Generation of Mucosal Anti-Human Immunodeficiency Virus Type 1 T-Cell Responses by Recombinant Mycobacterium smegmatis^▿

Abstract

A successful vaccine vector for human immunodeficiency virus type 1 (HIV-1) should induce anti-HIV-1 immune responses at mucosal sites. We have generated recombinant Mycobacterium smegmatis vectors that express the HIV-1 group M consensus envelope protein (Env) as a surface, intracellular, or secreted protein and have tested them in animals for induction of both anti-HIV-1 T-cell and antibody responses. Recombinant M. smegmatis engineered for expression of secreted protein induced optimal T-cell gamma interferon enzyme-linked immunospot assay responses to HIV-1 envelope in the spleen, female reproductive tract, and lungs. Unlike with the induction of T-cell responses, priming and boosting with recombinant M. smegmatis did not induce anti-HIV-1 envelope antibody responses, due primarily to insufficient protein expression of the insert. However, immunization with recombinant M. smegmatis expressing HIV-1 Env was able to prime for an HIV-1 Env protein boost for the induction of anti-HIV-1 antibody responses.

A vector that induces mucosal immunity will be an important component of an effective human immunodeficiency virus type 1 (HIV-1) vaccine. Recombinant bacillus Calmette-Guérin (rBCG) has been shown to induce systemic anti-HIV-1 T-cell responses in animals and humans (1, 11-13, 15, 24). However, in patients with immune deficiency, BCG can cause, albeit rarely, a clinically significant mycobacterial infection (4, 30). We have previously demonstrated that recombinant Mycobacterium smegmatis can induce systemic anti-HIV-1 T-cell responses to the wild-type HIV-1 group M consensus envelope (Env) epitope expressed as a surface protein (2).

A strategy to address HIV-1 diversity for the induction of broadly reactive anti-HIV-1 T-cell responses is the design of artificial group M consensus genes (6). As we have previously described, the HIV-1 group M consensus env gene CON6 induces anti-HIV-1 T-cell responses with a breadth equal to that of polyvalent HIV-1 immunogens (5, 28).

In this paper, we describe recombinant M. smegmatis vectors constructed to express the HIV-1 group M consensus env gene CON6 either as a surface, intracellular, or secreted protein and determine the insert expression and immunogenicity of these constructs. Moreover, we have determined the ability of HIV-1 Env expressed in recombinant M. smegmatis to induce mucosal T-cell immune responses after systemic immunizations. We found that recombinant M. smegmatis induced lower anti-HIV-1 T-cell responses than did a recombinant adenovirus (rAd)-recombinant vaccinia virus (rVV) prime/boost in the spleen but induced similar gamma interferon (IFN-γ) enzyme-linked immunospot (ELISPOT) assay responses to rAd-rVV in the female reproductive tract (FRT) and lungs. These data demonstrate that recombinant M. smegmatis is immunogenic for the induction of HIV-1 T-cell responses at mucosal surfaces.

MATERIALS AND METHODS

Bacterial strains and plasmids.

M. smegmatis mc² 155 was used for the generation of recombinant M. smegmatis. Table 1 summarizes the mycobacterial plasmids that were used to transform M. smegmatis mc² 155 for expression of the insert HIV-1 envelope (env) CON6-gp120 or CON6-gp140CF gene (23). Expression of the HIV-1 CON6 env gene in the pMV261 and pMV361 vectors is regulated by the Mycobacterium tuberculosis heat shock protein 60 (hsp60) expression cassette (26). pMV261 is a constitutive expression plasmid, while expression of Hsp60 fusion protein in pMV361 was increased in response to stress with heat, acid, and peroxide (26). pJH plasmids are kanamycin resistant and were designed to express foreign proteins in different locations under regulation of the M. tuberculosis α-antigen promoter. Plasmid pJH152 was used to express the CON6 env gene fused to the M. tuberculosis 19-kDa signal sequence as a surface protein. Plasmid pJH153 was used to express the CON6 env gene as an intracellular protein. Plasmid pJH154 was used to express the CON6 env gene fused to an M. tuberculosis α-antigen export signal sequence as a secreted protein. Plasmids pYUB2051, pYUB2052, and pYUB2053 are the same as plasmids pJH152, pJH153, and pJH154, respectively, except that the pYUB plasmids contain the hygromycin resistance gene instead of the kanamycin resistance gene.

TABLE 1.

