MTBVAC-Based TB-HIV Vaccine Is Safe, Elicits HIV-T Cell Responses, and Protects against Mycobacterium tuberculosis in Mice

Abstract

The tuberculosis (TB) vaccine MTBVAC is the only live-attenuated Mycobacterium tuberculosis (Mtb)-based vaccine in clinical development, and it confers superior protection in different animal models compared to the current vaccine, BCG (Mycobacterium bovis bacillus Calmette-Guérin). With the aim of using MTBVAC as a vector for a dual TB-HIV vaccine, we constructed the recombinant MTBVAC.HIVA^2auxo strain. First, we generated a lysine auxotroph of MTBVAC (MTBVACΔlys) by deleting the lysA gene. Then the auxotrophic MTBVACΔlys was transformed with the E. coli-mycobacterial vector p2auxo.HIVA, harboring the lysA-complementing gene and the HIV-1 clade A immunogen HIVA. This TB-HIV vaccine conferred similar efficacy to the parental strain MTBVAC against Mtb challenge in mice. MTBVAC.HIVA^2auxo was safer than BCG and MTBVAC in severe combined immunodeficiency (SCID) mice, and it was shown to be maintained up to 42 bacterial generations in vitro and up to 100 days after inoculation in vivo. The MTBVAC.HIVA^2auxo vaccine, boosted with modified vaccinia virus Ankara (MVA).HIVA, induced HIV-1 and Mtb-specific interferon-γ-producing T cell responses and polyfunctional HIV-1-specific CD8+ T cells producing interferon-γ (IFN-γ), tumor necrosis factor alpha (TNF-α), and CD107a in BALB/c mice. Here we describe new tools to develop combined vaccines against TB and HIV with the potential of expansion for other infectious diseases.

Keywords: BALB/c; BCG; HIV; HIVA; MTBVAC; SCID; mycobacterium; p2auxo; tuberculosis; vaccine.

Figure 1

Construction and Characterization of MTBVACΔlys Strain (A) lysA gene (gray arrow) from MTBVAC was inactivated using homologous recombination techniques by introducing a kanamycin resistance cassette (gray rectangle), flanked by two resolvase sites (white arrowheads) in order to allow the release of the resistance cassette. The inactivated phoP and fadD26 genes in the MTBVAC parental strain are also illustrated. (B) Genotypic characterization of the lysA gene inactivation by the kanamycin cassette (km) insertion, primers used, and expected sizes of the PCR products are indicated. MTBVAC sample was used in lanes 3, 5, and 7 and MTBVACΔlys samples in lanes 4, 6, and 8. Lane 1, molecular weight marker; lane 2, negative control; lanes 3 and 4, PCR product of the lysA gene using Lys-fw and Lys-rv primers; lanes 5 and 6, PCR of the 5′ insertion point of km expression cassette using Lys-fw and km-OUT-rv primers; and lanes 7 and 8, PCR of the 3′ insertion point of km expression cassette using km-OUT-fw and Lys-rv primers. Genes are represented as gray arrows; gray rectangles illustrate antibiotic resistance markers and white arrowheads depict resolvase recognition sequences or res sites. (C) Phenotypic lysine auxotrophic verification by plating MTBVACΔlys strain in 7H10-ADC with and without lysine supplementation.

Figure 2

Construction of MTBVAC.HIVA^2auxo (A) The HIVA gene was fused to the region encoding the 19-kDa lipoprotein signal sequence of the episomal p2auxo.Ø E. coli-mycobacterial shuttle plasmid to obtain p2auxo.HIVA plasmid. The BALB/c mouse T cell and Mab-Pk epitopes used in this study are depicted. P α-Ag (M. tuberculosis α-antigen promoter), PHSP60 (heat shock protein 60 gene promoter), and glyA- and lysA-complementing genes are used as markers for selection and maintenance in E. coli M15ΔGly and MTBVACΔlys, respectively. (B) Phenotypic characterization of the lysine auxotrophy and plasmid complementation of MTBVACΔlys and MTBVAC.HIVA^2auxo (MTBVACΔlys plated on lysine-supplemented 7H10, left; MTBVACΔlys plated on non-lysine-supplemented 7H10, center; and MTBVAC.HIVA^2auxo plated on non-lysine-supplemented 7H10, right). (C) Western blot of MTBVACHIVA^2auxo lysates. Lanes 1 and 2, MTBVAC.Ø^2auxo clones 1 and 2; lanes 3 and 5, MTBVAC.HIVA^2auxo; lane 6, BCG wild-type lysate (negative control). HIVA immunogen was detected using the anti-Pk monoclonal antibodies (mAbs) followed by horseradish peroxidase-protein A and enhanced chemiluminescence detection.

