ABSTRACT
In healthy individuals, influenza virus (IAV) infection generally remains localized to the epithelial cells of the respiratory tract. Previously, IAV-specific effector CD8 T cells found systemically during the course of IAV infection were thought to have been primed in lung-draining lymph nodes with subsequent migration to other tissues. However, little is known about whether other lymphoid sites participate in the generation of virus-specific CD8 T cells during localized IAV infection. Here, we present evidence of early CD8 T cell priming in the spleen following respiratory IAV infection independent of lung-draining lymph node priming of T cells. Although we found early indications of CD8 T cell activation in the lymph nodes draining the respiratory tract, we also saw evidence of virus-specific CD8 T cell activation in the spleen. Furthermore, CD8 T cells primed in the spleen differentiated into memory cells of equivalent longevity and with similar recall capacity as CD8 T cells primed in the draining lymph nodes. These data showed that the spleen contributes to the virus-specific effector and memory CD8 T cell populations that are generated in response to respiratory infection.
INTRODUCTION
Influenza virus (IAV) infection is usually restricted to the upper and lower respiratory tract. Lung antigen-presenting cells (APCs) acquire viral antigens from infected lung epithelial cells (1, 2) or through direct dendritic cell (DC) infection (3) and then undergo a maturation process that induces migration to local draining lymph nodes (LN) via the lymphatics (4, 5). These events generally restrict generation of the immune response locally to the cervical and mediastinal LN, which drain the respiratory tract (4, 6, 7). Although it has been shown that IAV may infect tissues other than the lung (8–10), this is rare in otherwise healthy individuals/organisms and is usually restricted to highly virulent virus strains (11, 12). The systemic appearance of virus-specific effector cells after IAV infection must therefore emerge from dissemination of locally expanded cells or could be derived from a previously unappreciated process of antigenic priming in nondraining sites. Whether the dissemination of virus, viral genetic material, or viral antigen is important for the generation of a more effective immune response is not known.
T cells play an important role in the control of primary IAV infections and memory T cells have been shown to mediate protection to infection with both homosubtypic and heterosubtypic virus strains (13–16). The ability of CD8 T cells to recognize conserved viral gene products provides the impetus to target vaccination to the CD8 T cell response to generate heterosubtypic immunity. Unlike the antibody/B cell memory conferred protection, which creates a systemic barrier to the virus, T cell-based immunity likely requires the presence of memory T cells at the site of infection (17). In fact, in experimental systems, the persistence of T cell-mediated protection from influenza virus infection has been shown to diminish over time coincident with the decrease in virus-specific T cells in the lung (18), even in the presence of systemic pools of virus-specific memory T cells. The site of initial priming of CD8 T cells may affect the localization of memory cells. The protective ability of memory T cells that are originally primed in systemic lymphoid sites must therefore be compared to T cells primed in local draining lymph nodes in order to predict the potential efficacy of vaccines administered by different routes.
In the present study we sought to define the sites of initial T cell encounter with viral antigen following respiratory IAV infection. We found that after respiratory IAV infection, viral antigen was transiently presented in the spleen, in addition to the lung-draining LN. Furthermore, our results showed that IAV-specific memory CD8 T cells generated in the spleen during primary infection demonstrated survival and effector abilities equivalent to those of mediastinal LN-primed memory CD8 T cells. Thus, these findings identified the spleen as a contributor to the immune response to respiratory infection and may provide the rationale for vaccine formulations that allow multisite priming of both T and B cells.
MATERIALS AND METHODS
Mice.C57BL/6 (CD45.2 and CD45.1) and BALB/c mice, 6 to 8 weeks of age, were purchased from Jackson Laboratories (Bar Harbor, ME) or Charles River Laboratories/National Cancer Institute (Wilmington, MA). TCR transgenic OT-I-RAG−/− mice (19), F5 mice (20), or TS1 mice (21) were bred in-house and used between the ages of 3 and 6 months. Animals were maintained in the University of Connecticut Health Center or Columbia University animal care facilities in standard pathogen free conditions. All protocols involving animals were approved by the University of Connecticut Health Center Animal Care Committee and Columbia University Institutional Animal Care and Use Committee.
Influenza virus infections.E61-13-H17 (A/HK/8/68 × A/PR/8/34) (H3N2) influenza virus and recombinant WSN influenza virus strains expressing SIINFEKL (WSN-OVA1) or SIINLEKL (WSN-OVA0) epitopes were generously provided by David Topham. Influenza virus (A/PR/8/34) (PR8) was grown and titered as described previously (16). WSN and E61-13-H17 virus stocks were grown in chicken eggs, titered, and stored as described previously (22). For influenza virus infections, mice were anesthetized by intraperitoneal (i.p.) injection of avertin (2,2,2-tribromoethanol) or isoflurane before inoculation with 300 EID50 of E61-13-H17, 1,000 PFU of WSN, or 200 50% tissue culture infective doses (TCID50) of PR8 influenza viruses. Influenza virus titers in lung and spleen homogenates were determined by performing a TCID50 assay as described previously (23).
