INFECTION AND IMMUNITY, Nov. 2004, p. 6255–6261
0019-9567/04/$08.00?0 DOI: 10.1128/IAI.72.11.6255–6261.2004
Copyright © 2004, American Society for Microbiology. All Rights Reserved.
Vol. 72, No. 11
Recognition of Mycobacterial Antigens Delivered by Genetically
Detoxified Bordetella pertussis Adenylate Cyclase by T Cells
from Cattle with Bovine Tuberculosis
H. Martin Vordermeier,1* Marcela Simsova,2Katalin A. Wilkinson,3
Robert J. Wilkinson,3R. Glyn Hewinson,1
Peter Sebo,2and Claude Leclerc4
TB Research Group, Veterinary Laboratories Agency, Weybridge, New Haw, Addlestone,1and Wellcome Centre for
Research in Clinical Tropical Medicine, Division of Medicine, Faculty of Medicine, Imperial College London,
Wright Fleming Institute, Paddington, London,3United Kingdom; Cell and Molecular Biology Division,
Institute of Microbiology of the Czech Academy of Sciences, Prague, Czech Republic2; and Unite
de Biologie des Regulations Immunitaires, INSERM E 352, Institut
Pasteur, Paris, France4
Received 3 June 2004/Returned for modification 15 July 2004/Accepted 3 August 2004
The exponential increase in the incidence of tuberculosis in cattle over the last two decades in the British
national herd constitutes a significant economic problem. Therefore, research efforts are under way to develop
cattle tuberculosis vaccines and specific diagnostic reagents to allow the distinction of vaccinated from infected
animals. Mycobacterial antigens like ESAT-6 and CFP10 allow this distinction. This study investigates whether
fusion protein of ESAT-6 or CFP10 with genetically detoxified Bordetella pertussis adenylate cyclase (CyaA) are
recognized by Mycobacterium bovis-infected cattle more effectively than conventional recombinant proteins are,
thus enhancing sensitivity or reducing the amount of antigens required. By measuring the frequencies of
gamma interferon (IFN-?)-producing cells, we were able to show that the presentation of CFP10 as a CyaA
fusion protein enhanced the molecular efficiency of its recognition 20-fold, while the recognition of ESAT-6 was
not improved by CyaA delivery. Furthermore, in the whole-blood IFN-? test currently used in the field, the
delivery of CFP10 and ESAT-6 by fusion to CyaA increased the amount of IFN-? produced and hence the
proportion of infected animals responding to CFP10. The improved T-cell recognition of CyaA336/CFP10 was
found to be dependent upon interaction with CD11b. In addition, presentation of CyaA336/CFP10 to CD4?T
cells was chloroquine sensitive, and CFP10 delivery by CyaA resulted in its accelerated presentation to T cells.
In conclusion, the use of CyaA fusion proteins with ESAT-6 and CFP10 has the potential to improve the
sensitivity of immunodiagnostic tests detecting bovine tuberculosis in cattle.
Bovine tuberculosis (BTB), caused by Mycobacterium bovis,
is a zoonotic disease and was the cause of approximately 6% of
total human deaths due to BTB in the 1930s and 1940s (5, 6).
The introduction of pasteurization of milk in developed coun-
tries in the 1930s dramatically reduced the transmission from
cattle to humans, although BTB is still a major human health
problem in developing countries. Compulsory BTB eradication
programs were introduced in many countries based on the
slaughter of infected cattle detected by the single intradermal
comparative tuberculin skin test. The implementation of this
control strategy resulted in the dramatic reduction of BTB in
Great Britain. However, possibly due to a wildlife reservoir,
the incidence of TB in cattle caused by M. bovis has exponen-
tially increased over the last two decades in the British national
herd. This increase constitutes a significant economic and po-
tential public health problem (17). To control this zoonotic
disease, better and more specific diagnostic reagents, as well as
effective vaccines for cattle, are urgently needed.
The diagnosis of BTB in cattle is at present almost exclu-
sively based on the use of tuberculin purified protein derivative
(PPD) in skin tests. In addition, a blood-based test measuring
tuberculin-induced production of gamma interferon (IFN-?) is
now also in limited field use (34). The specificity of these
tuberculin-based tests is limited due to the undefined and
cross-reactive nature of PPD. Furthermore, the specificity of
tuberculin-based reagents is also compromised following vac-
cination with the human TB vaccine M. bovis BCG (16). Di-
agnostic reagents allowing the differential diagnosis of M. bo-
precondition for the development of novel TB vaccines in
cattle (17). The specificity of diagnostic reagents can be im-
proved by using antigens that are highly expressed by M. bovis
yet whose genes have been deleted from the genome of BCG.
The antigens ESAT-6 and CFP10, which are encoded in the
RD1 region of M. bovis-Mycobacterium tuberculosis—which is
deleted in all strains of BCG (2, 19)—have shown particular
promise as such improved diagnostic reagents when used as
recombinant proteins or synthetic peptides in the IFN-? test
(3, 4, 28, 31). Such antigens not only allowed the differential
diagnosis of infected and BCG-vaccinated cattle but also im-
proved the specificity of the test per se compared to tuberculin
PPD in the absence of vaccination (22, 23, 31).
An attractive recent approach to effectively delivering pro-
* Corresponding author. Mailing address: VLA Weybridge, TB Re-
search Group, Woodham Lane, New Haw, Addlestone KT15 3NB,
United Kingdom. Phone: 44 1932 357 684. Fax: 44 1932 357 584.
teins to the immune system is through nonreplicating protein
vectors such as bacterial toxins (20). The Bordetella pertussis
adenylate cyclase (CyaA) is such a vector system that has
shown promise in mouse models (21, 25, 26). Indeed, we have
recently demonstrated that CD11b/CD18, a member of the
?2-integrin family, is a specific cell receptor for this toxin (14).
