Design of an LV encoding collectin scaffolds harboring Mtb antigens and a DC-targeting segment

MBL and SPD collectins possess four distinct segments: (i) an N-terminal cysteine-rich crosslinking domain, (ii) a collagen-like domain, (iii) an α-helical neck domain, and (iv) a CRD (Fig. 1a). The primary structural MBL or SPD unit consists of self-assembled collagen-like triple helix (Fig. 1b). SPD triple-subunits can tetramerize to form cross-shaped dodecamers (Fig. 1c). SPD or MBL triple-subunits can also hexamerize to form a “tulip-like nano-bouquet” octodecamer (Fig. 1d). The resulting secreted polymers are soluble9 and can be used as antigen carriers13,14. We generated an LV-based vaccine vector encoding MBL or SPD carriers, potentially usable for an infectious disease controllable by CD4+ T cells, using the EsxA, EspC (ESX-1 secretion-associated protein C), EsxH, PE19, hypoxic response protein 1 (Hrp1), and resuscitation promoting factor D (RpfD) immunogens from Mtb16,17 (Table S1).

Fig. 1: Structure of MBL- and SPD-based antigen carriers.

Schematic structure of the MBL and SPD collectin polymers. a Structural domains of MBL and SPD. CRD = carbohydrate-recognition domain. b MBL and SPD self-assembled collagen-like triple helixes formed by interchain cysteine bonds. c SPD cross-shaped dodecamer. d SPD and MBL “tulip-bouquet” octodecamers, adapted from9. e–f Schematic representation of the primary structure of the designed M40 (e) and S40 (f) monomers carrying selected Mtb antigens: crosslinking region (S), collagen-like region (Coll), neck region (N).

We first engineered the murine MBL to carry the complete sequences of: (i) EsxH alone, (ii) EsxH and EsxA, (iii) EsxH, EsxA, and PE19, or (iv) EsxH, EsxA, PE19, and EspC within the collagen-like region, while replacing its CRD with the murine CD40L115-260 ectodomain13,14. The prospective MBL polymers will be referred to as “M40-H”, “M40-HA”, “M40-HAP”, and “M40-HAPE”, respectively (Fig. 1e, Table S2, Table S3).

In parallel, we engineered SPD to carry: (i) EsxH alone, (ii) EsxH, EsxA, PE19, and EspC, or (iii) EsxH, EsxA, PE19, EspC, Hrp1, and RpfD42-154 within the collagen-like region and substituted its CRD with CD40L115-26013,14 (Fig. 1f, Table S2, Table S4). The expected SPD polymers are referred to as “S40-H”, “S40-HAPE”, and “S40-HAPEHR”, respectively. To enhance the DC-targeting potential of the resulting fusion protein, we also designed an S40-HAPEHR that carries the murine CCL2028-97 segment within the second collagen-like domain (“S40-HAPEHR-20”). CCL20 is the CCR6 ligand, largely involved in the migration and recruitment of immature DCs and lymphocytes18.

We also generated an LV encoding S40-H StrepII-tagged at its C-terminus (“LV::S40-H-StrepII”) to obtain insights into the polymerization state of antigen-bearing “S40” scaffolds. Supernatants and total lysates from LV::S40-H-StrepII-transduced or non-transduced HEK293T cells were analyzed by SDS-PAGE (Fig. 2a). Under reducing conditions, S40-H monomers were highly detectable in both supernatants and lysates from transduced cells. Under non-reducing conditions, relatively intense bands of S40-H dimers and tetramers and less intense bands corresponding to S40-H trimers were readily detectable in both supernatants and lysates from transduced cells. The lower amount of trimer is obviously related to its further successive multimerization, as the supernatants, and, to a lesser extent, the cell lysates from the transduced cells contained intense bands of high molecular weight. These results show that the engineered recombinant S40 scaffold maintains its multimerization capacity. In order to have a more precise quantification of the high molecular weight species, the multimers of a StrepII-tagged S40-H from a concentrated supernatant of HEK293T transduced cells were fractionated according to their size by exclusion chromatography, then quantified by Western blot (Fig. 2b). While the S40-H-StrepII protein was detected in fractions corresponding to the molecular masses of dimers, trimers and tetramers, it was not possible to detect higher molecular weight multimers. This underlines that the bands of high weight multimer observed in the Western blot are varied but in low amount compared to those of low weight.

