Pathogen Proliferation Governs the Magnitude but Compromises the Function of CD8 T Cells¹

The activation of CD8⁺ T cells occurs rapidly within the first few days of infection (1, 2, 3), which is followed by an intense phase of expansion of the CD8⁺ T cell response (4). Subsequently, the majority (>95%) of the primed CD8⁺ T cells are eliminated, and only a small portion of those T cells survive (<5%) for extended periods as memory cells (5, 6, 7, 8, 9). Memory CD4⁺ and CD8⁺ T cells have been segregated phenotypically and functionally into effector-memory and central memory populations based on the expression of CD62L and CCR7 (10, 11, 12). Tissue-homing “effector-memory T cells” (CD62L^lowCCR7⁻) are capable of immediate effector function, whereas the lymph node homing “central memory T cells” (CD62L^highCCR7⁺) are devoid of effector activity in vitro (12), but proliferate profoundly in vivo (8, 13, 14, 15).

Although CD8⁺ T cells have been considered to play an essential role mainly during viral infection models, they also play an important role in mediating protection against intracellular bacteria such as Listeria monocytogenes (LM)³ (16, 17, 18), Mycobacterium tuberculosis (19), and Salmonella enterica, ser typhimurium (ST) (20, 21). ST is a highly virulent pathogen that induces gastroenteritis in humans. In C57BL/6J strain of mice, ST (strain SL1344) induces a lethal infection even when used at doses as low as 10² i.v., and 100% of mice die within 7 days of infection. In contrast, ST induces a chronic, but nonlethal, infection in 129SvJ or B6.129F₁ mice where the infection is usually cleared around day 60–90. ST survives within macrophages and epithelial cells and the intracellular replication of ST is considered to be essential for virulence (22). The genes that are involved in Salmonella invasion are clustered at one location on the bacterial chromosome (centisome 63; Salmonella pathogenicity island-1 (SPI-1)) (23, 24, 25, 26). They encode several factors, including an operon encoding a type III secretory apparatus that exports specific proteins into the host cell. The invA mutant of ST is unable to invade epithelial cells and is attenuated for oral infection of mice (27). InvA is a putative inner membrane component of the SPI-1 type three section system (25). Two major virulence loci allow Salmonella to survive and replicate inside phagocytes (28). The two-component regulatory system phoP/phoQ which controls >40 different genes (29, 30) is involved in intracellular survival (31). Another pathogenicity island (SPI-2) encodes a second type III secretion system that mediates resistance to intracellular killing and is key to bacterial virulence (32, 33).

We have evaluated the role of pathogen proliferation in the differentiation and function of CD8⁺ T cells. Our results reveal that the magnitude of the CD8⁺ T cell response is governed by the proliferation of ST. However, the optimal function of CD8⁺ T cells is governed by the phenotype of CD8⁺ T cells, which is blunted by the chronic proliferation of ST.

The gene for OVA was introduced into virulent (SL1344) and the various mutants (aroA, phoP, ssaR, and invA) of ST. Plasmid pKK-OVA (34, 35) (10–100 ng of DNA) carrying the full-length OVA was electroporated into ST as described previously (36). Expression of OVA by recombinant wild-type (WT) ST-OVA and the various mutants of ST-OVA was determined by enhanced chemiluminescence detection system as described previously (36). ST-OVA were grown in liquid culture at 37°C under constant shaking in brain heart infusion (BHI) medium (Difco Laboratories). At mid-log phase (OD₆₀₀ = 0.8), bacteria were harvested and frozen at −80°C (in 20% glycerol). CFU were determined by performing serial dilutions in 0.9% NaCl, which were spread on BHI-streptomycin agar plates.

OVA-expressing LM, as described previously (37), was grown to OD_{600 nm} = 0.4. The bacteria were grown in BHI medium supplemented with 50 μg/ml streptomycin (Sigma-Aldrich). At mid-log phase (OD₆₀₀ = 1.0), bacteria were harvested and frozen in 20% glycerol and stored at −80°C. CFU were determined by performing serial dilutions in 0.9% NaCl, which were spread on BHI-streptomycin agar plates.