CD11b/CD18, also known as complement type 3 receptor or
MAC-1, is expressed on macrophages, neutrophils, dendritic
cells, and NK cells. Thus, the cellular specificity of CyaA allows
its selective targeting to CD11c?CD11bhighdendritic cells in
vivo (13). Peptide and small proteins can be inserted into this
protein and expressed as fusion proteins (11, 12, 21, 25, 27).
CyaA facilitates direct translocation of these inserted antigens
across the plasma membrane of target cells (15). Importantly,
it has been shown that CyaA vaccination can not only induce
major histocompatibility complex (MHC) class I-restricted
CD8?-T-cell responses (8, 10, 12, 21, 25, 27) but also result in
the presentation of CD4?-T-cell epitopes restricted by MHC
class II (8, 18). Thus, CyaA fusion proteins that contain my-
cobacterial antigens could constitute not only candidates for
subunit vaccines but also diagnostic antigens, particularly if
they will be recognized in cattle more effectively than conven-
tional recombinant proteins, thereby enhancing sensitivity, or
are recognized at lower protein concentrations. The latter con-
sideration could have major cost benefits because this could
significantly reduce the amount of antigen that would have to
be produced to implement testing. Consequently, the experi-
ments conducted in this study have been performed to deter-
mine whether CyaA-based recombinant proteins fused with
either ESAT-6 or CFP10 are recognized by bovine T cells more
efficiently than the corresponding nonfusion proteins and to
establish the mechanisms of this improved recognition.
MATERIALS AND METHODS
Antigens. Bovine (PPD-B) and avian (PPD-A) tuberculin were obtained from
the Tuberculin Production Unit at the Veterinary Laboratories Agency-Wey-
bridge (VLA Weybridge) and used in culture at 10 ?g/ml. Recombinant ESAT-6
was supplied by A. Whelan (VLA Weybridge), and recombinant CFP10 was
obtained from M. Singh, Gesellschaft fu ¨r BioTechnologische Forschung, Braun-
Escherichia coli XL1-Blue (Stratagene) was used throughout this work for
recombinant DNA construction and for expression of antigens inserted into
CyaA. Bacteria transformed with appropriate plasmids derived from pT7CACT1
were grown at 37°C in Luria-Bertani medium supplemented with 150 ?g of
ampicillin/ml. The open reading frames of M. tuberculosis H37Rv genes esat-6
and cfp-10 were amplified by PCR from the pYUB412 cosmid clone of the RD1
region with the following primers: Esat6-I, 5?-GATGTGTACACATGACAGA
GCAGCAGTGG-3?; Esat6-II, 5?-GATGTGTACACTGAGCGAACATCCCA
GTGACG-3?; CFP-10-I, 5?-CATGTGTACACATGGCAGAGATGAAGACC-
3?; CFP-10-II, 5?-CATGTGTACACTGAAGCCCATTTGCGAGGA-3?.
The PCR product was digested by bsrG1 at the sites incorporated into the PCR
primers, and the purified fragments encoding the antigens were inserted in-frame
between codons 335 and 336 of CyaA on the pT7CACT-336-BsrGI expression
vector (21). The exact sequence of the cloned inserts was verified by DNA
sequencing. The control detoxified mock CyaA and the recombinant CyaA
proteins carrying the ESAT-6 and CFP10 antigens, respectively, were produced
in E. coli, purified from inclusion bodies in 8 M urea–50 mM Tris-Cl (pH 8)–2
mM EDTA, and characterized as previously described. The resulting proteins
were free of any detectable CyaA enzymatic activity.
M. bovis-infected cattle and BCG vaccination. Calves were infected with an M.
bovis field strain from Great Britain (AF 2122/97) by intratracheal instillation of
between 5 ? 103and 5 ? 104CFU (28). Infection was confirmed by the presence
of tuberculous lesions in the lungs and lymph nodes of these animals as well as
by the culture of M. bovis from tissue collected at the postmortems performed 20
weeks after the infection. Heparinized blood samples were obtained at least 6
weeks after infection when strong and sustained in vitro tuberculin responses
were observed. Another group of 11 calves was vaccinated with BCG (Pasteur)
by subcutaneous injection of 106CFU. Blood was taken 3 weeks postvaccination
when peak responses were observed.
IFN-? ELISPOT assay. Peripheral blood mononuclear cells (PBMC) were
isolated from heparinized blood by Histopaque 1077 (Sigma) gradient centrifu-
gation and cultured in tissue culture medium (RPMI 1640; Life Technologies,
Paisley, Scotland, United Kingdom) supplemented with 5% controlled process
serum replacement type 1 (Sigma Aldrich, Poole, United Kingdom), nonessen-
tial amino acids (Sigma Aldrich), 5 ? 10?5M 2-mercaptoethanol, 100 U of
penicillin/ml, and 100 ?g of streptomycin sulfate/ml. Direct ELISPOTs were
enumerated, as described previously (28). Briefly, ELISPOT plates (Immuno-
bilon-P polyvinylidene difluoride membranes; Millipore, Molsheim, France)
were coated overnight at 4°C with the bovine IFN-?-specific monoclonal anti-
body (MAb) 2.2.1. Unbound antibody was removed by washing, and the wells
were blocked with 10% fetal calf serum in RPMI 1640 medium. PBMC (2 ? 105
to 5 ? 105/well) suspended in tissue culture medium (RPMI 1640 supplemented
with 5% controlled process serum replacement type 1) were then added and
cultured at 37°C and 5% CO2in a humidified incubator for 24 h. Spots were
developed with rabbit serum specific for IFN-? followed by incubation with an
alkaline phosphatase-conjugated MAb specific for rabbit immunoglobulin G
(IgG; Sigma Aldrich). The MAb 2.2.1 was kindly supplied by D. Godson (Vet-
erinary Infectious Disease Organization, Saskatoon, Saskatchewan, Canada).
