Synthetic Particulate Subunit Vaccines for the Prevention of Q Fever

Abstract

Coxiella burnetti is an intracellular bacterium that causes Q fever, a disease of worldwide importance. Q‐VAX^®, the approved human Q fever vaccine, is a whole cell vaccine associated with safety concerns. Here a safe particulate subunit vaccine candidate is developed that is ambient‐temperature stable and can be cost‐effectively manufactured. Endotoxin‐free Escherichia coli is bioengineered to efficiently self‐assemble biopolymer particles (BPs) that are densely coated with either strings of 18 T‐cell epitopes (COX‐BP) or two full‐length immunodominant antigens (YbgF‐BP‐Com1) all derived from C. burnetii. BP vaccine candidates are ambient‐temperature stable. Safety and immunogenicity are confirmed in mice and guinea pig (GP) models. YbgF‐BP‐Com1 elicits specific and strong humoral immune responses in GPs with IgG titers that are at least 1 000 times higher than those induced by Q‐VAX^®. BP vaccine candidates are not reactogenic. After challenge with C. burnetii, YbgF‐BP‐Com1 vaccine leads to reduced fever responses and pathogen burden in the liver and the induction of proinflammatory cytokines IL‐12 and IFN‐γ inducible protein (IP‐10) when compared to negative control groups. These data suggest that YbgF‐BP‐Com1 induces functional immune responses reducing infection by C. burnetii. Collectively, these findings illustrate the potential of BPs as effective antigen carrier for Q fever vaccine development.

Escherichia coli is engineered to assemble ambient‐temperaturestable biopolymer particles (BPs) coated with either COX epitopes or C.burnetii antigens YbgF and Com1. Respective BPs are safe and immunogenic inmouse and guinea pig (GP) models. Notably, YbgF‐BP‐Com1 immunisation effectivelyreducing C. burnetii infection. This study underscores the effectiveness of BPsas carriers in the development of Q fever vaccines.

1. Introduction

Q fever is an infectious zoonotic disease caused by the Gram‐negative bacterium Coxiella burnetii. The pathogen predominantly infects livestock and transmission to humans occurs through contact with infected animals or their birth products.^{[ 1 ]} The disease in humans initially manifests as acute Q fever, a flu‐like illness that can progress to chronic Q fever with endocarditis.^{[ 2 ]} C. burnetii is designated as a “bioterrorist agent” by the U.S Centre for Disease Control and Prevention (CDC) due to its ease of transmission by aerosols, low infectious dose, and ability to survive in the environment for a prolonged period.^{[ 3 ]} The only current human vaccine available for the prevention of Q fever is Q‐VAX, a formalin‐inactivated whole‐cell vaccine only licensed for use in Australia.^{[ 4 ]} The limited worldwide acceptance of Q‐VAX is due to delayed‐type hypersensitivity reactions (DTH) induced in previously sensitized individuals. For this reason, a pre‐screening test is required to determine pre‐existing immunity prior to vaccination.^{[ 5 ]} In addition, the propagation of phase I C. burnetii used for the preparation of the whole‐cell vaccine involves inoculation of chick embryos within the egg and poses challenges in terms of biosafety requirements, manufacturing costs and potential adverse reactions in egg‐allergic individuals. Hence there is a need for a safe and less reactogenic vaccine that can be cost‐effectively produced at scale.

Over recent years, many whole‐cell and subunit vaccine candidates have been explored to serve as potential Q fever vaccines, but none have progressed to the development of an approved vaccine. Although various vaccine candidates elicited specific immune responses, they failed to confer protective immunity against C. burnetii infection and often induced reactogenic responses.^{[ 6 ]} Subunit vaccines have attracted increasing interest with respect to the development of safe and less reactogenic vaccines.^{[ 7 ]} Subunit vaccines are formulations with immunodominant antigenic components often recombinantly produced and fused with a carrier and/or formulated with adjuvants. In this study, we developed a subunit vaccine with immunodominant antigens of C. burnetii biologically conjugated to a biopolymer. This vaccine approach involves bioengineering of the well‐known safe production host E. coli to synthesize and self‐assemble the biopolymer into spherical particles coated with epitopes and/or antigens of interest. Biopolymer synthesis requires the enzymatic conversion of acetyl‐Coenzyme A to R‐3‐hydroxybutyrl‐Coenzyme A, the biopolymer building block, which is subsequently polymerized by polyhydroxybutyrate synthase (PhaC) resulting in the biopolymer, polyhydroxybutyrate (PHB).^{[ 8 ]} The resulting biopolymer remains inside the cell and constitutes the hydrophobic core of the biopolymer particles (BPs) surrounded by a layer of protein PhaC that remains covalently attached.^{[ 9 ]} Translational fusion of PhaC with antigens/epitopes of interest results in the production of antigen/epitope‐coated BPs which performance as vaccines was demonstrated in the context of infectious diseases such as caused by the intracellular pathogen Mycobacterium tuberculosis or the viruses Hepatitis C and SARS‐CoV2.^{[ 10 ]} The BPs as an antigen delivery system offer many advantages over conventional vaccines including ambient‐temperature stability, the induction of long‐lasting strong humoral and T cell mediated protective immune responses combined with a cost‐effective high‐yield manufacturing process.^{[ 10} , ^{11 ]} Here the BP vaccine approach was applied to design and produce safe and immunogenic Q fever vaccine candidates. We designed two BP vaccine candidates (Figure  1a) comprising selected epitopes or immunodominant antigens of C. burnetii and assessed their safety and immunogenicity in two animal models. A guinea pig (GP) model of Q fever was used to evaluate vaccine efficacy.

Figure 1.

Production and characterization of BPs displaying Coxiella burnetti antigens. a) Schematic of BP vaccine production. Plasmid encoding PhaC synthase is translationally fused with epitopes/antigens of C. burnetii and transformed to endotoxin free production strain, where they self‐assemble to produce BPs expressing C. burnetii antigens/epitopes. The BPs were isolated by mechanical disruption and assessed for their immunogenicity in guinea pig and mouse models. b) Schematic representation of hybrid genes encoding fusion proteins for production of BPs, COX‐BP (where COX refers to the multiple C.burnetii epitope fusion candidate), and YbgF‐BP‐Com1. BP within the schematic represents the BP anchor protein PhaC, the polyhydroxybutyrate synthase. c) Protein profile analysis of whole‐cell lysates and the purified BPs separated by SDS‐PAGE. Lane 1: E. Coli harboring pET14b empty, negative control; Lane 2: pET14b‐BP, Lane 3: pET14b‐COX‐BP Lane 4: pET14b‐YbgF‐BP‐Com1 Lane 5: purified BP (64.3 kDa), Lane 6: purified COX‐BPs (101.1 kDa), Lane 7: Purified YbgF‐BP‐Com1 (127.1 kDa). d) Production, and accumulation of various BPs were analyzed by TEM in whole‐cells and purified BP fractions. Scale bar, 500 nm e) %w/w of PHB content over purified BP mass determined by HPLC‐MS and protein proportion of total BP mass breakdown determined by SDS‐PAGE and densitometry. BP, PhaC protein fraction. f) Zeta‐potential of BPs before and after formulation with alum adjuvant. g) Size distribution of BP vaccines before and after formulation with alum. Each data point of measurement represents the mean ± the standard error of mean (SEM). Measurements were done in triplicates.

