Lymphatic endothelial cells prime naïve CD8+ T cells into memory cells under steady-state conditions

Vokali, Efthymia; Yu, Shann S.; Hirosue, Sachiko; Rinçon-Restrepo, Marcela; V. Duraes, Fernanda; Scherer, Stefanie; Corthésy-Henrioud, Patricia; Kilarski, Witold W.; Mondino, Anna; Zehn, Dietmar; Hugues, Stéphanie; Swartz, Melody A.

doi:10.1038/s41467-019-14127-9

Download PDF

Article
Open access
Published: 27 January 2020

Lymphatic endothelial cells prime naïve CD8⁺ T cells into memory cells under steady-state conditions

Efthymia Vokali¹^na1,
Shann S. Yu ORCID: orcid.org/0000-0002-4461-793X^1,2^na1,
Sachiko Hirosue¹,
Marcela Rinçon-Restrepo¹,
Fernanda V. Duraes³,
Stefanie Scherer⁴,
Patricia Corthésy-Henrioud¹,
Witold W. Kilarski ORCID: orcid.org/0000-0003-0734-4749^1,2,
Anna Mondino⁵,
Dietmar Zehn ORCID: orcid.org/0000-0003-1393-8527⁴,
Stéphanie Hugues³ &
…
Melody A. Swartz ORCID: orcid.org/0000-0001-5484-1903^1,2,6

Nature Communications volume 11, Article number: 538 (2020) Cite this article

10k Accesses
43 Citations
6 Altmetric
Metrics details

Subjects

Abstract

Lymphatic endothelial cells (LECs) chemoattract naïve T cells and promote their survival in the lymph nodes, and can cross-present antigens to naïve CD8⁺ T cells to drive their proliferation despite lacking key costimulatory molecules. However, the functional consequence of LEC priming of CD8⁺ T cells is unknown. Here, we show that while many proliferating LEC-educated T cells enter early apoptosis, the remainders comprise a long-lived memory subset, with transcriptional, metabolic, and phenotypic features of central memory and stem cell-like memory T cells. In vivo, these memory cells preferentially home to lymph nodes and display rapid proliferation and effector differentiation following memory recall, and can protect mice against a subsequent bacterial infection. These findings introduce a new immunomodulatory role for LECs in directly generating a memory-like subset of quiescent yet antigen-experienced CD8⁺ T cells that are long-lived and can rapidly differentiate into effector cells upon inflammatory antigenic challenge.

Immunization-induced antigen archiving enhances local memory CD8+ T cell responses following an unrelated viral infection

Article Open access 21 March 2024

The precursors of CD8+ tissue resident memory T cells: from lymphoid organs to infected tissues

Article 03 September 2021

Requirements for the differentiation of innate T-bethigh memory-phenotype CD4+ T lymphocytes under steady state

Article Open access 06 July 2020

Introduction

The lymphatic endothelium is actively involved in shaping T cell responses, both directly and indirectly^1,2,3,4,5,6. In the lymph nodes (LNs), lymphatic endothelial cells (LECs) have been shown to present peripheral tissue-restricted antigens on MHC class I molecules (MHCI) to induce tolerance in autoreactive CD8⁺ T cells^7,8. In addition to endogenous antigen presentation, we and others have demonstrated that LECs abundantly scavenge and can cross-present exogenous antigens to naïve CD8⁺ T cells and drive proliferation, both in the context of a tumor⁹ as well as under homeostatic conditions¹⁰. These LEC-activated T cells lack effector cytokine secretion and some undergo early-generation apoptosis, suggesting a dysfunctional or anergic phenotype. LECs have therefore been established as antigen-presenting cells (APCs) that can sample the peripheral exogenous antigen repertoire and contribute to CD8⁺ T cell regulation and tolerance.

The induction of peripheral tolerance has been primarily assigned to professional APCs¹¹. Under homeostatic conditions, LN-resident immature dendritic cells (DCs) acquire self-antigens from peripheral tissues and cross-present them without co-stimulation, inducing T cell anergy or deletion¹². In this case, peripherally tolerized CD8⁺ T cells remain functionally impaired, even in response to pathogenic challenge¹³. In contrast, apparently tolerized CD8⁺ T cells that were primed by non-hematopoietic cells—such as intestinal epithelial cells and liver sinusoidal endothelial cells (LSECs)—can exit the non-responsive state and be functionally reactivated under pro-inflammatory conditions^14,15. We therefore asked whether quiescent LEC-educated CD8⁺ T cells can be functionally activated upon inflammatory challenge, thereby representing a memory-like state rather than an irreversibly anergic or tolerized state.

