Massive Number of Antigen-Specific CD4 T Cells during Vaccination with Live Attenuated Salmonella Causes Interclonal Competition

Mouse strains

SM1 Rag-deficient TCR-transgenic mice expressing the Thy1.1 or CD45.1 allele were produced by backcrossing the original C57BL/6 SM1 Rag-deficient line (25) to B6.PL-thy1a/Cy or B6.SJL-PtprcaPep3b/BoyJ mouse strains (The Jackson Laboratory, Bar Harbor, ME). OT-II TCR-transgenic mice (26) were generously provided by Dr. L. Lefrancois (University of Connecticut Health Center). C57BL/6 mice were purchased from the National Cancer Institute (Frederick, MD) and used at 8–16 wk of age. All mice were housed in specific pathogen-free conditions and cared for in accordance with the University of Connecticut Health Center and National Institutes of Health guidelines.

Adoptive transfer of SM1 T cells

Spleen and lymph node cells (inguinal, axillary, brachial, cervical, mesenteric, periaortic) were harvested from SM1 RAG-deficient, CD90.1 congenic (or CD45.1 congenic) TCR-transgenic mice, and a pooled single-cell suspension was generated. A small aliquot of cells was stained using Abs to CD4, CD90.1 (or CD45.1), and Vβ2 (BD Biosciences, San Diego, CA) and the percentage of SM1 T cells was determined by flow cytometry using a FACSCalibur (BD Biosciences). Cell numbers were adjusted accordingly and 2–5 × 10⁶ SM1 T cells were injected i.v. into recipient C57BL/6 or OT-II mice. In most experiments, SM1 cells were also stained with CFSE (27) before adoptive transfer.

Vaccination with live attenuated Salmonella or heat-killed S. typhimurium (HKST)³

Attenuated (AroA⁻D⁻) S. typhimurium (BRD509) (28) was provided by Dr. D. Xu (University of Glasgow, Glasgow, U.K.). Flagellin-deficient S. typhimurium (BC490) was generously provided by Dr. B. Cookson (University of Washington, Seattle, WA). Bacteria were routinely grown overnight in Luria-Bertani broth without shaking, and diluted in PBS after an estimation of bacterial concentration using a spectrophotometer. Aliquots of live bacterial suspensions were incubated at 75°C for 1 h to generate Heat-killed S. typhimurium (HKST). Live attenuated (5 × 10⁵) and heat-killed bacteria (1 × 10⁸), were administered i.v. into the lateral tail vein of mice. In all experiments, the live bacterial dose was confirmed by serial plating onto MacConkey agar plates. In preliminary experiments, the challenge doses chosen for this study were found to generate similar peak SM1 clonal expansion and were therefore chosen for further study. In some experiments, the recall response of SM1 T cells to flagellin peptide immunization was examined in immunized mice. Adoptively transferred and previously immunized mice were injected i.v. with 200 μg flagellin peptide and 25 μg LPS and clonal expansion of SM1 T cells was examined 3 days later.

Flow cytometric analysis

Spleen or lymph node cells were incubated on ice for 20–45 min in Fc block (spent culture supernatant from the 24G2 hybridoma, 2% rat serum, 2% mouse serum, and 0.01% sodium azide) in the presence of the relevant primary Abs. FITC-, PE-, CyChrome-, PE-Cy5-, allophycocyanin-, or biotin-conjugated Abs specific for CD4, CD11a, CD45.1, CD69, CD90.1, IL-2, IL-4, IL-10, TNF-α, IFN-γ, Vβ2, 3, 4, 5.1/5.2, 6, 7, 8.1/8.2, 8.3, 9, 10b, 11, 12, 13, 14, 17a, and isotype control Abs were purchased from BD PharMingen (San Diego, CA). Following staining, cells were analyzed by flow cytometry using a FACSCalibur. Data were analyzed using FlowJo software (TreeStar, San Carlos, CA).

Tracking SM1 T cells and Salmonella

At various times after infection, spleen or lymph node cells were harvested in Eagle’s Hanks’ amino acids medium (Biofluids, Rockville, MD) containing 2% FBS and 5 mM EDTA. Serial dilutions of each sample were plated onto MacConkey agar plates (Difco, Detroit, MI) at 37°C to determine bacterial colonization in each tissue. Cells from the same organ (5 × 10⁶/tube) were then processed and stained as described above.

