This information is current as
of June 16, 2017.
Adjuvanticity of the Cholera Toxin A1-Based
Gene Fusion Protein, CTA1-DD, Is Critically
Dependent on the ADP-Ribosyltransferase
and Ig-Binding Activity
Lena C. Ågren, Lena Ekman, Björn Löwenadler, John G.
Nedrud and Nils Y. Lycke
J Immunol 1999; 162:2432-2440; ;
http://www.jimmunol.org/content/162/4/2432
Subscription
Permissions
Email Alerts
This article cites 43 articles, 23 of which you can access for free at:
http://www.jimmunol.org/content/162/4/2432.full#ref-list-1
Information about subscribing to The Journal of Immunology is online at:
http://jimmunol.org/subscription
Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
Receive free email-alerts when new articles cite this article. Sign up at:
http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by
The American Association of Immunologists, Inc.,
1451 Rockville Pike, Suite 650, Rockville, MD 20852
Copyright © 1999 by The American Association of
Immunologists All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017
References
Adjuvanticity of the Cholera Toxin A1-Based Gene Fusion
Protein, CTA1-DD, Is Critically Dependent on the
ADP-Ribosyltransferase and Ig-Binding Activity1
Lena C. Ågren,* Lena Ekman,* Björn Löwenadler,† John G. Nedrud,‡ and Nils Y. Lycke2*
C
onfusing and conflicting information regarding the mechanisms of adjuvanticity of cholera toxin (CT)3 and Escherichia coli heat-labile toxin (LT) has recently been presented (1– 8). These structurally closely related and well defined
molecules consist of a single enzymatically active A1 subunit
linked to a pentamere of cell membrane binding B subunits via a
connecting A2 subunit (9 –11). Whereas some investigators find
intact adjuvanticity in mutants devoid of the ADP-ribosyltransferase activity, others have provided evidence suggesting that the
enzymatic function is critically important (3, 4, 12). Thus, the role
of the A1 subunit for the adjuvant effect is poorly understood. In
fact, mutant molecules with single amino acid changes in the A1
subunit, rendering the enzyme partially or even completely inactive, have demonstrated adjuvant function comparable with that of
the intact holotoxin (2, 3, 6, 7, 13). Therefore, it is not clear
whether it is the ADP-ribosylating ability of the A1 subunit in the
holotoxin or its structural role in the AB5 complex or both that is
mediating the adjuvant effect.
*Department of Medical Microbiology and Immunology, University of Göteborg,
Göteborg, Sweden; †Department of Molecular Biology, Astra Hässle AB, Mölndal,
Sweden; and ‡Institute of Pathology, Case Western Reserve University, Cleveland,
OH
Received for publication September 8, 1998. Accepted for publication November
10, 1998.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1
This study was supported by The Swedish Medical Research Council, the Swedish
Cancer Foundation, the WHO Transdisease Vaccinology Program, Astra Research
Center Boston, EU project B104-98-0505, and National Institute of Health Grant R01
AI40701.
2
Address correspondence and reprint requests to Dr. Nils Lycke, Department of
Medical Microbiology and Immunology, University of Göteborg, S-413 46 Göteborg,
Sweden. E-mail address: [email protected]
3
Abbreviations used in this paper: CT, cholera toxin, LT, Escherichia coli heat-labile
toxin; ARFs, ADP-ribosylation factors; KLH, keyhole limpet hemocyanin; cAMP,
cyclic adenosine 39:59-monophosphate; i.n., intranasal.
Copyright © 1999 by The American Association of Immunologists
Our laboratory achieved a major breakthrough in research on
immunomodulation and vaccine adjuvant design when we constructed a gene fusion protein that combined the enzymatic activity
of the A1 subunit of CT with a B cell targeting moiety derived
from an Ig-binding fragment (D) of Staphylococcus aureus protein
A. The novel immunomodulator, CTA1-DD, was found to exert
adjuvant activity comparable with that of the intact CT after systemic as well as mucosal immunizations (14). However, in contrast
to CT, and as a consequence of the lack of receptor-binding B
subunits, the CTA1-DD adjuvant was found to be completely nontoxic in vitro and in vivo. Thus, we had successfully separated the
adjuvant function from the toxic side effects of the cholera holotoxin in this novel fusion protein. Furthermore, the CTA1-DD
molecule not only represents a promising new strategy for vaccine
adjuvant design but also proves the concept that novel immunomodulators can be constructed as gene fusion proteins that target
powerful bacterial enzymes to selected groups of cells, thereby
avoiding harmful side effects. We believe that among many diverse applications, this strategy has an obvious applicability in the
treatment of autoimmune diseases as well as in future vaccine
development.
CT and LT are among the most powerful immunoenhancing
molecules we know of today, and their use in vaccines is currently
being exploited (8, 15). Since CT and LT have been intensely
studied for several decades, our information on the structure and
function of these toxins is extensive. For example, it is well documented that both CT and LT enter the cell via the ganglioside
GM1-receptor, present on most nucleated mammalian cells, and
that this binding appears to be central for immunogenicity as well
as adjuvanticity (9, 16 –18). Furthermore, the toxins act via the A1
subunits to ADP-ribosylate GTP-binding proteins, of which the
Gsa is thought to be most important, resulting in activation of
adenylate cyclase and the subsequent increase in intracellular cyclic adenosine 39:59-monophosphate (cAMP) (19, 20). In addition,
the toxins interact with yet other GTP-binding proteins known as
0022-1767/99/$02.00
Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017
The ADP-ribosylating enterotoxins, cholera toxin (CT) and Escherichia coli heat-labile toxin, are among the most powerful
immunogens and adjuvants yet described. An innate problem, however, is their strong toxic effects, largely due to their promiscuous binding to all nucleated cells via their B subunits. Notwithstanding this, their exceptional immunomodulating ability is
attracting increasing attention for use in systemic and mucosal vaccines. Whereas others have separated adjuvanticity from
toxicity by disrupting the enzymatic activity of the A1 subunit by site-directed mutagenesis, we have constructed a nontoxic
molecule that combines the full enzymatic activity of the A1 subunit with a B cell targeting moiety in a gene fusion protein, the
CTA1-DD adjuvant. Despite its more selective binding properties, we found comparable adjuvant effects of the novel CTA1-DD
adjuvant to that of CT. Here we unequivocally demonstrate, using a panel of mutant CTA1-DD molecules, that the immunomodulating ability of CTA1-DD is dependent on both an intact enzymatic activity and the Ig-binding ability of the DD dimer. Both
agents, CT and CTA1-DD, ADP-ribosylate intact B cells. However, contrary to CT, no increase in intracellular cyclic AMP in the
targeted cells was detected, suggesting that cyclic AMP may not be important for adjuvanticity. Most remarkably, CTA1-DD
achieves similar immunomodulating effects to CT using a ganglioside-GM1 receptor-independent pathway for
internalization. The Journal of Immunology, 1999, 162: 2432–2440.