Canonically, memory T cells are considered to arise during the contraction phase of an immune response (e.g., after primary infection or vaccination), differentiated from a subset of effector T cells. However, other models propose that memory T cells can arise directly from naïve cells, either as precursors to effector memory and effector T cells (i.e., the loss of memory correlates with terminal differentiation), or as asymmetric divisions reminiscent of stem cell behavior^16,17. The memory T cell pool is also highly heterogeneous, as their phenotypic state is tightly governed by the integration of numerous signals including antigen load and persistence, TCR affinity, co-stimulation, cytokine environment, and CD4⁺ T cell help^16,17,18. An emerging consensus suggests that higher levels of TCR–pMHC affinity, antigen load, co-stimulation, and pro-inflammatory cytokines promote differentiation into terminal effectors, while weaker activation signals promote a shift towards memory T cells¹⁹. This new understanding raises important considerations with regard to CD8⁺ T cell priming by LECs and other non-hematopoietic cells, which occurs in the context of lower levels of co-stimulation and pro-inflammatory cytokines and higher levels of co-inhibitory signals relative to priming by activated DCs¹⁰.

Here, we describe a new role for LECs in directly cross-priming naïve CD8⁺ T cells into long-lived memory T cells capable of potent effector differentiation following inflammatory challenges. This has implications for understanding T cell memory and also helps to reconcile the seemingly contradictory roles of lymphangiogenesis in tumors versus in chronic inflammatory diseases or transplantation. We suggest that LEC-educated T cells may function as a reserve of antigen-experienced T cells in a memory-like state, enhancing immune responses upon immunogenic challenge or promoting tolerance upon challenge under immunosuppressive conditions, as in the tumor microenvironment.

Results

Priming by stromal cells promotes CD8⁺ T cell memory

First, we sought to determine the relevance of antigen uptake and cross-presentation by LECs on CD8⁺ T cell responses in vivo. Creating mice in which MHCI presentation is restricted to LECs was not feasible, since there is no known truly lymphatic-specific gene (e.g., Lyve1 is expressed by macrophage subsets, Prox1 by hepatocytes, and podoplanin (Pdpn) by fibroblast subsets). Instead, we generated bone marrow chimeric mice in which donor CD45⁺ cells lacked MHCI (β₂m^–/–:WT, denoting the status of the hematopoietic population:non-hematopoietic remainder of the host), and therefore presentation of the H-2K^b-restricted SIINFEKL epitope after intradermal (i.d.) injection of OVA would be restricted to radioresistant lymph node stromal cells (LNSCs). LNSCs include fibroblastic reticular cells (FRCs), follicular DCs, and blood endothelial cells (BECs) as well as LECs. However, among these LNSC subtypes, LECs take up the majority of antigen in the draining LNs (dLNs) following i.d. administration under steady-state conditions¹⁰. Furthermore, when loaded with antigen ex vivo, FRCs induced negligible T cell proliferation in vitro compared to LECs (Supplementary Fig. 1). We therefore reasoned that LECs would be the primary Ag-presenting cells for OVA following i.d. administration in β₂m^–/–:WT chimeric mice, with little contribution from FRCs^9,20. As controls, we also generated inverse chimeras (WT:β₂m^–/–) as well as positive (WT:WT) and negative (β₂m^–/–:β₂m^–/–) controls.

We transferred naïve CFSE-labeled OT-I CD8⁺ T cells (CD45.1/2⁺) intravenously (i.v.) into the mice and the following day administered i.d. OVA (50 µg) or PBS (Fig. 1a). Five days later, we analyzed the OT-I cells in skin-dLNs and observed substantial levels of antigen-specific proliferation in β₂m^–/–:WT mice compared to those injected with PBS, suggesting that indeed, MHCI-antigen presentation by LNSCs can induce T cell proliferation in vivo (Fig. 1b). As expected, transferred OT-I cells proliferated the most in WT:WT mice and the least in β₂m^–/–:β₂m^–/– mice, while interestingly, proliferation was comparable in WT:β₂m^–/– and β₂m^–/–:WT mice (both ~70% of levels in WT:WT mice), indicating that LNSC participation in T cell priming under steady-state conditions is important and physiologically relevant compared to that induced by DCs and other hematopoietic cells.

**Fig. 1: CD8⁺ T cells educated by lymph node stromal cells display memory-like characteristics.**

Next, we evaluated the phenotype of the LNSC-educated OT-I cells. In all mice except β₂m^–/–:β₂m^–/– chimeras, the majority of transferred cells expressed CD44⁺ with both central memory (T_CM, CD44⁺CD62L⁺) and effector/effector memory (T_eff/EM, CD44⁺CD62L⁻) subsets in all groups (Fig. 1c). Interestingly, the T_CM fraction was highest in the β₂m^–/–:WT mice (Fig. 1d), where the ratio of T_CM to T_eff/EM was close to unity (Fig. 1e), while in mice lacking MHCI on LNSC (WT:β₂m^–/–), ~30% of total OT-I cells were T_CM. Upon ex vivo SIINFEKL re-stimulation, the OT-I cell fractions that were IFNγ⁺ or IL-2⁺ were similar among the groups (Fig. 1f, g), but the fraction of double-positive cells was lowest in WT:β₂m^–/– mice (Fig. 1h) and the fraction of IFNγ⁺ cells expressing IL-2 was highest in β₂m^–/–:WT mice (Fig. 1i). These results established that (i) LNSCs are capable of antigen presentation under non-inflammatory conditions to induce the proliferation and activation of CD8⁺ T cells, and (ii) T cell education by LNSCs promotes a memory phenotype.

