CD8 T cells enter the splenic T cell zones independently of CCR7 but the subsequent expansion and trafficking patterns of effector T cells after infection are dysregulated in the absence of CCR7 migratory cues

Abstract

CCR7 is an important chemokine receptor which regulates T cell trafficking and compartmentalization within secondary lymphoid organs. However, the T cell intrinsic role of CCR7 during infection in the spleen is not well understood. This study was designed to understand how CCR7 dependent localization and migration of CD8⁺ T cells in different compartments of spleen affected the primary and recall responses after infection. To this end we utilized adoptive transfer of naive antigen specific CD8 T cells (OT-I) that either lacked CCR7 or constitutively expressed CCR7 (CD2-CCR7) in mice that were subsequently infected intravenously with Listeria monocytogenes (Lm). We show that naive CCR7^−/− CD8⁺ T cells failed to enter the T cell zone, while CD2-CCR7 OT-I cells were exclusively confined to the T cell zones of the spleen. However, surprisingly CCR7^−/− OT-I cells entered the T cell zones after infection, but the entry and egress migratory pattern of these cells was dysregulated and very distinct compared to WT OT-I cells. Moreover, CCR7 deficient OT-I cells failed to expand robustly when compared to WT OT-I cells and were preferentially skewed towards a short-lived effector cells (SLEC) differentiation pattern. Interestingly, CCR7^−/−, CD2-CCR7 and WT OT-I memory cells responded equally well to rechallenge infection. These results highlight a novel role of CCR7 in regulating effector CD8 T cell migration in the spleen and demonstrate differential requirement of CCR7 for primary and secondary CD8 T cell responses to infection.

Introduction

Spleen is the most important priming site for generating CD8 T cell immunity against blood borne pathogens like Listeria monocytogenes (Lm). The anatomical structure of spleen is compartmentalized into White Pulp (WP), Which is a T cell rich area surrounded by B cell follicles, and the Red Pulp (RP), which is a blood filled area and contains populations of macrophages, dendritic cells (DC) and granulocytes (1). The WP and RP are separated by the marginal zone (MZ), where specific subsets of macrophages as well as CD21^hi B cells reside. The uptake of Lm by CD8α⁺ DCs and their entry into the white pulp is shown to be an important step in the initiation of the CD8 T cell immune response against Lm (2,3). CD8α⁺ DCs capture and transport the bacteria to the splenic white pulp where CD8 T cells encounter Lm derived antigens. A robust CD8⁺ T cell response is required for protective immunity against intracellular pathogens such as Lm.

Naïve T cells recirculate between peripheral blood and secondary lymphoid organs (SLO) in search of antigen (4). When Naïve CD8⁺ T cells encounter cognate antigen, they undergo an expansion phase giving rise to large number of effector cells at the peak of response (5). Subsequently majority of these effector CD8 T cells die and only a small heterogeneous population of memory CD8⁺ T cells survive. These memory CD8⁺ T cells can be further divided into effector memory CD8⁺ T cells (Tem), or central memory CD8⁺ T cells (Tcm) on the basis of CCR7 and CD62L surface protein expression. Tcm express CCR7 or CD62L and thus, mainly reside in lymph nodes and the splenic T cell zones. Conversely, Tem fail to express CCR7 or CD62L and consequently inhabit non lymphoid organs and the splenic RP (6,7). The generation of memory cells can be affected by many factors including; strength and duration of antigenic stimulus, inflammatory milieu, and modulation of chemokine and homing receptors (8–12).