The Journal of Immunology
Materials and Methods
DNA constructions
A DNA fragment consisting of 582-bp-encoding amino acids 1–194 of
CTA1 (24) with flanking HindIII and BamHI sites was cloned into the
pUC19 vector to generate pUC19-CTA1. This vector was used as a template for site-directed mutagenesis and subcloning of PCR products before
sequencing. PCR was performed for mutagenesis of R7K in CTA1 with
59-TTACGCCAAGCTTCTAATGATGATAAGTTATATAAGGCAGAT
TCGCGA-39 and 59-TTGCATGATCATAAAGGTTGATATTCA-39 as
forward and reverse primers, respectively. The CTA1 mutations D109A
and E112K were constructed using the following oligonucleotides; D109A,
59-ACCATCCATATATTTGGGAGTATGGAATCCCACCTAAAGCAG
AAACTTCTTGTTCAGCTGGATGAGGACTG-39 (reverse) and 59-ACA
GGAAACAGCTATGAC-39 (forward); E112K, 59-GATACCATCCA
TATATTTGGGAGTATGGAATCCCACCTAAAGCAGAAACTTTTTG
T39 (reverse) and 59-ACAGGAAACAGCTATGAC-39 (forward),
respectively. The mutated CTA1-fragments were cloned into the expression vector pCTA1-DD as described earlier (14) to generate the vectors
pCTA1-R7K-DD, pCTA1-D109A-DD, and pCTA1-E112K-DD. A
CTA1-DD fusion protein lacking Fc-binding ability was prepared by
substitution of Ile34 to Ala and Lys38 to Ala in both D fragments. These
amino acid substitutions had previously been shown to strongly reduce the
Fc-binding ability of the Z domain of staphylococcal protein A to Ig (25,
26). Primers used for mutagenesis were 59-ACTTAAACGAAGAG
CAACGCAATGGTTTCGCTCAAAGTCTTGCAGACGATCCA-39 (forward) and 59-ACAGGAAACAGCTATGAC-39 (reverse), respectively.
The mutated D fragments were cloned into the pCTA1-DD expression
vector (14), cut with BspMI and BpuAI, to generate the pCTA1-DDI34A/
K38A expression vector. To enable purification of these fusion proteins,
plasmids were equipped with a C-terminal his-tag (his), pCTA1-DDhis and
pCTA1-DDI34A/K38Ahis, by PCR using the following primers: 59GCTCCAAGATCATCGGGATCCGGAAG (forward) and 59-GATCGA
ACAGTTTCCTAATGATGATGATGATGGTGCACCTGACCTGCTAC
CTCGGGTTT (reverse), respectively. The sequence and all DNA
fragments obtained by PCR were confirmed (27) with a Taq DyeDeoxy
Terminator cycle sequencing kit (Applied Biosystems, Foster City, CA) on
an Applied Biosystems Model 373A DNA Sequencing system. Oligonucleotides were purchased from Scandinavian Gene Synthesis (Köping,
Sweden). Low temperature melting agarose (NuSieve GTG; FMC Bioproducts, Rockland, MA) was used for preparative work, and Multi
Purpose agarose (Boehringer Mannheim, Mannheim, Germany) was used
for DNA analysis. PCR amplifications were performed using the DNA
Thermal Cycler and Taq DNA polymerase (Perkin-Elmer, Foster City,
CA). Restriction enzymes and T4 DNA ligase (Boehringer Mannheim and
New England Biolabs, Beverly, MA) were used as recommended.
Expression and purification of fusion proteins
The CTA1-DD, CTA1-R7K-DD, CTA1-D109A-DD, and CTA1E112K-DD fusion proteins were produced in E. coli and affinity purified on
an IgG column as described previously (14). The CTA1-DDhis and CTA1DDI34A/K38Ahis fusion proteins were expressed (14) and purified under
reducing conditions (6 M guanidine-HCl). Denatured CTA1-DDhis and
CTA1-DDI34A/K38Ahis were recovered in the soluble fraction and subsequently purified using a nickel chelate column according to the manufacturer’s instructions (Novagen, Madison, WI). Purified CTA1-DDhis and
CTA1-DDI34A/K38Ahis were renatured by two sequential dialysis steps,
overnight dialysis at 4°C against 1 M guanidine-HCl followed by 8 h of
dialysis against 0.2 M acetic acid, pH 3. Typical yield per culture was 8 –16
mg of fusion protein. The endotoxin content was ,5 ng of LPS/mg of
protein. The refolded soluble material was stored in 0.2 M acetic acid at
4°C.
Mice
Female C57BL/6 mice, age 8 –12 wk, from B & K Universal AB, Sollentuna, Sweden were used for immunizations and nu/nu mice on the
C57BL/6 background (Bomholt, Bomholtsgard, Denmark) were used for
all B cell experiments in vitro.
Evaluation of adjuvanticity
Priming and booster immunizations were performed i.p. with purified CT
(List Biological Laboratories, Campbell, CA) used at 1–2 mg/dose or affinity-purified CTA1-DD, CTA1-R7K-DD, CTA1-D109A-DD, CTA1E112K-DD gene fusion proteins used at 10 mg/dose. The CTA1-DDhis and
CTA1-DDI34A/K38Ahis fusion proteins were used at a low dose which
was equivalent to 5 mg and a higher dose equivalent to 10 mg of native
CTA1-DD ADP-ribosyltransferase activity. As test Ags we used keyhole
limpet hemocyanin (KLH; Calbiochem, San Diego, CA) or OVA (Sigma,
St. Louis, MO), at 5 and 100 mg/dose, respectively. Putative adjuvants
were simply admixed with the probe Ags in PBS before injection. Five
mice per group were immunized at 10-day intervals, and the mice were
sacrificed 6 – 8 days after the final immunization. Adjuvanticity was evaluated as enhanced serum anti-KLH, anti-OVA, or anti-CTA (List Biologicals) log10 titers by ELISA essentially as described (14, 28).
Analysis of structure, binding, and conformation
A MiniProtean (Bio-Rad, Richmond, CA) system was used for SDS-PAGE
analysis according to the manufacturer’s instructions. Ten micrograms
of CTA1-DD, CTA1-R7K-DD CTA1-D109A-DD, CTA1-E112K-DD,
CTA1-DDhis, and CTA1-DDI34A/K38Ahis were dissolved in buffer and
analyzed on a 12% Novex gel (San Diego, CA) under reducing conditions
as described earlier (14). Protein concentrations were determined by the
Bio-Rad DC Protein Assay (Bio-Rad).
The binding ability to mouse Ig- of the CTA1-DD, CTA1-R7K-DD
CTA1-D109A-DD, CTA1-E112K-DD, CTA1-DDhis, or CTA1-DDI34A/
K38Ahis fusion proteins was assessed by ELISA. Briefly, polystyrene microtiter plates (Maxisorp Immunoplates; Nunc, Roskilde, Denmark) were
coated with 10 mg/ml of murine IgG or IgM (PharMingen, San Diego, CA)
or IgG Fab or IgG Fc fragments (Jackson ImmunoResearch Labs, West
Port, PA) in PBS. Plates were blocked with 0.1% BSA in PBS/0.05%
Tween 20 (PBST) for 1 h at 37°C, and serial 1:2 dilutions of the fusion
proteins starting with 100 mg/ml were added to duplicate wells. Plates were
incubated for 1 h at room temperature (room temperature), washed with
PBST, and alkaline phosphatase-conjugated chicken anti-staphylococcal
protein A Abs (Immunsystem AB, Stockholm, Sweden), was added to the
plates at 1/500 dilution. After 1h bound Abs were visualized using N-pnitrophenylphosphate (Sigma) substrate in diethanolamine buffer, and the
reaction was read at 405 nm using a Thermomax microplate reader (Molecular Devices, Sunnyvale, CA). The ability of mutant or native
CTA1-DD molecules to bind to naive B cells in vivo was assessed by
FACS. Before analysis, i.v injections of 20 mg of mutant or native
CTA1-DD fusion protein were performed. CTA1-DD on B cells was detected by double-labeling of the isolated splenocytes with phycoerythrinand FITC-conjugated anti-IgD (PharMingen) and anti-protein A (Immunsystem AB) Abs, respectively, using a FACscan (Becton Dickinson, San
Jose, CA) (14). Gates were set on IgD1 cells which were analyzed for
mean fluorescence intensity using histogram representation (14).
Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017
ADP-ribosylation factors, ARFs, that are thought to augment
ADP-ribosyltransferase activity as well as participate in endocytic
and exocytic trafficking (21, 22). For example, ARF1 is thought to
engage in transport between the endoplasmic reticulum and the
Golgi (21, 22). Also, after internalization, CT itself is transported
to the Golgi and endoplasmic reticulum, a process that is independent of the A subunit (23). Thus, CTA1 may potentially have several functions that do not necessarily involve an enzymatic activity
but merely are effects of other interactions that for example, might
facilitate intracellular transport.
With the demonstration of a strong adjuvant function of the
CTA1-DD molecule we have proved that CTA1, separate from the
rest of the holotoxin, exerts immunomodulating ability. However,
we did not assess whether the adjuvant activity was a result of
ADP-ribosyltransferase activity and/or the effect of structural or
binding properties of CTA1 in the fusion protein. To this end, we
have now used site-directed mutagenesis to construct three enzymatically inactive mutants, CTA1-R7K-DD, CTA1-D109A-DD,
and CTA1-E112K-DD, and compared their adjuvant function to
that of the native CTA1-DD molecule and CT (14). Furthermore,
we addressed the importance of the Ig-binding ability for the adjuvant function by introducing mutations in the Ig-binding D region of the CTA1-DD fusion protein. Using these newly constructed molecules, we now unequivocally demonstrate that the
ADP-ribosyltransferase activity as well as the Ig-binding ability
are critically required for the adjuvant function of the CTA1-DD
fusion protein. These results also confirm the importance of ADPribosylation activity for the adjuvant function in intact CT or LT.
2433
2434
TARGETED ADP-RIBOSYLTRANSFERASE ACTIVE ADJUVANTS
CTA1-specific Abs from CT-hyperimmunized mice (generated by three
i.p. immunizations with CT at 2 mg/dose) were used to detect conformational changes of CTA1 in the mutant fusion proteins. Briefly, microtiter
plates (Maxisorp Immuno plates; Nunc) were coated with 10 mg/ml of
human IgG (Jackson ImmunoResearch Labs) followed by CTA1-DD or
mutant proteins incubated at 100 mg/ml for 1 h at room temperature. Thereafter, hyperimmune CT-specific serum in serial dilutions was added to the
plates and allowed to incubate for 1 h at room temperature. The degree of
recognition of CTA1 in the different mutants by CTA-hyperimmune sera
was visualized by the addition of alkaline phosphatase-conjugated goat
anti-mouse IgG1 (Southern Biotechnology Associates, Birmingham, AL)
at 1/200 followed by N-p-nitrophenylphosphate substrate as described
above.
ADP-ribosyltransferase and cAMP activity
RESULTS
The ADP-ribosyltransferase activity of CTA1-DD is critical for
the adjuvant function
The CTA1-DD fusion protein provides a unique possibility to experimentally address the role of the enzymatic activity of CTA1
for the adjuvant effect of the bacterial enterotoxins, CT and LT
(14). A broad approach was taken to evaluate whether the
CTA1-DD molecule required ADP-ribosyltransferase activity or
not for the adjuvant function. We constructed three enzymatically
inactive mutants, CTA1-R7K-DD, CTA1-D109A-DD, and CTA1E112K-DD (Fig. 1). Using the cell-free NAD-agmatine assay, we
observed that while the mutants failed to demonstrate ADP-ribosyltransferase activity even at the highest doses tested (Fig. 2),
CTA1-DD exhibited an enzymatic activity that was roughly 50%,
on a molar level, of that of the holotoxin. Following an i.p. priming
and booster immunization with KLH or OVA in the presence or
absence of active CTA1-DD or the inactive mutants, serum Ab
formation was monitored. We found that the adjuvant effect of
CTA1-DD was comparable ( p , 0.05) with that of CT whereas
the enzymatically inactive mutants CTA1-R7K-DD, CTA1D109A-DD, and CTA1-E112K-DD all failed to significantly augment systemic immune responses above that seen with Ag alone
(Fig. 3). Thus, the ADP-ribosyltransferase activity was essential
for the adjuvant functions. Substitutions in the CTA1 molecule at
positions Arg7 to Lys, Asp109 to Ala and Glu112 to Lys completely
abolished the ADP-ribosyltransferase activity and concomitantly
the mutants lost their adjuvant function in vivo. The data strongly
suggest that CTA1 exerts adjuvant effect in association with enzymatic activity and that structurally related properties or func-
FIGURE 1. Schematic drawing of the CTA1 subunit and the amino acid
substitutions in the enzymatically active cleft that were tested in the study.
Ribbon model of the crystallographic structure of the a-chain of cholera
toxin (51). Mutants are shown as space-filling models.
tions such as the mere binding of CTA1 to a ligand did not suffice
for an adjuvant effect.
The enzymatically inactive CTA1-DD mutants seem to have a
preserved structure
It could be argued that even a single amino acid substitution in the
enzymatically active cleft of the CTA1-DD fusion protein might
not only abolish enzyme function but also result in structural
changes of the molecule. However, as illustrated by the representative data with the CTA1-R7K-DD mutant in Fig. 4, we were
unable to disclose any conformational or degradative changes in
the enzymatically inactive mutants: Mutants as well as the native
CTA1-DD molecule all behaved as monomeric fusion proteins 37
kDa in mass (Fig. 4A). CTA1-specific antiserum from CT-hyperimmunized mice recognized CTA1-DD and the mutants equally
well (Fig. 4B). Binding of the mutants to mouse IgD1 B cells after
i.v. injection (Fig. 4C), or to intact IgM or IgG, or to IgG Fc and
Fab fragments in solid phase in ELISA plates (Fig. 5) was retained.
FIGURE 2. ADP-ribosyltransferase activity. The NAD-agmatine assay
was used to detect enzymatic activity of serial 1:2-dilutions of CT (l),
CTA1-DD (f), CTA1-R7K-DD, CTA1-D109A-DD or CTA1-E112K-DD
(Œ) were assayed for ADP-ribosylagmatine formation through incorporation of [U-14C]adenine. The values represent mean counts per minute
(cpm) of three experiments with SD ,5%. The mean background activity
was 1987 cpm.
Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017
Determinations of ADP-ribosyltransferase and cAMP activity was performed as previously described (10, 14, 29). Briefly, the NAD:agmatine
assay was used to detect enzymatic activity in twofold dilutions, beginning
at 10 mg/ml, of CTA1-DD, CTA1-R7K-DD, CTA1-D109A-DD, CTA1E112K-DD, CTA1-DDhis, and CTA1-DDI34A/K38Ahis or CT by assessing the ADP-ribosylagmatine formation through incorporation of [U-14C]
adenine. A commercial cAMP-test kit was used according to the manufacturer’s instructions (Amersham International, Little Chalfont, U.K.).
