Optimal protection against Salmonella infection requires noncirculating memory

Abstract

While CD4 Th1 cells are required for resistance to intramacrophage infections, adoptive transfer of Th1 cells is insufficient to protect against Salmonella infection. Using an epitope-tagged vaccine strain of Salmonella, we found that effective protection correlated with expanded Salmonella-specific memory CD4 T cells in circulation and nonlymphoid tissues. However, naive mice that previously shared a blood supply with vaccinated partners lacked T cell memory with characteristics of tissue residence and did not acquire robust protective immunity. Using a YFP-IFN-γ reporter system, we identified Th1 cells in the liver of immunized mice that displayed markers of tissue residence, including P2X7, ARTC2, LFA-1, and CD101. Adoptive transfer of liver memory cells after ARTC2 blockade increased protection against highly virulent bacteria. Taken together, these data demonstrate that noncirculating memory Th1 cells are a vital component of immunity to Salmonella infection and should be the focus of vaccine strategies.

Keywords: CD4 T cell; Salmonella infection; protective immunity; tissue-resident memory; vaccines.

Fig. 1.

Immunization with LVS Salmonella confers long-lasting protection against Salmonella infection. (A and B) Mice were immunized with 2.5 × 10⁶ cfu/mouse intravenously (BRD509-2W1S). Mice were challenged orally 45 d later with 1 × 10⁹ cfu/mL SL1344 via water bottle exposure. Mice were euthanized on day 5 post challenge, and bacterial burdens were analyzed in spleen and liver. Data show mean ± SEM of two replicative experiments with three to six mice per group. (C and D) Mice were immunized with 5 × 10⁵ cfu/mouse intravenously (BRD509-2W1S), and on indicated days later mice were challenged with 1 × 10⁹ cfu/mL SL1344 via water bottle exposure. Mice were euthanized on day 5 post challenge, and bacterial burdens were analyzed in the spleen and liver. Data show mean ± SEM with four to eight mice per group per time point. ****P < 0.0001.

Fig. 2.

Circulating and resident Salmonella-specific CD4 memory T cells are generated after LVS immunization. (A) Generation of circulating tetramer+ CD4 T cells by LVS immunization. Blood and spleen samples were collected from naive and LVS immunized mice (2.5 × 10⁶ cfu/mouse intravenous BRD509-2W1S) 42 d after immunization. Lymphocytes were isolated, stained, and tetramer-positive cell events were quantified. Data show mean ± SEM of at least four mice per group. **P < 0.01. (B) Representative flow plot and quantitation of IFN-γ⁺T-bet^high CD4 T cells from mice immunized as in A. Data show mean ± SEM of pooled data from two replicative experiments. ****P < 0.0001. (C and D) Representative flow plots and pooled analysis from LVS-immunized mice (2.5 × 10⁶ cfu/mouse i.v.). On day 45 after immunization, mice were administered anti-CD45.2 antibody (3 µg/mouse) 3 min before euthanasia. Organs were harvested and digested before lymphocytes were stained and quantified. Data show mean ± SEM with six mice per group.

Fig. 3.

LVS immunization induces noncirculating Salmonella-specific memory CD4 T cells. Congenically marked CD45.2 mice previously immunized intravenously with 2.5 × 10⁶ cfu/mouse were surgically joined to CD45.1 mice for 28 d. Mice were separated, and tissues were harvested 14 d later. (A) Tetramer numbers in spleens from naive, naive parabiont, and LVS-immunized parabionts. Splenocytes were compared with naive and LVS-immunized parabionts for presence of the CD45.2 congenic marker. (B) Tetramer numbers in spleen (Left) and liver (Right) from naive and LVS-immunized parabionts. Tissue lymphocytes were compared between naive and LVS-immunized parabionts for presence of the CD45.2 congenic marker. Data show mean ± SEM from three mice per group. *P < 0.05.

Fig. 4.

Both tissue-resident and circulating memory are required for optimal protective immunity against Salmonella infection. (A) Congenically marked CD45.2 mice were immunized and boosted with 2.5 × 10⁶ cfu/mouse, and 60 d later were surgically joined to CD45.1 mice for 30 d. Mice were separated and challenged intravenously 14 d later with 1 × 10³ cfu/mouse SL1344. Mice were then euthanized 6 d post challenge. Spleens (B) and livers (C) were harvested, homogenized, diluted, and plated to determine bacterial burden in each tissue. Data show the mean of two replicative experiments ±SEM with five to seven mice per group. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 5.

CD69^Hi Th1 cells in the liver display markers of tissue residence. (A–H) IFN-γ-eYFP mice were immunized intravenously with 200 cfu of TAS2010, and 12 wk later spleens and livers were assessed by flow cytometry. (A and B) Representative plots and pooled percentages of CD69+ eYFP+ CD4+ T cells from immunized mice. (C–H) FACS plots comparing the expression of P2X7, ARTC2, CD101, LFA-1, CD103, and KLRG1 by eYFP+ CD69^hi (solid black line) and eYFP+ CD69^lo (gray shading) CD4+ T cells in the liver 12 wk after TAS2010 immunization. Data are representative of pools of (A–H) two independent experiments. (B) n = 8. (C–H) n = 6 per group. **P < 0.01, ****P < 0.0001.

Fig. 6.

Liver-associated IFN-γ⁺ CD4⁺ T cells protect against SL1344 infection. (A–C) Survival of Rag2^−/− × Il2rg^−/− mice receiving 5 × 10⁷ splenocytes or 1 × 10⁷ lymphocyte-enriched liver cells from C57BL/6 mice that were previously infected for 12 wk and received 50 µg S+16a or PBS by intravenous injection 15 min before harvest. (A) Recipients were challenged intravenously with 200 cfu SL1344 24 h after adoptive transfer. (B) At the time of cell transfer, recipients of S+16a-treated liver cells received intraperitoneal injections of 250 µg of isotype control or anti-CD4 (GK1.5) or 200 µg anti–IFN-γ (HB-170-15), and 20 h later, recipients were challenged intravenously with 200 cfu SL1344. Antibody treatments were repeated intraperitoneally every 3–4 d. (C) Distribution of CD69⁺ eYFP⁺ CD4⁺ T cells in Rag2^−/− × Il2rg^−/− mice that had been adoptively transferred with liver cells from C57BL/6 mice infected for 12 wk that received 50 µg S+16a or PBS. Data in A and B are representative of three independent experiments. Data in C are representative of two independent experiments. (B) n = 7–8. (C) n = 7–12 per group.