C57BL/6 and 129X1SvJ mice were obtained from The Jackson Laboratory. B6129F₁ mice were generated in house in our experimental animal facility by mating 129X1SvJ female mice with C57BL/6 male mice. OT-1 TCR transgenic mice were obtained from The Jackson Laboratory. For immunization, frozen stocks of bacteria were thawed and diluted in 0.9% NaCl. Mice were inoculated with 1 × 10³ organisms suspended in 200 μl of 0.9% NaCl, via the lateral tail vein (i.v.). In some experiments mice were injected first with 10⁴ OT-1 CD8⁺ T cells (i.v.) and challenged a few days later with ST-OVA (10³, i.v.).

Single cell suspensions were obtained from the spleens of infected mice in RPMI 1640. An aliquot of the suspension was lysed with water for 30 s and then evaluated for the numbers of viable bacteria. CFU were determined by plating 100-μl aliquots of serial 10-fold dilutions in 0.9% saline on BHI plates as above.

Bacterial cultures were grown to mid-log phase (OD = 0.6–0.8) shaking at 37°C in liquid BHI medium supplemented with 100 μg/ml ampicillin. Cells were harvested using RNAProtect bacteria reagent (Qiagen) as per manufacturer’s recommendation and snap frozen in a dry-ice/100% ethanol bath. Total RNA was extracted using the Qiagen RNeasy mini kit according to the manufacturer’s instructions along with rapid mechanical lysis. Mechanical lysis was done in 1 ml of lysis buffer in a Mini-BeadBeater 3110BX (BioSpec Products) using 0.1 mm glass beads (BioSpec Products). Total RNA from homogenates was extracted according to the manufacturer’s instructions. RNA was treated with RNase free Turbo DNase (Ambion) for 30 min at 37°C followed by LiCl precipitation of the RNA. Five micrograms of total RNA was taken for cDNA synthesis. cDNA was synthesized using N8 random primers purchased from Sigma and cDNA synthesis was performed as previously described (38). Identical samples not treated with Superscript II were also prepared as controls to measure DNA contamination. Remaining RNA template was hydrolyzed, neutralized, and purified as previously described (38). The number of amplicons was measured by quantitative RT-PCR using gene-specific primers and qPCR SYBR Green Supermix (ABgene). The primers used for the ST 16S rRNA cDNA were 5′-CGGGGAGGAAGGTGTTGTG-3′ and 5′-GAGCCCGGGGATTTCACATC-3′ and for the OVA cDNA within the various Salmonella mutants evaluated were 5′-CAACCTCACATCTGTCTTAATGG-3′ and 5′-GCCTCTGCTGACCCTACC-3′. Primers were designed using Primer Express 2.0. To obtain a standard curve for each primer-template set, 10-fold dilutions of known amounts of ST-OVA chromosomal DNA (0.01, 0.1, 1, 10, and 100 attomoles) were used as template DNA. This standard curve was run together with triplicate reactions of the uncharacterized samples. PCR conditions were optimized based on the dissociation curve of each primer and its target. PCR was performed in sealed tubes in a 96-well microtiter plate in an ABI Prism 7000 thermocycler (Applied Biosystems). The 25-μl reaction consisted of 12.5 μl qPCR SYBR Green Supermix, 2.5 μl of each primer (1.5 pmol/μl each), and 10 μl of template (0.15 ng/μl). Thermal conditions were as follows: activation at 95°C for 15 min, followed by 40 cycles of denaturation at 95°C for 30 s, annealing at 60°C for 30 s, and extension at 72°C for 1 min. Fluorescence was measured during the annealing step and plotted against the amplification cycle. Absolute quantitative analysis of the data was extrapolated from the standard curve. Primer efficiencies were between 100 and 99%.

Aliquots (10 × 10⁶) of spleen cells were incubated in 200 μl of PBS plus 1% BSA with anti-CD16/32 at 4°C. Cells were then incubated on ice for 30 min with anti-CD44-FITC, anti-CD62L-PE-Cy7, anti-CD8-PerCP-Cy5.5. Cells were washed and then incubated at room temperature with PE-H-2K^bOVA_257–264 tetramer. After 30 min, cells were washed with PBS and fixed in 0.5% formaldehyde and acquired on BD FACSCanto flow cytometer.

