Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Dendritic Cell Cross-Priming Is Essential for Immune Responses to Listeria monocytogenes

  • Anna T. Reinicke,

    Affiliation: The Biomedical Research Centre, Michael Smith Laboratories, Departments of Medical Genetics, Microbiology and Immunology, and Zoology, The University of British Columbia, Vancouver, British Columbia, Canada

  • Kyla D. Omilusik,

    Affiliation: The Biomedical Research Centre, Michael Smith Laboratories, Departments of Medical Genetics, Microbiology and Immunology, and Zoology, The University of British Columbia, Vancouver, British Columbia, Canada

  • Genc Basha,

    Affiliation: The Biomedical Research Centre, Michael Smith Laboratories, Departments of Medical Genetics, Microbiology and Immunology, and Zoology, The University of British Columbia, Vancouver, British Columbia, Canada

  • Wilfred A. Jefferies

    wilf@brc.ubc.ca

    Affiliation: The Biomedical Research Centre, Michael Smith Laboratories, Departments of Medical Genetics, Microbiology and Immunology, and Zoology, The University of British Columbia, Vancouver, British Columbia, Canada

Dendritic Cell Cross-Priming Is Essential for Immune Responses to Listeria monocytogenes

  • Anna T. Reinicke, 
  • Kyla D. Omilusik, 
  • Genc Basha, 
  • Wilfred A. Jefferies
PLOS
x

Abstract

Cross-presentation is now recognized as a major mechanism for initiating CD8 T cell responses to virus and tumor antigens in vivo. It provides an elegant mechanism that allows relatively few Dendritic cells (DCs) to initiate primary immune responses while avoiding the consumptive nature of pathogenic infection. CD8 T cells play a major role in anti-bacterial immune responses; however, the contribution of cross-presentation for priming CD8 T cell responses to bacteria, in vivo, is not well established. Listeria monocytogenes (Listeria) is the causative agent of Listeriosis, an opportunistic food-borne bacterial infection that poses a significant public health risk. Here, we employ a transgenic mouse model in which cross-presentation is uniquely inactivated, to investigate cross-priming during primary Listeria infection. We show that cross-priming deficient mice are severely compromised in their ability to generate antigen-specific T cells to stimulate MHC I-restricted CTL responses following Listeria infection. The defect in generation of Listeria-elicited CD8 T cell responses is also apparent in vitro. However, in this setting, the endogenous route of processing Listeria-derived antigens is predominant. This reveals a new experimental dichotomy whereby functional sampling of Listeria-derived antigens in vivo but not in vitro is dependent on cross-presentation of exogenously derived antigen. Thus, under normal physiological circumstances, cross-presentation is demonstrated to play an essential role in priming CD8 T cell responses to bacteria.

Introduction

Antigen processing and subsequent presentation to specific T cells is necessary for effective bacterial immune responses involving both CD4 and CD8 T cells. While CD4 helper T cell responses are necessary for limiting infection, CD8 cytotoxic T lymphocytes are required for clearance of bacteria [1]. MHC class-II antigen presentation is mainly restricted to internalized proteins that are processed and loaded in phagosomal/endosomal compartments for activation of CD4 T helper cells [2]. MHC I presentation is classically restricted to endogenously synthesized antigens of viral or self origin [3]. Such proteins are digested by the proteasome and translocated into the endoplasmic reticulum (ER) via the transporters associated with antigen processing, TAP1 and TAP2 for loading on MHC I. A second, less well-defined mechanism of processing has also been described and is referred to as cross-presentation whereby exogenously-derived antigens may be captured from other sources and following uptake are processed in such a way to access MHC I molecules for loading and presentation [4].

There are several mechanisms proposed to explain how exogenously-derived antigens that have been phagocytosed or macropinocytosed may be processed and loaded on MHC I molecules for presentation. In one model, exogenous antigens are proposed to be processed directly in a compartment within the endocytic pathway in a TAP and proteosome independent manner where they are degraded by proteases and cathepsins for loading [5][7]. In a second model, internalized antigens may be delivered to the cytosol where they can follow the classical proteosome-mediated degradation and entry to the ER via TAP for loading on nascent MHC I molecules [8], [9]. Membrane of the ER has been proposed to fuse with the phagosome during phagocytosis [10] that would allow the delivery of antigens into the cytosol. In addition, ER phagosome fusion has also led to studies showing that phagosomes are competent organelles for proteosome and TAP dependent cross-presentation [11][13]. Thus, the mechanisms proposed are non-mutually exclusive and the importance of one or other may depend on the source of antigen.

Many antigens have been shown to be cross presented in vitro including soluble proteins, cellular antigens, immune complexes, intracellular bacteria and parasites [14]. Indeed, the primary significance of cross-presentation is shown in vivo as a major mechanism for initiating CD8 T cell priming [5], [15]. A key role of cross-priming in vivo has been demonstrated to initiate CTL immunity during virus infection in a system where only non-hematopoetic cells were infected and bone marrow-derived APCs were shown to be required to capture viral antigens from infected cells by cross-priming [16].

Cross-presentation of bacterial antigens on MHC I has been demonstrated in vitro [17] and several groups have provided data for a mechanism whereby apoptosis of infected cells would provide the source of antigen and result in uptake of bacteria-encoded antigens by bystander DCs which have the capacity to activate naïve T cell responses [18][20]. In vivo, pathogen-infected APCs may be functionally compromised and DCs may not be directly infected. Thus, we pose that an exogenous pathway for processing bacterial-derived antigens to prime CD8 T cell responses would indeed be required; however, this has not been directly addressed. The contribution of cross-priming versus direct priming in immunity to such intracellular pathogens in vivo is therefore not well established owing to difficulties associated with distinguishing these mechanisms of initiating cellular immune responses.

Recent data from our laboratory demonstrated that MHC I trafficking is controlled by a highly conserved tyrosine located in the cytoplasmic tail of the class I molecule. Transgenic mice expressing either the wild type H-2Kb MHC I allele (KbWT), or H-2Kb containing a single point mutation substituting a phenylalanine residue for the conserved tyrosine in exon 6 of the molecule (ΔY) were generated and bred onto a H-2K background by backcrossing transgenic founders with C3H/He mice. Our results showed that the highly conserved MHC I cytoplasmic tyrosine residue forms part of an intracellular targeting motif that is required for routing of MHC I molecules through endolysosomal compartments in DCs where MHC I loading of cross-presented exogenous peptides occurs [5]. The functional consequences of the cytoplasmic tail mutation were demonstrated for two anti-viral CTL responses in vivo whereby the tyrosine mutation abrogated cross-priming of those viral epitopes. Here, we are using this transgenic model system to address the contribution of cross-priming during intracellular bacterial infection.

