Levine, D. Hone, M. Posada, R. A. Strugnell, and G. Dougan. 1992. Construction .... Townsend, A. R. M., T. Elliot, V. Cerundolo, L. Foster, B. Barber, and A. Tse.
INFECTION AND IMMUNITY, Dec. 1993, p. 5374-5380
Vol. 61, No. 12
0019-9567/93/125374-07$02.00/0 Copyright ) 1993, American Society for Microbiology
Salmonella typhimurium AaroA AaroD Mutants Expressing a Foreign Recombinant Protein Induce Specific Major Histocompatibility Complex Class I-Restricted Cytotoxic T Lymphocytes in Mice S. J. TURNER,1 F. R. CARBONE,* AND R. A. STRUGNELL2 Department of Microbiology, University of Melbourne, Parkville, Victoria 3152,2 and Department of Pathology and Immunology, Monash Medical School, Prahran, Victonia 3181,1 Australia Received 8 June 1993/Returned for modification 2 August 1993/Accepted 24 September 1993
Recombinant SalmoneUa typhimurium aroA aroD mutants which expressed ovalbumin were constructed. The two expression constructs used were based on either pUC18 or pBR322. The pBR322-based construct was more stable in vitro and in vivo than the pUC-based construct. Salmonellae containing the stable pBR322-based plasmid induced major histocompatibility complex (MHC) class I-restricted cytotoxic T lymphocytes (CTL), in contrast to salmonellae containing the pUC18-based expression construct. The priming of MHC class I-restricted CTL was increased by multiple immunizations. The study described in this report suggests that S. typhimurium Aaro mutants have the capacity to induce MHC class I-restricted CTL against carried antigens and that MHC class I-restricted CTL responses require stable in vivo expression of the target antigen. Further, the results indicate that the SalmoneUa typhi Aaro mutants currently undergoing evaluation in studies with humans may be good carriers of viral antigens with CTL determinants.
eases. For this to occur, activation of the appropriate T-lymphocyte responses is needed. Products expressed from the major histocompatibility complex (MHC) locus present peptides for recognition by circulating T lymphocytes. It is generally accepted that MHC class II molecules present peptides derived from exogenous antigens to CD4+ T helper lymphocytes (4, 26). Conversely, MHC class I molecules present peptides derived from endogenous antigens to CD8+ cytotoxic T lymphocytes (CTL) (34). Recognition, resulting in activation, occurs via a receptor found on the T lymphocytes which is specific for the MHC-peptide complex. Activation of CD4+ MHC class II-restricted T helper lymphocytes results in the production of a vast array of lymphokines, giving rise to both humoral and cell-mediated (delayed-type hypersensitivity) responses. It has been demonstrated that recombinant, attenuated vaccine strains of Salmonella typhimurium can induce both humoral and cellmediated responses to specific antigens (3). The model for activation of CD8+ MHC class I-restricted CTL requires the cytoplasmic localization of foreign proteins for processing and then presentation of peptides via the MHC class I molecule (25, 27). Viruses replicate intracellularly, and as a consequence, the viral products are processed and presented for recognition by CD8+ MHC class I-restricted CTL (6, 37). If recombinant attenuated bacteria were used to express viral antigens, then induction of an MHC class I-restricted response would be essential for protection against subsequent viral infection. According to past studies, the ability of recombinant S. typhimunium Aaro mutants to elicit a specific MHC class I-restricted response is equivocal. The purposes of our investigation were to determine whether S. typhimunium Aaro mutants carrying a heterologous antigen, ovalbumin, could elicit an MHC class I-restricted response specific for ovalbumin and to examine the kinetics of CTL induction.
