Lack of Expansion of Major Histocompatibility Complex Class Ib-Restricted Effector Cells following Recovery from Secondary Infection with the Intracellular Pathogen Listeria monocytogenes

Abstract

Sublethal infection of BALB/c mice with the intracellular bacterial pathogen Listeria monocytogenes leads to the development of antilisterial immunity with concurrent stimulation of major histocompatibility complex (MHC) class Ia- and Ib-restricted CD8⁺ effector T cells. Secondary L. monocytogenes infection is followed by an accelerated increase in the number of Listeria-specific CD8⁺ cells and rapid clearance of the bacterium from the murine host. Recovery from secondary infection is associated with increased levels of effector cell function, as measured by gamma interferon secretion following coculture of immune cells with L. monocytogenes infected APCs in vitro, as well as antilisterial cytotoxicity, as measured by effector cell recognition of L. monocytogenes-infected target cells. We assessed the frequency of L. monocytogenes-specific MHC class I-restricted cells following secondary infection by ELISPOT assays utilizing coculture of immune cells with L. monocytogenes-infected antigen-presenting cells that express MHC class Ia and/or Ib molecules. We found that the antilisterial Qa-1^b (MHC class Ib)-restricted effector subset is not detected as an expanded population following secondary infection compared to the frequency of this effector population as measured following recovery from primary infection. This is in contrast to the frequency of antilisterial H2-K^d (MHC class Ia)-restricted effector cells, which following recovery from secondary infection are detected as an expanded population, and appears to undergo a substantial expansion event 3 to 4 days post-secondary infection. These results are consistent with the conclusion that although Listeria-specific MHC class Ib-restricted effector cells are present following recovery from secondary infection, this subset does not appear to undergo the expansion phase that is detected for the MHC class Ia-restricted effector cell response.

Induction of a protective immune response against the intracytoplasmic pathogen Listeria monocytogenes is apparent following subclinical infection with viable hemolysin-secreting strains. Nonviable L. monocytogenes, as well as non-hemolysin-secreting mutant strains of this pathogen, is avirulent and does not trigger protective antilisterial immunity (1, 2, 11, 17). These findings are consistent with the observation that L. monocytogenes escapes the phagocytic vacuole due to secretion of the pore-forming hemolysin listeriolysin O (LLO) and then replicates within the cytoplasm of the host cell. It has been established that protective antilisterial immunity can be adoptively transferred with major histocompatibility complex (MHC) class I-restricted CD8⁺ T-cells (3, 13), an observation immunologically consistent for a pathogen that can exist and replicate within the cytoplasm of the infected cell. As a facultative intracellular pathogen, L. monocytogenes can infect and then replicate within professional MHC class II-positive phagocytes, as well as within MHC class II-negative cells, such as fibroblasts (21). Thus, a required involvement of the MHC class I-restricted CD8⁺ T-cell subset is consistent with clearance of a pathogen which can infect and survive within MHC class II-negative cells, since MHC class II-negative cells are not typically armed with the necessary mechanisms to kill intracellular bacteria.

L. monocytogenes-specific CD8⁺ T cells include effector cytotoxic T lymphocytes (CTL) restricted by both MHC class Ia and Ib molecules. For the MHC class Ia-restricted component, H2-K^d has been identified as a restricting element; in contrast, H2-D^d and H2-L^d molecules do not appear to be restricting elements for Listeria-specific CD8⁺ T cells (9, 18, 19). Four H2-K^d presented nonamer target peptides have been identified, corresponding to amino acids 91 to 99 of the L. monocytogenes-derived LLO, amino acids 217 to 225 and 449 to 457 of the murien hydrolase p60, and amino acids 84 to 92 of the L. monocytogenes-derived metalloprotease. A common feature of these proteins is that they are secreted products of the bacterium, thus facilitating entry of these molecules into the endogenous antigen processing pathway for ultimate MHC class I presentation. The MHC class Ib molecules that have been identified as restricting elements for Listeria-specific CD8⁺ T cells include both H2-M3 and Qa-1^b (7, 20). H2-M3-presented targets include peptides of shorter length (as short as five amino acids), contain a formylated methionine at the N terminus, and are derived from both structural and secreted proteins (12, 15, 22). A Qa-1^b associated target peptide has yet to be identified; however, a candidate peptide derived from the secreted LLO molecule awaits characterization (6).