Adoptive transfer of TCR transgenic T cells.CFSE (carboxyfluorescein diacetate succinimidyl ester) (24)-labeled OT-I (19), F5 (20), or TS1 (21) TCR transgenic T cells (105 to 106) were transferred by intravenous injection at the indicated times after infection.
Cell isolation.Mice were given a lethal dose of avertin i.p. and then perfused with phosphate-buffered saline (PBS; with heparin). The lungs, spleen, and inguinal, brachial, axillary, mesenteric, cervical, and mediastinal LN were then removed. For collagenase digestion, the spleen and lungs were cut into small pieces using dissection scissors, and the LN were disrupted using insulin syringe needles. The tissue fragments were then incubated in collagenase solution at 37°C for 20 min (spleen and LN) or 1 h (lung). EDTA was added, followed by incubation for a further 10 min. The remaining pieces were crushed between frosted glass slides and then passed through nylon mesh. Cells were washed and resuspended in Hanks balanced saline solution.
Flow cytometry.Ovalbumin (OVA)-specific cells were detected using H-2Kb tetramers containing the OVA-derived SIINFEKL peptide as previously described (25). NP- and PA-specific cells were detected using H-2Db tetramers with NP366-374 and PA224-233 peptides obtained from the National Institute of Health Tetramer facility. For staining, lymphocytes were suspended in PBS–0.2% bovine serum albumin–0.1% NaN3 at a concentration of ∼107 cells/ml. For tetramer staining, cells were incubated at room temperature for 1 h with tetramer-APC plus the appropriate dilution of anti-CD8 (clone 53.6-72) antibody and antibodies to other phenotypic surface molecules. Staining with antibodies specific for CD45.1, CD11a, CD127, KLRG-1, CD4, CD25, CD69, and CD62L (all monoclonal antibodies [MAbs] from BD Pharmingen or eBioscience) was done at 4°C for 20 min. After staining, the cells were washed and then fixed with 3% paraformaldehyde in PBS. The relative fluorescence intensities were measured with a FACSCalibur or LSR II machine (BD Biosciences, San Jose, CA). The data were analyzed using FlowJo Software (Tree Star, San Carlos, CA).
In situ tetramer staining and confocal microscopy.Tetramer staining was carried out as previously described (26). Briefly, the spleen was cut into sections (∼200 μm) using a scalpel. The sections were then stained for 12 h at 4°C with OVA:Kb tetramer, along with unconjugated anti-CD8 antibody in 2% fetal calf serum in PBS. Tissue sections were then washed four times (20 min each time) with cold PBS. Sections were stained with goat anti-APC antibody overnight and then washed as described above. Finally, the sections were stained with Alexa 546-conjugated rat anti-goat, CY5-conjugated anti-CD45.1, and Alexa 488-conjugated anti-B220 antibodies. Images were acquired using a Zeiss LSM 510 Meta (Carl Zeiss MicroImaging, Inc., Thornwood, NY) using a ×10 water immersion objective lens. Images were analyzed using Imaris 3D software (Bitplane, Inc.).
Intracellular detection of IFN-γ.Lymphocytes were isolated from the spleen and were cultured 5 h with 1 μg of Golgistop (BD Pharmingen)/ml, with or without 1 μg of the SIINFEKL and the NP peptides/ml. After culture, cells were stained for surface molecules and then fixed and permeabilized in Cytofix/Cytoperm solution (BD Pharmingen) and stained with anti-gamma interferon (IFN-γ)-phycoerythrin (PE) or control rat IgG1-PE. The cells were then washed, and the fluorescence intensity was measured on a FACSCalibur.
Magnetic cell separation.Single cell suspensions were labeled with allophycocyanin-conjugated anti-CD45.1 antibody for 20 min at 4°C. The cells were then washed with buffer and incubated with anti-APC magnetic beads (Miltenyi Biotec) for 15 min at 4 C. Next, the cells were washed and passed through magnetically activated cell-sorting columns in the presence of a strong magnetic field. The column was washed three times to elute the unlabeled fraction, the column was removed from the magnet, and the labeled fraction was eluted with buffer.