Plasmid constructs generated for expression of HIV-1 CON6-gp120 and -gp140 in M. smegmatis

Vector	Designed for:	Marker	Insert(s)	Cloning site
pMV261	Intracellular expression	Kanamycin	CON6-gp120 and CON6-gp140CF	BamHI/HpaI
pMV361	Chromosomal integrated vector	Kanamycin	CON6-gp120 and CON6-gp140CF	EcoRI/HpaI
pJH152	Surface expression	Kanamycin	CON6-gp120 and CON6-gp140CF	ApaI/HpaI
pJH153	Intracellular expression	Kanamycin	CON6-gp120 and CON6-gp140CF	ApaI/HpaI
pJH154	Secreted expression	Kanamycin	CON6-gp120 and CON6-gp140CF	ApaI/HpaI
pYUB2051	Intracellular expression	Hygromycin	CON6-gp120 and CON6-gp140CF	ApaI/HpaI
pYUB2052	Surface expression	Hygromycin	CON6-gp120 and CON6-gp140CF	ApaI/HpaI
pYUB2053	Secreted expression	Hygromycin	CON6-gp120 and CON6-gp140CF	ApaI/HpaI
pJH222	Surface expression	Kanamycin	CON6-gp140CF	NdeI/SpeI

Construction of the HIV-1 env gene in mycobacterial expression vectors.

To clone the group M consensus CON6 gp120 and CON6 gp140CF env genes (6) into various plasmids (Table 1) at the desired cloning sites, DNA fragments were generated by PCR with primers containing sequences of the corresponding restriction enzyme sites. Specifically, to clone the HIV-1 CON6 gp120 and gp140CF env genes in pMV261 and pMV361, CON6 gp120 and gp140CF env DNAs were amplified by PCR using Platinum Pfx polymerase (Invitrogen) with the forward primer 5′-ATGCGCGGATCCAGAATTCATGCGCGTGATGGGC-3′ and the backward primer set 5′-ATCGGCTCGCGTTAACTAGGCCTTGGTGGGG-3′ (for amplifying gp120) and 5′-ATCGGCTCGCGTTAACCTACCACAGCCAGTTGGT-3′ (for amplifying gp140CF). For the construction of gp120 and gp140CF in the pJH and pYUB series plasmids, the HIV-1 env DNAs were amplified by PCR using the forward primer 5′-ATGCGCGCGCGGGCCCAACCTGTGGGTGACCGTG-3′ and the backward primers described above. Enzymatic manipulations and cloning of the CON6 gp120 and CON6 gp140CF env genes into various plasmids were performed as described previously (22). All constructs were confirmed by DNA sequencing.

Generation of recombinant M. smegmatis expressing HIV-1 envelope.

All M. smegmatis cultures were grown in Middlebrook 7H9 broth (Difco) containing 10% albumin-dextrose saline (ADS)-0.5% glycerol-0.05% Tween 80 (7). For generation of recombinant M. smegmatis, M. smegmatis mc² 155 prepared in 10% glycerol was transformed with the constructed plasmids (Table 1) by electroporation with a Gene Pulsar (Bio-Rad) set at 2.5 kV and 25 μF and with the pulse controller resistance set at 1,000 Ω (23, 27). Transformed M. smegmatis organisms were selected on Middlebrook 7H10 (Difco, Sparks, MD) agar plates supplemented with 10% ADS containing either 30 μg/ml kanamycin or 20 μg/ml hygromycin. To monitor the expression of HIV-1 gp120 or gp140CF, individual colonies of recombinant M. smegmatis grown in Middlebrook 7H9-ADS-Tween broth in the presence of 30 μg/ml of kanamycin or hygromycin were harvested by centrifugation. After a rinse with sterile phosphate-buffered saline, mycobacterial cells were lysed by using the modified extraction buffer with 106-μm glass beads (Sigma) (7) and cell lysates were cleared by centrifugation. The lysate of recombinant M. smegmatis was fractionated by 4 to 20% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and blotted onto nitrocellulose filters (Schleicher & Schuell, Germany). HIV-1 Env protein bands were detected by incubation of the filters with the monoclonal antibody (MAb) T8, 7B9, or 7B2 at 1 μg/ml, followed by incubation with goat anti-mouse immunoglobulin G (IgG)- or goat anti-human IgG (Sigma, St. Louis, MO)-alkaline phosphatase. T8 is a mouse MAb that binds to the HIV-1 Env gp120 C1 region (a gift from Pat Earl, NIH, Bethesda, MD). 7B2 is a human MAb against the immunodominant region of the HIV-1 Env gp41 region and was a gift from James Robinson (Tulane Medical School, New Orleans, LA). 7B9 is a mouse IgG1 MAb made from murine B cells immunized with the 62.19 HIV-1 clade B V3 loop sequence (RPNNNTRKSIHIGPGRAFYATE) (18) that reacts with the CON6 gp120 V3 loop.

Mouse immunizations.