Listeria monocytogenes (Listeria) is a gram-positive facultative intracellular bacterial pathogen. It is the causative agent of listeriosis, a food borne disease characterized by systemic infection of Listeria disseminating from the intestine into the blood stream and organs. Listeriosis is the leading cause of death among pathogenic bacteria. Pregnant women, neonates, and elderly or immunocompromised individuals are particularly at risk from infection but apparently healthy individuals may also be affected. Listeriosis is normally sporadic with few cases per year but outbreaks of epidemic proportions still occur. Following entry into the host cell Listeria expresses a virulence factor, Listeriolysin O (LLO). This lyses the phagosomal membrane and allows the bacteria to escape the phagosome into the cytosol where it survives and replicates. Listeria infection can spread from cell to cell without entering the extracellular domain and therefore a functional CD8 T cell response is required for clearance of the bacteria and development of protective immunity [21][23]. Listeria-derived MHC I- restricted peptides may be generated from proteins secreted into the cytosol such as the virulence determinant, LLO [24]. Thus, secreted bacterial proteins may be directed through the classical endogenous route of processing for presentation on MHC I. However, following infection in vivo, bacteria are found preferentially within the cytosol of macrophages and hepatocytes in the spleen and liver [25]. Therefore, professional antigen presenting cells (pAPCs) such as DCs may internalize antigens from infected cells or bacterially-derived apoptotic debris for initiation of CD8 T cell responses.

DCs are now considered essential for priming of naïve CD8 T cell responses owing to their unique maturation and migration ability, high expression of co-stimulatory molecules, and unique pH sensing and NADPH oxidase 2 (NOX2) activity [26]. Indeed, DCs are extraordinarily well adapted to cross present antigens to CD8 T cells most efficiently [5], [27]. Further, DCs have been shown to be essential in immune responses to bacterial pathogens since they are critical in priming CD8 T cells to bacterial pathogens. In one study, a diphtheria toxin-based system was used to induce short-term ablation of DCs and showed that the anti-Listeria CTL response was critically dependent on DCs [28]. In a second study, Lenz et al. showed that the priming of CTL to Listeria is restricted to bm-derived APCs [29]. Thus, DC cross-priming of CD8 T cells may provide an important means of alerting cellular immune responses to Listeria monocytogenes bacterial infection. Here, we directly examine the hypothesis that cross-presentation is essential for initiating CD8 T cell responses and mount anti-bacterial responses in vivo.

Results

The strength of an antigen specific CD8 T cell response following Listeria-OVA infection in vivo differs depending on the route of infection

Following Listeria infection, bacteria are preferentially found within the cytosol of macrophages and hepatocytes of the spleen and liver [25]. Bacteria spread from cell to cell without leaving the intracellular compartment, thereby making CD8 T cell activation crucial for clearance of the bacteria and induction of protective immunity [30]. The route of administration of bacteria may greatly affect the strength and type of adaptive immune response generated. In particular, the contribution of cross-presentation may be influenced. Here, we examined primary CD8 T cell responses following i.v. and oral gavage infection of C57BL/6 mice to establish a model examining the initiation of adaptive immunity to Listeria. In order to analyse Ag-specific CD8 T cell responses to Listeria infection, a recombinant Listeria strain (Listeria-OVA) that expresses a well-defined model antigen, ovalbumin (OVA), was used. Specific T cell responses were measured by CTL killing assays of spleen preparations taken 9 days post infection and expanded in culture by in vitro boosting with MHC I immunodominant OVA(257–264) peptide. The ability of activated CD8 T cells to kill target cells presenting OVA(257–264) peptide in the context of H-2Kb was measured. Splenic CTLs from i.v. infection were found to have more robust killing capacity compared to those generated following oral infection (Fig. 1). However, specific responses could be measured following oral infection thereby allowing us to assess the contribution of cross-priming in an oral bacterial infection model, which is the natural route of infection of this bacterium.

thumbnail
Figure 1. Antigen specific CD8 T cell responses in C57BL/6 mice differ following Listeria-OVA oral and i.v. routes of infection.

C57BL/6 mice were infected with 1e9 cfu or 1e4 cfu Listeria-OVA by oral gavage or i.v. injection, respectively. Mock oral infections were performed with 0.1 M HEPES-PBS (Mock) while PBS alone was used for mock i.v. infections. At the peak of the cellular response, 9 days post infection, spleens were harvested and the cytolytic capacity of CD8 T cells was analysed in a standard 51-Cr release assay. T cells were expanded in culture with OVA(257–264)-specific peptide for 6 days. Effector T cells were then incubated with OVA(257–264) peptide pulsed, sodium chromate-labeled target cells and their ability to lyse targets was measured at indicated ratios. Experiments were performed in triplicate with 3 mice per group.

http://dx.doi.org/10.1371/journal.pone.0007210.g001

ΔY mice are deficient in generating Listeria-derived specific CD8 T cell responses in vivo

To gain insight into the role of cross-presentation for priming T cell responses following bacterial infection, we evaluated CD8 T cell responses following Listeria-OVA infection in the H-2Kb transgenic models. Cross-presentation is uniquely inactivated in the ΔY mice thereby allowing the contribution of cross-priming during oral infection to be tested specifically. Interestingly, transgenic mice showed increased susceptibility to Listeria-OVA infection and thus lower infection doses were administered for sub-lethal infection.

