corroborate the idea that CT is dependent on activated T cells to exert adjuvant function. We reported earlier that oral immunization with CT adjuvant caused an increase in CD40L-expressing CD4+ T cells in Peyer’s patches, a phenomenon that was dependent on an intact B7-CD28 interaction (31). Thus, recruitment of activated CD4+ T cells appears to be a prerequisite for an adjuvant effect of CT. This could be achieved in many ways, but CT has been found to strongly up-regulate the expression of B7.2 on B cells as well as other APC, which could enhance CD4+ T cell priming and prevent apoptosis in B cells (9, 57, 58). Whether it is the antiapototic or the T cell priming effect that is responsible for the adjuvant effect of CTA1-DD is not known. We have previously shown that CTA1-DD enhances CD4+ T cell priming, and recently it was found to also augment priming of Ag-specific CTL in vivo (9, 59). CTA1-DD shares the effect of CT on B cells by up-regulating B7.2 expression (9), which may indicate that it is the enhancing effect on CD4+ T cell priming that is the key mechanism for adjuvanticity of both CTA1-DD and CT (3). Further studies addressing these issues are clearly much warranted.


   Acknowledgments
 
We are grateful for the help provided by Fredrik Knoop in preparing Figs. 2Go and 5Go.


   Footnotes
 
1 This work was supported by the Swedish Medical Research Council, the Swedish Cancer Foundation, the World Health Organization Transdisease Vaccinology Program, EU Grant BIO-CT0505, AstraZeneca Boston, and National Institutes of Health Grant R01AI40701. Back

2 Address correspondence and reprint requests to Dr. Nils Lycke, Department of Medical Microbiology and Immunology, University of Göteborg, S-413 46 Göteborg, Sweden. Back

3 Abbreviations used in this paper: CT, cholera toxin; LT, heat-labile toxin; IC, immune complex; CTB, nontoxic B subunits of cholera toxin; CTA1, A1 fragment of cholera toxin; DD, Ig-binding D region of staphylococcal protein A; TI, thymus independent; TD, thymus dependent; GC, germinal center; Dx, dextran; CSA, chicken serum albumin; KLH, keyhole limpet hemocyanin; PAP, peroxidase-antiperoxidase; PNA, peanut agglutinin; CD40L, CD40 ligand. Back