Female BALB/c mice (6 to 8 weeks of age) were purchased from Charles River Laboratory (Raleigh, NC) and housed in the Duke Cancer Center Isolation Facility under AALAC guidelines with animal use protocols approved by the Duke University Animal Use and Care Committee and the Duke University Institutional Biosafety Committee. Mice were immunized either intraperitoneally (i.p.) or intradermally (i.d.) with recombinant M. smegmatis expressing CON6 gp120 or gp140CF using doses ranging from 10⁷ to 10¹⁰ CFU/mouse in 100 μl of phosphate-buffered saline containing 0.05% Tween 80. For boosting immunizations with CON6 Env protein, 50 μg of recombinant CON6-gp140CF oligomer (6) was mixed in an equal volume with RiBi adjuvant (Sigma) and administrated i.p. and subcutaneously (s.c.). After immunization, serum was collected for antibody assays every 2 weeks. Mouse spleen, lung, and FRT samples were harvested for evaluation of T-cell responses 10 days after the final immunization.

ELISA and antibody isotype analysis.

Serum antibody titers against HIV-1 CON6-gp140CF or M. tuberculosis whole-cell lysate (WCL) (Colorado State University, Fort Collins, CO) were measured in a standard enzyme-linked immunosorbent assay (ELISA). For isotype analysis of serum antibodies, we used goat anti-mouse IgG (heavy chain specific), IgG1, IgG2a, IgG2b, IgG3, or IgM (Southern Biotechnology) as alkaline phosphatase-conjugated antibodies. Antibody endpoint titers were determined as the reciprocal of the highest dilution of the serum assayed against recombinant HIV-1 envelope proteins or M. tuberculosis WCL giving an optical density reading of experiment/control of ≥3 times over the background activity of prebleed serum samples. Geometric mean titers were determined based on the endpoint titer of each mouse serum. Statistical significance was assessed by Student's t test.

IFN-γ ELISPOT assays.

ELISPOT assays were performed as described previously (20). Briefly, 96-well flat-bottom plates (Millipore) were coated with an anti-mouse IFN-γ capture antibody (Pharmingen). After the plates were blocked and washed, lymphocytes freshly isolated from the spleen, lungs, and FRT of immunized mice were added and incubated with a mixture of either 10 CON6-gp140 Env 15-mer peptides (for determining T-cell responses to the HIV-1 Env insert) that have been previously shown to represent CON6 Env epitopes recognized in BALB/c mice (6, 28; E. Weaver, Z. Lu, H. X. Liao, B. Ma, M. S. Alam, R. M. Scearce, L. L. Sutherland, J. M. Decker, Z. Hartman, A. Amalfitano, B. T. Korber, B. H. Hahn, D. C. Montefiori, and B. Haynes, AIDS Vaccine 2003 Meet., abstr. 43, 2003) or M. tuberculosis WCL (for determining T-cell responses to the M. smegmatis vector itself) for 24 h at 37°C in a 5% CO₂ incubator. For the counting of spot-forming cells (SFC), ELISPOT plates were scanned into an ImmunoSpot Series I analyzer and spots were quantitated with ImmunoSpot 2.1 software (CTL Analyzers, Cleveland, OH). Cells cultured in medium in the absence of peptides or M. tuberculosis WCL were used as controls. For analysis of mucosal T-cell responses, extracted lungs or FRT were pooled from three to four mice and assayed as one data point. At least three data points representing three groups of mice were analyzed. Statistical significance was assessed by Student's t test.

RESULTS

Expression of CON6 gp120 and gp140CF HIV-1 group M consensus Env proteins in M. smegmatis mc² 155.

We used various mycobacterial expression vectors for the expression of CON6 gp120 Env or gp140CF Env as a surface, intracellular, or secreted protein. HIV-1 CON6 Env protein expression in M. smegmatis was detected by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and Western blot analysis (Fig. 1A). No HIV-1-specific protein bands were detected in untransformed M. smegmatis (Fig. 1A, lane 2) or recombinant M. smegmatis transformed with empty plasmid (lane 3). Figure 1A shows that the expression of the intact CON6 gp120 was detected in recombinant M. smegmatis transformed with either the surface expression plasmid pJH152 (Fig. 1A, lane 4), the intracellular expression plasmid pJH153 (lane 5), or the secreted expression plasmid pJH154 (lane 6). Expression of CON6 gp140CF was also detected in recombinant M. smegmatis transformed with the surface expression plasmid pJH152 (lane 7) or pJH222 (lane 8). Both the intact and partially cleaved gp140 products were identified with the gp120-specific MAb T8 (anti-C1 gp120 region), the gp41-specific MAb 7B2, and V3 loop-specific MAb 7B9. The calculated nonglycosylated molecular masses based on amino acid sequences were 53 kDa for CON6 gp120 and 71 kDa for CON6 gp140CF. However, the protein bands of CON6 gp120 and CON6 gp140CF expressed in recombinant M. smegmatis were approximately 80 kDa and 90 kDa, respectively, suggesting that the CON6 gp120 and CON6 gp140CF proteins expressed in recombinant M. smegmatis were partially glycosylated.

