A protein subunit vaccine elicits a balanced immune response that protects against Pseudomonas pulmonary infection

Abstract

The opportunistic pathogen Pseudomonas aeruginosa (Pa) causes severe nosocomial infections, especially in immunocompromised individuals and the elderly. Increasing drug resistance, the absence of a licensed vaccine and increased hospitalizations due to SARS-CoV-2 have made Pa a major healthcare risk. To address this, we formulated a candidate subunit vaccine against Pa (L-PaF), by fusing the type III secretion system tip and translocator proteins with LTA1 in an oil-in-water emulsion (ME). This was mixed with the TLR4 agonist (BECC438b). Lung mRNA sequencing showed that the formulation activates genes from multiple immunological pathways eliciting a protective Th1-Th17 response following IN immunization. Following infection, however, the immunized mice showed an adaptive response while the PBS-vaccinated mice experienced rapid onset of an inflammatory response. The latter displayed a hypoxic lung environment with high bacterial burden. Finally, the importance of IL-17 and immunoglobulins were demonstrated using knockout mice. These findings suggest a need for a balanced humoral and cellular response to prevent the onset of Pa infection and that our formulation could elicit such a response.

Fig. 1. Kinetics of serum IgG and IgA.

CD-1 mice were vaccinated on days 0, 14 and 28 and their sera were assessed for anti-PcrV and anti-PopB immunoglobulins as a function of time and boosting. Anti-PcrV IgG (A) and IgA (C), along with anti-PopB IgG (B) and IgA (D), are shown. Individual titers are represented as EU/ml. Each point denotes a mean and error bars represent SD for each group (n = 10 mice/group).

Fig. 2. Serum immunoglobulins possess opsonophagocytic killing (OPK) activity.

The day 42 sera from Fig. 1 were heat-treated at 56 °C for 30 min for subsequent use in OPK assays. A Relative OPK activities are shown for sera collected from each CD-1 mouse group are shown. B A correlation between OPK and either anti-PcrV or anti-PopB immunoglobulin titers is shown. C A correlation between OPK and specific subclasses of IgG against PcrV and PopB is shown. The points in (A) represent individual OPK values obtained from the experiment (error bars represent SD for each group), whereas the points in (B) and (C) represent correlation values obtained between OPK and pooled D42 serum. Pearson’s r coefficient and simple linear regression (95% confidence level) were calculated for (B) and (C). *p < 0.05, **p < 0.01, ***p < 0.001. Exact p values can be found in Supplementary Tables 1, and 2.

Fig. 3. IL-17 and IFN-γ secretion following ex vivo stimulation of lung cells.

Cell suspensions were prepared and treated with either PcrV or PopB or left untreated (media). IL-17 (A) and IFN-γ (B) secreting cells were quantified after 24 h stimulation and are shown as secreting cells per million total cells. Values were plotted as individual points ± SD (n = 5/group). Error bars represent SD. Statistical significance was calculated by comparing PBS vaccinated group with each of the other vaccinated groups using Dunnett’s test. *p < 0.05, **p < 0.01, ***p < 0.001.

Fig. 4. The in vivo protective efficacy of each vaccine formulation was determined.

Mice were challenged with 1 × 10⁸ CFU/30 µl/mouse of mPa08-31 and lung burden was determined after 16 h post infection (HPI) (n = 5). The points represent individual CFU/lung values, and the SDs are denoted by error bars. Lung burden was compared between PBS and all the other groups using a Dunnett’s test. *p < 0.05.

Fig. 5. Correlation between lung burden and lung cytokines.

Cytokine determination assays were performed to allow for identification of potential correlations between Pa lung burden and PcrV- or PopB-stimulated pre-challenge lung cytokines (IL-17A (A), IFN-γ (B), TNF-α (C) and IL-6 (D)). Fold changes between pre- and post-challenge untreated lung cytokines were evaluated to analyze the correlation with lung burden (E). Pearson’s r coefficient and simple linear regression (95% confidence level) were calculated. The exact values can be found in Supplementary Tables 3, and 4. The separation of points in A., B., and C. are readily seen in Supplementary Fig. 3A, B, and C, respectively.

Fig. 6. Correlation between in vitro (OPK) and in vivo reduction in lung burden.

Groups with high in vitro killing ability harbor fewer lung bacteria in vivo. Pearson’s r coefficient and simple linear regression (95% confidence level) were calculated. r = −0.7759, 95% confidence interval = −0.9572 to −0.1572, R squared = 0.6020, p value (two-tailed) = 0.024.

Fig. 7. Pattern of gene up/down regulation following 16 HPI with mPa08-31 in 10 μg L-PaF BECC/ME vaccinated and PBS vaccinated mice lung cells.

CI_1 = Lung cells from control or PBS vaccinated, infected group_sample #1, CI_2 = Lung cells from control or PBS vaccinated, infected group_sample #2, II_1 = Lung cells from immunized or 10 μg L-PaF BECC/ME infected group_sample #1, II_2 = Lung cells from immunized or 10 μg L-PaF BECC/ME infected group_sample #2. Upregulated genes in immunized or 10 μg L-PaF BECC/ME mice are shown in (A), downregulated genes in immunized or 10 μg L-PaF BECC/ME mice are shown in (C). B Pathways upregulated in infected 10 μg L-PaF BECC/ME (or downregulated in infected PBS vaccinated) mice shown via a cnet plot. Upregulated genes were found in cytokine-cytokine interaction, TCR signaling, chemokine signaling, JAK-STAT signaling, Th17, Th1/Th2 differentiation, cell adhesion molecules, antigen processing, NLR signaling pathways. D Pathways downregulated in infected 10 μg L-PaF BECC/ME (or upregulated in infected PBS vaccinated) mice shown via a cnet plot. Downregulated genes were found in cytokine-cytokine interaction, IL-17 signaling, TNF signaling, HIF-1 signaling and MAPK signaling pathways. Fold increase/decrease in (B and D), have been shown by different color according to the log2 scale shown in the figure. Size of the gene clusters have also been shown.

Fig. 8. In vivo protective efficacy study in knock out (KO) mice.

A IL-17 KO or (B) muMt^- KO mice were infected with 1 × 10⁸ CFU/30 µl/mouse and their lungs processed at 16 HPI to evaluate lung bacterial burden (n = 3). The points represent individual CFU/lung values, and the SDs are denoted by error bars. Lung burden was compared between uWT and other groups using Dunnett’s test. *p < 0.05, **p < 0.01, ***p < 0.001. uWT = PBS vaccinated wild type; vWT = 10 μg L-PaF BECC/ME vaccinated WT; u il17 KO = PBS vaccinated IL-17 K; v il17 KO = 10 μg L-PaF BECC/ME vaccinated IL-17 KO; u muMt KO = PBS vaccinated muMt KO; v muMt KO = 10 μg L-PaF BECC/ME vaccinated muMt KO mice.