The spots were visualized with 5-bromo-4-chloro-3-indolylphosphate–nitroblue
tetrazolium substrate (Sigma Aldrich).
The involvement of CD11b was determined by addition (50 ?l of ELISPOT
plate/well) of the mouse MAbs CC94 and ILA15 (both IgG1; kindly provided by
C. Howard, Institute for Animal Health, Compton, United Kingdom) to 2 ? 105
PBMC dispensed in 100 ?l. After 30 min of preincubation at 37°C, serial dilu-
tions of CyaA336/CFP10 were added and the cultures were incubated for 24 h as
described above, after which time ELISPOT analysis was performed.
CD4?- and CD8?-T-cell subpopulations were depleted by magnetic negative
selection with the anti-bovine CD4- or CD8-specific MAbs CC30 and CC58 (C.
Howard, Institute for Animal Health) in conjunction with the MACS system
(goat anti-mouse IgG-coated beads, LS separation columns; Miltenyi Biotec
Ltd., Bergisch-Gladbach, Germany) as described previously (30).
IFN-? assay. Whole-blood cultures were performed in 96-well plates in 0.2-
ml/well aliquots by mixing 0.1 ml of heparinized blood with an equal volume of
antigen-containing solution. Supernatants were harvested after 24 h of culture,
and the levels of IFN-? present were determined using the Bovigam enzyme
immunoassay (EIA) kit (CSL, Melbourne, Australia) (29, 34). The data are
expressed as units of optical density at 450 nm (?OD450) (OD450? 1,000).
Background levels obtained after stimulation with control CyaA were subtracted
from values obtained after stimulation with CyaA336/ESAT-6 and CyaA336/
CFP10. The cutoff for positivity was determined by assessing responses of BCG-
vaccinated animals (cutoff, 125 ?OD450units).
CFP10-specific bovine CD4?-T-cell line. A long-term T-cell line was estab-
lished from PBMC collected from a calf experimentally infected with M. bovis
(32). Briefly, PBMC (2 ? 106/well of 24-well plates in a 1-ml aliquot) were
cultured in the presence of recombinant CFP10 (2 ?g/ml) for 7 days. Viable cells
were then isolated by Histopaque 1077 (Sigma) gradient centrifugation, and cells
were rested for 1 week in the presence of mitomycin C (Sigma)-treated freshly
prepared autologous PBMC. Cells (2 ? 105/well) were then restimulated in the
presence of CFP10 protein (2 ?g/ml) and mitomycin C-treated PBMC (106/ml)
as a source of antigen-presenting cells (APC) for 5 days, after which cells were
collected and rested as described above. After three such stimulation and rest
cycles, the specificity of this T-cell line was established by incubating 2 ? 104T
cells with 105mitomycin C-treated PBMC in 0.2-ml portions in the presence of
antigen. IFN-? production was determined in culture supernatants harvested 2
days later as described above by using the Bovigam EIA kit. The line was
composed of CD4?T cells because magnetic depletion of CD4?cells reduced
antigen-specific IFN-? production by 99.7%, whereas depletion of CD8?T cells
or WC1??? T cells did not have any effects (data not shown).
Inhibition of T-cell recognition by chloroquine. PBMC were depleted of CD4?
T cells by using the bovine CD4-specific MAb CC30 (kindly donated by C.
Howard) in conjunction with the Miltenyi magnetic sorting system (goat anti-
mouse IgG-coated beads, LS separation columns) and mitomycin C treatment.
These APC were plated at 105/well in 96-well microtiter plates and incubated
with CFP10 or CyaA336/CFP10 (at 3.7 nM) for 0, 60, 120, and 240 min before
the addition of chloroquine (Sigma) to a final concentration of 5 mM (33) and 2
? 104cells of the CFP10-specific T-cell line described above per well. Superna-
tants were harvested after 2 days of culture, and their IFN-? contents were
determined by IFN-? EIA as described above.
6256VORDERMEIER ET AL.INFECT. IMMUN.
Statistical analysis. Statistical analysis was performed using Instat v3.0a
(GraphPad, San Diego, Calif.) on an iMac personal computer. Data were ana-
lyzed using the one- or two-tailed Wilcoxon signed rank matched pairs test. See
the figure legends for further details.
IFN-? responses of experimentally infected cattle. PBMC
were prepared from experimentally infected cattle and incu-
bated with serial dilutions of antigens (recombinant ESAT-6,
CFP10, CyaA336/ESAT-6, CyaA336/CFP10, and CyaA con-
trol), and the antigen-induced IFN-? responses were deter-
mined after 24 h of culture by ELISPOT assay. To illustrate
how the data are subsequently expressed, representative re-
sults for CFP10 and CyaA336/CFP10 tested in one calf are
given in Fig. 1. Responses to CyaA alone in this cow were ?14
spot-forming cells (SFC)/2 ? 105cells at all concentrations
tested (data not shown), and these values were subtracted from
the CyaA336/CFP10 responses. CyaA336/CFP10 both induced
a higher peak response than did recombinant CFP10 (as shown
by comparison of values indicated by horizontal lines a and b in
Fig. 1) and was recognized more effectively as indicated by the
vertical lines d and e, representing the concentrations required
for half-maximum (50% of peak responses) responses induced
with the recombinant protein (line c).
