CD8(+) T cells specific to a single Yersinia pseudotuberculosis epitope restrict bacterial replication in the liver but fail to provide sterilizing immunity

Abstract

CD8(+) T cells use contact-dependent cytolysis of target cells to protect the host against intracellular pathogens. We have previously shown that CD8(+) T cells and perforin are required to protect against the extracellular pathogen Yersinia pseudotuberculosis. Here we establish an experimental system where CD8(+) T cells specific to a single model antigen are the only memory response present at time of challenge. Using mice immunized with a vaccine strain of Listeria monocytogenes that expresses secreted ovalbumin (Lm-OVA), we show that OVA-specific CD8(+) T cells are generated and provide limited protection against challenge with virulent OVA(+)Y. pseudotuberculosis. Perforin expression by OVA-specific CD8(+) T cells was required, as Lm-OVA-immunized perforin-deficient mice showed higher bacterial burden as compared to Lm-OVA-immunized perforin-sufficient mice. Surprisingly, antigen-specific T cell protection waned over time, as Lm-OVA-immune mice eventually succumbed to Yersinia infection. Kinetic analysis of infection in mice with and without OVA-specific CD8(+) T cells revealed that bacterial numbers increased sharply in OVA-naïve mice until death, while OVA-immune mice held bacterial burden to a lower level throughout the duration of illness until death. Clonal analysis of bacterial populations in OVA-naïve and OVA-immune mice at distinct time points revealed equivalent and severe bottle-neck effects for bacteria in both sets of mice immediately after intravenous challenge, demonstrating a dominant role for other aspects of the immune system regardless of CD8(+) T cell status. These studies indicate that CD8(+) T cells against a single antigen can restrict Y. pseudotuberculosis colonization in a perforin-dependent manner, but ultimately are insufficient in their ability to provide sterilizing immunity and protect against death.

Keywords: Bottle-neck; CD8(+) T cells; Ovalbumin; Perforin; Yersinia pseudotuberculosis; YopE.

Figure 1. YopE-OVA fusion protein is secreted by Yersinia, and is processed and presented in vivo to stimulate OVA-specific CD8⁺ T cell responses

(A) The indicated Y. pseudotuberculosis strains were grown in Yop secretion conditions (Materials and Methods) and processed for western blotting of whole cell lysates and supernatant proteins, with anti-ovalbumin polyclonal antisera. (B) ksgA⁻ strains expressing YopE₁₈::OVA or YopE₁₃₈::OVA were delivered to animals via intravenous inoculation and 10 days later, spleens removed, single cell suspensions generated and cells incubated with MHC Class I H2-K^b pentamers specific to OVA residues 257–264 and antibodies against CD8 and CD19. Shown are representative histograms showing pentamer and CD8 labeling of CD8⁺ CD19⁻ cells. (C) Immunization with L. monocytogeneses ΔactA ΔplcB ActA₁₀₀::OVA generates OVA-specific CD8⁺ T cells. 3 × 10⁷ CFU of noted L. monocytogeneses strains were delivered to animals via intravenous inoculation. 7 days later spleens were removed, single cell suspensions generated and cells were incubated with H2-K^b OVA_257-264 pentamers and antibodies against CD8 and CD19. Shown are representative scatter plots of 2 separated experiments showing pentamer and CD8 labeling of CD8⁺ CD19⁻ cells (n=3 mice per experiment).

Figure 2. OVA-specific CD8⁺ T cells are sufficient to protect against Y. pseudotuberculosis expressing a YopE-OVA fusion protein and required for Lm-OVA immune mice to reduce Yptb-OVA burden

Mice were immunized with 3 ×10⁷ CFU L. monocytogenes ΔactA ΔplcB (Lm, A, B) or the same strain carrying the integrated ActA::OVA fusion protein (Lm-OVA, A, B, C), then 60 days later, challenged with 100 CFU virulent Y. pseudotuberculosis expressing YopE_1-138::OVA (A, C), YopE_1-18::OVA (C) or the parental strain YPIII(pIB1) (Yptb, B, C). (A, B) Mice were either sacrificed at day 5 and bacterial burden per gram tissue of spleen and liver determined or (C) observed until day to death. Mice were immunized with either Lm or Lm-OVA and at days 60 post-immunization challenged with Yptb-OVA. At days −3, -1, +1, +3 and +5 relative to challenge, mice were intraperitoneally injected with 200μg anti-CD8 monoclonal antibody or PBS. At day 6 post-challenge, mice were sacrificed and bacteria burden per gram (D) spleen or (E) liver determined. Bars indicate median values.

Figure 3. Perforin is required for OVA-specific CD8⁺ T cells to protect against Y. pseudotuberculosis and OVA-specific CD8⁺ T cells are present in Lm-OVA-immunized PKO animals and increase after exposure to Yptb-OVA

C57BL/6 and PKO mice were immunized with Lm-OVA, then 60 days later, challenged with Yptb YopE_1-138::OVA (hereafter called Yptb-OVA). Mice were sacrificed on day 5 and the bacterial burden per gram tissue of (A) spleen and (B) liver determined. C57BL/6 and PKO mice were immunized with 3 × 10⁷ CFU Lm-OVA, then 60 days later a subset of the animals challenged with Yptb-OVA. 7 days post-challenge all animals were sacrificed, spleens harvested and single cell suspensions stained with pentamer specific to H2-K^b OVA_257-264 and antibodies against CD8 and CD19. Shown are the calculated percent (C) and number (D) of splenic pentamer⁺ CD8⁺ cells.

Figure 4. Lm-OVA-immune mice restrict Yptb-OVA burden early but eventually succumb to infection

C57BL/6 mice were immunized with 3 ×10⁷ CFU Lm or Lm-OVA, then 60 days later challenged with 100 CFU Yptb-OVA. Mice were sacrificed upon observation of morbidity and (A) day of death noted. Results shown are pooled from 3 separate experiments. Lm or Lm-OVA-immune mice were challenged with 150 clones of a Lm-OVA clone library, sacrificed at days 1, 7 and output bacterial burden (B, C) and clone numbers (D, E) in the spleen (B, D) and liver (D, E) determined. ND, not detected. Results shown are from two separate experiments.