|Home | About | Journals | Submit | Contact Us | Français|
Tuberculosis remains a global health threat, and there is dire need to develop a vaccine that is safe and efficacious and confers long-lasting protection. In this study, we constructed recombinant attenuated Salmonella vaccine (RASV) strains with plasmids expressing fusion proteins consisting of the 80 amino-terminal amino acids of the type 3 secretion system effector SopE of Salmonella and the Mycobacterium tuberculosis antigens early secreted antigenic target 6-kDa (ESAT-6) protein and culture filtrate protein 10 (CFP-10). We demonstrated that the SopE-mycobacterial antigen fusion proteins were translocated into the cytoplasm of INT-407 cells in cell culture assays. Oral immunization of mice with RASV strains synthesizing SopE–ESAT-6–CFP-10 fusion proteins resulted in significant protection of the mice against aerosol challenge with M. tuberculosis H37Rv that was similar to the protection afforded by immunization with Mycobacterium bovis bacillus Calmette-Guérin (BCG) administered subcutaneously. In addition, oral immunization with the RASV strains specifying these mycobacterial antigens elicited production of significant antibody titers to ESAT-6 and production of ESAT-6- or CFP-10-specific gamma interferon (IFN-γ)-secreting and tumor necrosis factor alpha (TNF-α)-secreting splenocytes.
The World Health Organization reported that there were 9.4 million new cases of tuberculosis (TB) in 2009. This infectious disease causes more deaths worldwide than any other infection caused by a single bacterial pathogen, mostly in developing countries (80). Mycobacterium tuberculosis, the causative agent of TB, may cause acute infection or the bacteria may persist in infected individuals for years by switching to a nonreplicating (dormant) state. Reactivation of the dormant bacteria to active growth depends on epidemiological, host, and bacterial factors (11). Although there are effective antibiotics for treating TB, the lengthy treatment of the infection frequently results in compliance failures. Strains of M. tuberculosis that are resistant to multiple drugs have arisen and continue to increase in incidence due to insufficient control measures (1). The live attenuated M. bovis bacillus Calmette-Guérin (BCG) vaccine has been in use for over 80 years. BCG has displayed efficacy in protecting newborns and young children against serious complications of the disease, e.g., meningitis, but does not confer long-lasting protection against infection. However, the efficacy of BCG against pulmonary TB is variable in adults, ranging from 0 to 80% in different trials (2, 29, 76). Therefore, new approaches to controlling TB are essential and will result from understanding the biology of M. tuberculosis and its interactions with its host. Such understanding is required both for the development of new drugs to extend the range of TB treatments and for the development of a new generation of vaccines.
Attenuated Salmonella enterica has been used both as a homologous vaccine and as a delivery system for recombinant heterologous antigens to induce protective immunity against several infectious diseases and tumor sources in animal models (10, 19, 24, 27, 35, 48, 53, 65, 68, 70). Oral administration of Salmonella allows infection of Peyer's patches via M cells, as well as phagocytosis by dendritic cells sampling the gut mucosa and colonization of the mesenteric lymph nodes, liver, and spleen, generating mucosal, humoral, and cellular immune responses against Salmonella and its heterologous antigens (10, 19, 24, 49, 77, 81). We have reported the advantages of using new-generation recombinant attenuated Salmonella vaccine (RASV) strains that are phenotypically similar to the wild-type strain at the time of oral vaccination as an alternative for vaccination (23, 24, 52, 79). These RASV strains are able to colonize and persist in the lymphoid tissue without causing disease symptoms when carrying heterologous antigens, thereby inducing higher protective mucosal and systemic immune responses against a number of infectious diseases (27, 45, 47, 48, 68, 74, 83). Additionally, several approaches have been employed to improve the ability of Salmonella to survive in the gastrointestinal tract and to reach the lymphoid tissues. The deletion of Salmonella genes that encode enzymes involved in the biosynthesis of the peptidoglycan layer of the bacterial cell wall (e.g., aspartate β-semialdehyde dehydrogenase [Asd]) allows the use of plasmid systems harboring the gene encoding this enzyme to be maintained without the use of antibiotic resistance markers (60). Additionally, use of plasmids with different copy numbers is used to attain a better balance between plasmid replication and the synthesis of heterologous protective antigens (45, 60, 74). A series of expression vectors harboring chimeric fusions between the antigen to be analyzed and different types of secretion signal sequences (e.g., a β-lactamase signal sequence to allow protein secretion into the periplasm or extracellular compartment) was constructed to enhance the immune responses to the antigens (45, 81).
S. enterica employs different mechanisms to colonize, replicate, and survive within the eukaryotic host cells, such as the specialized type 3 secretion system (T3SS) encoded in Salmonella pathogenicity island 1 (SPI-1). The T3SS forms a multiprotein needle-like apparatus that injects proteins (effectors) into host cells to modulate a variety of cellular functions (34). One of these effector proteins is SopE, encoded within the genome of a cryptic bacteriophage located at centisome 61 of the S. enterica serovar Typhimurium chromosome (40). SopE is a Rho GTPase activator that interacts with Cdc42 and Rac1, resulting in membrane ruffling and actin cytoskeletal reorganization, thereby promoting the internalization of Salmonella into host cells (39, 46, 49, 50). Moreover, SopE has been shown to be rapidly ubiquitinated and processed by the eukaryotic proteasome degradation pathway (49). The signals for secretion and translocation of SopE by the T3SS are located within the amino-terminal region (between residues 1 and 78) of the protein (46, 50). The SopE secretion and translocation domain is a tool to explore the use of the Salmonella T3SS as a means for delivery of M. tuberculosis antigens into the eukaryotic cell cytosol. Antigens delivered into the cell cytosol become accessible to the major histocompatibility complex (MHC) class I-restricted antigen pathway, which is a prerequisite for efficient stimulation of CD8+ T-cell responses required to confer complete protection against intracellular pathogens, such as M. tuberculosis (13). Several groups of investigators have used the SPI-1 T3SS effector protein SopE or SptP fused to viral (simian immunodeficiency virus [SIV] and lymphocytic choriomeningitis virus [LCMV]) or protozoan (Eimeria acervulina and E. tenella) antigens to deliver these antigens to the host cell cytosol (27, 48, 68). These investigators demonstrated that delivery of antigens by the T3SS stimulated antigen-specific cytotoxic T-cell responses, antigen-specific CD8+ memory T cells, and protection against challenge with viral or Eimeria pathogens (27, 48, 68).
In the present study, we designed a TB vaccine based on an RASV harboring an Asd-positive (Asd+) vector that contains the gene sequence encoding SopE amino-terminal region residues 1 to 80 (SopENt80) to stimulate a T-cell immune response by employing the SPI-1 T3SS as a delivery system to secrete and translocate M. tuberculosis major T-cell antigens into the eukaryotic cytosol. The antigens included are the early secreted antigenic target 6-kDa (ESAT-6) protein and the culture filtrate protein 10 (CFP-10) (5, 72), which are encoded by genes in region of difference 1 (RD-1) of the M. tuberculosis chromosome (3, 54) and have been shown to be useful vaccine candidates against TB (9, 69).
The bacterial strains and plasmids used in this study are listed in Table 1. Lennox broth (51) supplemented with 0.3 M NaCl was used to stimulate the expression of the components associated with T3SS (32). Luria-Bertani (LB) broth (4) or LB broth supplemented with 0.05% arabinose was used to grow the Salmonella strains for the immunizations. LB broth and LB agar (1.5% agar) or MacConkey agar (Difco) were used for propagation and plating of Salmonella. For the growth of noncomplemented ΔasdA strains and plasmid stability tests, 50 μg/ml diaminopimelic acid (DAP) was added to the growth medium (60). Middlebrook 7H9 broth and Middlebrook 7H11 agar (Difco), each supplemented with 10% oleic acid-albumin-dextrose-catalase enrichment (Difco), were used to grow M. tuberculosis and M. bovis BCG.