A further six experimentally M. bovis-infected calves were
then tested. Responses to control CyaA in these calves were
below 10% of those determined with the fusion proteins and
were subtracted from the responses of CyaA336/CFP10 and
CyaA/ESAT-6. As demonstrated by the comparison of peak
responses induced by CFP10 and by CyaA336/CFP10 and the
reduced concentration needed for 50% maximal responses, the
CyaA336/CFP10 fusion protein was again superior to its non-
fusion counterpart (Fig. 2). CyaA336/CFP10 peak responses
were about twice as high as those observed with CFP10 (me-
dian responses: CyaA336/CFP10, 157 SFC; CFP10, 75 SFC; P
? 0.03), and CyaA336/CFP10 was recognized about 20 times
more efficiently than was CFP10 (50% maximum concentra-
tions: CyaA336/CFP10, 0.3 nM; CFP10, 6.25 nM; P ? 0. 017).
The IFN-? responses induced by ESAT-6 and by the CyaA336/
ESAT-6 fusion proteins were not significantly different (Fig.
2). Recombinant ESAT-6 was about 70 times more efficiently
recognized than CFP10 was (median of 50% maximum con-
centrations: 0.09 with ESAT-6 compared to 6.25 with CFP10).
FIG. 1. Dose-response relationship of in vitro IFN-? production after
stimulation with CFP10 (squares) and CyaA336/CFP10 (triangles). The
readout system was the IFN-? ELISPOT assay. SFC numbers from cul-
tures with medium alone were subtracted from all values. In addition, the
numbers of spots after incubation with CyaA alone were subtracted from
the SFC induced by CyaA336/CFP10 stimulation. Tests were performed
in duplicate with 2 ? 105PBMC/well isolated from an M. bovis-infected
calf. Horizontal lines indicate the maximum SFC numbers (peak values)
induced after stimulation with CyaA336/CFP10 (a) and CFP10 (b); line c
indicates the half-maximum SFC induced after CFP10 stimulation (50%
maximum values). Vertical lines indicate the CyaA336/CFP10 (d) and
CFP10 (e) concentrations required to induce 50% of CFP10-induced
peak responses (50% maximum concentration).
FIG. 2. Comparison of abilities of CyaA fusion proteins and recombinant proteins to stimulate in vitro IFN-? production by PBMC from M.
bovis-infected cattle. Left panel: 50% maximum concentrations determined as illustrated in Fig. 1. Right panel: peak values determined as illustrated in
Fig. 1. The readout system was the IFN-? ELISPOT assay. SFC numbers from cultures with medium alone were subtracted from all values. In addition,
were performed in duplicate with 2 ? 105PBMC/well. *, P ? 0.05 (two-tailed Wilcoxon signed rank matched pairs test).
VOL. 72, 2004MYCOBACTERIAL ANTIGENS DELIVERED BY INACTIVATED CyaA 6257
Performance of CyaA336/ESAT-6 and CyaA336/CFP10 in
whole-blood IFN-? tests (Bovigam assay). A bovine IFN-? test
is applied as a diagnostic assay in the field in the format of a
whole-blood assay (Bovigam test). In this format heparinized
blood is incubated in the presence of antigens for 24 h and the
amount of antigen-induced IFN-? in plasma supernatants is
determined by enzyme-linked immunosorbent assay (ELISA).
To determine the performance of the CyaA fusion proteins
with ESAT-6 and CFP10 in the Bovigam assay, blood was
obtained from 11 BCG-vaccinated animals and eight calves
experimentally infected with M. bovis. These blood samples
were stimulated with bovine tuberculin PPD (PPD-B), as well
as ESAT-6, CFP10, CyaA336/ESAT-6, and CyaA336/CFP10.
Strong IFN-? responses were demonstrated in 11 of 11 vacci-
nated calves after stimulation with PPD-B. Stimulation with
ESAT-6 and CFP10 resulted in no or only marginal IFN-?
production (Table 1), nor did stimulation with CyaA336/
ESAT-6 or CyaA336/CFP10 result in IFN-? production (Table
1). After assessment of the data by receiver operating charac-
teristic analysis (including the data from the infected animals
discussed below), a cutoff for positivity for this assay was set at
125 ?OD450units. Applying this cutoff, all 11 animals tested
were classified positive after PPD-B stimulation, while none of
the BCG-vaccinated animals tested positive after stimulation
with ESAT-6, CFP10 at a 20 nM antigen concentration, and
CyaA336/ESAT-6 or with CyaA336/CFP10 when used at 4 nM
(Table 1). Only 1 of 11 animals tested positive after stimulation
with CyaA336/ESAT-6 and CyaA336/CFP10 at 20 nM (Table
1). Taken together, these results confirmed the specific nature
of these antigens.
The results of the ELISA conducted in experimentally in-
fected calves are shown in Fig. 3. As shown above for PBMC
responses measured by ELISPOT, significantly stronger IFN-?
responses were observed with CyaA336/CFP10 at both test
concentrations than with recombinant CFP10 protein (P ?
0.0078 at both concentrations). These increased responses
were particularly evident when the blood was stimulated at 4
nM antigen concentrations. While the responses for CyaA336/
ESAT-6 and ESAT-6 were not significantly different at 20 nM,
significantly elevated responses were observed after stimula-
tion with CyaA336/ESAT-6 at 4 nM (P ? 0.015).
When the diagnostic outcome was evaluated using the com-
monly applied cutoff of 125 ?OD450units for the Bovigam
assay, six of eight tested animals were deemed positive for BTB
by using ESAT-6 and CyaA336/ESAT-6 applied at both test
concentrations (Fig. 3). In contrast, the use of CyaA336/CFP10
improved the sensitivity of CFP10, because, respectively, seven
of eight and six of eight animals tested positive at 20 and 4 nM
test concentrations with CyaA336/CFP10, whereas six of eight
and three of eight, respectively, were classified as positive after
stimulation with recombinant CFP10 at corresponding test
concentrations (Fig. 3).