The intracellular proliferation of ST was evaluated on thioglycollate elicited macrophages as described previously (22). After thioglycollate injection, peritoneal cells were harvested from B6.129F₁ mice on day 10 and seeded (in RPMI 1640 plus 8% FBS) into 24-well plates (2 × 10⁵) and infected with 2 × 10⁶ ST-OVA. After 15 min, cells were washed and extracellular bacteria were removed after incubation in medium containing gentamicin (50 μg/ml). Each incubation step was conducted at 37°C in a CO₂ incubator. At 2 h, cells were washed with medium without gentamicin and the numbers of intracellular bacteria were enumerated by lysing macrophages and plating serial dilutions on BHI agar plates. After 2 h, aliquots of cells were cultured in RPMI 1640 plus 8% FBS medium containing 5 μg/ml gentamicin to allow intracellular, but not extracellular, proliferation of ST. At 18 h, the numbers of intracellular bacteria were enumerated by lysing cells and plating serial dilutions on BHI agar plates.

Aliquots of spleen cells (10 × 10⁶/ml) were stained with anti-CD8 Ab and H-2K^bOVA_257–264 tetramer for 30 min as described above. Cells were then washed, reconstituted in R8 medium (RPMI 1640 plus 8% FBS), plated into 96-well plates (2 × 10⁶/well), and stimulated with OVA_257–264 peptide (1 μg/ml) in the presence of GolgiStop (BD Biosciences). After 1 h, cells were harvested, washed, permeabilized, and stained for intracellular IL-2, TNF-α, and IFN-γ using the staining kit (obtained from BD Biosciences). Cells were acquired on BD Biosciences FACSCanto analyzer.

Cytolytic activity was determined after stimulation of spleen cells for 5 days in vitro with peptide expressing irradiated OVA-expressing target cells (EG7 cells) and IL-2 as described previously (37).

We wanted to determine whether mutation in key virulence genes of ST would influence the priming and function of OVA-specific CD8⁺ T cells. To this end, we cloned the gene for OVA into ST so that CD8⁺ T cell responses against OVA could be measured specifically. OVA was cloned into virulent WT ST (SL1344), and into ST that were defective in aromatic amino acid synthesis (aroA) (39), defective in invasion into epithelial cells (invA) (23, 24, 25), defective in intracellular replication (ssaR and phoP) (32, 33), and defective in inhibition of phagosomal maturation (phoP) (29, 30, 31). The expression of OVA by recombinants was measured in aliquots of actively growing bacteria. Samples were normalized for the number of bacteria (1.5 × 10⁷) and loaded on SDS-12% polyacrylamide gels. Expression of OVA was measured by Western blotting using an ECL-based detection system. As shown in Fig. 1, WT as well as the various mutants of ST express similar levels of OVA.

We then performed quantitative RT-PCR analysis to determine the relative levels of OVA expressed by the virulent vs the mutants of ST-OVA. The amount of OVA mRNA was correlated to the total bacterial RNA (16S rRNA). The relative ratio of OVA mRNA to the total bacterial RNA (16S) was similar for WT and the various mutants of ST-OVA (Fig. 1 B), indicating that the various mutants of ST-OVA express similar levels of OVA.

To determine whether the mutations of key virulence genes in ST influence the clearance of ST in vivo, we infected the susceptible (C57BL/6J) and resistant (129X1SvJ) strains of mice with 10³ ST-OVA. Mice were infected through the i.v. route so that the virulent as well as the various mutants of ST-OVA would gain entry into the lymphoid organs at the same time. In C57BL/6J mice, infection with the WT as well as the invA mutant resulted in massive bacterial burden initially culminating in fatality within the first 7 days (Fig. 2,A). All the other mutants did not induce a fatal infection and the bacterial burden was reduced in all mice. In the resistant mice, 129X1SvJ, or in F₁ hybrids between resistant and susceptible mice, bacterial burdens in the spleens (WT > invA > aroA > ssaR > phoP) displayed varying degrees of attenuation, with the phoP mutant displaying the greatest degree of attenuation. We also measured the influence of mutations on the survival of B6.129F₁ mice after infection with 10³ or 10⁵ dose of ST-OVA. At the 10³ dose of ST-OVA, a fraction of mice infected with WT ST-OVA were moribund (Fig. 2 B). In contrast, there was no fatality in mice that were infected with the mutants. When the dose of ST-OVA was increased to 10⁵, mice infected with WT ST-OVA became more susceptible progressively, which resulted in complete fatality by day 40. A significant proportion of mice infected with the invA mutant also succumbed to infection between days 30 and 50. In contrast, none of the mice that were infected with the other mutants of ST-OVA became moribund.