Salmonella spp. are intracellular pathogens of humans and animals. The organisms usually reside and replicate within phagosomes of professional phagocytes, especially macrophages (22). Immunity against Salmonella infections is predominantly cell mediated, as antibody responses which arise following vaccination with killed organisms often afford only limited protection (13, 17). The variable efficacies of killed whole-cell vaccines and problems with vaccine site reactions have fostered the search for new anti-Salmonella vaccines. The traditional human and veterinary salmonellosis vaccines will almost certainly be replaced by live, rationally attenuated Salmonella spp. which have the added benefit of being able to express heterologous antigens (7, 16). Salmonella spp. with deletion and/or insertion mutations in genes encoding enzymes in the prechorismate pathway have been generated for this purpose and studied extensively with mice (3, 16), cattle (29), chickens (8, 14), and, more recently, humans (31). These so-called Aaro mutants are unable to synthesize chorismate, which is a key intermediate in the synthesis of aromatic amino acids and compounds such as dihydroxybenzoate, which may act as a siderophore, and p-aminobenzoate. Significantly, Aaro mutants of Salmonella spp. remain attenuated in animals with decreased immunocompetence because their attenuation is mediated not by the immune system but rather by a requirement for compounds which are not synthesized by higher eucaryotes (16). With their potential as anti-Salmonella vaccines well established, Salmonella spp. Aaro mutant vaccines are also being evaluated for the ability to carry heterologous antigens to the immune system, with a view to developing multivalent vaccines (7). Live bacterial carriers of heterologous antigens must be capable of stimulating immune responses which are appropriate for protection against both viral and bacterial dis*
Corresponding author. 5374
S. TYPHIMURIUM MUTANTS INDUCE CLASS I-RESTRICTED CTL
VOL. 61, 1993
TABLE 1. Bacterial strains and plasmids Organism or plasmid
Strains LB5010 BRD509
Plasmids pUGova pKK233.2 pKKova
pDEL2 pDEL2ova
Bacteriophage P22HT/int-
Mutations or feature(s)
Reference or source
Rough LPS, restriction- and
30
modification' Wild-type (smooth) LPS, AaroA AaroD
G. Dougan
Plasmid containing cDNA encoding ovalbumin Ampr, ptrc promoter pKK containing ovalbumin cDNA aroC Ampr Kanr, based on pUC18 pDEL2 containing ptrcpromoted cDNA from pKKova Defective integration
25 Pharmacia This study 30
This study
G. Dougan
MATERIALS AND METHODS Bacterial strains and plasmids. The bacterial strains and plasmids used in this study are shown in Table 1. Expression of ovalbumin by recombinant S. typhimurium. Plasmids pKKova and pDEL2ova were electroporated into electrocompetent S. typhimunum LB5010 as described by Binotto et al. (2) with a Bio-Rad electroporation apparatus. The conditions of electroporation were 2.5 kV, 200 fl of resistance, 25 ,uF of capacitance, and an electric plate width of 0.2 cm (field strength of 12.5 kV/cm, pulse length of 5 ms). Transduction of the plasmids was performed by using the Salmonella-specific bacteriophage P22HT/int- (24). P22 lysates of the LB5010 strains were used to infect S. typhimurium BRD509 grown in 1 mM CaCl2, and transductants were selected on L agar plates (Luria broth supplemented with 1% bacteriological agar) containing 50 ,ug of ampicillin per ml with or without 30 ,ug of kanamycin per ml. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was carried out on a 7.5% acrylamide gel as described by Laemmli (21) with a Mini Protean II gel apparatus (Bio-Rad). Western immunoblotting was carried out according to the method of Towbin et al. (33). Transfer onto nitrocellulose was performed at 100 V for 1 h. Ovalbumin expression was detected with a polyclonal rabbit antiovalbumin serum, followed by swine anti-rabbit immunoglobulin conjugated to horseradish peroxidase (Dako), and was visualized with the chromogenic substrate 4-chloronaphthol (Bio-Rad) (60 mg of 4-chloro-1-naphthol, 15 ml of methanol, 100 ml of distilled H20, 60 pl of 30% H202). In vitro and in vivo stability studies. S. typhimunium BRD509 containing the plasmids pKK233.2, pKKova, pDEL2, and pDEL2ova was grown overnight at 370C with the appropriate antibiotic selection. These cultures were then diluted 10-6 on consecutive days without antibiotics and allowed to reach stationary phase by incubation overnight at 37°C. On each day the approximate number of generations was calculated and the percentage of cells carrying plasmids was determined by plating dilutions of the cells onto media with or without antibiotics. The relative stability of a plasmid was expressed as a plot of the propor-
