Systemic CD8 T-Cell Memory Response to a Salmonella Pathogenicity Island 2 Effector Is Restricted to Salmonella enterica Encountered in the Gastrointestinal Mucosa

Bacterial strains. S. enterica serovar Typhimurium strain ATCC 14028s was used as the background for the construction of isogenic vaccine strains. The sspH2::SIINFEKL chromosomal translational fusion was constructed by following the one-step mutation procedure originally described by Datsenko and Wanner (9). The forward primer contained 60 bp of DNA homologous to the sspH2 target gene, followed by a sequence encoding amino acids 257 to 264 of the ovalbumin peptide (OVA_257-264; SIINFEKL) and the upstream FLP recombination target site of the pKD13 plasmid. DNA sequencing verified the insertion of the sequence encoding the OVA_257-264 peptide SIINFEKL in frame between nucleotides encoding residues 453 and 455 of SspH2 (Fig. 1A). The sspH2::SIINFEKL allele was inserted into Salmonella strains CL1509 (aroA) (40) and AJB82 (aroA invA) (4) to yield strains AV2199 (aroA sspH2::SIINFEKL⁺) and AV2200 (aroA invA sspH2::SIINFEKL⁺). The ΔsseB::aphT mutation that inactivates the SPI2 translocon was moved from strain HH102 (12) into strain AV2200, yielding strain AV2201 (aroA invA sseB sspH2::SIINFEKL⁺).

• Open in new tab
• Download powerpoint

FIG. 1.

A model for studying CD8 systemic responses to oral Salmonella vaccines. (A) Gel showing PCR fragments of wild-type sspH2 and the sspH2::SIINFEKL translational fusion. Below, the OVA_252-267 peptide sequence is shown in bold, flanking SspH2 domains lie within the green boxes, and the scar left after the excision of pKD13 is underlined. (B) The SIINFEKL peptide (black rectangle) is injected into the cytosol of APC via the SPI2 TTSS as an SspH2 chimera. After digestion, the SIINFEKL peptide is transported by TAP-1 into the endoplasmic reticulum (ER) for binding with major histocompatibility complex class I molecules (red). OT1 CD8 αβ T-cell responses to the SIINFEKL peptide serve as surrogate markers of Salmonella-specific immunity. (C) Ly5.1⁺ C57BL/6 mice were orally immunized with an invasion-deficient Salmonella vaccine strain (aroA invA sspH2::SIINFEKL⁺) 1 to 3 days after the grafting of naïve Ly5.2⁺ CD8⁺ OT1 cells. The expression of Ly5.2 by CD8⁺ splenocytes was measured by flow cytometry. The SPI2 dependency of cytosolic translocation was studied by expressing the chimera in sseB mutant Salmonella cells. (D and F) Levels of IFN-γ in the sera of orally immunized mice (D) and in the supernatants of splenocytes cultured with the SIINFEKL peptide or Salmonella parboiled antigens (F) were measured by ELISAs. Because Salmonella cells disseminating extraintestinally reach the spleen and liver in identical numbers (45), the bacterial burden was quantified by plating all tissue from livers macerated in a stomacher. (E) The role of the cytosolic pathway in the SspH2::SIINFEKL-dependent stimulation of Ly5.1⁺ OT1 cells was tested with Ly5.2⁺ TAP-1 knockout (KO) mice. All responses were measured 7 days after oral boosting. The data are representative of five to eight independent observations from at least two separate experiments.

Mice.Wild-type Ly5.1⁺ C57BL/6 and Ly5.2⁺ TAP-1-deficient mice (43) and OVA transgenic Ly5.2⁺ OT1 mice expressing the Vβ5.1 T-cell receptor, which was originally derived from the K^b-restricted OVA_257-264-specific clone 149.42 (14), were purchased from Jackson Laboratories (Bar Harbor, ME). The OT1 transgenic allele in a Ly5.2⁺ background was backcrossed into wild-type Ly5.1⁺ C57BL/6 mice in our animal facility. Mice were housed and treated in accordance with National Institutes of Health guidelines and Institutional Animal Care and Use Committee protocols. Oral immunization was initiated with 6- to 8-week-old mice.