We then determined whether the plasmid that contains the gene for OVA is maintained by ST-OVA in the long term. To this end, we enumerated ST-OVA burden in the spleens of infected mice in the presence or absence of ampicillin, because the plasmid carrying the gene for OVA also expresses ampicillin resistance gene, and any discordance in the number of colonies in the absence or presence of ampicillin would be indicative of antigenic loss variants of ST-OVA. As is evident from Fig. 3, at day 45 after ST-OVA infection, no significant differences were noted in the number of ST-OVA colonies in the absence or presence of ampicillin.

B6.129F₁ mice were injected with OT-1 TCR transgenic CD8⁺ T cells and challenged with ST-OVA (10³, i.v.). The numbers of OVA-specific CD8⁺ T cells were evaluated by flow cytometry after staining spleen cells with anti-CD8 Ab and H-2K^b OVA tetramers. Although the infection of mice with the WT ST-OVA resulted in progressive increase in the numbers of OVA-specific CD8⁺ T cells, infection of mice with the mutants resulted in a greatly reduced OVA-specific CD8⁺ T cell response (Fig. 4). Similar results were noted for the response in peripheral blood (data not shown). The reduced OVA-specific CD8⁺ T cell response in mice infected with the mutants of ST could be explained by the reduced bacterial burden in such mice (Fig. 2,A); however, the bacterial burden in case of mice infected with the invA mutant was only slightly reduced (Fig. 2 A).

To determine whether the magnitude of the CD8⁺ T cell response correlates to the intracellular proliferation of ST-OVA, we measured the intracellular proliferation of ST-OVA within peritoneal macrophages using the gentamicin resistance assay (40). At 2 h postinfection, invA-infected macrophages had significantly reduced numbers of intracellular bacteria in comparison with other mutants or WT ST-OVA (Fig. 5,A). From 2 to 18 h, WT ST-OVA underwent maximal proliferation within macrophages (Fig. 5 B). invA and aroA mutants displayed moderate intracellular proliferation, whereas phoP and ssaR displayed minimal proliferation. These results indicate that the intracellular proliferation of ST-OVA correlates to the magnitude of the CD8⁺ T cell response.

To determine whether pathogen proliferation influences the phenotype of primed CD8⁺ T cells, we evaluated the expression of CD62L and IL-7Rα on OVA-specific CD8⁺ T cells. Infection with the least virulent mutants (phoP and ssaR) resulted in the development of increased numbers of OVA-specific CD8⁺ T cells expressing high levels of CD62L (Fig. 6,A) and IL-7Rα (Fig. 6,B). An inverse correlation was noted for intracellular proliferation of ST (Fig. 5) vs the CD62L/IL-7Rα expression on the OVA-specific CD8⁺ T cells (Fig. 6,A and B). Mutants that displayed less intracellular proliferation-induced CD8⁺ T cells expressing higher levels of CD62L and IL-7Rα. Even at late time points, OVA-specific CD8⁺ T cells induced against WT ST-OVA displayed mainly effector (CD62L^lowIL-7Rα^low) and effector-memory phenotype (CD62L^lowIL-7Rα^high), whereas the OVA-specific CD8⁺ T cells induced against the phoP and ssaR mutants of ST-OVA displayed mainly central memory phenotype (CD62L^highIL-7Rα^high). We also measured the cytolytic activity of OVA-specific CD8⁺ T cells at day 30 after infection. OVA-specific CD8⁺ T cells generated against mutants that displayed lowest intracellular proliferation (phoP and ssaR) exhibited the most potent cytolytic CD8⁺ T cell response (Fig. 6,C). OVA-specific CD8⁺ T cells generated against the ssaR and phoP mutants expressed higher levels of intracellular IL-2 in comparison with CD8⁺ T cells generated against WT ST-OVA (Fig. 6,D). The expression of intracellular IFN-γ and TNF-α by OVA-specific CD8⁺ T cells was not influenced by the intracellular proliferation of ST (Fig. 6 D).