Briefly, spleen B cells (107 cells/ml) from nu/nu mice or enriched peritoneal macrophages (4 3 106 cells/ml) were incubated with or without various concentrations of CT, rCTB CTA1-DD, or the various mutants and the
level of cAMP induction was determined after 1 h or at later time points
and calculated from a standard curve.
cAMP-independent B cell proliferation was determined essentially as
described (30). Briefly, triplicate cultures of nu/nu B cells (105 cells/well)
were incubated in Iscove’s medium containing 10% FCS with 0.1 mg/ml
CT, 10 mg/ml CTA1-DD, 10 mg/ml CTA1-R7K-DD, or 10 mg/ml rCTB
together with 1 mM ionomycin with or without 10 mM 3-aminobenzamide
(Sigma-Aldrich, St. Louis, MO) (31). After 3 days, [3H]thymidine (1 mCi;
Amersham) was added for the final 18 h of culture. The cells were harvested on glass fiber filters, and the [3H]TdR uptake was assessed using a
b-scintillation counter (1450 MicroBetaTriLux, Wallac, FL).
Statistical analysis. We used the t test for independent samples for analysis of significance.
The Journal of Immunology
These results are in agreement with unaltered structural and Igbinding properties of the mutant CTA1-DD fusion proteins.
In addition, we used CTA1-specific serum Abs as an internal
standard to assess altered immunoenhancing activity of the mutated CTA1-DD molecules in vivo. CTA1-specific serum Abs
were determined after i.p. immunizations. Both CT and CTA1-DD
exerted strong immunoenhancing effect on CTA-specific responses, whereas, by contrast, the mutated CTA1-DD, represented
by the CTA1-R7K-DD and CTA1-E112K-DD mutations in Fig. 6,
stimulated poor or no ( p , 0.05) CTA1-specific serum Ab titers.
Thus, the lack of enzymatic activity dramatically reduced the immunogenicity of the CTA1-moiety itself.
Adjuvanticity of CTA1-DD is dependent on the ability to bind to
Ig
Next we investigated the importance of the Ig-binding ability of
the DD dimer for the adjuvant function of CTA1-DD. On the basis
of earlier findings, we constructed a fusion protein that contained
two mutations in the D fragments: Ile34 to Ala and Lys38 to Ala,
CTA1-DDI34A/K38Ahis mutant (25, 26, 32). To enable purification of the non-Ig-binding fusion proteins, we added a C-terminal
his-tag consisting of six histidine residues (His)6 to the construct,
and purification was performed using a nickel chelate column. For
FIGURE 4. A comparative analysis of the structural and conformational
integrity of the mutant and native CTA-DD molecules. A, The CTA1R7K-DD fusion protein (lane 1) was analyzed after purification on a 12%
Novex gel under reducing conditions and compared with native CTA1-DD
(lane 2). Lane M shows m.w. markers (kilodaltons). Similar results were
obtained with CTA1-D109A-DD or CTA1-E112K-DD. B, CTA1-specific
epitope recognition was determined by ELISA by incubating the fusion
proteins with serial dilutions of hyperimmune anti-CT serum. The analysis
was performed in duplicates, and values are shown for the CTA1-R7K-DD
mutant and expressed as mean absorbance at 405 nm of three independent
experiments. SD was ,5%. C, Ability of mutant and native CTA1-DD to
bind to naive IgD1 B cells. Freshly isolated splenocytes were isolated 4 h
after an i.v. injection of 20 mg/ml of the fusion protein. Isolated splenocytes
were double-labeled with anti-staphylococcal protein A-FITC and antiIgD-PE and analyzed for DD binding by FACS by setting live gates on
IgD1 lymphocytes. The fluorescence intensity was monitored by histogram
representation. Background labeling of splenocytes from control, PBS-injected mice, is seen in gray. This is a representative experiment of at least
three using the CTA1-R7K-DD mutant. Similar results were obtained with
the other two mutants.
comparison, we also constructed a fully competent Ig-binding
CTA1-DDhis fusion protein that was similarly purified. After assessment of the ADP-ribosylating capacity of the various constructs, we compared their immunoenhancing ability using doses
of equivalent enzymatic activity. We found that the CTA1DDI34A/K38Ahis mutant exhibited significantly ( p , 0.01) lower
immunoenhancement of KLH-specific serum Ab responses than
did the native Ig-binding CTA1-DDhis fusion protein (Fig. 7).
With a low dose of CTA1-DDI34A/K38Ahis mutant, no adjuvant
effect was achieved, while the Ig-binding CTA1-DDhis gave substantial enhancement. With a higher dose, the immunoenhancing
Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017
FIGURE 3. Enzymatically inactive mutants of CTA1-DD lack adjuvant
activity. Mice were immunized twice i.p. with 5 mg/dose of KLH (A) or
100 mg/dose of OVA (B) in the absence or presence of CT at 2 mg/dose or
CTA1-DD, CTA1-R7K-DD, CTA1-D109A-DD, or CTA1-E112K-DD at
10 mg/dose. Ag-specific serum Abs were determined by ELISA and expressed as total Ig mean log10-titers 6 SD of three experiments. The
CTA1-DD and CT-enhanced OVA and KLH titers are significantly different from titers generated in the presence of ADP-ribosyltransferase-inactive mutants of CTA1-DD; p , 0.005.
2435
2436
TARGETED ADP-RIBOSYLTRANSFERASE ACTIVE ADJUVANTS
FIGURE 5. Native or enzymatically inactive mutants of CTA1-DD
bind equally well to mouse Igs in solid phase. The ability of the inactive
mutants to bind to mouse IgG, IgM, or Fab and Fc fragments of IgG was
analyzed in duplicates by ELISA and compared with that of native CTA1DD. This depicts the representative results of the CTA1-R7K-DD mutant
and gives the mean absorbance values at 405 nm of three independent
experiments with SD ,5%.
effect was almost 10-fold lower than the native Ig-binding CTA1DDhis. Thus, the adjuvant effect of CTA1-DD was dramatically
reduced when the Ig-binding ability of the D fragments were
changed.
However, the loss of adjuvanticity using the high dose of the
CTA1-DDI34A/K38Ahis mutant was not complete compared with
the PBS control (Fig. 7). A detailed analysis of the Ig-binding
ability revealed that although the Fc binding of the CTA1DDI34A/K38Ahis fusion protein was lost, it still had some residual binding to Fab fragments (Fig. 8). The diminished binding was
particularly evident with low doses of CTA1-DDI34A/K38Ahis
mutant, whereas higher doses bound Fab fragments. This observation would sufficiently explain the residual adjuvant effect of the
CTA1-DDI34A/K38Ahis mutant observed with the high dose in
vivo. Nevertheless, these results documented a close relationship
between the Ig-binding ability and the adjuvant effects of the
CTA1-DD molecule.
The adjuvant effect of CTA1-DD appears to be cAMP
independent
In the present study, we have shown that the adjuvant effect of
CTA1-DD was comparable with that of CT and required the ADP-
ribosyltransferase activity, as exemplified by the lack of adjuvanticity of, e.g., the CTA1-R7K-DD mutation. Since CT acts by
ADP-ribosylating G-proteins, primarily the Gsa, that activate adenylate cyclase, leading to increased intracellular cAMP levels, we
investigated the effects of CTA1-DD on intracellular cAMP levels.
