Hypothesis / aims of study
The increasing rate of resistance of urinary tract infection (UTI) against front-line antibiotics requires the advent of next-generation treatment strategies. The development of vaccines against uropathogenic bacteria has been difficult, in part because the genetic diversity of bacteria means there are many potential target antigens and little way to know a priori which will prove effective at inducing protective immunity. The painstaking process of selecting appropriate antigens could be avoided with whole-cell bacteria; however, whole-cell formulations typically fail to produce long-term and durable immune responses. These complications are one reason why vaccines against uropathogenic E. coli (UPEC), the primary causative agent of UTI, have yet to be successfully clinically translated. It is our hypothesis that the immunogenicity of UPEC whole-cell vaccines will be enhanced by forming a slow-releasing depot through biomimetically mineralizing bacteria within a metal coordination polymer called a Zeolitic Imidazole Framework (ZIF). This works tests the application of ZIFs towards the production of inactivated, ZIF-encapsulated whole-cell UPEC vaccines and evaluates their efficacy against fatal sepsis in a mouse urosepsis model.
Study design, materials and methods
ZIF encapsulation was performed by adding 1 mg of CFT073 to 500 uL of 1.6 M 2-methylimidazole (HMIM), followed by the addition of 500 μL of 20 mM zinc acetate dihydrate (ZnOAC). The reaction was allowed to incubate at room temperature for 20 min. Bacterial viability following ZIF encapsulation was determined by CFU assay. Injection site residency studies were performed according to a previously published protocol (1) using CFT073 expressing the far-red fluorescent protein smURFP (sCFT). Briefly, we fed 10 BALB/c mice a non-fluorescent diet and injected with 100 µL of saline, or equivalent amounts of sCFT073 prepared as formalin-inactivated CFT073 (sCFT-fixed), or sCFT@ZIF. Fluorescence was assessed hourly for the first 24hrs and every 24h thereafter until the fluorescence decayed back to the level of saline-injected mice. Vaccination and urosepsis challenge experiments were performed as described in Mellata et al (2) with the following modifications. We inoculated mice with (i) CFT-fixed - formalin-fixed , (ii) CFT-heat - thermally inactivated, (iii) CFT@ZIF - ZIF inactivated bacteria, or (iv) saline. The vaccination schedule consisted of doses on days 0, 7, and 14. At day 21, mice were infected with a lethal dose of CFT073 (108 CFU, intraperitoneally) and monitored for 48 hours. Mice were euthanized when they became moribund or after 48 hours and blood, spleen, liver, kidney, lung, heart, and skin at the administration site were collected for histopathological analysis, and/or determination of endpoint immunoglobulin titers and bacterial CFUs. Endpoint immunoglobulin titers were measured by an enzyme-linked immunosorbent assay following a previously published method.78 Lyophilized CFT073 was used as the capture antigen. Bacterial load in the spleen, liver, and blood was determined by CFU assay.
Results
We identified suitable synthetic conditions to coat ZIF-8 on the surface of urosepsis isolate CFT073 (CFT@ZIF). We found that the crystalline layer is thermally stable but can be removed completely with EDTA (Fig 1A). We determined that the bacteria died within an hour of encapsulation, as regrowth did not occur following re-plating or inoculation in culture media for 72 hours. We injected mice with CFT@ZIF and performed a time-dependent fluorescence at the injection site and measured steady and extended-release of CFT@ZIF from the site of injection compared to unencapsulated control (Fig 1B, C). We then conducted vaccination and urosepsis challenge experiments comparing traditionally inactivated whole-cell formulations—either formalin fixation or thermal inactivation to ZIF-encapsulation. After 48 hours only the CFT@ZIF mice were alive and apparently healthy. Endpoint anti-CFT immunoglobin titers revealed that CFT@ZIF induced the strongest IgG response (Fig 1D). Histopathological analysis of the injection site showed no differences to that of control mice that received subcutaneous injections of saline (Fig 1E). All mice were given three doses in the same anatomical location over the course of a week, with no changes in tissue histology. This supports that ZIF-8 is not acutely toxic in mice. Quantification of bacterial CFUs in organs showed that the lowest bacteria count in the liver and spleen were in the CFT@ZIF-vaccinated mice (Fig 1F). IFN-y levels were slightly elevated vs fixed formulations (Fig 1G), possibly a result of greater LPS release in the ZIF formulation. This may account for why 6/7 CFT@ZIF vaccinated mice survived past the 48-hour observation period vs CFT-fixed (median 32h survival), CFT-heat-inactivated vaccinated (median 20h survival), or unvaccinated mice (median 16h survival) after infection (Fig 1H).
Interpretation of results
ZIF-8 is a non-toxic and biodegradable zinc-based crystalline material that fundamentally differs from existing slow-release agents. The injection of CFT@ZIF creates a long-lasting depot that slowly introduces antigens and enhances immune responses. CFT@ZIF showed significantly enhanced anti-CFT073 immunoglobulin production, significantly reduced bacterial CFUs in the blood, liver, and spleen all resulting in a conspicuous improvement in survivability in a fatal mouse model of urosepsis. These results suggest non-mutually exclusive hypotheses for the improved protection conferred by CFT@ZIF vaccination. One possible explanation is that a stronger antibody response arises from ZIF-formulated UPEC because of surface antigen structure, and therefore antigenicity is maintained within the ZIF matrix. Secondly, the extended-release of the ZIF-encapsulated CFT073 from the injection site may induce stronger memory responses.
Concluding message
In this work, we demonstrated that ZIF-8 effectively inactivated a urosepsis strain of UPEC, creating a persistent vaccine depot that considerably outperformed current whole cell inactivation methods in survival following lethal challenge in a sepsis model. Formulation of the inactivated bacteria does not use toxic compounds, involves the mixture of only three components, and the total reaction time is under 30 minutes. Importantly, our biomimetic growth strategy is likely translatable across different bacterial species, presenting an opportunity for a generalizable approach toward whole-cell bacterial vaccine formulation. Our ZIF-encapsulation approach may be an ideal platform for rapid vaccine production from patient-derived bacterial strains, opening the potential for production of patient-specific prophylactic and therapeutic vaccine formulations for those suffering from recurrent UTI or at increased risk for complicated UTI.