To further determine whether the reduced CD8⁺ T cell response observed with the mutants of ST-OVA was due to decreased bacterial burden, we infected one group of mice with 10³ WT ST-OVA and compared the CD8⁺ T cell response to that generated in other groups of mice that received substantially higher doses of mutants of ST-OVA. The highest doses that could be tolerated by mice were used. Bacterial burden and OVA-specific CD8⁺ T cell response was measured at various time intervals. As is evident in Fig. 7,A, infection with 10⁵ doses of aroA and ssaR resulted in greater bacterial burden compared with WT at day 3 and 7 after infection. However, the CD8⁺ T cell response induced against the mutants was still greatly reduced in comparison with that induced against the WT ST-OVA (Fig. 7,B). These results suggest that pathogen proliferation, not persistence, governs the magnitude of CD8⁺ T cell response. Similarly, as described in Fig. 6,A and B, infection of mice with mutants resulted in increased numbers of cells expressing high levels of CD62L and IL-7Rα (Fig. 7 C and D).

To further determine whether the continued proliferation of ST-OVA after day 3 of infection is necessary for the development of CD8⁺ T cell response, we treated one group of infected mice with antibiotics (ciprofloxacin) from day 3 onwards. Commencement of antibiotic treatment at day 3 of infection resulted in significant and persistent reduction in bacterial burden (Fig. 8,A). Antibiotic treatment also resulted in massive reduction in the magnitude of the OVA-specific CD8⁺ T cell response (Fig. 8 B and C). OVA-specific CD8⁺ T cells in antibiotic-treated mice expressed high levels of CD62L (Fig. 8,D) and IL-7Rα (Fig. 8 E). In contrast, OVA-specific CD8⁺ T cells in control mice expressed low levels of CD62L and IL-7Rα.

As we noted that WT ST-OVA displaying increased intracellular proliferation induced greater numbers of OVA-specific CD8⁺ T cells with little cytolytic activity in vitro, we therefore determined whether this was due to the defect in OVA-specific CD8⁺ T cells induced against virulent ST-OVA, or due to some inhibitory activity present in the spleens of ST-OVA-infected mice. At day 15 and 30 post infection, spleen cells from LM-OVA-infected mice mediated potent OVA-specific cytolytic T cell response in contrast to the spleen cells from ST-OVA-infected mice (Fig. 9,A and D). Purified CD8⁺ T cells from ST-OVA-infected mice failed to mediate cytolytic CD8⁺ T cell response indicating a direct impairment in CD8⁺ T cell function (Fig. 9,B and E). Furthermore, spleen cells from ST-OVA-infected mice enhanced the response of CD8⁺ T cells purified from LM-OVA-infected mice (Fig. 9 C and F). Taken together, these results indicate that there is no inhibitory activity in the spleens of ST-OVA-infected mice, and the lack of cytolytic CD8⁺ T cell response in the spleens of ST-OVA-infected mice can be attributed directly to the defect in CD8⁺ T cells.

The next question we addressed was whether the OVA-specific CD8⁺ T cells induced against virulent and mutant ST-OVA would display differences in their ability to proliferate in response to antigenic encounter in vivo. OVA-specific CD8⁺ T cells induced against WT ST-OVA, which were mainly effector in phenotype, displayed poor recall response upon antigenic encounter in vivo (Fig. 10). In contrast, OVA-specific CD8⁺ T cells induced against the ssaR and phoP mutants, which were mainly central in phenotype, displayed potent recall response in response to antigenic encounter in vivo. Taken together, these results indicate that the intracellular proliferation of ST compromises the function of CD8⁺ T cell response.

Although memory CD8⁺ T cells have been considered to play an essential role mainly during viral infection models, they also play an important role in mediating protection against intracellular bacteria such as LM (16, 17, 18), M. tuberculosis (19), and ST (20, 21). Pathogens vary enormously in their mechanisms of virulence (41) and it is unclear how the pathogen virulence influences the generation, maintenance, and function of CD8⁺ T cell memory. Using a murine infection model wherein mice are infected with virulent (WT) ST or with ST in which key virulence genes are deleted, we have addressed the influence of the proliferation of ST on the magnitude and function of CD8⁺ T cell response. Our results indicate that although the most potent CD8⁺ T cell responses are induced only against WT ST, the most functional CD8⁺ T cell responses are induced against ST that do not display massive intracellular proliferation.