Fig. 2: Multimerization of S40-based carriers.

a Western-blot analysis of supernatants and cell lysates of HEK293T cells, either transduced at an MOI = 100 of LV::S40-H-StrepII or non-transduced, in reducing or non-reducing conditions. β-Actin staining was use of as loading control. b Chromatogram of column calibration using a set of purified recombinant proteins on a Superose 6 FPLC column. X axis correspond to MW equivalence observed during elution and Y axis to UV absorbance (280 nm). c Chromatogram of 50x-concentrated supernatant of HEK293T cells transduced with LV::S40-H-StrepII. d Quantification of S40-H-StrepII protein by measuring mean pixel intensity of the band of each well, corresponding to the molecular weight of monomeric S40-H-StrepII. e Western blot analysis of elution fractions in reducing conditions.

Induction of MHC-II-restricted antigen presentation by LV::M40 and LV::S40

DCs (H-2d or H-2b) were directly transduced with LV::M40-H, -HA, -HAP, or -HAPE using a composite β2m-CMV promoter (“BCUAG”) comprised of the human β2-microglobulin (β2 m) promoter19 and the human cytomegalovirus immediate early enhancer and promoter (CMV)20 (Fig. S1A, B). Control DCs were transduced with a conventional LV encoding EsxH under the same promoter, without insertion into an engineered scaffold. Three days post-transduction, the DCs were co-cultured with T-cell hybridomas specific for the immunodominant epitopes of each of the Mtb antigens. The DCs transduced with either LV were largely able to induce the presentation of EsxH via MHC-I (Fig. 3a). In contrast to the conventional LV::EsxH, the LV encoding M40-H, -HA, -HAP, or -HAPE induced MHC-II-restricted presentation of EsxH and EsxA when these antigens were included. No MHC-II presentation of PE19 or EspC was detected in this context, which, in the case of EspC, can be explained by the relatively weak sensitivity of the T-cell hybridoma (Fig. S2). We evaluated whether M40 or S40 carriers secreted by transduced cells induce presentation through MHC-II by incubating DCs with successive dilutions of M40-H-, -HA-, -HAP-, or -HAPE-containing supernatants from HEK293T cells transduced at a multiplicity of infection (MOI) of 1 of the corresponding LVs (Fig. 3b). On day 1 after incubation, co-culture of the DCs with T-cell hybridomas showed the DCs to be unable to present EsxH via MHC-I, strongly suggesting that endocytosis/micropinocytosis or CD40-mediated cell entry of the M40 carrier does not allow their access to the MHC-I machinery, in contrast to observations made by others21. In net contrast to MHC-I, DCs incubated with M40-H-, -HA-, -HAP-, or -HAPE-containing supernatants were highly efficient at inducing presentation of the respective antigens via MHC-II, including EspC. The level of antigen presentation tended to decrease with a growing number of antigens carried by the M40 scaffold (Fig. 3a, b). In a mutually non-exclusive manner, this may result from: (i) slight structural instability of the carriers with the insertion of an increasing number of antigens or (ii) competition among the multiple T-cell epitopes for the available MHC presentation sites.

Fig. 3: Properties of backbones at recruiting APC and inducing antigenic presentation by both the MHC-I and-II pathways.