Mounting a protective immune response is critically dependent on the orchestrated movement of cells within lymphoid organs. Secondary lymphoid organ structure is one of the underlying regulators of immune responses and is responsible for promOT-Ing interactions between cells as well as between cells and extracellular matrix. How local migratory properties of naïve or memory CD8 T cells affect priming or recall responses respectively is not well understood. In this vein the role of CCR7 is of particular interest given that CCR7 is a critical regulator of naive, effector as well as memory CD8 T cell migration. CCR7 plays an important role in migration of T cells to and within secondary lymphoid organs. Naive CD8 T cells fail to migrate to lymph nodes and Peyer’s patches of CCR7 deficient mice. Although, CCR7^−/− CD8 T cells migrate to the spleen, they however fail to travel to the splenic T cell zones and remain in the RP (13–16). Conversely, constitutive expression of CCR7 (i.e. in CD2-CCR7 transgenic mice) restricts naive CD8⁺ T cells exclusively to the splenic white pulp (16). In the current study we took advantage of the aforementioned unique properties of CCR7^−/− and CD2-CCR7 transgenic mice to understand how the local compartmentalization of naïve or memory CD8 T cells within the spleen impacts the priming and migration of naïve, or reactivation of memory CD8 T cells following primary or secondary infection respectively. To this end, we investigated whether naive CD8 T cells, which are found in both the splenic T cell zones and in the RP are capable of encountering antigen in the RP or WP, and if the location of the initial priming site within the splenic tissue alters the magnitude or phenotype of effector CD8 T cell expansion and differentiation. In addition, since Tem CD8 T cells fail to express CCR7 or CD62L and thus remain in the RP, we wanted to determine whether these cells migrate to the white pulp to encounter antigen or if they can encounter antigen within the RP after a secondary infection. As with naive CD8 T cells we also investigated whether the location of the initial antigen re-exposure affects memory CD8 T cell recall response. We addressed these question by comparing Ova-specific OT-I TCR transgenic CD8 T cells devoid of CCR7 (CCR7^−/− OT-I), which restrict themselves to RP of the spleen and CD2-CCR7 OT-I cells, which constitutively express CCR7 under CD2 promoter and therefore are exclusively found in the splenic T cell zones. Our results unexpectedly showed that in response to primary or secondary infection CCR7^−/− naïve or memory CD8⁺ T cells in fact migrated to the splenic T cell and B cell zones. Although, early after infection CCR7^−/− effector T cells were capable of entering the T cells zones, the anatomical program followed by these cells following antigen presentation was very distinct from that of the WT effector CD8 T cells. CCR7^−/− effector OT-I cells exited the T cell zones with accelerated kinetics and failed to utilize the bridging channels; apparently migrating through the B cell follicles into the RP (17). The efficiency of the primary CD8 T cell immune response was dependent on CCR7 mediated migratory cues, since CCR7^−/− CD8 T cells failed to expand as well as WT CD8 T cells. Interestingly, in contrast to naive CD8 T cell response, memory CD8 T cell response to a secondary infection was not as dependent on CCR7 mediated migratory cues since CCR7^−/− memory CD8 T cells expanded normally after rechallenge. Thus, our study uncovered novel mechanisms by which CCR7 regulates effector CD8 T cell egress kinetics in the spleen and influences antigen specific CD8 T cell responses to an infection.

Material and Method

Mice

C57BL/6J mice were purchased from Charles River through the NCI animal program. The OT-I mouse line was generously provided by Dr. W.R. Heath (Walter and Eliza Hall Institute, Parkville, Australia) and Dr. F. Carbone (Monash Medical School, Prahan, Australia) and was maintained as a C57BL/6-CD45.1-RAG^−/− line. The transgenic CD2-CCR7 OT-I mouse line was generated by crossing C57BL/6-CD45.1-RAG^−/− line with CD2-CCR7 transgenic line generously provided by Dr. Andrew D. Luster (Massachusetts General Hospital, Harvard University, Boston, US).

Infections

Recombinant L. monocytogenes-producing ovalbumin (LM-OVA) was produced as previously described (Foulds et al., 2002; Pope et al., 2001). The actA⁻ LM-OVA was a generous gift of Dr. John T. Harty (University of Iowa, Iowa City, IA). Unless otherwise mentioned, for all experiments mice were immunized i.v. with 10⁵ colony forming units (c.f.u) actA⁻ L. monocytogenes.

Isolation of lymphocytes

Single-cell suspensions were prepared from spleens after infections at different time-points, lymphocytes were released from the tissue by digestion with 100 U/ml collagenase (Invitrogen, Carlsbad, CA) in RPMI 1640 containing 1 mM MgCl₂, 1 mM CaCl₂, and 10% FCS at 37°C for 30 min. Red blood cells were lysed with ammonium chloride and lymphocytes were filtered through cell strainers.

Immunofluorescence analysis

At the indicated times after infection, lymphocytes were isolated and OT-I cells were detected using a CD45.1-specific mAb. For staining, lymphocytes were suspended in PBS/ 2% FCS/ 0.1% NaN₃/ 2mM EDTA at a concentration of 1x10⁶ to 1x10⁷ cells per milliliter, followed by incubation at 4°C for 30 min with appropriate dilutions and combinations of Abs specific for CD8, CD127, KLRG1, CD69, CD62L, CD11a, CD45.1(Biolegend, San Diego, CA; BD Biosciences, San Jose, CA; or eBioscience, San Diego, CA). Relative fluorescence intensities of up to eight fluorochromes in a single stain were measured with a LSR II (BD Biosciences). Data were analyzed using FlowJo software (Tree Star, Ashland, OR).

CFSE labeling and detection

OT-Is were harvested from spleens of different naïve CD45.1⁺ OT-I mice. Cells were stained with 1 μM of CFDA SE dye in PBS (Vybrant carboxyfluorescein diacetate, succinimidyl ester dye from Life technology). Cells were incubated in 37°C water bath for 12–15 minutes. 2–3 times volume of media containing serum was added. CFSE labeled cells were spun down and suspended in PBS. 0.5 million CFSE labeled cells were transferred to naïve congenic CD45.2⁺ mice, one-day prior to infection with 10⁵ LM-OVA (actA⁻). Spleens were harvested at different days and analyzed for CFSE and CD45.1 by flow cytometry using LSR II.

