Boosting BCG-primed responses with a subunit Apa vaccine during the waning phase improves immunity and imparts protection against Mycobacterium tuberculosis

Heterologous prime–boosting has emerged as a powerful vaccination approach against tuberculosis. However, optimal timing to boost BCG-immunity using subunit vaccines remains unclear in clinical trials. Here, we followed the adhesin Apa-specific T-cell responses in BCG-primed mice and investigated its BCG-booster potential. The Apa-specific T-cell response peaked 32–52 weeks after parenteral or mucosal BCG-priming but waned significantly by 78 weeks. A subunit-Apa-boost during the contraction-phase of BCG-response had a greater effect on the magnitude and functional quality of specific cellular and humoral responses compared to a boost at the peak of BCG-response. The cellular response increased following mucosal BCG-prime–Apa-subunit-boost strategy compared to Apa-subunit-prime–BCG-boost approach. However, parenteral BCG-prime–Apa-subunit-boost by a homologous route was the most effective strategy in-terms of enhancing specific T-cell responses during waning in the lung and spleen. Two Apa-boosters markedly improved waning BCG-immunity and significantly reduced Mycobacterium tuberculosis burdens post-challenge. Our results highlight the challenges of optimization of prime–boost regimens in mice where BCG drives persistent immune-activation and suggest that boosting with a heterologous vaccine may be ideal once the specific persisting effector responses are contracted. Our results have important implications for design of prime–boost regimens against tuberculosis in humans.


specific T cells after mucosal or parenteral BCG vaccination.
Mice were vaccinated once with BCG by the i.n. or s.c. route as described in Figure 1. At different time points after vaccination, as indicated, mice were euthanized (n = 4 mice/time point/group) and their lungs and spleen were harvested. Cytokine expression by T cells was measured by an intracellular cytokine staining and a polychromatic flow cytometry following the stimulation of spleen or lung cells (pools) with a medium alone (no antigen), nApa or WCL. (a) Kinetic changes in the frequencies of nApa-specific CD4 + T cells expressing IFN-γ, IL-2 or TNF-α. Data are percentages (%) of nApaspecific cytokine-producing cells among CD4 + T cells plotted for an individual cytokine after subtracting no antigen stimulation control values. Data at week 12, 32, 52 and 78 are means ± s.e.m. of 3-4 independent mouse experiments, while data (means) at week 3, 6 and 104 are from one experiment evaluated in duplicate. *P < 0.05, **P < 0.01 and ***P < 0.001 comparing responses of i.n. and s.c. BCG using one-way ANOVA followed by Bonferroni's post-test. (b) Representative flow-cytometric 3 plots from one experiment showing nApa or WCL-specific IFN-γ, IL-2 and/or TNF-α expression by CD8 + T cells in the lung at the peak of response (week 32) after s.c. BCG vaccination. s.c., subcutaneous; i.n., intranasal; Ag, antigen; WCL, whole cell lysate of Mtb. Figure S2. Timing, route and form of a subunit-Apa-boost in the BCG-primed mice impact the cellular booster response Mice were primed with BCG vaccine by the i.n. or s.c. route and boosted with a subunit nApa or rApa (10µg) vaccine using a DDA/MPL adjuvant by a homologous or heterologous route at the peak (8 months) or contraction-phase (16 months) of BCG-elicited response. In one experiment, i.n. BCGprimed mice were boosted with a DDA/MPL adjuvant alone by a homologous i.n. route during the contraction-phase. Three weeks after a subunit vaccine-or adjuvant-boost, frequencies of antigenspecific cytokine-producing cells in the lung, CLN or spleen (pooled) were determined using a cultured ELISPOT assay. (a-b) Influence of the timing and route of a subunit-rApa-boost in the BCG-primed 4 mice on the cellular booster response. The nApa-(upper panels) and WCL-specific (lower panel) IFN-γ, IL-17 and IL-4 SFU per million organ cells following a single i.n. subunit-rApa-boost in (a) i.n.