a, c BM-DCs from BALB/c (H-2d) or C57BL/6 (H-2b) mice were transduced (MOI = 20) with LV::M40-H, -HA, -HAP, or -HAPE (a) or LV::S40-H, -HAPEHR, or -HAPEHR-20 (c) under the transcriptional control of the BCUAG promoter. Control cells were transduced with LV::EsxH alone. b, d BM-DCs from BALB/c or C57BL/6 mice were incubated with successive dilutions of supernatants of HEK-293T cells transduced (MOI = 20) for 48 h with each of the indicated LVs. On day 3 after addition of the LVs or day 1 after incubation with the HEK293T cell supernatants, the presentation of MHC-I- or -II-restricted epitopes of the EsxH, EsxA, PE19, or EspC mycobacterial antigens by DCs was assessed by their co-culture with T-cell hybridomas specific for EsxH:20-28 (YB8 cell line, restricted by Kd), EsxH:74-88 (1G1 cell line, restricted by I-Ad), EsxA:1-20 (NB11 cell line, restricted by I-Ab), PE:19:1-18 (IF6 cell line, restricted by I-Ab), or EspC:45:54 (IF1 cell line, restricted by I-Ab). Results are presented as the concentration of IL-2 produced by the T-cell hybridomas 24 h after the beginning of the co-cultures. The amount of IL-2 found in the co-culture supernatants is proportional to the efficacy of antigenic presentation by DCs and TCR triggering. This assay does not measure the physiological response induced in the T cells but only provides an indicator of the stimulation of the T-cell hybridoma TCR. e Impact of HAPEHR and HAPEHR-20 proteins on migration of BM-DCs in a transwell system. Supernatant of HEK293T cells transduced (MOI = 100) with LV::S40-HAPEHR or LV::S40-HAPEHR-20 (n = 6). Statistical significance was evaluated using the Mann–Whitney test (* = p < 0.05, ** = p < 0.01, *** = p < 0.001, ns non-significant).

Direct transduction of DCs with LV::S40-HAPE, -HAPEHR or -HAPEHR-20 also induced efficient MHC-I- and -II-restricted presentation of the selected Mtb antigens (Fig. 3c). Incubation of DCs with successive dilutions of supernatants from HEK293T cells transduced at an MOI of 1 with LV::S40-HAPE, -HAPEHR, or -HAPEHR-20 induced MHC-II-restricted presentation of the Mtb antigens (Fig. 3d). In the absence of identified T-cell epitopes and T-cell hybridomas specific to Hrp1 and RpfD, we studied the immunogenicity of these antigens in the context of the developed vectors in vivo, as detailed below. Insertion of 4–6 Mtb antigens to the S40 carrier gave the highest levels of antigenic presentation via MHC-II by DCs directly transduced with LV::S40 (Fig. 3c). In addition, in contrast to the results obtained with M40 (Fig. 3b), the supernatants from HEK293T cells transduced at the same MOI with various LV::S40 constructs with an increasing number of Mtb antigens did not show a progressive decrease in the efficiency of antigen presentation via MHC-II (Fig. 3d). Assays performed with DCs incubated with synthetic peptides for the homologous T-cell epitopes showed the same sensitivity as the T-cell hybridoma-based presentation assay (Fig. S2). At last, impact of the insertion of the CCL20 protein sequence into the HAPEHR backbone was evaluated by use of supernatants from HEK293T cells transduced with LV::S40-HAPEHR or LV::S40-HAPEHR-20 in a transwell assay. This experiment demonstrated an improved ability of HAPEHR-20, compared to the native HAPEHR, to mobilize bone-marrow derived DCs in vitro (Fig. 3e).

These results show that, in net opposition to conventional LVs, this new generation of LVs encoding secreted scaffolds that can incorporate numerous antigens and immune mediators possess a strong capacity to induce MHC-II-restricted antigen presentation and thus provide a valuable platform for both CD4+ and CD8+ T-cell induction.