Primary immunization with L. monocytogenes results in an initial period (the first 3 to 4 days) of bacterial replication, followed by a decline in CFU with sterilizing immunity by days 7 to 8 postinfection (4). Animals that have recovered from the initial infection exhibit long-term antilisterial immunity, and immune animals given a second infection of L. monocytogenes quickly eliminate the bacterium. It is clear that MHC class I-restricted memory cells are rapidly activated and that, following recovery, this subset is increased numerically in numbers. Analyses of CTL frequencies of H2-K^d-restricted cells (limiting dilution, ELISPOT, and tetramer staining) indicate that this subset expands 5- to 20-fold following recovery from reinfection (8, 14). However, similar analyses of the H2-M3-restricted response show that the frequency of peptide-specific H2-M3-restricted cells does not increase as a consequence of the recall response. This suggests a dichotomy of recall responses for MHC class Ia- and Ib-restricted cells. We show here that, although Qa-1^b-restricted cells are evident following primary infection with L. monocytogenes, this component of the MHC class I-restricted pool does not appear to expand in numbers following secondary infection with L. monocytogenes. Thus, the lack of memory expansion as previously reported for H2-M3-restricted cells is also evident for the Qa-1^b-restricted effector subset, an observation that supports the premise that MHC class Ib-restricted cells do not possess the same capacity to undergo memory cell expansion that is characteristic of the MHC class Ia-restricted response in antilisterial immunity.

MATERIALS AND METHODS

Bacteria.

L. monocytogenes 10403 serotype 1 was utilized as the immunogen for these studies. Virulence was maintained by repeated passages in BALB/c mice.

Mice and immunization.

Six-week-old female BALB/c mice were purchased from Jackson Labs (Bar Harbor, Maine) and provided unrestricted access to food and water. Eight-week-old BALB/c mice were immunized by injection via the tail vein with approximately 400 CFU of viable L. monocytogenes in 0.2 ml of phosphate-buffered saline (PBS). For the secondary immunization, mice were given approximately 4,000 CFU in 0.2 ml of PBS by the same route.

Cell lines and reagents.

The J774 cell line was maintained by culture in Dulbecco modified Eagle medium (DMEM; Gibco, antibiotic-free) supplemented with nonessential amino acids (Gibco) and 5% fetal calf serum (FCS; Tissue Culture Biologicals, Tulare, Calif.). The LtK, Lg37 (Qa-1^b-expressing), and LtK-K^d (K^d-expressing) cell lines, kindly provided by James Forman (University of Texas Southwestern Medical School), were maintained in DMEM supplemented with 10% FCS (7).

Stimulation of immune cells for assessment of cytokine secretion.

Sixteen-to-eighteen hours before infection, 1.5 × 10⁷ J774 cells were added to a 25-cm² tissue culture flasks in 10 ml of antibiotic-free DMEM plus 10% FCS. At the time of the infection, 7 ml of medium was removed, and the J774 cells were infected with log-phase cultures of wild-type L. monocytogenes at a multiplicity of infection (MOI) of 2 to 5 or an LLO⁻ strain of L. monocytogenes (DP-L215 [7]) at an MOI of 5 to 10. The J774 monolayers were washed with warm (37°C) PBS 1 h later, and 5 ml of DMEM supplemented with 5% FCS and 40 μg of gentamicin per ml was added to each flask. Two hours later, the J774 cells were collected from the flasks, washed once, and then added to 24-well plates at 10⁵ cells/well in 0.5 ml of DMEM with 5% FCS and gentamicin. Spleen cells from mice immunized 10 days previously with L. monocytogenes (either as a primary or secondary injection) were obtained, and a single cell suspension was prepared and added to the 24-well plates in 0.5 ml of DMEM supplemented with 5% FCS and 100 U of penicillin and 100 μg of streptomycin per ml at effector/target (E/T) ratios ranging from 20:1 to 5:1. The cells were cultured for 24 h, after which the supernatants were collected and stored frozen (−80°C) for later analyses of the relative concentrations of gamma interferon (IFN-γ). The levels of IFN-γ were assessed using mouse IFN-γ enzyme-linked immunosorbent assay (ELISA) kits from Biosource (Camarillo, Calif.).

CFU reduction assay.