LECs uniquely generate CD8⁺ memory

Although LECs represented the dominant LNSC subset that both take up i.d.-delivered antigen¹⁰ and can cross-present antigen to naïve CD8⁺ T cells (Supplementary Fig. 1), we could not rule out the unlikely possibility that migratory DCs such as Langerhans cells, which are less radiation-sensitive than circulating DCs²¹, could contribute to the memory skewing that we saw (although previous in vitro studies using co-cultures of such cells with CD8⁺ T cells did not recapitulate the phenotype we observed^21,22,23). To address this, we first exposed LN-resident cells to antigen in vivo and then tested their ability to activate cognate T cells ex vivo. Specifically, we injected WT mice i.d. with OVA or PBS, and after 2 h, isolated the dLNs and sorted for four subsets: DCs (CD45⁺ CD11c⁺), macrophages (Mφ; CD45⁺ F4/80⁺), LECs (CD45⁻ gp38⁺ CD31⁺), and FRCs (CD45⁻ gp38⁺ CD31⁻). These subsets (or antigen-pulsed, LPS-activated bone marrow-derived DCs (BMDCs) as a positive control) were immediately placed into culture with naïve OT-I CD8⁺ T cells for 3 days (Fig. 2a).

**Fig. 2: LECs that had taken up antigen in vivo can activate T cells ex vivo.**

We reasoned that after 2 h, injected antigens would have already drained to and perfused the sinuses of the dLNs, whereas migratory DCs that picked up antigen from the skin would not yet have arrived^24,25,26. Consistent with this, 71 ± 4% of LECs were positive for antigen at this time, versus 14 ± 4% of LN-resident DCs (Fig. 2b, c). Nevertheless, the antigen-positive DCs had taken up more antigen per cell than the antigen-positive LECs, showcasing their propensity to sample and internalize antigen (Fig. 2d).

Without antigen, overall OT-I cell survival was highest when co-cultured with LECs and lowest with FRCs, consistent with previous reports that LECs are the major source of the pro-survival cytokine IL-7 in the LN^27,28. With antigen, LECs were the only subset that increased T cell numbers (Fig. 2e) and activated T cells (Fig. 2f) relative to PBS. There was no appreciable increase in T cell numbers when cultured with DCs, presumably due to inefficient in vivo targeting of DCs compared to LECs after 2 h (Fig. 2c). As expected, the overall differences seen in OVA-treated vs. PBS-treated mice for all groups were small compared to when antigen pulsing is done in vitro, where all of the cells would be exposed to antigen (Supplementary Fig. 1).

In addition, antigen education by LECs increased intracellular IFNγ⁺ in OT-I CD8⁺ T cells (Fig. 2g). Interestingly, this was not reflected in the levels of IFNγ in the supernatant, which were low in all cultures compared to BMDC-educated T cells (Fig. 2h). On the other hand, IL-2 levels were significantly elevated in LEC and DC co-cultures from OVA-treated vs. PBS-treated mice. These results suggest that LEC education activates the CD8⁺ T cells to a state where they are poised to generate IFNγ without secreting it, while increasing secretion of IL-2 to support homeostatic proliferation and survival¹⁸.

We next explored the phenotype of these activated OT-I cells. Interestingly, the IFNγ⁺ cells activated by LECs isolated from OVA-treated vs. PBS-treated mice occurred mainly in CD62L⁺ cells (Fig. 2i); indeed, the T_CM compartment accounted for ~55% of IFNγ⁺ T cells (Fig. 2j). These IFNγ⁺ OT-I cells also stained positively for other markers of memory cells, including the receptors to IL-2, IL-7, and IL-2/15 (CD25, CD127, and CD122, respectively) as well as the co-stimulatory receptor CD27 (Fig. 2k, l). Collectively, these data suggest that in the absence of inflammatory signals, antigen cross-presentation by LECs under steady-state conditions may directly promote CD8⁺ T cell memory.

To further test whether LECs can indeed promote the acquisition of a memory phenotype by CD8⁺ T cells in vivo, we designed a method to introduce β₂m-competent LECs into β₂m^–/– mice, such that the transplanted LECs would be the sole source of peptide-MHCI systemically. For such a system, we used LECs isolated from Actb-mOVA mice, where expression of membrane-bound OVA is driven by the β-actin promoter in all cells. We confirmed that in vitro, these LECs could stimulate the acquisition of CD44⁺CD62L⁺ phenotype by co-cultured naïve OT-I cells (Fig. 3a; 28 ± 13%, relative to 7.6 ± 0.4% with wild-type, unpulsed LECs), albeit to a lesser degree than OT-I cells co-cultured with OVA-pulsed control LECs (84 ± 2%).