2. Results

2.1. Design, Production, and Characterization of Synthetic BP Vaccine Candidates against Coxiella burnetii

To produce the multiple T‐cell epitope BP vaccine candidate (COX‐BP) (Figure 1b) predicted HLA class I and class II epitopes of C. burnetii were selected based on their HLA binding capacity and ability to induce cellular responses in mice and induce IFN‐γ recall responses in humans exposed to a Q fever outbreak (Table S1, Supporting Information).^{[ 12 ]} A total of 18 epitopes were translationally fused to the N terminus of PhaC that serves as covalent BP anchor. For the YbgF‐BP‐Com1 fusion vaccine candidate, the PhaC domain was flanked by the two complete antigen sequences. YbgF was fused at the N terminus and Com1 to the C terminus (Figure 1b). Empty BPs, COX‐BP and YbgF‐BP‐Com1 were generated by expressing the respective hybrid genes in E. coli (Tables S2 and S3, Supporting Information). The corresponding recombinant plasmids were expressed in an endotoxin‐free E. coli strain engineered to produce the precursor, (R)−3‐hydroxybutyrl‐CoA, required for BP assembly. The endotoxin‐free production strain was cultivated in a bioreactor using animal component‐free synthetic media with glucose as carbon source. To confirm production of antigen‐coated BPs, cells were harvested, disrupted, and subjected to BP purification. BP containing fractions were subjected to sodium dodecyl sulfate‐polyacrylamide gel electrophoresis (SDS‐PAGE) analysis, which confirmed the presence of recombinant proteins displayed on the BPs. The prominent bands corresponded to the expected theoretical molecular weights of the antigens fused to the BP anchor, namely COX‐BP (101.1 kDa), YbgF‐BP‐Com1 (127.1 kDa), and BP (64.3 kDa) (Figure 1c). Protein identity was further confirmed by using matrix‐assisted laser desorption/ionization (MALDI)‐Mass spectrometry (MS)‐quadrupole‐time‐of‐flight‐(Q‐Tof) (Table S4, Supporting Information). TEM analysis revealed images showing the formation and accumulation of BPs inside E. coli, as well their morphology after purification from E. coli (Figure 1d). The presence of the BP and PHB was confirmed by high‐performance liquid chromatography‐mass spectrometry (HPLC‐MS). PHB contributed 83%, 69%, 54% of dry weight of purified BP, COX‐BP, and YbgF‐BP‐Com1 respectively. Densitometry analysis was done to determine the fusion protein percentage over BP mass (Figure 1e).

BPs were further characterized to measure their size and surface charge (Figure 1f,g). Size and charge are crucial factors for efficient uptake by antigen‐presenting cells.^{[ 13 ]} Zeta potential analysis demonstrated that all BPs had anionic surface charges (Figure 1f). Size analysis revealed that the BPs were polydisperse with variations within a sample as previously reported,^{[ 14 ]} however, the polydispersity index of all BPs was < 30% (Figure S1, Supporting Information) indicating narrow size distribution. The effect of adjuvant on BP size and charge was analyzed and all BPs once formulated with alum showed a cationic surface charge with slight deviations from their original size (Figure 1g).

2.2. Immunological Properties of BP‐Based Q fever Vaccine Candidates

Three animal trials were conducted in this study. As GPs mimic the clinical course of C. burnetii infection in humans and are more susceptible to infection they are considered as an effective model for Q fever vaccine assessment.^{[ 15 ]} Initial safety, immunogenicity, and efficacy assessment of the vaccine candidates was performed in a female GP model. The best performing candidate was further assessed in a mouse model to confirm reproducibility of immunogenicity in a different animal model. The final evaluation of the best performing candidate was performed in the male GP model of C. burnetii infection. Testing vaccine performance in female and male GPs allowed to assess gender‐specific effects on immune responses.

To evaluate safety and immunogenicity of BP vaccine candidates, female GPs were immunized with empty BPs, COX‐BP, and YbgF‐BP‐Com1. The BP vaccine candidates were formulated with alum and without alum. Alum alone served as the mock vaccinated group for this study. All GPs were immunized subcutaneously with 20 µg of antigen followed by two boosters (Figure  2a). No adverse reactions were observed. Blood was collected on day 21 and day 72 following the initial vaccine dose to assess C. burnetii antigen‐specific antibodies induced by BP vaccine candidates. Enzyme‐linked immunosorbent assay (ELISA) plates were coated with respective BPs and sera from each group were analyzed to assess the induction of IgG antibodies against respective antigens. Empty BPs induced low IgG background levels both on day 21 and day 72 (Figure 2b), with the alum group inducing significantly less or non‐detectable IgG levels against the BPS. GPs immunized with COX‐BP (+Alum) induced an EC 50 value of 2650 on day 21 and this increased by ≈14‐fold on day 72 to an average EC 50 value of 35 553 (Figure 2c). Although, COX‐BP (‐Alum) immunized GPs had a significantly higher EC 50 value on day 21 (p<0.05) compared to GPs Immunized with COX‐BP (+ Alum), the EC50 values were markedly less in the non‐adjuvanted (‐Alum) group on day 72 (Figure 2c). GPs immunized with YbgF‐BP‐Com1 induced high levels of IgG both on day 21 and day 72 (Figure 2d). This was demonstrated using formulations with and without alum. The EC50 values of YbgF‐BP‐Com1 were significantly higher compared to empty BPs on day 21 and day 72 (p <0.05) (Figure 2d).

Figure 2.

Immunogenicity study of C. burnetii antigen‐coated BPs in female guinea pigs. Antigen‐specific antibody responses in guinea pigs vaccinated with various BP vaccines coated with selected antigens/epitopes and the empty BP were measured by ELISA and represented as EC50 values. a) Immunization schedule of female guinea pig trial. b) IgG response against empty BPs; plates were coated with the empty BPs and tested against serum of all groups collected on Day 21 and Day 72. c) IgG response against COX‐BPs. d) IgG response against YbgF‐BP‐Com1. e) IgG response against BP‐Com1. f) IgG response against BP‐YbgF. Each data point of measurement represents the results from ≥2 guinea pigs ± SEM. Statistical significance for multiple group comparison were assessed by one way ANOVA with Dunnett's correction for multiple comparison; to assess statistical significance between groups on day 21 and day 72 Wilcoxon and Mann‐Whitney U test were performed where appropriate. The BP group and alum group were combined as control groups for statistical analysis. (*p < 0.05 ** p < 0.01).

As COX‐BP includes epitopes derived from Com1 and YbgF, GPs immunized with this vaccine candidate produced anti‐Com1 and anti‐YbgF antibodies, which are detected in the ELISAs against YbgF‐BP‐Com1. This was further confirmed by testing the serum of COX‐BP group against individual BP‐Com1 and BP‐YbgF coated microtiter plates using ELISA (Figure 2e,f). Accordingly, GPs immunized with YbgF‐BP‐Com1 produced antibodies cross‐reacting with COX‐BP when compared to empty BPs (Figure 2d). To assess which antigen (YbgF or Com1) of YbgF‐BP‐Com1, induced greater IgG responses, serum from this group was tested against individual BP‐Com1 and BP‐YbgF using ELISA. Significantly higher EC50 values were found for antibodies binding BP‐Com1 than BP‐YbgF, which indicated Com1 induced a dominating IgG response (Figure 2e). To demonstrate antigen‐specific antibody responses, immunoblots were performed. Whole‐cell lysates of respective BPs harboring host cell protein and purified BPs were separated by SDS‐PAGE followed by immunoblotting analysis. Pooled serum from respective groups was used to assess the specific binding of antibodies to the antigens coating the BPs. Both BP candidates induced antibodies against antigens displayed on the BPs. For YbgF‐BP‐Com1, induced antibodies specifically bound to Com1, YbgF, and COX, the multi epitope fusion. COX‐BP vaccination induced antibodies directed against, COX, YbgF and Com1 (Figure S2, Supporting Information). Both BP vaccine candidates did not induce detectable antibodies binding to the anchor protein PhaC or to host cell proteins (Figure S2).