Unexpectedly, we found that the intracellular cAMP levels in
freshly isolated B cells or macrophages after treatment for 1 h with
CTA1-DD did not change relative to control cultures, while they
were substantially increased, in a dose-dependent manner, after
treatment with CT (Table I). The CTA1-DD fusion protein did not
affect cAMP levels even in doses 10,000-fold in excess of the
FIGURE 7. The adjuvant effect of CTA1-DD requires binding to Ig via
the DD-dimer. Mice were immunized twice with 5 mg/dose of KLH in the
absence or presence of CTA1-DDhis or CTA1-DDI34A/K38Ahis in doses
of comparable ADP-ribosyltransferase activity to native CTA1-DD. The
low dose was equivalent to 5 mg, and the higher dose was equivalent to 10
mg of CTA1-DD ADP-ribosyltransferase activity. KLH-specific Ab titers
in serum were determined by ELISA. Low dose: the CTA1-DDI34A/
K38Ahis KLH-titers are significantly different from CTA1-DDhis but not
from KLH alone (p , 0.01) (asterisk). High dose: the CTA1-DDI34A/
K38Ahis KLH titers are significantly different from CTA1-DDhis and
KLH alone (p , 0.05) (asterisk in oval).
Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017
FIGURE 6. Determinations of CTA1-specific Abs is a sensitive marker for
strong immunoenhancement associated with enzymatically active CTA1-DD.
Anti-CTA Abs in serum was assessed after i.p. immunization with CT, CTA1DD, or the enzymatically inactive mutants CTA1-R7K-DD and CTA1E112K-DD. Mice were immunized twice with CT at 2 mg/dose or CTA1-DD,
CTA1-R7K-DD, and CTA1-E112K-DD at 10 mg/dose. CTA1-specific Ab
titers in serum were determined by ELISA, and the results are expressed as
total Ig mean 6 SD log10 titers of three experiments. Similar results were
obtained with the CTA1-D109A-DO mutant. p , 0.05.
The Journal of Immunology
2437
Table I. Stimulation of intracellular cAMP levels a
Immunomodulator
(mg/ml)
B Cells
CT
0.1
0.01
0.001
0.0001
.16
.16
8.9
4.4
.16
.16
14.2
3.6
r-CTB
10
1.0
1.9
1.8
3.1
2.9
CTA1-DD
10
1.0
0.1
0.01
3.1
2.6
2.5
2.3
3.4
3.2
3.0
2.6
CTA1-R7K-DD
10
1.0
0.1
0.01
2.6
2.1
2.3
2.3
NT
NT
NT
NT
Medium
2.5
3.4
Macrophages
A dose-response analysis of induction of intracellular cAMP was performed with
spleen B cells from nu/nu mice or enriched peritoneal macrophages (PMF) from
C57BL/6 mice. Cells were incubated for 1h in the presence or absence of indicated
concentrations of CT, r-CTB. CTA1-DD, or CTA1-R7K-DD. cAMP values are expressed in picomols per 107 B cells or 4 3 106 PMF and are representative of seven
experiments giving similar results. rCTB, recombinant CTB subunit; NT, not tested.
FIGURE 8. The CTA1-DDI34A/K38Ahis mutant lacks Fc-binding
ability but has partly retained the Fab-binding ability to mouse Igs. The
ability of CTA1-DD with mutations in the D fragments to bind to mouse
Ig in solid phase was investigated by ELISA. The CTA1-DDhis and CTA1DDI34A/K38Ahis were analyzed in duplicates, and the results represent
the mean results of three experiments with SD of ,5%.
CTA1-DD- and CT-stimulated B cell proliferation was effectively
inhibited to near background levels, while no effect of 3-aminobenzamide was seen on B cell proliferation in the presence of the
CTA1-R7K-DD mutant or rCTB (Fig. 9). These results unequivocally demonstrated that the CTA1-DD fusion protein acts on intact B cells by exerting ADP-ribosyltransferase activity.
active dose of CT (Table I), nor did we observe increases in intracellular cAMP at later times, whereas CT promoted a dosedependent increase of intracellular cAMP at 1, 2 4, 8, or 21 h (not
shown). On the other hand, as expected, the CTA1-R7K-DD mutant or rCTB both failed to affect intracellular cAMP levels (Table
I). Consequently, the mechanism by which CTA1-DD exerts its
immunoenhancing effect appears to be independent of cAMP and
may not involve Gsa, which is believed to be the main target for
the enzymatic activity of CTA1 in the intact holotoxin.
Induction of B cell proliferation confirms that CTA1-DD exerts
ADP-ribosyltransferase activity on intact cells
Although CTA1-DD adjuvant was shown to act through ADPribosylation of target proteins in the cell-free NAD-agmatine system (Fig. 2), we did not know whether the fusion protein indeed
also could act on intact cells. To this end, we took advantage of a
previous study demonstrating that CT could exert cAMP-independent stimulatory effects on B cell proliferation in the presence of
ionomycin (30, 33). As illustrated in Fig. 9, both CT and CTA1DD, but not rCTB or CTA1-R7K-DD, stimulated significant B cell
proliferation in the presence of ionomycin (Fig. 9). The effect of
CTA1-DD was dramatic and dose dependent, while that of CT was
more modest, even in higher doses (not shown). Neither the
CTA1-R7K-DD mutant nor rCTB could significantly support
ionomycin-driven B cell proliferation. Moreover, when we added
3-aminobenzamide, a blocker of ADP-ribosylation (31), the
FIGURE 9. CTA1-DD acts on intact B cells as assessed by the ability
to promote ionomycin-driven ADP-ribosyltransferase-dependent proliferation. Enriched splenic B cells from nu/nu mice were incubated (2 3 105
cells/well) in Iscove’s total medium containing 10% FCS with 0.1 mg/ml
CT, or 10 mg/ml rCTB, CTA1-DD, or CTA1-R7K-DD, together with 1
mM ionomycin (black bars) or with 1 mM ionomycin and 10 mM 3-aminobenzamide (striped bars), a blocker of ADP-ribosyltransferase activity.
B cell proliferation was assessed by [3H]thymidine uptake on the third day
of incubation using a b-scintillation counter. Data are mean counts/minute
(cpm) 6 SD of triplicate cultures and representative of three independent
experiments. Values for cultures with CT or CTA1-DD are significantly
different from those with the enzymatically inactive mutant, CTA1-R7KDD, rCTB, or medium (asterisk in oval) or those obtained in the presence
of 3-aminobenzamide (p , 0.05) (asterisk).
Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017
a
2438
TARGETED ADP-RIBOSYLTRANSFERASE ACTIVE ADJUVANTS
Discussion
Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017
The mechanism for the adjuvant effect of the bacterial ADP-ribosylating enterotoxins, CT and LT, is still poorly understood (8, 34).
A current controversy in the field is the degree of adjuvanticity
contributed by the GM1 receptor-binding B subunits relative to that
contributed by the enzymatically active A1 subunit (8, 35, 36).
Here we show, using our newly constructed CTA1-DD fusion protein (14), that ADP-ribosyltransferase activity of the CTA1 is a
prerequisite for the adjuvant function. Moreover, adjuvanticity of
the fusion protein is dependent on its ability to bind to Ig via the
DD dimer. Thus, contrary to recently published work from several
laboratories using mutant holotoxins, we can unequivocally demonstrate that the ADP-ribosyltransferase active A1 subunit has potent immunomodulating ability of its own (1–3, 13). By excluding
the B subunits and via direct targeting of the full enzymatic activity of the CTA1 subunit to B cells, and perhaps other APC, we
have achieved full adjuvanticity with no toxicity. We believe that
the CTA1-DD fusion protein, therefore, represents an important
advancement in immunomodulation and vaccine adjuvant research.