One of the eight animals was skin test negative and pre-
sented without tuberculous lesions at postmortem examina-
tions carried out several months after this experiment was
performed; in addition M. bovis could not be detected by cul-
ture from tissue samples taken at the postmortem examination.
Taken together, this suggests that the experimental infection in
this animal was contained and did not result in disease. Con-
sistent with this diagnosis, no IFN-? was induced in the blood
of this calf after stimulation with PPD-B, CyaA336/ESAT-6,
TABLE 1. IFN-? responses in BCG-vaccinated calvesa
No. of calves
959 ? 218
32 ? 11
20 ? 11
98 ? 48
3 ? 15
59 ? 32
12 ? 48
aHeparinized blood from 11 BCG-vaccinated calves was incubated with the
antigens. IFN-? production was determined by ELISA.
bThe results are expressed as mean ?OD450units ? standard errors of the
means (OD450? 1,000) with ?OD450values from cultures with medium alone
subtracted from all values including wells stimulated with CyaA alone. In addi-
tion, the ?OD450values with CyaA were subtracted from wells with CyaA336/
CFP10 or CyaA336/ESAT-6 stimulation.
cCutoff, 125 ?OD450units.
FIG. 3. Performance of CyaA fusion proteins and recombinant
ESAT-6 and CFP10 in the whole-blood Bovigam IFN-? assay. Hepa-
rinized blood from eight M. bovis-infected calves was incubated with
the antigens at 4 and 20 nM test concentrations. IFN-? in plasma
culture supernatants was determined by ELISA. The results are ex-
pressed as ?OD450units (OD450? 1,000) with ?OD450values from
cultures with medium alone subtracted from all values. In addition, the
?OD450values after incubation with CyaA alone were subtracted from
wells with CyaA336/CFP10 or CyaA336/ESAT-6 stimulation. The hor-
izontal line indicates cutoff for positivity (?125 OD450units). Cultures
were performed in duplicate in 96-well flat-bottomed plates. *, P ?
0.05; **, P ? 0.01 (two-tailed Wilcoxon signed rank matched pairs
6258VORDERMEIER ET AL.INFECT. IMMUN.
CyaA336/CFP10, ESAT-6, or CFP10, thus highlighting the
specificity of these reagents (Fig. 3).
Mechanisms of the enhanced responses. To determine
whether the presentation of CyaA336/CFP10 to bovine T cells
is mediated via a CD11b-dependent mechanism, PBMC from
an infected calf were stimulated with CyaA336/CFP10 in the
presence of two MAbs of the same isotype (IgG1) specific for
bovine CD11b. One of these MAbs (ILA15) interfered with
the interaction of CyaA336/CFP10 with CD11b, as the number
of SFC was reduced significantly, whereas the nonblocking
isotype control MAb (CC94) did not modify this response (Fig.
4, P ? 0.02 for each concentration tested; one-tailed Wilcoxon
matched pairs test). These results therefore provide evidence
that interaction between CyaA and CD11b on APC is required
in order to increase the CFP10-specific T-cell responses.
To determine whether CD11b-mediated uptake of CyaA
fusion proteins also resulted in accelerated processing and
presentation of the antigen, we investigated the effects of in-
hibiting the MHC class II processing pathway. Chloroquine
was added between 0 and 240 min after the coculture of
CyaA336/CFP10 or CFP10 and APC to stop further antigen
processing. Then cells from a CFP10-specific CD4?-T-cell line
were added and IFN-? production was determined. As ex-
pected, the recognition of both CFP10 and CyaA336/CFP10 by
CD4?T cells was chloroquine sensitive, as addition of chloro-
quine at the same time as the proteins completely inhibited
IFN-? production (Fig. 5). However, CyaA336/CFP10 was pre-
sented at an accelerated rate compared to that of CFP10 be-
cause a significant IFN-? response to CyaA/CFP10 was ob-
served when processing was inhibited after 120 to 240 min
whereas processing of CFP10 was not sufficient at these time
points to allow IFN-? production (Fig. 5).
Based on the experiments described above, we conclude that
fusion of ESAT-6 and CFP10 with CyaA improved antigen
recognition. These results extend previous observations made
in the murine model (18) where 100-times-higher molar effi-
ciency has been described for the CyaA fusion protein. This
constitutes a novel approach to the generation of potential in
vitro diagnostic reagents for the detection of BTB in cattle. In
particular the observation that they can be used in a whole-
blood format is encouraging, as this is the format of choice
when contemplating in vitro diagnosis of BTB. Our results
demonstrate that, compared to the recombinant protein, pre-
sentation in vitro of CFP10 in the form of a CyaA fusion
protein resulted in a significant improvement of IFN-? re-
sponses. A recombinant CyaA carrying a human melanoma
epitope was also shown to be 100-fold more efficient than the
synthetic peptide in inducing the presentation of the epitope to
specific human T cells by dendritic cells (7). In addition, a
recent study testing human TB patients with CyaA336/ESAT-6
and CyaA336/CFP10 also showed 10-times-higher molar effi-
ciencies of recognition compared to those of purified recom-
binant CFP10 and ESAT-6 (K. A. Wilkinson et al., unpub-
lished data). It has to be highlighted that CyaA336/CFP10
when used in the whole-blood test format not only increased
the strength of IFN-? responses but also resulted in increased
test sensitivity by detecting more of the truly infected animals
than recombinant CFP10 alone did. Again this result is in
agreement with the results obtained with human TB patients
(K. A. Wilkinson et al., unpublished). Interestingly, and in
contrast to the results with human TB patients, CyaA336/
ESAT-6 did not increase the molar efficiency of its recognition
in cattle compared to the recombinant ESAT-6 protein. Re-
combinant ESAT-6 was about 70 times more efficiently recog-
FIG. 4. Involvement of CD11b in the recognition of CyaA336/
CFP10. Cultures were performed in the presence of a blocking CD11b-
specific IgG1 MAb (ILA15) and an isotype control (CC94). The read-
out system was the IFN-? ELISPOT assay. SFC numbers from cultures
with medium alone were subtracted from all values. Tests were per-
formed in duplicate with 2 ? 105PBMC/well isolated from one in-
fected calf. Responses are significantly different (P ? 0.02) for each
concentration tested except lowest concentration (one-tailed Wilcoxon
signed rank matched pairs test).