2.3. BP Vaccines Confer Protective Immune Responses against C. Burnetii Challenge in a Female Guinea pig Model

Here, we evaluated the efficacy of BP vaccine candidates by challenging the immunized GPs with the phase I Nine Mile strain of C. burnetii. GPs were intranasally infected with 10⁶ genomic equivalents of C. burnetii 14 days post‐last booster. Body temperatures were measured daily for 21 days using implanted temperature transponders (Figure  3 ). Body temperature of ≥ 39.5 °C was defined as fever.^{[ 6} , ^{16 ]} Mock vaccinated and empty BP groups showed similar trends and hence were combined for comparison as the placebo group. Most GPs from the “placebo group” developed fever (body temperature of > 39.5 °C) by day 11 post‐challenge (Figure  5a) out of six GPs, four developed fever (Figure 3a). Four out of five GPs from the YbgF‐BP‐Com1 (+Alum) group developed borderline fevers (temperature = 39.5 °C) on day 10 or 11 post‐challenge except for two GPs which had fever for 2 days (Figure 3a). In contrast, most GPs from the YbgF‐BP‐Com1 (‐Alum) group did not develop fever except for one GP that developed fever for a single day (Figure 3a). GPs immunized with COX‐BP (+ Alum) and COX‐BP (‐Alum) exhibited similar fever responses with two GPs from each group developing low‐grade fever on day 11 (Figure 3a). On average YbgF‐BP‐Com1 (‐Alum) and COX‐BP (+ Alum) groups developed reduced fever responses post‐challenge in comparison to the placebo group (Figure 3b,c). Body weights were measured every three days over 21 days post‐challenge (Figure 3d). A significant weight loss was observed on day 11 post‐challenge in the placebo group (Figure 3d). No significant weight reduction was exhibited by the vaccine groups. GPs immunized with COX‐BP(+Alum) showed a reduction in body weight on day eleven after which they recovered by day 18, whereas the reduction was sustained until day 21 in the placebo group (Figure 3d). All GPs from YbgF‐BP‐Com1 (+Alum), YbgF‐BP‐Com1 (‐Alum), and COX‐BP (‐Alum) exhibited similar body weights with no detrimental weight loss over 21 days (Figure 3d,e).

Figure 3.

Clinical signs of C. burnetii challenged female guinea pigs. Body temperatures were recorded daily from the day of infection until day 21 post infection. The placebo group represents the combined temperatures of GPs in BP: Alum and Alum group. a) Temperatures were plotted for each individual animal per time point. Fever was defined as ≥ 39.5 °C and is indicated by the dotted red line. b) Combined body temperatures of GPs immunized with COX‐BP: Alum and COX‐BP. c) Combined body temperatures of GPs immunized with YbgF‐BP‐Com1: Alum and YbgF‐BP‐Com1. Each data point represents mean of ≥2 guinea pigs ± standard error of mean (SEM). d) Percentage reduction of body weights in C. burnetii infected female guinea pigs. Combined body weights gained/lost in GPs immunized with COX‐BP: Alum, COX‐BP. e) Combined body weights gained/lost in GPs immunized with YbgF‐BP‐Com1: Alum, YbgF‐BP‐Com1. Body weights were recorded every three days post challenge with C. burnetii. (day 0) until 3 weeks (day 21). Body weight change is presented as percentage decrease/increase of body weight compared to day −1 post challenge. Each data point represents mean body weight change of ≥2 guinea pigs ± SEM. Statistical significance between independent groups were assessed by Mann‐Whitney U test (*p < 0.05).

Figure 5.

Immunogenicity study of C. burnetii antigen‐coated BPs in male guinea pigs. Antigen‐specific antibody responses in guinea pigs vaccinated with YbgF‐BP‐Com1 and empty BP were measured by ELISA and represented as EC50 values. a) Immunization schedule of male guinea pig trial. b) IgG response against empty BPs; plates were coated with the empty BPs and tested against serum of all groups collected on Day 21 and Day 72. c) IgG response against YbgF‐BP‐Com1. d) IgG response against BP‐Com1. e) IgG response against BP‐YbgF. f) IgG response against Q‐VAX^® presented as IgG titers. YbgF‐BP‐Com* represents the uninfected control group. g) Avidity Index of antibodies induced by Q‐VAX^® and YbgF‐BP‐Com1 and empty BP against itself. For YbgF‐BP‐Com1 induced antibodies additional avidity indices were obtained against the individual antigens. Each data point of measurement represents the results from 8 guinea pigs ± the standard error of mean (SEM). Statistical significance for multiple group comparison were assessed by one way ANOVA with Dunnett's correction for multiple comparison; to assess statistical significance between groups on day 21 and day 72 Wilcoxon or Mann‐Whitney U test were performed respectively where appropriate. (*p < 0.05 ** p < 0.01 ***p < 0.001, ****p < 0.0001).

2.4. YbgF‐BP‐Com1 Immunogenicity Assessment in a Mouse Model

As the GP trial data suggested that YbgF‐BP‐Com1 seemed to show the most promising immunological properties, this vaccine candidate was further characterized. The vaccine was subjected to a safety and immunogenicity study in mice to further corroborate its ability to elicit strong and specific humoral responses in another animal model. Female C57BL/6 mice were subcutaneously vaccinated with 20 µg of YbgF‐BP‐Com1 adjuvanted with alum. Two boosters were administered with a time interval of two weeks between each administration (Figure  4a). A challenge study was excluded in the mouse trial, therefore the vaccination schedule implemented in the mouse study differed from that of the GP study. Alum alone served as the negative control for this study. No adverse reactions were observed. Blood was collected on day 21 and day 42 to assess humoral response using ELISA. YbgF‐BP‐Com1 immunized mice elicited significantly high IgG antibodies against YbgF‐BP‐Com1 on day 21 and 42 post‐immunization, compared to empty BPs and mice immunized with alum (Figure 4b). Besides total IgG, IgG subtypes were determined. IgG1 and IgG2c mirrored a similar response to total IgG with significantly higher EC50 values on days 21 and 42 compared to mice immunized with alum alone (Figure 4b). To further evaluate the specificity of antibody responses, immunoblots were performed using pooled sera (Figure 4c). Specificity was confirmed as antibodies only detected YbgF and Com1 antigens and not any BP carrier components or host cell proteins.

Figure 4.

Immunogenicity study of C. burnetii antigen‐coated BPs in female C57BL/6. a) Immunization schedule of mouse trial. b) Antigen‐specific antibody responses in mice vaccinated with YbgF‐BP‐Com1 and alum were measured by ELISA and represented as EC50s. The plates were coated with the empty BPs, and YbgF‐BP‐Com1 respectively. Serum collected on day 21 and day 42 were serially diluted and assessed for antigen specific response. Total IgG response, IgG1 response, IgG2c response presented as EC 50s. Each data point of measurement represents the mean of 10 mice ± SEM. c) The specific antibody responses were analyzed by Western blot using pooled serum samples from mice immunized with YbgF‐BP‐Com1. Bis‐tris gel (left) shows the order of BPs used for the immunoblots. Immunoblots (right) demonstrate antigen specific response elicited by YbgF‐BP‐Com1. Lane 1: E. coli harboring pET14b empty, negative control; Lane 2: pET14b‐BP; Lane 3: pET14b‐YbgF‐BP‐Com1; Lane 4: pET14b‐BP‐Com1; Lane 5: pET14b‐BP‐YbgF; Lane 6: purified BPs (64.3 kDa); Lane 7: Purified YbgF‐BP‐Com1 (127.1 kDa); Lane 8: Purified BP‐Com1 (93.3 kDa); Lane 9: Purified BP‐YbgF (99.9 kDa). Statistical significance between independent groups were assessed by Mann‐Whitney U test. (*p < 0.05 ** p < 0.01).