5375
tion of recovered bacteria which were resistant to antibiotics against the approximate generation number. Mice were infected orally with S. typhimunum Aaro containing either pKKova or pDEL2ova and were killed on day 3, 5, 7, or 10 and splenectomized. Pelleted single-cell suspensions were made from the spleens and resuspended in distilled water, after which dilutions were plated onto media with or without antibiotics. Following overnight incubation, the percentage of cells carrying the plasmid was estimated for each construct. Immunization of mice and measurement of CTL responses. Female C57BL/6 mice 6 to 8 weeks of age were infected with 5 x 109 to 5 x 1010 salmonellae, which had been grown at 37°C overnight without shaking, by inoculation with a gastric lavage needle. Mice used as positive controls in CTL assays were injected intravenously with 1 x 107 to 2.5 x 107 splenocytes which had been loaded with ovalbumin (Sigma Chemical Co., St. Louis, Mo.) by osmotic lysis of pinosomes (5, 25). CTL assays were performed as previously described (25). Activation and proliferation of ovalbumin-specific CTL were done by first incubating splenocytes from infected mice with 3 x 10' irradiated (2 x 104 rads) EG.7-Ova cells (EL.4 cells transfected with an ovalbumin expression construct [25]). After 5 days, the spleen cells were mixed with 104 1Crlabelled EL.4 or EG.7-Ova cells at various effector:target ratios and incubated for 4 h, after which the supernatant was assayed for released 51Cr by gamma counting. Maximal release and spontaneous release were determined by incubation of labelled target cells with 1% Triton X-100 or RP10 medium alone. The percent specific lysis was calculated as follows: % specific lysis = [(sample count - spontaneous release)/(maximal release - spontaneous release)] x 100. Antibody assays. Mice were bled at weekly intervals from the orbital sinus after inoculation with 0.5 x 1010 to 3.5 x 1010 salmonellae. The serum fraction was reacted in an enzyme-linked immunosorbent assay (ELISA) in wells coated with either 1 jig of S. typhimurium lipopolysaccharide (LPS) (Sigma) per ml or 10 ,ug of ovalbumin (Sigma) per ml. Bound antibodies were detected with a biotinylated sheep anti-mouse immunoglobulin antibody and streptavidin-horseradish peroxidase conjugate, both titrated for optimal reactivity. The A492 was determined for each well after the addition of substrate.
RESULTS Cloning and expression of ovalbumin in S. typhimurium. The cDNA encoding ovalbumin (23) was subcloned into the procaryotic expression vector pKK233.2, yielding pKKova (Fig. 1). The expression cassette comprising the trc promoter, ribosome-binding site, and ovalbumin cDNA was excised from pKKova by digestion with EcoRI and ligated into the AaroC chromosomal replacement vector pDEL2, yielding pDEL2ova. Both orientations of the inserted expression cassette were obtained in pDEL2, but subsequent attempts to insert the expression cassette into the S. typhimurium chromosome at aroC by homologous recombination were unsuccessful (data not shown). The pDEL2 construct was subsequently used as an expression plasmid. The plasmids pKKova and pDEL2ova were electroporated into LB5010 and then transduced, by using bacteriophage P22MT/int- into strain BRD509 (24). Ovalbumin expression from pDEL2ova and pKKova in BRD509 was determined by Western immunoblotting (Fig. 2). Western immunoblots of S. typhimunium AaroA AaroD
5376
TURNER ET AL.
INFECT. IMMUN.
BamH EcoRI v \
Ncol Ncol
Ncol
digest pUGova with Ncol(partial)/Pstl and ligate ova into Ncol/PstI digested pKK 233.2
-
EcoRI Pstl
digest EcoRI, ligate ova/ptrc into pDEL2 Hindlil
Drall
Hindlil Hindill
Pstl
Smal
FIG. 1. Construction of pKKova and pDEL2ova. The cDNA encoding ovalbumin was cloned into pKK after restriction with NcoI (partial) and PstI. The ptrc promoted ovalbumin was subcloned from pKKova by using EcoRI and was ligated into the polylinker of the chromosomal integration vector pDEL2, forming pDEL2ova. Both pKKova and pDEL2ova were then electroporated into S. typhimurium BRD666 prior to P22HT/int--mediated transduction into S. typhimurium BRD509.
A
B
C
D
97.4 --P 66.2 -o45.0
-P j!''!
qmp
4 11
aft
3 1 -0 21.5 14.5
FIG. 2. Western blot of S. typhimurium BRD509 containing either pKKova or pDEL2ova. BRD509 recombinants containing either pDEL2 (A), pKKova (B), or pDEL2ova (C) were grown statically in L broth containing ampicillin (50 jig/ml). The lysates were first electrophoresed with an SDS-7.5% PAGE gel and then transferred to nitrocellulose. Ovalbumin expression was detected by using mouse antisera raised against ovalbumin. Migration of molecular standards is shown, with ovalbumin being detected at 45.0 kDa (lane D).