RESULTS
Recently activated virus-specific CD8 T cells are in the lung-draining lymph nodes and spleen.To study the kinetics of naive CD8 T cell recruitment to influenza virus-infected sites, 105 naive CFSE-labeled CD45.1 TCR transgenic OT-I cells (specific for the SIINFEKL peptide derived from ovalbumin [OVA]) were adoptively transferred into CD45.2 C57BL/6 mice. The mice were then intranasally (i.n.) infected with a sublethal dose (1,000 PFU) of WSN-OVA1 influenza virus, which encodes the SIINFEKL epitope (29). At days 2, 3, and 4, lymphocytes were harvested from the mediastinal LN (medLN), spleen, and pLN (inguinal, brachial, axillary, cervical, and mesenteric LN) and enriched for CD45.1+ cells by magnetic cell separation. The number of transferred cells found in each of these compartments was quantified by using flow cytometric analysis. Over the first 3 days after infection (a period preceding cell proliferation), the transferred cells left the pLN and accumulated in the spleen and the medLN (data not shown). Cells recovered from the medLN expressed the early activation marker CD69 as early as day 2 postinfection, while cells from the nondraining pLN did not (Fig. 1A), indicative of recent antigen encounter in the lung-draining LN but not the pLN. Surprisingly, we observed CD69 expression on OT-I cells recovered from the spleen at day 3 after IAV infection, indicating that viral antigen was also present in this compartment. Furthermore, the transferred cell populations harvested from the spleen and medLN, but not from the pLN, demonstrated evidence of blastogenesis by increased size as measured by forward light scatter at day 4 postinfection (Fig. 1A and B). These results suggested that naive CD8 T cells encountered influenza virus antigen not only in the lung-draining LN but also in the spleen.
The presence of viral antigen was determined in the medLN and spleen after respiratory infection. A total of 100,000 naive CD45.1 OT-I cells were transferred into CD45.2 C57BL/6 mice, which were infected with 1,000 PFU of WSN-OVA1 influenza virus by i.n. infection. At days 2, 3, and 4, the cells were harvested from the medLN, spleen, and pLN (inguinal, brachial, axillary, and mesenteric lymph nodes) and enriched for CD45.1 by magnetic cell separation. Single cell suspensions were stained with fluorescently labeled antibodies to CD45.1, CD8, and CD69 and analyzed by flow cytometry. (A) Histograms show the CD69 profiles and forward-scatter profiles of gated CD8+ CD45.1+ cells. (B) Graphs represent the means and standard errors of the mean (SEM) of the frequency of CD69+ and forward scatterhi cells (n = 3) at each time point. Two-way analysis of variance (ANOVA) of medLN versus pLN and spleen versus pLN was performed (*, P < 0.05; **, P < 0.01; ****, P < 0.0001).
Viral antigen is present in the lung-draining lymph nodes and spleen.After initial antigen encounter, naive CD8 T cells undergo prolonged interactions with the APCs presenting their cognate antigens. These interactions are thought to last ca. 36 to 72 h before cells begin to leave the secondary lymphoid organs (SLO) (30). As a consequence of these tight interactions mediated by TCR and enhanced adhesion molecule expression, responding T cells become “locked” in SLO during the early stage of the immune response (31–34). These findings argue against the possibility of cells migrating from their original site of antigen encounter to other lymphoid organs during this early phase of antigen presentation. Nonetheless, the possibility remained that the splenic CD69+ CD8 T cells encountered antigen in the medLN and then migrated to the spleen. To address the possibility that newly primed influenza virus-specific CD8 T cells redistributed to the spleen during the early stage of the immune response, CFSE-labeled OT-I cells were transferred into previously infected mice. The mice were then sacrificed ∼1 day later, and cells from various sites were cultured in vitro. This protocol allows time for naive CD8 T cells to encounter their cognate antigens but insufficient time for migration to other sites (35). In this system, CD8 T cell division in vitro indicates tissue-specific influenza virus antigen encounter in vivo. Spleen, mediastinal, cervical, inguinal, and brachial LN were harvested, and the cells cultured for 3 days in interleukin-2-supplemented media. OT-I cells transferred 3 days postinfection into WSN-OVA1-infected mice proliferated only when they were harvested from the medLN (Fig. 2A). There was no indication of antigen encounter in the spleen or any other LN. However, when OT-I cells were transferred at day 5 postinfection, proliferation was seen in cells harvested from both the medLN and the spleen but not from the inguinal LN (Fig. 2A). Interestingly, viral antigen presentation in the spleen was transient as there was little to no proliferation in the spleen when OT-I cells were transferred into animals at 11 days postinfection, whereas, consistent with previous reports (36–38), there was evidence of continued antigen presentation in the medLN at this late time point. These results supported our conclusion that viral antigen was presented in the spleen after respiratory IAV infection but that presentation was temporally delayed compared to medLN.