Adoptive transfers of OT1 cells and oral immunization.About 2 × 10⁶ nylon wool column-purified T cells isolated as described previously (33) from among splenocytes of OT1 transgenic mice were inoculated intravenously into naïve or immunized C57BL/6 mice. Mice receiving OT1 cell grafts were immunized orally with approximately 5 × 10⁹ CFU of Salmonella strain AJB82, AV2200, or AV2201 and were given a booster dose 2 weeks later. Selected groups of naïve or immunized mice received grafts of memory cells isolated from immunized animals 40 to 110 days after oral immunization. Memory CD8 T cells were isolated from splenocytes by nylon wool column purification, and the cell population was further enriched with memory CD8 T cells by positive magnetic-bead sorting using biotinylated anti-CD8 monoclonal antibodies and magnetic bead-labeled streptavidin (Miltenyi Biotech, Auburn, CA). CD8 memory cells were adoptively transferred into naïve controls or mice immunized 40 to 50 days earlier. Selected groups of mice were orally immunized a few days after grafting, whereas the water of some orally immunized mice was treated during the last 3 weeks with nalidixic acid before memory cells were grafted.

Immunoassays.SIINFEKL peptide-specific CD8 splenic responses were analyzed by flow cytometry by dually labeling CD8 and Ly5.1 or Ly5.2 in a lymphocytic gate designed according to forward and side scattering. Memory cells, identified on the basis of CD44^high expression, were classified further as central or effector memory lymphocytes according to their high or low levels of CD62L expression, respectively (37). The production of gamma interferon (IFN-γ) by splenocytes cultured in RPMI medium supplemented with 10% fetal calf serum (BioWhittaker, Walkersville, MD), 2 mM l-glutamine, 15 mM HEPES, 1 mM Na pyruvate (Sigma-Aldrich, St. Louis, MO), 5 μM 2-mercaptoethanol, 100 U of penicillin/ml, and 100 mg of streptomycin (Cellgro, Herndon, VA)/ml in the presence or absence of 0.50 μg of the SIINFEKL peptide/ml was measured by sandwich enzyme-linked immunosorbent assays (ELISAs). All the reagents for flow cytometry and the ELISAs were purchased from Caltag (Burlingame, CA), BD Pharmingen (San Diego, CA), or eBiosciences (San Diego, CA).

Isolation of APC.Antigen-presenting cells (APC) were isolated from the spleens, Peyer's patches, or mesenteric lymph nodes (MLN) of C57BL/6 mice. Spleens were macerated and the red blood cells were eliminated by using Gey's lysis buffer. MLN and Peyer's patches were resected aseptically and digested in Dulbecco's phosphate-buffered saline, without Ca²⁺ or Mg²⁺ (Sigma-Aldrich), containing 0.2% collagenase I and IV (Sigma-Aldrich) and 2% heat-inactivated fetal calf serum for 5 min at 37°C in a 5% CO₂ incubator. The specimens were placed on ice for 10 min in 5 mM EDTA diluted in Dulbecco's phosphate-buffered saline supplemented with 2% heat-inactivated fetal calf serum. Cells from MLN were pelleted and resuspended in RPMI medium as described above for splenocytes.

Antigen presentation.Cells (4 × 10⁵) isolated from spleens, Peyer's patches, and MLN were added to 24-well plates containing 10⁶ OT1 cells. OT1 cells were isolated from mice immunized orally between 70 and 110 days earlier with an aroA invA sspH2::SIINFEKL⁺Salmonella vaccine strain. OT1 cells contained in the splenocyte samples were nylon wool-purified, and the samples were further enriched with OT1 cells by positive selection for CD8 T cells by magnetic-bead cell sorting according to the indications of the bead manufacturer (Miltenyi Biotech). APC and OT1 cells were cultured for 3 days with 0.50 μg of the SIINFEKL peptide/ml at 37°C in a 5% CO₂ incubator. The percentages of CD8⁺ (fluorescence parameter FL1) and Ly5.1⁺ (fluorescence parameter FL3) OT1 T cells contained in a lymphocyte gate were determined by flow cytometry after Topro-3⁺ (fluorescence parameter FL4) dead cells were eliminated from the analysis.