Intracellular detection of cytokines

Lymphocytes were isolated from the indicated tissues and cultured for 5 h with 1 mg/ml BD GolgiStop (BD Biosciences), with or without 1 mg/ml of the OVA-derived peptide SIINFEKL. After culture, cells were stained for surface molecules and then fixed, and cell membranes were permeabilized in BD Cytofix/Cytoperm solution (BD Biosciences) and stained with anti–IFN-γ, anti–IL-2, or anti–TNF-α mAbs or the appropriate corresponding isotype control rat IgG (BD Biosciences). Cells were then washed, and the fluorescence intensity was measured on a LSRII.

Whole mount confocal laser microscopy

Mice were sacrificed at the indicated times after infection, and the spleens were excised and processed for staining. Briefly, thick sections of whole spleen tissue were cut using a vibratome. Tissues were fixed in 2% paraformaldehyde for 2 h at 4°C. Tissues were subsequently washed and incubated overnight at 4°C in round-bottom 96-well plates with fluorescently labeled anti-CD45.1, B220, CD8 (Biolgened), Biotin-CD169 (Serotec) and anti-mouse CD11c (eBiosciences) diluted in 2% normal goat serum and FCS/PBS solution. On the next day, tissues were washed and incubated overnight at 4°C with Alexa Fluor 488-labeled goat anti-hamster IgG (Invitrogen) and Streptavidin-Cy3 (Biolegend). On the next day, tissues were washed, mounted, and analyzed using a LSM-780 confocal microscope (Zeiss, Oberkochen, Germany). Image analysis was performed using Imaris software (Bitplane, St. Paul, MN).

Memory transfer experiments

Memory OT-I CD8 T cells were generated by transferring 10³ naive CD45.1 cells to naive mice and 1 day later, the mice were infected with 10⁵ LM-Ova. CD8 T cells were enriched by positive selection using AutoMacs (Miltenyi). CD45.1⁺ memory CD8 T cells were FACS sorted from enriched CD8 cells using a FACS Aria (BD Biosciences). For experiments naive mice received 5x10⁴ sorted memory CD45.1 OT-I cells and 24 hours later infected with 10⁵ LM-OVA.

Results

CCR7 is important for naive CD8⁺ T cells migration into the splenic T cell zones

To confirm that CCR7 regulates naive CD8⁺ T cells trafficking to the splenic T cell zones we adoptively transferred 0.5x10⁶ CD45.1⁺ wild type (WT), CCR7^−/− or CD2-CCR7 OT-I cells into uninfected congenic CD45.2⁺ recipient mice intravenously (iv). Mice were sacrificed two days after transfer and the spleens were analyzed by flow cytometry and confocal microscopy. CD2-CCR7 OT-I expressed higher levels of CCR7 on its surface compared to WT OT-I, and CCR7^−/− OT-I failed to express CCR7 (Supplemental Fig. 1A). Within the spleen, WT OT-I cells were observed in both WP and RP, while the CCR7^−/− OT-I were located primarily in the RP (Fig. 1; pie charts). As expected, CD2-CCR7 OT-I cells were located exclusively in the T cell zones. (Fig. 1). The analysis of the peripheral tissues revealed (Supplemental Fig. 1B) that CD2-CCR7 OT-I cells migrated and were retained exclusively in the spleen and lymph nodes, and not in any other peripheral non-lymphoid organs or in circulation. CCR7^−/− OT-I cells were present in the splenic RP and non-lymphoid tissues but not in the lymph nodes. As expected, WT OT-I CD8⁺ T cells were detected in all organs tested (Supplemental Fig. 1B).

Fig. 1. CCR7 is important for the migration of CD8 T cell to the PALS.

Naive mice received 0.5 million CCR7^−/− OT-I, CD2-CCR7 OT-I or WT OT-I cells. Spleens were harvested and sectioned with vibrotome. 400 μm sections were stained for B220, CD8, CD45.1 and moma-1 and analyzed by confocal microscopy. Pie charts show the percentage of OT-I (CD45.1+) CD8 T cells in different compartments of spleen. Light blue arrows point to OT-I CD8 T cells inside the T cell zones whereas orange arrows point to OT-I CD8 T cells in red pulp. WP, White Pulp; RP, Red Pulp; BC, Bridging channel ; T, T cell zone; B, B cell zone. The data are representative of at least 2–3 experiments with n = 3–4 mice for each group.