2.5. YbgF‐BP‐Com1 Vaccine Confers Protection against C. Burnetii Challenge in a Male Guinea pig Model

Here, we extended the vaccine candidate evaluation study to the more widely used male GP model of C. burnetii infection and with the inclusion of Q‐VAX vaccinated GPs as the positive control group. Male GPs were subcutaneously vaccinated with 20 µg of YbgF‐BP‐Com1 formulated with alum, and the immunization schedule was similar to the female GP study (Figure 5a). Q‐VAX and alum served as positive and negative controls, respectively. In addition, a vaccinated: un‐challenged group was included as a vaccination control. No adverse reactions were observed. Empty BP induced antibody responses at low detectable levels (Figure 5b). As previously demonstrated the YbgF‐BP‐Com1 vaccine induced strong humoral immune responses (Figure 5c) both on days 21 and 72. To assess which antigen (YbgF or Com1) of the YbgF‐BP‐Com1 induced higher IgG levels, sera were tested against individual Com1 and YbgF using ELISA. Responses to Com1 were significantly greater than to YbgF (Figure 5d,e). Interestingly, serum from the YbgF‐BP‐Com1 vaccinated group contained C. burnetii‐specific antibodies that cross‐reacted with Q‐VAX, i.e., whole cell C. burnetii, in ELISA (Figure 5f). An avidity assay was performed to determine the quality of antibodies elicited by YbgF‐BP‐Com1. The anti‐Com1 and anti‐YbgF antibodies induced by the YbgF‐BP‐Com1 vaccine were of high avidity of 85% and 75%, respectively, when compared to the avidity against dual antigen fusion (68%) (Figure 5g). Immunoblots confirmed the specificity of the immune responses elicited by the vaccine candidate (Figure S3, Supporting Information).

Fever responses and loss of body weight post‐challenge with C. burnetii was monitored as in the previous trial (Figure  6 ). The non‐challenged vaccine group did not exhibit fever over the monitoring period (Figure 6a). Within the group of BP vaccinated GPs, six out of eight exhibited fevers, while two GPs were euthanized due to the severity of infection. Most GPs in the alum group developed fever by day seven which lasted for four days. Except for 3 GPs, the rest of the GPs in the Q‐VAX group did not develop any fever responses. Five out of eight GPs in the YbgF‐BP‐Com1 vaccine challenge group experienced low‐grade fever (< 40 °C); one additional GP in this group was the exception, with a high fever of 40.4 °C on day 7 post‐challenge (Figure 6a). The YbgF‐BP‐Com1 showed varied fever duration due to borderline fever responses from most of the GPs; However, fever responses tended to be less compared to the alum group (Figure 6b,c). A reduction in body weight occurred by day five in all infected groups. On days five and seven, the BP and Q‐VAX groups exhibited a significant decrease in body weight compared to the vaccine group (Figure 6d). Q‐VAX‐immunized GPs started to recover from day seven, whereas this was day ten for the GPs in the vaccine group. A similar trend was observed for GPs from the alum and BP groups, with the greatest weight reduction observed in the BP group (Figure 6e). To determine the efficacy of the YbgF‐BP‐Com1 vaccine candidate to reduce C. burnetii burden in organs, C. burnetii DNA in the lungs, liver, kidney, and spleen of GPs was determined by quantitative PCR (Figure 6f; Figure S4, Supporting Information). C. burnetii DNA was detected in all organs in GPs from Q‐VAX, alum, and YbgF‐BP‐Com1 vaccine groups (Figure S4, Supporting Information). The vaccine unchallenged group did not have any detectable C. burnetii DNA as expected. Out of 8 GPs, 5 GPs in the Q‐VAX group exhibited measurable spleen loads, among which three had high loads (Figure S4a, Supporting Information). Only four GPs from the YbgF‐BP‐Com1 vaccine group showed detectable loads, with the rest four displaying marginal or no detectable C. burnetii (Figure S4a, Supporting Information). Reduced C. burnetti DNA was detected in the liver of Q‐VAX and the YbgF‐BP‐Com1 vaccine group compared to alum, with the majority of GPs (five out of eight GPs) from the YbgF‐BP‐Com1 group exhibiting negligible loads (Figure 6f). Accordingly, C. burnetii clearance in the liver by the YbgF‐BP‐Com1 vaccine group was significant (p<0.05) compared to the alum group (Figure 6f). No significant reduction in bacterial loads was found in the kidney and lungs of Q‐VAX, YbgF‐BP‐Com1, and alum group (Figure S4a, Supporting Information). However, only three GPs from the YbgF‐BP‐Com1 vaccine group exhibited detectable lung bacterial burden compared to five out of eight GPs in the Q‐VAX group (Figure S4a, Supporting Information), suggesting some reduction of C. burnetti lung burden by YbgF‐BP‐Com1. The reduced overall loads in the kidney and lungs of GPs from alum group were likely due to the natural clearance of C. burnetti at the time the tissues were harvested (day 26 post‐challenge). Collectively our results indicate the potential of YbgF‐BP‐Com1 to provide substantial C. burnetti clearance in the liver with moderate clearance in the lungs. To further characterise the protective response elicited by YbgF‐BP‐Com1, cytokine expression in the spleen of immunized GPs post infection were measured (Figure S5, Supporting Information). Q‐VAX and YbgF‐BP‐Com1 immunized GPs mirrored similar trends in cytokine expression. GPs from these groups exihibited upregulated expression of IP‐10 and IL‐8, compared to the unchallenged group, whereas IL‐12 expression was reduced. Intrestingly, the uninfected vaccine group revealed a contrasting IP‐10 expression to the placebo group indicating a differential cytokine profile between infected and uninfected groups.

Figure 6.

Clinical signs of C. burnetii challenged male guinea pigs. Body temperatures were recorded daily from the day of infection until day 17 post infection. a) Temperatures were plotted for each individual animal per time point. Fever was defined as ≥ 39.5 °C and is indicated by the dotted red line. The uninfected YbgF‐BP‐Com1 control group is presented as “Unchallenged” b) Combined body temperatures of GPs immunized with Q‐VAX^®, YbgF‐BP‐Com1 and alum, YbgF‐BP‐Com1. Each data point represents mean of eight guinea pigs ± standard error of mean (SEM). c) Fever duration of GPs in vaccine groups. d) Percentage reduction of body weights in C. burnetii infected male guinea pigs. Body weights were recorded every 2–3 days post challenge with C. burnetii. Body weight change is presented as percentage decrease/increase of body weight compared to day −1 post challenge. Each data point represents mean body weight change of 8 guinea pigs ± SEM. e) Survival analysis of GPs with a probability of >10% weight loss. The lines at the top of the graph represent individual GPs that did not have weight loss >10% and each step down represents when a GP first had weight loss of >10%. f, C. burnetti load in male GPs liver post challenge. Copy number of C. burnetti gene Com1 was normalized against GAPDH. Data are represented as mean ± SEMs. Statistical significance between groups were assessed by Wilcoxon or Mann‐Whitney U test and one‐way ANOVA with Dunnett's correction where appropriate. Statistical significance for body weight was determined using a restricted maximum likelihood linear mixed model (REML) incorporating repeated measures and was compared against YbgF‐BP‐Com1 group. (*p < 0.05, ** p < 0.01).

2.6. BP Q Fever Vaccine Candidates are Stable Under Ambient‐Temperature Conditions

Vaccines that are ambient‐temperature stable are desirable as they exhibit increased shelf‐life stability and enable facilitated dissemination. The stability of BP vaccine candidates at different temperatures was assessed following treatment at 4, 25, 37, and 56 °C for 28 days (Figure  7 ). We performed ELISAs, to demonstrate the retained antigenicity of various temperature treated BP vaccine candidates (COX‐BP, YbgF‐BP‐Com1). Plates were coated with the vaccine candidates and tested against the serum of GPs that induced a high IgG response in the respective groups from the female GP trial (Figure 7a,b). The EC50 values of different temperature treated BP vaccine candidates on day seven were compared with the EC50 values of temperature treated BP vaccine candidates on days 14, 21, and 28. No significant differences in EC50 values were seen between day seven and days 14, 21, and 28 in different temperature treated BP candidates. We also assessed the BPs size distribution and zeta potential to confirm retention of inherent BP properties following incubation at various temperatures over 28 days (Figure 7c,d; Figure S7, Supporting Information). In addition, to confirm the retention of antigens the intensity of bands on SDS‐PAGE of treated BP vaccine candidates was quantified (Figure S6a,b, Supporting Information). A gradual decrease in protein intensity was observed with YbgF‐BP‐Com1 candidate, whereas COX‐BP maintained a similar intensity at all temperatures over 28 days.