DNA manipulations were carried out using standard procedures (66). Plasmid DNA was isolated using a QIAprep Spin miniprep kit (Qiagen, Valencia, CA). Restriction enzymes were used as recommended by the manufacturer (New England BioLabs, Inc., Ipswich, MA). Plasmid constructs were verified by DNA sequencing (Arizona State University facilities).
The Asd+ vectors pYA3869 and pYA3870, which contained the pSC101 (14) and p15A (17) replication origins, respectively, were constructed for delivery of heterologous antigens by the Salmonella T3SS. Both plasmids pYA3869 and pYA3870 were briefly described earlier (43) and have been evaluated by other members of our group for delivering Eimeria antigens (48). Plasmid pYA3869 was constructed from pYA3337 (20) by excising a 1-kb HpaI-NcoI fragment and replacing it with a 600-bp HpaI-NcoI fragment containing the carboxy-terminal region of the asd gene. The S. Typhimurium sopE promoter (PsopE), the Shine-Dalgarno sequence, and the nucleotides encoding the first 80 amino acids of SopE were then cloned into intermediate plasmids to finally generate pYA3869 (Table 1; see Fig. S1 in the supplemental material). The 80 N-terminal amino acids of SopE are essential for secretion and translocation of SopE by the T3SS (46, 50) and are designated SopENt80 in the plasmids used in this study.
The pYA3870 Asd+ vector was generated from pYA3332 (22) by excising a 1,004-bp HpaI-PstI fragment containing the carboxy-terminal region of the asd gene and the Ptrc promoter and replacing this sequence with a 1,096-bp HpaI-PstI fragment encoding the carboxy-terminal region of the asd gene, the S. Typhimurium sopE promoter (PsopE), and the SopENt80 secretion and translocation signal, which were excised from pYA3869, to generate pYA3870 (Table 1; see Fig. S1 in the supplemental material).
The DNA fragment containing the esxA and esxB genes, which encode the ESAT-6 and CFP-10 proteins, respectively, was PCR amplified from the M. tuberculosis H37Rv chromosome using the primer set CFP10-F1 (GGTAAAGAGAGAAGGTACCCCAGCATGGCAGAG) and ESAT6-R1 (GCTATTCTACGCGAACTAAGCTTTGCCCTATGCG).
The resulting 530-bp PCR product was digested with KpnI-HindIII and cloned into the pBK-CMV (Stratagene) plasmid digested with the same enzymes to obtain pYA3933. Plasmid pYA3933 was used for codon substitution of the M. tuberculosis esxA and esxB genes with the most frequently found codons in Salmonella. The esxB codons 20 (AGG to CGT) and 85 (CGG to CGT) and the esxA codons 20 (GGA to GGT) and 74 (CGG to CGT) were substituted by using a QuikChange site-directed mutagenesis kit (Stratagene). The resulting recombinant plasmid containing all of the optimized sequences from esxA and esxB was named pYA3934 (Table 1).
Two plasmids, pYA4221 and pYA4222, harboring two copies of esxA or with three copies of esxA fused in tandem, respectively, were constructed. The DNA fragment encoding three copies of esxA fused in tandem (E3) was excised from pYA4222 by digestion with EcoRI-HindIII and subcloned into pYA3869 or pYA3870 digested with the same enzymes to generate pYA4248 and pYA4251 (Table 1; see Fig. S1 in the supplemental material), respectively. Plasmids pYA4248 and pYA4251, containing sopENt80-esxA-esxA-esxA, were used to express the chimeric protein referred to as SopENt80-E3 (with E3 designating three copies of ESAT-6) in this study.
The 350-bp fragment containing M. tuberculosis esxB was PCR amplified from pYA3934 and cloned into XhoI-HindIII-digested pYA4221 to obtain pYA4224. The 891-bp fragment containing esxA-esxA-esxB was excised from pYA4224, digested with EcoRI and HindIII, and cloned into pYA3869 and pYA3870 digested with the same enzymes to generate pYA4254 and pYA4257, respectively (Table 1; see Fig. S1 in the supplemental material). Both the pYA4254 and pYA4257 plasmids contained sopE80Nt80-esxA-esxA-esxB and were used to express the chimeric protein referred to as SopENt80-E2C (with E2C designating two copies of ESAT-6 and one copy of CFP-10) in this study.
To detect chimeric protein SopENt80-E3 or SopENt80-E2C, both proteins were tagged independently with a sequence encoding the AU1 epitope tag (42). Additionally, to allow the specific identification of the chimeric proteins that were translocated into the eukaryotic cytoplasm, these chimeric proteins were tagged with an Elk tag that consists of the simian virus 40 large tumor antigen nuclear localization signal (NLS) (64) fused to amino acid residues 375 to 392 of the eukaryotic transcription factor Elk-1 (25), which are recognized and phosphorylated at serine 383 by eukaryotic protein kinases. The AU1 epitope fused to the Elk tag is referred to as the AU1E tag in this study. The DNA fragment containing the nucleotide sequence esxA-esxA-esxA-AU1E was subcloned into pYA3869 and pYA3870 to generate pYA4249 and pYA4252, respectively (see Fig. S1 in the supplemental material). The DNA fragment containing esxA-esxA-esxB-AU1E was subcloned into pYA3869 and pYA3870 to generate pYA4255 and pYA4258, respectively.
The plasmids pYA4249, pYA4252, pYA4255, and pYA4258 were digested with PvuII to remove the DNA fragment encoding the Elk tag. Each plasmid was self-ligated independently to obtain pYA4250, pYA4253, pYA4256, and pYA4259, respectively (Table).
The DNA fragment containing the promoter region and the nucleotide sequence encoding the entire S. Typhimurium sopE gene was PCR amplified from S. Typhimurium SL1344 and cloned into pYA3869 and pYA3870 to generate pYA4260 and pYA4263, respectively, to express SopE. A DNA fragment containing the AU1E nucleotide sequences was PCR amplified from pYA4226 and subcloned into pYA4260 and pYA4263 to obtain pYA4261 and pYA4264, respectively, encoding SopE-AU1E. pYA4261 and pYA4264 were digested with PvuII to remove the DNA fragment encoding the Elk tag. Each plasmid was self-ligated independently to obtain pYA4262 and pYA4265, respectively, to express SopE-AU1.
The esxB gene was PCR amplified from M. tuberculosis H37Rv, digested with EcoRI and BamHI, and cloned into the pYA3870 vector digested with the same enzymes. The resulting plasmid was digested with BamHI-HindIII, and a 294-bp BamHI-HindIII fragment containing the esxA gene, which was PCR amplified from M. tuberculosis H37Rv, was cloned into the plasmid to generate pYA3950, expressing SopENt80–CFP-10–ESAT-6, referred to in this work as SopENt80-C-E.
The plasmid encoding recombinant 6×His-tagged CFP-10 was constructed as follows: the nucleotide sequence encoding CFP-10 was PCR amplified from M. tuberculosis H37Rv chromosomal DNA, digested with KpnI and HindIII, and cloned into the pBAD-HA vector (Invitrogen). The resulting plasmid was digested with NcoI and HindIII to release a 300-bp DNA fragment, which was subcloned into the pET28a+ vector (Novagen, EMD4 Biosciences, San Diego, CA) digested with the same enzymes to generate pYA3815.
The χ8916 strain harboring an invA deletion mutation was generated by using the suicide vector constructed as follows: a nucleotide sequence encoding 250 amino acid (aa) residues of the invE carboxy-terminal region and 15 aa of the invA amino terminus was PCR amplified from S. Typhimurium chromosomal DNA, digested with KpnI-BamHI, and cloned into the suicide vector pRE112 (26). The resulting plasmid was digested with BamHI-SacI and was cloned with a 700-bp DNA fragment, PCR amplified, and digested with the same enzymes encoding the last 22 aa of invA and the first 79 aa of invC to obtain pYA4141.