In vivo cytokine stimulation and intracellular staining

For intracellular cytokine experiments examining SM1 T cells, mice were injected i.v. with 200 μg of flagellin peptide 427–441 (VQNRFNSAITNLGNT) (29) plus 25 μg of LPS (Sigma-Aldrich, St. Louis, MO). For intracellular cytokine experiments examining endogenous Salmonella-specific CD4 T cells, mice were injected i.v. with 1 × 10⁸ HKST. To examine cytokine production directly ex vivo, spleens from such injected mice were harvested at 2 h after injection. Spleen cells were rapidly surface stained, fixed with formaldehyde, permeabilized using saponin (Sigma-Aldrich), and stained intracellularly using the Abs described above.

Vaccination with live Salmonella increases the number of CD4 memory cells in the spleen

Although a number of studies have examined the clonal burst size of CD4 cells (12, 13, 14, 15, 16, 17, 18, 19, 20, 21), these have used viral or bacterial models where CD8 T cell expansion is likely to be required for protective immunity. It is possible that the CD4 burst size is far greater in models where CD4 T cells are the primary effector cell mediating pathogen clearance.

Vaccination of C57BL/6 mice with attenuated S. typhimurium provides protective immunity to challenge with virulent bacteria that is dependent upon CD4 T cells (24). In initial experiments, we noted that the total number of splenocytes increased markedly 1 wk after vaccination with attenuated Salmonella, largely due to an increase in the number of CD11b- and Gr1-expressing non-T cells (Fig. 1⇓A and data not shown). Although this splenomegaly decreased the percentage of CD4 and CD8 T cells (Fig. 1⇓B), the absolute number of splenic CD4 and CD8 T cells actually increased ∼2-fold in the infected spleen (Fig. 1⇓C). Phenotypic analysis of the CD4 population demonstrated that this 2-fold increase was due to a large increase in the number of cells with an activated phenotype, CD4⁺CD11a^highCD44⁺ (Fig. 1⇓D and data not shown). Furthermore, a sizable percentage of CD4 T cells from vaccinated mice produced IFN-γ and TNF-α directly ex vivo (Fig. 1⇓D). CD4 cells from vaccinated mice did not produce IL-4 or IL-10 (data not shown), although a small proportion produced IL-2 (Fig. 1⇓D). All cytokine production by CD4 cells was confined to the CD4⁺CD11a^high population (Fig. 1⇓E). A number of non-CD4, non-CD8 cells in vaccinated mice stained positive with all Abs, including isotype control Abs (Fig. 1⇓D). This nonspecific staining is likely to result from the increased number of CD11b⁺ cells in the spleen of vaccinated mice.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Vaccination with live Salmonella causes a large expansion of activated CD4 T cells in vivo. C57BL/6 mice were vaccinated with 5 × 10⁵ live attenuated Salmonella (vaccinated) or were left untreated (naive). One week later, spleen cells were stained and the total cell number (A), percentage of CD4 cells (B), total number of CD4 cells (C) were determined. Data show the mean ± SD of three mice per group and are representative of four individual experiments. D, The surface expression of CD11a and the production of intracellular cytokines were examined in CD4 cells from the spleen of naive and vaccinated mice. Data show representative plots of spleen cells from at least three individual mice. Numbers indicate the percentage of spleen cells within the boxed gate. E, CD11a and intracellular staining after gating on live CD4 T cells from naive or infected mice.

We examined the in vivo kinetics of the CD4⁺CD11a^high response. Vaccination with live Salmonella caused a transient increase in total splenic CD4 numbers that had largely dissipated 2 mo later (Fig. 2⇓A, total CD4 cells). The increase in CD4 numbers was due to a large expansion of the CD4⁺CD11a^high T cell population (Fig. 2⇓A, total CD11a⁺ CD4 cells). This kinetics of the CD4⁺CD11a^high response had three distinct phases: 1) an expansion phase between day 0 and days 7–14, when the absolute number of these cells increased ∼10-fold; 2) a plateau until days 21–28, when the number of these cells remained relatively constant at a high level; 3) a contraction phase after days 21–28, when the number of cells declined steadily (Fig. 2⇓A).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Kinetics of CD4 CD11a⁺ and IFN-γ⁺ expansion after vaccination with attenuated Salmonella. C57BL/6 mice were vaccinated with 5 × 10⁵ live attenuated Salmonella and the number of CD4 cells, CD11a⁺CD4 cells, and IFN-γ⁺CD4 cells (A) or number of viable Salmonella (B) determined in the spleen at various time points later. C, CD11a-positive and IFN-γ- positive CD4 cells expressed as a percentage of all CD4 cells in the spleen of infected mice. Data show the mean ± SD of two to four mice per time point.