The potential of M40 and S40 to induce DC maturation

We evaluated the potential of M40 and S40 carriers to induce DC maturation by incubating BM-DCs with supernatants from HEK293T cells transduced at an MOI of 1 with LV::M40-H or LV::S40-H. In parallel, DCs were incubated with supernatants from HEK293T cells transduced with the conventional LV::H, as a negative control, or infected with Mtb, as a positive control. The surface expression of co-stimulatory and MHC molecules was assessed for CD11b+ CD11c+ cells on day 1 post incubation (Fig. 4a, b). We detected no increase in CD40 surface expression for DCs incubated with M40-H or S40-H, likely due to the direct interaction of CD40 with M40-H or S40-H (Fig. 4a, b). CD80 upregulation was only detected for DCs incubated with S40-H, whereas CD86 upregulation and an increase in the percentage of MHC-Ihi and MHC-IIhi cells was observed for DCs incubated with M40-H or S40-H. Very slight functional maturation was also induced by S40-H, as shown by the secretion of minute levels of IFN-α, IL-6, and CCL5, but not IFN-β, IL-1α, IL-1β, IL-10, or TNF-α, detected by a multiplex ELISA assay applied to the supernatants of the same DCs (Fig. 4c). Therefore, through the induction of M40 or S40 secretion, this new generation of LVs is able to induce a certain degree of DC maturation, which is instrumental for appropriate T-cell activation. Taken together, these results indicated that soluble S40-H resulted in a higher degree of DC phenotypic and functional maturation than M40-H (Fig. 4).

Fig. 4: Phenotypic Maturation of DCs Induced by M40 or S40.

a Phenotypic maturation of BM-DCs from C57BL/6 mice infected at an MOI of 5 with Mtb, as a positive control, or incubated with supernatants from HEK-293T cells transduced (MOI = 20) with LV::EsxH alone (Ctrl), LV::M40-H, or LV::S40-H. Expression of co-stimulatory or MHC molecules on the surface of CD11b+ CD11c+ cells at 24 h post-immunization was assessed by cytometry. b Heatmaps showing the mean fluorescence intensity (MFI) of CD40 and CD80 surface expression and the percentage of CD86hi, MHC-I hi, and MHC-II hi DCs. c Quantitation of inflammatory mediators in the culture supernatants of the same DCs. Results are representative of at least two independent experiments.

T-cell immunogenicity of LVs encoding M40 or S40 carrying a single or multiple Mtb immunogens

We next assessed the immunogenicity of this new generation of LVs. BALB/c mice (n = 3/group) were immunized s.c. with LV::M40-H carrying the β2 m22, CMV20, or composite BCUAG promoter (Fig. S1) to obtain insights on the possible consequences of distinct antigen carrier transcription profiles on the induction of immune responses. Control mice were immunized s.c. with a conventional LV::EsxH. On day 13 post injection (dpi), the splenocytes from immunized mice were stimulated with EsxH:20–28 (MHC-I) or EsxH:74–88 (MHC-II) peptides23,24. Although the conventional LV::EsxH induced antigen-specific MHC-I-restricted CD8+ T cells, it was unable to induce antigen-specific MHC-II-restricted CD4+ T cells, as detected by ELISPOT (Fig. S3). The LV::M40-H vectors all induced both CD8+ T and CD4+ T cells (Fig. S3). Intracellular cytokine staining (ICS) showed the multifunctional properties of these CD8+ and CD4+ (Fig. S4A, B) T cells. Functional CD8+ T cell effectors were mainly distributed among IFN-γ+ (single positive), IFN-γ+ TNF-α+ (double-positive), and IFN-γ+ TNF-α+ IL-2+ (triple positive) subsets, whereas CD4+ T cells were essentially IFN-γ+ (single positive) or IFN-γ+ TNF-α+ IL-2+ (triple positive). In contrast to CD8+ T-cell responses, no EsxH-specific CD4+ T-cell responses were detected in the mice immunized with the conventional LV::EsxH. No consistent quantitative or qualitative differences were detected in the T-cell responses in mice immunized with LV::M40 carrying each of the distinct promoters.