containing either pKKova or pDEL2ova with antiovalbumin antisera showed slightly increased expression of ovalbumin from pKKova compared with that from pDEL2ova. The molecular mass of ovalbumin expressed by both Escherichia coli (data not shown) and S. typhimurium was approximately 3 kDa less than that of the native protein, which probably resulted from the inability of the bacteria to glycosylate ovalbumin. Expression of ovalbumin from either plasmid was constitutive in S. typhimunum, as LacI, the repressor of ptrc, is not produced by Salmonella spp., which lack the lactose operon. Stability studies of pKKova and pDEL2ova. The stability of S. typhimurium containing pKKova and pDEL2ova was determined by growth in vitro in the absence of antibiotics. Bacteria were subcultured for 140 generations, and the percentage of cells retaining the plasmids was determined by plating of cultures onto media with and without antibiotic selection (Fig. 3a). These studies showed that pKKova was more stable than pDEL2ova in S. typhimurium. Approximately 50% of bacteria harboring pKKova retained the plasmid after 100 generations, while only 1% still carried pDEL2ova. Groups of mice were infected orally with strain BRD509 containing either pKKova or pDEL2ova. On days 3, 5, 7,
S. TYPHIMURIUM MUTANTS INDUCE CLASS I-RESTRICTED CTL
VOL. 61, 1993
a
infected with S. typhimurium expressing ovalbumin from either pKKova or pDEL2ova was determined by release of 51Cr from labelled EL.4 cells transfected with a recombinant plasmid expressing ovalbumin (EG.7-Ova). Various mouse infection regimens were examined (Fig. 4). The CTL activity in spleen cell populations from infected mice was compared with that in spleen cell populations from animals injected with ovalbumin-loaded lymphocytes (positive control). The specificity of CTL-mediated lysis was determined by using untransfected EL.4 cells as targets. Only splenocytes from mice infected with S. typhimurium AaroA AaroD expressing ovalbumin from pKKova had measurable CTL activity. The amount of specific lysis increased with the number of inoculations the mice received, though animals infected once and left for 28 days produced detectable responses equivalent to 50% of the positive controls. In contrast, no priming of an ovalbumin-specific, class I-restricted CTL response occurred even after three inoculations with S. typhimurium AaroA AaroD expressing ovalbumin from pDEL2ova. Very low-level lytic responses were observed in splenocytes from animals infected once and left for 7 days. No lysis of untransfected control EL.4 cells was observed at any time point. Antibody levels in mice infected with recombinant salmonellae. Antibody levels were determined in Salmonellainfected animals by ELISA against both S. typhimurium LPS and ovalbumin. Control mice (injected with phosphatebuffered saline [PBS]) housed with infected animals did not produce antibodies against either antigen. All mice infected with S. typhimurium alone or recombinant salmonellae expressing ovalbumin from either pKKova or pDEL2ova produced antibodies reactive with S. typhimurium LPS (Fig. 5). The antibodies were first detected between 2 and 4 weeks postinfection. No antibodies against ovalbumin were detected in any animal infected with S. typhimunum expressing ovalbumin.
1001
4-0
0, In
.0 cn Cu
o
bn
Generations
0L)
n
4) c 0) a .) ON
0
2
4
6
5377
8
10
12
Day of infection
FIG. 3. Stability of pKKova and pDEL2ova in S. typhimunum BRD509. (a) In vitro plasmid stability was determined by plating bacteria for resistance carried on the plasmids pKK233.2, pKKova, and pDEL2ova after various numbers of generations in the absence of antibiotic selection. The number of resistance colonies is expressed as a percentage of the bacteria which grew on media without antibiotics. (b) Plasmid stability was examined in vivo in mice infected for various periods with recombinant salmonellae. Bacteria isolated from the spleens of mice infected with strain BRD509 carrying pKKova or pDEL2ova were plated on media with and without antibiotic resistance.
and 10, animals were killed, their spleens were removed, and the number of ampicillin-resistant (i.e., plasmid-containing) bacteria present in the spleen cell suspension was determined as a percentage of the total bacterial count (Fig. 3b). These studies confirmed the in vitro observation that pKKova was more stable in S. typhimurium than pDEL2ova. S. typhimurium isolated from mice infected with organisms containing pDEL2ova showed increasing losses of the plasmid over time. In contrast, pKKova was detected in 80 to 100% of S. typhimurium organisms isolated at each time point. CTh assays with spleens removed from mice infected with S. typhimurium. The ovalbumin-specific CITL response in mice
DISCUSSION Current models of antigen processing and presentation, and subsequent induction of T-lymphocyte responses, suggest that two mutually exclusive pathways are used (26, 34). Generally, antigen can be presented either by MHC class II molecules to CD4+ T lymphocytes or by MHC class I molecules to CD8+ T lymphocytes. An antigen which is exogenous, i.e., derived from outside the cell, is endocytosed and directed to the lysosomal compartment for processing. The resultant peptides are subsequently loaded onto MHC class II molecules and transported to the cell surface for presentation to class II-restricted CD4+ T lymphocytes (15). Interaction between peptides and MHC class I molecules occurs in the endoplasmic reticulum. Antigens which ultimately combine with MHC class I molecules are degraded in the cytoplasm prior to their transport into the endoplasmic reticulum, where they bind to the nascent MHC class I polypeptide (10, 27). Interaction between peptides and MHC class I molecules stabilizes the complex and allows transport through to the cell membrane (35). Salmonella spp. are thought to reside and replicate within the phagosomes of professional phagocytes and prevent phagolysosome formation (18). As a result, peptides from salmonellae should be presented by MHC class II molecules and should induce a predominantly CD4+ T-lymphocyte, i.e., T-helper, response. This study found evidence that S. typhimurium aro mutants induce an MHC class I-restricted CTL response. Studies by others of CTL induction in
5378
TURNER ET AL.