We also examined the kinetics of the CD4⁺IFN-γ⁺ response, the magnitude and kinetics of which was slightly different (Fig. 2⇑A, total IFN-γ⁺ CD4 cells). An increased frequency of CD4⁺IFN-γ⁺-producing cells was detected as early as 3 days after live vaccination. This response peaked at around 1 wk and stayed relatively constant until 3 wk following vaccination, after which it disappeared completely. Interestingly, the kinetics of the IFN-γ-producing CD4 cell response closely mirrored the kinetics of the bacterial burden in the spleen (Fig. 2⇑B). Numbers of Salmonella increased during the first week, stayed at relatively high levels for 2–3 wk, and then rapidly disappeared.

Most CD4⁺CD11a^high T cells in the spleen are Salmonella specific

These data suggested that the Salmonella burden in the spleen could be directly responsible for the specific activation of a large number of CD4 T cells to produce IFN-γ ex vivo. However, even if this were true, it still would not account for the vast majority of the expanded CD4⁺CD11a^high population found in the spleen after live vaccination. IFN-γ-producing CD4 cells represent <20% of the total CD4⁺CD11a^high population during the response to live vaccination (Fig. 2⇑C). From these data, it was not possible to distinguish whether non-IFN-γ-producing CD4⁺CD11a^high cellsrepresented Salmonella-specific CD4 T cells that had not been recently activated in vivo, or were simply CD4⁺CD11^high T cells that were not specific for Salmonella.

To discriminate between these possibilities, we modified an in vivo peptide pulse assay that has been used to examine the cytokine production of TCR-transgenic CD4 T cells in vivo (30, 31). This assay typically involves the i.v. injection of peptide to synchronously activate TCR-transgenic T cells in vivo, followed by analysis of intracellular cytokine production ex vivo, only 2 h later. Although the Ag specificity of the endogenous CD4 response to Salmonella is unknown, we hypothesized that most target Ags would be present in a HKST Ag preparation. C57BL/6 mice were therefore vaccinated with live attenuated Salmonella and, 1 wk later, some mice were injected with HKST. Two hours after this HKST pulse, spleens were harvested and rapidly processed to examine intracellular cytokine production.

As in our previous experiments, an increased frequency of IFN-γ- and TNF-α-producing CD4 T cells was detected in vaccinated mice (Fig. 3⇓A, naive vs vaccinated). These cytokine-producing cells were only found in the CD4⁺CD11a^high population (Fig. 3⇓B). A very small increase in CD4 IFN-γ and TNF-α production was detected in naive C57BL/6 mice that had received the HKST pulse (Fig. 3⇓A, naive + HKST), and these cells also expressed CD11a^high. It is possible that these cells are cross-reactive CD4 memory cells specific for the endogenous bacterial flora, but we have not examined this hypothesis in detail. A massive increase in the fraction of CD4 T cells producing IFN-γ and the amount of IFN-γ per cell was detected in vaccinated mice that had been pulsed with HKST (Fig. 3⇓A, vaccinated + HKST). Again, this cytokine response only occurred in the CD11a^high population and actually accounted for the vast majority of all CD4⁺CD11a^high T cells (Fig. 3⇓B).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Pulsing infected mice with HKST allows the identification of endogenous Salmonella-specific CD4 T cells in vivo. A, Groups of C57BL/6 mice were vaccinated with 5 × 10⁵ live attenuated Salmonella and 1 wk later some mice were injected i.v. with 1 × 10⁸ HKST. Spleens were harvested 2 h later and intracellular production of IFN-γ and TNF-α by CD4 T cells was determined. Data show cytokine production among live spleen cells and are representative of two to three mice per group and four individual experiments. Numbers represent the percentage of spleen cells falling within the boxed gate. B, CD11a expression and cytokine production from the same groups after gating on CD4 T cells. C, Groups of C57BL/6 mice were vaccinated with 5 × 10⁵ live attenuated Salmonella and 3 wk later injected i.v. with 1 × 10⁸ HKST. Spleens were harvested 2 h later and intracellular production of IFN-γ and expression of TCR Vβ receptors were determined among CD4 T cells. IFN-γ-negative and -positive cells were defined using a boxed gate similar to that shown in A. Data show the percentage of CD4 cells expressing a particular Vβ chain after gating on IFN-γ-positive or -negative CD4 cells. D, Groups of C57BL/6 mice were vaccinated with 5 × 10⁵ live attenuated Salmonella and 1 wk later some mice were injected i.v. with 1 × 10⁸ HKST or 1 × 10⁸ HKLM. Spleens were harvested 2 h later and intracellular production of IFN-γ was determined. Numbers represent the percentage of spleen cells falling within the boxed gate.