We evaluated the immunogenic potential of the developed poly-antigenic LV::M40-HAPE by immunizing C57BL/6 mice (n = 3/group) s.c. with LV::M40-HAPE carrying the β2 m, CMV, or BCUAG promoter. On 14 dpi, CD8+ and CD4+ T-splenocyte responses specific to EsxH:3–11 (MHC-I), EsxA:1-20 (MHC-II), PE19:1-18 (MHC-II), or EspC:45-54 (MHC-I and -II)25 were detected in all mice, as assessed by ELISPOT (Fig. S5). The conventional LV::HAPE induced CD8+, but not CD4+, T cells. ICS analysis of the splenocytes from the same mice showed the multifunctional properties of the induced CD8+ (Fig. S6A, C) and CD4+ (Fig. S6B, D) T cells. Functional CD8+ T cell effectors were again mainly distributed among IFN-γ+ single-positive, IFN-γ+ TNF-α+ double-positive, and IFN-γ+ TNF-α+ IL-2+ triple-positive subsets. CD4+ T cells specific to EsxA, PE10, or EspC antigen were preferentially distributed among IFN-γ+ single-positive, IFN-γ+ TNF-α+ double-positive, and or IFN-γ+ TNF-α+ IL-2+ triple-positive subsets (Fig. S6D). No consistent quantitative or qualitative differences were detected in the T-cell responses in mice immunized with LV::M40-HAPE carrying each of the distinct promoters. Again, in contrast to CD8+ T-cell responses, no antigen-specific CD4+ T-cell responses were detected in the mice immunized with the conventional LV::HAPE.

We then focused on the immunogenicity of LV::S40 vectors. We established the induction of both CD8+ and CD4+ T cells specific to EsxH, EsxA, PE19, and EspC in C57BL/6 mice (n = 3/group) immunized s.c. with LV::S40-HAPEHR or LV::S40-HAPEHR-20 (Fig. 5a). The immunogenicity of Hrp-1 and RpfD was assessed by mapping their epitopes using splenocytes from LV::S40-HAPEHR-immunized mice by ELISPOT (Fig. 5b). The Hrp-1:77-91 (SIYYVDANASIQEML), RpfD:57-71 (IAQCESGGNWAANT), and RpfD:87-101 (SNGGVGSPAAASPQQ) immunogenic regions were identified. Although these epitopes did not induce CD8+ T cells (not shown), they did trigger a CD4+ T-cell response, as assessed by ICS (Fig. 5c). Overall, these results provide evidence of the induction of robust, polyfunctional CD4+ T-cell responses by immunization with this new generation of LVs.

Fig. 5: Immunogenicity of Multi-Antigenic LV::S40-HAPEHR or LV::S40-HAPEHR-20.

a IFN-γ T-cell responses, as assessed by ELISPOT on day 13 post-immunization, in the spleens of individual C57BL/6 mice (n = 3) immunized s.c. with 1 × 108 TU/mouse of LV::S40-HAPEHR or LV::S40-HAPEHR-20. The frequency of responding T cells was determined following in vitro stimulation with the indicated synthetic peptides. Quantitative differences between the groups of mice immunized with LV::S40-HAPEHR or LV::S40-HAPEHR-20, were not statistically significant (Mann–Whitney test). b Epitope mapping of Hrp-1 and RfpD, as determined using pooled splenocytes from 3 C57BL/6 mice/group injected with PBS or immunized s.c. with 1 × 108 TU/mouse of LV::S40-HAPEHR prior to stimulation with each of the individual peptides from the Hrp-1- or RpfD-derived overlapping 15-mers offset by five amino acids. c Cytometric analysis of intracellular IFN-γ vs IL-2 staining of CD4+ T splenocytes after stimulation with 10 µg/ml of the indicated peptides encompassing the immunodominant epitopes identified in b. Pooled splenocytes from three mice/group were used.

Given the better results of: (i) antigenic presentation (Fig. 3), (ii) DC maturation (Fig. 4), and (iii) CD4+ T-cell immunogenicity against six Mtb antigens provided by LV::S40 vectors (Fig. 5), we chose to continue with S40 carrier for the remainder of the study.