CCR7^−/− OT-I cells enter the T cell zones early after infection, however, CCR7 migratory cues are required for ordered effector CD8 T cells egress from the splenic T cell zones

To visualize the early activation of antigen specific CD8 T cells we transferred 0.5 x 10⁶ WT, CCR7^−/−, or CD2-CCR7 OT-I cells into naïve mice that were infected intravenously with 10⁵ c.f.u. actA⁻ LM-OVA (from here on referred to as LM-OVA). Spleens were analyzed at 10, 24, 48 and 72 hours after infection for the presence of OT-I cells. As shown in Fig. 2A, in contrast to WT OT-I cells, virtually every CCR7^−/− OT-I cell was either present in MZ or RP at 10 hours (orange arrows). By 24 hours post infection (PI), WT OT-I cells were still localized to the T cell zones (blue arrows), however, CCR7^−/− OT-I cells appeared to migrate from the MZ/RP to the B cell zones (white arrowheads) and surprisingly, several cells were also detected in the T cell zones (blue arrows). By 48 and 72 hours PI, the migratory properties of WT and CCR7^−/− OT-I cells were strikingly distinct. At 48hrs PI majority of the WT OT-I cells were exclusively positioned in the T cell zones (Fig 2A & B). In contrast, a large percentage of CCR7^−/− OT-I cells appeared to have undergone peripheral migration within the white pulp and were primarily located in the B cell zones or at the borders of the B-T cell zones, while a smaller percentage (~18%) were located in the RP. In addition, small clusters of CCR7^−/− OT-I cells could be observed in the RP indicating antigen was also being presented to T cells in the splenic RP (Boxes). By 72 hours PI, WT OT-I cells had undergone a dramatic expansion and could be observed exiting the T cell zones via the bridging channels (BC; Fig 2A) (17). While, CCR7^−/− OT-I cells had proliferated, albeit at a reduced rate compared to WT OT-I cells and were localized in the T and B cell zones. In addition, a greater number of CCR7^−/− OT-I cells were now in the RP. Thus, these data show that in the absence of CCR7 the migratory cues followed by antigen specific CD8 T cells following infection are dramatically distinct when compared to the migratory patterns of WT CD8 T cells. Although CCR7^−/− OT-I cells migrated to the T cell zones from the RP, they failed to use the bridging channels to either enter (24hr PI) or exit (48–72hrs PI) the splenic T cell zones. In fact CCR7^−/− OT-I cells migrated through the B cell zones and majority of them primarily remained in the periphery of the T cell zones and appeared to exit the T cell zones with greater alacrity when compared to WT OT-I cells. The unique localization pattern of CCR7^−/− OT-I cells suggest that even when CCR7 is downregulated on activated T cells, some level of CCR7 signaling continues to guide the entry and egress kinetics of T cells in the spleen. These data clearly show that even in the absence of CCR7, OT-I cells entered the T cell zones, however, majority of these T cells stayed in the periphery of the PALS or entered the B cell zones. The chemokine receptor CXCR3 has been shown to allow CD8 T cell entry into reactive lymph nodes even in the absence of CCR7 (18). Furthermore CXCR3 has been implicated in the peripheral localization of central memory CD8 T cells within the lymph nodes (19) (20). However, its role in trafficking of CD8 T cells in the spleen is not well understood. Moreover, the chemokine receptor CXCR5 is required for entry of T cells (particularly CD4 T cells) to the B cell zones where the ligand CXCL13 is produced. Thus, it is likely that in the absence of CCR7, CXCR3 and CXCR5 may play a role in allowing for the distinct trafficking pattern followed by CCR7^−/− OT-I cells. Indeed at 2 days after infection both CXCR3 and CXCR5 are upregulated in both WT and CCR7^−/− OT-I cells (Supplemental Fig. 2). However, in the case of WT cells, CCR7 mediated cues may yet remain dominant and prevent the disordered peripheral migration of WT CD8 T cells into the B cell zones, as seen with CCR7^−/− OT-I cells.

Fig. 2. CCR7^−/− OT-I cells enter the T cell zones early after infection, however, CCR7 migratory cues are required for ordered effector CD8 T cells egress from the splenic T cell zones.

(A) Naïve mice received 0.5 million CCR7^−/− OT-I and WT OT-I cells and were infected next day with 10⁵ LM-OVA. Mice were sacrificed at 10, 24, 48 and 72 hours after infection and spleens were harvested for confocal microscopy. (B) Pie charts show percentages of OT-I cells (of total OT-I cells) in different splenic compartments. Light blue arrows point to OT-I CD8 T cells inside the T cell zone whereas orange arrows point to OT-I CD8 T cells in red pulp. White arrows represent the OT-I cells present in B cell zone. WP, White Pulp; RP, Red Pulp; BC, Bridging channel ; T, T cell zone; B, B cell zone. The data are representative of at least 2–3 experiments with n= 3 mice for each group.