**Fig. 3: LECs induce memory phenotype in cognate CD8⁺ T cells in vivo.**

In our hands, injections (i.d., intraperitoneal, intravascular) of LEC cell suspensions produced poor survival and recovery even within 1 day post-administration, so we opted to produce 3D spheroid cultures for implantation, based on the observation that many endothelial cells maintain better structure and function in the presence of self-derived extracellular matrix proteins. Spheroids (18 ± 3 μm diameters) were implanted into i.d. pockets formed between the ear cartilage and the ventral skin (Fig. 3b), where they survived more than 8 days in fully immunocompetent mice (Fig. 3c). β₂m^–/– mice implanted with WT or OVA-expressing LEC spheroids later received a co-transfer of OT-I and bystander CD8⁺ T cells at a 1:5 ratio (Fig. 3d). While there were no differences in the overall numbers of OT-I or bystander CD8⁺ T cells recovered in the blood at 1 week post-transfer (Fig. 3e), there were fewer naïve (CD44⁻CD62L⁺) cells and more memory-like (CD44⁺CD62L⁺) cells among OT-I cells in mice that had received OVA-expressing LECs (Fig. 3f). Moreover, among CD8⁺ T cells isolated from the implant site, mice that received OVA-expressing LECs specifically expanded OT-I cells, and a higher frequency of those OT-I cells expressed CD69 (Fig. 3e, g, h). This increased activation was only seen at the implant site, rather than the superficial cervical LNs draining the implant site (Fig. 3g, h). Together with the data shown in figs. 1 and 2, these data indicate that antigen presentation by LECs in vivo results in the acquisition of memory phenotype by cognate CD8⁺ T cells.

LECs prime naïve CD8⁺ T cells into T_CM/T_SCM-like states

To further examine APC-intrinsic differences in the education and activation of cognate CD8⁺ T cells, we turned to in vitro activation studies using CD8⁺ T cells from OT-I mice in co-culture with primary LN LECs or LPS-matured BMDCs (herein referred to as mDC). For antigen, we used a nanoparticle delivery platform for SIINFEKL with a reduction-sensitive linkage, NP-ss-COVA_250–264 (Fig. 4a–h, Supplementary. Fig. 2a–e), which our group has extensively characterized as a tool to detect efficient cross-presentation^10,29. Within 3 days, LEC-educated CD8⁺ T cells proliferated robustly and increased in size compared to naïve cells, indicative of their activated state, although they did not acquire the large blastoid morphology of mDC-educated cells (Fig. 4a, b). Interestingly, we observed that LEC-educated cells primarily expressed a CD44⁺CD62L⁺ phenotype, which was clearly distinct from the effector-like phenotype of mDC-educated cells. This was also distinct from the phenotype of OT-I cells co-cultured with LN FRCs in the presence of OVA_257–264, which proliferated significantly less than LEC-educated cells and also induced some effector cells (Supplementary Fig. 1). Therefore, OT-I CD8⁺ T cells co-cultured in vitro with LECs, but not with FRCs, acquired a similar phenotype as those educated in vivo in the β₂m^–/–:WT chimeras, further confirming that among the major non-hematopoietic LNSC populations, LECs rather than FRCs promote the acquisition of memory phenotype in CD8⁺ T cells.

**Fig. 4: Surviving LEC-primed CD8⁺ T cells express T_CM/T_SCM-like phenotype.**

We next analyzed the kinetics of T cell phenotype of LEC-educated versus mDC-educated OT-I cells, and we observed that CD62L was initially downregulated on both LEC-educated and mDC-educated cells (Fig. 4b, day 1), consistent with reports that CD62L is cleaved from the cell surface following TCR engagement³⁰. However, as LEC-educated cells proliferated, they predominantly expressed CD62L, whereas the mDC-educated T cells followed the opposite trend. Moreover, mDC-educated cells expanded continously, while LEC-educated cell numbers fluctuated over time (Fig. 4c). In agreement with our earlier findings for skin-derived, immortalized LECs¹⁰, a higher fraction of LN LEC-educated T cells were positive for the apoptosis marker AnnexinV (AnV) compared to those educated by mDCs on day 2, when both central and effector memory subsets are detectable (Fig. 4d). However, when we compared the expression of CD62L per generation among AnV⁺ versus AnV⁻ cells, apoptotic LEC-educated OT-I cells were enriched for the CD62L⁻ phenotype, whereas both apoptotic and non-apoptotic mDC-educated cells shared similar frequencies of CD62L⁻ cells (Fig. 4e). Conversely, the T_eff/EM-like cells were significantly more apoptotic than the T_CM-like cells (Fig. 4f). This observation implies that LEC education induces two divergent populations among CD8⁺ T cells: apoptotic effector-like cells (which are presumably deleted) and surviving predominantly T_CM-like cells.