The results of the present study are in keeping with findings
reported by Rappuoli and coworkers (6, 7), who demonstrated that
mutants of CT or LT holotoxin with partial ADP-ribosylating activity, CTS106 and LTR72, were significantly stronger adjuvants
than enzymatically inactive mutants. However, the inactive mutants, CTK63 or LTK63, had significantly better adjuvant ability
than the CTB or LTB subunits given alone. This finding has raised
the possibility that the AB5 complex may exert an adjuvant effect
that is independent of the enzymatic activity but dependent on
structural or binding properties of the A1 subunit. The authors put
forward a hypothesis that adjuvanticity of the bacterial holotoxins
is composed of two elements, enzymatic activity and, secondly,
nontoxic AB5 complex-associated effects (7). Possible targets for
this second effect could be the cell membrane interaction of the B
subunits or molecules that are bound by the A1 subunit. Because
CTB and LTB are generally poor immunogens and adjuvants (6,
37– 40) as compared with the nontoxic mutants, one might speculate that it is the interaction of the A1 subunit with some undefined cellular proteins that is the most interesting. It is known that
the enzymatic activity of CT and LT is enhanced by the interaction
of these proteins with intracellular ARFs, and it is conceivable that
an enzymatically inactive mutant A1 subunit might structurally
interact with these GTP-binding proteins (9, 21, 41). Although this
hypothesis is attractive, our results using the CTA1-DD fusion
protein do not support this explanation. In fact, we failed to demonstrate adjuvant activity in three different inactive CTA1-DD mutants. A single amino acid substitution in the active cleft disrupted
the enzymatic activity but did not seem to alter the conformation
or structure of the A1 subunit, thus probably still allowing binding
to ARFs or other molecules. Thus, although our current results do
not preclude a “structural” effect of CTA1, we believe our observations point in a different direction.
We would like to hypothesize that it is the ADP-ribosyltransferase activity of the A1 subunit per se that is important for the
immunomodulating effect. Our study clearly has demonstrated that
the enzymatic activity of the A1 subunit constitutes a unique and
separate immunomodulating ability of the bacterial enterotoxin.
The strength of this immunomodulation can easily be appreciated
by the fact that we achieved an adjuvanticity with the CTA1-DD
fusion protein of equal magnitude to that of the intact CT. The dose
required for such an effect was roughly threefold higher than that
of CT (14), supporting the idea that the holotoxins, indeed, also
host additional nontoxic AB5 complex-associated adjuvant prop-
erties. By inference, although not yet demonstrated, we predict that
it would be possible to replace or omit the A1 subunit from the
AB5 complex and still achieve stronger adjuvanticity than that obtained with the B subunits alone. No documentation is available on
the adjuvant role of the A2 subunit in the AB5 complex. Perhaps
a deletion of A1 would generate a molecule that still matches the
AB5 complex in adjuvant capability. Such a molecule has been
found an efficient delivery system for mucosal immune responses
to epitopes replacing the A1 subunit epitopes (42).
Thus, with regard to the second immunomodulating effect of the
holotoxins mediated by the AB5 complex, we firmly believe that it
is separate from the enzymatic activity of the A1 subunit. The
question is whether this effect is operating through a single or
multiple mechanisms. If the AB5 complex is more stable in vivo,
as been suggested (43), or even if it has a broader spectrum of
interactions with cell membrane or cytosolic substructures than the
B subunits alone, it is probable that immunomodulation dependent
on the B subunit.
Such a notion can explain why a great variability in adjuvanticity has been reported for the nontoxic mutants and the B subunits. In particular, the route of administration is critically influencing the degree of adjuvanticity. No doubt the oral route is most
demanding, whereas intranasal (i.n.) immunization appears to be
more permissive (6, 37). Indeed, several investigators have reported substantial adjuvant effects by CTB or LTB alone given i.n.,
whereas CTB or LTB, for the most part, has been found inactive
given perorally as adjuvants (35, 40, 44, 45). Interestingly,
Yamamoto et al. (3, 46) recently reported that the E112K mutant
CT, which is the same substitution tested by us in the CTA1E112K-DD fusion protein, gave enhancing effects on immune responses after peroral or s.c. administration similar to that observed
with the intact CT. Currently, we have no explanation for this
discrepancy, but in the present study we specifically used the i.p.
route for immunization to avoid possible involvement of variability in adjuvant and Ag uptake, as would be expected with the
peroral or i.n. routes. An earlier study from our laboratory also
failed to show adjuvant function of LT-E112K after oral immunization, whereas de Haan et al. (1, 40) found that nasal administration of this mutant gave good adjuvant effects. Taken together,
these observations indicate that the route of adjuvant administration and type of Ag used for evaluation are crucial elements to
consider when assessing the immunoenhancing effects of the mutant holotoxins.
Nevertheless, in practical terms all AB5 complex-dependent systems appear to have the same limitation, namely the requirement
for uptake via the ganglioside GM1 receptor. Since this receptor is
expressed on most nucleated cells, these adjuvant molecules can
never be selective in their action. Although Douce et al. have,
using the CTK63 and LTK63 mutants, provided evidence to suggest that LT may have broader immunogenic and adjuvant effects
than CT, most investigators have found that both molecules lack
these abilities in the mutated non-GM1 ganglioside-binding forms
(16, 17, 36). This is despite the fact that LT, as opposed to CT, can
bind several other carbohydrate ligands (47). Thus, adjuvanticity
of CT and LT appears to depend on the ganglioside GM1 receptor
pathway (16, 17). It should be mentioned, however, that at variance with these observations de Haan et al. (1) have reported a
non-GM1-binding LT mutant, LT-G33D, as well as a mutant that
lack both binding and ADP-ribosyltransferase activity, LT-E112K/
G33D, with retained adjuvant function after i.n. administration.
This latter observation is controversial, but, together with the reports of strong adjuvant effects of the B subunits given i.n., it
clearly demonstrates current problems in evaluation of adjuvant
function using the i.n. route of administration.
The Journal of Immunology
Acknowledgments
We thank Karin Schön for skilled technical assistance and Martin Norin for
help in preparing Fig. 1.
References
1. de Haan, L., I. K. Feil, W. R. Verweij, M. Holtrop, W. G. Hol, E. Agsteribbe, and
J. Wilschut. 1998. Mutational analysis of the role of ADP-ribosylation activity
and GM1-binding activity in the adjuvant properties of the Escherichia coli heatlabile enterotoxin towards intranasally administered keyhole limpet hemocyanin.
Eur. J. Immunol. 28:1243.
2. Dickinson, B. L., and J. D. Clements. 1995. Dissociation of Escherichia coli
heat-labile enterotoxin adjuvanticity from ADP-ribosyltransferase activity. Infect.
Immun. 63:1617.
3. Yamamoto, S., Y. Takeda, M. Yamamoto, H. Kurazono, K. Imaoka,
M. Yamamoto, K. Fujihashi, M. Noda, H. Kiyono, and J. R. McGhee. 1997.
Mutants in the ADP-ribosyltransferase cleft of cholera toxin lack diarrheagenicity
but retain adjuvanticity. J. Exp. Med. 185:1203.