J774 target cells were deposited, at 10⁵ cells/well, in 24-well tissue culture plates in 1.0 ml of antibiotic-free DMEM supplemented with nonessential amino acids and 5% FCS 18 h prior to infection with L. monocytogenes (obtained from a log-phase culture) at an MOI of 2 to 5 (2, 5). At 60 min after infection, the monolayers were washed twice with sterile 37°C PBS and covered with 0.5 ml of DMEM containing 5% FCS and 80 μg of gentamicin sulfate per ml. Pooled spleen cells from similarly immunized mice within a group that were stimulated in vitro for 72 h with viable L. monocytogenes were used as effector cells (5) and were added in 0.5 ml of DMEM with 5% FCS at 3 to 4 h after initiation of the infection. The assays were terminated 4 h later, and the numbers of intracellular bacteria remaining in each well were determined. Specifically, the medium was aspirated and replaced with 1 ml of distilled water. Five minutes later, dilutions were plated onto brain heart infusion agar plates that were incubated for 24 h at 37°C, and the numbers of CFU were determined. The data are presented as follows: the percentage of CFU reduction = [1 − (CFU in target monolayers incubated with effector cells)/(CFU in target monolayers incubated without effector cells)] × 100.

ELISPOT analysis for enumeration of IFN-γ-secreting cells.

At 16 to 18 h before use, 96-well nitrocellulose plates (Millipore, Bedford, Mass.) were coated with 100 to 500 ng of anti-mouse IFN-γ capture antibody (Pharmingen, San Diego, Calif.) per well diluted in PBS and added in a volume of 100 μl. At 1 h prior to use the plates were washed with sterile medium or sterile PBS and then blocked with cell culture medium (RPMI or DMEM) containing 5 to 10% FCS.

For enumeration of L. monocytogenes reactive effector cells 3 × 10⁶ Ltk, Lg37, LtK-K^d, or J774 cells were cultured in 10 ml of antibiotic-free DMEM supplemented with 10% FCS in petri dishes (Falcon 1029 plates) the day before infection. On the day of the assay, the cells were infected with L. monocytogenes (log-phase culture, MOIs of 2 to 5:1 for J774 cells and 10:1 for the LtK, LtK-K^d, and Lg37 cells). After 1 h of infection, the monolayers were washed once with room temperature PBS and then cultured in 10 ml of DMEM supplemented with 10% FCS and 10 μg of gentamicin per ml. After an additional 3 h, the monolayers were washed once with room temperature PBS, and then 5 ml of cold PBS (4°C) was added. The cells were placed at 4°C for 5 min, and then the cells were removed by aspiration with a pipette and the cell recoveries were determined. The infected or noninfected cells were then added to ELISPOT plates at 40,000 cell/well in a volume of 100 μl in DMEM supplemented with 10% FCS.

Single-cell suspensions of immune cells from mice previously immunized with a primary or a secondary infection of L. monocytogenes were prepared and added to the ELISPOT plates at 25,000 to 100,000 cells/well containing the L. monocytogenes-infected target cells in the presence of 200 U of penicillin and 200 μg of streptomycin per ml and 60 U of human recombinant IL-2. After 24 h of incubation at 37°C the plates were washed four times with 0.05% Tween 20-PBS, and a biotinylated anti-mouse IFN-γ detection antibody (Pharmingen) was added at 500 ng/well in a volume of 100 μl. The plates were then incubated overnight at 4°C and washed four times with 0.05% Tween 20-PBS, and 100 μl of a 1:1,000 dilution of streptavidin AKP (Pharmingen) was then added. After 1 h at room temperature, the plates are washed four times with 0.05% Tween 20-PBS, and then 200 μl of the detection substrate BCIP (5-bromo-4-chloro-3-indolylphosphate)-nitroblue tetrazolium (KPL, Gaithersberg, Md.) was added to each well. After 5 to 20 min, the plates were washed with distilled H₂O and allowed to dry. Spots were enumerated with a Zeiss microscopy unit equipped with KElispot software (Zeiss).

RESULTS

The antilisterial recall response is associated with increased levels of cytokine secretion and CTL activity.

Injection of BALB/c mice with a subclinical dose L. monocytogenes results in uncontrolled replication for the first 48 to 72 h, after which bacterial numbers begin to decline, with sterilizing immunity evident by days 7 to 8 (4). Immune mice given a second injection with L. monocytogenes rapidly clear the bacterium, with essentially sterilizing immunity by 48 h (data not shown). This accelerated elimination of L. monocytogenes is due in part to the recall response within the immune CD8⁺ T-cell compartment. In order to determine if this rapid elimination of L. monocytogenes is associated with increased antilisterial effector cell activity within the splenic population, mice were immunized with L. monocytogenes and injected again 30 days later with L. monocytogenes. The levels of cytokine production and cytotoxic activity exhibited by spleen cells obtained from these mice were compared to those of immune spleen cells obtained from mice following a primary infection.