Viral antigen is transiently presented to CD8 T cells in the spleen following respiratory influenza virus infection. (A) CD45.2 C57BL/6 mice were administered 1,000 PFU of WSN-OVA1 by i.n. infection and at days 3, 5, or 11 106 CD45.1 CFSE-labeled OT-I cells were adoptively transferred. At 22 h after transfer, the spleen, mediastinal, inguinal, mesenteric, and cervical lymph nodes were harvested, and 2 × 106 total cells were placed in culture for 3 days. The CFSE profiles of gated populations of CD45.1+ transferred cells are shown. The numbers represent the means and SEM of two or three animals per group. The results are representative of three independent experiments. (B) A total of 106 CFSE-labeled naive F5 cells were transferred into animals previously infected with 500 EID50 E61-13-H17 strain of influenza virus. The cells were harvested 24 h later and then placed in culture for 3 days. Histograms represent the CFSE profiles of transferred cells from individual tissues. The numbers represent the means and SEM of two or three animals per group. The results are representative of two independent experiments.
The WSN strain of influenza virus has been shown to be more promiscuous than other IAV strains in terms of the cell types and tissues that it can infect (39). This promiscuity has been linked to a mutation in the HA protein enabling cleavage by a wider range of proteases than other IAV strains (39). To rule out the possibility that the presentation of viral antigen after respiratory IAV infection was specific to the WSN strain, we repeated the short-term transfer experiments after infection with the E61-13-H17 IAV strain and used F5 TCR transgenic CD8 T cells, specific for NP, as detectors of viral antigen. Similar to WSN-OVA1 infection, at day 3 postinfection, there was evidence of early antigen presentation only in the medLN. At day 5, however, there was proliferation by the CD8 T cells harvested from the spleen, as well as the medLN, which indicated the presence of viral antigen in these sites (Fig. 2B). By day 5, antigen was also present in the cervical lymph nodes (CLN) after E61-13-H17 infection, which was not observed with the WSN-OVA1 strain (Fig. 2). This result was consistent with our previously published work showing long-term viral antigen persistence in the medLN and CLN following E61-13-H17 infection but only within the medLN following WSN infection (36). Similar to what was observed with WSN infection, antigen presentation in the spleen was only transient, disappearing by day 11 but persisting in the medLN and CLN (Fig. 2B). These results suggested that antigen presentation in the spleen following respiratory IAV infection may be generalizable to multiple influenza virus strains.
Antigen encounter in the T-cell areas of splenic follicles.To determine the scope of antigen presentation in the spleen following respiratory IAV infection, we used confocal microscopy to visualize OT-I cell distribution in the spleen after antigen encounter. A total of 5 × 105 CD45.1 OT-I cells were adoptively transferred into mice infected i.n. with either WSN-OVA1 (SIINFEKL-expressing) or WSN-OVA0 (SIINLEKL-expressing) influenza virus 5 days before. At 24 h after transfer, the spleen was harvested, sectioned, and stained with OVA:Kb tetramer and antibodies to B220 and CD45.1. OT-I cells transferred into WSN-OVA1-infected mice were detected primarily in the T cell areas (PALS) of splenic follicles and not in the red pulp (Fig. 3A). OT-I cell clusters and TCR reorientation, reminiscent of synapse formation, as seen by tetramer staining (26), was observed in only the WSN-OVA1-infected animals and not in uninfected or WSN-OVA0-infected animals (Fig. 3A, magnified regions). Although the precise identity of the relevant antigen-presenting cells (APCs) is unknown, purified CD11c+ dendritic cells (DCs) from the spleen of WSN-OVA1-infected mice were able to induce OT-I proliferation in vitro (data not shown). Thus, CD8 T cells were interacting with DCs in the spleen, in a location previously shown to be important for CD8 T cell activation following systemic infection (26).
Antigen encounter in the spleen occurs in the T-cell areas of splenic follicles. (A) CD45.2 C57BL/6 mice were inoculated i.n. with 1,000 PFU of WSN-OVA1 or WSN-OVA0 influenza virus. At 5 days after infection, naive CD45.1 OT-I cells were transferred intravenously, and 24 h later the spleens were harvested, sectioned, and stained with OVA:Kb tetramer (red) and fluorescently labeled antibodies specific for B220 (green) and CD45.1 (blue). Tissue sections were then analyzed by confocal microscopy (×100 total magnification). (B) In parallel experiments, the lymphoid organs were harvested, and the transferred cell population in each tissue analyzed for CD69 expression (left panel) and forward scatter (right panel) by flow cytometry. The statistical method used was one-way ANOVA/Dunnett post-analysis (*, (P < 0.05). The results are representative of two independent experiments. (C) CD90.2 BALB/c mice were inoculated intranasally with 200 TCID50 of PR8 influenza virus. At 4 days after infection, naive CD90.1 TS1 cells were transferred into these animals; 24 h later, the lymphoid organs and lungs were harvested, and the transferred cell population was analyzed for CD69 and CD25 expression by flow cytometry. The spleens and lungs were harvested from control or PR8-infected mice at day 4 and homogenized, and the virus titers were determined by a MDCK TCID50 titration assay. The means and SEM for three mice per infected group are shown (uninfected group, n = 2).