FIG. 1.

Expression of HIV-1 CON6 gp120 and gp140CF envelope proteins in M. smegmatis. (A) Western blot analysis of CON6 Env protein expression in various M. smegmatis designs, including molecular weight (Mr.) standards (lane 1), untransformed M. smegmatis (lane 2), and recombinant M. smegmatis transformed with either empty plasmid (lane 3), the surface expression plasmid pJH152-CON6 gp120 (lane 4), the intracellular expression plasmid pJH153-CON6 gp120 (lane 5), the secreted expression plasmid pJH154-CON6 gp120 (lane 6), or the surface expression plasmid pJH152-CON6 gp140 (lane 7) or pJH222-CON6 gp140 (lane 8). Both the intact (89 kDa for gp140, 79 kDa and 74 kDa for gp120) and partially cleaved (49 kDa and 34 kDa) CON6 Env protein products, as indicated by arrows, were found with gp120 MAb T8 (anti-C1 gp120 region), gp41-specific MAb 7B2, and V3 loop-specific MAb 7B9. (B) Schematic representation of the full-length (top) and cleaved (bottom) gp140, indicating the binding locations of MAbs T8, 7B2, and 7B9. Full-length gp140 could be detected by T8, 7B2, or 7B9, while the cleaved 34-kDa Env protein band could be detected only by MAb T8. A 50-kDa protein band could be detected by both MAbs 7B2 and 7B9. (C) Western blot analysis of supernatants from untransformed M. smegmatis (lane 1) and recombinant M. smegmatis transformed with the empty plasmid (lane 2) or pJH154-CON6gp140CF (lane 3) for detecting secreted CON6 Env protein expression. A truncated gp140 protein (lane 3) could be detected by 7B9, as indicated by the arrow. A high-molecular-weight protein band, as indicated by an arrowhead, was nonspecific, as it was also detected in recombinant M. smegmatis transformed with pJH154-CON6gp140CF (lane 3) and in the negative control lanes 1 and 2.

Figure 1B shows schematic representations of the full-length (top) and cleaved (bottom) CON6 gp140, with indications of the binding locations of MAbs T8, 7B2, and 7B9. In the cleaved gp140, MAb T8 reacted only with the 34-kDa protein band, while MAbs 7B2 and 7B9 reacted with the 49-kDa protein band (Fig. 1B). These data suggest that the 49-kDa cleaved protein contained portions of the V3 loop region and the gp41 domain. In addition, analysis of the supernatant of recombinant M. smegmatis expressing secreted CON6 gp140 that was grown in Sauton liquid medium revealed that HIV-1 CON6 Env was produced as a truncated soluble protein containing the V3 loop in tissue culture supernatant, as detected by MAb 7B9 (Fig. 1C).

Detection of HIV-1 envelope-specific IFN-γ SFC in the spleen.

Mice were immunized with recombinant M. smegmatis expressing HIV-1 gp120 as a secreted insert at doses ranging from 10⁷ to 10¹⁰ CFU at week 0 and week 8 via i.p. injections. Unimmunized mice were used as negative controls. Animals that were primed with rAd5 expressing CON6 Env (rAd-CON6) and boosted with rVV expressing CON6 Env (rVV-CON6) were used as positive controls. Splenocytes from animals primed once with recombinant M. smegmatis i.p. yielded no CON6 gp120-specific responses (<5 SFC/10⁶ cells). The lack of response to a single immunization with recombinant M. smegmatis was consistent regardless of the concentration of recombinant M. smegmatis used (10⁹, 10⁸, 10⁷, or 10⁶ CFU). When mice were boosted i.p. with recombinant M. smegmatis 8 weeks following the recombinant M. smegmatis prime, HIV-1 CON6 gp120-specific responses in spleens were easily detectable (Fig. 2A). Mice primed and boosted with 10⁹ CFU of recombinant M. smegmatis produced 66 ± 6 SFC/10⁶ cells, a significantly greater (P < 0.01) number of SFC than mice primed and boosted with lower doses (10⁸, 10⁷, or 10⁶ CFU). Mouse splenocytes in positive-control mice primed with rAd-CON6 and boosted with rVV-CON6 Env produced 358 ± 155 IFN-γ SFC/10⁶ cells.

FIG. 2.