Positive Control
80'
60~
60'
40'
40
20'
20
100
Sn
0n I
I
Sl
S.S:1
16.7:1
loo- f
e
_
50:1
16.7:1
(pKKOva)
1oo
d
c
80
80
60'
60
40'
40
20,
20
Xa F 0-II _ 5.S:1 150:1
S0:1
l
o
16.7:1
S.S:1
100
80
80
60.
60
60
40
40
40
40
20
\20
20
20
16.7:1
-090 150.1 Sl S.S:1
80'
O* 40'
20
16.7:1 S.5S: 1
0 ~~~~~0
150:1
501
16.7:1
S.S:1
100
100'
80'
80'
60
60'
40
40'
20j
20'
7
14 21 28
L l l IJ 150 1
ioo1
60
S0:1
Day 0
I (0 tt _
80-
100' i U(
1 Sl
BRD509 (pDEL20va)
BRD509
80
150:1
a.
b
100'
80g
150.1
_~
BRD509
a
OI
INFEC-F. IMMUN.
s_ 501
16.7:1
1 S.S:1
h
Day 0
7
14 21 28
l II l l AA 150.1
501
16.7:1
4
5.5:1
k
Day 0
7
14 21 28
l II l l
AA
A4
lb
50:150l 1
16.7:1
1S0.1
5.5:1
S0.1
16.7:1
S.S:1
lS0:1
1001
100
80'
80g
80'
60'
60'
60'
40.
40
40a
20'
20'
20.
S0 1
16.7:1
S.S:1
100-
m
n
Day 0
7
14 21 28
l
l
l
A
*
l
0-
150:1
50:1
16.7:1
S.S:1
S0:1
S0:1
16.7:1
S.S:1
150 .1
501
16.7:1
S.S:1
effector:target ratio FIG. 4. CTL responses in mice infected with recombinant salmonellae. The spleens were removed from mice infected orally with recombinant salmonellae at indicated times ( t ). Mice were infected at indicated times (A), and the positive control animals were primed with ovalbumin-loaded lymphocytes at designated times (*). The responses of mice given ovalbumin-loaded lymphocytes (a, e, i, and 1), BRD509 (b, f, j, and m), BRD509 containing pKKova (c, g, and n), or BRD509 containing pDEL2ova (d, h, and k) are shown. Splenocytes from infected or primed animals were reacted with 5"Cr-labelled EL.4 cells transfected with ovalbumin (0) or EL.4 cells alone (-). EL.4 cells carry only MHC class I (H-2b). The amount of 51Cr released was determined by counting cell supernatants and comparing release with that obtained by using 1% Triton X-100. In all cases, spontaneous release was less than 20% of maximal release.
Salmonella-infected mice have yielded conflicting results. Some investigations found the induction of MHC class I-restricted (presumably CD8+) CTL (1, 12), while another indicated that the cytotoxicity was mediated by CD4+ MHC class II-restricted lymphocytes (32). Our results showed that class I-restricted CTL were induced when the carried antigen was expressed by S. typhimurium in a stable manner from pKK233.2, which was constructed by using the pBR322 variant of the ColEl replicon. S. typhimunium containing the less stable pDEL2ova construct, expressed from the pUC variant of ColEl, failed to induce MHC class I-restricted CTL. Cytotoxicity observed in our study must be MHC class I-restricted, as EL.4, the transfected target cell line used in these studies, expresses only MHC class I.
