Development and Testing of a Spray-Dried Tuberculosis Vaccine Candidate in a Mouse Model

Abstract

Converting a vaccine into a thermostable dry powder is advantageous as it reduces the resource burden linked with the cold chain and provides flexibility in dosage and administration through different routes. Such a dry powder presentation may be especially useful in the development of a vaccine towards the respiratory infectious disease tuberculosis (TB). This study assesses the immunogenicity and protective efficacy of spray-dried ID93+GLA-SE, a promising TB vaccine candidate, against Mycobacterium tuberculosis (Mtb) in a murine model when administered via different routes. Four administration routes for the spray-dried ID93+GLA-SE were evaluated along with relevant controls-1) reconstitution and intramuscular injection, 2) reconstitution and intranasal delivery, 3) nasal dry powder delivery via inhalation, and 4) pulmonary dry powder delivery via inhalation. Dry powder intranasal and pulmonary delivery was achieved using a custom nose-only inhalation device, and optimization using representative vaccine-free powder demonstrated that approximately 10 and 44% of the maximum possible delivered dose would be delivered for intranasal delivery and pulmonary delivery, respectively. Spray-dried powder was engineered according to the different administration routes including maintaining approximately equivalent delivered doses of ID93 and GLA. Vaccine properties of the different spray-dried lots were assessed for quality control in terms of nanoemulsion droplet diameter, polydispersity index, adjuvant content, and antigen content. Our results using the Mtb mouse challenge model show that both intranasal reconstituted vaccine delivery as well as pulmonary dry powder vaccine delivery resulted in Mtb control in infected mice comparable to traditional intramuscular delivery. Improved protection in these two vaccinated groups over their respective control groups coincided with the presence of cytokine-producing T cell responses. In summary, our results provide novel vaccine formulations and delivery routes that can be harnessed to provide protection against Mtb infection.

Keywords: ID93+GLA-SE; dry powder vaccine; in vivo murine model; nose-only inhalation device; particle engineering; respiratory delivery; tuberculosis; vaccine adjuvant formulation.

FIGURE 1

(A): Schematic of the aerosol delivery system, consisting of the RBG system (Phillips et al., 2017) connected to the NOID. Spray-dried powder is loaded into the RBG system, which then aerosolizes the powder. The powder aerosol traverses the NOID to be inhaled through the nose of the mice. Red arrows represent the flow of clean air, and green arrows represent the flow of aerosolized powder. (B): Simplified schematic of the system used to assess aerosol particle size at the outlet of the aerosol delivery system. Aerosol particle size at the outlet of the RBG-NOID was measured using an aerodynamic particle sizer (APS). Abbreviations and nomenclature: RBG–Rotating Brush Generator, NOID–Nose-Only Inhalation Device, APS–Aerodynamic Particle Sizer, Q _o–flow rate at the outlet of the RBG-NOID apparatus (8.33 L/min), Q _s–sampling flow rate of the APS (5 L/min), Q _f–flow rate of the air exiting the filter from the tee fitting. Figures created with BioRender.com.

FIGURE 2

SEM images of the trileucine-containing C1 (A) and C2 (B) spray-dried powders. These vehicle powders were designed to not have the antigen or adjuvant system but still be representative of the spray-dried powders developed for the mouse study dry powder administration routes.

FIGURE 3

SEM images of the spray-dried powders prepared for the mouse model experiments. The lots containing trileucine that were designed for delivery of dry powder (A-D-N, V-D-N, A-D-NL/1, A-D-NL/2, and V-D-NL) demonstrate a more rugose particle morphology as compared to the lots intended for delivery after reconstitution (A-L, V-L/1, and V-L/2). The lots designed for dry powder deposition in the nose (A-D-N and V-D-N) are significantly larger than the lots designed for dry powder deposition in the nose and lungs (A-D-NL/1, A-D-NL/2, and V-D-NL). Smaller particles were designed to promote greater airway penetration. Scale bars are based on the respective images. Nomenclature: V–Vaccine; A–Adjuvant; L–Liquid; D–Dry powder; N–Nose; NL–Nose and Lung.