HIV-1 envelope-specific and vector-specific T-cell responses in the spleen, lungs, and FRT of mice immunized with recombinant M. smegmatis. HIV-1-specific (A to C) and vector-specific (D to F) T-cell responses in the spleen (A and D), lungs (B and E), and FRT (C and F) were assessed by IFN-γ ELISPOT assays. Mice that were primed with rAd-CON6 and boosted with rVV-CON6 were used as positive controls. Unimmunized mice were used as negative controls. Empty bars represent immunization groups that received prime/boost immunizations. Solid bars represent immunization groups that received priming immunization only. Mean (± standard error of the mean) numbers of SFC per 10⁶ cells are shown on the y axis. The indicated immunization groups and doses are shown on the x axis. Assays were performed and analyzed with pooled lung and FRT samples as described in Materials and Methods.

Mucosal anti-HIV-1 T-cell responses induced by recombinant M. smegmatis.

In addition to antigen-specific responses to CON6 Env peptides in splenocytes, we also examined mucosal antigen-specific T-cell responses in lungs and FRT of immunized mice. Mice immunized once with recombinant M. smegmatis produced no mucosal CON6 gp140 peptide-specific IFN-γ SFC regardless of the number of doses. However, upon boosting of mice with the same initial dose of recombinant M. smegmatis, IFN-γ ELISPOT assay responses were detected 2 weeks following boosting. In the lungs, a prime/boost with 10⁸ CFU of recombinant M. smegmatis gave the highest frequency (89 ± 2 SFC/10⁶ cells) (Fig. 2B) of IFN-γ SFC, a significantly greater number than that seen with 10⁹ CFU (52 ± 2 SFC/10⁶ cells), 10⁷ CFU (8 ± 4 SFC/10⁶ cells), or 10⁶ CFU (6 ± 2 SFC/10⁶ cells) (P < 0.05, P < 0.01, and P < 0.01, respectively). Mice primed with rAd-CON6 and boosted with rVV-CON6, produced 201 ± 89.6 IFN-γ SFC/10⁶ cells in lungs.

We also found that after a prime and boost of recombinant M. smegmatis, CON6 peptide-specific responses were detectable in the mouse FRT (Fig. 2C). Lymphocytes isolated from the FRT of mice that were primed and boosted with 10⁹ CFU of recombinant M. smegmatis produced 63 ± 4 IFN-γ SFC/10⁶ cells, a response comparable to that induced by the rAd prime/rVV boost immunizations (77 ± 36 IFN-γ SFC/10⁶ cells).

Vector-specific T-cell responses.

Responses to the M. smegmatis vector were also monitored in the spleen, lungs, and FRT. Immunizations of mice with 10⁹, 10⁸, 10⁷, and 10⁶ CFU of recombinant M. smegmatis pJH154-gp120 induced 832 ± 111, 738 ± 75, 591 ± 142, and 154 ± 44 SFC/10⁶ cells in splenocytes (Fig. 2D), respectively. Optimal concentrations for the induction of M. tuberculosis WCL-specific immunity were 10⁸ CFU in FRT (923 ± 212 SFC/10⁶ cells) and 10⁷ CFU in the lungs (1,169 ± 102 SFC/10⁶ cells) (Fig. 2E and F). Unlike the HIV Env-specific T-cell responses, detectable T-cell responses to the vector in the spleen, lungs and FRT were induced with a single priming immunization with recombinant M. smegmatis.

IFN-γ SFC responses were optimal in mice immunized with constructs that secreted the HIV-1 CON6 protein.

We next compared the immunogenicity of recombinant M. smegmatis constructs designed to express HIV-1 Env as a surface, intracellular, or secreted antigen. Mice were first immunized with recombinant M. smegmatis at doses ranging from 10⁷ to 10¹⁰ CFU at week 0 and week 10 by i.p. or i.d immunization. Untransformed M. smegmatis and recombinant M. smegmatis transformed with empty plasmid (pJH222) were used as negative controls. Immunogen groups were recombinant M. smegmatis expressing HIV-1 gp120 as full-length proteins (pJH152-gp120, pJH153-gp120, and pJH154-gp120) and recombinant M. smegmatis expressing HIV-1 gp140CF as full-length proteins and in cleaved forms (pJH152-gp140CF, pJH222-gp140CF, and pJH154-gp140CF). We found that immunization via the i.p. route (Fig. 3A and C) was optimal for the induction of T-cell responses in comparison with i.d. immunization (Fig. 3B and D). In the spleen, optimal IFN-γ SFC responses were seen in mice immunized i.p. with pJH154-gp120 or pJH154-gp140CF constructs, both engineered to secrete insert protein (Fig. 3A and B). An rAd-CON6 prime/rVV-CON6 boost was used for comparison and showed an approximately fourfold-greater IFN-γ ELISPOT response than did recombinant M. smegmatis pJH154-gp120 or pJH154-gp140 (Fig. 3A and B). In addition to this positive control, one group of mice was immunized with only the CON6 Env protein at the time all other groups were boosted with the CON6 Env protein. We found that immunization once with recombinant gp140CF Env induced no significant IFN-γ SFC (7 ± 3 SFC/10⁶ cells), similar to what occurred with mice immunized with recombinant M. smegmatis transformed with empty plasmid and boosted with CON6 Env proteins. Thus, protein immunization did not augment T-cell responses to the recombinant M. smegmatis prime and boost.