CCR7 mediated migratory cues determine the magnitude and the differentiation phenotype of responding CD8 T cells

We next determined if the altered migration and localization pattern of CCR7^−/− OT-I cells within the splenic tissue affects the magnitude of effector CD8 T cell expansion and/or the differentiation program of the responding effector CD8 T cells following Lm infection. We reasoned that WT OT-I cells will be primed primarily in the splenic T cell zones and will remain in the splenic T cell zones for the appropriate length of time. Conversely, CCR7^−/− CD8 T cells will likely be primed mainly in the splenic RP and those T cells that do gain access to the T cell zones will exhibit a disordered egress pattern characterized by premature exit from the T cell zones. In addition, CD2-CCR7 OT-I will be primed exclusively in the T cell zones. Therefore, we adoptively transferred 10³ naïve WT, CCR7^−/−, or CD2-CCR7 OT-I in mice. 24 hrs later these mice were infected with LM-OVA. At days 5 and 7 PI the spleen from each mouse was cut in two equal halves with one half used for imaging studies and the other for flow cytometric comparison. As shown in Fig. 3A, at both 5 and 7 days after infection WT OT-I cells were located in both WP and RP, CCR7^−/− OT-I were found largely in red pulp of spleen, while, CD2-CCR7 OT-I were strikingly confined to the T cell zones and failed to exit the splenic WP. Although CCR7^−/− OT-I cells expanded equally at 5 days PI (data not shown), by 7 days the expansion of these cells was significantly reduced when compared to WT or CD2-CCR7 OT-I cells (Fig. 3B). The observed reduced expansion of CCR7^−/− OT-I cells in the spleen was not due to increased expansion of these cells in the peripheral tissues, since we did not find increased numbers of these cells in the lungs or liver (Supplemental Fig. 3A); the expansion in the peripheral organs of CCR7^−/− OT-I cells was also significantly decreased compared to WT OT-I cells. Interestingly, the percentage of CD2-CCR7 OT-I cells in the peripheral tissue was severely reduced, which was likely due their inability to migrate out of the spleen. Although, at the peak of the immune response the expansion of CCR7^−/− OT-I cells was significantly decreased compared to WT OT-I cells, the percentage of CCR7^−/− CD8 T cells capable of secreting IFN-γ was equal to WT or CD2-CCR7 OT-I cells (Fig. 3C). To determine if the initial expansion and replication of OT-I cells in the absence of CCR7 contributes to their poor expansion we evaluated the ability of each OT-I cell population to proliferate early after infection. Indeed, the initial expansion of CCR7^−/− OT-I cells was compromised when compared to WT or CD2-CCR7 OT-I cells (Supplemental Fig. 3B) as judged by CFSE loss at day 2 PI. However, 24 hrs later (at day 3 PI) virtually all groups of T cells present in the spleen exhibited comparable loss of the CFSE stain. Similarly, BrdU incorporation at day 3 PI was comparable for all three types of OT-I cells (Supplemental Fig. 3C). There are many factors, which affect the balance between SLECs (short-lived effector cells, KLRG1^highCD127^low) and MPECs (Memory precursor effector cells, KLRG1^lowCD127^high) formation at early time points after infection. These include pro-inflammatory cytokines, strength of stimulus and its duration. We wanted to analyze whether priming location and the subsequent migratory cues of CD8⁺ T cells in different splenic compartments can have any effect on effector T cell differentiation. To this end, differentiation profile of effector CD8 T cells at earlier time points after infection was analyzed by evaluating the expression of KLRG1 and CD127. Interestingly, significantly more OT-I cells that were not directed by CCR7 mediated migratory cues (CCR7^−/− OT-I) and are primarily located in the RP (at 5 and 7 days PI) had differentiated into SLECs and less MPECs in comparison to WT or CD2-CCR7 OT-I cells (Fig. 3D).

Fig. 3. CCR7 mediated migratory cues determine the magnitude, localization and the differentiation phenotype of responding CD8 T cells.

Naive mice received 10³ different types OT-I cells, one day prior to infection. Spleens were harvested at days 5 and 7 post infection and divided into two halves. (A) One half of the spleens were sectioned for microscopic analysis. Light blue arrows point to OT-I CD8 T cells inside the T cell zone whereas orange arrows point to OT-I CD8 T cells in red pulp and marginal zone. (B) Other half of the spleen was analyzed by flow cytometry at day 7 post infection. (C) Mice received 10³ of the indicated types of OT-I cell one day prior to infection. Spleens were harvested at day 7 post infection and lymphocytes were processed for intracellular cytokine staining. (D) Spleens of mice that received the different types of OT-I cells were analyzed for SLEC (KLRG1^highCD127^low) and MPEC (KLRG1^lowCD127^low) populations at day 5 and 7 post infection. The bar graph shows the percentages of indicated population of total OT-I cells. The data are representative of at least 2–3 experiments with n = 3–4 mice for each group.