(homologous route) or (b) s.c. (heterologous route) BCG-primed mice are plotted. Number in the parenthesis indicates a fold increase in total IFN-γ, IL-17 and IL-4 SFU in rApa-boosted mice over corresponding non-boosted BCG-primed controls. (c-d) The form of Apa vaccine (native versus recombinant) influences IL-2 booster response. The IL-2 SFU per million organ cells following nApa or rApa subunit-boost at the contraction-phase of BCG-response using (c) a homologous (i.n.-i.n.) or (d) heterologous (s.c.-i.n.) route prime-boost regimen. (e) Boosting BCG-primed responses with an adjuvant alone does not afford a stronger antigen-specific boost. Error bars in (a-e) represent s.d. of cytokine response using pooled organ cells in triplicate cultures (n= 4mice/group/time point). Significant using ANOVA followed by Bonferroni's post-test. *P < 0.05, **P < 0.01 and ***P < 0.001. s.c., subcutaneous; i.n., intranasal; vac, vaccine; WCL, whole cell lysate of Mtb; SFU, spot forming units. (a) The magnitude and kinetics of cytokine-producing cells in rApa vaccinated mice. Six-eight-week old mice were subunit immunized three times with rApa (10µg/dose) in a DDA-MPL adjuvant at 2-week intervals by the i.n. route, and IFN-γ, IL-2, IL-4 and IL-17 responses were investigated in the lung and spleen as indicated using a cultured ELISPOT assay. Data are mean ± s.d. of triplicate cultures using 5 pooled cells (n = 4 mice/time point). The IL-2 response was not investigated at the 2-week time point.
Significant waning of cytokine response at the week 32 compared to the 2-week time point using a oneway ANOVA and Bonferroni's test. (b-c) Boosting rApa vaccine-primed responses with BCG. Mice were immunized three times with rApa in a DDA-MPL adjuvant as described in (a). Eight months after the last subunit dose, a group of subunit-vaccinated or age-matched naïve mice received a single BCG vaccine-boost (10 6 CFU) by a homologous i.n. route. Three weeks following a boost, mice (n = 4 mice/group) were euthanized and frequencies of (c) nApa-and (d) WCL-specific IFN-γ, IL-17 and IL-4producing cells in the lung and spleen (pooled) were determined using a cultured ELISPOT assay. Data are mean + s.d. of triplicate cultures. Number in the parenthesis indicates a fold change in the total SFU in BCG-boosted group over levels in the respective control group. Significant using ANOVA and Bonferroni's test compared to the respective control group. *P < 0.05, **P < 0.01 and ***P < 0.001. i.n., intranasal; vac, vaccine; WCL, whole cell lysate of Mtb; SFU, spot forming units. Figure S4. Synthetic peptide screening in the BCG-primed-Apa-boosted mice reveals the breadth of specific booster response and confirms a differential influence of Apa-form on IL-2 response. 6 Mice were vaccinated once with BCG using the i.n. or s.c. route. Sixteen months later, mice received two booster doses of nApa or rApa (1µg) in a DDA/MPL adjuvant by a homologous route as depicted in triplicate culture. ANOVA followed by Bonferroni's test. *P < 0.05, **P < 0.01 and ***P < 0.001. s.c., subcutaneous; mo, months; vac, vaccine; WCL, whole cell lysate of Mtb; SFU, spot forming units. 7 Figure S5. T-cell clones generated by a subunit nApa vaccination express an IL-2-producing phenotype.