FIG. 3.

Antigen-specific T-cell responses induced by different modes of insert expression by recombinant M. smegmatis. HIV-1 envelope-specific (A and B) and vector-specific (C and D) T-cell responses were assessed by IFN-γ ELISPOT assays. Mice were immunized i.p. (A and C) or i.d. (B and D) with various designs of recombinant M. smegmatis or untransformed recombinant M. smegmatis, as indicated. Mean (± standard error of the mean) numbers of SFC per 10⁶ cells derived from five mice per group are shown on the y axis. Mouse groups immunized with the indicated doses and designs of recombinant M. smegmatis or control constructs are shown on the x axis. The mouse group indicated as CON6 was immunized with CON6 protein only. The mouse group indicated as rAd-rVV was primed with rAd-CON6 and boosted with rVV-CON6 as a positive control.

Responses to the recombinant M. smegmatis vector itself in the spleen were also monitored and showed that i.p. immunization was superior to the i.d. route for induction of vector responses (Fig. 3C and D). While boosting once with M. smegmatis was effective in boosting an M. smegmatis prime, we found that boosting with recombinant M. smegmatis a second time had no effect on either anti-Env antibody levels or IFN-γ T-cell responses (not shown).

Recombinant M. smegmatis does not induce anti-HIV-1 Env antibody.

Serum samples were collected 2 weeks after each recombinant M. smegmatis immunization and assayed by ELISA for antibody titers against recombinant CON6 gp140CF. We found that neither a prime nor a boost of recombinant M. smegmatis alone could induce any detectable anti-HIV-1 Env antibody (not shown).

To determine whether recombinant M. smegmatis gp120 or recombinant M. smegmatis gp140CF could prime for an Env protein boost of antibody responses to the HIV-1 envelope, mice were boosted once with recombinant CON6 gp140CF protein 4 weeks after the second immunization with recombinant M. smegmatis. We found that mice primed with recombinant M. smegmatis and boosted with CON6 gp140CF protein had significantly higher anti-Env antibody titers than mice immunized with CON6 gp140CF protein alone or mice immunized with empty recombinant M. smegmatis plus Env protein (Table 2). Mice immunized with recombinant M. smegmatis expressing CON6 gp120 primed better for an Env protein boost than those immunized with recombinant M. smegmatis expressing CON6 gp140CF (P < 0.01). Also, mice immunized with recombinant M. smegmatis expressing secreted CON6 envelope immunogens primed optimally for Env protein boosting responses compared to those immunized with other modes of Env expression in recombinant M. smegmatis (P < 0.05) (Table 2).

TABLE 2.

Antibody responses to CON6 envelope protein after recombinant M. smegmatis CON6 prime and recombinant HIV-1 CON6-gp140 protein boost^a

Prime	Expression	Boostb	GMTc
M. smegmatis	Control	CON6-gp140	667
Empty pJH222	Control	CON6-gp140	560
pJH152-gp120	Surface, full length	CON6-gp140	368
pJH153-gp120	Intracellular, full length	CON6-gp140	5,164
pJH154-gp120	Secreted, full length	CON6-gp140	7,211
pJH152-gp140CF	Surface, full length	CON6-gp140	700
pJH222-gp140CF	Surface, cleaved form	CON6-gp140	1,622
pJH154-gp140CF	Secreted, cleaved form	CON6-gp140	4,498
		CON6-gp140	759d

^aDose of recombinant M. smegmatis, 10⁸ or 10⁹ CFU.

^bCON6 gp140CF protein with RiBi adjuvant.

^cGMT, anti-HIV-1 gp140 geometric mean titer (M. smegmatis-M. smegmatis-CON6-gp140CF) (n = 5).

^d1× CON6.