CD8 T-cell proliferation.Selected groups of naïve or memory OT1 cells were labeled as previously described (21) with 5 μM carboxyfluorescein diacetate succinimidyl ester (CFSE) prior to the adoptive transfer. Levels of proliferating CD8⁺ Ly5.1⁺ T cells were estimated by using a four-color FACSCalibur flow cytometer (Becton Dickinson, Mountain View, CA) to monitor the loss of CFSE 5 to 21 days after transfer. Dead cells were eliminated from the analysis after the gating out of Topro-3⁺ (FL4) cells.

Stimulation of antigen-specific CD8 systemic T-cell responses by an orally administered Salmonella vaccine strain expressing the sspH2::SIINFEKL translational fusion.The acquired response against Salmonella involves B cells and γδ T cells as well as CD4 and CD8 αβ T cells (13, 19, 23-25, 29, 30). Several investigators have successfully measured acute CD8 T-cell responses to Salmonella expressing OVA from a plasmid (11, 39, 49). However, in the absence of selective pressure, plasmids are frequently lost and thus are less reliable for long-term studies. To examine long-term systemic responses to oral Salmonella vaccines, we engineered a stable OVA_257-264 SIINFEKL construct expressed from the Salmonella chromosome as a chimera with an SPI2 effector that allows long-term expression in vivo (Fig. 1A). Because substrates of the type III secretion system (TTSS) are frequently poor immunogens, SPI2 TTSS effectors were evaluated for homology to the SPI1 SptP effector that is known to stimulate T-cell responses (35). Protein sequence alignments to determine homology to SptP domains identified the SspH2 effector, which is secreted into the host cell cytosol in an SPI2-dependent fashion (27, 28). The SIINFEKL peptide inserted in frame in a predicted flexible loop of SspH2 (Fig. 1B) was expressed in an aroA invA Salmonella vaccine strain (45).

Similar to CD8 T cells involved in the acute responses triggered by Salmonella expressing high levels of OVA (49), antigen-specific Ly5.2⁺ CD8⁺ T cells accumulated in the spleen shortly after oral immunization with the Salmonella vaccine strain expressing the sspH2::SIINFEKL chromosomal translational fusion (Fig. 1C). Further analysis verified that the Ly5.2⁺ CD8⁺ T cells expressed the Vβ5.1 allele characteristic of OT1 transgenic cells (data not shown). To demonstrate dependency on the SPI2 TTSS for the delivery of the SspH2::SIINFEKL chimera into the cytosol, the sspH2::SIINFEKL allele was expressed in an isogenic aroA invA vaccine strain carrying the sseB::aphT mutation that inactivates all intracellular SPI2-mediated translocation. The sseB mutant strain triggered significant (P < 0.01) Salmonella-specific IFN-γ systemic responses, although it was less immunogenic than the corresponding SPI2-proficient control (Fig. 1D, left panel). In sharp contrast to its IFN-γ-stimulating capability, the sseB mutant vaccine strain failed to trigger any significant (P = 0.32) expansion of the population of Ly5.2⁺ CD8⁺ cells (Fig. 1C). The lack of a CD8 T-cell response cannot be explained by reduced survival and extraintestinal dissemination of the sseB-deficient vaccine strain, since mice inoculated orally with either sseB-proficient or sseB-deficient vaccine strains harbored identical hepatic bacterial loads (P = 0.87) (Fig. 1D, right panel). The lack of SIINFEKL peptide-specific CD8 T-cell responses in mice immunized orally with the aroA invA sseB sspH2::SIINFEKL⁺ vaccine strain is consistent with the SPI2-dependent translocation of SspH2 into the cytosol (27, 28).