CTL killing assays were performed to test the efficiency of activated T cells from infected spleens of transgenic mice to recognize and lyse target cells. Equal numbers of splenic CTLs from KbWT and ΔY were used in the assays and differed significantly in their ability to kill peptide-pulsed targets at infection doses of 5e7 cfu (p<0.05) and 1e6 cfu (p<0.005) tested (Fig. 2A). Quantification of Listeria-specific CD8 T cells was performed on spleen and draining mesenteric lymph nodes (MLN) from KbWT and ΔY mice following oral gavage infection. Splenocytes were re-stimulated in vitro with MHC I immunodominant OVA(257–264) peptide and tetramer staining performed to detect H-2KbOVA(257–264) specific CD8 T cells. ΔY mice had significantly lower percentage of H-2KbOVA(257–264) CD8 T cells in spleen (Fig. 2B) indicating that ΔY mice are deficient in developing antigen specific CD8 T cell responses to Listeria infection. Quantification of Listeria-OVA-specific CD8 T cells was also performed on MLN directly ex vivo following oral gavage infection transgenic mice. Again, ΔY mice generated significantly fewer H-2Kb OVA(257–264)-specific CD8 T cells following Listeria-OVA infection (Fig. 2C). The results indicate that antigen-specific CD8 T cells from ΔY mice are deficient in developing cytotoxic T cell responses following infection.

thumbnail
Figure 2. ΔY mice are deficient in generating specific CD8 T cell responses following Listeria-OVA infection.

KbWT and ΔY mice were orally infected with 5e7 or 1e6 cfu Listeria-OVA or mock infected with 0.1 M HEPES-PBS. Spleens and MLN were harvested 9 days post infection and analysed directly ex vivo or specific T cells were expanded in culture by incubation with OVA(257–264) peptide A, Cytotoxicity assays were performed whereby OVA(257–264) peptide pulsed, sodium chromate-labeled target cells were incubated with restimulated effector T cells from spleens of infected mice. The ability of effector T cells to lyse targets was measured as chromium release at indicated ratios. 2 mice per group were analysed for each dose of infection tested. Cytotoxicity experiments were performed in triplicate. B, C, Listeria-derived OVA(257–264)-specific CD8 T cells from spleens or MLN of 2 mice per group were quantified by tetramer staining following oral infection with 5e7 cfu Listeria-OVA and restimulation of the T cells in vitro (B) or analysed directly ex vivo (C). Results are presented as bar graphs of fold increases in % tetramer/CD8 T cells stained plus SD; Student's t test *, p<0.05.

http://dx.doi.org/10.1371/journal.pone.0007210.g002

Notably, the percentage of total CD8 T cells generated in spleen following Listeria infection was observed lower in ΔY spleen compared to mock, KbWT or C57BL/6 infections. Indeed, T cell depletion following Listeria infection has been reported but only in the first 48 h of infection [31], [32]. To examine whether Listeria bacterial infection resulted in significant depletion of T cells in spleen, 9 days post infection, analysis of CD8 T cell numbers following in vitro stimulation was carried out. The numbers of CD8 T cells were significantly lower in ΔY mice compared to KbWT while the numbers of CD4 T cells did not significantly differ (Fig. 3A, B).

thumbnail
Figure 3. Depletion of CD8 T cells in ΔY spleen following Listeria-OVA infection.

KbWT and ΔY mice were orally infected with 5e7 cfu Listeria-OVA or mock infected with Hepes buffered PBS. Infected spleens were harvested 9 days post infection and OVA-specific T cells expanded in culture for 5 days by incubation with specific OVA(257–264) peptide. Cells were then counted by trypan blue exclusion and CD8 and CD4 T cells were stained and analysed by flow cytometry. Numbers of T cells were calculated and are presented as a bar graph plus SD of 2 mice per group infected; student's t test *, p<0.05.

http://dx.doi.org/10.1371/journal.pone.0007210.g003

ΔY DCs are deficient in stimulating CD8 T cell responses following Listeria infection in vitro

Dendritic cells from KbWT and ΔY mice are used to extend the original findings that the highly conserved tyrosine residue in the cytoplasmic domain of MHC I is responsible for endolysosomal trafficking in DCs thereby allowing processing of exogenously-derived antigen for cross-presentation [5]. To address the function of H-2Kb molecular trafficking on antigen processing for presentation and activation of T cell responses, we evaluated T cell activation and proliferation following Listeria-OVA infection of DCs from transgenic mice in vitro.

Immature DCs were generated from bone marrow (bmDCs) of transgenic KbWT and ΔY mice by 10 day culture with GM-CSF. DCs from wild type and mutant transgenic mice both express equivalent amounts of the DC-specific marker, CD11c, MHC I and class II molecules on their surface (data not shown).

DCs were infected with Listeria-OVA at MOIs, as indicated. Infections were terminated after 4 h by washing to remove extracellular bacteria and resuspending the cells in media with 10 µg/ml chloramphenicol to kill any remaining extracellular bacteria. DCs were allowed to process and present endogenously-derived (from direct infection) and any remaining exogenously-derived (from dead bacteria) bacterial antigens overnight. After 18 h, DCs were incubated with B3Z T cells, a T cell hybridoma that is activated by the recognition of H-2Kb in association with OVA(257–264) peptide. KbWT and ΔY DCs infected with Listeria-OVA were equally adept to process Listeria-derived antigens to activate B3Z T cells in a dose-dependent manner (p<0.005, p<0.0005) while infected TAP1−/− DCs were unable to activate B3Z (Fig. 4A). Interestingly, ΔY DCs showed reduced ability to activate B3Z T cells compared to KbWT DCs when differences in background were taken into account (Fig. 4b), although the statistical significance was smaller (p<0.05) compared to the dose dependant activation found for both KbWT and ΔY DCs, indicating a smaller contribution of cross-presentation, probably from extracellular dead bacteria, compared to direct presentation in vitro. In addition, cross-presentation of soluble whole OVA by KbWT but not ΔY DCs was observed as had been previously shown (Fig. 4C).

thumbnail
Figure 4. ΔY DCs have decreased ability to activate T cells following in vitro infection with Listeria-OVA.