To further investigate the phenotype of LEC-educated T_CM-like cells in terms of canonical markers of memory versus effector cell differentiation, we measured a variety of memory-related surface markers using flow cytometry (Fig. 4g, Supplementary Fig. 2a, b). LEC-educated cells expressed CD127 (IL-7R), CD122 (IL-2Rβ), and CCR7 at levels similar to or higher than in mDC-educated cells. These markers are notable as CD127 and CD122 functionally mediate IL-7 and IL-15-mediated signaling for the homeostatic proliferation of memory cells, while CCR7 allows homing of T_CM cells to secondary lymphoid organs³¹. Furthermore, the relative expression in LEC-educated T cells of the transcription factors (TFs) Eomes and Tbet (Tbx21), which are mutual antagonists that counter-regulate memory vs. effector differentiation³², was consistent with a T_CM-like phenotype (increasing Eomes and decreasing Tbet) (Fig. 4g, Supplementary Fig. 2b). Interestingly, LEC-educated cells expressed additional markers that may delineate a subset of stem cell-like memory T cells (T_SCM)^33,34,35, including stronger expression of Sca1, B cell lymphoma 2 (Bcl2), CD27, CD95 (Fas), CXCR3, and LFA1 compared to mDC-educated cells.

Further illustrating phenotypic differences between LEC-educated and mDC-educated CD8⁺ T cells (Fig. 4h), most LEC-educated cells were CD44⁺CD62L⁺ (T_CM/SCM-like) with the largest subset expressing high levels of CD127, CD122, and Sca1, whereas mDC-educated cells were mostly CD44⁺CD62L⁻CD127⁺ (T_eff/_EM-like). We further observed a distinct cytokine milieu in LEC- versus mDC-CD8⁺ T cell co-cultures (Supplementary Fig. 2c), with increased G-CSF, eotaxin (CCL11), CXCL1, CXCL10, CCL7, IL-31, CCL2, and CXCL2, and decreased IFNγ, IL-13, IL-17, IL-2, IL-22, TNFα, GM-CSF, and RANKL. In both LEC and mDC co-cultures with OT-I cells, IL-7 and IL-15, which are known to contribute to the induction of T_SCM cells³⁶, were similarly present, but at the mRNA level, LEC expression of these cytokines were increased upon cognate T cell activation (Supplementary Fig. 2d, e).

LEC-educated CD8⁺ T cells similarly acquired memory characteristics regardless of whether we used free or nanoparticle-bound SIINFEKL peptide (NP-ss-COVA_250-264) or full-length ovalbumin (Fig. 3a, Supplementary Figs. 1, 2f). We also examined if TCR affinity may affect memory phenotype acquisition by adapting the co-culture experiments to naïve OT-3 CD8⁺ T cells, which possess TCRs with a lower affinity for the SIINFEKL peptide than OT-I CD8⁺ T cells³⁷. TCR affinity accounted for a reduction in the overall frequency of activated OT-3 cells (~50% when educated by LECs and ~80% by mDCs compared to ~98% of OT-I cells for both education conditions). Nonetheless, OT-3 cells preferentially acquired a memory phenotype when educated by LECs (Supplementary Fig. 2g). Importantly, these same observations also held true in other antigen-transgenic TCR systems, such as with Pmel CD8⁺ T cells co-cultured with gp100 peptide-pulsed LECs (Fig. 4i, j).

Because naïve OT-I mice often harbor a significant population of CD44⁺ CD8⁺ T cells, or so-called ‘virtual memory’ (VM) cells³⁸, we wanted to rule out the possibility that the activated phenotype we saw in LEC-educated OT-I cells in these studies was not due to those cells being re-activated, rather than from priming of naïve cells. To this end, we sorted OT-I splenocytes based on naïve (CD44⁻CD62L⁺) or VM phenotype (CD44⁺) prior to co-culture with LECs or mDCs pre-pulsed with full-length OVA protein (Fig. 4k, l, Supplementary Fig. 3). We found that >80% of LEC-educated naïve OT-I cells acquired a CD44⁺CD62L⁺ phenotype, compared to ~57% for LEC-reactivated ‘VM’ OT-I cells (Fig. 4k, l, Supplementary Fig. 3b). This correlated with increased IL-2 production (Supplementary Fig. 3c), further confirming the T_CM-like phenotype of these cells.

Collectively, these data indicate that LEC-educated CD8⁺ T cells phenotypically show characteristics of both T_CM and T_SCM subsets, independently of the transgenic TCR tested (i.e., CD8⁺ T cells from OT-I, OT-3, and Pmel mice) or the antigen formulation (peptide, full-length protein, or nanoparticle-bound), and are distinct from the T_eff/EM-like phenotype acquired by DC-educated cells under inflammatory conditions.