Figure 7.

Stability assessment of temperature treated BP vaccines over 28 days. Antigenicity of temperature‐treated BPs analyzed by ELISA with serum from immunized GPs and antibody titers presented as EC50s. Serum from female guinea pigs immunized with YbgF‐BP‐Com1 and COX‐BP was tested against temperature‐treated YbgF‐BP‐Com1, COX‐BP and plain BPs. a) Serum from GPs immunized with YbgF‐BP‐Com1 was tested against temperature‐treated YbgF‐BP‐Com1 and plain BPs. b) Serum from GPs immunized with COX‐BP were tested against temperature‐treated COX‐BP and plain BPs. The experiment was done in triplicates and the measurement represents the mean ± SEM).c) Particle size distribution of temperature treated YbgF‐BP‐Com1. d) Particle size distribution of temperature treated COX‐BP. Statistical significance, were assessed by one way ANOVA with Dunnett's correction for multiple comparison (*p < 0.05).

3. Discussion

Q‐VAX is the current preventive vaccine available for Q fever. However, worldwide distribution is restricted due to its undesirable reactogenic responses and cumbersome pre‐screening process. Q‐VAX is an inactivated whole cell preparation of Phase I C. burnetti, which induces a high amplitude of secondary responses manifested as DTH reactions in previously exposed individuals. In addition, there are safety concerns associated with whole cell vaccine use and manufacture. Hence there is need for a safe and equally efficacious but less reactogenic vaccine that can be efficiently produced at scale for worldwide use.

In this study, we aimed to develop a safe and immunogenic subunit vaccine by incorporating immunodominant antigens or epitopes of C. burnetii into the surface of bacterially assembled BPs. This eliminates potential non‐specific immunogens present in the whole cell vaccine, which are responsible for eliciting secondary DTH reactions in sensitized individuals. Furthermore, our synthetic subunit vaccine design offers the advantage of an underlying cost‐effective manufacturing process without the need for tightly regulated Biosafety Level‐3 (BSL‐3) conditions. Antigen‐coated BPs are ambient‐temperature stable and are efficient inducers of humoral and cellular responses as was previously demonstrated in the context of other infectious diseases.^{[ 10} , ^{11 ]} Here, we developed two subunit BP vaccines against C. burnetii, an epitope‐based vaccine, that included HLA class I and class II restricted T cell epitopes identified by immunoinformatic analysis and evaluated in mice for their immunogenicity. These epitopes were further validated by their ability to induce IFN‐γ recall responses in a human population exposed to a Q fever outbreak.^{[ 12 ]} Further class I epitopes were selected based on mouse protection studies (Table S1, Supporting Information).^{[ 12b ]} The second vaccine candidate was a two‐antigen vaccine containing two complete sequences of the C. burnetii immunogenic antigens, YbgF and Com1. Com1 is a conserved outer membrane protein of C. burnetii. It is widely detected by antibodies sourced from individuals with Q fever. Furthermore, it has been shown as a potential vaccine candidate inducing protective responses in vaccination studies.^{[ 12} , ^{17 ]} YbgF is a component of the tot‐pal system protein complex that maintains outer membrane stability.^{[ 18 ]} Previously, it was shown to induce protective immunity in mice^{[ 12} , ^{19 ]} and it was bioinformatically predicted to be a potential vaccine target.^{[ 20 ]} These vaccine candidates either incorporating two antigens or multiple epitopes, were effectively produced in one‐step production process using our engineered endotoxin‐free E. coli mutant. Purified vaccine candidates were characterized to evaluate their potential as particulate vaccine candidates. The antigen composition of BPs was confirmed by SDS‐PAGE, HPLC‐MS and MALDI‐MS‐Qtof analysis (Figure 1). HPLC‐MS analysis of purified BPs showed a PHB content (%w/w) of 60–80% indicative of their purity. Surface charges mediate cellular uptake by antigen‐presenting cells, with cationic surfaces facilitating the internalization and expression of chemokines for dendritic cell (DC) recruitment.^{[ 21 ]} Anionic charges drive the response toward a balanced T_H1 and T_H2 response, whereas cationic charges drive toward a T_H1 response.^{[ 22 ]} Our BP vaccines are anionic when formulated with alum, but cationic by themselves, indicating their ability to induce a mixed T_H1 and T_H2 response (Figure 1f). All BPs were ≤1 µm in size, with a polydispersity index (PDI) of < 30%, suggesting polydispersity with a narrow size distribution (Figure 1g; Figure S1, Supporting Information). The size determines the cellular uptake and migration pathway of the vaccine particles, with particles of larger size (500 nm −2 µm) being efficiently taken up by migratory DCs that present epitopes to T cells for cell‐mediated immune responses.^{[ 23 ]} The detection of particulate antigens by B cells in lymph nodes involves intricate interactions with DCs and macrophages. DCs expressing elevated levels of mannose receptors have been linked to the active transport of antigens into B‐cell follicles.^{[ 13} , ^{24 ]} In addition, recent findings indicate that particulate antigens accumulate in the macrophage‐rich area at the boundary of lymph node follicles, where cognate B cells acquire the antigen, thus initiating cellular responses.^{[ 24} , ^{25 ]}

BP vaccine candidates were evaluated in two animal models, mouse and GP, for their safety and immunogenicity, and in a GP infection model to evaluate efficacy. In the female GP trial, both vaccines induced stronger humoral immune responses than empty BPs, with YbgF‐BP‐Com1 inducing a superior response (Figure 2). As a result of the inclusion of Com1 and YbgF epitopes in the COX‐epitope BP vaccine, detectable antibodies were detected in YbgF‐BP‐Com1 serum. The anti‐YbgF and anti‐Com1 antibodies provoked by the YbgF‐BP‐Com1 could detect YbgF and Com1 specific epitopes within the COX BP candidate. However, YbgF‐BP‐Com1 outperformed COX‐BP in both the presence and absence of alum. It elicited an EC50 of 180709 with alum and an EC50 of 96360 without alum on day 21, which slightly peaked on day 72 without alum and marginally decreased with alum. The humoral response elicited by YbgF‐BP‐Com1 without alum was comparable to that elicited by YbgF‐BP‐Com1 formulated with alum. Furthermore, the lack of reduction in body weight post‐infection in this non adjuvanted group indicates that YbgF‐BP‐Com1 possesses inherent self‐adjuvating properties (Figure 3). This corroborates previous studies with BP based vaccine candidates, such as those designed to prevent Neisseria meningitides infection.^{[ 14b ]}

The findings from the initial female GP trial revealed that YbgF‐BP‐Com1 is a promising candidate in terms of safety, immunogenicity, and protective efficacy, hence, to confirm the safety and immunogenicity of YbgF‐BP‐Com1, it was tested in a different animal model, the mouse model. The vaccine induced high EC50 values of antigen‐specific antibodies in mice on days 21 and day 42, corroborating our finding in female GPs. Accordingly, no immune response to the carrier, i.e., empty BPs, or host cell proteins was detectable, suggesting that neither carrier nor host cell protein impurities could induce immune suppression. YbgF‐BP‐Com1 induced various subtypes of IgG, IgG1, and IgG2c in mice, indicating the ability of YbgF‐BP‐Com1 to steer isotype switching (Figure 4). The production of these antibody isotypes suggests induction of both T_H1 and T_H 2 immune responses. Typically, the T_H1 subtype of CD4⁺ T cells steers the humoral response towards the IgG2c subtype, and T_H 2 CD4⁺ T cells drive the production of the IgG1 subtype. The findings of our study indicate that YbgF‐BP‐Com1 can direct the response toward both T_H1 and T_H 2 CD4⁺ T cell responses.​ While a T_H1 skewed response is crucial for C. burnetii clearance, the IgG1 isotype induced by T_H2 is known to recognize immunodominant antigens present in phase I and phase II C. burnetii, assisting clearance.^{[ 26 ]} Additionally, C. burnetii infection is associated with elevated IgG1 levels in patients with Q fever, possibly indicating that this isotype may correlate with protection.^{[ 27 ]}