Bone marrow preparations from KbWT and ΔY mice were cultured in vitro with GM-CSF to develop bmDCs. After 10 days in culture, cells were labeled with CD11c and H-2Kb and analysed by flow cytometry. A, Day 10 bmDCs from KbWT, ΔY or TAP1−/− mice were infected with increasing doses of Listeria-OVA for 4 h. After 4 h infection time, extracellular bacteria were removed and DCs were subsequently incubated overnight to allow processing of internalized bacteria. DCs were then incubated with the B3Z T cell hybridoma for 18 h and T cell activation was measured using a CPRG chemiluminescent assay. Uninfected bmDCs from of KbWT, ΔY or TAP1−/− mice served as negative controls. B, Data is presented as a fold increase in T cell activation by dividing CPRG absorbance values by background C. uninfected bmDCs from KbWT, ΔY or TAP1−/− mice were incubated with indicated doses of soluble whole OVA. Processing of internalized OVA was allowed to take place overnight followed by incubation with B3Z T cell hybridoma for 18 h before measuring T cell activation. BmDCs were also pulsed for 1 h with 1 µM OVA(257–264)-specific peptide as positive control. Experiments were performed in triplicate and values are presented as means plus SD with Student's t tests * p<0.05; ** p<0.005; ***p<0.0005.

http://dx.doi.org/10.1371/journal.pone.0007210.g004

ΔY APCs are deficient in stimulating naïve T cell proliferation following Listeria-OVA infection

Antigen processing and presentation to stimulate naïve T cells was next tested. For this, freshly isolated splenocytes were counted and infected with Listeria-OVA for 4 h before washing and resuspending cells in media plus antibiotics to terminate the infection. Eighteen hours later, cells were incubated with naïve CD8 OT-I T cells which recognize APCs presenting OVA(257–264) peptide in the context of H-2Kb molecules. 3H-[Thy] incorporation by proliferating OT-I in response to activation was measured 48 h later. Similarly to what was found in bmDC infections, both KbWT and ΔY splenocytes processed and presented Listeria-derived antigen to stimulate naïve CD8 T cell proliferation in a dose dependant manner (p<0.005, p<0.0005) (Fig. 5A). Significant differences between ΔY and KbWT splenocytes to activate OT-I T cells were also observed (p<0.05, p<0.005) (Fig. 5B) indicating the role of cross-presentation for processing and presenting Listeria-OVA antigens.

thumbnail
Figure 5. ΔY APCs have decreased ability to stimulate naive T cell proliferation following in vitro infection with Listeria-OVA.

A. Splenocytes from KbWT and ΔY mice were infected with increasing doses of Listeria-OVA. Following 4 h infection, extracellular bacteria were removed and infected cells were incubated overnight. APCs were then cultured with OT-I T cells for 48 h. [3H]-Thymidine was then added to the cultures and 24 h later, proliferation was measured by [3H]-Thy incorporation. B. Data is presented as fold increases in T cell proliferation. Values are calculated by dividing [3H]-Thy count values by background. Experiments were performed in triplicate and values are presented as means plus SD with Student's t tests * p<0.05; ** p<0.005; ***p<0.0005.

http://dx.doi.org/10.1371/journal.pone.0007210.g005

Loading and Presentation of Listeria-OVA-derived antigens on ΔY and KbWT APCs

In vitro infection of DCs and APCs indicated that the deficiency observed in ΔY mice to generate CD8 T cell responses against Listeria-OVA lies in its APCs, which are defective in stimulating fully functioning T cells. To address the mechanism of the defect, the antigen processing and loading capacity following Listeria-OVA infection was investigated in ΔY and KbWT splenocytes. Following 4 h infection with Listeria-OVA, splenocytes were incubated overnight and surface stained with the mAb 25.D1.16 which specifically recognizes H-2Kb-OVA(257–264) complexes. Both ΔY and KbWT splenocytes showed loading and presentation of Listeria-OVA derived antigen (Fig. 6A). Marginal differences in loading and presentation were observed between ΔY and KbWT splenocytes at MOI 0.001 and MOI 0.1 but were not significant by students' t test (Fig. 6B). Interestingly, infected ΔY splenocytes pulsed with OVA(257–264) peptide resulted in significantly less presentation of H-2Kb-OVA(257–264) complexes (Fig. 6B). Expression of H-2Kb molecules is normally similar or higher on ΔY APCs compared to KbWT APCs [6]. The latter result indicates that aberrant recycling of MHC I molecules on ΔY APCs which has been shown previously [6], may contribute to the decreased overall levels of surface H-2Kb-OVA(257–264) complexes following infection and peptide pulsing.

thumbnail
Figure 6. Exchange of H-2KbOVA peptide is affected by MHC I tyrosine mutation on ΔY APCs.

Splenocytes from KbWT and ΔY mice were infected with Listeria-OVA between 0.001–0.1 MOI. Following 4 h infection, extracellular bacteria were removed and infected cells were incubated overnight. A. APCs were surface stained with a monoclonal antibody (mAb 25.D1.16) that specifically recognizes H-2Kb-OVA(257–264) complexes and analysed by flow cytometry. Uninfected APCs or infected APCs pulsed with 1 µM OVA(257–264) peptide for 1 h were used as negative and positive controls, respectively. B. Data is represented as a bar graph of the mean fluorescence intensity (MFI) of surface H-2KbOVA(257–264) complexes. Experiments were performed in triplicate and values are presented as average MFI plus SD with Student's t tests * p<0.05.

http://dx.doi.org/10.1371/journal.pone.0007210.g006

Discussion

The physiological relevance of cross-presentation in vivo has been related to the need to generate CTL immunity to virus-infected cells and tissue-specific viruses [5], [16] or following major virus and tumor escape mechanisms resulting in down-regulation of antigen presentation [33], [34]. Cross-presentation is now established under various investigational conditions as a fundamental pathway of activating CD8 T cells but understanding its significance for mounting immune responses in vivo has been limited by experimental means to distinguish cross-priming from overall MHC I priming. Here, we have taken advantage of a transgenic mouse model in which a tyrosine motif in the cytoplasmic tail of MHC I has been mutated. This mutation abrogates the targeting of MHC I molecules to endolysosomal compartments where they may undergo peptide exchange and acquire exogenously derived antigens for cross-presentation [5], [6]. In this context, we have used the mouse model to examine the role of cross-presentation of bacterially-derived antigens following Listeria infection in vivo.

During naturally acquired infection, bacteria traverse the epithelial-cell layer and disseminate in the bloodstream to other organs, such as the spleen and liver, [30]. We show that the ability to generate Listeria-specific CD8 T cells with cytotoxic capacity is affected by the route of infection. Systemic infection induced by i.v. injection of Listeria-OVA produced specific CD8 T cells with more robust ability to perform cytotoxic killing. However, we were still able to detect significant OVA-specific T cells with killing capacity following oral gavage infection, mimicking the natural route of infection of the bacterium. Interestingly, transgenic mice showed increased susceptibility to infection compared to wild type C57BL/6 regardless of the route of administration and therefore required a 20-1000 fold reduction in the dose of bacteria used for infection. This may be due in part to the different background strain of the transgenic mice since the expression of the H-2Kb and ΔY H-2Kb mutant genes have been bred to C3H/He animals and particular strains of mice have been shown to have distinct differences in resistance or susceptibility to Listeria [35].