The deletion-insertion mutations ΔasdA33, ΔPphoPQ176::TT araC PBAD phoPQ, ΔPcrp527::TT araC PBAD crp, and ΔaraBAD23 (where TT is transcription terminator, P stands for promoter, and the subscripted number refers to a composite deletion and insertion of the indicated gene) (23, 52) were introduced into the S. Typhimurium χ3761 wild-type strain by allelic exchange using suicide vectors and/or by transduction using bacteriophage P22HTint to yield χ9879 using standard protocols (44, 45). The artB13::MudJ allele (31), which directs constitutive expression of β-galactosidase, was introduced into S. enterica serovar Typhimurium χ8916 by transduction using a bacteriophage P22HTint lysate from Salmonella χ4574 harboring this allele, resulting in strain χ9930. Transductants with kanamycin-resistant and β-galactosidase phenotypes were selected and grown on Evans blue uridine agar plates to confirm that the transductants were phage free and not P22 lysogens (6). The invAC genes were deleted in S. Typhimurium χ8916 by allelic exchange using the suicide vector pYA4141 to generate χ11406. The ΔinvAC deletion mutation impairs the ability of Salmonella to invade cells of the intestinal epithelium (33).
Stability of the recombinant plasmids expressing the SopENt80 fusion proteins was determined for approximately 50 generations of growth under selective and nonselective conditions (presence of DAP), as described previously (74).
S. Typhimurium strains χ8916, χ9930, and χ11406 were transformed independently by electroporation with each Asd+ SopENt80 plasmid derivative and were grown under conditions to stimulate the expression of the T3SS and chimeric proteins (32). Briefly, a single colony of each transformed Salmonella strain was inoculated in Lennox broth (3 ml in a 13- by 100-mm tube) and grown at 37°C on an Orbit 1000 shaker (Labnet International, Edison, NJ) at 30 rpm overnight. On the next day, aliquots were taken from each overnight culture to inoculate fresh Lennox broth (5 ml in a 16- by 150-mm tube) at an optical density at 600 nm (OD600) of 0.17. The cultures were grown with shaking at 100 rpm (on the same shaker described above) for 3 h at 37°C. The cultures were centrifuged at 12,500 × g for 2 min. The bacterial pellets were washed with phosphate-buffered saline (PBS), resuspended and lysed with 150 μl of lithium dodecyl sulfate (LDS) sample buffer (Invitrogen), and then stored at −70°C. The supernatants were filtered using a 0.22-μm-pore-size filter (Corning Gilbert, Inc., Glendale, AZ) and precipitated with 10% trichloroacetic acid (TCA), and the pellet was resuspended in 100 μl of LDS sample buffer. Then, 37 μl of the pellet sample and 25 μl of the supernatant sample were boiled for 5 to 10 min and analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on 4 to 20% gels (Bio-Rad) and immunoblotted.
Secreted chimeric proteins were identified by immunoblotting using rabbit anti-ESAT-6 serum or anti-AU1 epitope tag serum (Bethyl Laboratories, Montgomery, TX), followed by alkaline phosphatase-conjugated goat anti-rabbit IgG (Sigma-Aldrich, St. Louis, MO). All of these antibodies were used at a 1:5,000 dilution. Mouse anti-β-galactosidase monoclonal antibody (1:1,000; clone Gal-40; Sigma-Aldrich) was used for detection of β-galactosidase, followed by alkaline phosphatase-conjugated goat anti-mouse IgM (Sigma-Aldrich). All experiments were performed three times.
The translocation assays were conducted according to the procedures described previously (15). Briefly, INT-407 cells were seeded into 100-mm tissue culture plates containing 12 ml of Dulbecco's modified Eagle medium (DMEM) with 10% fetal bovine serum (FBS) without antibiotics at a density of 2 × 107 cells per plate and allowed to adhere for 24 h. Before infection, cell monolayers were washed twice with 5 ml of Hanks' balanced salt solution (HBSS) at 37°C, and 2.5 ml of DMEM without FBS was added. The proteosome inhibitors MG132 (final concentration, 10 μM) and lactacystin (final concentration, 1 μM) (Calbiochem, EMD4 Biosciences, San Diego, CA) were added to the cells 15 min prior to infection. For translocation assays in the presence of cytochalasin D (Sigma-Aldrich), cytochalasin D was added to the cells 30 min before infection at a final concentration of 5 μg/ml. The cells were infected independently with each RASV strain harboring an Asd+ SopENt80 plasmid derivative expressing a chimeric protein (grown under conditions to stimulate the expression of the SPI-1 T3SS, centrifuged, and resuspended in HBSS) at a multiplicity of infection (MOI) of 50 CFU/eukaryotic cell for 1 h, and an MOI of 100 CFU/eukaryotic cell was used when cytochalasin D was previously added to the cells. Subsequently, the culture medium was removed, and the cells were washed three times with 10 ml Dulbecco's phosphate-buffered saline (DPBS) and treated with 10 μg/ml proteinase K for 15 min at 37°C. Afterwards, 2 mM phenylmethylsulfonyl fluoride (PMSF; Sigma-Aldrich) was added, and the cells were centrifuged at 600 × g for 5 min. The cellular pellet was lysed in 1 ml 10 mM Na2HPO4 (pH 7.2) containing 0.1% Triton X-100, 10 μg/ml DNase, 10 μg/ml RNase, 1 mM PMSF, 0.1% (vol/vol) protease inhibitor (P-8340), and 0.01% (vol/vol) phosphatase inhibitor (P-2850) cocktails (Sigma-Aldrich) and then incubated for 15 min at 4°C. Finally, the lysates were centrifuged for 30 min at 12,500 × g and 4°C, and the pellet obtained (P), containing the unbroken cells, membranes, and bacteria that had adhered and had been internalized, was resuspended in 200 μl of LDS sample buffer. The supernatant of the cytoplasmic fraction (C) containing the eukaryotic cytoplasm and the translocated recombinant proteins was filtered as indicated above and then precipitated with 10% TCA and resuspended in 210 μl of LDS sample buffer. The proteins of each sample were analyzed by 10% SDS-PAGE and immunoblotted. The chimeric proteins were identified using rabbit anti-ESAT-6 or anti-AU1 epitope tag sera, followed by peroxidase-conjugated goat anti-rabbit antibody (Sigma-Aldrich), using a chemiluminescent detection system (ECL; Pierce, Rockford, IL).
The Escherichia coli BL21(DE3) strain (Novagen) was transformed with the expression vector pYA3815 (His-tagged CFP-10) or pMRLB7 (His-tagged ESAT-6), and cells were grown at 37°C with aeration in LB broth containing 30 μg/ml kanamycin or 100 μg/ml ampicillin, respectively, to 0.5 OD600 unit. Production of the recombinant proteins was induced with 0.5 mM isopropyl-β-d-thiogalactopyranoside (IPTG) for 3 h at 37°C. Each protein was purified by nickel-nitrilotriacetic acid agarose chromatography. Eluted fractions containing purified 6×His–CFP-10 or 6×His–ESAT-6 were selected after SDS-PAGE, pooled, dialyzed against PBS, and then concentrated. Protein concentration was determined by the Bradford assay (8) using bovine serum albumin as a standard. Endotoxins were removed using Detoxi-Gel endotoxin-removing gel (Pierce, Rockford, IL). The amount of endotoxin contamination in the recombinant proteins was measured quantitatively with the Limulus amebocyte lysate assay (Cambrex Bio Science Walkersville, Inc., Walkersville, MD), which was used according to the manufacturer's instructions. The amount of endotoxin found was <0.01 endotoxin unit (EU) per μg of recombinant protein. The purified recombinant proteins were used for production of the antisera in rabbits for enzyme-linked immunosorbent assay (ELISA) and enzyme-linked immunospot (ELISPOT) assay. Salmonella outer membrane proteins (SOMPs) were obtained using the sonication and Triton X-100 extraction procedure described previously (56).