TAP-1-deficient mice were used to test whether CD8 T-cell responses elicited by the Salmonella vaccine require the translocation of the SIINFEKL peptide from the cytosol into the endoplasmic reticulum. Because TAP-1-deficient mice express the Ly5.2 allele, the OT1 transgene was backcrossed into an Ly5.1⁺ C57BL/6 background. Ly5.1⁺ OT1 transgenic mice were used for this and all subsequent experiments. Figure 1E shows that the TAP-1 transporter is critical for the expansion of populations of antigen-specific CD8⁺ T cells derived in response to the orally administered sspH2::SIINFEKL⁺Salmonella vaccine strain. In agreement with the results of other studies (1, 41), an endogenous Ly5.1⁻ CD8⁺ population consisting mainly of γδ T cells was nevertheless observed in the spleens of immunized TAP-1-deficient mice. In contrast, αβ T cells represented most of the Ly5.1⁻ CD8⁺ lymphocytes in wild-type mice. Splenocytes isolated from orally vaccinated TAP-1-deficient mice did not secrete IFN-γ in response to the SIINFEKL peptide but synthesized amounts of IFN-γ similar to those synthesized by TAP-1-competent controls in response to parboiled Salmonella antigens (Fig. 1F). Together, these findings are consistent with a model in which the SIINFEKL peptide is injected into the cytosol via the SPI2 TTSS as a chimera with SspH2, translocates into the endoplasmic reticulum for loading into major histocompatibility complex class I molecules, and is highly immunogenic. This strategy was used to evaluate long-term systemic CD8 T-cell memory responses evoked by an oral Salmonella vaccine.

Oral immunization elicits systemic effector memory CD8 T-cell responses.The kinetics of systemic CD8 T-cell responses to oral immunization with Salmonella were studied. In contrast to the relatively slow responses seen when OVA is delivered into the phagosome by systemically administered wild-type Salmonella (20), a six- to sevenfold increase in antigen-specific CD8 T cells in mice receiving OT1 cell grafts was already detected 7 days after oral exposure to sspH2::SIINFEKL⁺Salmonella (Fig. 2A). The percentage of SIINFEKL peptide-specific T cells increased 15-fold 2 weeks after oral immunization and had increased about 65-fold by 7 days after oral boosting. The increase in Ly5.1⁺ CD8⁺ T cells was antigen specific and not the product of the expansion of populations of bystander T cells, because populations of OT1 T cells did not expand in mice immunized orally with an isogenic vaccine strain (i.e., an aroA invA strain) expressing the sspH2 wild-type allele (Fig. 2A). To determine whether the expansion of populations of antigen-specific CD8 T cells is a consequence of T-cell proliferation, donor OT1 cells were labeled with CFSE before grafting. As shown in Fig. 2B, all the grafted cells proliferated 3 weeks after oral immunization. Over 90% of the cells that proliferated in the spleens exhibited the CD44^high memory phenotype, whereas 80% of them were CD62L^low (Fig. 2C and D). Predominantly CD62L^low expression by CD44⁺ cells suggests that oral Salmonella vaccines preferentially trigger early effector memory T-cell responses.

• Open in new tab
• Download powerpoint

FIG. 2.

Effector memory CD8⁺-T-cell responses to an oral Salmonella vaccine. (A) The percentage of CD8⁺ Ly5.1⁺ SIINFEKL peptide-specific T cells recovered from the spleens of mice receiving grafts of OT1 cells was determined at different time points after oral immunization with aroA invA sspH2::SIINFEKL⁺Salmonella. Naïve mice and mice inoculated orally with aroA invA Salmonella were used as controls. 7d, 7 days. (B) The proliferation of lymphocytes in the CD8⁺ Ly5.1⁺ gate was visualized as the loss of CFSE among cells from naïve or orally immunized (immune) mice 21 days after immunization. (C and D) The abundance of central memory cells was expressed as the percentages of CD8⁺ Ly5.1⁺ splenocytes that were CD44^high (C) and CD62L^high (D). Regions were drawn by using isotype controls (data not shown) according to the CD44 and CD62L expression patterns of naïve OT1 cells. The mice in these experiments were immunized orally with the aroA invA sspH2::SIINFEKL⁺Salmonella vaccine strain. The data are representative of nine independent observations from three separate experiments.