The novel findings that antigen-specific T cells are deficient in number and in function in ΔY mice following Listeria infection in vivo demonstrates that development of functional Ag-specific T cells is severely impaired following infection due to the MHC I cytoplasmic tail mutation and implies a dependence on MHC I targeting and cross-presentation of bacterial antigens for loading and subsequent priming of CD8 T cells. The failure of the CD8 Ag-specific T cells from ΔY mice to expand following infection could also be detected as a significantly lower total number of CD8 T cells recovered after infection. A model of early events in the T cell response to Listeria infection has been described whereby almost all peripheral T cells become partially activated within one day. This is followed by massive T cell depletion of nonspecific T cells while Ag-specific T cells begin to divide and become fully activated to become the majority of the peripheral T cells by days 7–9, at the peak of the response [31], [32]. In this context, we propose that the cross-presentation deficiency in ΔY mice results in T cells receiving inadequate activation signals which do not clonally expand and instead undergo apoptosis with the manifestation of a large reduction in CD8 T cell number at the peak of Listeria infection response. Further, ΔY DCs can only present endogenously. Therefore, if they become directly infected with Listeria then they would become targets for any CTL produced, contributing to the loss in CD8 T cell number.

Cross-presentation of many types of antigens has been reported [14]. In this work, we have used the MHC I cytoplasmic tail transgenic mice to dissect the importance of the mechanism of cross-presentation within the overall context of MHC I presentation following bacterial infection. We observed that dose-dependent T cell responses are activated following infection of KbWT and ΔY bmDCs, indicating that the endogenous route of processing is predominant following direct infection. Following cellular invasion, Listeria initially occupies an endosome/phagosome [36], [37] but upon acidification of the phagosome, LLO, a cholesterol-dependent pore-forming toxin, is secreted and blocks phagosome-lysosome fusion by generating small pores that uncouple pH and calcium gradients across the phagosome membrane [38]. This acts to destroy the phagosomal membrane and allows escape of bacteria into the cytosol where they can replicate. In this way, Listeria-derived proteins are expected to deliver predominantly to the endogenous route of processing. However, we also observed that DCs from ΔY mice showed significantly reduced ability to activate T cells following direct infection with Listeria-OVA in vitro indicating the contribution of cross-presentation, probably due to the presence of extracellular dead bacteria. In addition, DCs have been shown to have a special ability to alkalinize phagosomes through the activity of NADPH oxidase [39]. LLO-mediated escape of Listeria to the cytosol can be inhibited by alkalinization of the phagosome [37] and bacterial escape from vesicles is approximately 10-fold less active at neutral compared to acidic pH [40]. Thus, the pH of DC Listeria-containing phagosomes may influence the timing of bacterial escape and also contribute to the exogenous route of processing following direct infection with Listeria.

Our previous studies demonstrated that MHC I tyrosine-based motif is required for constitutive internalization of MHC I molecules from the cell surface into early endosomes and lysosomes where loading of exogenously-derived antigen can occur [6]. It is proposed that surface KbWT molecules form an abundant source of MHC I for endosomal and lysosomal peptide-loading of exogenously-derived bacterial antigens. We examined loading of Listeria-OVA-derived antigens on APCs following direct infection of APCs and staining of MHC I-OVA complexes and despite significant differences between ΔY and KbWT APCs to stimulate T cell responses, were unable to detect significant differences in processing and loading of Listeria-derived OVA antigen. However, significant differences in the level of surface H-2Kb-OVA(257–264) complexes could be detected following infection and peptide pulsing. Considering our previous results, we propose that following infection, MHC molecules internalizing from the surface do not sort to compartments compatible with cross-presentation for recycling to the surface, causing an overall decrease in the level of H-2Kb-OVA(257–264) complexes present. The equilibrium of surface levels of MHC I molecules is a balance between new synthesis and constitutive internalization for degradation or recycling. Therefore, we conclude that aberrant recycling of MHC I molecules on ΔY APCs following infection may contribute to the decreased overall levels of surface H-2Kb-OVA(257–264) complexes, the differences of which are only detectable following peptide pulsing.

Overall, the contribution of cross-presentation is likely to differ depending on mode of infection i.e. direct infection of APCs versus oral infection in vivo. Figure 7 describes mechanisms of cross-presentation that may occur depending on the route of infection. Protease-mediated degradation of the bacterium following direct infection in vitro within endolysosomal compartments competent for cross-presentation would provide peptides capable of binding recycling MHC I molecules. However, our data suggest that escape of the bacterium into the cytosol is predominant following infection in vitro whereby the bacterium is subsequently targeted for proteosome –mediated degradation and endogenous processing. Conversely, our results establish that, indeed cross-priming of CD8 T cells is an essential mechanism required to generate primary cellular immune responses following in vivo infection with Listeria. A mechanism for priming of CD8 T cells following bacterial oral infection across the gastrointestinal barrier is likely through bystander acquisition of bacterial antigen. We, and others have shown that Listeria infection causes substantial depletion of T cells by apoptosis. These cells may also be infected [41] necessitating bystander DCs to cross-present and prime CTL. Hereby, we propose that directly infected cells may undergo apoptosis in vivo, thereby providing antigenic material for uptake and processing in cross presenting endolysosomal compartments by bystander DCs. Indeed, cross-priming initiated by pathogen-induced apoptosis has been described following infection with viruses and bacteria [18][20], [42] suggesting that transfer of antigens from apoptotic vesicles to bystander APCs is a general mechanism of cross-priming of CD8 T cells in infectious disease.The rules of engagement for initiating and priming immune responses are now being written. These new edicts will eventually find their application in modulating autoimmune processes and together with novel molecular adjuvants, will promote the efficacy of emerging vaccines. The depth and breath of applying this knowledge is only beginning to be tapped.

thumbnail
Figure 7. Proposed mechanisms of Listeria-derived antigen sampling following in vitro or in vivo infection.