CCR7 migratory cues during the priming phase are required for proper memory CD8 T cell generation

Our data thus far clearly showed that CD8 T cell activation and trafficking was dramatically altered during the effector phase of the anti-microbial immune response. Thus in the next set experiments we sought to determine if quantity and quality of memory CD8 T cell was also affected by the absence of T cell intrinsic CCR7. To this end, we first analyzed the localization of all three groups of memory OT-I cells. As shown in Fig.4A, early memory OT-I cells in mice sacrificed at day 30 PI were localized in distinct areas of the spleen. WT OT-I memory cells were widely distributed throughout the different splenic compartments, however, majority were located in the RP and B cell zones (yellow arrows), while a smaller percentage were in T cell zones (blue arrows). CCR7^−/− OT-I memory cells were exclusively found in the splenic RP (orange arrows), whereas CD2-CCR7 memory OT-I cells were exclusively located in the T cell zones. The expression of CCR7 on these early WT memory cells was low (data not shown).

Fig. 4. CCR7 migratory cues during the priming phase are required for proper memory CD8 T cell generation.

(A) Naive mice received 10³ different types of OT-I cells one day prior to infection. Spleens were harvested at day 30 post infection and divided into two halves. (A) One half of the spleen was sectioned for confocal microscopic analysis. Light blue arrows point to OT-I CD8 T cells inside the T cell zone whereas orange arrows point to OT-I CD8 T cells in red pulp and marginal zones. (B) The other half of the spleen was used for flow cytometric analysis. (C) Naive mice received 10³ different types of OT-I cells one day prior to infection. Spleens were harvested at 30 days post infection and lymphocytes were processed for and intracellular cytokine staining. The data are representative of 2–3 experiments with n = 3–4 mice for each group.

Flow cytometric analysis of splenocytes revealed that mice that received WT and CD2-CCR7 OT-I cells harbored a healthy population of memory CD8 T cells. However, memory T cell generation was severely compromised in mice that received CCR7^−/− OT-I cells (Fig. 4B). To confirm that the reduced memory CCR7^−/− OT-I numbers was not due to the increased migration of these cells to the peripheral organs, we assayed various organs for OT-I cells and found that the numbers of CCR7^−/− OT-I cells in the indicated organs was similarly low (Supplemental Fig. 4A). Interestingly, the number of CD2-CCR7 OT-I cells was also dramatically reduced in non-lymphoid organs such as the liver and lung suggesting that after activation the failure to downregulate CCR7 in these cells prevented the migration of effector cells from lymphoid organs to the periphery. Although number of memory CCR7^−/− OT-I cells was dramatically reduced in the spleen, they were functionally as capable of secreting cytokines as the WT memory OT-I cells (Fig. 4C). These data suggest that CCR7 directed migration and localization early during the primary phase of T cell activation is essential for adequate formation of a memory T cell pool.

Expression of CCR7 and localization of memory CD8 T cells within the spleen does not affect the magnitude of recall responses

Since memory CCR7^−/− OT-I cells were localized primarily in splenic RP and memory CD2-CCR7 OT-I cells were exclusively positioned in splenic T cell zones, we next wanted to determine how CCR7 expression and the localization of memory CD8 T cells in different compartments of Spleen would affect recall responses after challenge infection. Mice that were transferred with the three different types of OT-I cells and infected with LM-OVA 30 days earlier were re-challenged with LM-OVA and spleens were imaged at 5 days post recall (PR) As shown in Fig. 5A, WT OT-I cells were observed both in the splenic RP and WP, with majority of the located in the MZ or RP. In contrast, CCR7^−/− OT-I cells were primarily located in the splenic RP. In remarkable contrast, CD2-CCR7 OT-I cells were completely confined to the T cell zones of the spleen (Fig. 5A). Interestingly, at 5 days PR the expansion of all three memory OT-I cells was robust (Fig. 5B), however, the number of CCR7 ^−/− OT-I cells was slightly reduced (~2 fold) compared to WT and CD2-CCR7 OT-I cells. This reduction in responding CCR7^−/− OT-I memory cells was likely due to the reduced numbers of memory cells generated in CCR7^−/− mice (~20 fold reduction; Fig. 4B). To further confirm that the reduction in memory CCR7^−/− OT-I cell expansion was not due to increased migration of these OT-I cells to peripheral organs, we analyzed these organs for the presence of CCR7^−/− OT-I cells and found that their numbers were similarly reduced when compared to the WT OT-I cells (Supplemental Fig. 4B).