Studies of S. typhimurium as a carrier in mice have found that the level of heterologous expression affects the ability of immunized mice to mount a humoral immune response (11). Many carrier studies have used expression vectors based on pBR322, which our study has shown to be more stably maintained than pUC-based vectors. While both pUC and pBR322 are variants of the ColEl-type plasmid pMB1, the copy numbers of the two vectors vary considerably. Plasmid pBR322 usually replicates at 18 to 20 copies per cell; in contrast, pUC plasmids usually attain at least 100 copies per cell (9, 36). Total stability of expression may be obtained by integration of the foreign gene and appropriate transcription and translation signals into the chromosome. Integration vectors which specifically target sites which should not
VOL. 61, 1993
S. TYPHIMURIUM MUTANTS INDUCE CLASS I-RESTRICTED CTL 0
A
0.80
B
0.80 '
0
0 *
0.60
EC
0.
0
0.60
0
0
0)
0.40
0
S *
-
0
0 0.204
0.204
_ T 0.c
0.0
6
4
0.80
E
c 0 0)
q. 0
8
0
c
*41- 1
00.60
0
2
0
D
0.80
8
6
4
0.60 0
*
.|
-
0.400
0.40
0 S
0
,o 0.20.
0.20
-
I
2
4
.
6
-
I
8
0.0 0
2
4
6
8
Weeks after infection FIG. 5. Antibody responses in Salmonella-infected mice. Mice infected with S. typhimurium BRD509 (B), BRD509 carrying pKK ova (C), and BRD509 carrying pDEL2ova (D) and mice inoculated with PBS (A) were bled at regular intervals. The serum was assayed for antibodies against ovalbumin (Cl) or S. typhimurium LPS (0) (30) by ELISA.
further attenuate the carrier (i.e., other genes in the prechorismate pathway, e.g., aroC) have been used, though the level of expression from a single copy of the gene has usually been insufficient to induce antibody responses. Multicopy plasmid vectors such as pBR322 which are carried in a stable manner should facilitate increased foreign antigen expres-
5379
escape from the endosome, they do extend the in vitro experiments by Pfeifer et al. (28) by showing that ovalbuminexpressing bacteria can also access the MHC class I presentation pathway for effective in vivo CTL priming. None of the mice infected in this study produced antibody against ovalbumin expressed by salmonellae. The priming of T lymphocytes by S. typhimurium Aaro mutants expressing heterologous antigens in the absence of measurable antibody was inferred in a recent study in which partial protection against murine pertussis was achieved by using recombinant salmonellae expressing Bordetella pertussis antigen P.69 (30). The results reported here suggest that the primed T lymphocytes may have included CD8+ CTL and show that in both instances (this study and reference 30) the recombinant proteins probably entered the T-lymphocytes presentation pathway without being released into the general circulation, thus failing to trigger a detectable humoral response. The use of rationally attenuated pathogens as carriers of heterologous antigens requires that the carrier be capable of inducing appropriate effector responses. We have shown here that S. typhimunium AaroA AaroD mutants can induce specific CTL reactive with heterologous antigens, and this suggests that induction requires stable expression of the foreign antigen. The kinetics of the response indicate that CTL specific for carried antigens are present within 2 weeks of immunization and that reimmunization augments the CIL response. The role of CD8+ CTL in resistance to bacterial infections other than experimental listeriosis is unknown. The techniques reported here will be used to examine CTL induction by other auxotrophic (pur), regulatory gene (ompR), and protease (htrA) salmonella mutants to determine whether the capacity to induce CTL correlates with protection, or lack of protection, by these experimental vaccine strains. By comparison, the relevance of CTL to resistance against viral infections is well documented, and our results suggest that the S. typhi Aaro vaccines currently undergoing evaluation with humans may be excellent carriers of viral CTL determinants.
sion.
