Complete Protection against Pneumonic and Bubonic Plague after a Single Oral Vaccination

Abstract

Background: No efficient vaccine against plague is currently available. We previously showed that a genetically attenuated Yersinia pseudotuberculosis producing the Yersinia pestis F1 antigen was an efficient live oral vaccine against pneumonic plague. This candidate vaccine however failed to confer full protection against bubonic plague and did not produce F1 stably.

Methodology/principal findings: The caf operon encoding F1 was inserted into the chromosome of a genetically attenuated Y. pseudotuberculosis, yielding the VTnF1 strain, which stably produced the F1 capsule. Given orally to mice, VTnF1 persisted two weeks in the mouse gut and induced a high humoral response targeting both F1 and other Y. pestis antigens. The strong cellular response elicited was directed mostly against targets other than F1, but also against F1. It involved cells with a Th1-Th17 effector profile, producing IFNγ, IL-17, and IL-10. A single oral dose (108 CFU) of VTnF1 conferred 100% protection against pneumonic plague using a high-dose challenge (3,300 LD50) caused by the fully virulent Y. pestis CO92. Moreover, vaccination protected 100% of mice from bubonic plague caused by a challenge with 100 LD50 Y. pestis and 93% against a high-dose infection (10,000 LD50). Protection involved fast-acting mechanisms controlling Y. pestis spread out of the injection site, and the protection provided was long-lasting, with 93% and 50% of mice surviving bubonic and pneumonic plague respectively, six months after vaccination. Vaccinated mice also survived bubonic and pneumonic plague caused by a high-dose of non-encapsulated (F1-) Y. pestis.

Significance: VTnF1 is an easy-to-produce, genetically stable plague vaccine candidate, providing a highly efficient and long-lasting protection against both bubonic and pneumonic plague caused by wild type or un-encapsulated (F1-negative) Y. pestis. To our knowledge, VTnF1 is the only plague vaccine ever reported that could provide high and durable protection against the two forms of plague after a single oral administration.

Fig 1. F1 capsule production.

(A) V674, V674pF1, and VTnF1 bacteria grown at 37°C were re-suspended in India ink and observed by microscopy. (B) To evaluate the stability of F1 production, isolated colonies of Y. pestis CO92, or Y. pseudotuberculosis V674pF1, and VTnF1 were obtained after three subcultures in vitro. Isolated VTnF1 colonies were also obtained by culturing an homogenate of Peyer’s patches taken from mice previously inoculated with VTnF1 (10⁸ CFU i.g.; noted "in vivo"). Serial dilutions of bacteria were tested using an F1-specific ELISA and the F1 index shown was calculated as 1/CFU yielding DO_450nm = 1 in the ELISA. The un-encapsulated V674 was used as a negative control. (C) F1 production by VTnF1 was compared to that of Y. pestis CO92 after growth at 28°C or 37°C, using an F1-specific ELISA, and CO92Δcaf was used as negative control.

Fig 2. In vivo dissemination of VTnF1 after oral vaccination.

Groups of mice were inoculated orally with the VTnF1 vaccine (10⁸ CFU) and were sacrificed at the indicated times to evaluate the VTnF1 loads in: (A) feces (two pellets/mouse), (B) Peyer’s patches (two patches/mouse), (C) the spleen (whole organ), (D) the liver (whole organ), and (E) mesenteric lymph nodes (all). Samples were minced and dilutions were plated on selective agar plates containing kanamycin to count colonies, with a detection limit of 10 CFU/sample. Shown are individual values from 7–14 mice per condition. The horizontal line indicates the median. The Mann Whitney test was used for statistical analysis: *: p ≤0.05, **: p <0.01, ***: p<0.001.

Fig 3. Humoral immune response induced by vaccination.

(A) The antibody production induced by VTnF1 vaccination (10⁸ CFU) was quantified at various days post vaccination using an ELISA measuring IgG directed against purified F1 antigen. Shown are the means ± sem of titers observed in two experiments performed during the first month post-vaccination (Exp.1, 7 mice) and during the six-months post-vaccination period (Exp.2, 14 mice). Mean IgG titers at each time point were compared statistically to that on day zero using the Mann-Whitney test, and all were significantly higher (p<0.001), except those on day 1. (B) The different Ig mouse isotypes present in sera were analyzed at sequential times. (C) The IgG response against antigens other than F1 was analyzed by ELISA against a Y. pestis CO92Δcaf sonicate, and (D) by western blotting. The Y. pestis CO92 wild type or CO92Δcaf (noted F1-) were used as source of blotted antigens and sera pooled either from naive mice or mice vaccinated 30 days earlier (noted VTnF1). The Caf1 band is indicated by an asterisk and the Caf1M band by a triangle. (E) The response against purified Yops was also analyzed by ELISA. Shown are results from 14 individual animals (dots), and group medians (horizontal line). The Mann-Whitney test was used for statistical analysis: ***: p<0.001.

Fig 4. Cellular immune response of vaccinated mice.

Splenocytes isolated from mice vaccinated orally with VTnF1 (10⁸ CFU) 30 days (dark red boxes), or 180 days (black boxes) earlier, or from unvaccinated (naive) mice (blue boxes), were stimulated in vitro with 5 μg/ml of either an antigenic preparation of Y. pestis CO92Δcaf (noted Y.p.F1^-) or purified F1 antigen, or the mitogen Concanavalin A (ConA, 1 μg/ml) as positive control. Supernatants taken three days after stimulation were tested for the presence of IFNγ (A), IL-17 (B), and IL-10 (C) by ELISA. Shown are the mean ± s.e.m. of 14 mice per condition (two pooled experiments). The unpaired Mann Whitney test was used for statistical analysis: *: p<0.05, **: p<0.01, ***: p<0.001, ns: not significant. Comparable results were obtained in two other experiments using naive and 30-days vaccinated mice (16 per group).

Fig 5. Protection against bubonic and pneumonic plague of mice vaccinated with VTnF1.

Mice having received a single oral dose of VTnF1 (10⁸ CFU) were challenged via i.n. or s.c. routes, with various doses of bacteria as indicated. (A-D) Animals were challenged four weeks after vaccination by injection of (A, B) Y. pestis CO92, or (C, D) the un-encapsulated CO92Δcaf. (E) Vaccinated mice were challenged six months after vaccination. Mouse survival was recorded daily for 21 days. The number of mice surviving / number of animals tested is indicated above the corresponding bar for each condition. The Fisher Exact test was used for statistical analysis: *: p≤0.05; ***: p<0.001.

Fig 6. Vaccination with VTnF1 prevents Y. pestis dissemination from the site of infection.

(A) Mice vaccinated with VTnF1 (10⁸ CFU i.g.) were infected four weeks later by s.c. injection of 10³ CFU of the bioluminescent Y. pestis CO92::Tn7-Pail-lux in the ventral skin. Whole body bioluminescence of the animals was recorded at regular time intervals using an IVIS camera. Shown are white light photographs merged with the bioluminescence signal, scaled in the right margin. (B) The bioluminescence intensity emitted at the site of injection was measured for each animal and shown are means ± s.e.m. for each group. Animals missing at time point 92h died from infection before this bioluminescence recording. The paired Mann-Whitney test was used for statistical analysis: **: p<0.01.