In the final study, we employed the more widely used male GP model of C. burnetii infection and included Q‐VAX as a positive control for benchmarking purposes. Antibodies induced by YbgF‐BP‐Com1 cross‐reacted with Q‐VAX, i.e., whole cells of C. burnetii, indicating their functionality to bind to the native antigen counterparts, enabling opsonization of C. burnetii cells toward complement‐mediated killing (Figure 5). It is noteworthy that no adverse effects at the injection site were observed even after multiple boosters across all three trials, suggesting that no sensitization occurred, as is the case for Q‐VAX.^{[ 5} , ^{28 ]} Hence, the YbgF‐BP‐Com1 vaccine candidate is deemed non‐reactogenic. In addition, Q‐VAX generated antibodies with an avidity index of 93%, as demonstrated in previous studies^{[ 29 ]} while YbgF‐BP‐Com1 induced antibodies with an index of 68% (Figure 5g). The ability of the two‐antigen vaccine YbgF‐BP‐Com1 to induce high avidity antibodies further underpins functionality of the immune response.

Fever responses and reduction in body weight post‐challenge mirrored similar trends to those observed in the female GP trial (Figure 6a–e). The mortality of the two GPs in the BP group provides compelling evidence for the inadequate protection conferred by BP alone. Moreover, this finding was reinforced by the notable weight loss observed among members of this group (Figure 6e). The reduction of clinical signs in vaccinated challenged groups compared to the placebo group is further explained by differential cytokine expression in the spleens of these GPs (Figure S5, Supporting Information). The decreased expression of IFN‐γ in the Q‐VAX and YbgF‐BP‐Com1 immunized GPs suggests a potential outcome of expedited clearance of C. burnetii, resulting in a shorter duration of cytokine induction. This observation is further supported by the downregulation of IL‐12, a pro‐inflammatory cytokine known to stimulate the production of IFN‐γ by T cells. The IL‐8 response in the Q‐VAX and YbgF‐BP‐Com1 challenged groups indicate an inflammatory response. IL‐8, a neutrophil chemotactic factor that recruits neutrophils, was significantly elevated in YbgF‐BP‐Com1 vaccinated GPs post‐infection compared to that in the uninfected group. These findings are supported by studies that have shown an increased frequency of granulocytic cells in animals vaccinated with Coxevac, an inactivated whole cell Q fever veterinary vaccine, and the involvement of neutrophils in vaccine‐induced protection against Q fever in mouse models.^{[ 30 ]} Nevertheless, future multiple timepoint assessments of differential cytokine expression will generate more conclusive data.

Several experimental vaccines developed for Q fever can be classified as whole cell extracts and recombinant protein vaccines. Whole‐cell extracts despite providing significant protection as characterized by the reduced bacterial burden in tissues induce reactogenicity when tested on sensitized GPs.^{[ 29} , ^{31 ]} In contrast, recombinant proteins, when formulated with adjuvants such as TLR agonists augment immunogenicity with less reactogenic responses.^{[ 32 ]} Our subunit vaccine, apart from being a potent inducer of humoral immune responses, did not induce a reactogenic responses as shown by no adverse reactions after three immunizations. When compared with the only alum group, YbgF‐BP‐Com1 mediated a significant reduction in bacterial burden in the liver, which was not observed in the Q‐VAX group (Figure 6f). The role of antibodies in the protection against C. burnetii has not been investigated in detail, as cell‐mediated immunity is considered crucial for its clearance. Antibody‐mediated immunity is extracellular and includes opsonization, complement‐mediated killing, and antibody‐dependent cellular cytotoxicity (ADCC).^{[ 33 ]} These functions aid in controlling C. burnetii for the duration of its extracellular phase following inhalation and prior to evading immune cells and residing as an intracellular pathogen. Considering this initial extracellular growth phase, YbgF‐BP‐Com1 ability to induce high titer of specific antibodies suggest its potential for protection against early stages of infection by C. burnetii. However, cellular immunity is required for clearance and controlling C. burnetii once it invades the cells.^{[ 26} , ^{34 ]} As BP‐based vaccines were previously shown to induce both humoral and T cell immune responses, these immunological properties might aid in prevention of C. burnetii infection. Since YbgF‐BP‐Com1 induced protective immunity and reduced febrile responses and pathogen burden in the liver, it likely engaged T‐cell immune responses.

Assessment of C. burnetii burden in GP organs confirmed their dissemination to systemic organs (Figure 6f; Figure S4, Supporting Information). It can be noted that GPs from the alum, Q‐VAX and YbgF‐BP‐Com1 groups with increased C. burnetii burden in lungs in general developed higher body temperatures post‐infection when compared to GPs with less burden. GPs with increased pathogen burden in all four organs (kidney, liver, lungs and spleen) were likely to develop fever and with increased duration (Figure S4b, Supporting Information). In the YbgF‐BP‐Com1 group, the two GPs with C. burnetii burden in all four organs developed only borderline fevers. This suggests that the significant humoral responses induced by YbgF‐BP‐Com1 have played a role in the reduction of clinical disease, as was previously demonstrated in mouse models.^{[ 26 ]} Hepatitis is a frequent presentation of acute Q fever pathology and histological analysis revealed “donut‐like granulomas” with minimal necrosis. The fact that our vaccine demonstrated reduced C. burnetii loads in the liver suggests the ability to inhibit systemic dissemination of C. burnetii.^{[ 35 ]} No significant clearance was seen in the kidney and lungs with any of the groups, which could be due to vaccine‐independent clearance at the day of organ collection (day 26 post‐challenge) as further supported by low bacterial loads in the organs of GPs in the alum group. However, a more detailed analysis of cell‐mediated mechanisms is required to explain the protective mechanisms initiated by our BP vaccine candidate. In addition, the utilization of the GP model poses significant challenges owing to its outbred nature and the considerable heterogeneity observed among individuals within a group, as demonstrated by the differential C. burnetii loads within a group. In this study, we only explored the subcutaneous route of injection, other routes such as intranasal should be explored as C. burnetii is an air‐borne pathogen, and induction of mucosal immunity provides superior protection in comparison to other routes.^{[ 36 ]} The importance of mucosal immunity in protection against C. burnetii was previously investigated with promising results^{[ 37 ]} and this could be an area for future exploration in developing an effective Q fever vaccine. It should here be noted that BP vaccines coated with multiple Pseudomonas aeruginosa antigens prevented acute infection by P. aeruginosa when intranasally administered.^{[ 11a ]} This suggested suitability of intranasal administration for BP vaccines.

In conclusion, the two C. burnetii antigens YbgF and Com1 attached to BPs were safe and immunogenic inducing immune responses correlating with protective immunity. YbgF‐BP‐Com1 might be developed into a stable Q fever subunit vaccine that can be cost‐effectively manufactured at scale to control this infectious disease.

4. Experimental Section

Bacterial Strains and Growth Conditions

All bacterial strains and plasmids used in this study are listed in Table S1 (Supporting Information). For plasmid propagation, E. coli XL1‐Blue strain (San Diego, CA, USA) was grown in Luria Broth (LB) medium (Difco, Detroit, MI) at 37 °C with the appropriate selection antibiotic ampicillin (Amp), 100 µg/mL). For BP production, E. coli strain Clear Coli (Lucigen, Middleton, WI, USA), an endo‐toxin‐free variant, was grown in LB miller medium (LB supplemented with 1% Sodium Chloride (NaCl)) with selection antibiotics (100 ug mL⁻¹ Ampicillin, 50 ug mL⁻¹ Chloramphenicol). Primers for sequencing were synthesized by Integrated DNA Technologies (IDT, Coralville, IA, USA).