The finding that S. typhimunum, which is thought to remain within phagosomes, induces MHC class I-restricted CTL is in conflict with the current antigen presentation and processing models (4, 34). Listeia monocytogenes, which is known to escape from endosomes through the action of listeriolysin, is an efficient inducer of CD8+ CTL (19). Induction of CD8+ CTL by S. typhimunum suggests either that the organism escapes from the phagosome or that antigens present within the phagosome can interact with MHC class I molecules. Our in vivo data are compatible with a recent in vitro study of antigen presentation by cells infected with salmonellae and E. coli which suggested that bacteria may have access to an alternative pathway of antigen processing and presentation (28). Since this alternative presentation pathway was resistant to inhibition by cycloheximide and brefeldin A, known inhibitors of classical class I presentation, it appears likely that access to this mechanism relies solely on bacterial escape into the cell cytoplasm (28). Furthermore, there is increasing evidence which suggests that particulate forms of exogenous antigens such as ovalbumin and bacterial 3-galactosidase can efficiently access the class I presentation pathway (5, 20). Consequently, ovalbumin expressed by S. typhimurium may simply represent an alternative form of a particulate antigen. While our current data do not formally exclude the possibility that this phenomenon relies on some level of bacterial
ACKNOWLEDGMENTS This study was funded by the N.H. and M.R.C. of Australia, whose financial assistance is gratefully acknowledged. The generous gift of LB5010, BRD509, and P22HT/intC by Gordon Dougan, Imperial College London, is also acknowledged. REFERENCES 1. Aggarwal, A., S. Kumar, R. Jaffe, D. Hone, M. Gross, and J. Sadoff. 1990. Oral Salmonella: malaria circumsporozoite recombinants induce specific CD8+ cytotoxic T cells. J. Exp. Med. 172:1083-1090. 2. Binotto, J., P. R. MacLachan, and K. E. Sanderson. 1991. Electrotransformation in Salmonella typhimurium LT2 strains. Can. J. Microbiol. 37:474-477. 3. Brown, A., C. E. Hormaeche, R. D. de Hormaeche, M. Winther, G. Dougan, D. J. Maskell, and B. A. D. Stocker. 1987. An
attenuated aroA Salmonella typhimurium vaccine strain elicits humoral and cellular immunity to cloned 3-galactosidase in mice. J. Infect. Dis. 155:86-92. 4. Carbone, F. R., and M. J. Bevan. 1989. Major histocompatibility complex control of T cell recognition, p. 541-571. In W. E. Paul (ed.), Fundamental immunology. Raven Press, New York. 5. Carbone, F. R., and J. M. Bevan. 1990. Class I restricted processing and presentation of exogenous cell-associated antigen in vivo. J. Exp. Med. 171:377-387. 6. Cerundolo, V., J. Alexander, K. Anderson, C. Lamb, P. Creswell, A. McMichael, F. Gotch, and A. Townsend. 1990. Presentation of viral antigen controlled by a gene in the major
5380
TURNER ET AL.
histocompatibility complex. Nature (London) 345:449-452. 7. Chatfield, S. N., N. Fairweather, I. Charles, D. Pickard, M. Levine, D. Hone, M. Posada, R. A. Strugnell, and G. Dougan. 1992. Construction of a genetically defined Salmonella typhi Ty2 aroA, aroC mutant for the engineering of a candidate oral typhoid-tetanus vaccine. Vaccine 10:53-60. 8. Cooper, G. L., L. M. Venables, R. A. J. Nicholas, G. A. Cullen, and C. E. Hormaeche. 1992. Vaccination of chickens with chicken-derived Salmonella enteritidis phage type 4 aroA live oral salmonella vaccines. Vaccine 10:247-254. 9. Covarrubias, L., L. Cervantes, A. Covarrubias, X. Soberon, I. Vichio, A. Blanco, Y. M. Kupersztoch-Portnoy, and F. Bolivar. 1981. Construction and characterization of new cloning vehicles. V. Mobilization and coding properties of pBR322 and several deletion derivatives including pBR327 and pBR328. Gene 13:25-35. 10. Cox, J. H., J. W. Yewdell, L. C. Eisenlohr, P. R. Johnson, and J. R. Bennink. 1990. Antigen presentation requires transport of MHC class I from the endoplasmic reticulum. Science 247:715718. 11. Fairweather, N. F., S. N. Chatfield, A. J. Makoff, R. A. Strugnell, J. Bester, D. J. Maskell, and G. Dougan. 1990. Oral vaccination of mice against tetanus by use of a live attenuated Salmonella carrier. Infect. Immun. 58:1323-1326. 12. Flynn, J. L., et al. 1990. Generation of a cytotoxic T-lymphocyte response using a Salmonella antigen-delivery system. Mol. Microbiol. 4:2111-2118. 13. Germanier, R. 1972. Immunity in experimental salmonellosis. III. Comparative immunization with viable and heat-inactivated cells of Salmonella typhimurium. Infect. Immun. 5:792-797. 14. Griffin, H. G., and P. A. Barrow. 1993. Construction of an aroA mutant of Salmonella serotype Gallinarum: its effectiveness in immunization against experimental fowl cholera. Vaccine 11: 457-462. 15. Harding, C. V., D. S. Collins, J. W. Slot, H. J. Gueze, and E. R. Unanue. 1991. Liposome-encapsulated antigens are processed in lysosomes, recycled and presented to T cells. Cell 64:393-401. 16. Hoiseth, S. K., and B. A. D. Stocker. 1981. Aromatic-dependent Salmonella typhimurium are non-virulent and effective as live vaccines. Nature (London) 291:238-239. 17. Hsu, H. S. 1989. Pathogenesis and immunity in murine salmonellosis. Microbiol. Rev. 53:390-409. 18. Ishibashi, Y., and A. Toshihiko. 1990. Specific inhibition of phagosome-lysosome fusion in murine macrophages mediated by Salmonella typhimurium infection. FEMS Microbiol. Lett. 64:35-44. 19. Kaufmann, S. H. E., E. Hug, and G. DeLibero. 1986. Listeria monocytogenes-reactive T lymphocyte clones with cytolytic activity against infected target cells. J. Exp. Med. 164:363-368. 20. Kovacsovicsbankowski, M., K. Clark, B. Benacerref, and K. L. Rock. 1993. Efficient major histocompatibility complex class-I presentation of exogenous antigen upon phagocytosis by macrophages. Proc. Natl. Acad. Sci. USA 90:4942-4946. 21. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London)
INFECT. IMMUN.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35. 36.