FIGURE 4

Comparison of the physicochemical properties before and after spray drying of the lots manufactured for the mouse model in terms of (A) nanoemulsion droplet diameter, (B) polydispersity index, (C,D) squalene content, and (E–G) GLA content. Measurements of the feedstock liquid are given in black, whereas measurements of the reconstituted powder are given in grey, for each lot. Results are reported as the mean ± standard deviation. The target range for each lot is demonstrated by the red dashed lines. Target range was a nanoemulsion droplet diameter measured as 120 ± 40 nm, polydispersity index <0.2, and squalene and GLA content ±20% of the target feedstock concentration. Nomenclature: V–Vaccine; A–Adjuvant; L–Liquid; D–Dry powder; N–Nose; NL–Nose and Lung.

FIGURE 5

IM, IN, and aerosol vaccinations with ID93+GLA-SE induce cellular immune responses in vaccinated mice. B6 (n = 7–8) mice were either vaccinated with ID93+GLA-SE or only GLA-SE through intramuscular (IM), intranasal (IN), nasal aerosol delivery of large particle dry powder, or pulmonary aerosol delivery of small particle dry powder, at day 0 and day 21. Mice spleens were harvested at 4 weeks post final immunization and stimulated with ID93 antigen (2 μg/ml for 48 h). Levels of splenic (A) IFN-γ and (B) IL-5 were detected in the supernatant by ELISA. Splenocyte ELISA data shown are from one experiment, and similar patterns of response were evident in the repeat experiment. Frequency of (C) CD4⁺GMCSF⁺, (D) CD4⁺TNF-α⁺, (E) CD4⁺IL-17⁺, (F) CD4⁺IFN-γ⁺, (G) CD4⁺CD154⁺, (H) CD4⁺IL-5⁺, (I) CD4⁺IL-2⁺, (J) CD8⁺GMCSF⁺, (K) CD8⁺TNF-α ⁺, (L) CD8⁺IL-17⁺, (M) CD8⁺IFN-γ⁺, (N) CD8⁺CD154⁺, (O) CD8⁺IL-5⁺, and (P) CD8⁺IL-2⁺ T cells in the lung were detected using flow cytometry. Mice lungs were harvested at 4 weeks post final immunization for detection of immune cell components. Flow cytometry data shown are from one experiment and were not repeated in the second mouse experiment. For both flow cytometry and ELISA readouts, responses in unstimulated controls were subtracted from the stimulated samples, and any resulting negative values were assigned as zero. *p < 0.05, **p < 0.01, ***p < 0.001, and p < 0.0001 by one-way ANOVA or Welch’s ANOVA with Sidak’s or Dunnet’s T3 correction, respectively, for multiple comparisons between selected groups. The Kruskal-Wallis non-parametric test with Dunn’s correction for multiple comparisons was used when multiple comparisons were not possible with Welch’s ANOVA due to one experimental group having constant values.

FIGURE 6

IM vaccination with ID93+GLA-SE induces greater serum and mucosal antibody responses than alternative routes/presentations. B6 (n = 7–8) mice were vaccinated with ID93+GLA-SE or only GLA-SE through intramuscular (IM), intranasal (IN), nasal aerosol delivery of large particle dry powder, or pulmonary delivery of small particle dry powder, at day 0 and day 21. Antigen-specific IgA, total IgG, IgG2c, and IgG1 antibody responses in serum and bronchoalveolar lavage (BAL) were measured 1 and 4 weeks following the second immunization by ELISA. Data presented as mean ± SD. For simplicity, all statistical differences are represented as *p < 0.05 even when lower p-values were achieved. Each group immunized by IN or aerosol routes is compared to their adjuvant alone controls. The IM-immunized group comparison (red asterisk) represents statistical significance compared to all 3 alternative routes/presentations of ID93+GLA-SE. Statistical evaluation was conducted by two-way ANOVA with Tukey’s correction for multiple comparisons between selected groups.

FIGURE 7

IN delivery of reconstituted liquid ID93+GLA-SE or pulmonary aerosol delivery of small particle size spray-dried powder ID93+GLA-SE confers protection in a mouse model of Mtb. B6 mice (n = 18–20) were vaccinated with ID93+GLA-SE or only GLA-SE through reconstituted liquid intramuscular (IM), reconstituted liquid intranasal (IN), nasal aerosol delivery of large particle dry powder, or pulmonary delivery of small particle dry powder, at day 0 and day 21. All groups of B6 mice were rested for 4 weeks after which mice were challenged with Mtb H37Rv (100 CFU). Mtb CFU was determined at 4 weeks post-infection. Data presented are combined results from two identical experiments, showing mean ± SD. *p < 0.05 and **p < 0.01 by one-way Welch’s ANOVA with Dunnett’s T3 correction for multiple comparisons between selected groups.