An oral Salmonella vaccine triggers a sustained effector memory systemic response.A large population of antigen-specific CD8 T cells remained in orally immunized mice 40 days after boosting (Fig. 3A). Systemic boosting with the SIINFEKL peptide-expressing Salmonella vaccine strain did not lead to further increases in the CD8⁺ Ly5.1⁺ T-cell pool, which remained at a level of around 1 to 3% of the total lymphocyte population for at least 3 weeks postchallenge. Further analysis indicated that CD62L^low effector memory T cells were predominant, accounting for about 75% of the CD8⁺ memory pool (Fig. 3B). Splenocytes isolated from immune mice secreted IFN-γ in response to the SIINFEKL peptide (Fig. 3C). Since Salmonella reaches similar burdens in livers and spleens (45), the hepatic Salmonella burden was determined as an indicator of systemic infection. The Salmonella burden in livers peaked 7 days after the intraperitoneal challenge of orally immunized mice but dropped to very low levels after 21 days.

• Open in new tab
• Download powerpoint

FIG. 3.

Effector memory responses to oral Salmonella vaccines. (A) The frequencies of CD8⁺ Ly5.1⁺ cells in naïve mice (N) and orally immunized mice 40 days postimmunization (I) and those in orally immunized mice after intraperitoneal boosting with the aroA invA sspH2::SIINFEKL⁺Salmonella vaccine strain (I-B) were determined. (B and D) The percentages of CD62L^high central memory (CM) and CD62L^low effector memory (EM) cells (B) and the hepatic bacterial burdens (D) in animals 40 days postimmunization after intraperitoneal boosting with ∼100 CFU were determined. (C) Splenocytes from immunized mice were pulsed in vitro with the SIINFEKL peptide and analyzed for their IFN-γ-secreting capacities. The data in all panels are representative of six independent observations from two separate experiments.

Memory T cells responding to Salmonella are capable of homeostatic and antigen-driven proliferation.To test whether memory cells are maintained by foci of infection, Ly5.1⁺ splenocytes isolated from orally vaccinated mice 40 days after immunization were grafted into orally immunized mice or naïve controls. About 20% of immunized mice that received grafts of memory OT1 cells harbored CD8⁺ Ly5.1⁺ cells at levels of approximately 1% of the lymphocytic populations in the spleens (Fig. 4A). Mice with large numbers of CD8⁺ Ly5.1⁺ cells (i.e., those with CD8⁺ Ly5.1⁺ cells accounting for about 1% of the CD8 αβ T cells) were classified as high responders. The observations that antibiotic treatment abolished the occurrence of high CD8⁺ Ly5.1⁺ responders and that populations of naïve T cells expanded in mice that had been immunized 50 days before grafting (Fig. 4B) suggest the existence of foci of infection. Exposure to antigenic stimulation for extended periods of time may lead to lymphocyte exhaustion, a condition in which antigen-specific T cells are eventually depleted (46). Therefore, the ability of memory cells to sustain homeostatic and antigen-driven proliferation was studied (Fig. 4C). The proliferation of CFSE-labeled Ly5.1⁺ memory cells in naïve mice revealed that oral Salmonella vaccines elicit a population of memory cells capable of homeostatic turnover. When results were corrected for absolute numbers, memory cells were found to proliferate to a fivefold-higher level when recipients were immunized orally shortly after grafting. These findings rule out the possibility that Salmonella vaccines elicit oral tolerance by either depletion or anergy of antigen-specific cells and suggest that the suppression recently attributed to Salmonella (42) does not appear to block antigen-specific CD8 T-cell recall responses.

• Open in new tab
• Download powerpoint

FIG. 4.