LEC-primed CD8⁺ T cells express a memory-like transcriptome

To determine whether the phenotypic differences we saw between LEC-educated and mDC-educated T cells are reflected at the transcriptomic level, we performed RNA sequencing to characterize and quantify the transcriptional profiles of naïve, LEC-educated and mDC-educated CD8⁺ T cells after up to 3 days of co-culture. Unsupervised principal component analysis (PCA) (Fig. 5a) and hierarchical clustering (Supplementary Fig. 4a) were performed to visualize the kinetics of T_CM-like and T_EM-like phenotype acquisition by LEC-educated and mDC-educated CD8⁺ T cells, respectively, and to potentially identify sets of genes that were modulated by differences in the APCs across time. The first and second principal components (PC1 and PC2) explained ~52% and 18%, respectively, of the total variance in gene expression profiles. Interestingly, while LEC and mDC-educated cells clustered closely to one another on day 1, both shifted along the PC2 axis over the 3 days co-culture period, with LEC-educated cells diverging on the last day from mDC-educated cells along the PC1 axis toward naïve cells (Fig. 5a). Therefore, we reasoned that PC1 is likely to primarily contain genes that differentiate antigen-experienced from naïve T cells, while PC2 likely accounted for genes associated with memory/effector differentiation.

**Fig. 5: Transcriptome and metabolic signaling in LEC-educated CD8⁺ T cells is consistent with T_CM/T_SCM-like phenotype.**

Focusing solely on the expression of memory/effector signature genes, we saw that LEC-educated cells (Fig. 5b, Supplementary Fig. 4b) displayed higher expression of genes related with the T_CM/T_SCM subsets (Sell, Cd27, Cd28, Ly6e, Ccr7) by d3, whereas mDC-educated cells expressed higher levels of genes consistent with an effector signature (Gzm’s, Prf1, Ifng, Il2ra, Cxcr5, Sema7a), with the exception of Pdcd1 (encoding PD-1), which was highly upregulated in LEC-educated cells as previously reported¹⁰. Notably, Sell (CD62L) expression was initially downregulated in LEC-educated cells on day 1, but then upregulated to levels similar to those in naïve cells, consistent with our observations at the protein level (Fig. 4b). This trend was also observed in LEC-educated cells for other memory-associated genes (Cd27, Cd28, Ly6e, Spn, Il7r, and Ccr7), while mDC-educated cells either failed to upregulate or upregulated them to a lesser extent (Fig. 5b, c, Supplementary Fig. 4b). Finally, consistent with our observations that some LEC-educated cells differentiate into apoptotic effectors, we saw some genes related to effector function (Il2ra, Gzmb, Gzmc, Ifng) upregulated at d1, but expression dropped drastically by d3.

T cell differentiation and functionality are regulated by the tightly tuned expression of several TFs^11,32,39, and over 3 days in co-culture LEC-educated T cells expressed increasing levels of genes encoding TFs favoring long-term survival and proliferative potential (e.g., Id3, Bcl6, Tcf7, Eomes, Elf4, Klf2, and Irf7), and low levels of TFs associated with effector differentiation (e.g., Irf4, Zbtb32, Tbx21, Zeb2, Id2, Rora, Prdm1). In contrast, mDC-educated cells exhibited the opposite trends (Fig. 5d, e, Supplementary Fig. 4c). This divergence appeared most pronounced at d3, when LEC-educated and mDC-educated cells displayed diametrically opposed expression patterns for genes encoding TFs known to act in mutually antagonistic pairs (Eomes-Tbx21, Bcl6-Prdm1, Id3-Id2) to coordinate memory versus effector differentiation, respectively³² (Fig. 5e). Moreover, d3 LEC-educated T cells clustered closer to naïve T cells than to mDC-educated T cells by hierarchical clustering analysis (Fig. 5c), supporting their relatedness to resting cells with the potential for long-term survival.

To identify the unique gene expression signature in LEC-educated cells, we filtered our dataset for genes that were most differentially regulated (mean |log₂FC| ≥ 2.89 and p < 0.01) between day 3 LEC-educated versus mDC-educated cells, arriving at a subset of ~200 genes (Fig. 5f, Supplementary Table 1). Among the genes that mDC-educated cells upregulated the most were those associated with T cell activation and effector potential (Sema7a, Ccr1, Nrp2, Zbtb32, Prdm1, Id2) and effector function (Ifng, Gzma, Gzmd, Gzmg, Gzmf). In contrast, the most differentially expressed genes in LEC-educated cells were correlated with the T_CM-specific signature, including Bcl6, a key signal for memory T cell differentiation⁴⁰; Dapl1, overexpressed in early memory cells⁴¹; and the small GTPase Arl4d, linked with limited early-effector CD8⁺ T cell development upon viral infection⁴² (Fig. 5g). Additionally, LEC-educated OT-I cells also expressed signatures of pharmacologically induced T_SCM cells⁴³ including interferon-responsive genes (Ifi27l2a, Ifi27l2b, Ifit1, Ifit3, Ifit3b), Selp and Isg15 (all important for antiviral innate immunity), Irf7 and Il9r; members of the tripartite motif-containing protein 30 (Trim30) family—which fine-tune the threshold to effector function⁴⁴; as well as Klf2—which can activate and maintain CD62L and CCR7 expressions⁴⁵. Surprisingly, LEC-educated cells also upregulated several neuron-related genes, such as Cnr2, Slfn5, and Nsg2 (Supplementary Fig. 4d), which, together with Sell and Xcl1, comprise a gene expression pattern of naïve and late memory T cells⁴⁶. In addition, a number of genes previously identified as differentially expressed between T_SCM and T_CM cells³³ were also expressed at higher levels in LEC-educated cells (Supplementary Fig. 4e).

