Resident memory CD8 T cells (TRM) are a nonrecirculating subset positioned in nonlymphoid tissues to provide early responses to reinfection. Although TRM are associated with nonlymphoid tissues, we asked whether they populated secondary lymphoid organs (SLO). We show that a subset of virus-specific memory CD8 T cells in SLO exhibit phenotypic signatures associated with TRM, including CD69 expression. Parabiosis revealed that SLO CD69+ memory CD8 T cells do not circulate, defining them as TRM. SLO TRM were overrepresented in IL-15–deficient mice, suggesting independent regulation compared with central memory CD8 T cells and effector memory CD8 T cells. These cells were positioned at SLO entry points for peripheral Ags: the splenic marginal zone, red pulp, and lymph node sinuses. Consistent with a potential role in guarding SLO pathogen entry points, SLO TRM did not vacate their position in response to peripheral alarm signals. These data extend the range of tissue resident memory to SLO.
Memory CD8 T cells are widely distributed throughout the body and are partitioned into subsets based on trafficking properties (1, 2). Central memory CD8 T cells (TCM) recirculate through secondary lymphoid organs (SLO). Effector memory CD8 T cells (TEM) recirculate through nonlymphoid tissues (NLT). Resident memory CD8 T cells (TRM) are a newly recognized subset that patrols nonlymphoid frontline sites of pathogen exposure without recirculating (3–5). CD69 is expressed upon TCR stimulation by early effectors and T cells responding to chronic infections. CD69 is also constitutively expressed by TRM after Ag clearance, suggesting a separate pathway for regulation (6, 7). Tissue microenvironments induce constitutive CD69 expression by TRM, tying this phenotype to location (4, 6–9). Moreover, CD69 expression is associated with retention in tissues, suggesting that this marker plays a defining functional role in TRM differentiation (8).
TRM play a specialized sentinel role and are positioned to rapidly alert the host in the event of reinfection and accelerate pathogen clearance (4, 5, 10, 11). Thus, their anatomic localization at initial sites of Ag exposure is critical for the conceptualization of TRM function. Although these early responders are tightly associated with NLT distribution, SLO are compartmentalized and also contain “frontline” sites of initial Ag encounter (12). For instance, the subcapsular and medullary sinuses of lymph nodes (LN) capture lymph-borne pathogens derived from tissues (13, 14). In spleen, hematogenous pathogens are first encountered in the red pulp or trapped in the marginal zone (15). Thus, we hypothesized that TRM also may exist in the “frontline” portions of lymphoid tissue.
Materials and Methods
Mice and infections
All mice were used in accordance with the guidelines of the Institutional Animal Care and Use Committees at the University of Minnesota. C57BL/6J mice were purchased from The Jackson Laboratory. Thy1.1+ P14, CD45.1+ OT-I, and CD45.2+ OT-I mice were fully backcrossed to C57BL/6J mice and maintained in our animal colony. P14 immune chimeras were generated by transferring 5 × 104 naive transgenic Thy1.1+ P14 T cells into naive wild-type or IL-15−/− C57BL/6J mice and infecting with 2 × 105 PFU lymphocytic choriomeningitis virus (LCMV, Armstrong strain) i.p. the next day. Memory OT-I cells were generated by transferring 5 × 104 naive transgenic CD45.1+ OT-I T cells into memory P14 chimeras. The next day, recipients were infected with 1 × 106 PFU recombinant vesicular stomatitis virus expressing OVA (VSV-OVA) i.v., as described (16). Naive CD45.2+ OT-I T cells were injected into CD45.1+ RAG−/− mice, as indicated. For local rechallenge experiments, 50 μg gp33 peptide was delivered transcervically, as described (10), in a volume of 35 μl delivered by modified gel loading pipette.