Antigen Selection

For the T‐cell epitope vaccine, HLA Class I and Class II epitopes from immunodominant antigens of C. Burnetii were selected based on their immunogenicity in mice and humans.^{[ 12 ]} For the YbgF‐BP‐Com1 vaccine, the complete protein sequence of antigens ComQ1 (CBU_1910) and YbgF (CBU_0092) was obtained from the NCBI database.

Molecular Cloning for Production of BP Vaccines

Cloning techniques were done as previously described.^{[ 38 ]} E. coli codon‐optimized DNA fragments purchased from Biomatik (Kitchener, ON N2C 1N6, Canada) were used to construct two plasmids in this study: (1) pET14b‐COX‐PhaC and (2) pET14b‐YbgF‐PhaC‐Com1. The DNA fragments (Biomatik) were restriction digested by XbaI, NotI, BamHI, XhoI depending on the gene location, and fragment separation was done using agarose gel electrophoresis with GelRed stain (Biotium, USA). Subsequently, the target bands were excised and subjected to gel purification (New England Biolabs, USA). The targeted bands were ligated to the pET14b vector with T4 DNA ligase to generate the recombinant plasmids whose sequence was confirmed by DNA Sequencing Center (DE3) (Lucigen, USA) (Griffith University, Nathan, Australia). The successful plasmid was transformed to a chemically competent production host E. coli strain Clear Coli for the respective BP production.

Production and Purification of BP Vaccines

E. coli strain Clear Coli BL21 (DE3) (Lucigen) production hosts harboring pMCS69 plasmid containing recombinant fusions were grown in Bio‐Flow 32 bioreactor (Eppendorf Germany).^{[ 39 ]}

BP harboring cells were centrifuged at 8000 x g for 15 min at 4 °C to collect the sediment. Cells were mechanically disrupted using a microfluidizer M‐110P (Microfluidics, USA). The cell lysate was centrifuged at 8 000 x g at 4^°C for 20 min to collect the BPs. BPs were washed to remove cellular debris.¹⁶ Purified sterile BPs were stored at 4  °C.

Characterization of Antigens Expressed on BPs

Detection and analysis of antigens expressed on BP were done by SDS‐PAGE as described elsewhere.^{[ 40 ]} The target protein was excised and amino acid sequence confirmed by Q‐TOF‐MS (Mass Spectrometry Facility, University of Queensland, Australia).The antigen concentration expressed on the BP particles was determined by densitometry using bovine serum albumin (BSA) standards ranging from 62.5 ng to 500 ng as previously described.^{[ 14a ]} The images were captured using gel doc (BioRad Laboratories, Hercules, CA, USA) and analyzed with Image Lab software (BioRad Laboratories, Hercules, CA, USA).

Characterization of BP Vaccines

Purified BPs and Whole cells were characterized by Transmission Electron Microscopy (TEM) for morphology analysis to confirm the production, shape, and size of biopolymers produced. Purified BPs were analyzed for particle size distribution and surface charge (Zeta potential) using LiteSizer 500 (Anton Paar) at 25  ^°C. Purified BPs were diluted 1000‐fold in Tris buffer (pH 7), sonicated for 2 min and particle size was evaluated by Dynamic light scattering (DLS) analysis. The zeta potential was measured by electrophoretic light scattering coupled with phase analysis light scattering. Three technical replicates were performed for each vaccine. PHB content analysis was performed based on method described by Karr et al.^{[ 41 ]} Briefly, BPs were freeze dried and weighed. The dry BPs were boiled in 1 mL of concentrated sulphuric acid for one hour at 90 ^°C for acid hydrolysis to crotonic acid. The samples were cooled before addition of water and filtered using a 0.22 µm filter (Merk Millipore). Diluted samples (1:500) were analyzed for PHB concentration by high‐performance liquid chromatography (HPLC) using an Agilent 1200 HPLC system. Crotonic acid was used as a standard and PHB %w/w was calculated using the amount of crotonic acid produced in the sample.

Stability Assessment of BP Vaccines

The stability of BP vaccines at different temperatures (4, 25, 37, and 56 ^°C) was assessed by Enzyme‐linked immunosorbent assay (ELISA) to confirm the retention of antigenicity; DLS to assess the size and charge variation, and SDS‐PAGE to assess the retention of antigens following temperature treatment over 28 days. ELISA was performed with the temperature‐treated BPs coated at 5 µg mL and tested against the serum of immunized female GPs (serum with the highest EC50). ELISA procedure is described in the section below. The final antibody titers were represented as EC50s.

Animal Ethics Statement

Animal experiments were approved by the Griffith University animal Ethics committee (Queensland, Australia), with ethics numbers GRIDD/06/21AEC and GRIDD/02/20AEC.

Vaccine Formulation and Immunization

Vaccines were formulated with or without the adjuvant, Aluminum hydroxide (Invivo Gen, USA), with 20 µg of antigen in a final volume of 100 µL in sterile tris‐buffered saline at pH 7.5. Empty BPs and aluminum hydroxide alone served as respective controls. Three animal trials with two animal models were performed.

Animal Trials

In the first trial, IMVS‐colored female GPs were used. GPs were acclimated for at least a week before interventions. GPs were housed five per cage and administered with high fiber pellets (Mi‐Feed premium micronized grains: Mi‐Hobbi Farm mix), Barley Hay, and vitamin C‐supplemented water. Group numbers were equaled to five apart from the alum group which had only two GPs. In this trial, GPs were immunized subcutaneously with and without adjuvant. The booster vaccinations were administered after 2 and 14 ‐week intervals post first vaccination. Blood was collected through lateral saphenous vein puncture on Days 0, 14, and 72 for immunogenicity assessment. Challenge infection was performed on day 14 post last booster by intranasal injection of 1×10⁶ CFU/ml of Coxiella Burnetii Nine‐mile strain.

In the second trial 4–6‐week‐old female C57BL/6 mice were vaccinated subcutaneously with 20 ug of antigen with adjuvant. Boosters were administered after 2‐week intervals and blood was collected on Days 0, 21, and 48 through submandibular bleeds. All mice were humanely euthanized on day 48 following blood collection via cardiac puncture. No challenge studies were performed for this trial.

For the third trial IMVS male GPs were used. The housing conditions were kept similar to the first trial. Group numbers were equaled to eight GPs per group including the alum control. Positive control with Q‐VAX was included. The immunization schedule, and the route of administration were similar to the first GP trial. However, challenge infection was performed on day 26 post last booster.

Immunogenicity Assessment–Blood Collection

Whole blood was collected at each time point by saphenous vein puncture, centrifuged at 1 000 g for 10 min, and serum stored at −20 ^°C until analysis. Serum was stored at −80 ^°C for long‐term storage.

Enzyme‐Linked Immunosorbent Assay (ELISA)

GP serum collected post‐vaccination was assessed for total IgG against antigens by ELISA. Briefly, high‐binding ELISA plates (Nunc MaxiSorp) were coated with 100 µl of 5ug mL⁻¹ of respective vaccines: BP, YbgF‐BP‐Com, BP‐Com, BP‐YbgF, BP‐COX diluted in PBS buffer (pH 7.5) and incubated overnight at 4 ^°C. The plates were blocked with 200 µl of 3% skimmed milk in PBST (PBS; 0.05% (v/v) Tween 20: pH 7.5) at 37 ^°C for one h. After washing three times with PBST, serum from individual animals was added at 1:200 dilution, followed by twofold serial dilution, and incubated at 37 ^°C for one h. Following three washes with PBST, plates were incubated with HRP‐conjugated secondary anti‐GP IgG (AB6908, Abcam) diluted at 1: 10 000 for one h at 37 ^°C. After three washes, SIGMAFAST OPD (o‐Phenylenediamine dihydrochloride) tablets (Sigma‐Aldrich) were added and incubated for 15 min in dark. Plates were read at 450 nm using BioTek Synergy 2 microplate reader (BioTek). Serum titers were represented as Effective concentration (EC50). The EC50 values were calculated as described elsewhere.^{[ 42 ]}