4. Douce, G., C. Turcotte, I. Cropley, M. Roberts, M. Pizza, M. Domenghini,
R. Rappuoli, and G. Dougan. 1995. Mutants of Escherichia coli heat-labile toxin
lacking ADP-ribosyltransferase activity act as nontoxic, mucosal adjuvants. Proc.
Natl. Acad. Sci. USA 92:1644.
5. Grant, C. C., R. J. Messer, and W. Cieplak, Jr. 1994. Role of trypsin-like cleavage
at arginine 192 in the enzymatic and cytotonic activities of Escherichia coli
heat-labile enterotoxin. Infect. Immun. 62:4270.
6. Douce, G., M. Fontana, M. Pizza, R. Rappuoli, and G. Dougan. 1997. Intranasal
immunogenicity and adjuvanticity of site-directed mutant derivatives of cholera
toxin. Infect. Immun. 65:2821.
7. Giuliani, M. M., G. Del Giudice, V. Giannelli, G. Dougan, G. Douce,
R. Rappuoli, and 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.
8. Lycke, N. 1997. The mechanism of cholera toxin adjuvanticity. Res. Immunol.
148:504.
9. Rappuoli, R., and M. Pizza. 1991. Structure and evolutionary aspects of ADPribosylating toxins. In Sourcebook of Bacterial Protein Toxins. J. E. A. a. J. H.
Freer, ed. Academic Press, London, p. 1.
10. Spangler, B. D. 1992. Structure and function of cholera toxin and the related
Escherichia coli heat-labile enterotoxin. Microbiol. Rev. 56:622.
11. Merritt, E. A., and W. G. Hol. 1995. AB5 toxins. Curr. Opin. Struct. Biol. 5:165.
12. de Haan, L., W. R. Verweij, I. K. Feil, T. H. Lijnema, W. G. Hol, E. Agsteribbe,
and J. Wilschut. 1996. Mutants of the Escherichia coli heat-labile enterotoxin
with reduced ADP-ribosylation activity or no activity retain the immunogenic
properties of the native holotoxin. Infect. Immun. 64:5413.
13. Tsuji, T., T. Yokochi, H. Kamiya, Y. Kawamoto, A. Miyama, and Y. Asano.
1997. Relationship between a low toxicity of the mutant A subunit of enterotoxigenic Escherichia coli enterotoxin and its strong adjuvant action. Immunology 90:176.
14. Ågren, L., L. Ekman, B. Löwenadler, and N. Lycke. 1997. A genetically engineered non-toxic vaccine adjuvant that combines B cell targeting with immunomodulation by cholera toxin A1 subunit. J. Immunol. 158:3936.
15. Elson, C. O. 1996. Cholera toxin as a mucosal adjuvant. In Mucosal Vaccines.
Academic Press, Inc., New York.
16. Guidry, J. J., L. Cardenas, E. Cheng, and J. D. Clemens. 1997. Role of receptor
binding in toxicity, immunogenicity, and adjuvanticity of Escherichia coli heatlabile enterotoxin. Infect. Immun. 65:4943.
17. Nashar, T. O., H. M. Webb, S. Eaglestone, N. A. Williams, and T. R. Hirst. 1996.
Potent immunogenicity of the B subunits of Escherichia coli heat-labile enterotoxin: receptor binding is essential and induces differential modulation of lymphocyte subsets. Proc. Natl. Acad. Sci. USA 93:226.
18. Holmgren, J. 1981. Actions of cholera toxin and the prevention and treatment of
cholera. Nature 292:413.
19. Field, M., M. C. Rao, and E. B. Chang. 1989. Intestinal electrolyte transport and
diarrheal disease. N. Engl. J. Med. 321:800.
20. Moss, J., and M. Vaughan. 1988. ADP-ribosylation of guanine nucleotide-binding regulatory proteins by bacterial toxins. Adv. Enzymol. 61:303.
21. Donaldson, J. G., and R. D. Klausner. 1994. ARF: a key regulatory switch in
membrane traffic and organelle structure. Curr. Opin. Cell Biol. 6:527.
22. Moss, J., and M. Vaughan. 1995. Structure and function of ARF proteins: Activators of cholera toxin and critical components of intracellular vesicular transport
processes. J. Biol. Chem. 270:12327.
23. Sofer, A., and A. H. Futerman. 1996. Rate of retrograde transport of cholera toxin
from the plasma membrane to the Golgi apparatus and endoplasmic reticulum
decreases during neuronal development. J. Neurochem. 67:2134.
24. Mekalanos, J. J. 1983. Duplication and amplification of toxin genes in Vibrio
cholerae. Cell 35:253.
25. Jendeberg, L., B. Persson, R. Andersson, R. Karlsson, M. Uhlen, and B. Nilsson.
1995. Kinetic analysis of the interaction between protein A domain variants and
human Fc using plasmon resonance detection. J. Mol. Recognit. 8:270.
26. Cedergren, L., R. Andersson, B. Jansson, M. Uhlen, and B. Nilsson. 1993. Mutational analysis of the interaction between staphylococcal protein A and human
IgG1. Protein Eng. 6:441.
27. Sanger, F., S. Niclen, and A. R. Coulson. 1977. DNA sequencing with chain
terminating inhibitors. Proc. Natl. Acad. Sci. USA 74:5463.
28. Bromander, A. K., L. Ekman, M. Kopf, J. G. Nedrud, and N. Y. Lycke. 1996.
IL-6 deficient mice exhibit normal mucosal IgA responses to local immunizations
and Helicobacter felis infection. J. Immunol. 156:4290.
29. Tsuji, T., T. Inoue, A. Miyama, and M. Noda. 1991. Glutamic acid-112 of the A
subunit of heat-labile enterotoxin from enterotoxigenic Escherichia coli is important for ADP-ribosyltransferase activity. FEBS Lett. 291:319.
Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017
The unique strength of the CTA1-DD system is that it targets the
full enzymatic activity of the CTA1 to cells that are engaged in the
formation of an immune response, leaving all other cells at rest.
The inherent problem with the mutant holotoxins using the ganglioside GM1 pathway, incurred by the risk of toxicity, is the required balancing of adjuvanticity against enzymatic activity. Elegant studies have demonstrated that partially active LTR72 mutant
can give similar adjuvant effects to the intact holotoxin in doses
50 –100-fold higher (7). Such dose increases appeared to be safe
and nontoxic in mice, but for clinical use, adjuvant active, but
nontoxic, doses may be difficult to achieve. Ongoing clinical trials
with some of these mutants will hopefully answer this critical
question (48 –50).
What is most remarkable with the CTA1-DD adjuvant, however, is the ability to circumvent the natural ganglioside GM1 pathway for entrance of CTA1 into the cell via binding through the DD
dimer. Although we have not completely defined the ligands for
the DD dimer, we strongly believe that the B cell receptor is the
most likely candidate (14). Previous and ongoing studies point to
the idea that B cell targeting rather than binding to dendritic cells
and macrophages occurs after injection (14). Whether this effect is
responsible for the adjuvanticity or not awaits to be proved.
Clearly, the fact that admixed CTA1-DD, without conjugation to
the Ag, achieves strong immunoenhancement in CD41 T cell
priming and humoral immunity (14) argues in favor of an involvement in Ag presentation. We have also reported that up-regulation
of the costimulatory molecule CD86 is a primary event after treatment of naive B cells with CTA1-DD (14).