LEC priming preferentially engages memory-related pathways

We next made use of genome-wide functional annotations provided by the Gene Ontology Consortium⁴⁷ and mapped genes corresponding to specific biological functions onto volcano plots comparing naïve, LEC-educated or mDC-educated cells (Fig. 5h, i and Supplementary Fig. 4f). Although both LEC/mDC-educated cells upregulated genes related to T cell cytotoxicity relative to naïve OT-I cells, mDC-educated cells upregulated them to a much greater extent, highlighting the limited effector function of LEC-educated cells. On the other hand, LEC-educated cells upregulated genes associated with response to IFNβ and defense response to viral infections (Fig. 5h).

Since metabolism and T cell differentiation state are interrelated processes⁴⁸, we examined the expression of genes related to respiratory chain and fatty-acid metabolic programs, typically associated with resting naïve and memory T cells, and glycolysis, which is dominant in proliferating effector cells (Fig. 5i). Both LEC/mDC-educated cells upregulated genes related to these metabolic programs compared to naïve cells, while glycolysis-associated genes were more upregulated in mDC-educated cells, confirming that LEC-educated cells are less prone to effector differentiation.

Crucially implicated in the regulation of these metabolic processes is the mammalian target of rapamycin (mTOR) pathway, which integrates intracellular and extracellular signals, leading to the formation of either of two multiprotein complexes, mTORC1 or mTORC2, which propagate different downstream signals. Blockade of mTORC1 was shown to favor memory T cell generation⁴⁹ and induce T_SCM differentiation⁴³, while suppression of mTORC2 activity enhanced memory CD8⁺ T cell formation⁵⁰. We detected consistently lower levels of mTOR in LEC-educated compared to mDC-educated cells at all time points examined (Fig. 5j, Supplementary Fig. 4g). Furthermore, while they both expressed similar levels of Rictor, LEC-educated cells displayed significantly lower Rptor expression on d3, suggesting a lower induction of the mTORC1 complex. To further confirm these observations at the protein level, we assessed the phosphorylation of mTOR (pmTOR^S2448) and Akt (pAkt^S473, indicative of mTORC2 activity) by flow cytometry (Fig. 5k, l). Within 2 h of co-culture, mDC education was superior to LEC education in inducing pmTOR^S2448+ and pAkt^S473+ in OT-I cells, suggesting that both mTORC1 and mTORC2 were indeed less active and that PI3K–Akt–mTOR activity is less sustained in LEC-educated cells.

Altogether, these data validate the T_CM/T_SCM-like phenotypic properties of LEC-educated CD8⁺ T cells, which exhibit transcriptional and metabolic programs consistent with a memory-like differentiation state, distinct from mDC-educated cells.

In vitro LEC-primed CD8⁺ T cells have memory-like LN-homing

The elevated expression levels of CD62L in LEC-educated cells prompted us to investigate whether they exhibit preferential migration to secondary lymphoid organs, because CD62L enables naïve and T_CM cells to localize to lymphoid tissue⁵¹. To this end, we transferred naïve or LEC/mDC-educated OT-I cells into mice and analyzed their homing into various organs 1 week later (Fig. 6a). LEC-educated cells homed primarily to secondary lymphoid organs (LN and spleen, 53%) whereas mDC-educated cells migrated mainly to the periphery (lungs and liver, 67%) (Fig. 6b). A smaller fraction of mDC-educated cells was found in lymphoid tissues (33%), to which naïve cells almost exclusively homed (91%), as expected.

**Fig. 6: CD62L expression in LEC-educated CD8⁺ T cells correlates with LN homing after in vivo transfer.**

LEC-educated cells that homed to the LNs and spleen primarily expressed a T_CM-like (CD44⁺CD62L⁺) phenotype (Fig. 6c, d), as well as in the rest of the organs examined (Supplementary Fig. 5a). In contrast, mDC-educated cells were predominantly T_eff/EM-like (CD44⁺CD62L⁻) in all organs except in the LN, where they were mostly T_CM. LEC-educated cells possessed a higher ratio of T_CM/T_eff/EM than mDC-educated cells in all organs examined, with the highest ratio in the LN, while for mDC-educated cells, the ratio T_CM/T_eff/EM was <1 in all organs except in the LN, underlining the phenotypic divergence due to their education (Fig. 6e, Supplementary Fig. 5b, c).

To determine whether LEC-educated cells preferentially localized to specific sub-anatomic sites in the LN, we co-transferred LEC-educated and naïve CD8⁺ T cells, each labeled with different fluorophores, and after a week, imaged thick sections of the brachial LN using confocal microscopy. We observed the naïve cells primarily in the paracortical T cell zone (Fig. 6f), while the LEC-educated cells localized mostly near the cortical and medullary sinuses, around B cell follicles and particularly in the interfollicular areas. These homing patterns are consistent with recent descriptions for naïve and memory CD8⁺ T cells^52,53.