FIG. 5. Accelerated presentation kinetics of CyaA336/CFP10 to
CD4?T cells. CyaA336/CFP10 and CFP10 (both at 3.7 nM) were
incubated with CD4-depleted PBMC as a source of APC. At the
indicated time points, chloroquine was added to stop further process-
ing. APC were then cultured for 48 h in the presence of 2 ? 104cells
from a CFP10-specific CD4?-T-cell line added. IFN-? in culture su-
pernatants was determined by ELISA. Triangles, CFP10; circles,
CyaA336/CFP10; open symbols, no addition of chloroquine; closed
symbols, chloroquine was added at indicated time points. Box, addition
of chloroquine and antigens simultaneously. Results are expressed as
means of triplicate determinations ? standard errors.
VOL. 72, 2004 MYCOBACTERIAL ANTIGENS DELIVERED BY INACTIVATED CyaA6259
nized than CFP10 was, and this difference in the efficiency of
recognition between those two proteins might explain why an
additional benefit of presenting ESAT-6 as a CyaA fusion
protein could not be realized in these experiments. Interest-
ingly, preliminary results in our laboratory suggest that
ESAT-6 does not need to be lysosomally processed to induce a
response in vitro, as we observed strong responses (?50% of
responses observed without addition of chloroquine) when we
prevented processing by adding chloroquine prior to antigen
addition (H. M. Vordermeier and A. O. Whelan, unpublished
data). This observation, if confirmed by further studies, could
therefore also account for the failure of CyaA336/ESAT-6 to
improve in vitro recognition. Generation of MHC class II-
restricted epitopes by extracellular processing has been de-
scribed for epitopes of, e.g., hepatitis ? antigen (1).
A number of previous reports have conclusively shown that,
in mice, CyaA fusion proteins facilitate the delivery of CD8?-
T-cell epitopes directly into the cytosol of dendritic cells both
in vitro and in vivo (13, 14). CyaA fusion proteins are therefore
excellent vehicles to trigger class I-restricted CD8?T cells in
vitro and in vivo (7, 11, 12, 15, 21, 25, 27). In addition, murine
studies have shown that MHC class II-restricted CD4?-T-cell
epitopes could also be efficiently delivered by CyaA fusion
proteins (8, 18). This was confirmed in the present study where
we detected predominantly responses of CD4?T cells (Fig. 5
and data not shown). We also demonstrated that this enhance-
ment of T-cell responses by CyaA fusion proteins depends
upon their interaction with CD11b in agreement with the re-
sults obtained in mice (9). Furthermore, we demonstrate that
the processing and presentation of CyaA fusion proteins to
CD4?T cells via the MHC class II pathway are chloroquine
sensitive and that CyaA fusion proteins are presented at ac-
celerated kinetics compared to the nonfusion proteins (Fig. 5).
Taken together, the CD11b-mediated uptake and improved
presentation kinetics could account for the superior recogni-
tion of CyaA336/CFP10 compared to recombinant CFP10.
We also determined whether CyaA336/CFP10 was recog-
nized by bovine CD8?T cells. This was analyzed by depleting
either subpopulation with magnetic beads. However, the cattle
used in this study were at early stages of BTB, and it has been
reported that such animals display only weak or undetectable
CD8?-T-cell responses (24; H. M. Vordermeier, unpublished
observation). Consequently, we observed significant PPD-B-
and CFP10-specific CD8?-T-cell responses only in one of four
cows tested. Nevertheless, the results obtained indicated that
CyaA336/CFP10 induced higher in vitro CD8?-T-cell re-
sponses than did the recombinant protein (data not shown).
These results are also in line with results obtained using the
same CyaA constructs to test human TB patients (K. A.
Wilkinson et al., unpublished).
One important aspect to determine when considering the
practical diagnostic application of the CyaA fusion proteins
described in the present study is whether background re-
sponses to the CyaA backbone protein can be found in cattle.
We determined this by inclusion of the CyaA toxoid as vector
control in our assays and by stratifying the data obtained with
the fusion proteins by subtracting IFN-? responses obtained
with this control. Interestingly, only one cow tested displayed
significant responses to CyaA. Therefore, we conclude that
background responses to CyaA are no obstacle to the practical
application of CyaA fusion proteins in cattle, as they can be
easily controlled and distinguished from ESAT-6- or CFP10-
specific responses by the inclusion of CyaA alongside the fu-
In conclusion, the present study has demonstrated that the
fusion of mycobacterial proteins to CyaA improved their rec-
ognition in vitro by increasing both signal strength and molec-
ular efficacy of recognition. In addition, when applied to the
whole-blood IFN-? test format used routinely to diagnose BTB
in cattle, the CyaA336/CFP10 fusion protein increased test
sensitivity compared to conventional CFP10, although precise
cutoffs have to be defined to confirm these results in larger
groups of naturally infected animals. Increased sensitivity and
reduction in the amount of antigens needed to perform diag-
nostic tests are important attributes for immunodiagnostic re-
agents and prioritize these reagents for further evaluation to
establish their performance in a larger field trial.