Memory CD8 T cells are capable of both homeostatic and antigen-driven proliferation. (A and B) Ly5.1⁺ splenocytes isolated 40 days after animals were orally immunized (mem; panel A) or Ly5.1⁺ cells isolated from naïve controls (naïve; panel B) were grafted into naïve mice or mice that had been orally immunized 45 days earlier (immune). Some immune animals were treated with nalidixic acid in drinking water 3 weeks before grafting (immune+ATB). The frequency of SIINFEKL peptide-specific Ly5.1⁺ CD8⁺ splenic T cells was estimated by flow cytometric analysis. (C) CSFE-labeled naïve splenocytes or Ly5.1⁺ splenocytes taken from immune mice 45 days postimmunization (mem) were grafted into either orally immunized mice 45 days postimmunization (imm) or naïve recipients. Selected groups of mice were orally inoculated with the sspH2::SIINFEKL⁺Salmonella vaccine strain 2 days after grafting (imm+boost). Cell division was evaluated by flow cytometry as the loss of the CFSE label among the CD8⁺ Ly5.1⁺ population. All responses were measured 7 days after grafting. The data are representative of 5 to 10 independent observations from at least two separate experiments.

Oral Salmonella vaccines elicit CD8 T-cell memory responses restricted to antigens encountered in the gastrointestinal tract.We found that 80% of orally immunized animals maintained a robust CD8 T-cell memory pool (i.e., they were high responders) 50 days postimmunization (Fig. 5A); however, the frequency of high responders dropped to 40% by 110 days. The administration of oral booster doses increased (P < 0.01) the number of high responders among mice at 110 days postimmunization (Fig. 5A). In contrast, the number of high responders did not significantly increase (P = 0.65) after the administration of parenteral booster doses to mice at late time points postimmunization. Collectively, these data suggest that the reconstitution of the pool of antigen-specific splenic CD8 T cells in orally immunized mice is restricted to antigens encountered in the gastrointestinal tract. The patterns of IFN-γ production by splenocytes isolated from mice at late time points postimmunization (Fig. 5B) further support the enteric restriction of memory CD8 T-cell responses elicited by oral Salmonella vaccines. The early ability of splenocytes to secrete IFN-γ in response to the SIINFEKL peptide was dramatically lessened at later time points (e.g., 50 versus 110 days post-oral immunization). However, treating mice with oral booster doses at late time points postimmunization effectively increased the incidence of IFN-γ producers (Fig. 5B, right panel). Although both effector and central memory cells synthesized IFN-γ, the majority of the memory cells expressed a CD62L^low effector phenotype (Fig. 5C). By 110 days postimmunization, the majority of the high responders expressed a CD62L^low effector memory phenotype, whereas a central memory CD62L^high cell population predominated in low responders (Fig. 5D). Of note, both populations of memory cells expressed high levels of the interleukin-7 receptor.

• Open in new tab
• Download powerpoint

FIG. 5.

The antigen-driven expansion of CD8 memory cells is restricted to Salmonella cells encountered in the oral mucosa. (A) The percentages of CD8⁺ Ly5.1⁺ splenocytes in mice receiving grafts of OT1 cells were determined 50 days (50d) and 110 days (110d) after oral vaccination. An analysis was also performed 7 days after mice received oral (110d+po) or parenteral (110d+ip) booster vaccines at 110 days postimmunization. (B) The IFN-γ production in splenocytes isolated 50 or 110 days after oral immunization was measured. The SIINFEKL peptide (SIIN) was added to selected groups of splenocytes. The right panel shows the IFN-γ response 7 days after animals received oral booster doses at 110 days postimmunization. (C) Cells isolated from high responders (immune) were analyzed by flow cytometry for intracellular IFN-γ. OT1 cells transferred into naïve mice were used as controls. (D) CD8⁺ Ly5.1⁺ cells isolated from mice with elevated or reduced numbers of OT1 cells (i.e., high responders [H.R.] and low responders [L.R.], respectively) were examined for the expression of the interleukin-7 receptor (IL7R) and CD62L 110 days postimmunization. The data are representative of four to seven independent observations from two independent experiments.