Fig. 5. Expression of CCR7 and localization of memory CD8 T cells within the spleen does not affect the magnitude of recall responses.

(A) Previously infected mice that harbored memory OT-I cells were recalled with LM-OVA and spleens were harvested at day 30 post infection. (A) One half of the spleen was sectioned for confocal microscopic analysis. (B) The other half of the spleen was analyzed by flow cytometry. (C) Memory OT-I cells were purified by FACS sorting and 50,000 sorted memory OT-I cells were transferred in to naïve mice. These mice were then recalled with 10⁴ LM-OVA. Spleens were harvested at day 5 and analyzed for the presence of (C) OT-I cells by staining for CD45.1 and (D) SLECs by identified by staining for KLRG1 and CD127. The data are representative of 2–3 experiments with n = 3–4 mice for each group.

To definitively understand how localization of memory CD8⁺ T cells within the spleen influences the recall response we transferred equal numbers of CCR7^−/−, WT or CD2-CCR7 memory OT-I cells (5x10⁴ cells) to naïve mice and challenged the naive mice one day later with LM-OVA. The spleens were isolated at day 5 after infection and the OVA-specific memory T cell response was analyzed. Interestingly, CCR7^−/− OT-I cell expansion was as robust as the WT or CD2-CCR7 OT-I cells (Fig. 5C), suggesting that the CCR7 dependent differential localization of memory CD8⁺ T cells within the spleen did not affect the recall response. This was in contrast to the primary response where proper T cell expansion was heavily dependent on CCR7 migratory cues. We also analyzed the differentiation pattern of recalled memory OT-I cells. SLEC formation of recalled OT-I cells was not affected by the differential expression of CCR7 (Fig. 5D), which was also in contrast to the primary response.

Following re-infection early memory CD8 T cells enter the splenic T cell zones independently of CCR7 through the B cell zones

Since we found that CD8 T cells enter the splenic T cell zones independently of CCR7 at earlier time points after primary infection. Since early memory CD8 T cells also have lower surface expression of CCR7, we wanted to know whether early WT memory OT-I cells or CCR7^−/− OT-I memory cells can enter the T cell zones through the B cell zones. Equal numbers of WT or CCR7^−/− memory OT-I cells were purified from mice infected 30 days previously, and transferred into naive mice that were recalled with LM-OVA. The spleens were excised at day 2 after infection and sections were analyzed for OT-I cells by confocal microscopy. As shown in Fig. 6, even memory CCR7^−/− OT-I cells were observed in B cell zones and T cell zones, which was similar to the primary response. Moreover, not only CCR7^−/−, but also WT OT-I memory CD8 T cells could be observed in the B cell zones. These results suggested that CCR7 low memory CD8 T cells can enter the T cell zones independently of CCR7 and the trafficking pattern of both WT and CCR7^−/− memory CD8 T cells is similar, which is in contrast to what was observed early after primary infection.

Fig. 6. Following re-infection early memory CD8 T cells enter the splenic T cell zones independently of CCR7.

Memory OT-I cells were purified by FACS sorting and 50,000 sorted memory OT-I cells were transferred in to naïve mice that were recalled with 10⁴ LM-OVA. Spleens were sacrificed at day 2 post infection. Splenic sections were analyzed for OT-I by staining with congenic marker CD45.1. White arrow heads represent the OT-I CD8⁺ T cells present in B cell zone. WP, White Pulp; RP, Red Pulp; T, T cell zone; B, B cell zone. All images are acquired with 10X or 20x objectives. The data are representative of 2 experiments with 3 mice for each group.