Immunogenicity of the poly-antigenic multistage LV::S40 at the mucosal level

We next evaluated the immunogenicity of LV::S40-HAPEHR and LV::S40-HAPEHR20 in C57BL/6 mice immunized intranasally (i.n.) with 1 × 108 TU. Intravenous (i.v.) injection of the immunized mice 14 dpi with PE-anti-CD45 mAb 3 min before sacrifice showed massive T-cell recruitment to the lung interstitium distinct from that in the vasculature26. The lung interstitial (CD45i.v.−) CD4+ (Fig. 6a) or CD8+ (Fig. 7a) T cells of LV::S40-HAPEHR- or LV::S40-HAPEHR20-vaccinated mice contained an elevated number of CD27− CD45RB− CD62L− migrant effectors and CD69+ CD103+ resident cells relative to their PBS-injected counterparts (Fig. 6b, Fig. 7b). Most of the CD69+ CD103+ CD4+ or CD8+ T cells were CD44+ CXCR3+. Some degrees of CD4+ and CD8+ T-cell recruitment to the lungs were also detected with a conventional LV::HAPE (cLV), compared to PBS, albeit the percentages of interstitial CD69+ CD103+ inside the CD4+ subset (Fig. S7A), or those of interstitial CD27− CD45RB- activated cells inside the CD8+ subset (Fig. S7B) in cLV-treated mice did not reached those in LV::S40-HAPEHR- or LV::S40-HAPEHR20-immunized mice (Fig. 6b and Fig. 7a). ICS analysis of these cells showed the presence of (poly)functional CD4+ (Fig. 6c) and CD8+ (Fig. 7c) antigen-specific T cells, essentially located in the lung interstitium. cLV administered i.n. induced mucosal antigen-specific CD8+ (Fig. 7c), but not mucosal CD4+ (Fig. 6c), T cells. We also thoroughly investigated the impact of the i.n. administration of LV::S40-HAPEHR or LV::S40-HAPEHR20 on the composition of the lung innate immune cell subsets on day 1 (Fig. S8A) and day 2 (not shown) post injection. We detected no significant differences in the proportions of various cell subsets vs total lung CD45+ cells relative to PBS-treated mice (Fig. S8B). This observation demonstrates that, despite the capacity of the developed vectors to induce robust T-cell immunity, their administration via the mucosal route has no notable impact on the features of mucosal innate immune cells, indicating the absence of an adverse inflammatory effect after mucosal administration of these vectors.

Fig. 6: Features of mucosal CD4+ T cells triggered by i.n. immunization with LV::S40-HAPEHR or LV::S40-HAPEHR-20.

C57BL/6 mice (n = 3/group) were immunized i.n. with 1 × 108 TU of LV::S40-HAPEHR or LV::S40-HAPEHR-20. On 14 dpi, lung CD4+ T cells were distinguished by their location within the interstitium (CD45i.v−) or vasculature (CD45i.v+) by an i.v. injection of PE-anti-CD45 mAb. a Profile of CD27 vs CD62L or CD45RB, and b CD103 vs CD69, or CD44 vs CXCR3 of lung CD4+ T cells of the interstitium or vasculature. c Percentage of (poly)functional CD4+ T cells specific to EsxA, PE19, or EspC in the lung interstitium or vasculature, as determined by ICS. Results, representative of two independent experiments, were obtained using pooled lungs for each group (n = 3/group). Individual points represent technical duplicates. Results are representative of at least three independent experiments.

Fig. 7: Features of mucosal CD8+ T cells induced by i.n. immunization with LV::S40-HAPEHR or LV::S40-HAPEHR-20.

The immunized C57BL/6 mice are those studied in the Fig. 6. a Shown are lung CD8+ T cells, distinguished for their location within the interstitium (CD45i.v−) or in the vasculature (CD45i.v+). Profile of CD27 vs CD62L or CD45RB, and b CD103 vs CD69, or CD44 vs CXCR3 of the lung CD8+ T cells from the interstitium or vasculature. c Recapitulative percentages of (poly)functional CD8+ T cells specific to EsxH or EspC in the lung interstitium (top) or vasculature (bottom), as determined by ICS. Results are representative of at least two independent experiments.

Although LV::S40-HAPEHR and LV::S40-HAPEHR20 displayed similar efficiency in inducing T cells, based on the capacity of LV::S40-HAPEHR20 to better mobilize DCs in vitro (Fig. 3e), for further experiments we continued with LV::S40-HAPEHR20 vector.