Transferred cells from the spleen and LN were also analyzed for CD69 expression and size by flow cytometry 5 days postinfection. The majority of transferred cells harvested from the medLN and spleens of WSN-OVA1-infected animals were CD69 positive (Fig. 3B) and enlarged in size, indicating the presentation of viral antigen in these sites. Importantly, CD69 upregulation and blastogenesis was not seen in transferred cells harvested from the medLN or spleens of animals infected with the WSN-OVA0 virus, indicating that activation of the transferred cells was antigen dependent and not a bystander effect (Fig. 3B). This phenomenon was not influenza virus strain or mouse specific or only applicable to CD8 T cells, since adoptively transferred HA-specific TCR transgenic CD4 T cells in BALB/c mice also exhibited increased CD69 and CD25 expression in the spleen and medLN but not in the pLN after infection with PR8 influenza virus (Fig. 3C). Moreover, virus was detected in the lung but not in the spleen 5 days after infection (Fig. 3C), indicating that viral replication (at least not above the limit of detection) was likely not the source of splenic antigen. These results support the hypothesis that both major histocompatibility complex class I and II virus-derived peptides were presented in the spleen as a consequence of respiratory IAV infection.
Sequestration of virus-specific cells in the medLN does not affect the expansion of splenic virus-specific CD8 T cells.To begin to define the relative contribution of splenic versus LN priming to the overall immune response and to further exclude the possibility that antigen-specific T cells circulated to the spleen from the medLN, we used FTY720, an S1P receptor agonist to sequester lymphocytes in the LN (40) after IAV infection. C57BL/6 mice were infected with WSN-OVA1 and 3 days later treated with 1 mg of FTY720/kg. Day 4 was chosen for the start of FTY720 treatment in order to minimize the effect of the drug on the migration of APCs to and from the infected tissue. At 7 days postinfection, the medLN, lungs, and spleen were harvested, and the number of virus-specific CD8 T cells (NP, PA, and OVA specific) were quantified by tetramer staining. FTY720 treatment after IAV infection resulted in an ∼2-fold increase in the total number of virus-specific cells in the medLN at day 7 postinfection (Fig. 4A). There was a similar decrease in the magnitude of the virus-specific CD8 T cell population in the lungs of FTY720-treated mice, but no difference was noted in the spleen compared to control animals (Fig. 4B). Thus, these results supported the likelihood that priming and expansion of virus-specific cells in the spleen occurred independently of the medLN immune response.
Sequestration of virus-specific cells in the medLN does not affect the expansion of the splenic virus-specific CD8 T cell population. C57BL/6 mice were inoculated i.n. with 1,000 PFU of WSN-OVA1 and then treated with 1 mg of FTY720/kg or PBS at 4 days after infection. At day 7, the lung, spleen, and medLN were harvested, and the cells were stained with MAbs to CD8 and CD11a and with PA224-233:Db, NP366-374:Db, and OVA:Kb tetramers. (A) Representative plots are gated on CD8 T cells, with percentages shown representing populations of CD11ahi cells that are NP, PA, or OVA specific. (B) Graphs show the means and SEM of the total number of NP-, PA-, or OVA-specific cells present at day 7 p.i. in the indicated tissues after PBS or FTY720 treatment. **, P < 0.01; ns, P > 0.05 (as determined by two-way ANOVA).