This study was funded by the Department for Environment, Food
and Rural Affairs, United Kingdom, with additional support from the
Wellcome Trust. P.S. was supported in part by grant no. S5020311
from the Academy of Sciences of the Czech Republic.
We express our appreciation to the staff of the Animal Services Unit
at VLA, in particular Derek Clifford, for their dedication to the wel-
fare of test animals. We also thank C. Howard, Institute for Animal
Health, Compton, United Kingdom, for the supply of MAbs, and M.
Singh, Lionex Ltd., Braunschweig, Germany, for the supply of recom-
1. Accapezzato, D., R. Nisini, M. Paroli, G. Bruno, F. Bonino, M. Houghton,
and V. Barnaba. 1998. Generation of an MHC class II-restricted T cell
epitope by extracellular processing of hepatitis delta antigen. J. Immunol.
2. Behr, M. A., M. A. Wilson, W. P. Gill, H. Salamon, G. K. Schoolnik, S. Rane,
and P. M. Small. 1999. Comparative genomics of BCG vaccines by whole-
genome DNA microarray. Science 284:1520–1523.
3. Buddle, B., N. A. Parlane, D. L. Keen, F. E. Aldwell, J. M. Pollock, K.
Lightbody, and P. Andersen. 1999. Differentiation between Mycobacterium
bovis BCG-vaccinated and M. bovis-infected cattle by using recombinant
mycobacterial antigens. Clin. Diagn. Lab. Immunol. 6:1–5.
4. Buddle, B. M., A. R. McCarthy, T. J. Ryan, J. M. Pollock, H. M. Vorder-
meier, R. G. Hewinson, P. Andersen, and G. W. de Lisle. 2003. Use of
mycobacterial peptides and recombinant proteins for the diagnosis of bovine
tuberculosis in skin test-positive cattle. Vet. Rec. 153:615–620.
5. Cosivi, O., J. M. Grange, C. J. Daborn, M. C. Raviglione, T. Fujikura, D.
Cousins, R. A. Robinson, H. F. Huchzermeyer, I. de Kantor, and F. X.
Meslin. 1998. Zoonotic tuberculosis due to Mycobacterium bovis in develop-
ing countries. Emerg. Infect. Dis. 4:59–70.
6. Cosivi, O., F. X. Meslin, C. J. Daborn, and J. M. Grange. 1995. Epidemiology
of Mycobacterium bovis infection in animals and humans, with particular
reference to Africa. Rev. Sci. Technol. 14:733–746.
7. Dadaglio, G., S. Morel, C. Bauche, Z. Moukrim, F. A. Lemonnier, B. J. Van
Den Eynde, D. Ladant, and C. Leclerc. 2003. Recombinant adenylate cyclase
toxin of Bordetella pertussis induces cytotoxic T lymphocyte responses against
HLA*0201-restricted melanoma epitopes. Int. Immunol. 15:1423–1430.
8. Dadaglio, G., Z. Moukrim, R. Lo-Man, V. Sheshko, P. Sebo, and C. Leclerc.
2000. Induction of a polarized Th1 response by insertion of multiple copies
of a viral T-cell epitope into adenylate cyclase of Bordetella pertussis. Infect.
9. El-Azami-El-Idrissi, M., C. Bauche, J. Loucka, R. Osicka, P. Sebo, D.
Ladant, and C. Leclerc. 2003. Interaction of Bordetella pertussis adenylate
cyclase with CD11b/CD18: role of toxin acylation and identification of the
main integrin interaction domain. J. Biol. Chem. 278:38514–38521.
10. Fayolle, C., D. Ladant, G. Karimova, A. Ullmann, and C. Leclerc. 1999.
Therapy of murine tumors with recombinant Bordetella pertussis adenylate
cyclase carrying a cytotoxic T cell epitope. J. Immunol. 162:4157–4162.
11. Fayolle, C., A. Osickova, R. Osicka, T. Henry, M. J. Rojas, M. F. Saron, P.
Sebo, and C. Leclerc. 2001. Delivery of multiple epitopes by recombinant
detoxified adenylate cyclase of Bordetella pertussis induces protective antivi-
ral immunity. J. Virol. 75:7330–7338.
12. Guermonprez, P., C. Fayolle, G. Karimova, A. Ullmann, C. Leclerc, and D.
6260 VORDERMEIER ET AL.INFECT. IMMUN.
Ladant. 2000. Bordetella pertussis adenylate cyclase toxin: a vehicle to deliver
CD8-positive T-cell epitopes into antigen-presenting cells. Methods Enzy-
13. Guermonprez, P., C. Fayolle, M. J. Rojas, M. Rescigno, D. Ladant, and C.
Leclerc. 2002. In vivo receptor-mediated delivery of a recombinant invasive
bacterial toxoid to CD11c?CD8alpha-CD11b high dendritic cells. Eur.
J. Immunol. 32:3071–3081.
14. Guermonprez, P., N. Khelef, E. Blouin, P. Rieu, P. Ricciardi-Castagnoli, N.
Guiso, D. Ladant, and C. Leclerc. 2001. The adenylate cyclase toxin of
Bordetella pertussis binds to target cells via the alpha(M)beta(2) integrin
(CD11b/CD18). J. Exp. Med. 193:1035–1044.
15. Guermonprez, P., D. Ladant, G. Karimova, A. Ullmann, and C. Leclerc.
1999. Direct delivery of the Bordetella pertussis adenylate cyclase toxin to the
MHC class I antigen presentation pathway. J. Immunol. 162:1910–1916.
16. Hewinson, R. G., H. M. Vordermeier, and B. M. Buddle. 2003. Use of the
bovine model of tuberculosis for the development of improved vaccines and
diagnostics. Tuberculosis (Edinburgh) 83:119–130.