Schematic representation of the mechanisms involved in functional sampling of Listeria-derived antigens depending on the route of infection. A. Listeria directs its own phagocytic uptake into the cell. Within the phagosome, protease-mediated degradation of the bacterium generates peptides capable of binding MHC I molecules recycling from the surface, a process dependant on an intact tyrosine (Y) residue in the cytoplasmic tail of the MHC I molecule. Peptide exchange occurs allowing bacterially-derived antigens to be loaded on MHC I molecules and trafficked to the surface for cross-presentation. Predominantly, toxin-induced pore-forming of the phagosome occurs to allow the escape of the bacterium into the cytosol. Within the cytosol, Listeria is targeted for proteosome –mediated degradation and antigen transport to the ER where bacterial peptides are loaded on newly synthesized MHC I molecules. Loaded MHC I molecules follow the secretory pathway through the trans golgi network (TGN) to the surface for presentation. B. Following infection in vivo, directly infected cells undergo apoptosis, providing antigenic material for uptake by bystander DCs. In this scenario, apoptotic material containing bacterially-derived antigens is phagocytosed into a compartment competent for cross-presentation, named the endolysosomal compartment (ELC). MHC I molecules recycle from the surface into this compartment, peptide exchange occurs and bacterially-derived peptide antigens are loaded and recycled to the cell surface for cross-presentation.

http://dx.doi.org/10.1371/journal.pone.0007210.g007

Materials and Methods

Mouse strains and bacterial infections in vivo

C3H/He mice transgenic for the expression of wild-type MHC I allele H-2Kb (KbWT) or cytoplasmic tail tyrosine mutant (ΔY) were used in these studies as described previously [5]. C57BL/6 mice of H-2Kb haplotype were purchased from Charles River (St. Constant, QE, Canada). The transporter associated with antigen processing 1 knock out (TAP1−/−) and transgenic OT-I mice expressing a TCR specific for OVA(257–264) peptide on a C57BL/6 background were purchased from Jackson Laboratory (Br Harbor, ME). Mice were maintained under specific pathogen-free conditions and fed the normal mouse diet ad libitum. All mouse protocols were approved and performed in accordance with the requirements set out by the Canadian Council on Animal Care.

A recombinant form of Listeria monocytogenes encoding Ovalbumin and an erythromycin-resistance marker was used for these studies (Listeria-OVA). An Antigen expression cassette consisting of the coding sequence of OVA fused to the signal sequence and promoter of the hly gene and an erythromycin resistance gene was introduced into Listeria, and double-crossed onto the Listeria chromosome by homologous recombination, as previously described [43].

Mice were infected with Listeria-OVA by oral gavage with 5e6-1e9 colony forming units (cfu) in 100 µl 0.1 M Hepes-PBS, as indicated or by i.v. with 1×104 cfu in 100 µl PBS. Actual cfu were calculated following infection by plating dilutions of the inoculum. Mock oral infections were performed with 0.1 M Hepes-PBS. PBS alone was used for mock i.v. infections.

Detection of CD8-specific T cell responses to Listeria-OVA

At the indicated times following infection, lymphocytes were isolated by Ficoll-paque (Amersham Biosciences) centrifugation and OVA-specific CD8 T cells were detected directly using a H-2Kb/OVA(257–264) tetramer (immunomics-BeckmanCoulerTM). Otherwise, splenocytes were cultured for 5 days in RPMI 1640 medium (Sigma-Aldrich) supplemented with 10% FCS (HiClone), 100 U/ml penicillin, 100 µg/ml streptomycin (Invitrogen Life Technologies), nonessential amino acids (0.1 mM), sodium pyruvate (1 mM), 2-mercaptoethanol (50 muM), L-glutamine (2 mM) and 1 µM OVA(257–264) (SIINFEKL) H-2Kb-restricted peptide. Following in vitro boosting, CD8 T cells were tetramer stained and analysed by flow cytometry.

Cytotoxic T lymphocyte (CTL) assays were performed using a standard 51Chromimum (51Cr)-release assay. Splenocytes, following in vitro boosting with SIINFEKL peptide at 1 µM concentration for 5 days, were washed and used as effector cells. RMA-S incubated with 1 µM OVA(257–264) and labeled with sodium chromate (100 µCi, Amersham) for 1 h were used as target cells. The CTL activity of activated CD8 T cells was assessed at various effector: target (E:T) ratios in a 4 h 51Cr-release assay at 37°C. Percent specific lysis was calculated as 100% x (cpm [experimental] - cpm [spontaneous release])/(cpm [maximum release] - cpm [spontaneous release]). All assays were performed in triplicate.

Bone marrow-derived DC culture

Bone marrow-derived DCs (bmDCs) were prepared according to the procedure developed by Lutz et al. [44]. Cells flushed from femurs and tibias were plated in RPMI medium (Sigma-Aldrich) with 10% FBS (HiClone), 100 U/ml penicillin, 100 µg/ml streptomycin (Invitrogen Life Technologies), nonessential amino acids (0.1 mM), sodium pyruvate (1 mM), 2-mercaptoethanol (50 muM), L-glutamine (2 mM) and 1% GM-CSF containing supernatant from the X-63Ag8-GM-CSF-transfected cell line [45]. Cells were washed on days 3 and 6, 8, with culture media and 0.5% GM-CSF. DCs were routinely 95% CD11c+ and displayed low to intermediate levels of CD40, CD80, CD86 and MHC II, while expressing intermediate levels of MHC I, characteristic of immature DCs.

Bacterial infection in vitro

bmDCs or splenocytes were harvested and plated in 48-well plates at a density of 5e5/well and infected with Listeria-OVA at 0.001-1 multiplicities of infection (MOI) for 4 h in antibiotic-free medium. After 4 h cells were washed twice in HBSS, resuspended in media with antibiotics (chloramphenicol, 10 µg/ml; Sigma), and infected cells were cultured for an additional 18 h. As positive controls for cross-presentation, bmDCs were incubated with soluble OVA at 10 and 30 mg/ml concentrations overnight. BmDCs were also incubated with 1 µM H-2Kb-restricted OVA(257–264) peptide for 1 h as positive control for B3Z assay.

B3Z T cell activation assay

B3Z is a T cell hybridoma expressing a TCR that specifically recognizes OVA(257–264) (SIINFEKL) in the context of H-2Kb. The cells carry a beta-galactosidase (lacZ) construct driven by nuclear factor of activated T cells elements from the interleukin 2 promoter [46]. Processing and subsequent presentation of Listeria-derived OVA or soluble OVA was measured using Chlorophenol Red Galactopyranoside (CPRG, Calbiochem) as a chemiluminescent substrate for the detection of lacZ activity in B3Z lysates.