To assess the levels of cytokine secretion by L. monocytogenes-immune populations, donor spleen cells were cocultured with L. monocytogenes-infected J774 target cells as antigen-presenting cells (APC), the supernatant was collected 24 h later, and the levels of IFN-γ secretion in the supernatants were measured by ELISA. The experiments (Fig. 1, left panel) showed that coculture of immune cells from mice that received two injections of L. monocytogenes secrete increased levels of IFN-γ compared to similarly cocultured immune cells obtained from mice that received only a single L. monocytogenes injection. This is most apparent at E/T ratios of 10:1 and 5:1, since immune spleen cells from mice receiving a primary infection show little if any IFN-γ secretion at these E/T ratios. These results are indicative of an increased frequency of IFN-γ-secreting effector cells present in immune spleen cell populations obtained from mice that have recovered from a secondary L. monocytogenes infection.

FIG. 1.

The level of INF-γ secretion by immune spleen cells is enhanced following secondary infection with L. monocytogenes. BALB/c mice were immunized with approximately 400 CFU of L. monocytogenes and then 30 days later injected with approximately 4,000 CFU of L. monocytogenes. On day 10 following a primary or secondary injection, the immune spleen cells were obtained and cocultured with 10⁵ J774 cells infected with wild-type L. monocytogenes (left panel) or with an LLO⁻ strain of L. monocytogenes (right panel), in duplicate, at the indicated E/T ratios. After 24 h, the supernatants were collected, and the levels of IFN-γ were determined by ELISA. The data are representative of three independent experiments.

As a control, the same populations of immune spleen cells were cocultured with J774 cells infected with a LLO⁻ strain of L. monocytogenes. This LLO⁻ strain of L. monocytogenes is unable to escape to the cytoplasm of the infected cell and thus remains within a membrane-bound vacuole (21). J774 cells are not bacteriocidal and do not kill the phagosome-imprisoned bacteria. The results presented in Fig. 1 (right panel) show that, upon coculture of J774 target cells infected with an LLO⁻ strain of L. monocytogenes with immune cells obtained 10 days following primary immunization, they do not secrete appreciable levels of IFN-γ at all of the E/T ratios tested. Coculture of J774 target cells infected with an LLO⁻ strain of L. monocytogenes with immune cells obtained 10 days following secondary immunization leads to a low level of IFN-γ at an E/T of 20:1; however, at an E/T of 10:1, IFN-γ is detected at background levels.

The data presented in Fig. 1 show enhanced levels of IFN-γ secretion with immune cells obtained from animals given a secondary L. monocytogenes infection. In order to determine if these populations also exhibit enhanced levels of L. monocytogenes-specific CTL activity, immune cells obtained from mice injected either once or twice with L. monocytogenes were stimulated in vitro, and the CTL activity of the recovered cells was assessed against L. monocytogenes-infected targets in a standard CFU reduction assay (2, 5). The results of these experiments (Fig. 2) showed that, following primary immunization, E/T ratios of 25:1 and 3:1 resulted in 55 and 10% CFU reduction, respectively, in the infected target cell population. In contrast, culture-stimulated immune cells obtained from mice injected twice with L. monocytogenes caused 90 and 50% CFU reduction in this same target population at E/T ratios of 25:1 and 1.5:1, respectively. The CTL activity of these same immune populations was also assessed against target cells infected with the LLO⁻ strain of L. monocytogenes, and no reduction in CFU was observed at all E/T ratios tested (data not shown). These results demonstrate that the magnitude of the cytolytic response is enhanced following secondary injection with L. monocytogenes. These results are in support of data shown in Fig. 1 showing an increased frequency of antilisterial effector cells in immune spleen cell populations obtained from mice that have recovered from a secondary L. monocytogenes infection.

FIG. 2.

Antilisterial CTL activity of immune spleen cells is enhanced following secondary infection with L. monocytogenes. BALB/c mice were immunized with approximately 400 CFU of L. monocytogenes and then 30 days later injected with approximately 4,000 CFU of L. monocytogenes. Immune spleen cells, obtained 10 days following either a primary (hatched bars) or secondary (solid bars) infection were stimulated in culture with wild-type L. monocytogenes. After 72 h of culture, the antilisterial CTL activity of the recovered cell was assessed by a CFU reduction assay as described in Materials and Methods. The data are representative of five independent experiments.