Parabiosis surgeries were performed as described (10). Briefly, mice were shaved along opposite lateral flanks. Skin was wiped clean of fur with alcohol prep pads and further cleaned with Betadine solution and 70% isopropyl alcohol. Mirrored incisions were made on lateral aspects of both mice, and 5-0 VICRYL thread was used to suture skin to conjoin the mice. Additional sutures were placed through the olecranon and knee joints to secure the legs.
Immunofluorescence microscopy and flow cytometry
For immunofluorescence microscopy, tissues were frozen in 2-methylbutane surrounded by dry ice. Frozen blocks were cut into 7-μm sections, fixed in acetone, blocked in a 5% BSA PBS solution for 1 h, and stained with DAPI (Invitrogen) and Abs specific for Thy1.1 (OX-7; eBioscience), CD45.1 (A20; eBioscience), CD169 (AbD Serotec), CD69 (goat polyclonal; R&D Systems), and Collagen IV (rabbit polyclonal; Acris). Jackson ImmunoResearch secondary Abs conjugated to various fluorochromes were used to stain unconjugated Abs. Tiled images were acquired with an automated Leica DM5500B microscope, and analysis was performed with ImageJ and Adobe Photoshop. Cell isolations and flow cytometry were performed as previously described (17). Quantification of microscopy images was performed as previously described (10).
The p values were determined by a two-tailed, unpaired Student t test. Differences between groups were considered significant if p ≤ 0.05.
Results and Discussion
A subset of memory CD8 T cells in SLO phenotypically resembles TRM
To determine whether a population of memory CD8 T cells within SLO expressed canonical markers of TRM, we analyzed Thy1.1+ memory P14 CD8 T cells 8 wk after LCMV Armstrong infection (see 2Materials and Methods). P14 small intestine intraepithelial lymphocytes were used as a positive control because they were shown to be TRM in this infection model (3). Flow cytometric analysis revealed a subset of CD69+ memory CD8 T cells within both the LN (data not shown) and the spleen (Fig. 1A). CD69+ memory CD8 T cells within the SLO lacked CD62L, a cardinal feature of recirculating TCM. This unusual SLO population was distinct from canonical TEM (CD69− CD62L−) and TCM (CD69− CD62L+) but was similar to NLT-derived TRM, by expression of Ly6C, KLRG1, and Eomes (Fig. 1B). Unlike TRM that have been characterized in or near epithelial surfaces, CD69+ memory CD8 T cells within the SLO did not express CD103 (i.e., αeβ7 integrin; data not shown). It should be noted that CD69 expression was upregulated after clearance of LCMV, suggesting that persistent foreign Ag was not driving expression (Fig. 1C). To support this interpretation, we transferred naive Rag−/− OT-I CD8 T cells to lymphopenic RAG−/− mice, which results in homeostatically expanded memory-like T cell differentiation without pathogen-derived Ag (6). Under these conditions, we also observed a CD8 T cell subset within SLO that exhibited a TRM-like phenotype (Fig. 1D).
CD69+ SLO memory CD8 T cells are TRM positioned at common Ag entry sites
We wished to test whether CD69+ memory CD8 T cells within SLO were TRM. The vascular systems of LCMV-immune mice (generated as in Fig. 1) were surgically conjoined to naive mice via parabiosis, which equilibrates TCM and TEM between hosts without equilibrating TRM (5, 18). Fig. 2A shows that the CD69+ SLO memory CD8 T cell population was specifically absent in the naive parabiont, demonstrating that this population was indeed resident within SLO. Immunofluorescence microscopy localized these TRM preferentially to the marginal zone and red pulp of the spleen and the subcapsular and medullary sinuses of the LN (Fig. 2B–E), which represent the primary initial sites of Ag entry into these SLO.