In the present study, we extend the evidence for an immunomodulating effect of CTA-DD on binding to the B cell, by showing
that B cell proliferation in the presence of ionomycin is greatly
enhanced. In this regard, the effect of CTA1-DD mimics that of
CT, whereas rCTB or the enzymatically inactive mutant
CTA1-DD failed to affect this system. Benzamide, a known inhibitor of ADP-ribosyltransferase activity, substantially abrogated the
proliferation induced by CTA1-DD and CT, clearly indicating that
these agents use ADP-ribosyltransferase activity for the effect (30).
Although CTA1 in the holotoxin is thought to ADP-ribosylate the
Gsa protein, leading to an activation of adenylate cyclase and subsequent increase of intracellular cAMP, we failed to show an effect
on cAMP levels. In spite of numerous experiments using different
conditions, high doses of the fusion protein and different time
points, using isolated B cells or macrophages or two different murine B cell lines (X16, CH12LX), we were unable to mimic the
effect of CT on cAMP levels. This observation was unexpected
and highly surprising.
A speculation derived from the findings mentioned above is that
CTA1 in the form of a fusion protein uses a pathway distinct from
the classical Gsa-adenylate cyclase-cAMP pathway for its adjuvant effect. In fact, we believe our report is the first to suggest that
CT exerts adjuvant activity independently of cAMP. Even though
the pathway for entrance of CTA1 into the target cell differs between the fusion protein and the holotoxin, the present study indicates that a substrate (or substrates), which is ADP-ribosylated,
in both cases is targeted by CTA1. The nature of this substrate
may, thus, not be Gsa but one or several other G-proteins, or
alternatively, it may be that the Gsa, in a different subcellular
location, is not acting on adenylate cyclase. Future studies in our
laboratory will address these important and challenging questions.
2439
2440
TARGETED ADP-RIBOSYLTRANSFERASE ACTIVE ADJUVANTS
41. Moss, J., and M. Vaughan. 1995. Structure and function of ARF proteins: activators of cholera toxin and critical components of intracellular vesicular transport
processes. J. Biol. Chem. 270:12327.
42. Hajishengallis, G., S. K. Hollingshead, T. Koga, and M. W. Russell. 1995. Mucosal immunization with a bacterial protein antigen genetically coupled to cholera
toxin A2/B subunits. J. Immunol. 154:4322.
43. Magagnoli, C., R. Manetti, M. R. Fontana, V. Giannelli, M. M. Giuliani,
R. Rappuoli, and M. Pizza. 1996. Mutations in the A subunit affect yield, stability, and protease sensitivity of nontoxic derivatives of heat-labile enterotoxin.
Infect. Immun. 64:5434.
44. Russell, M. W., and H.-Y. Wu. 1991. Distribution, persistence, and recall of
serum and salivary antibody responses to peroral immunization with protein antigen I/II of Streptococcus mutans coupled to the cholera toxin B subunit. Infect.
Immun. 59:4061.
45. Tamura, S., Y. Samegai, H. Kurata, T. Nagamine, C. Aizawa, and T. Kurata.
1988. Protection against influenza virus infection by vaccine inoculated intranasally with cholera toxin B subunit. Vaccine 6:409.
46. Yamamoto, S., H. Kiyono, M. Yamamoto, K. Imaoka, K. Fujihashi,
F. W. Van Ginkel, M. Noda, Y. Takeda, and J. R. McGhee. 1997. A nontoxic
mutant of cholera toxin elicits Th2-type responses for enhanced mucosal immunity. Proc. Natl. Acad. Sci. USA 94:5267.
47. Fukuta, S., J. L. Magnani, E. M. Twiddy, R. K. Holmes, and V. Ginsburg. 1988.
Comparison of the carbohydrate-binding specificities of cholera toxin and Escherichia coli heat-labile enterotoxins LTh-I, LT-IIa, and LT-IIb. Infect. Immun.
56:1748.
48. Michetti, P., C. Kreiss, K. Kotloff, N. Porta, J. L. Blanco, D. Bachman,
P. F. Sadlinger, I. Corthesy-Theulaz, G. Losonsky, R. Nichols, M. Stolte,
T. Monath, S. Ackerman, and Blum. 1997. Oral immunization of H. pylori infected adults with recombinant urease and LT adjuvant. Gastroenterology 112:
A1042 (Abstr.).
49. Conference Coverage (Interscience Conference on Antimicrobial Agents and
Chemotherapy). 1997. Vaccine Wkly. p. 4.
50. Levine, M. M., J. B. Kaper, R. E. Black, and M. L. Clemens. 1983. New knowledge on pathogenesis of bacterial enteric infections as applied to vaccine development. Microbiol. Rev. 47:510.
51. Zhang, R. G., D. L. Scott, M. L. Westbrook, S. Nance, B. D. Spangler,
G. G. Shipley, and E. M. Westbrook. 1995. The three-dimensional crystal structure of cholera toxin. J. Mol. Biol. 251:563.
Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017
30. Francis, M. L., I. Okazaki, J. Moss, A. Kurosky, L. M. Pecanha, and J. J. Mond.
1995. cAMP-independent effects of cholera toxin on B cell activation. III. Cholera toxin A subunit-mediated ADP-ribosylation acts synergistically with ionomycin or IL-4 to induce B cell proliferation. J. Immunol. 154:4956.
31. Rankin, P. W., E. L. Jacobson, R. C. Benjamin, J. Moss, and M. K. Jacobson.
1989. Quantitative studies of inhibitors of ADP-ribosylation in vitro and in vivo.
J. Biol. Chem. 264:4312.
32. Ljungberg, U. K., B. Jansson, U. Niss, R. Nilsson, B. E. Sandberg, and
B. Nilsson. 1993. The interaction between different domains of staphylococcal
protein A and human polyclonal IgG, IgA, IgM and F(ab9)2: separation of affinity
from specificity. Mol. Immunol. 30:1279.
33. Klaus, G. G., K. Vondy, and M. Holman. 1987. Selective effects of cholera toxin
on the activation of mouse B cells by different polyclonal activators. Eur. J. Immunol. 17:1787.
34. Elson, C. O., and M. T. Dertzbaugh. 1994. Mucosal adjuvants. In Handbook of
Mucosal Immunology. P. L. Ogra, J. Mestecky, M. E. Lamm, W. Strober,
J. R. McGhee, and J. Bienenstock, eds. Academic Press, San Diego, p. 391.
35. 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. Marcel
Dekker, New York, p. 253.
36. de Haan, L., W. Verweij, E. Agsteribbe, and J. Wilschut. 1998. The role of
ADP-ribosylation and GM1-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.
37. Blanchard, T. G., N. Lycke, S. J. Czinn, and J. Nedrud. 1998. Recombinant
cholera toxin B subunit is not an effective mucosal adjuvant for oral immunization of mice against H. felis. Immunology 93:22.
38. Lycke, N., and J. Holmgren. 1986. Strong adjuvant properties of cholera toxin
on gut mucosal immune responses to orally presented antigens. Immunology
39:301.
39. Van der Heijden, P. J., A. T. Bianchi, M. Dol, J. W. Pals, W. Stok, and
B. A. Bokhout. 1991. Manipulation of intestinal immune responses against
ovalbumin by cholera toxin and its B subunit in mice. Immunology 72:89.
40. Lycke, N., T. Tsuji, and J. Holmgren. 1992. The adjuvant effect of Vibrio cholerae and Escherichia coli heat-labile enterotoxins is linked to their ADP-ribosyltransferase activity. Eur. J. Immunol. 22:2277.