Next, HIV-1-specific IgG subclass responses to recombinant M. smegmatis were determined. Immunization of mice with both rAd-rVV and CON6 protein alone induced predominantly IgG1 antibodies, characteristic of a Th2-type response: the IgG1/IgG2a ratios were 4.44 and 3.59, respectively (Fig. 4). Untransformed (empty) M. smegmatis plus CON6 Env oligomer boosting also resulted in antibody responses that were biased toward IgG1, with an IgG1/IgG2a ratio of 2.20 (Fig. 4). Interestingly, antibody responses primed by immunizations with recombinant M. smegmatis pJH154-gp120 or pJH154-gp140CF showed a dominance of IgG2a over IgG1 antibodies (IgG1/IgG2a ratios of 0.82 and 0.49, respectively) (Fig. 4). Thus, priming with recombinant M. smegmatis expressing HIV-1 Env or Env protein alone changed the HIV-1 isotype profile from IgG1 predominant to IgG2a predominant.

FIG. 4.

Isotype of anti-HIV envelope antibodies induced by pJH154-gp120 and -gp140CF. IgG subclasses, including IgG1 (empty bars), IgG2a (gray bars), IgG2b (hatched bars), and IgG3 (solid bars), of HIV-1 envelope-specific serum antibodies induced by immunization with various immunization regimens (shown on the x axis) were determined by ELISA. The geometric mean log endpoint antibody titers (± standard errors of the means; n = 5) are plotted on the y axis.

Mechanism of limitations of recombinant M. smegmatis for induction of anti-HIV-1 antibody.

To determine the factors associated with the lack of induction of anti-Env antibodies by recombinant M. smegmatis, we first determined whether live recombinant M. smegmatis was immunosuppressive for Env insert antibody responses. Live recombinant M. smegmatis was mixed with 50 μg of CON6 Env oligomer protein and used to immunize mice. This group of mice was compared to groups that were immunized either with 50 μg CON6 Env oligomer protein alone or with heat-killed recombinant M. smegmatis plus 50 μg CON6 Env. We found that recombinant M. smegmatis did not suppress anti-Env antibody responses, as the three immunogens induced similar levels of antibody responses (Fig. 5). Immunization with recombinant M. smegmatis formulated in incomplete Freund's adjuvant (IFA) did not induce anti-HIV antibody responses, while the formulation of 50 μg CON6 Env in IFA did induce anti-Env antibody. Finally, killed recombinant M. smegmatis either alone or with CON6 Env oligomer immunization did not augment HIV-1 Env antibody induction. Thus, recombinant M. smegmatis was not suppressive of antibody responses; rather, HIV-1 Env was not expressed at sufficiently high levels in recombinant M. smegmatis for an induction of anti-Env antibody.

FIG. 5.

Induction of antibody responses with recombinant M. smegmatis is limited by the amount of HIV-1 Env expressed in recombinant M. smegmatis pJH154-gp120. To evaluate the immunogenicity of recombinant M. smegmatis expressing CON6 env in mice, mice were immunized once with either killed or live recombinant M. smegmatis expressing CON6 gp120 with or without IFA adjuvant. The geometric mean log endpoint antibody titers (± standard errors of the means; n = 5) are shown on the y axis. Abbreviations: rMsmeg, recombinant M. smegmatis expressing HIV-1 CON6 Env; CON6, recombinant CON6 gp140CF protein.

DISCUSSION

In this study, we have shown that recombinant M. smegmatis can express full-length hypoglycosylated HIV-1 Env and that, when engineered for insert secretion, it can induce HIV-1 IFN-γ T-cell responses in the spleen, FRT, and lungs. While unable to directly induce anti-HIV-1 Env B-cell responses, recombinant M. smegmatis was able to prime for an Env protein boost.

M. smegmatis is a rapidly growing nonpathogenic mycobacterial species that has properties that may make it a practical vector, particularly for priming for heterologous recombinant vectors or for protein boosting. M. smegmatis has been used as a model system for M. tuberculosis and for high-level expression of foreign genes (23). Unlike other mycobacterial species, such as BCG, that survive in host cells by inhibiting phagosome maturation, M. smegmatis is rapidly destroyed by phagolysosomal proteases in the phagosomes of infected cells (16, 19), facilitating rapid uptake of expressed antigens in M. smegmatis and cross-presentation of antigen. Recombinant M. smegmatis has been tested experimentally as a vaccine candidate for M. tuberculosis (21), as an alternative gene expression system to BCG or M. tuberculosis (3), and as an anticancer immunotherapy with cytokine-expressing M. smegmatis (29). Recently, we demonstrated that recombinant M. smegmatis expressing HIV-1 HXBc2 env can generate effector, effector memory, and central memory HIV-specific CD8 T cells (2). We also found that preexisting immunity to BCG may have only a marginal effect on the immunogenicity of recombinant M. smegmatis (2).