17. Krebs, J. R. 1997. Bovine tuberculosis in cattle and badgers. Ministry of
Agriculture, Fisheries and Food Publications, London, United Kingdom.
18. Loucka, J., G. Schlecht, J. Vodolanova, C. Leclerc, and P. Sebo. 2002.
Delivery of a MalE CD4?-T-cell epitope into the major histocompatibility
complex class II antigen presentation pathway by Bordetella pertussis adenyl-
ate cyclase. Infect. Immun. 70:1002–1005.
19. Mahairas, G. G., P. J. Sabo, M. J. Hickey, D. C. Singh, and C. K. Stover.
1996. Molecular analysis of genetic differences between Mycobacterium bovis
BCG and virulent M. bovis. J. Bacteriol. 178:1274–1282.
20. Moron, G., G. Dadaglio, and C. Leclerc. 2004. New tools for antigen delivery
to the MHC class I pathway. Trends Immunol. 25:92–97.
21. Osicka, R., A. Osickova, T. Basar, P. Guermonprez, M. Rojas, C. Leclerc,
and P. Sebo. 2000. Delivery of CD8?-T-cell epitopes into major histocom-
patibility complex class I antigen presentation pathway by Bordetella pertussis
adenylate cyclase: delineation of cell invasive structures and permissive in-
sertion sites. Infect. Immun. 68:247–256.
22. Pollock, J. M., and P. Andersen. 1997. The potential of the ESAT-6 antigen
secreted by virulent mycobacteria for specific diagnosis of tuberculosis. J. In-
fect. Dis. 175:1251–1254.
23. Pollock, J. M., R. M. Girvin, K. A. Lightbody, R. A. Clements, S. D. Neill,
B. M. Buddle, and P. Andersen. 2000. Assessment of defined antigens for the
diagnosis of bovine tuberculosis in skin test-reactor cattle. Vet. Rec. 146:
24. Pollock, J. M., D. A. Pollock, D. G. Campbell, R. M. Girvin, A. D. Crockard,
S. D. Neill, and D. P. Mackie. 1996. Dynamic changes in circulating and
antigen-responsive T-cell subpopulations post-Mycobacterium bovis infection
in cattle. Immunology 87:236–241.
25. Saron, M. F., C. Fayolle, P. Sebo, D. Ladant, A. Ullmann, and C. Leclerc.
1997. Anti-viral protection conferred by recombinant adenylate cyclase tox-
ins from Bordetella pertussis carrying a CD8?T cell epitope from lymphocytic
choriomeningitis virus. Proc. Natl. Acad. Sci. USA 94:3314–3319.
26. Sebo, P., C. Fayolle, O. d’Andria, D. Ladant, C. Leclerc, and A. Ullmann.
1995. Cell-invasive activity of epitope-tagged adenylate cyclase of Bordetella
pertussis allows in vitro presentation of a foreign epitope to CD8?cytotoxic
T cells. Infect. Immun. 63:3851–3857.
27. Sebo, P., Z. Moukrim, M. Kalhous, N. Schaft, G. Dadaglio, V. Sheshko, C.
Fayolle, and C. Leclerc. 1999. In vivo induction of CTL responses by recom-
binant adenylate cyclase of Bordetella pertussis carrying multiple copies of a
viral CD8(?) T-cell epitope. FEMS Immunol. Med. Microbiol. 26:167–173.
28. Vordermeier, H. M., M. A. Chambers, P. J. Cockle, A. O. Whelan, J. Sim-
mons, and R. G. Hewinson. 2002. Correlation of ESAT-6-specific gamma
interferon production with pathology in cattle following Mycobacterium bovis
BCG vaccination against experimental bovine tuberculosis. Infect. Immun.
29. Vordermeier, H. M., P. C. Cockle, A. Whelan, S. Rhodes, N. Palmer, D.
Bakker, and R. G. Hewinson. 1999. Development of diagnostic reagents to
differentiate between Mycobacterium bovis BCG vaccination and M. bovis
infection in cattle. Clin. Diagn. Lab. Immunol. 6:675–682.
30. Vordermeier, H. M., P. J. Cockle, A. O. Whelan, S. Rhodes, M. A. Chambers,
D. Clifford, K. Huygen, R. Tascon, D. Lowire, M. J. Colston, and R. G.
Hewinson. 2001. Effective DNA vaccination of cattle with the mycobacterial
antigens MPB83 and MPB70 does not compromise the specificity of the
comparative intradermal tuberculin skin test. Vaccine 19:1246–1255.
31. Vordermeier, H. M., A. Whelan, P. J. Cockle, L. Farrant, N. Palmer, and
R. G. Hewinson. 2001. Use of synthetic peptides derived from the antigens
ESAT-6 and CFP-10 for differential diagnosis of bovine tuberculosis in
cattle. Clin. Diagn. Lab. Immunol. 8:571–578.
32. Vordermeier, M., A. O. Whelan, and R. G. Hewinson. 2003. Recognition of
mycobacterial epitopes by T cells across mammalian species and use of a
program that predicts human HLA-DR binding peptides to predict bovine
epitopes. Infect. Immun. 71:1980–1987.
33. Wilkinson, K. A., F. Hudecz, H. M. Vordermeier, J. Ivanyi, and R. J. Wilkin-
son. 1999. Enhancement of the T cell response to a mycobacterial peptide by
conjugation to synthetic branched polypeptide. Eur. J. Immunol. 29:2788–
34. Wood, P. R., and J. S. Rothel. 1994. In vitro immunodiagnostic assays for
bovine tuberculosis. Vet. Microbiol. 40:125–135.
Editor: J. T. Barbieri
VOL. 72, 2004MYCOBACTERIAL ANTIGENS DELIVERED BY INACTIVATED CyaA6261