Central memory cells restricted to MLN drive systemic effector memory cells in response to oral immunization.The selective expansion of populations of OT1 memory cells in response to sspH2::SIINFEKL⁺Salmonella encountered orally but not systemically suggests that memory CD8 T cells selected by oral vaccines are restricted to cognate antigens encountered in the gastrointestinal tract. To test this idea, the capacity of memory grafts to expand in response to oral or systemic restimulation was studied. Consistent with the data presented in Fig. 5, populations of memory CD8 T cells in naïve mice expanded after oral but not parenteral immunization with sspH2::SIINFEKL⁺Salmonella (aroA invA) (Fig. 6A). CD62L^high central and CD62L^low effector memory cells were isolated by flow cytometric sorting from vaccinated mice 110 days after oral immunization. Grafts consisting of CD62^low effector memory CD8 T cells survived in low numbers in mice that received oral booster doses (Fig. 6B, left panel); however, these cells disappeared in naïve hosts (Fig. 6B). Conversely, central memory cells survived after grafting into naïve mice (Fig. 6B, central panel), possibly reflecting the capacity of this population of memory cells to undergo homeostatic turnover (15, 47). In contrast to effector memory cells, central memory grafts (i.e., CD62L^high) expanded in response to SspH2::SIINFEKL⁺Salmonella administered through the oral mucosa (Fig. 6B, right panel). Central memory cells restricted to orally administered antigens repopulated both CD62L^low and CD62L^high memory subsets.

• Open in new tab
• Download powerpoint

FIG. 6.

Central memory CD8 T cells restricted to orally administered antigens drive systemic effector memory responses. (A) Ly5.1⁺ cells isolated from mice orally immunized with Salmonella 70 to 110 days earlier were grafted into naïve mice after positive selection over a magnetic bead column. The expansion of the Ly5.1⁺ CD8⁺-T-cell population was tested 7 days after the intraperitoneal (ip; middle panel) or oral (po; right panel) administration of an aroA invA sspH2::SIINFEKL⁺Salmonella vaccine strain. (B) CD62L^low and CD62L^high CD8⁺ Ly5.1⁺ cells isolated by flow cytometric sorting from mice orally immunized 110 days earlier with aroA invA sspH2::SIINFEKL⁺Salmonella were grafted into naïve mice. A day after grafting, selected groups of mice were either intraperitoneally or orally immunized with aroA invA sspH2::SIINFEKL Salmonella. The abundance of CD62L CD8⁺ T cells in a Ly5.1⁺ gate was analyzed by flow cytometry 7 days after Salmonella challenge. The data are representative of five independent observations from two separate experiments.

The compartmentalization of the recall response to the oral mucosa was further examined by comparing the expansion of populations of memory CD8⁺ Ly5.1⁺ T cells in response to APC from Peyer's patches, MLN, or spleens. As shown in Fig. 7, the greatest expansion of Ly5.1⁺ CD8⁺ transgenic cells isolated from orally immunized mice was seen upon coculture with SIINFEKL peptide-primed APC from MLN. In contrast, APC from spleens, Peyer's patches, and MLN supported the expansion of populations of naïve OT1 cells (data not shown). About 2% of the SIINFEKL peptide-specific OT1 cells selected at 110 days after oral immunization expressed α4β7 gut-homing receptors, whereas at late time points after oral immunization with S. enterica serovar Typhi, most of the CD8 T cells coexpress this receptor (36).

• Open in new tab
• Download powerpoint

FIG. 7.

Selective expansion of central memory cells restricted to antigen administered orally in response to cells isolated from MLN. APC from Peyer's patches, MLN, or spleens were cocultured for 7 days with OT1 memory CD8 T cells isolated from spleens of mice immunized orally 85 to 90 days earlier. The T-cell population was enriched with OT1 cells by CD8 magnetic-bead sorting of nylon wool-purified T cells. The expansion of transgenic cells was quantified by flow cytometric analysis by monitoring the frequency of CD8⁺ Ly5.1⁺ cells. The data are representative of 12 independent observations from three separate experiments.