The frequency of L. monocytogenes specific effector cells is increased following recovery from secondary infection with L. monocytogenes.

The results presented above show that the antilisterial effector function exhibited by immune cells, as measured by either IFN-γ secretion or cytolytic activity, is clearly enhanced following secondary infection with L. monocytogenes. We next sought to determine if the frequency of L. monocytogenes-specific effector cells are enhanced following recovery from secondary infection; for these studies, immune spleen cells from mice infected once or twice with L. monocytogenes were stimulated in the presence of L. monocytogenes-infected J774 cells, and the numbers of IFN-γ-secreting cells were measured in an ELISPOT assay. Control target cells included J774 cells infected with the LLO⁻ strain of L. monocytogenes, as well as noninfected J774 cells. The data presented in Fig. 3 show that the frequency of L. monocytogenes-specific IFN-γ-secreting cells as assessed 10 days following a single injection with L. monocytogenes is approximately 75 per 50,000 immune cells. In contrast, 10 days following a second injection with L. monocytogenes the frequency of IFN-γ-secreting cells increases approximately fourfold to 300 per 50,000 immune cells. Coculture of these immune cells with noninfected J774 cells or with J774 cells infected with the LLO⁻ strain of L. monocytogenes did not result in IFN-γ secretion as assessed by ELISPOT assay.

FIG. 3.

The frequency of antilisterial specific IFN-γ-secreting cells is increased following secondary L. monocytogenes injection. BALB/c mice were immunized with approximately 400 CFU of L. monocytogenes and then 30 days later injected with approximately 4,000 CFU L. monocytogenes. Immune spleen cells, obtained 10 days following either a primary or secondary injection, were cocultured with J774 cells infected with either wild-type L. monocytogenes (solid bars) or an LLO⁻ mutant of L. monocytogenes (shaded bars) or left uninfected (cross-hatched bars) in an ELISPOT assay. After 24 h of culture, the numbers of IFN-γ-secreting cells were determined as described in Materials and Methods. The data are representative of three independent experiments.

The frequency of Qa-1^b-restricted effector cells is not enhanced following recovery from secondary infection with L. monocytogenes.

The data presented in Fig. 3 shows that the frequency of IFN-γ-secreting effector cells is increased following recovery from secondary infection with L. monocytogenes. Although the MHC class I-restricted pool contains both Ia- and Ib-restricted subsets, it has recently been shown that the H2-M3 (MHC class Ib)-restricted cells do not expand following stimulation provided by a secondary L. monocytogenes infection (14). Thus, we were interested in determining if the Qa-1^b-restricted component of the MHC class Ib-restricted pool is or is not expanded under these same conditions. In order to assess the frequency of the Qa-1^b-restricted effector response following L. monocytogenes injection, we utilized L. monocytogenes-infected Qa-1^b-expressing Lg37 fibroblasts as the APC in ELISPOT assays. We have demonstrated previously that L. monocytogenes-infected Qa-1^b-expressing Lg37 target cells are lysed by L. monocytogenes-immune effectors, while L. monocytogenes-infected parent LtK cells (which do not express Qa-1^b) were not lysed (7). Thus, the use of L. monocytogenes-infected Qa-1^b-expressing Lg37 cells or the parent LtK cells as APC in the ELISPOT assay would allow the differential determination of the Qa-1^b-restricted response. BALB/c mice were immunized with L. monocytogenes and 30 days later were given a secondary injection. Ten days later the spleen cells were collected and assessed for a Qa-1^b-specific effector response by ELISPOT assay. The data presented in Fig. 4 show that 10 days following a primary L. monocytogenes immunization, the number of Qa-1^b-restricted IFN-γ-secreting cells is approximately 50 per 50,000 L. monocytogenes immune spleen cells. Furthermore, 10 days following secondary L. monocytogenes injection, the frequency of Qa-1^b-restricted cells remains approximately 50 per 50,000 recovered cells. This frequency of Qa-1^b-restricted cells 10 days following secondary infection is also similar to that observed in mice 40 days following primary immunization. These data show that Qa-1^b-restricted cells do not disappear following clearance of the bacterium after primary immunization and, in addition, this subset is not increased in frequency following the recall antilisterial response. Immune spleen cells cocultured in the presence of L. monocytogenes-infected Ltk cells (which do not express Qa-1^b) or noninfected Lg37 cells do not secrete IFN-γ as measured by the ELISPOT assay.

FIG. 4.