It seemed likely that the HKST pulse assay allowed for the identification of Salmonella-specific T cells and that these cells accounted for a surprisingly high fraction of the CD4 population, in mice that were vaccinated with live Salmonella. However, it was possible that the response to the HKST pulse was either a nonspecific inflammatory response or a specific response to an unknown Salmonella superantigen (SAg). To address the latter concern, we analyzed the TCR Vβ repertoire of CD4 T cells producing IFN-γ and those not producing IFN-γ after the HKST pulse, as SAg responses involve the restricted activation of certain Vβ-expressing T cells (32). Although a 2-fold increase in Vβ9-, 12-, and 17a-expressing CD4 T cells was noted among IFN-γ-positive cells, there was no obvious bias in the Vβ repertoire that would indicate a SAg response (Fig. 3⇑C). To examine the effect of nonspecific bacterial inflammation on CD4 IFN-γ production, we pulsed vaccinated mice with HKST or heat-killed Listeria monocytogenes (HKLM). Although HKST induced CD4 IFN-γ production, pulsing vaccinated mice with HKLM did not increase the frequency of IFN-γ-producing CD4 cells detected in the spleen (Fig. 3⇑D and data not shown). We conclude that the in vivo pulse assay allows the detection of endogenous Salmonella-specific CD4 T cells in vivo and that these cells comprise a surprisingly large fraction of the CD4 population in vaccinated mice.

Live vaccination causes a massive increase in Salmonella-specific CD4 T cells but vaccination with HKST does not

We used the HKST pulse assay to examine the frequency and kinetics of the endogenous Salmonella-specific CD4 response. These data demonstrate that a huge clonal expansion of Salmonella-specific CD4 T cells occurred in the first week after live vaccination (Fig. 4⇓A, total IFN-γ⁺ CD4 cells). The number of these cells persisted at a very high level for 3–4 wk, after which there was a significant decline. Importantly, the kinetics of IFN-γ-producing cells closely mirrored the kinetics of the CD11a⁺ population in response to vaccination (Fig. 4⇓A). Indeed, it was possible to account for all of the increase in the CD4⁺CD11a^high population as being due to an increase in Salmonella-specific CD4 T cells (Fig. 4⇓B). At the peak of the response, Salmonella-specific CD4 T cells accounted for ∼50% of all CD4 T cells (Fig. 4⇓B). Among Salmonella-specific CD4 cells, ∼30–40% also produced TNF-α (Fig. 4⇓B) and ∼5% produced IL-2 (data not shown). No production of IL-4 or IL-10 by CD4 T cells was detected at any time point using this assay (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Massive expansion of Salmonella-specific CD4 T cells after vaccination with live Salmonella but not HKST. A, C57BL/6 mice were vaccinated with 5 × 10⁵ live attenuated Salmonella and injected with 1 × 10⁸ HKST at various time points later. Two hours after HKST injection, spleen cells were harvested and the number of CD4 cells, CD11a⁺ CD4 cells, and IFN-γ⁺ CD4 cells was determined. Data show the mean ± SD of two to four mice per time point. B, CD11a-positive and IFN-γ-positive CD4 cells expressed as a percentage of all CD4 cells in the spleen of infected mice. C, C57BL/6 mice were vaccinated with 1 × 10⁸ HKST and 11 days later some mice were injected with 1 × 10⁸ HKST to stimulate cytokine production. Two hours later, spleen cells were harvested and cytokine production of CD4 T cells was examined by intracellular staining. The number of CD4 cells, CD11a⁺ CD4 cells, and IFN-γ⁺ CD4 cells was determined. Numbers indicate the percentage of spleen cells within the boxed gate.