Inhibition of T cell priming in the medLN has no effect on the expansion of virus-specific CD8+ T cells in the spleen.The presentation of viral antigen in the spleen in the context of a respiratory viral infection may serve as a secondary source of effector cells for the clearance of virus. Considering the large number of cells constantly trafficking through the spleen, it is quite likely that, after an i.n. IAV infection, a fraction of naive virus-specific CD8 T cells first encounter influenza virus antigen within the spleen. Activation and expansion of CD8 T cells in the spleen may provide an alternate population of effector cells, which are then directed to the lung to participate in viral clearance. To determine whether splenic-primed CD8 T cells acquired effector functions and migrated to the lung, we used the Mel-14 (anti-CD62L) MAb to inhibit CD8 T cell priming in the medLN and analyzed the population of CD8 T cells primed in the spleen. The Mel-14 MAb inhibits trafficking of lymphocytes into LN via the high endothelial venules (27). Pretreating animals for 6 days resulted in an ∼10-fold reduction in the cellularity of LN and a smaller increase in spleen cellularity (data not shown). Mice were pretreated with Mel-14 or IgG2a control MAb and then infected with WSN-OVA1. At 7 days postinfection, virus-specific CD8 T cells were quantified from the lung, spleen, and medLN. Mel-14 treatment resulted in a 10-fold reduction in the number of both OVA- and NP-specific CD8 T cells in the medLN and in the lung compared to IgG-treated animals (Fig. 5). As with FTY720 treatment, no change was observed in virus-specific CD8 T cell numbers in the spleen, suggesting that inhibiting CD8 T cell priming in the medLN had no effect on virus-specific T cell expansion in the spleen. Together with the data from the FTY720 experiment, these results suggested that substantial numbers of effector CD8 T cells are generated in the spleen and may contribute to the pool of virus-specific CD8 T cells in the lung.
Inhibiting priming of virus-specific cells in the medLN has no effect on the expansion of virus-specific CD8 T cells in the spleen. C57BL/6 mice were pretreated with three doses of 250 μg of Mel-14 or IgG2a MAb 3 days apart. At the time of the third dose, animals were infected intranasally with 1,000 PFU of WSN-OVA1. At 7 days after infection, the lung, spleen, and medLN were harvested, and the cells were stained with MAbs specific for CD8 and CD11a and with NP366-374:Db and OVA:Kb tetramers. (Top) Representative plots showing the frequency of NP-specific and OVA-specific cells of total CD8 T cells in the indicated tissues of mice treated with Mel-14 or IgG MAb. (Bottom) Total number of OVA-specific and NP-specific CD8 T cells in the spleen, lung, and mediastinal lymph node were measured 7 days postinfection. The graph indicates the means and SEM for three animals per group. **, P < 0.01; *, P < 0.05 (as determined by two-way ANOVA).
CD8 T cells primed in the medLN and the spleen generate functional memory.The environment in which a naive cell first encounters cognate antigen determines the differentiation fate of its daughter cells (41–45). Due to the differences in proximity of the medLN and the spleen to the infected tissue, there may be differences in the amount and nature of inflammatory and costimulatory signals and availability of antigen in the two sites. In order to determine the ability of cells primed in the medLN or the spleen to survive and generate memory cells, we utilized the Mel-14 treatment protocol to ensure splenic priming. To this end, CD45.1 C57BL/6 mice were given three doses of Mel-14 MAb or IgG2a MAb 3 days apart. On the day of the final dose, mice were infected i.n. with WSN-OVA1. At 7 days postinfection, cells from the spleens of Mel-14-treated mice and the medLN of the IgG-treated mice were enriched for CD8+ T cells by negative selection, and the total spleen and medLN CD8 T cells, which contained 104 virus-specific (NP- and OVA-specific) CD8 T cells, were transferred intravenously into uninfected CD45.2 C57BL/6 mice (Fig. 6A). On days 14 and 49 after transfer the animals were challenged with WSN-OVA1, and at 7 days after infection the number of tetramer-positive donor cells in the medLN, lungs, and spleen were quantified. Challenge at day 14 after transfer resulted in the appearance of donor-derived virus-specific CD8 T cells in the medLN, lungs, and spleen, regardless of whether the donor cells were spleen or medLN derived (Fig. 6B). The number of total donor cells recovered from animals that received spleen-derived cells was 2- to 5-fold greater than that from animals that received medLN-derived cells, reflecting the original number of cells that was transferred to each group. More importantly, the numbers of transferred, virus-specific CD8 T cells after recall were equal between animals that received either medLN- or splenic-primed CD8 T cells (Fig. 6C). This result was reiterated when transferred mice were challenged 49 days later (Fig. 6C). Thus, both splenic- and medLN-primed CD8 T cells were able to generate long-lived memory cells. These results indicated that antigen presentation in the spleen after respiratory influenza virus infection primed CD8 T cells as effectively as did antigen presentation in the medLN to drive memory T cell development.