Received for publication December 16, 1999. Accepted for publication April 3, 2000.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Woodrow, G. C.. 1997. An overview of biotechnology as applied to vaccine development. , , , , ed. New Generation Vaccines 25. Marcel Decker, New York.
  2. Edelman, R.. 1997. Adjuvants for the future. , , , , ed. New Generation Vaccines 173. Marcel Decker, New York.
  3. Lycke, N.. 1997. The mechanism of cholera toxin adjuvanticity. Res. Immunol. 148:504.[Medline]
  4. Elson, C. O.. 1999. Mucosal adjuvants. , , , , , , ed. Mucosal Immunology 817. Academic Press, San Diego.
  5. Levine, M. M., J. B. Kaper, R. E. Black, M. L. Clemens. 1983. New knowledge on pathogenesis of bacterial enteric infections as applied to vaccine development. Microbiol. Rev. 47:510.[Free Full Text]
  6. Spangler, B. D.. 1992. Structure and function of cholera toxin and the related Escherichia coli heat-labile enterotoxin. Microbiol. Rev. 56:622.[Abstract/Free Full Text]
  7. Holmgren, J.. 1981. Actions of cholera toxin and the prevention and treatment of cholera. Nature 292:413.[Medline]
  8. Merritt, E. A., W. G. Hol. 1995. AB5 toxins. Curr. Opin. Struct. Biol. 5:165.[Medline]
  9. Agren, L. C., L. Ekman, B. Lowenadler, N. Y. Lycke. 1997. Genetically engineered nontoxic vaccine adjuvant that combines B cell targeting with immunomodulation by cholera toxin A1 subunit. J. Immunol. 158:3936.[Abstract]
  10. Agren, L., B. Lowenadler, N. Lycke. 1998. A novel concept in mucosal adjuvanticity: the CTA1-DD adjuvant is a B cell-targeted fusion protein that incorporates the enzymatically active cholera toxin A1 subunit. Immunol. Cell Biol. 76:280.[Medline]
  11. Burnette, W. N., V. L. Mar, B. W. Platler, J. D. Schlotterbeck, M. D. McGinley, K. S. Stoney, M. F. Rhode, H. R. Kaslow. 1991. Site-specific mutagenesis of the catalytic subunit of cholera toxin: substituting lysine for arginine 7 causes loss of activity. Infect. Immun. 59:4266.[Abstract/Free Full Text]
  12. Douce, G., M. Pizza, M. Roberts, R. Rappuoli, and G. Dougan. 1998. Mutant pertussis and Escherichia coli heat-labile toxins as adjuvants for enhancing local and systemic immune responses to coadministered, nonliving antigens. In New Generation Vaccines. G. C. W. M. M. Levine, J. B. Kaper, and J. S. Cobon, eds. p. 253.
  13. Freytag, L. C., J. D. Clements. 1999. Bacterial toxins as mucosal adjuvants. Curr. Top. Microbiol. Immunol. 236:215.[Medline]
  14. Yamamoto, S., Y. Takeda, M. Yamamoto, H. Kurazono, K. Imaoka, M. Yamamoto, K. Fujihashi, M. Noda, H. Kiyono, J. R. McGhee. 1997. Mutants in the ADP-ribosyltransferase cleft of cholera toxin lack diarrheagenicity but retain adjuvanticity. J. Exp. Med. 185:1203.[Abstract/Free Full Text]
  15. Douce, G., M. M. Giuliani, V. Giannelli, M. G. Pizza, R. Rappuoli, G. Dougan. 1998. Mucosal immunogenicity of genetically detoxified derivatives of heat labile toxin from Escherichia coli. Vaccine 16:1065.[Medline]
  16. de Haan, L., W. Verweij, E. Agsteribbe, J. Wilschut. 1998. The role of ADP-ribosylation and G(M1)-binding activity in the mucosal immunogenicity and adjuvanticity of the Escherichia coli heat-labile enterotoxin and Vibrio cholerae cholera toxin. Immunol. Cell Biol. 76:270.[Medline]
  17. Dickinson, B. L., J. D. Clements. 1995. Dissociation of Escherichia coli heat-labile enterotoxin adjuvanticity from ADP-ribosyltransferase activity. Infect. Immun. 63:1617.[Abstract]
  18. Giuliani, M. M., G. Del Giudice, V. Giannelli, G. Dougan, G. Douce, R. Rappuoli, M. Pizza. 1998. Mucosal adjuvanticity and immunogenicity of LTR72, a novel mutant of Escherichia coli heat-labile enterotoxin with partial knockout of ADP-ribosyltransferase activity. J. Exp. Med. 187:1123.[Abstract/Free Full Text]
  19. Agren, L. C., L. Ekman, B. Lowenadler, J. G. Nedrud, N. Y. Lycke. 1999. Adjuvanticity of the cholera toxin A1-based gene fusion protein, CTA1-DD, is critically dependent on the ADP-ribosyltransferase and Ig-binding activity. J. Immunol. 162:2432.[Abstract/Free Full Text]