Infection of mice with ST provides a stringent model in which the influence of pathogen virulence on the development of acquired immune responses can be experimentally addressed. Infection with very low doses of ST results in complete fatality in C57BL/6J mice. In contrast, 129X1SvJ mice, bearing the same MHC (H-2K^b) are resistant, but they harbor a chronic infection and some mice still succumb to infection in the long term. Furthermore, the host-pathogen interactions during ST infection, and the influence of pathogenicity islands and virulence genes of ST has been extensively studied (27, 42). All the mutants that we have used in this study expressed similar levels of the recombinant protein OVA as determined by Western blotting and qRT-PCR analysis, which indicates that the differences in CD8⁺ T cell responses that we have observed could not be due to disproportionate expression of OVA by mutants.

We used F₁ hybrids between susceptible and resistant mice because they were also resistant to infection and they allowed us to study the fate and phenotype of adoptively transferred OT-1 TCR transgenic CD8⁺ T cells (36). Our results are valid even without the transfer of OT-1 cells, because the numbers of OVA-specific CD8⁺ T cells (measured by ELISPOT assay) generated at day 30 against a dose of 10³ WT ST-OVA were still >10-fold greater than those generated against a 100-fold higher dose of aroA ST-OVA. Furthermore, the functional evaluation of cytolytic CD8⁺ T cell response (Figs. 6 and 8) was done in mice without OT-1 transfer.

Although Ag presentation occurs rapidly within, but not after, the first few days of infection in most models (2, 3), we have previously reported that Ag presentation and consequent generation of OVA-specific CD8⁺ T cell response during infection with WT ST-OVA infection, is greatly delayed (36). Thus, when antibiotics were used to eliminate ST-OVA from day 3 onwards, the development of OVA-specific CD8⁺ T cell response was significantly blunted (Fig. 8) further indicating that Ag presentation during infection with ST-OVA does not occur within the first few days in vivo. Even when dendritic cells or macrophages were pulsed with WT ST-OVA in vitro, they failed to activate CFSE-labeled OT-1 cells within 96 h (<5% of OT-1 cells down-regulated CFSE expression). Similarly, dendritic cells and macrophages pulsed with the various mutants of ST-OVA did not cause any appreciable down-regulation (<5%) of CFSE expression on OT-1 cells. In contrast, when dendritic cells and macrophages were pulsed with LM-OVA, >90% of OT-1 cells down-regulated CFSE expression within 72 h. Similar results were noted if IFN-γ production by OT-1 cells was measured. LM-OVA infection in vitro induced ∼25 ng/ml IFN-γ production by OT-1 cells, whereas WT ST-OVA and the various mutants of ST-OVA failed to elicit any IFN-γ production (<0.1 ng/ml) by OT-1 cells. Considering that LM replicates in the cytosol of infected cells whereas ST replicates in the phagosomes of infected cells, these results indicate that the phagosomal residence of ST may be a deterrent to rapid Ag presentation.

Mice display differences in the relative clearance of various mutants of ST from the spleens. This is expected because different virulence genes would influence the pathogenesis of ST differently (27). It is difficult to know the relative proliferation of the pathogen in vivo based on measurement of bacterial burdens. When the ssaR mutant of ST was injected into mice, there was an increase in bacterial burden in the spleen during the first week of infection, despite failure of the mutant to grow within macrophages in vitro. ST could be persisting without proliferating in vivo, or some ST could be proliferating while others could be dying. Furthermore, ST can also survive and proliferate within epithelial cells, and such cells/bacteria could be trafficking from one organ to the other.

We noted similar CD8⁺ T cell response induced against the aroA- vs the invA-mutant of ST-OVA despite the apparent differences between the virulence of these two mutants, implying that CD8⁺ T cell response may not always correlate to pathogen virulence. In contrast, the CD8⁺ T cell responses correlated very well with the intracellular replication of WT and the various mutants of ST-OVA. It has been previously reported that the intracellular replication of ST is essential for its virulence (22). Pathogen virulence is a highly complex phenomenon, and it is quite likely that pathogen proliferation is one among the many mechanisms that contribute to virulence. In other infection models, virulent mycobacteria were reported to have faster doubling times in vivo (43, 44); however, faster doubling times may not be the only mechanism that enables virulent pathogens to escape the developing immune response. Virulent but not avirulent mycobacteria were shown to selectively overcome the growth inhibitory action of a Th1-dependent, NOS2-independent immune response (45). It has been estimated that ∼4% of the genome of ST is needed for causing fatality in mice (46), suggestive of a highly complex mechanism(s) of virulence.