Discussion

In the present study, we investigated how CCR7 regulates naive, effector and memory CD8 T cell migration after bacterial infection in the spleen. Although the role of CCR7 in regulating CD8 T cell trafficking into the lymph nodes has been extensively studied (12,21), less is known about its effect on T cell migration in the spleen. Since intravenous inoculation of mice with Lm results in an infection which is largely confined to the spleen (22) we were able to specifically visualize CD8 T cell migration within different splenic compartments. As expected, CCR7 was required for the localization of naive T cells in the splenic T cell zones. Initially we reasoned that using adoptive transfer of CCR7 transgenic and deficient OT-I CD8 T cells we will be able to exploit the unique positioning of these antigen specific naive CD8 T cells to ask how CD8 T cell activation is affected when these cells are exclusively primed in T cell zones or in the red pulp of the spleen. After infection CD8α⁺ DCs transport Lm to the splenic T cell zones (2,23) and therefore, it is believed that CD8 T cells encounter antigen and are subsequently primed within T cell zones. Thus, we hypothesized that CCR7^−/− CD8 T cells that fail to enter the T cell zones will be poorly activated. However, to our surprise our results demonstrated that CCR7^−/− OT-I cells in fact migrated to the white pulp at early time points after infection. Since the naive CCR7^−/− CD8 T cells are primarily located in the splenic red pulp, these cells likely migrated to the white pulp after infection and following the transport of Lm to the T cell zones. In this respect the infected splenic T cell zone represents a ‘reactive’ area where bacteria induced inflammation likely results in the production of several different chemokines other than CCL19/21 that attract the CCR7^−/− OT-I cells to migrate to the splenic T cell zones. The identity of this chemokine is currently unknown, but CXCR3 may be a possible candidate (18). However, the main CXCR3 ligand CXCL9 is not expressed in the T cell zones of the spleen, but is primarily produced in the marginal zone (9). These data suggest that CXCR3 may not be responsible for allowing CCR7^−/− OT-I cells to migrate to the T cell zones. Thus, we hypothesize that CCR7^−/− CD8 T cells migrate to the ‘reactive’ T cell zones in the spleen where majority of these T cells are primed. However, we cannot exclude the possibility that a small proportion of these cells encountered antigen very early within hours after infection in the red pulp before expanding and exiting the spleen via the red pulp. Our data clearly showed that even in the absence of CCR7 OT-I CD8 T cells migrated to the splenic T cell zones where they were primed, however, after infection, the migratory and activation patterns of OT-I cells in the absence of CCR7 were dramatically distinct when compared to WT OT-I cells. Our data clearly demonstrated that although the entire population of the transferred CCR7^−/− OT-I cells in spleen had been primed (as judged by CD69 expression at 24hrs after infection; data not shown) these cells failed to expand with the same magnitude as the WT or CD2-CCR7 OT-I cells. This result could be partially explained by the fact that a greater percentage of CCR7^−/− OT-I cells exhibited the SLEC phenotype and thus, were terminally differentiated. Interestingly, at early time points (day 3 PI), both WT and CCR7^−/− OT-I cells exhibited comparable CFSE dilution, and at 5 days PI the number of WT and CCR7^−/− OT-I cells in the spleen was similar. However, by 7 days PI the number and frequency of CCR7^−/− OT-I cells in the spleen was dramatically reduced. This could be due to a combination of two reasons: since a greater frequency of CCR7^−/− OT-I cells are terminally differentiated, these cells are likely undergoing early apoptosis. Moreover, the failure to expand adequately is likely also related to the inability of CCR7^−/− OT-I cells to be sequestered within the T cell zones long enough to receive the adequate amount of signals from antigen presenting cells leading to poor proliferation and survival. As noted earlier, CCR7^−/− OT-I cells entered the T cell zones, however, majority of these T cells stayed in the periphery of the PALS or entered the B cell zones. These cells expressed high levels of CXCR3 and CXCR5 and thus, it is likely that in the absence of CCR7, both CXCR3 and CXCR5 may play dominant roles in mediating the disordered peripheral positioning and migration of CCR7^−/− OT-I cells through the B cell zones. This is in contrast to the protracted sequestration of WT OT-I cells in the splenic T cell zones and the ordered egress of WT OT-I cells from the T cell zones into the rep pulp via the bridging channels (17) following infection. Thus, our study clearly demonstrated that although the cell surface expression of CCR7 is reduced on WT OT-I cells after activation some level of tonic CCR7 mediated cues may yet remain dominant and prevent the disordered peripheral migration of WT CD8 T cells into the B cell zones, as observed with CCR7^−/− OT-I cells. Moreover, the migratory and localization cues provided by CCR7 are essential for the adequate expansion and differentiation of CD8 T cells early after infection in the spleen.

Interestingly, CCR7 mediated signals were not important for the proper expansion of memory CD8 T cells. This may be related to the possibility that memory CD8 T cells exhibit a lower threshold of activation and do not require protracted sequestration in the splenic T cell zones after rechallenge for proper expansion and differentiation. Moreover, even WT memory CD8 T cells at 30 days post infection (early memory CD8 T cells) fail to express high levels of CCR7 and in this respect they are similar to CCR7^−/− and thus, they appear to exhibit similar trafficking patterns that are very distinct from primary effector CD8 T cells.

In summary our study showed that alterations in CCR7 expression can dramatically influence localization, migration and activation of antigen specific CD8 T cells in the spleen following a bacterial infection. Thus, our results will advance the understanding of how CCR7 mediated guidance cues affect the distribution and localization of CD8 T cells within the splenic microenvironment and how these guidance cues shape the differentiation and function of effector and memory CD8 T cells following infection.

Abbreviations

MZ

Marginal zone

RP

Red pulp

WP

White pulp

PI

Post infection

PR

Post recall

Lm

Listeria monocytogenes

OVA

Ovalbumin

SLO

Secondary lymphoid organs