Female C57BL/6 mice that were 6 to 7 weeks old were purchased from Charles River Laboratories (Wilmington, MA). The Arizona State University Animal Care and Use Committee approved all of the animal procedures. Mice were acclimated for 7 days before starting the experiments. Groups of 5 to 6 mice were vaccinated subcutaneously with a single dose of 5 × 104 CFU of M. bovis BCG at day 0. For immunization with RASV strains, mice were deprived of food and water for 6 h before oral immunization. The RASV χ8916 strains independently harboring each of the Asd+ SopENt80 plasmid derivatives pYA3870, pYA4252, and pYA4257 were grown statically for 18 h in 5 ml LB broth at 37°C. These cultures were used to inoculate 100 ml of LB broth and then grown at 37°C with aeration to an OD600 of 0.8. The χ9879 strains independently harboring each of the Asd+ SopENt80 plasmid derivatives pYA3870, pYA4251, pYA4254, and pYA4257 were grown in the same way as described above, but in LB broth containing 0.05% arabinose. Cells from each culture were pelleted by centrifugation at room temperature (4,000 × g for 15 min), and each pellet was resuspended in 1 ml of buffered saline containing 0.01% gelatin (BSG) (18). Dilutions of the vaccine strains were plated onto LB agar plates or LB agar plates supplemented with 0.2% arabinose to determine bacterial titers. Four groups of 16 mice each (for vaccination with χ8916 strains independently harboring pYA3870, pYA4252, or pYA4257 or with BSG alone) were orally inoculated with 20 μl of the respective RASV strains resuspended in BSG containing 1 × 109 CFU or with BSG alone on days 0, 21, and 49. Another five groups of 16 mice each were orally inoculated as described above with χ9879 strains harboring the respective Asd+ SopENt80 plasmid derivatives or with BSG alone on days 0, 7, and 49. Water and food were returned to the mice 30 min after immunization. Blood samples were obtained by submandibular vein puncture 2 days before vaccination for all of the groups of mice. Blood samples were also collected at days 20 and 48, for the mice immunized with χ8916 harboring the control vector pYA3870 or its derivative pYA4252 or pYA4257 or the BSG-dosed mice, and at days 21 and 65, for the mice immunized with χ9879 harboring the control vector pYA3870 or its derivative pYA4254, pYA4257, or pYA4251 or the BSG-dosed mice. The blood was incubated at 37°C for 60 min. Afterwards, the blood was centrifuged at 4,000 × g for 5 min, and the serum was removed. Sera obtained from mice in the same experimental group were pooled and stored at −70°C.
To assess the protective effects of the RASV strains against M. tuberculosis infection in immunized mice, groups of 5 to 6 immunized mice or those from the M. bovis BCG and buffered saline (BSG) control groups were infected 4 weeks after the last immunization with M. tuberculosis H37Rv delivered as an aerosol by a Glas-Col aerosolization chamber (Glas-Col LLC, Terre Haute, IN), programmed to deliver approximately 100 bacilli per lung. The mice were euthanized 6 weeks after the challenge, and immediately the lungs and the spleen were aseptically collected. The numbers of bacteria in the lungs and spleens were determined by serial dilution of individual whole-organ homogenates in sterile PBS. Serial dilutions of the samples were plated on Middlebrook 7H11 agar supplemented with 10% oleic acid-albumin-dextrose-catalase enrichment in duplicate. The colonies were counted after 4 weeks of incubation at 37°C. All experiments involving live M. tuberculosis H37Rv were conducted under biosafety level 3 laboratory conditions.
Total IgG, IgG2b, and IgG1 antibody titers against ESAT-6 and SOMPs and total IgG antibody titers against CFP-10 from vaccinated mice and controls were determined by ELISA, using standard protocols (75). Briefly, Nunc Immunoplate Maxisorb F96 plates (Nalge Nunc, Rochester, NY) were coated with purified ESAT-6 at 1 μg/well or SOMPs at 0.5 μg/well suspended in 0.05 M carbonate-bicarbonate buffer, pH 9.6. Sera from mice orally immunized with RASV χ8916 harboring the Asd+ SopENt80 plasmid pYA3870 or its derivative pYA4252 or pYA4257 were pooled and serially diluted by 2-fold dilutions from an initial dilution of 1:100 in PBS. Aliquots of 100 μl were added to duplicate wells and incubated overnight at 4°C. Plates were washed as indicated before and treated with horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG, IgG1, or IgG2b (1:4,000 dilution; Southern Biotechnology Inc., Birmingham, AL). Wells were washed and developed with o-phenylenediamine dihydrochloride (OPD) at 0.4 mg/ml in phosphate-citrate buffer with H2O2 (Sigma) at 200 μl/well. Color development was stopped by the addition of 50 μl of 3 M H2SO4 per 200 μl of reaction solution. Absorbance was recorded at 492 nm using an automated ELISA plate reader (Labsystems Multiskan MCC/340). Endpoint titers were expressed as the last sample dilution with an absorbance of 0.1 OD unit above that for the negative controls after 1 h of incubation.
The ELISPOT assay was performed to enumerate the gamma interferon (IFN-γ), tumor necrosis factor alpha (TNF-α), interleukin-4 (IL-4), and IL-10 cytokine-secreting cells (CSCs) in the spleens of immunized and naïve mice to determine the potential cellular immune response to immunization. This was performed using the ELISPOT assay kits (mouse IFN-γ, TNF-α, IL-4, and IL-10 ELISPOT sets; eBioscience) according to the manufacturer's instructions. Briefly, ELISPOT assays for χ8916 harboring the Asd+ SopENt80 plasmid derivatives were conducted at 1 week after the last immunization with spleens from three mice per group, and the analysis was conducted on the splenocytes from individual mice in triplicate. The spleen cells were incubated with the recombinant antigen at 1 μg/well for IFN-γ-secreting cells for 48 h at 37°C in a humidified (5% CO2-in-air) incubator. ELISPOT assays for χ9879 harboring the Asd+ SopE80 plasmid derivatives were conducted at 3 weeks after the last immunization with the pool of spleens from three mice of the same group. The spleen cells from all groups of mice were incubated with the recombinant antigen at 1 μg/well for 24 h (for IFN-γ- and TNF-α-secreting cells) or 48 h (for IL-4- and IL-10-secreting cells) as described above. The spots were counted using an automated ELISPOT assay plate reader (CTL analyzers; Cellular Technology Ltd., Cleveland, OH).
Differences in antibody responses, cytokine secretion levels measured by ELISPOT assay, and bacterial loads in the lungs and spleens between groups were determined by one-way analysis of variance (ANOVA), followed by Tukey's multiple-comparison test. Differences with P values of <0.05 were considered significant. Statistical analysis was performed using GraphPad Prism software (GraphPad Software, San Diego, CA).
We constructed and evaluated two plasmid vectors to analyze the effect of the antigen amount synthesized and delivered by the Salmonella vaccine strain on the immune response. The first plasmid, pYA3869, harbors the very-low-copy-number pSC101 replication origin, and the second plasmid, pYA3870, harbors the low-copy-number p15A replication origin. Both pYA3869 and pYA3870 plasmids contain the asdA gene encoding an enzyme involved in the biosynthesis of DAP, which is an integral component of the peptidoglycan layer of the bacterial cell wall. The asdA gene was used as a selectable marker to complement the chromosomal ΔasdA mutation in the RASV strains (60). To stimulate protective immunity against intracellular pathogens such as M. tuberculosis, it is also necessary that the mycobacterial antigens delivered by Salmonella vaccines become accessible to the MHC class I-restricted antigen-processing pathways. Thus, both plasmid vectors contain the secretion and translocation signals of the SPI-1 T3SS effector protein SopE, which are specified by the first 80 amino acids in the amino-terminal region of SopE (designated SopENt80), and allow SopE to be specifically transported via the SPI-1 T3SS. The nucleotide sequences encoding the potent immunogen ESAT-6 fused in triplicate were cloned downstream and in frame with sopENt80 into pYA3869 (to generate pYA4248) or pYA3870 (to generate pYA4251) (Table 1; see Fig. S1 in the supplemental material). The nucleotide sequences encoding ESAT-6 and CFP-10, fused in tandem, were also cloned downstream and in frame with sopENt80 into pYA3869 (to generate pYA4254) and pYA3870 (to generate pYA4257) (Table 1; see Fig. S1 in the supplemental material).