Figure 3

Genetic Stability of p2auxo.HIVA Plasmid DNA (A) In vitro. Serial passages of the working vaccine stock (WVS) were performed weekly (+1 to +6), and HIVA PCRs were used to check stability of the plasmid DNA. Lane 1, WVS MTBVAC.HIVA^2auxo; lanes 2–4, passages +4, +5, and +6 WVS MTBVAC.HIVA^2auxo; lane 5, H₂0 (negative control); lane 6, positive control; lane 7, molecular weight marker. (B) In vivo. Spleens from SCID mice inoculated with 10⁶ CFU MTBVAC.HIVA^2auxo and used for safety experiments were harvested and plated on complete 7H10 supplemented with Lys and Km. The presence of p2auxo.HIVA plasmid in the colonies from these mice was analyzed by specific PCR using the pairs of primers to detect HIVA (19kDss-fw/HIVA-rv). Each number represents one colony and numbers with # symbol indicate colonies from the same animal. Minus and plus symbols indicate negative and positive controls of PCR, respectively. Plasmid maintained in vivo was calculated as the percent of positive colonies with respect to total colonies analyzed.

Figure 4

Induction of HIV-1- and Mtb-Specific T Cell Responses by the MTBVAC.HIVA^2auxo Prime MVA.HIVA Boost Regimen (A) Adult (7-week-old) mice were either left unimmunized or primed with 10⁶ CFU MTBVAC.HIVA^2auxo or MTBVAC.Ø^2auxo (intradermally) and boosted with 10⁶ plaque-forming units (PFUs) of MVA.HIVA (intramuscularly) 6 weeks post-MTBVAC inoculation. Mice were sacrificed 2 weeks later for T cell analysis. (B) Analysis of IFN-γ and CD107 vaccine elicited HIV-1-specific CD8+ T cell responses. The frequencies of cells producing cytokines are shown. Data are presented as means (SD; n = 7 for groups A, B, and D and n = 6 for group C). (C) The functionality of vaccine-induced CD8+ T cell responses was assessed in a multicolor intracellular cytokine staining assay. The group mean frequencies of single, double, and triple cytokine-producing P18-I10-specific cells are shown for the four vaccination groups. (D) Elicitation of specific T cell responses was assessed in an ex vivo IFN-γ enzyme-linked immunosorbent spot (ELISPOT) assay using the immunodominant P18-I10 CD8+ T cell epitope peptide. The median spot-forming units (SFUs) per 10⁶ splenocytes for each group of mice (n = 7 for groups A, B, and D and and n = 6 for group C) as well as individual animal responses are shown. (E) Purified protein derivative (PPD)-specific T cell responses elicited by MTBVAC.HIVA^2auxo. Immune responses to mycobacteria were assessed in an ex vivo IFN-γ ELISPOT assay using PPD as the antigen. The median SFUs per 10⁶ splenocytes for each group of mice (n = 7 for groups A, B, and D and and n = 6 for group C) as well as individual animal responses are shown. Statistical analysis was performed by ANOVA plus Bonferroni multiple comparisons test (*p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001).

Figure 5

MTBVAC.HIVA^2auxo Prime and MVA.HIVA Boost Safety in Adult Mice Adult mice were either left unimmunized or immunized with 10⁶ CFU of MTBVAC.HIVA^2auxo by intradermal route and subsequently given a booster dose of 10⁶ PFU of MVA.HIVA at week 12 by intramuscular route. The body mass was recorded over time, and the mean for each group of mice is shown (n = 5). Data from control mice are presented as mean ± 2 SD (dashed lines). The weight differences between vaccinated and naive mouse group were analyzed at the final time point by ANOVA.

Figure 6

Efficacy of MTBVAC.HIVA^2auxo Vaccine against M. tuberculosis C57BL/6 mice were vaccinated subcutaneously with 10⁶ CFU of the strains indicated, MTBVAC, MTBVAC.HIVA^2auxo, and MTBVACΔlys, or naive (unvaccinated as control). At 8 weeks post-vaccination, mice were challenged by intranasal route with 200 CFU H37Rv. Bacterial burden was assessed in lungs (A) and in spleen (B) 4 weeks post-challenge. Data are expressed as mean ± SEM and compared by 2-way ANOVA test, using Bonferroni multiple comparison post-test (**p < 0.01).

Figure 7

MTBVAC-HIVA^2auxo Safety in SCID Mice SCID mice were inoculated by intraperitoneal route with 10⁶ CFU BCG, MTBVAC, MTBVAC-HIVA^2auxo, MTBVACΔlys, or naive (unvaccinated as control). Analysis of survival was done applying the Mantel-Cox test (*p < 0.05 and **p < 0.01).