CD8 T cells primed in the medLN and in the spleen have a similar capacity to generate a recall response. (A) Schematic of the protocol used to generate and transfer medLN-primed and splenic-primed cells. Briefly, C57BL/6 mice were pretreated with three doses of 250 μg of Mel-14 or IgG2a MAb 3 days apart. At the time of the third dose, animals were infected i.n. with 1,000 PFU of WSN-OVA1. At 7 days after infection, the mediastinal lymph nodes (medLN) were harvested from IgG2a-treated mice, while the spleens were harvested from Mel-14-treated animals, and the cells enriched for CD8 by magnetic cell separation. Total CD8-enriched medLN and spleen cells were transferred into uninfected mice at numbers that contained 10,000 NP/OVA tetramer-positive cells. (B) At 14 days (left panel) and 49 days (right panel) after transfer, recipient mice were challenged with WSN-OVA1, and 7 days later, the spleen, medLN, and lung harvested and stained with fluorescently labeled anti-CD45.1 antibody and NP366-374:Db and OVA:Kb tetramers. Shown are representative plots of gated CD8+ cells with percentages showing the frequency of NP- and OVA-specific CD45.1+ transferred cells. (C) Graphs represent the means and SEM of the absolute numbers of NP-and OVA-specific CD45.1+ CD8 T cells after recall at 14 days (left) and 49 days (right) after transfer (n = 4 to 5 animals per group). Open bars, medLN cells transferred; solid bars, spleen cells transferred.
CD8 T cells primed in the medLN and the spleen have similar capacity to produce IFN-γ upon recall.To compare the functional capacity of memory virus-specific CD8 T cells from the spleen and medLN, we measured IFN-γ production by these cells after in vitro peptide stimulation. Spleen- and medLN-primed cells were transferred and, 20 days later, recipient mice were infected with WSN-OVA1. Eight days later, CD8 T cells isolated from the spleen, medLN, and lungs were stimulated in vitro for 5 h with SIINFEKL and NP366-374 peptides in the presence of brefeldin A and analyzed for intracellular cytokine expression. Equal numbers of splenic-primed and medLN-primed cells were able to produce IFN-γ upon in vitro peptide restimulation (Fig. 7). These data indicated that cells primed in distant locations exhibited not only similar memory cell-generating potential but also equivalent effector functions after recall.
CD8 T cells primed in the medLN and in the spleen have similar capacity to produce IFN-γ upon recall. Mice were treated with Mel-14 or IgG2a and infected as in Fig. 6A. The medLN were harvested from IgG-treated mice, while the spleens were harvested from Mel-14-treated animals, and the cells were enriched for CD8 by magnetic cell separation as in Fig. 6A. Whole CD8-enriched medLN and spleen cells were transferred into naive mice at numbers which contained 20,000 NP/OVA tetramer-positive cells. (A) At 20 days after transfer, recipient mice were infected with WSN-OVA1 and, 8 days later, the spleen, medLN, and lung were harvested and stimulated for 5 h with SIINFEKL peptide (1 μg/ml) and NP366-374 (1 μg/ml) in a standard in vitro restimulation assay. The cells were then stained with fluorescently labeled CD11a, CD8, and CD45.1 antibodies and then permeabilized and stained with fluorescently labeled anti-IFN-γ antibody. Shown are representative plots of gated CD8+ CD45.1+ transferred cells with percentages showing the frequency of IFN-γ-producing cells. (B) Graphs represent the total number of IFN-γ-producing transferred cells in the indicated tissues after recall (n = 4 per group).
DISCUSSION
Understanding more about the pathogenesis of viral infections and the immune response to infections is important for the design and implementation of more effective vaccination strategies aimed at preventing viral infections. Antiviral vaccines combine different antigenic components and adjuvants designed to activate naive CD8 T cells in a manner that will cause their expansion and generate long-lived memory cells. Recent data suggest that a threshold of memory CD8 T cell numbers must be reached to provide protection against infection (46, 47). This threshold likely varies for different pathogens, but failure to reach this threshold may result in poor vaccine efficacy (48, 49). Thus, in an effort to understand the kinetics and localization of T cell priming, we analyzed the activation of labeled transferred transgenic CD8 T cells in secondary lymphoid tissues following respiratory influenza virus infection. We observed evidence of CD8 T cell priming in the medLN as early as 2 days after infection. However, to our surprise, CD8 T cells in the spleen also appeared to have encountered flu antigen. This observation led us to more closely analyze the different sites where virus-specific CD8 T cells encountered antigen after a respiratory IAV infection. Using a short-term transfer assay followed by in vitro culturing, we discovered that virus-specific CD8 T cells were in fact encountering viral antigen in the spleen, as well as the lung-draining medLN, of influenza virus-infected animals. Our results also showed that this phenomenon extended to multiple influenza virus strains and included the presentation of antigen to splenic CD4 T cells.