Others have fused the V3 epitope of HIV-1 Env with mycobacterial antigen 85B and expressed the V3-85B fusion protein in BCG as an HIV immunogen (11, 15). The anti-HIV-1 envelope V3 neutralizing antibody induced by immunization of rhesus macaques with BCG expressing HIV-1 envelope V3 antigen was able to protect against homologous T-cell-line-adapted simian-human immunodeficiency virus type 1 (SHIV-1) but not against heterologous SHIV challenge (24). Moreover, we and others have recently shown the limitation of anti-V3 loop immunogens for practical use as a HIV-1 vaccine immunogen in vivo (8). We found that insufficient levels of HIV-1 CON6 envelope protein expression in M. smegmatis were likely the cause of the lack of induction by a recombinant M. smegmatis prime and boost of anti-Env antibody responses. Thus, our current efforts are targeted at enhancing the expression of HIV-1 Env by incorporation of novel promoters in recombinant M. smegmatis and by insertion of new HIV-1 env genes in recombinant M. smegmatis and screening for optimal Env expression.

To address the issue of genetic diversity in HIV-1 vaccine design, we have begun to explore the immunogenicity of artificial consensus HIV-1 genes (6, 28; Weaver et al., AIDS Vaccine 2003 Meet.). We have previously described the 1999 CON6 envelope and shown that it induced a greater breadth of cross-clade T-cell responses than did a polyvalent subtype A, B, and C envelope immunogen (5, 6). In addition, CON6 protein was equally well recognized by patient sera of multiple HIV-1 subtypes but induced a limited spectrum of anti-HIV neutralizing antibodies to subtype B primary isolates (5, 6).

We have now produced a second-generation HIV-1 consensus Env protein, CON-S, that has shortened variable loops and induces a greater breadth of neutralizing antibodies than CON6 Env to subtype A, B, and C primary isolates (18). Our present study was undertaken to determine whether full-length consensus CON6 Env could be expressed in recombinant M. smegmatis and, if so, whether the expressed Env was immunogenic. While anti-HIV-1 antibodies were not induced by immunization with recombinant M. smegmatis alone, it is important that recombinant M. smegmatis was able to prime for an Env oligomer boost. The experiment performed here sought to answer the question of whether fully glycosylated Env protein can boost antibody responses to conserved Env regions in hypoglycosylated Env expressed in recombinant M. smegmatis. HIV-1 neutralization assays are technically difficult to perform with mouse sera due to serum volumes and natural inhibitors of HIV-1 in mouse sera. It will be of interest to test the breadth of neutralizing antibodies induced by an M. smegmatis prime protein boost in similar experiments with guinea pigs.

To address the issue of the optimal mode of insert expression in recombinant M. smegmatis, we have tested the HIV-1 CON6 env gene that was expressed either as a surface, intracellular, or secreted antigen and determined the comparative expression levels and immunogenicities of these constructs. Differential localization of HIV envelopes within mycobacteria could affect both the timing of the immune response and the pathway of presenting antigen to the immune system. Since mycobacteria are intraphagosomal, expressed antigens in mycobacteria can be delivered into phagolysosomes and preferentially activate CD4⁺ T cells via the major histocompatibility complex class II pathway (9). Proteins produced within the mycobacterial cytoplasm could remain encapsulated inside the bacterium and their presentation to the host immune system delayed as long as the bacterium persists within the macrophage (14). However, membrane-anchored and secreted antigens from M. smegmatis in the phagosome could result in earlier presentation to the immune system. The most commonly studied secreted sequence is the mycobacterial α-antigen (also known as MPT59 or antigen 85B), which we used to fuse the HIV-1 envelope. The other common sequence is a 19-kDa lipoprotein signal sequence that has been used to deliver the outer surface protein A (OspA) antigen of Borrelia burgdorferi to a membrane-associated lipoprotein (25). Others have shown that protein export via α-antigen is inefficient and that the majority of proteins fused with α-antigen are associated with cell membranes (10, 17). In our study, we demonstrated that anti-HIV-1 T-cell responses in splenocytes were optimal in mice immunized with recombinant M. smegmatis expressing HIV-1 CON6 Env engineered as secreted antigens (Fig. 3). In addition, antibody responses against HIV-1 envelope induced by priming with recombinant M. smegmatis and boosting with CON6 Env protein were also optimal when HIV gp120 and gp140CF were expressed by recombinant M. smegmatis as secreted antigens (Table 2).

Taken together, our data are proof of the concept that recombinant M. smegmatis vectors can induce mucosal cellular responses and prime for protein boosts of serum antibody. If the expression levels of recombinant M. smegmatis vector inserts are optimized, increased levels of immunogenicity should be attainable, and recombinant M. smegmatis may be a candidate to serve as a prime for a heterologous vector boost in an HIV-1 candidate vaccine formulation.