Antibody avidity was measured by a modified ELISA with potassium thiocyanate (Sigma‐Aldrich) as a chaotropic agent as described elsewhere.^{[ 43 ]} Briefly, ELISA plates were coated overnight at 4 °C with 5 ug mL⁻¹ of plain BPs, YbgF‐BP‐Com1, and Q‐VAX respectively. The plates were blocked with 3% skim milk. Following three washes with PBST, serially diluted serum from individual animals was added to each well and incubated at 37 ^°C for one h. After three washes of 100 uL of 1.5 M, Potassium thiocyanate (KSCN) was added to one half of the plate and PBS was added to the other half. Following 15 min incubation at 37 °C, plates were washed three times with PBST and incubated with HRP‐conjugated secondary anti‐GP IgG or secondary goat anti‐mouse IgG/IgG1/IgG2c (anti‐GP ab6908; anti‐mouse IgG ab205719; anti‐mouse IgG1 ab97240; anti‐mouse IgG2c ab97255; Abcam) diluted at 1: 10  000 for one h at 37 ^°C. After three washes, SIGMAFAST OPD (o‐Phenylenediamine dihydrochloride) tablets (Sigma‐Aldrich) were added and incubated for 15 min in dark. Plates were read at 450 nm using BioTek Synergy 2 microplate reader (BioTek). The avidity index (AI) was calculated for each serum dilution complying with OD _+KSCN < OD _‐KSCN. The avidity index was calculated as described elsewhere.^{[ 29 ]}

Immunoblot

Immunoblots were performed to confirm the antigen‐specific IgG responses, Pooled sera from Day 42 were tested against BP vaccines. First, antigens were separated by SDS‐PAGE and transferred to the Nitrocellulose membrane (ThermoFisher Scientific, USA), using iBlot 2 Drying Blotting System (Invitrogen, USA). The membrane was blocked with 5% BSA for one h at room temperature. Following three washes with TBST (Tris‐buffered saline: 0.1% Tween pH 7.5), pooled serum samples from immunized animals were diluted 1:2 000 with TBST and incubated with the membrane at 25 ^°C for one h. After three washes with TBST, membranes were incubated with secondary goat anti‐GP IgG HRP or secondary goat anti‐mouse IgG/IgG1/IgG2c (anti‐GPAB6908; anti‐mouse IgG ab205719; anti‐mouse IgG1 ab97240; anti‐mouse IgG2c ab97255; Abcam) diluted at 1:10 000 for one h at 25 ^°C. Super‐Signal West Pico Stable Peroxide Solution and Super Signal West Pico Luminol/Enhancer Solution (Thermos Scientific, USA) were mixed for 5 min at 25 ^°C for signal development. Immunoblots were imaged using Odyssey (LI‐COR Biosciences, USA).

Guinea pig Challenge Study

One week before challenge temperature transponders (BMDS, IPTT‐300, Yuasa BioSystems Co Ltd., Okayama City, Japan) were implanted subcutaneously into the dorsal interscapular region of GPs. Temperatures were read using a reading wand (BMDS Reader model DAS‐8027‐IUS). On the day of the challenge Immunized GPs were lightly sedated with a ketamine (10–20 mg k⁻¹g)/xylazine (1–2 mg k⁻¹g) combination subcutaneous injection and, while in dorsal recumbency, intranasally infected with 10⁷ C. Burnetii Nine‐mile strain Phase I. Body weights, body temperatures, and any behavioral changes were observed for three weeks post‐challenge. Any severely affected GPs were humanely euthanized by lethal injection of pentobarbitone.

Quantification C. Burnetii DNA in Guinea pig Tissues

A section of the spleen, lungs, kidney, and liver were collected from male GPs thirty days post‐challenge to quantify C. burnetti DNA by Real‐time PCR. The tissues were weighed to a final weight of 0.5 g ± 0.1 and homogenized in 2 mL of PBS. DNA extractions were performed using Dneasy Blood and Tissue Kit (Qiagen, Hilden, Germany) according to the manufacturer's protocol. Purified genomic DNA was amplified using a duplex qPCR assay with C. Burnetii‐specific Com1 primers and probe (F5’‐AAAACCTCCGCGTTGTCTTCA‐3′, R5’GCTAATGATACTTTGGCAGCGTATTG‐3′probe5’‐/FAM/AGAACTGCC/ZEN/CATTTTTGGCGGCCA/3IABkFQ/−3′) (Integrated DNA Technologies, IDT, USA). GP glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH) was used as the housekeeping gene. (F 5′ ‐AATGGGAAGCTCACAGGTATGG‐3′, R 5′ ‐ATGTCATCGTATTTGGCCGGT‐3′, probe 5′‐/5Cy5/TCCAGGCGG/TAO/CAGGTCAGATCCACA/3IAbRQSp/−3′). The qPCR was conducted on MIC Realtime PCR platform under the conditions 50 ^°C for 2 min, 95 °C for 2 min, and fast cycling of 95 °C for 10 s and 60 °C for 20 s with fluorescence acquisition for 45 cycles. Results were presented as copy numbers of the Com1 gene against GAPDH.

Gene Expression Analysis

Spleens were collected from GPs 26 days post challenge to measure the gene expression of cytokines using qPCR. RNA isolation was performed using RNAeasy kit (Qiagen, Germany) according to manufacturer's instructions. RNA purity and quantification was determined by NanoDrop (Isogen Life, Netherlands). cDNA was synthesized using SensiFAST cDNA synthesis kit (Meridian Bioscience, TN). The resulting cDNA was diluted 1/100 for qPCR reactions. The primer sets for amplification of housekeeping genes and cytokines were designed using NCBI Primer‐Blast tool across intron‐exon boundaries to exclude DNA amplification and are listed in Table S5 (Supporting Information). qPCRs were performed using SYBR (PowerUp SYBR) Green PCR master mix (ThermoFisher Scientific). Three genes (GAPDH, ACTB, ATP5B) were selected as the reference genes. The gene expression fold change was evaluated using the common base method,^{[ 44 ]} which takes into account the primer efficiency. The results were presented as fold change normalized against all three reference genes.

Statistical Analysis

All statistical analyses were performed using Prism version 8 (GraphPad) or GenStat (22nd Edition, VSN International Ltd). Statistical differences were assessed using one‐way analysis of variance (ANOVA) with Dunnett's correction for multiple comparisons. Two‐tailed Mann‐Whitney T‐test (unpaired) or Wilcoxon t‐test (paired) were performed where appropriate. No adjustments were made for comparisons in the female GP trial due to the small group size involved. Statistical significance with weight data was determined using a restricted maximum likelihood linear mixed model (REML) incorporating repeated measures. A p‐value of <0.05 was defined as statistically significant. (*p < 0.05, ** p < 0.01, *** p < 0.001).

Conflict of Interest

The authors declare the following competing financial interest(s): B.H.A.R. is a cofounder and shareholder of PolyBatics Ltd. The remaining authors declare no conflict of interest.

Author Contributions

G.S. J.S., S.R.G. and B.H.A.R. conceived the study and designed the animal trials. B.H.A.R. and G.S. designed the vaccines and most experiments. G.S. manufactured and characterized the vaccine candidates. G.S. conducted most of the animal trials. S.C. A.I., K.P. and M.W. assisted with animal trials. K.P. and G.S. analyzed the bacterial burden in organs and cytokine production. G.S. did the stability analysis and conducted most of the immune response analysis work. B.H.A.R. and G.S. wrote the manuscript. G.S., S.R.G, J.S., B.H.A.R. and I.M. coordinated the animal trials. All authors contributed to editing of the manuscript.

Supporting information

Supporting Information

Sam G., Plain K., Chen S., Islam A., Westman M. E., Marsh I., Stenos J., Graves S. R., Rehm B. H. A., Synthetic Particulate Subunit Vaccines for the Prevention of Q Fever. Adv. Healthcare Mater. 2024, 13, 2302351. 10.1002/adhm.202302351

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.