BmDCs were incubated with B3Z cells at 1∶1 ratio for 18 h. Individual cultures were lysed by addition of CPRG lysis buffer (100 mM 2-ME, 9 mM MgCl2, 0.125% Nonidet P-40, 0.15 mM chlorophenol red β-galactoside in PBS). Following 18 h incubation at room temperature, in the dark, absorption was read at 570 nm, with 650 nm as the reference wavelength. Uninfected bmDCs incubated with B3Z served as background controls. Fold induction index was calculated by dividing induced activity by background.

OT-I T cell proliferation assay

Splenocytes were isolated from KbWT and ΔY transgenic mice and were infected with varying MOIs of Listeria-OVA for 4 h before the infection was terminated by washing cells with HBSS and resuspending cells in media with 10 µg/ml of chloramphenicol. Infected splenocytes were incubated overnight to allow processing and presentation of Listeria-derived antigens. Next day, splenocytes were cocultured at a 1∶1 ratio with OT-I T cells expressing a transgenic TCR that specifically recognizes H-2Kb in complex with OVA(257–264) [47]. T cell proliferation was determined 48 h later by 3H-thymidine incorporation. Uninfected splenocytes incubated with OT-I T cells served as background controls and fold increase in proliferation was calculated by dividing induced activity by background.

Detection of H-2KbOVA(257–264) loaded complexes

Processing and loading of Listeria-OVA-derived antigens on MHC I was detected with a monoclonal antibody (mAb 25.D1.16) that specifically recognizes H-2Kb-OVA(257–264) complexes [48].

APCs were blocked anti-mouse Fc receptor (BD) and stained with biotin labeled antibody mAb 25.D1.16 followed by phycoerythrin-conjugated rat anti-mouse IgG1 (BD). Cells were then washed and analysed by flow cytometry. Uninfected APCs or infected APCs pulsed with 1 µM OVA(257–264) peptide for 1 h were used as negative and positive controls, respectively.

Statistical Analysis

A student's T test was used to compare numbers and percentages of CD4/CD8 or H-2KbOVA(257–264) tetramer labeled cells from KbWT and ΔY transgenic mice. Statistical differences in B3Z T cell activation between KbWT, ΔY and TAP1−/− mice were also analyzed by student's T tests. Differences between two populations were considered statistically different if *p<0.05, **p<0.005 or ***p<0.0005.

Acknowledgments

The authors gratefully acknowledge the assistance of Dr. Robyn P. Seipp in manuscript preparation.

Author Contributions

Conceived and designed the experiments: ATR WAJ. Performed the experiments: ATR KDO GB. Analyzed the data: ATR WAJ. Contributed reagents/materials/analysis tools: ATR WAJ. Wrote the paper: ATR WAJ.