Initially, we observed that S. Typhimurium strain χ8916 harboring pYA3950, which contained a single copy of the DNA sequences encoding CFP-10 and ESAT-6 fused to sopENt80 (sopENt80-esxB-esxA), expressed the chimeric protein SopENt80–CFP-10–ESAT-6 (SopENt80-C-E) at very low levels. This fusion protein was barely detectable as a 30.1-kDa product in whole-cell lysates and supernatants of the cultured cells by immunoblotting using anti-ESAT-6 serum (data not shown). However, when several copies of the gene encoding ESAT-6 were cloned in tandem three times (sopENt80-esxA-esxA-esxA) or were cloned twice along with the DNA fragment encoding CFP-10 (sopENt80-esxA-esxA-esxB), the expression and secretion of the chimeric proteins (SopENt80-E3 and SopENt80-E2C, respectively) by the RASV strains were improved substantially (see expression details below). The AU1 epitope tag was added to the chimeric recombinant proteins at their carboxy termini, generating pYA4250 and pYA4253 (SopENt80-E3-AU1) or pYA4256 and pYA4259 (SopENt80-E2C-AU1), or the proteins were fused with the AU1E tag (AU1 epitope tag-Elk tag) to generate pYA4249 and pYA4252 (SopENt80-E3-AU1E) or pYA4255 and pYA4258 (SopENt80-E2C-AU1E). Thus, the mycobacterial protective antigens ESAT-6 and CFP-10 expressed by the Salmonella vaccine strains from the Asd+ SopENt80 plasmid derivatives are tagged to detect their secretion and translocation to the host cell cytosol.
The data included in the figures are results of experiments performed using some of the plasmids described above to illustrate the points that we will describe throughout the Results section. Results obtained with other plasmids described in Table 1 were not included in order to reduce the complexity and amount of data presented.
To analyze and verify the secretion and translocation of the chimeric proteins to the cytoplasm of eukaryotic cells, the isogenic S. Typhimurium strains χ9930 and χ11406, derived from strain χ8916, were constructed as described in Materials and Methods. To evaluate the immunogenicity of the RASV strains synthesizing and delivering the heterologous antigens ESAT-6 and CFP-10, we employed S. Typhimurium χ8916 and χ9879. Strain χ8916, derived from χ8768 (7), harbors the deletion mutations ΔasdA16 and ΔphoP233. The phoP and phoQ genes form a two-component regulatory system that regulates the transcription of several operons and genes necessary for virulence (63). Salmonella ΔphoP mutants have been described to be safe and immunogenic, and they elicit a predominantly cellular immune response (13). Strain χ9879, which exhibits regulated delayed attenuation in vivo, was also used in this study. In this strain, the promoter sequence of the phoPQ virulence genes was replaced with the improved tightly arabinose-regulated araC PBAD activator promoter cassette (37, 47). Strain χ9879 harbors the deletion-insertion mutations ΔPphoPQ176::TT araC PBAD phoP and ΔPcrp527::TT araC PBAD crp (where TT is transcription terminator, P stands for promoter, and the subscripted number refers to a composite deletion and insertion of the indicated gene). The phoP gene is transcribed from the PBAD promoter, which is activated by the AraC protein in the presence of arabinose. Thus, this strain is phenotypically wild type and able to colonize the host tissues when grown in the presence of arabinose. However, since arabinose is not available in the host tissues, this strain becomes attenuated as a result of the dilution of the gene products during cell division (23, 24). To downregulate translational efficiency, the Shine-Dalgarno (SD) and phoP start codons were modified to prevent overexpression, which could result in a hyperattenuation with a decrease in immunogenicity. The crp gene encodes the cyclic AMP (cAMP) receptor protein necessary for virulence (21), and like phoP, its transcription is arabinose regulated from the PBAD promoter (23, 24). This strain also harbors the asdA33 deletion mutation. Finally, to delay the onset of attenuation, the ΔaraBAD23 deletion mutation was included in χ9879 to prevent consumption of the arabinose available in the bacterial cytoplasm during oral immunization and to preclude acid formation during growth of the culture for oral inoculation (23, 24).
All plasmids were 100% stable in χ8916 and χ9879 throughout 50 generations of growth under both selective and nonselective conditions (data not shown).
To analyze the secretion of chimeric ESAT-6 proteins by the SPI-1 T3SS, single colonies of χ8916 harboring different ESAT-6 expression plasmids, pYA4248 or pYA4251 (SopENt80-E3), pYA4250 or pYA4253 (SopENt80-E3-AU1), pYA4249 or pYA4252 (SopENt80-E3-AU1E), pYA4254 or pYA4257 (SopENt80-E2C), pYA4256 or pYA4259 (SopENt80-E2C-AU1), and pYA4255 or pYA4258 (SopENt80-E2C-AU1E), or the SopE expression plasmids pYA4262 or pYA4265 (SopE-AU1) and pYA4261 or pYA4264 (SopE-AU1E) were grown under conditions to induce the expression of the T3SS and chimeric proteins, as described in Materials and Methods. The chimeric proteins SopENt80-E3 (with a molecular mass of 40.3 kDa) and SopE-E2C (40.4 kDa), either of which was fused with an AU1 epitope tag, SopENt80-E3-AU1 (41.1 kDa) and SopE-E2C-AU1 (41.4 kDa), or with an AU1E tag, SopENt80-E3-AU1E (44.9 kDa) and SopENt80-E2C-AU1E (45.4 kDa), as well as SopE-AU1 (27.1 kDa) and sopE-AU1E (30.9kDa) were detected in whole-cell lysates (pellet) and supernatants by immunoblotting using rabbit anti-ESAT-6 serum or anti-AU1 epitope tag serum (Fig. 1A and B). The data in Fig. 1 (and unpublished results) demonstrate that the p15A ori plasmids and the pSC101 ori plasmids produce similar amounts of the recombinant proteins. Moreover, the presence of the AU1 or AU1E epitope tags on the recombinant proteins does not impair secretion. To determine if the recombinant fused proteins were effectively secreted via the T3SS rather than being released in the culture supernatant due to cell lysis, single colonies of the RASV χ9930 strain (with the artB13a::MudJ mutation-insertion, which results in constitutive synthesis of β-galactosidase ) harboring p15A ori plasmids pYA4251 (SopENt80-E3), pYA4252 (SopENt80-E3-AU1E), pYA4257 (SopENt80-E2C), pYA4258 (SopENt80-E2C-AU1E), pYA3950 (SopENt80-C-E), pYA4264 (SopE-AU1E), and pYA4265 (SopE-AU1) or the control vector (pYA3870) were grown under conditions to stimulate expression and secretion by the T3SS. β-Galactosidase was detected in the pellets (as a protein of approximately 116 kDa) and was undetectable in the supernatants (Fig. 2), indicating that the recombinant chimeric proteins in the supernatants were secreted by the T3SS. To confirm this result, single colonies of χ11406 (defective in the expression of the T3SS due to deletion of invAC) harboring the same plasmids were grown in broth culture under conditions to stimulate expression of the T3SS. The recombinant antigens were detected only in the pellets and were undetectable in the supernatants of the cultures (Fig. 3). These results indicated that the Salmonella ΔinvAC mutant χ11406 was unable to secrete the recombinant proteins and confirmed that the chimeric antigenic proteins expressed from the Asd+ SopENt80 plasmid derivatives are secreted by the T3SS.