227:680-685. 22. Mackaness, G. B., R. V. Blanden, and F. M. Collins. 1966. Host-parasite relations in mouse typhoid. J. Exp. Med. 124:573583. 23. MacReynolds, L., B. W. O'Malley, A. D. Nisbet, J. E. Fothergill,
37.
D. Givol, S. Fields, M. Robertson, and G. G. Brownlee. 1978. Sequence of chicken ovalbumin mRNA. Nature (London) 273: 723-728. Millar, I. A., S. Chatfield, G. Dougan, L. De Silva, H. S. Joysey, and C. Hormaeche. 1989. Bacteriophage P22 as a vehicle for transducing cosmid gene banks between smooth strains of Salmonella typhimurium: use in identifying a role for aroD in attenuating virulent Salmonella strains. Mol. Gen. Genet. 215: 312-316. Moore, M. W., F. R. Carbone, and M. J. Bevan. 1988. Introduction of soluble protein into the class I pathway of antigen processing and presentation. Cell 54:777-785. Morrison, L. A., A. E. Lukacher, V. L. Braciale, D. P. Fan, and T. J. Braciale. 1986. Differences in antigen presentation to MHC class I and class II restricted influenza virus specific cytotoxic T lymphocyte clones. J. Exp. Med. 63:903-921. Nuchtern, J. G., J. S. Bonifacino, W. E. Biddison, and R. D. Klausner. 1989. Brefeldin A implicates egress from endoplasmic reticulum in class I restricted antigen presentation. Nature (London) 339:223-226. Pfeifer, J. D., M. J. Wick, R. L. Roberts, K. Findlay, S. J. Normark, and C. V. Harding. 1993. Phagocytic processing of bacterial antigens for class I MHC presentation to T cells. Nature (London) 361:359-362. Stocker, B. A. D., S. K. Hoiseth, and B. P. Smith. 1983. Aromatic-dependent Salmonella spp. as live vaccines in mice and cattle. Dev. Biol. Stand. 53:47-54. Strugnell, R., G. Dougan, S. Chatfield, I. Charles, N. Fairweather, J. Tite, J. L. Li, J. Beesley, and M. Roberts. 1992. Characterization of a Salmonella typhimurium aro vaccine strain expressing the P.69 antigen of Bordetella pertussis. Infect. Immun. 60:3994-4002. Tacket, C. 0., D. M. Hone, R. Curtiss III, S. M. Kelly, G. Losonsky, L. Guers, A. M. Harris, R. Edelman, and M. M. Levine. 1992. Comparison of the safety and immunogenicity of AaroC AaroD and Acya Acrp Salmonella typhi strains in adult volunteers. Infect. Immun. 60:536-541. Tite, J. P., X. M. Gao, C. M. Hughes-Jenkins, M. Lipscombe, D. O'Callaghan, and G. Dougan. 1990. Anti-viral immunity induced by recombinant nucleoprotein to the immune system using attenuated Salmonella typhimurium as a live carrier. Immunology 70:540-546. Towbin, H., T. Staehelin, and J. Gordon. 1979. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. USA 76:4350-4354. Townsend, A. R. M., and H. Bodmer. 1989. Antigen recognition by class I restricted T lymphocytes. Annu. Rev. Immunol. 7:601-624. Townsend, A. R. M., T. Elliot, V. Cerundolo, L. Foster, B. Barber, and A. Tse. 1990. Assembly of MHC class I molecules analyzed in vitro. Cell 62:285-295. Weber, P. C., M. Levine, and J. C. Glorioso. 1989. A rapid and sensitive method for plasmid copy number comparisons. Nucleic Acids Res. 17:5866. Zinkernagel, R. M., and P. C. Doherty. 1974. Immunological surveillance against altered self components by sensitised T lymphocytes in lymphocytic choriomeningitis. Nature (London) 251:547.