Our results reveal a dichotomy between the magnitude and function of CD8⁺ T cells as it relates to the proliferation of ST-OVA. Virulent ST-OVA induced maximum numbers of OVA-specific CD8⁺ T cells that displayed poor cytolytic activity, whereas mutants of ST-OVA induced muted numbers of OVA-specific CD8⁺ T cells that displayed potent cytotoxicity. This can be easily explained when one takes into consideration the phenotype of CD8⁺ T cells induced against WT vs avirulent ST-OVA. CD8⁺ T cells induced against WT ST-OVA in contrast to mutants were mainly effector in phenotype. Effectors die in culture, are apoptotic, and proliferate poorly (4, 8, 36). Such cells would therefore not respond efficiently during the 5-day re-stimulation that is used to detect cytotoxicity of CD8⁺ T cells (4). Mainly effector-phenotype CD8⁺ T cells that are induced against virulent ST-OVA produce little IL-2, which would result in their reduced responsiveness (and proliferation) in response to re-stimulation for measurement of cytotoxicity. In contrast, mainly central memory phenotype CD8⁺ T cells that are induced against the mutants produce IL-2, which facilitates their proliferation in response to stimulation in vitro for measurement of cytotoxicity. It has been previously reported that chronic viral infection causes CD8⁺ T cell exhaustion, which manifests in impaired IL-2, cytotoxicity, but not IFN-γ expression (47). Thus, it appears that IFN-γ expression is least influenced by pathogen chronicity.

It is conceivable that the fluorochrome PE-Cy7, which was used in this study, is dissociated and consequently gives signal in the PE channel. We have obtained similar results if PE-Cy7 was not used to stain cells, or if alternative fluorochromes were used to stain cells. Additionally, even when the magnitude of CD8⁺ T cell response was evaluated by non-flow cytometry-based assays (such as ELISPOT), similar results were obtained.

Although the central (CD62L^high) vs effector (CD62L^low) subsets of memory CD8⁺ T cells have been shown to display differences in homing properties and effector function (11, 12), the central subset of memory CD8⁺ T cells has been shown to be most effective in mediating protective efficacy (13). This has been ascribed mainly to the unique ability of central memory CD8⁺ T cells to proliferate (13, 48). Our results are in agreement with this interpretation because infection with the nonvirulent mutants of ST resulted in the generation of mainly central phenotype CD8⁺ T cells, which displayed reduced contraction, higher responsiveness to antigenic encounter, increased IL-2 expression, and cytolytic response in vitro. We have previously reported that Ag presentation during virulent ST infection is prolonged, which explains the persistent effector phenotype of CD8⁺ T cells generated against WT ST (36). Increased intracellular proliferation of WT ST, coupled with the induction of chronic infection, would result in chronic activation of CD8⁺ T cells and preclude the development of central phenotype cells. Indeed, we have previously reported that central phenotype CD8⁺ T cells are generated under conditions of reduced stimulation of naive CD8⁺ T cells (49).

In this report we have shown that the increased intracellular proliferation of ST enhances the magnitude but compromises the function, of CD8⁺ T cells. It is quite likely that this paradigm operates mainly against pathogens that induce a chronic infection. LM is also a virulent intracellular bacterium that induces a potent CD8⁺ T cell response both in terms of magnitude and function. However, because LM induces an infection that lasts only for a few days, this does not impair the function of primed CD8⁺ T cells. In case of ST, as the bacterium causes a chronic infection, chronic Ag presentation may impair the functionality of CD8⁺ T cells (50). Although chronic pathogens have been shown to induce dysfunctional CD8⁺ T cells (50, 51, 52), our results indicate that this may be due to the intracellular proliferation of the pathogen. Increased intracellular proliferation of the pathogen would result in the generation of increased antigenic levels, which may cause a persistent increase in Ag presentation and consequent development of the defective CD8⁺ T cell response.

Attenuated mutants have been considered as vaccine candidates against various diseases (53, 54). However, the emphasis has often been on the measurement of the magnitude of the response, as increased response may translate to better protective efficacy. Our results indicate that evaluation of the quality of the response is more important than the magnitude of the response.

The authors have no financial conflict of interest.