Initially, we used the phosphospecific antibody against the phosphorylated Elk peptide contained in the AUIE tag of some chimeric proteins to detect if these chimeric proteins were translocated to the eukaryotic cells by the T3SS. The Elk peptide, derived from the eukaryotic transcription factor Elk-1, is recognized and phosphorylated by eukaryotic protein kinases at serine 383 (25). However, due to the high background obtained with this antibody, we examined the translocation of the chimeric proteins using anti-ESAT-6 antibody, as described in Materials and Methods. To analyze whether the RASV strains χ9930(pYA4251), χ9930(pYA4252), χ9930(pYA4257), and χ9930(pYA4258) were able to translocate the expressed recombinant chimeric proteins via the T3SS to the cytoplasm of the eukaryotic cells, assays were conducted in parallel with the isogenic strain χ11406 (ΔinvAC) transformed independently with each of the same plasmids. The chimeric proteins SopENt80-E3, SopENt80-E3-AU1E, SopENt80-E2C, and SopENt80-E2C-AU1E, which were synthesized and secreted by χ9930, were detected as proteins of 40.3 kDa, 44.9 kDa, 40.4 kDa, and 45.4 kDa, respectively, in the cellular fraction of the INT-407 cells by immunoblotting with anti-ESAT-6 polyclonal antibody (Fig. 4, lanes 2, 4, 6, and 8). However, proteins of the same size from strain χ11406 harboring the same plasmids were undetectable in the cytoplasmic fraction of the INT-407 cells (Fig. 4, lanes 3, 5, 7, and 9). To evaluate the possibility that the internalized bacteria had moved to the cytoplasm of the INT-407 cells rather than had secreted the chimeric proteins into the cytoplasm of the INT-407 cells, the INT-407 cells were analyzed by immunoblotting cytoplasmic fractions using a mouse monoclonal antibody against β-galactosidase (expressed constitutively in the bacteria). We observed no detectable β-galactosidase in the cytoplasmic fractions of the INT-407 cells (Fig. 4). To validate these results, the translocation assay was performed with strain χ9930(pYA4252) in the presence of cytochalasin D, which inhibits the internalization of S. Typhimurium by RAW264-7 cells (59). As expected, the recombinant protein SopENt80-E3-AU1E was detected in the cytoplasmic fraction of the INT-407 cells in the presence of cytochalasin D (Fig. 5, lanes 8 and 9). The results depicted in Fig. 4 and and55 suggest that recombinant chimeric mycobacterial proteins ESAT-6 and CFP-10 fused with the amino-terminal and secretory domains of SopENt80 are expressed and translocated by RASV strains into the cytosol of eukaryotic cells, which should facilitate generation of a specific T-cell immune response against M. tuberculosis. Other investigators who have used SopE or SptP fusions with viral antigens have demonstrated translocation of those fusion proteins into the cytosol of INT-407 cells and have subsequently shown that such fusions generated specific T-cell responses to the viral antigens (27, 65, 68).
To investigate whether the secreted SopENt80-E3-AU1E and SopENt80-E2C recombinant proteins induced IgG antibody responses, we orally immunized groups of C57BL/6 mice on days 0, 21, and 49 with RASV χ8916 harboring the control Asd+ SopENt80 vector (pYA3870) or its pYA4252 or pYA4257 (p15A ori) derivative. Serum IgG responses to ESAT-6 and Salmonella outer membrane proteins (SOMPs) from immunized mice were measured by ELISA. Total IgG responses to ESAT-6 in the groups of mice vaccinated with RASV strains χ8916(pYA4252) and χ8916(pYA4257) had significantly higher anti-ESAT-6 antibody titers at days 20 and 48 (P < 0.001) (Fig. 6A). This induced immune response was further examined by measuring the levels of IgG isotype subclasses IgG1 and IgG2b in preimmune serum at day 20 and at day 48. In mice, IgG1 is associated with a Th2-like response, while a Th1 response is associated with the induction of IgG2a and IgG2b (36). Since the gene coding for IgG2a is deleted in C57BL/6 mice (55), the IgG2b isotype was used as an indicator of a Th1 response in this study. The sera from mice immunized with RASV strain χ8916(pYA4252) or χ8916(pYA4257) had anti-ESAT-6 IgG2b titers higher than the anti-ESAT-6 IgG1 titers at days 20 and 48 (Fig. 6B). These data indicate that χ8916 synthesizing ESAT-6 or ESAT-6–CFP-10 chimeric proteins induces a Th1-related IgG2b antibody response bias (36). Total IgG responses to CFP-10 in mice immunized with χ8916(pYA4257) had significant anti-CFP-10 antibody titers at days 20 and 48 (P < 0.001), although the antibody titers were lower than those to ESAT-6 (Fig. 6C). Significant total IgG responses to SOMPs were observed at 21 days after the first immunization (P < 0.01), and increased levels were observed at day 48 (P < 0.001) (Fig. 6D). Analogous to what has been observed in BALB/c mice immunized with RASV strains (45, 47), the IgG2b titers against the SOMPs were higher than the IgG1 titers (Fig. 6E), suggesting that a Th1 response to these proteins had occurred.
To assess the effect of the genotype of the RASV strain on stimulation of antibody responses to ESAT-6, we orally immunized groups of C57BL/6 mice on days 0, 7, and 49 with Salmonella χ9879 harboring the control Asd+ SopENt80 vector (pYA3870, p15A ori) or its pYA4251 (SopENt80-E3) or pYA4257 (SopENt80-E2C) derivative or with χ9879 harboring the pSC101 ori plasmid pYA4254 (SopENt80-E2C). As was observed in the mice immunized with χ8916 harboring the Asd+ SopENt80 plasmid derivatives, total IgG responses to ESAT-6 were observed in the sera of mice immunized with Salmonella χ9879 harboring pYA4254, pYA4257, or pYA4251, which had significantly higher anti-ESAT-6 IgG levels at day 21 than mice dosed with the vector control strain χ9879(pYA3870) or BSG-dosed mice (P < 0.001) (Fig. 7A). A significant anti-ESAT-6 IgG response was still observed on day 65, after three immunizations, although the titers of total IgG were not as high as those observed in mice after two immunizations with RASV χ8916 delivering the same antigens (Fig. 6A). The levels of the IgG isotype subclasses IgG1 and IgG2b were measured in preimmune serum and at days 21 and 65, and higher anti-ESAT-6 IgG2b titers than anti-ESAT-6 IgG1 titers were observed in all of the groups vaccinated with the RASV strains expressing chimeric antigenic proteins (Fig. 7B). Significant total IgG responses to SOMPs were observed at 21 and 65 days (P < 0.001) (Fig. 7C). Moreover, significant anti-CFP-10 total IgG responses were observed in mice immunized with χ9879(pYA4257) compared with those of preimmune serum and in mice immunized with the pYA3870 vector control (Fig. 7D).
ELISPOT assays were used to compare ESAT-6 or CFP-10 stimulation of IFN-γ (Th1-associated) production by spleen cells from immunized and control C57BL/6 mice (Fig. 8A and B). Splenic lymphocytes isolated from mice immunized with strain χ8916(pYA4252), synthesizing recombinant SopENt80-E3-AUIE, or χ8916(pYA4257), synthesizing SopENt80-E2C, and analyzed 1 week after the last immunization produced significantly more IFN-γ spot-forming units (SFU) than spleen cells from BSG-dosed mice (P < 0.01 and P < 0.001, respectively) or mice from the χ8916(pYA3870) control group (P < 0.05) (Fig. 8A). Similar results were obtained for CFP-10-specific IFN-γ-secreting cells from mice immunized with strain χ8916(pYA4257) compared to the BSG-dosed control group (P < 0.05), although the number of SFU was much lower than the number of ESAT-6-specific IFN-γ-secreting cells (Fig. 8B). These results indicate that the Salmonella-vector system designed to deliver ESAT-6 and CFP-10 as fused proteins by the T3SS was able to stimulate ESAT-6- and CFP-10-specific IFN-γ-secreting cells.