BALB/c mice were vaccinated using a single dose of nApa (40 µg) in the Freund's incomplete adjuvant and the T-cell hybridomas were generated using a draining lymph node cells. Individual T-cell clones responding to nApa were purified by a serial dilution and tested for a reactivity to the native or recombinant form of Apa using an IL-2 capture ELISA, after a co-culture with the syngeneic antigen presenting cells (APCs) pulsed with the antigen. Of the total 17 Apa-reactive hybridoma clones developed, 10 clones recognized only nApa and not rApa (a), while 7 clones recognized both nApa and rApa (b) and produced a positive IL-2 response. The IL-2 response of one clone that recognized both nApa and rApa but stopped responding consistently is not shown. The antigen dose response curve was 8 carried out using the antigen-pulsed bone marrow derived dendritic cells as APCs. For clones 5F9 and 5D8 data is available using only one antigen dose (2 µg) per well and BALB/c derived B-cell line A20 as APCs. The horizontal line indicates a median response. None of the 10 nApa-reactive clones recognized any of the synthetic, overlapping, nonmodified peptides of Apa. Of the 10 clones recognizing only nApa, one clone, 4C3, was tested further for the epitope specificity. The T cell clone 4C3 reacted exclusively with the N-terminal O-mannosylated glycopeptide fraction (residues p40-145) of trypsin-digested nApa following fractionation by a reversed phase-HPLC column chromatography and produced a positive IL-2 response using 1µg of nApa or glycopeptide fraction (inset). All clones were negative for the IFN-γ production (O.D. 450 < 0.2) testing the same supernatants as in the IL-2 assay, 24 h after in vitro stimulation. Regardless of the timing of boost, decreased specific IgG1 antibody levels were found in the sera of nApa-boosted mice relative to levels in the corresponding non-boosted controls. Interestingly, nApaboost at the contraction-phase (16 months) increased the specific serum IgG2a and IgG2b levels, indicating the influence of the timing of a subunit nApa-boost after BCG-priming on the magnitude and quality of antibody response (b) Serum IgG response of the s.c.-i.n. regimen (in Figure 3) using ELISA.
The WCL-specific IgG1, IgG2a and IgG2b antibody levels are plotted for the nApa-boosted groups (i.e. boosted 8 months or 16 months after BCG-priming) in comparison with the respective BCG-primed non-boosted controls. No increase was found in the serum antibody levels following nApa-boost at the peak of BCG-response (8 months), where higher anti-WCL IgG subclass levels preexisted. On the contrary, nApa-boost during the contraction-phase (16 months) effectively boosted WCL-specific IgG1 and IgG2a but not IgG2b levels, which collectively suggest that the route of prime-boost regimen differentially influences the quality of antibody response. s.c., subcutaneous; i.n., intranasal; vac, vaccine; WCL, whole cell lysate of Mtb.  Figure S8. Poor IL-2 response in the BCG-primed-rApa vaccine-boosted mice is not due to the inherent inability of a subunit rApa vaccine to induce specific IL-2 + CD4 + T cells.
Six-eight-week old mice were vaccinated with 3 doses of nApa or rApa (1µg/dose) in DDA/MPL adjuvant at 4-week intervals by a s.c. route on the hind legs. Magnitude (histograms) and quality (pie charts) of nApa-and WCL-specific total cytokine-producing (IFN-γ, IL-2 and/or TNF-α) cells among CD4 + T cells of spleen, ILN and lung were determined 4 weeks after the last subunit dose (prechallenge) and 6 weeks after Mtb challenge (post-challenge) using a flow cytometry. Significant using ANOVA followed by Bonferroni's test. Arcs surrounding pie charts denote IL-2-producing cells in any combinations of nApa-or WCL-specific CD4 + T cells (n = 4 individually analyzed mice/group). WCL, whole cell lysate of Mtb; vac, vaccine.

Supplementary methods
Vaccinations and experimental infections. Mtb H37Rv was grown to mid-log phase at 37 o C in liquid glycerol-alanine-salts (GAS) medium and culture filtrate was harvested. The nApa was purified from the culture filtrate by combinations of column chromatography techniques using Con-A-Sepharose and Phenyl-Sepharose columns, while rApa was expressed in E. coli BL21 (DE3) and purified from lysates by Nickel chromatography with endotoxin removal step followed by DEAE-Sepharose chromatography. 14,17 Purity was confirmed by SDS-PAGE, N-terminal sequencing and by Western blot Approximately 500,000 lymphocytes were acquired on the LSRII system (BD Immunocytometry Systems) and analyzed using FlowJo software (Treestar, Inc., San Carlos, CA). Lymphocytes were identified based on their scatter patterns, and CD3 + , CD8 − , CD4 + cells were considered to be CD4 T cells, while CD3 + , CD8 + , CD4 − cells were considered CD8 T cells. These CD4 + and CD8 + T cells were then gated for cells positive for the respective cytokines. Boolean combination gating was performed to calculate the frequencies of expression profiles corresponding to the seven different combinations of 15 cytokines by using the FlowJo software. After subtracting the background values, frequencies of individual or total cytokine secreting CD4 + or CD8 + T cells were plotted for each antigen.