  20. Sverremark, E., C. Fernandez. 1998. Role of T cells and germinal center formation in the generation of immune responses to the thymus-independent carbohydrate dextran B512. J. Immunol. 161:4646.[Abstract/Free Full Text]
  21. Sverremark, E., C. Fernandez. 1998. Germinal center formation following immunization with the polysaccharide dextran B512 is substantially increased by cholera toxin. Int. Immunol. 10:851.[Abstract/Free Full Text]
  22. Takai, T., M. Ono, M. Hikida, H. Ohmori, J. V. Ravetch. 1996. Augmented humoral and anaphylactic responses in Fc{gamma}RII-deficient mice. Nature 379:346.[Medline]
  23. Takai, T., M. Li, D. Sylvestre, R. Clynes, J. V. Ravetch. 1994. FcR{gamma} chain deletion results in pleiotrophic effector cell defects. Cell 76:519.[Medline]
  24. Rickert, R. C., J. Roes, K. Rajewsky. 1997. B lymphocyte-specific, Cre-mediated mutagenesis in mice. Nucleic Acids Res 25:1317.[Abstract/Free Full Text]
  25. Seppala, I., O. Makela. 1989. Antigenicity of dextran-protein conjugates in mice: effect of molecular weight of the carbohydrate and comparison of two modes of coupling. J. Immunol. 143:1259.[Abstract]
  26. Hunter, R.. 1970. Standardization of the chloramine-T method of protein iodination. Proc. Soc. Exp. Biol. Med. 133:989.[Medline]
  27. Fernandez, C., G. Moller. 1990. Serum antibody and cellular immune response in mice to dextran B512. Cell. Immunol. 131:41.[Medline]
  28. Ravetch, J. V.. 1997. Fc receptors. Curr. Opin. Immunol. 9:121.[Medline]
  29. Johansson, A., T. Sundqvist, T. Skogh. 1988. Elimination kinetics of circulating antigens and immune complexes. III. Elimination of intravenously injected model antigens and soluble IgG immune complexes in mice. J. Clin. Lab. Immunol. 27:19.[Medline]
  30. Johansson, A. G., T. Lovdal, K. E. Magnusson, T. Berg, T. Skogh. 1996. Liver cell uptake and degradation of soluble immunoglobulin G immune complexes in vivo and in vitro in rats. Hepatology 24:169.[Medline]
  31. Gardby, E., P. Lane, N. Y. Lycke. 1998. Requirements for B7-CD28 costimulation in mucosal IgA responses: paradoxes observed in CTLA4-H {gamma}1 transgenic mice. J. Immunol. 161:49.[Abstract/Free Full Text]
  32. Lees, A., F. Finkelman, J. K. Inman, K. Witherspoon, P. Johnson, J. Kennedy, J. J. Mond. 1994. Enhanced immunogenicity of protein-dextran conjugates. I. Rapid stimulation of enhanced antibody responses to poorly immunogenic molecules. Vaccine 12:1160.[Medline]
  33. Coutinho, A., G. Moller, W. Richter. 1974. Molecular basis of B-cell activation. I. Mitogenicity of native and substituted dextrans. Scand. J. Immunol. 3:321.[Medline]
  34. Fernandez, C., G. Moller. 1977. Immunological unresponsiveness to thymus-independent antigens: two fundamentally different genetic mechanisms of B-cell unresponsiveness to dextran. J. Exp. Med. 146:1663.[Abstract/Free Full Text]
  35. Liu, Y. J., C. Arpin. 1997. Germinal center development. Immunol. Rev. 156:111.[Medline]
  36. Wang, D., S. M. Wells, A. M. Stall, E. A. Kabat. 1994. Reaction of germinal centers in the T-cell-independent response to the bacterial polysaccharide {alpha}(1->6)dextran. Proc. Natl. Acad. Sci. USA 91:2502.[Abstract/Free Full Text]
  37. Kosco-Vilbois, M. H., H. Zentgraf, J. Gerdes, J. Y. Bonnefoy. 1997. To ‘B’ or not to ‘B’ a germinal center?. Immunol. Today 18:225.[Medline]
  38. Zheng, B., S. Han, G. Kelsoe. 1996. T helper cells in murine germinal centers are antigen-specific emigrants that downregulate Thy-1. J. Exp. Med. 184:1083.[Abstract/Free Full Text]
  39. Stedra, J., J. Cerny. 1994. Distinct pathways of B cell differentiation. I. Residual T cells in athymic mice support the development of splenic germinal centers and B cell memory without an induction of antibody. J. Immunol. 152:1718.[Abstract]
  40. Secord, E. A., L. V. Rizzo, E. W. Barroso, D. T. Umetsu, G. J. Thorbecke, R. H. DeKruyff. 1996. Reconstitution of germinal center formation in nude mice with Th1 and Th2 clones. Cell. Immunol. 174:173.[Medline]
  41. Roberts, T., E. C. Snow. 1999. Cutting edge: recruitment of the CD19/CD21 coreceptor to B cell antigen receptor is required for antigen-mediated expression of Bcl-2 by resting and cycling hen egg lysozyme transgenic B cells. J. Immunol. 162:4377.[Abstract/Free