For the experiments whose results are depicted in Fig. 9, lymphocytes cultured from spleens isolated from each group of C57BL/6 mice 3 weeks after the last immunization were subjected to ELISPOT assays to compare production of the proinflammatory Th1 cytokines IFN-γ and TNF-α and the anti-inflammatory Th2 cytokines IL-4 and IL-10 (67). The splenocytes were restimulated for 24 h with 1 μg/well of recombinant ESAT-6 or medium for IFN-γ and TNF-α and for 48 h for IL-4 and IL-10. The number of ESAT-6-specific IFN-γ SFU from the splenocytes of mice immunized with strain χ9879(pYA4257) or χ9879(pYA4254), synthesizing SopENt80-E2C, or χ9879(pYA4251), synthesizing SopENt80-E3, was significantly higher than that in the BSG-dosed mouse group (P < 0.001) (Fig. 9A). However, only splenocytes from mice vaccinated with χ9879(pYA4257) showed a significantly higher number of IFN-γ SFU than the control group dosed with χ9879(pYA3870) (P < 0.05). The number of ESAT-6-specific TNF-α SFU from splenocytes from all of the groups of mice vaccinated with χ9879(pYA4251), χ9879(pYA4254), and χ9879(pYA4257) was significantly higher than that in the BSG control group (P < 0.001), and the highest production was detected in mice vaccinated with χ9879(pYA4257) (Fig. 9B). The number of TNF-α SFU produced by this group was not significantly different from that of the vector control group dosed with χ9879(pYA3870). Very low levels of ESAT-6-specific IL-4 SFU were detected from splenocytes of mice vaccinated with χ9879 expressing any of the SopE-M. tuberculosis fusion proteins (data not shown). Production of IFN-γ and TNF-α SFU by the vector control group in response to ESAT-6 was surprising, and we do not understand the basis for these results, since ESAT-6 is an M. tuberculosis-specific protein. Production of IL-10 from splenocytes was not detected in any of the groups of vaccinated mice (data not shown). These results indirectly suggested a Th1 immune response, characterized by the secretion of IFN-γ and TNF-α in mice immunized with the strains expressing the recombinant antigens, which was significantly higher for the strain χ9879 harboring pYA4257 (p15A ori) and synthesizing SopENt80-E2C than for χ9879 harboring pYA4254 (pSC101 ori) and synthesizing the same recombinant protein or for χ9879(pYA4251) synthesizing SopENt80-E3 (Fig. 9A and B).
To examine the protective efficacy of Salmonella RASV–ESAT-6 vaccines against M. tuberculosis infection, the immunized mice were challenged with a low aerosol dose (100 bacteria per lung) of virulent M. tuberculosis H37Rv at 4 weeks after the last immunization. Five mice per group were euthanized at 6 weeks postchallenge, and protection was measured by enumeration of M. tuberculosis CFU in the lungs and spleens. The group of mice orally immunized with the χ8916(pYA3870) vector control did not show a reduction in the bacterial load in the lungs and spleens compared with the buffered saline (BSG)-treated control group (Fig. 10A and B). The group of mice immunized with χ8916(pYA4252), synthesizing SopENt80-E2C-AUIE, showed a modest reduction of the bacterial load in the lungs, but not in the spleen, compared with the χ8916(pYA3870) control group. The group of mice vaccinated with χ8916(pYA4257), synthesizing SopENt80-E2C, showed a significant reduction in the bacterial load in the lungs and spleens compared with mice immunized with χ8916(pYA3870) (P < 0.05), but the load was not reduced to the extent that it was in mice immunized with M. bovis BCG (positive vaccine control) (Fig. 10A and B).
To examine the protective efficacy of the arabinose-regulated RASV strain against M. tuberculosis infection, groups of C57BL/6 mice immunized with χ9879 expressing the SopENt80-mycobacterial antigens were challenged as described above with virulent M. tuberculosis H37Rv. Six mice per group were euthanized at 6 weeks postchallenge, and the protection was measured as described above. The group of mice orally immunized with the χ9879(pYA3870) vector control or χ9879(pYA4254) (pSC101 ori), synthesizing SopENt80-E2C, showed reductions of bacterial loads in the lungs and spleens compared with the buffered saline (BSG)-treated control group, but this difference was not statistically significant (Fig. 11A and B). The groups of mice immunized with χ9879 bearing the p15A ori plasmid pYA4257, which synthesizes SopENt80-E2C, or pYA4251, which synthesizes SopENt80-E3, showed greater reductions in bacterial loads in the lungs and spleens than the BSG-dosed control groups (P < 0.05) but not the M. bovis BCG-immunized groups (Fig. 11A and B).
Rüssmann et al. (65) first showed that viral antigen epitopes could be delivered by the Salmonella T3SS to elicit protective immune responses. Shams et al. (68) demonstrated that antigens of the lymphocytic choriomeningitis virus (LCMV) fused to SptP and thus delivered by the T3SS elicited specific antiviral cyotoxic T-lymphocyte (CTL) responses, as well as induced production of specific CD8+ memory T cells, following intragastric immunization of mice with an attenuated S. Typhimurium isolate producing the SptP-LCMV antigen fusions. These investigators also demonstrated long-lasting protection of immunized mice against intracranial infection with live LCMV, indicating successful stimulation of cell-mediated immunity as a consequence of antigen delivery by the Salmonella T3SS (68). Konjufca et al. (48) delivered antigens from two Eimeria species as fusion proteins with the SPI-1 effector proteins SptP and SopE, using S. Typhimurium RASV strain χ8879 to orally immunize chickens. Like M. tuberculosis, Eimeria is an intracellular pathogen and cell-mediated immunity is necessary to provide protection to chickens against infection. Konjufca et al. demonstrated that immunization of chickens with RASV χ8879 producing the Eimeria antigens, which were delivered by the T3SS, induced protection against infection with E. acervulina, a significant pathogen causing coccidiosis in chickens (48). SopE has been fused to fragments of the simian immunodeficiency virus (SIV) Gag antigen and delivered to rhesus macaques by attenuated S. Typhimurium strains via the T3SS (27). These investigators demonstrated Gag-specific CTL responses, which were enhanced when the macaques received a booster immunization with vaccinia virus Ankara producing the SIV Gag protein (27). Gag-specific CD8+ T-cell responses were detected in the peripheral blood and in lymphocytes isolated from the colons of the immunized macaques (27). Heterologous antigens delivered by the T3SS to secrete and translocate the Gag protein from the human immunodeficiency virus (HIV) have also been described to be a potential vaccine (13). The Gag protein was fused to the secretion and translocation signals of SopE and was effectively translocated to the cell cytosol to be presented by MHC class I (13). Oral immunization of mice with 108 CFU of S. enterica serovar Typhimurium ΔphoP ΔphoQ expressing the optimized HIV Gag protein, followed by an intraperitoneal boost with 104 CFU of an S. Typhimurium Δasd strain expressing the same antigen 4 weeks later, elicited T-cell responses with significantly large numbers of Gag-specific CD8+ T cells (13). Therefore, we hypothesized that the ability to inject effectors of the T3SS could be utilized for the cytosolic delivery of M. tuberculosis antigens by attenuated Salmonella strains. Protection against intracellular pathogens such as M. tuberculosis depends on the induction of cell-mediated immunity. CD4+ T cells play a central role in protection against infections caused by M. tuberculosis (57), and CD8+ T cells are also important for protection against this pathogen (28, 73). In this study, two plasmid vectors were genetically engineered to encode the secretion and translocation signals of Salmonella T3SS effector SopE protein for cytosolic injection of M. tuberculosis T-cell antigens to induce cellular immunity against M. tuberculosis in orally immunized mice.