We examined whether this expansion of Salmonella-specific CD4 cells could be observed after vaccination with Salmonella Ags or whether it only occurred following vaccination with live bacteria. C57BL/6 mice were immunized i.v. with HKST and 6 and 11 days later, IFN-γ and TNF-α production of CD4 T cells was examined using the HKST pulse assay exactly as described above. CD4 T cells in the spleen of HKST-vaccinated mice did not produce large quantities of cytokines in response to the in vivo pulse (Fig. 4⇑C and data not shown)

Salmonella-specific TCR-transgenic cells do not develop into memory cells following vaccination with live Salmonella

We sought to examine the clonal expansion and development of Salmonella-specific memory CD4 cells in more detail using an adoptive transfer approach. We previously generated the SM1 TCR-transgenic mouse specific for Salmonella flagellin (25), and therefore decided to use this model to examine the CD4 response to live vaccination. A small population of SM1 T cells(CD4⁺CD90.1⁺) was detected in the spleen, mesenteric lymph nodes and axillary-brachial-linguinal (ABI) nodes of C57BL/6 mice after adoptive transfer (Fig. 5⇓). In response to live vaccination, SM1 T cells expanded in the spleen but not in the mesenteric lymph nodes or peripheral ABI lymph nodes (Fig. 5⇓). This restricted activation of SM1 T cells is reminiscent of our previous observation that SM1 T cell expansion is limited to the gut-associated lymphoid tissue following oral Salmonella infection (25) and likely reflects an intrinsic property of the immune response to Salmonella flagellin. However, the most surprising finding was that by 10 days after vaccination very few SM1 T cells could be detected in the spleen or any other lymphoid tissue examined (Fig. 5⇓). Therefore, despite the massive clonal expansion of adoptively transferred SM1 T cells at days 3–5, they are not found at an elevated frequency at later time points. Indeed, by day 10, frequencies of SM1 T cells typically drop below the low level detected in transfer only mice (Fig. 5⇓).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

SM1 T cells do not persist after live attenuated Salmonella vaccination. C57BL/6 mice were adoptively transferred with 2 × 10⁶ SM1 T cells and groups of mice were vaccinated the following day with 5 × 10⁵ live attenuated Salmonella or 1 × 10⁸ HKST. A, Spleen cells were harvested 5 and 10 days later and the percentage of SM1 T cells was determined. Plots are representative of three mice per group and three similar experiments. B, In another similar experiment, spleen, mesenteric lymph nodes, and ABI lymph nodes were harvested at various time points following vaccination with live Salmonella or HKST and the percentage of SM1 T cells in each tissue was determined by flow cytometry.

It was possible that the unusual inability to generate SM1 memory cells was an intrinsic property of SM1 T cells or of flagellin-specific T cells in general. We therefore attempted to generate memory SM1 T cells after immunization with HKST or flagellin peptide and LPS. After immunization with HKST or peptide/LPS, SM1 T cells expanded in the spleen, mesenteric lymph nodes, and ABI lymph nodes (Fig. 5⇑ and data not shown). By day 10 or 20 after immunization, an increased frequency of SM1 T cells was detected in all lymphoid tissues analyzed (Fig. 5⇑ and data not shown). In some experiments, the peak expansion of SM1 cells to HKST was greater than that of live vaccination (Fig. 5⇑B). However, this did not explain the ability of these cells to persist in vivo. The relative decrease from the peak frequency was always severalfold greater following live vaccination than after HKST vaccination. Thus, the inability to generate an elevated frequency of SM1 T cells is a specific property of live vaccination and is not due to an intrinsic defect in SM1 TCR-transgenic cells.