The frequency of antilisterial Qa-1^b-restricted cells is not increased following recovery from secondary L. monocytogenes infection. BALB/c mice were immunized with approximately 400 CFU of L. monocytogenes and then 30 days later injected with approximately 4,000 CFU of L. monocytogenes. On day 40 following a primary injection, day 10 following a primary injection, or day 10 following a secondary injection immune spleen cells were cocultured with L. monocytogenes-infected Ltk cells (solid bars), Lg37 cells (shaded bars), or noninfected Lg37 cells (cross-hatched bars) in ELISPOT assays. After 24 h, the numbers of IFN-γ-secreting cells were determined as described in Material and Methods. The data are representative of three independent experiments.

This result for Qa-1^b-restricted cells (MHC class Ib response) is in sharp contrast to the finding that MHC class Ia-restricted responses (as measured with peptide-pulsed targets or tetramer staining) are increased following recovery from secondary L. monocytogenes infection (10, 14). In order to assure that the lack of any observed expansion of the Qa-1^b-restricted subset is not due to the nature of the L. monocytogenes-infected transfected L cells, immune spleen cells from mice injected either once or twice with L. monocytogenes were cocultured with L. monocytogenes-infected L cells expressing K^d (LtK-K^d) as the APC in ELISPOT assays. In experiments utilizing this infected APC population, we found that the frequency of H2-K^d-restricted IFN-γ-secreting effector cells was increased at least threefold when assessed 10 days following secondary infection compared to the frequency of H2-K^d-restricted cells obtained 10 days following a primary infection (data not shown). Thus, the lack of expansion as detected for the Qa-1^b-restricted population is not attributable to the nature of the L. monocytogenes-infected L-cell targets utilized as APC in the ELISPOT assays.

Kinetics of H2-K^d-and Qa-1^b-restricted responses following secondary L. monocytogenes infection.

The results presented in Fig. 4 show that following recovery from secondary L. monocytogenes infection, the Qa-1^b-restricted subset is not detected as an expanded population. It is possible that following secondary infection, the Qa-1^b subset does in fact expand and then decreases in number by day 10 following infection; thus, any expansion would not be detected in the experiments as described here. Previous studies assessing the kinetics of MHC class I-restricted responses following secondary infection have found that the H2-K^d-restricted component is detected as an expanded population 3 to 5 days post-secondary infection, whereas the H2-M3-restricted pool does not expand during this or any time post-secondary infection (10, 14). Thus, it was of interest to assess whether the Qa-1^b-restricted subset can be detected as an expanded population at earlier time points following secondary infection. In order to assess the kinetics of the Qa-1^b-restricted subset following secondary infection, BALB/c mice were immunized with L. monocytogenes and then 30 days later given a secondary injection. Immune spleen cells were collected at several time points post-secondary infection, and the frequencies of Qa-1^b-restricted and H2-K^d-restricted effector cells were assessed by ELISPOT assays using L. monocytogenes-infected targets as APC. The data presented in Fig. 5 show that at 3 days post-secondary infection H2-K^d-restricted effector cells are detected and that at 4 days post-secondary infection the H2-K^d-restricted subset is present as an expanded population, measured as a fivefold increase. This observation is in contrast to the Qa-1^b-restricted subset, which can also be detected 3 days post-secondary infection, and at 4 days post-secondary infection the Qa-1^b-restricted subset has not increased in frequency. The Qa-1^b-restricted subset is also not increased on day 5 or 7 post-secondary infection, whereas the H2-K^d-restricted subset continues to be detected as a numerically increased population (data not shown). Thus, following secondary infection, the Qa-1^b-restricted T-cell population does not increase numerically, an observation distinct from that for the T-cell population responding to H2-K^d-presented antigens.

FIG. 5.

H2-K^d- but not Qa-1^b-restricted responses increase in frequency following secondary L. monocytogenes infection. BALB/c mice were immunized with approximately 400 CFU of L. monocytogenes and then 30 days later injected with approximately 4,000 CFU of L. monocytogenes. On the days indicated post-secondary infection, immune spleen cells were cocultured with L. monocytogenes-infected Ltk-K^d cells (K^d expressing) or Lg37 cells (Qa-1^b expressing) in ELISPOT assays. After 24 h, the numbers of IFN-γ-secreting cells were determined as described in Materials and Methods.