Live attenuated Salmonella strains have been shown to induce mucosal, humoral, and cell-mediated immune responses to heterologous antigens (13, 19, 24, 27, 48, 68). Other vaccines against M. tuberculosis based on attenuated Salmonella strains have shown induction of immune responses to M. tuberculosis antigens, such as antigen 85B (Ag85B ) or fusions of Ag85B and ESAT-6 (38, 78). One study using S. Typhimurium SL7207 aroA harboring the plasmid pMO6esat, which expresses and secretes ESAT-6 through an HlyA secretion system, delivered intravenously reduced the numbers of M. tuberculosis H37Rv CFU in the lungs of mice challenged intravenously with a single dose of 5 × 105 CFU of H37Rv (58). The secretion of antigens by the Salmonella strain was important to elicit immune responses against M. tuberculosis. However, these studies were conducted using a Salmonella strain with a single attenuating mutation (12), and both immunization and challenge were administered by intravenous inoculation, which is not the normal route of entry of either Salmonella or M. tuberculosis. Intravenous injection of either of these bacteria has been shown to elicit different patterns of immune responses than inoculation by normal routes of entry (12, 19, 24, 61). In the present study, we evaluated the potential of RASVs harboring Asd+ SopENt80 vectors with different replication origins, including those with low copy numbers (p15A) and very low copy numbers (pSC101), and encoding two mycobacterial protective antigens, ESAT-6 (fused in tandem two or three times) and CFP-10, to be delivered into the cell cytosol of the immunized host by the Salmonella T3SS to elicit specific immune responses against M. tuberculosis. Two schedules were employed for oral immunization of mice. When the vaccine strain χ8916 was used, mice were immunized at days 0, 21, and 49. For the χ9879 strain, mice were immunized at days 0, 7, and 49. We reasoned that the second vaccination given at day 7 could result in a better colonization of the host tissues by the vaccine strain than immunization at day 21, when an immune response against the Salmonella vaccine was increasing. Interestingly, we did not observe differences in the antibody responses elicited by either Salmonella vaccine strain, even with the variation of the schedule of immunization. However, with the second immunization schedule using χ9879, significantly lower levels of IFN-γ were produced by splenocytes of mice immunized with the RASVs expressing the mycobacterial antigens. These lower levels of IFN-γ could be due to the fact that the ELISPOT assay was performed at a different time than it was with χ8916 (at week 3 instead of week 1, after the last immunization) or to the genotype of χ9879, which harbors several deletion-insertion mutations and delayed attenuation, or to both conditions. We also examined the effect of different copy numbers of the plasmid vectors on the immune response. We observed that the serum IgG responses to ESAT-6 were similar in χ9879 harboring either the Asd+ SopENt80 plasmid derivative pYA4254 (pSC101 ori) or the derivative pYA4257 (p15A ori) and encoding the same chimeric protein, SopENt80-E2C. The levels of total IgG in the sera were generally not affected by the vaccine strain or the copy number of the plasmid vector, although χ8916 harboring pYA4257 elicited high levels of anti-ESAT-6 for a longer period than the other RASV constructs. The IgG subclass distribution is dependent on several factors, including the cytokine environment, the type of cells that are presenting the antigen, and the dose of antigen (71). The IgG subclass distribution was not affected by the use of vaccine vectors with low (p15A ori) or very low (pSC101 ori) copy numbers. The predominant subclass of anti-ESAT-6 was IgG2b, which is characteristic of a Th1 response (35). These data suggest that the lower dose of antigen synthesized and delivered by the Salmonella vaccine strain harboring pYA4254 (pSC101 ori) also presented a bias toward a Th1 immune response. However, the levels of anti-ESAT-6 IgG2b and IgG1 were lower in the mice immunized with χ9879 harboring the pSC101 ori plasmid than in the mice immunized with χ9879 harboring the p15A ori plasmid (Fig. 7B).
The antigen-specific stimulation of cytokine production was influenced by the copy number of the vaccine plasmids. Our results showed that both strains χ8916 and χ9879 harboring pYA4257 (p15A ori), synthesizing the chimeric SopENt80-E2C protein, induced significantly higher IFN-γ in splenocytes from vaccinated mice than χ9879 harboring pYA4254 (pSC101 ori) and synthesizing the same antigenic protein. The exception was with χ9879 harboring pYA4251 (p15A ori), synthesizing SopENt80-E3 protein, which induced IFN-γ production similar to that for χ9879 harboring pYA4254. Low but significant CFP-10-specific IFN-γ production was also induced in mice immunized with both χ8916 and χ9879 harboring pYA4257. Production of IFN-γ and TNF-α is crucial to eukaryotic cells for fighting infections caused by intracellular pathogens such as M. tuberculosis and may play a role in the clearance of bacteria from the infected host (16, 30). We detected very low numbers of IL-4-secreting cells, but IL-10 secretion was not detected in mice vaccinated with any RASV strain. These results (higher levels of IgG2b antibodies and elicitation of IFN-γ) suggest that the RASV-M. tuberculosis vaccines are stimulating a Th1 response and are in agreement with observations from previous studies in which oral immunization of mice with Salmonella induced a bias toward a Th1 immune response (62), delivery of viral and parasite antigens via the Salmonella T3SS elicited cell-mediated immune responses to those antigens (13, 27, 48, 68), and vaccination with the Salmonella phoP mutant elicited IFN-γ-dependent cellular immunity (13, 65).
In this study, we observed that the protective efficacy conferred by the Salmonella vaccines in orally immunized mice was influenced by the dose of antigen expressed and delivered from the vaccine strains harboring plasmids with different copy number replication origins. A significant decrease in the number of M. tuberculosis CFU was observed in the lungs and spleens of mice immunized with χ9879(pYA4257) or χ9879(pYA4251), containing the p15A replication origin, compared to mice immunized with χ9879(pYA4254), containing the pSC101 replication origin. Both of the strains (χ8916 and χ9879) with pYA4257 were the RASVs that induced the highest stimulation of IFN-γ and TNF-α secretion, resulting in significant protection of mice against M. tuberculosis infection, although neither achieved the level of protection observed in the mice immunized with M. bovis BCG by a subcutaneous route (the “gold standard” for TB vaccines). Additionally, strains χ8916(pYA4252), synthesizing SopENt80-E3-AU1E, and χ9879(pYA4254), synthesizing SopENt80-E2C, showed a modest reduction in the number of M. tuberculosis CFU that was not significant compared with the number for the BSG control. Interestingly, strain χ9879(pYA4251), synthesizing SopENt80-E3, which induced lower IFN-γ and TNF-α production, similar to that observed with χ9879(pYA4254), was able to confer the same level of protection against M. tuberculosis infection as χ8916(pYA4257).
Based on the results of other investigators who have used SopE fusions with viral and parasite antigens and demonstrated elicitation of protective cell-mediated immune responses (13, 27, 48, 68) and our demonstration that our SopE-M. tuberculosis fusion proteins are translocated into the cytoplasm of INT-407 cells, we hypothesize that the RASV strains harboring Asd+ SopENt80 plasmid derivatives allowed successful secretion and translocation of the mycobacterial antigenic chimeric proteins into the host cell cytosol to become accessible to the MHC class I-restricted processing pathways to likely stimulate T-cell immune responses. Thus, the RASV strains harboring Asd+ SopENt80 vaccine plasmid pYA4257 or pYA4251 administered orally to mice produced a significant reduction of the bacterial load in the lungs and spleens of mice orally vaccinated and challenged with aerosol-delivered M. tuberculosis. The protection conferred by the RASV strains harboring the Asd+ SopENt80 plasmids expressing M. tuberculosis antigens could be improved by using a heterologous prime-boosting strategy with antigens delivered as a subunit vaccine and an adjuvant or combined with M. bovis BCG vaccination. However, since M. bovis BCG lacks the esxA and esxB genes, immunizing first with BCG would necessitate at least two immunizations with RASVs expressing ESAT-6 and CFP-10 or a combination of RASV immunization followed by immunization with a subunit vaccine or inclusion of genes expressing additional antigens of M. tuberculosis that are shared with M. bovis BCG in the RASV strains. Taken together, the data produced here encourage the use of the RASV strains harboring Asd+ SopENt80 vaccine plasmids to deliver protective T-cell antigens by the T3SS to induce the T-cell responses required for protection against M. tuberculosis TB.
We thank Ascencion Torres-Escobar and Praveen Alamuri for their valuable suggestions and critical reviews of the manuscript.
This research was supported by National Institutes of Health grant AI 56289.
Published ahead of print 5 December 2011
Supplemental material for this article may be found at http://iai.asm.org/.