Vaccination with live Salmonella or HKST induces SM1 T cell populations with similar phenotypic and functional properties

Immunizing mice with live Salmonella or HKST generated a similar frequency of splenic SM1 cells on day 5, but a very different frequency by day 10. To understand this difference, we examined the phenotypic and functional properties of SM1 T cells at day 5. Analysis of later time points was difficult due to the rapid loss of the SM1 population in mice vaccinated with live Salmonella (Fig. 5⇑). On day 5, SM1 cells in the spleen of both groups of mice had undergone multiple rounds of cell division and increased surface expression of CD11a (Fig. 6⇓A) and CD44 (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

SM1 T cells are functionally responsive to peptide in mice administered live Salmonella. C57BL/6 mice were adoptively transferred with 2 × 10⁶ CFSE-labeled SM1 T cells and groups of mice were vaccinated the following day with 5 × 10⁵ live attenuated Salmonella or 1 × 10⁸ HKST. A, Spleen cells were harvested 5 days later and CFSE staining was examined by flow cytometry after gating on CD4⁺CD90.1⁺ SM1 T cells. Data are representative of three individual mice and two separate experiments. B, C57BL/6 mice were adoptively transferred with 2 × 10⁶ CFSE-labeled SM1 T cells and groups of mice were vaccinated the following day with 5 × 10⁵ live attenuated Salmonella or 1 × 10⁸ HKST. Five days after vaccination, some mice were injected i.v. with 200 μg of flagellin peptide and 25 μg of LPS (+ peptide), and some mice were left untreated (− peptide). Two hours after peptide injection, spleen cells were harvested and stained using various different Abs. Data show surface staining for CD69 and intracellular staining for IL-2 or IFN-γ after gating on splenic CD4⁺CD90.1⁺ SM1 T cells. Plots from individual mice are shown and are representative of two individual mice and two individual experiments. Numbers represent the percentage of SM1 T cells in the boxed gate. C, C57BL/6 mice were adoptively transferred with 2 × 10⁶ SM1 T cells and groups of mice were vaccinated the following day with 5 × 10⁵ live attenuated Salmonella or 1 × 10⁸ HKST. Two weeks later, groups of mice were injected i.v. with 200 μg of flagellin peptide and LPS. Three days later, spleen cells were harvested and the percentage of SM1 T cells was determined. Numbers represent the percentage of SM1 T cells in the spleen, as defined by the boxed gate.

Recent reports suggest that IFN-γ-producing CD4 T cells may be short-lived and unable to develop into memory T cells (33). It was possible that live Salmonella induce a more pronounced IFN-γ response in SM1 T cells, leading to their inability to survive at later time points. We therefore examined the cytokine production of SM1 T cells directly ex vivo after vaccination with live Salmonella or HKST. Five days after vaccination, SM1 T cells were uniformly CD69^low, IL-2⁻, and IFN-γ⁻ (Fig. 6⇑B). We next sought to examine the functional response of these SM1 T cells after in vivo stimulation with flagellin peptide and LPS. Five days after vaccination, transfer only, live Salmonella, or HKST mice were injected i.v. with peptide/LPS to activate SM1 cells. Intracellular cytokine production was examined 2 h later. SM1 T cells in transfer only, live Salmonella, or HKST mice increased surface expression of CD69 in response to in vivo peptide stimulation, indicating that all of these cells encountered peptide (Fig. 6⇑B). Approximately 20% of SM1 T cells transfer-only mice produced IL-2 in response to peptide stimulation, and this was increased in mice exposed to live Salmonella or HKST (Fig. 6⇑B). Very few naive SM1 cells in transfer-only mice produced IFN-γ in response to peptide stimulation (Fig. 6⇑B). A higher percentage of SM1 cells produced IFN-γ in mice administered live Salmonella or HKST, but there was no noticeable difference between the vaccination regimens (Fig. 6⇑B). These data indicate that SM1 T cells in mice immunized with HKST or live Salmonella are functionally similar and that excess IFN-γ production is unlikely to explain the loss of SM1 T cells observed at later time points.

Next, we examined whether the few SM1 T cells that remain at later time points in mice vaccinated with live Salmonella are actually functionally responsive. C57BL/6 mice were adoptively transferred with SM1 T cells and vaccinated with HKST or live Salmonella or were left unimmunized (transfer only). Two weeks later, some mice were immunized i.v. with flagellin peptide/LPS and clonal expansion of SM1 T cells was examined 3 days later. A small population of SM1 T cells was detected in the spleen of transfer-only mice, and these cells expanded in response to peptide/LPS stimulation (Fig. 6⇑C). In agreement with previous experiments, HKST-immunized mice had a higher frequency of SM1 T cells in the spleen than transfer-only mice and these cells also expanded in response to peptide/LPS challenge (Fig. 6⇑C). Although live Salmonella-vaccinated mice had a very low frequency of SM1 T cells in the spleen, these cells expanded after exposure to peptide/LPS. Therefore, the few SM1 T cells remaining in the spleen of mice administered live Salmonella are functionally responsive.