DISCUSSION

Injection of BALB/c mice with a sublethal dose of L. monocytogenes leads to the development of cell-mediated immunity that ultimately results in the eradication of the bacterium. Upon subsequent infection, the anti-Listeria specific memory cells that are present are expanded and activated upon reinfection, thus facilitating the rapid elimination of this pathogen (10). The results presented here are in support of these findings, since the antilisterial effector response is increased in magnitude (as assessed by measuring the release of proinflammatory cytokines, as well as measuring the CTL activity) following secondary infection with L. monocytogenes compared to the activity of immune cells obtained following primary immunization. Although the H2-K^d-restricted component of the antilisterial immune response is markedly enhanced following a recall response (at least fivefold), the results presented here show that the antilisterial Qa-1^b-restricted subset of the MHC class I-restricted pool is not increased following such a recall response. This finding suggests a basic difference in the nature of the Ia-versus Ib-restricted recall responses within the MHC class I-restricted pool of antilisterial effector cells.

A general role for MHC class Ib-restricted effector cells within the cell-mediated immune response against viral or bacterial pathogens has yet to be determined. MHC class Ib-restricted effector cells are clearly stimulated as a consequence of infection with L. monocytogenes, with H2-M3 and Qa-1^b shown to be restricting elements (7, 20). Mice lacking MHC class Ia molecules are reported to resolve a L. monocytogenes infection, indicating that MHC class Ib-restricted effectors can mediate the clearance of this intracellular bacterial pathogen (24). Further studies revealed that L. monocytogenes immune CD8⁺ cells from MHC class Ia-deficient donors adoptively transferred protection to syngeneic recipients. However, the relevant contribution of isolated H2-M3- or Qa-1^b-restricted cell populations to the adoptively transferred protective response remains to be addressed. Thus, the degree to which H2-M3- or Qa-1^b-restricted subsets independently mediate bacterial clearance in vivo is presently unknown. An in vivo role for bacterial clearance mediated by H2-M3-restricted effector cells is suggested in a recent report showing that the numbers of L. monocytogenes CFU are diminished in recipients of an adoptively transferred peptide-specific H2-M3-restricted clone when compared to controls (23). However, in this study, immunization with the H2-M3-restricted peptide does not protect against L. monocytogenes infection; thus, an in vivo role for the H2-M3-restricted subset remains unclear.

Recently, it was reported that MHC class Ib-restricted CTL are stimulated following experimental infection with Salmonella enterica serovar Typhimurium in a murine model (16). Direct ex vivo analysis demonstrated that MHC class Ib-restricted CTL activity can be detected 1 week following the challenge of vaccinated mice with virulent serovar Typhimurium. Additional studies defined Qa-1^b as the relevant MHC class Ib-restricting element and revealed that the Qa-1^b-restricted subset contributed approximately 50% of the total MHC class I-restricted pool. As a pathogen, serovar Typhimurium remains membrane bound within the host cell, and whether this contributes to the apparent high frequency of Qa-1^b-restricted cells remains to be studied. Although serovar Typhimurium-specific Qa-1^b-restricted cell lines were established in this study, the in vivo activity of these Qa-1^b-restricted populations was not reported. These studies collectively suggest that MHC class Ib-restricted effector cells are important for the elimination of intracellular bacterial pathogens. However, definitive adoptive transfer studies showing the precise contribution of MHC class Ib-restricted cells to antibacterial immunity await completion.

An H2-M3 tetramer reagent has recently been produced with the Listeria-derived peptide MIGWIIA (14). Using this reagent, H2-M3-restricted cells appear to be a major component of the response following the initial infection, with the numbers of H2-M3-restricted tetramer-positive CD8⁺ cells in the spleen being approximately 3%. In that study it was established that the H2-M3-restricted cell subset does not appear to expand during or following clearance of the secondary infection inoculum. It was also established that H2-M3-restricted CTL could not be detected by direct ex vivo CTL assays during or following clearance of the reinfection inoculum. Thus, in contrast to what is observed for the K^d-restricted antilisterial response, H2-M3-restricted cells do not appear as an expanded effector population as a consequence of the recall response.