References

  1. 1. Kaufmann SH, Schaible UE (2005) Antigen presentation and recognition in bacterial infections. Curr Opin Immunol 17: 79–87.
  2. 2. Hsing LC, Rudensky AY (2005) The lysosomal cysteine proteases in MHC class II antigen presentation. Immunol Rev 207: 229–241.
  3. 3. Yewdell JW (2007) Plumbing the sources of endogenous MHC class I peptide ligands. Curr Opin Immunol 19: 79–86.
  4. 4. Rock KL, Gamble S, Rothstein L (1990) Presentation of exogenous antigen with class I major histocompatibility complex molecules. Science 249: 918–921.
  5. 5. Lizee G, Basha G, Tiong J, Julien JP, Tian M, et al. (2003) Control of dendritic cell cross-presentation by the major histocompatibility complex class I cytoplasmic domain. Nat Immunol 4: 1065–1073.
  6. 6. Basha G, Lizee G, Reinicke AT, Seipp RP, Omilusik KD, et al. (2008) MHC class I endosomal and lysosomal trafficking coincides with exogenous antigen loading in dendritic cells. PLoS ONE 3: e3247.
  7. 7. Shen L, Sigal LJ, Boes M, Rock KL (2004) Important role of cathepsin S in generating peptides for TAP-independent MHC class I crosspresentation in vivo. Immunity 21: 155–165.
  8. 8. Kovacsovics-Bankowski M, Rock KL (1995) A phagosome-to-cytosol pathway for exogenous antigens presented on MHC class I molecules. Science 267: 243–246.
  9. 9. Rodriguez A, Regnault A, Kleijmeer M, Ricciardi-Castagnoli P, Amigorena S (1999) Selective transport of internalized antigens to the cytosol for MHC class I presentation in dendritic cells. Nat Cell Biol 1: 362–368.
  10. 10. Gagnon E, Duclos S, Rondeau C, Chevet E, Cameron PH, et al. (2002) Endoplasmic reticulum-mediated phagocytosis is a mechanism of entry into macrophages. Cell 110: 119–131.
  11. 11. Houde M, Bertholet S, Gagnon E, Brunet S, Goyette G, et al. (2003) Phagosomes are competent organelles for antigen cross-presentation. Nature 425: 402–406.
  12. 12. Guermonprez P, Saveanu L, Kleijmeer M, Davoust J, Van Endert P, et al. (2003) ER-phagosome fusion defines an MHC class I cross-presentation compartment in dendritic cells. Nature 425: 397–402.
  13. 13. Ackerman AL, Kyritsis C, Tampe R, Cresswell P (2003) Early phagosomes in dendritic cells form a cellular compartment sufficient for cross presentation of exogenous antigens. Proc Natl Acad Sci U S A 100: 12889–12894.
  14. 14. Heath WR, Belz GT, Behrens GM, Smith CM, Forehan SP, et al. (2004) Cross-presentation, dendritic cell subsets, and the generation of immunity to cellular antigens. Immunol Rev 199: 9–26.
  15. 15. Bevan MJ (2006) Cross-priming. Nat Immunol 7: 363–365.
  16. 16. Sigal LJ, Crotty S, Andino R, Rock KL (1999) Cytotoxic T-cell immunity to virus-infected non-haematopoietic cells requires presentation of exogenous antigen. Nature 398: 77–80.
  17. 17. Pfeifer JD, Wick MJ, Roberts RL, Findlay K, Normark SJ, et al. (1993) Phagocytic processing of bacterial antigens for class I MHC presentation to T cells. Nature 361: 359–362.
  18. 18. Yrlid U, Wick MJ (2000) Salmonella-induced apoptosis of infected macrophages results in presentation of a bacteria-encoded antigen after uptake by bystander dendritic cells. J Exp Med 191: 613–624.
  19. 19. Schaible UE, Winau F, Sieling PA, Fischer K, Collins HL, et al. (2003) Apoptosis facilitates antigen presentation to T lymphocytes through MHC-I and CD1 in tuberculosis. Nat Med 9: 1039–1046.
  20. 20. Winau F, Weber S, Sad S, de Diego J, Hoops SL, et al. (2006) Apoptotic vesicles crossprime CD8 T cells and protect against tuberculosis. Immunity 24: 105–117.
  21. 21. Unanue ER (1997) Studies in listeriosis show the strong symbiosis between the innate cellular system and the T-cell response. Immunol Rev 158: 11–25.
  22. 22. McGregor DD, Koster FT, Mackaness GB (1970) The short lived small lymphocyte as a mediator of cellular immunity. Nature 228: 855–856.
  23. 23. Ladel CH, Flesch IE, Arnoldi J, Kaufmann SH (1994) Studies with MHC-deficient knock-out mice reveal impact of both MHC I- and MHC II-dependent T cell responses on Listeria monocytogenes infection. J Immunol 153: 3116–3122.
  24. 24. Villanueva MS, Sijts AJ, Pamer EG (1995) Listeriolysin is processed efficiently into an MHC class I-associated epitope in Listeria monocytogenes-infected cells. J Immunol 155: 5227–5233.
  25. 25. Gregory SH, Sagnimeni AJ, Wing EJ (1997) Internalin B promotes the replication of Listeria monocytogenes in mouse hepatocytes. Infect Immun 65: 5137–5141.
  26. 26. Savina A, Amigorena S (2007) Phagocytosis and antigen presentation in dendritic cells. Immunol Rev 219: 143–156.
  27. 27. Wilson NS, Villadangos JA (2005) Regulation of antigen presentation and cross-presentation in the dendritic cell network: facts, hypothesis, and immunological implications. Adv Immunol 86: 241–305.
  28. 28. Jung S, Unutmaz D, Wong P, Sano G, De los Santos K, et al. (2002) In vivo depletion of CD11c(+) dendritic cells abrogates priming of CD8(+) T cells by exogenous cell-associated antigens. Immunity 17: 211–220.
  29. 29. Lenz LL, Butz EA, Bevan MJ (2000) Requirements for bone marrow-derived antigen-presenting cells in priming cytotoxic T cell responses to intracellular pathogens. J Exp Med 192: 1135–1142.
  30. 30. Pamer EG (2004) Immune responses to Listeria monocytogenes. Nat Rev Immunol 4: 812–823.
  31. 31. Merrick JC, Edelson BT, Bhardwaj V, Swanson PE, Unanue ER (1997) Lymphocyte apoptosis during early phase of Listeria infection in mice. Am J Pathol 151: 785–792.
  32. 32. Jiang J, Lau LL, Shen H (2003) Selective depletion of nonspecific T cells during the early stage of immune responses to infection. J Immunol 171: 4352–4358.
  33. 33. Spiotto MT, Rowley DA, Schreiber H (2004) Bystander elimination of antigen loss variants in established tumors. Nat Med 10: 294–298.
  34. 34. Basta S, Bennink JR (2003) A survival game of hide and seek: cytomegaloviruses and MHC class I antigen presentation pathways. Viral Immunol 16: 231–242.
  35. 35. Cheers C, McKenzie IF (1978) Resistance and susceptibility of mice to bacterial infection: genetics of listeriosis. Infect Immun 19: 755–762.
  36. 36. Alvarez-Dominguez C, Roberts R, Stahl PD (1997) Internalized Listeria monocytogenes modulates intracellular trafficking and delays maturation of the phagosome. J Cell Sci 110 (Pt 6): 731–743.
  37. 37. Beauregard KE, Lee KD, Collier RJ, Swanson JA (1997) pH-dependent perforation of macrophage phagosomes by listeriolysin O from Listeria monocytogenes. J Exp Med 186: 1159–1163.
  38. 38. Shaughnessy LM, Hoppe AD, Christensen KA, Swanson JA (2006) Membrane perforations inhibit lysosome fusion by altering pH and calcium in Listeria monocytogenes vacuoles. Cell Microbiol 8: 781–792.
  39. 39. Savina A, Jancic C, Hugues S, Guermonprez P, Vargas P, et al. (2006) NOX2 controls phagosomal pH to regulate antigen processing during crosspresentation by dendritic cells. Cell 126: 205–218.
  40. 40. Glomski IJ, Gedde MM, Tsang AW, Swanson JA, Portnoy DA (2002) The Listeria monocytogenes hemolysin has an acidic pH optimum to compartmentalize activity and prevent damage to infected host cells. J Cell Biol 156: 1029–1038.
  41. 41. McElroy DS, Ashley TJ, D'Orazio SE (2009) Lymphocytes serve as a reservoir for Listeria monocytogenes growth during infection of mice. Microb Pathog 46: 214–221.
  42. 42. Albert ML, Sauter B, Bhardwaj N (1998) Dendritic cells acquire antigen from apoptotic cells and induce class I-restricted CTLs. Nature 392: 86–89.
  43. 43. Shen H, Slifka MK, Matloubian M, Jensen ER, Ahmed R, et al. (1995) Recombinant Listeria monocytogenes as a live vaccine vehicle for the induction of protective anti-viral cell-mediated immunity. Proc Natl Acad Sci U S A 92: 3987–3991.
  44. 44. Lutz MB, Kukutsch N, Ogilvie AL, Rossner S, Koch F, et al. (1999) An advanced culture method for generating large quantities of highly pure dendritic cells from mouse bone marrow. J Immunol Methods 223: 77–92.
  45. 45. Stockinger B, Zal T, Zal A, Gray D (1996) B cells solicit their own help from T cells. J Exp Med 183: 891–899.
  46. 46. Shastri N, Gonzalez F (1993) Endogenous generation and presentation of the ovalbumin peptide/Kb complex to T cells. J Immunol 150: 2724–2736.
  47. 47. Hogquist KA, Jameson SC, Heath WR, Howard JL, Bevan MJ, et al. (1994) T cell receptor antagonist peptides induce positive selection. Cell 76: 17–27.
  48. 48. Porgador A, Yewdell JW, Deng Y, Bennink JR, Germain RN (1997) Localization, quantitation, and in situ detection of specific peptide-MHC class I complexes using a monoclonal antibody. Immunity 6: 715–726.