Endogenous Salmonella-specific T cells prevent the development of SM1 memory cells

Our data suggested that a massive expansion and maintenance of endogenous Salmonella-specific T cells occurs in response to live vaccination. In contrast, SM1 T cells expand but are unable to survive at later time points. We therefore examined the possibility that these two populations of cells compete with each other during live vaccination. We decided to examine SM1 T cell persistence in a model where endogenous CD4 T cell competition cannot occur in vivo. SM1 T cells were therefore adoptively transferred into C57BL/6 mice or OT-II TCR-transgenic mice. All endogenous CD4 T cells in the periphery of OT-II mice are specific for OVA and should be unable to respond during vaccination with live Salmonella. Three days following live Salmonella vaccination, bacterial counts in the spleen of both recipient groups were similar (data not shown). SM1 T cells in both C57BL/6 and OT-II recipients expanded in response to vaccination, increased expression of CD11a, and underwent multiple rounds of cell division, as observed by CFSE staining (Fig. 7⇓A, day 3). SM1 T cells in OT-II mice expanded slightly more than those in C57BL/6 mice. As in previous experiments, SM1 T cells did not persist in the spleen of C57BL/6 mice vaccinated with live Salmonella (Fig. 7⇓A, day 10). In this particular experiment, SM1 T cells had not fully diluted CFSE in response to live Salmonella, allowing clear visualization of divided and undivided SM1 cells. The failure of SM1 T cells to persist at day 10 corresponded with the preferential loss of the most CFSE-dilute cells (Fig. 7⇓A, C57BL/6, day 3 vs day 10). In contrast, SM1 T cells persisted in OT-II recipient mice at day 10, and this included the survival of the most CFSE-dilute cells (Fig. 7⇓A, OT-II, day 3 vs day 10). We conclude that SM1 T cells fail to persist due to competition from the endogenous Salmonella-specific CD4 T cell response to live Salmonella.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 7.

Salmonella-specific endogenous CD4 T cells hinder the persistence of SM1 T cells in vivo. A, C57BL/6 or OT-II mice were adoptively transferred with CFSE-labeled 2 × 10⁶ SM1 T cells and groups of mice were vaccinated the following day with 5 × 10⁵ live attenuated Salmonella. Spleen cells were harvested 3 and 10 days later, and the percentage of SM1 T cells was determined. CFSE plots are gated only on SM1 T cells, as defined by the boxed gate shown. Numbers represent the percentage of SM1 T cells in the spleen. B, Groups of C57BL/6 mice were vaccinated with 5 × 10⁵ live attenuated Salmonella and 1 wk later some mice were injected i.v. with 1 × 10⁸ HKST prepared from flagellated (HKST) or aflagellate bacteria (aflagellate HKST). Spleens were harvested 2 h later and intracellular production of IFN-γ was determined. Data show the percentage of CD4 T cells producing IFN-γ as a percentage of all CD4 T cells in the spleen.

Flagellin is not a major target Ag of endogenous Salmonella-specific CD4 T cells

It still was not clear why endogenous CD4 T cells would have a selective survival advantage over SM1 T cells. In fact, previous adoptive transfer experiments have reported that an increased frequency of TCR-transgenic cells actually blocks the clonal expansion of endogenous T cells (3, 6, 34). Although flagellin has been identified as a target of Salmonella-specific CD4 T cell clones (29, 35), it remained possible the majority of the endogenous CD4 response was directed at other unknown proteins and that flagellin-specific CD4 T cells do not compete well with these other specificities. We therefore sought to use the HKST pulse assay to estimate the fraction of endogenous Salmonella-specific CD4 T cells that are specific for flagellin after live vaccination. HKST was prepared from flagellate or aflagellate Salmonella and these Ag preparations compared for their ability to activate endogenous CD4 T cells in vaccinated mice. Surprisingly, we found that the endogenous CD4 response to flagellate HKST was roughly equivalent to that of aflagellate HKST (Fig. 7⇑B), suggesting that flagellin is not a major target of Salmonella-specific CD4 T cells in vivo. Therefore, it is likely that SM1 T cells do not compete well with the endogenous Salmonella-specific response due to differences in the Ag specificity of these two populations.