The origin of flu antigen present in the spleen is not known. One possibility is that there is a period of transient viremia during the period of peak virus titers in the lung. Alternatively, phagocytes or APC carrying viral particles or antigen from the lung could potentially migrate to the spleen. In either case, antigenic material would then be either directly presented or passed to splenic APCs for presentation to CD8 T cells located in the spleen. With viremia, splenic APC could acquire blood-borne viral particles leading to temporally limited antigen cross-presentation. A productive infection may not occur in the spleen given the cell-type-specific constraints on influenza virus replication, and we did not detect replicating virus in the spleen at the time that antigen was present. Indeed, antigen presentation in the spleen was transient, lasting for several days and then dissipating by 11 days after infection. The latter was not the case for the lung-draining LN, where antigen persisted for several weeks as we and others have previously shown (36, 38, 50). Furthermore, our finding that CD8 T cells encountered antigen primarily in the T cell zones of the spleen implies that DC are the likely candidate for APCs. The pathways that transport antigen/pathogens to the splenic T cell zones from the blood are not fully delineated, but in some cases require particular specialized DC subsets (51, 52). A recently described subset of APCs, termed late-activator APCs, migrate into the lung following pulmonary influenza virus infection and subsequently migrate to the spleen and the draining LN with delayed kinetics (53). However, these cells appear in the spleen 6 days after infection, and their numbers peak 8 days after infection and appear to preferentially drive Th2 development (53, 54). Thus, these cells could contribute to T cell activation in the spleen, although our data suggest that splenic priming may begin prior to the arrival of such APCs.
The phenomenon of localized inoculations generating both a local and distant immune response has been previously proposed by the McHeyzer-Williams laboratory (55). Their data showed that two qualitatively different responses are generated in the draining LN and in the spleen after peptide vaccination. Thus, high-affinity CD4 T cells preferentially differentiate into follicular helper T cells in the draining LN, while the response in the spleen is comprised of CD4 T cell clones that are of relatively lower affinity. In our system, we investigated whether antigen presentation in the medLN versus the spleen produced qualitatively different outcomes for CD8 T cells. Cells primed either in the medLN or in the spleen were tested for their ability to secrete IFN-γ, and we observed no quantitative or obvious qualitative differences between the recalled CD8 T cells from the different tissues of origin. This observation was interesting in light of the fact that influenza virus infection induced robust APC activation in the medLN but not in the spleen (data not shown). However, as discussed above, it is not precisely known which APC-presented antigen in the spleen, while the medLN APCs have been extensively characterized after influenza virus infection (7, 56–58).
We also noted that the appearance of antigen in the spleen occurred ∼1 day after antigen appearance in the medLN. This result indicated that T cell priming would be temporally disconnected between the two priming sites. CD8 T cells recruited into the immune response at different time points may receive quantitatively and qualitatively different antigenic stimulation, as previously suggested (59, 60). Naive T cells recruited into the response throughout the course of the infection are likely to encounter decreasing levels of antigen and inflammation and will face increasing competition for antigen and resources as responding T cells expand. Effective T cell lineage differentiation likely requires multiple encounters with APCs for optimal expansion and differentiation (34, 61–63), which in turn may be affected by the time of recruitment. For example, the memory potential of CD4 T cells is inversely proportional to the number of cell divisions undertaken (59), while CD8 T cells that encounter antigen later after infection undergo fewer rounds of cell division but are preferentially recruited into the memory pool (60, 64). In any case, our findings indicated that despite the temporal difference in the initiation of priming in the spleen versus medLN, memory CD8 T cells were efficiently produced in the spleen. Determining the relative contribution of splenic versus LN priming to the overall T cell response is difficult. While splenectomy could be used, compensatory mechanisms and alterations in cell trafficking make the interpretation of results problematic. A very early paper showed that splenectomy had little effect on the outcome of the response (65), although quantitation of T cell responses was less precise at that time. Another report showed that splenectomy, along with Mel-14 treatment, resulted in a substantially diminished CD8 T cell response in the lungs after influenza virus infection, although the underlying mechanisms involved were not identified (66). In addition, in mice lacking LN, the spleen contributes to the overall inflammatory response (67). Interestingly, a recent study using mathematical modeling concluded that the spleen is a major source of effector CD8 T cells after influenza virus infection (68). That study considered the possibility that antigen presentation might occur in the spleen but did not directly measure this parameter. Thus, in conjunction with these studies, our data provide further support that the spleen is an important site of antigen presentation, T cell priming, and memory generation in response to IAV infection and perhaps in other respiratory infections.
ACKNOWLEDGMENTS
This study was supported by NIH grants AI41576 and AI76457 to L.L. D.L.F. was supported by NIH grant AI1083022.
We thank Linda Cauley (University of Connecticut Health Center) for providing some of the virus stocks.
FOOTNOTES
- Received 11 December 2012.
- Accepted 29 January 2013.
- Accepted manuscript posted online 6 February 2013.
- Copyright © 2013, American Society for Microbiology. All Rights Reserved.