ELISPOT assay. IFN-γ, IL-2, IL-4 (BD-Biosciences) or IL-17A (eBioscience) ELISPOT assay was performed using a commercially available mouse reagent set as described previously 8,17 using Mtb antigens for in vitro stimulations of organ cells at 37°C in the presence of 5% CO2 for 40 h. In brief, 96well ELISPOT plates were coated with 100 µl of 5 µg/ml respective capture antibody in PBS (pH 7.2) and incubated overnight at 4°C. Free binding sites were blocked with RPMI-1640 medium containing 10% FCS (Atlas Biologicals) for 2 h at room temperature. Mouse lung, spleen or LN cells were suspended in different dilutions starting at 1 or 2 × 10 5 cells per well in supplemented RPMI-1640 medium containing 10% FCS. Cells were stimulated with 10 µg/ml of nApa, rApa, WCL, rESAT-6, rCFP-10 or synthetic Apa peptides for 40 h. Stimulation with Con-A (1µg/ml; Sigma-Aldrich) was used as a positive control for cell viability and reactivity. After incubation at 37°C, the wells were washed with PBS-Tween-20 and the site of cytokine secretion was detected with a biotin-labeled detection antibody and horseradish peroxidase-conjugated streptavidin. The enzyme reaction was developed using 3-amino-9-ethylcarbazole (AEC) substrate reagent set (BD-Biosciences, San Diego, CA). The number of SFU per well were counted using an ELISPOT reader (Cellular Technology Limited, Cleveland, OH).
The number of spots specific for each antigen preparation was calculated by subtracting the number of spots that formed in the absence of added antigen from the number that formed in its presence. The cytokine response ≥ 50 SFU/10 6 organ cells after subtracting no antigen control values was considered a positive response.

Generation of nApa-specific T-cell hybridomas and cytokine ELISA. T-cell hybridomas specific for
Mtb nApa were generated as previously described 17 . Briefly, four BALB/c mice were each vaccinated with 40 µg of nApa in FIA by injecting 25 µl into each hind footpad and the remainder at the base of the tail. Five days after the vaccination, the draining lymph nodes (popliteal, inguinal and periaortic) were harvested to obtain LN cells. The primed LN cells were re-stimulated in vitro with syngeneic bone marrow derived dendritic cells (BMDCs) that had been pulsed with 10 µg/ml nApa per well.
Approximately 1×10 6 primed LN cells were added per well in two 24-well tissue culture plates in a total volume of 1 ml of complete RPMI 1640 medium (RPMI 1640 supplemented with 10% FCS, 5×10 −5 M 2-ME (Sigma) plus a nutrient cocktail). After two days of culture, the cells were harvested and pooled from all 48 wells, washed, fused with the T cell fusion partner BWα − β − , and plated out into ten 96-well plates. Clones that grew in individual wells were screened using either BMDCs, as above, or a BALB/c mouse-derived B cell lymphoma line A20 pulsed with 0.5µg/well nApa. The selection of responding T cell hybridomas was performed by assaying 24 h culture supernatants using paired rat mAbs specific for mouse IL-2 or IFN-γ (Pharmingen/BD-Biosciences) in a capture ELISA. For antigen presentation studies, micro-culture wells were prepared containing 250 µl of culture medium, 5×10 4 each of T cell hybridoma and antigen presenting cells (APC), and a known amount of nApa or rApa (intact or trypsin digested), in flat-bottomed 96-well micro-titer wells. Dose response curves with native and recombinant antigens were performed to determine which T cell hybridomas respond to nApa alone, or to both native and recombinant form of Apa. T cell hybridomas were further tested with a panel of synthetic, overlapping, nonglycosylated Apa peptides or trypsin digested and RP-HPLC separated nApa fractions 17 to determine the peptide epitopes that were recognized. Antigen-specific IL-2 responses of the clones were comparable regardless of APC type used for antigen presentation (i.e., bone-marrow-derived macrophages (BMDMs) versus BMDCs; data not shown).