The original studies from the Pamer lab showing that the H2-M3-restricted component of the MHC class Ib pool is not expanded following recovery from a secondary infection with L. monocytogenes infection (14) prompted an assessment of whether this observation described a property of MHC class Ib-restricted cells in general or is unique to the H2-M3-restricted subset. We have previously determined that Qa-1^b-restricted cells are also a component of the MHC class Ib-restricted pool (7); thus, we assessed the fate of this subset following recovery from a secondary infection. For the analysis of the frequencies of Qa-1^b-restricted cells following primary or secondary infection, we utilized an ELISPOT assay with Listeria-infected Qa-1^b-expressing Lg37 line cells as the APC. (The antilisterial Qa-1^b-presented peptide has not been identified as yet, thus precluding any tetramer-based assays.) The ELISPOT assay measures functional cytokine-secreting cells and thus allows a direct enumeration of Qa-1^b -restricted effector cells. Although we found that Qa-1^b-restricted cells are detected at 40 days following immunization, this MHC class Ib-restricted population does not increase in frequency as a consequence of a secondary infection. Furthermore, the results presented here on the kinetic development of MHC class I-restricted effector cells show that, at the time the H2-K^d-restricted component has expanded in frequency in response to the secondary infection, the frequency of the Qa-1^b-restricted component has not changed (Fig. 5). These data for antilisterial Qa-1^b-restricted responses, as detected with L. monocytogenes-infected targets as the APC in ELISPOT assays, are consistent with the nature of the H2-M3-restricted antilisterial responses as defined by tetramer staining. The data presented here utilized BALB/c mice and, in a preliminary study, we have found that antilisterial Qa-1^b-restricted cells derived from C57BL/6 mice also do not increase in frequency following reinfection with L. monocytogenes (data not shown). We are currently assessing whether the H2-K^b-restricted component of the MHC class I response in the C57BL/6 strain undergoes similar levels of expansion as that seen for H2-K^d-restricted cells in BALB/c mice.

The H2-M3-restricted response in BALB/c mice is of a lesser magnitude compared to H2-M3-restricted antilisterial response that develops in C57BL/6 mice (18). In our studies with BALB/c-derived antilisterial effector cells and infected bone marrow macrophage from congenic strains as targets, we have been unable to detect an H2-M3-restricted CTL response (16). The data presented here is in further support of this apparent strain difference. The transfected L cells utilized as the APC population for ELISPOT analysis express H2-M3, and yet we were not able to detect the presence of an H2-M3-restricted antilisterial effector. We did not observe any increase in the frequency of IFN-γ-secreting effector cells following coculture of immune cells with L. monocytogenes-infected parent L cells (LtK) compared to the noninfected Qa-1^b-expressing L-cell controls (Fig. 4). If H2-M3-restricted cells were present, we would have expected a greater number of IFN-γ-secreting effector cells with immune cells cocultured with L. monocytogenes-infected L cells. Thus, the data presented here are consistent with previous reported results showing the H2-M3 subset to be minimal in the BALB/c strain.

The observation that MHC class Ib-restricted effectors do not appear to undergo memory cell expansion may reflect (i) the inability to further stimulate MHC class Ib memory cells following infection or (ii) that the rapid clearance of L. monocytogenes during the recall response does not allow for rapid expansion of the MHC class Ib-restricted subset (14). The results presented in Fig. 4 showing that L. monocytogenes-specific Qa-1^b-restricted cells are evident 40 days post-primary immunization suggest persistence of this subset of MHC class Ib-restricted cells. In addition, we can find MHC class Ib-restricted effector CTL by direct ex vivo analysis of immune spleen cells during the recovery phase following reinfection with L. monocytogenes, with CTL activity detected between days 2 and 3 postreinfection (data not shown). This ex vivo CTL activity correlates directly with the decline in numbers of L. monocytogenes CFU in the spleens of immune mice. We are currently assessing the contribution of the MHC class Ib-restricted subsets to this direct ex vivo CTL response. Collectively, our observations would not argue strongly for either of the proposed models. An alternative possibility is one in which the L. monocytogenes-infected APC may participate in determining the development of memory cells. For example, dendritic cells are well documented to be potent stimulators of effector and memory responses (25). How this APC population functions regarding the stimulation of L. monocytogenes effector and memory responses for MHC class Ib-restricted responses is currently under investigation.

Subunit-based vaccines that target MHC class Ib-restricted responses may be of benefit to a broad range of individuals, since MHC class Ib molecules are much less polymorphic. Since MHC class Ib-restricted effector cells have been shown in a number of model systems, it may be reasonable to exploit this arm of the MHC class I-restricted repertoire as a vaccine strategy. However, recently published data, as well as data presented here, would suggest a more cautious approach regarding MHC class Ib-targeted vaccines since it is not clear if the memory component of an existing MHC class Ib-directed response is able to expand following antigen reencounter and thereby assist in either short- or long-term disease prevention.