Reduced IL-15 dependence of CD8 TRM in SLO
Previous reports showed that maintenance of TCM and TEM is critically dependent on IL-15 (19). We found that CD69+ TRM within SLO express reduced levels of CD122 (IL-15Rβ, Fig. 3A), similar to what was reported for TRM located within the epithelium of the small intestine (17). We hypothesized that a distinguishing feature between SLO TRM and recirculating memory CD8 T cells could be dependence upon IL-15. To test this hypothesis, we compared memory P14 differentiation and maintenance in wild-type and IL-15–deficient LCMV-immune chimeras (see 2Materials and Methods). Sixty days after LCMV infection, we found that CD69+ SLO TRM were overrepresented in the absence of IL-15 (Fig. 3B, 3C). These data suggest that SLO TRM are less dependent upon IL-15 for their survival compared with recirculating primary TCM and TEM, which further defines them as a unique subset.
SLO TRM do not redistribute in response to peripheral alarm signals
Upon Ag recognition within the female reproductive mucosa, reactivated local memory CD8 TRM precipitate alarm signals (10). These signals recruit bystander resting memory CD8 T cells, which redistribute from spleen and blood to the site of reactivation in the mucosa (10). We asked whether this redistribution function was restricted to recirculating memory CD8 T cells or whether CD8 TRM in the SLO also would mobilize to the female reproductive tract in this context. To this end, we transferred naive OT-I CD8 T cells into P14 LCMV-immune chimeras. The next day, we infected these mice with VSV-OVA. Thirty days later, these mice contained populations of both OT-I and P14 memory CD8 T cells. They were then rechallenged transcervically with gp33 peptide, as described (10), to reactivate P14 within the female reproductive tract and precipitate local alarm signals that induce redistribution of memory CD8 T cells. CD69+ and CD69− OT-I CD8 T cells were enumerated within the spleen (2 d after gp33 peptide rechallenge) (Fig. 4). Unlike recirculating memory OT-I CD8 T cells, many of which vacated SLO to respond to the site of peripheral P14 CD8 T cell reactivation, the number of CD69+ memory OT-I CD8 T cells was unaffected. This indicates that, unlike recirculating memory CD8 T cell subsets, TRM maintain their presence in the SLO even in the context of peripheral alarm signals.
TRM have been identified in many NLT, as well as the thymus, where they are positioned to rapidly provide protection against pathogens re-encountered at anatomic barriers (4, 5, 7, 11, 20). TRM express unique phenotypic signatures, including CD69 expression, that distinguish them from recirculating TCM and TEM (4, 8, 17). TRM differentiation is induced by local environmental cues; thus, TRM ontogeny is intrinsically coupled with location (6, 8). Therefore, it is surprising that we defined a population of bona fide TRM via parabiosis experiments, which also could be identified by CD69 expression, present within lymphoid tissues. It should be noted that CD69+ CD8 T cells also have been detected in human SLO (21, 22).
It remains possible that local inductive cues also regulate TRM ontogeny within SLO microenvironments. Indeed, SLO TRM occupied specific niches within lymphoid organs, particularly the subcapsular and medullary sinuses of LN and the marginal zones of spleen, which showed minimal overlap with TCM and TEM. These compartments also represent major portals for pathogen and Ag entry into lymphoid organs (12). In light of these findings, there may be operational similarities between SLO TRM and those in NLT tissues: as guardians of frontline sites of pathogen exposure. It is telling that, unlike CD69− recirculating memory CD8 T cell populations, SLO TRM did not abandon the lymphoid tissue in response to distant inflammatory cues. Perhaps they fulfill a specialized immunosurveillance role that, like NLT TRM, accelerates alarm signals and pathogen control in response to reinfection.
This work was supported by National Institutes of Health Grants R01AI084913-01 (to D.M.), DP2OD006467 (by the Office of the Director) (to D.M.), T32AI007313, and F30DK100159-01 (both to J.M.S.).
Abbreviations used in this article:
lymphocytic choriomeningitis virus
secondary lymphoid organ
central memory CD8 T cell
effector memory CD8 T cell
resident memory CD8 T cell
vesicular stomatitis virus expressing OVA.
The authors have no financial conflicts of interest.