Protective effect of an LV::S40-HAPEHR-20 boost against Mtb infection

Prime-boost strategies using BCG or an improved live-attenuated vaccine for priming and subunit vaccine candidates for boosting is a promising approach to improve the incomplete efficiency of BCG. We assessed the booster potential of LV::S40-HAPEHR-20 by immunizing C57BL/6 mice s.c. at week 0 with 1 × 106 CFU of a genetically improved BCG, i.e., BCG::ESX-1Mmar vaccine candidate27 or leaving them unvaccinated (Fig. 8a). This approach provided the opportunity to perform a prime-boost with the developed LV vaccine, as this live-attenuated vaccine actively secretes EsxA and EspC. A group of BCG::ESX-1Mmar-primed mice was boosted s.c. with 1 × 108 TU of LV::S40-HAPEHR-20 on week 5 and then again boosted i.n. on week 10 with the same LV to recruit the induced immune effectors to the lung mucosa. The choice of the i.n. route of the last boost, after a primary immunization by the systemic route, was based on our observations in the fields of Mtb11,28 and SARS-CoV-222,29,30 and those of many other studies31 showing the importance of boost immunization by the mucosal i.n. route in the immune control of respiratory pathogens. The aim was to target the immune arsenal toward the primary site of infection. On week 12, mice were challenged with ≈200 CFU of Mtb H37Rv via aerosol and lung mycobacterial burdens were determined on week 17 (Fig. 8b). The average lung Mtb load in the primed-boosted mice was ≈2.5 log10 lower that of unvaccinated controls (Mann–Whitney test, p = 0.0005) and ≈1 log10 lower than that of their BCG::ESX-1Mmar-vaccinated counterparts (Mann–Whitney test, p = 0.0415).

Fig. 8: Protective potential of an optimized poly-antigenic LV as a booster against Mtb.

a Timeline of the prime-boost-challenge performed in C57BL/6 mice (n = 5–8 mice/group). b Mtb burden as quantitated by CFU counting in the lungs of BCG::ESX-1Mmar-primed and LV::S40-HAPEHR-20-boosted mice on week 5 post challenge. c Whole-lung section of the left lobe and d quantification of the number and size of lung granulomatous lesions per mouse in each experimental group. e Mtb burden quantified by CFU counting in the lungs of BCG-primed and LV::S40-HAPEHR-20-boosted C57BL/6 mice following the timeline indicated in a, but using Danish BCG for the prime immunization and i.n. Mtb challenge. The pooled results from two independent experiments are shown (n = 6–8 mice/group in each experiment). The significance of the differences was determined using the Mann–Whitney test. ns not significant.

Granulomas were present in all non-vaccinated and BCG::ESX-1Mmar-vaccinated animals but in only 4/9 primed and boosted mice (Fig. 8c, d). They consisted primarily of lymphocytes and epithelioid or foamy macrophages. Polynuclear cells were rare. In non-vaccinated animals, the granulomas were diffuse and disorganized, with a tendency to fuse to each other. In the lungs of vaccinated animals, the granuloma were smaller and neatly delimited from the rest of the parenchyma and assumed a lymphocyte-inside/macrophage-outside organization. Remarkably, five mice in the primed-boosted group had no detectable granulomatous lesions. All histology preparations exhibited at least some degree of alveo-interstitial inflammatory lesions (Fig. 8c), characterized by a mononuclear cell infiltrate, sometimes interspersed with polynuclear cells. In vaccinated mice, the interstitial involvement was generally mild. Infiltrated alveoli were present throughout the sections and were generally separated by free alveoli, although in some preparations, pre-consolidation areas of limited size were observed.

Following the same prime-boost timeline (Fig. 8a), but using Danish BCG for the prime instead of BCG::ESX-1Mmar and i.n. instead of aerosol Mtb challenge, the largely significant protective booster effect of LV::S40-HAPEHR-20 was also confirmed in the lungs of C57BL/6 mice (Mann–Whitney test, p = 0.0025) (Fig. 8e). These results reinforce the interest of this optimized LV in inducing or boosting T-cell immunity against infectious diseases.