Collectively, these data suggest that CD62L expression in LEC-educated cells is functional and corresponds to preferential migration to secondary lymphoid organs.

Re-stimulation of LEC-educated CD8⁺ T_CM generates cytotoxic lymphocytes

The most important function of memory T cells is their reactivation potential, so we next assessed their response to a secondary challenge. To model this process in vitro, OT-I cells were educated by LECs or mDCs for 3 days, after which they were CFSE-labeled and further co-cultured with mDCs loaded with SIINFEKL (mDC-recall, +Ag) or without antigen (−Ag) (Fig. 7a). After 24 h, we uniquely detected low levels of proliferation upon antigen recall in LEC-educated cells (Fig. 7b, Supplementary Fig. 6a), consistent with the proliferation potential of T_CM cells. More importantly, reactivated LEC-educated cells exited their non-responsive state and produced even higher levels of TNFα and IL-2 compared to mDC-educated cells (Fig. 7c, Supplementary Fig. 6b), and acquired more of an effector-like phenotype with downregulation of CD62L and increased expression of granzyme-B (GzB) (Fig. 7d–f), although not to the same extent as mDC-educated cells.

**Fig. 7: LEC-educated CD8⁺ T cells expand rapidly and differentiate into effector cells upon secondary challenge.**

Because LEC-educated cells responded to an in vitro antigen recall with a strong, multifunctional response, we asked whether this translates to rapid recall responses following secondary challenge under inflammatory conditions in vivo. For this purpose, we transferred CFSE-labeled LEC-educated or mDC-educated OT-I cells into mice, and after 5 days, we administered OVA with LPS (+RECALL) or saline (−RECALL) s.c. (Fig. 7g). Both LEC-educated and mDC-educated CD8⁺ T cells (which may have been induced to divide at different rates before the time of CFSE labeling) expanded significantly within the LNs and spleens of recalled recipients, even to the point of completely diluting CFSE before the endpoint (Fig. 7h, i). Based on cell recovery, compared to non-recalled recipients, recalled LEC-educated T cell numbers increased four-fold in the LN and the spleen, while recalled mDC-educated cells increased 0.5-fold.

This surprising difference may be attributable to the tendency of LEC-educated T cells to be enriched in memory-like cells, which in theory, possess superior expansion potential than the effector-like cells generated by mDC education. Additionally, LEC-educated cells were particularly enriched in the LN compared to the spleen, a result attributable to the early time point following challenge and their homing properties (Fig. 6). Along with the vigorous expansion, and in agreement with our in vitro observations, LEC-educated cells also exhibited the functional potential expected of memory cells by upregulating effector cytokines (Fig. 7j, k). Following ex vivo re-stimulation, we noted increased production of IFNγ and TNFα in antigen-recalled mice, and furthermore, the transferred cells acquired effector function, expressing GzB and demonstrating cytolytic capacity (Fig. 7l), although mDC-educated CD8⁺ T cells were superior by all measures.

Because true memory cells persist even in the absence of antigen, we repeated the above experiment but with re-challenge after >5 weeks of LEC-educated or mDC-educated OT-I cell transfer (Fig. 7g, long-term recall). Both LEC-educated and mDC-educated T cells expanded following recall (Fig. 7m), with LEC-educated cells giving rise to effector-like cells in both the LN and spleen (Fig. 7n, o), though to a less extent than mDC-educated T cells. Similar frequencies of LEC-educated and mDC-educated cells acquired the short-lived effector phenotype (KLRG1⁺CD127⁻), which was predominantly expressed by OT-I cells in the spleen, especially by mDC-educated cells (Fig. 7o). Overall, the phenotypes of LEC-educated T cells resembled those of mDC-educated cells, although LEC-educated cells achieved significantly lower frequencies of T_eff/EM (CD44⁺CD62L⁻) cells (Fig. 7o, Supplementary Fig. 6c). The differences seen in the magnitude of responses in LEC-educated vs. DC-educated T cells upon recall are likely due to differences in their phenotypes and degree of activation at the time of adoptive transfer.

Taken together, these data suggest that LEC-educated CD8⁺ memory-like T cells can acquire cytotoxic function and effector phenotype upon antigen re-encounter. The ability of antigen-experienced T cells to proliferate and differentiate into cytotoxic lymphocytes upon secondary challenge constitutes a hallmark of immunological memory, adapted to long-term immune protection in the host.

LEC-educated CD8⁺ T cell memory protects against infection

These findings raised the question of whether LEC-educated CD8⁺ T cells can mediate protective immunity against a pathogen in a manner comparable to mDC-educated cells. To first measure the cellular response to memory recall, we co-transferred LEC-educated (CD45.1/2⁺) and mDC-educated (CD45.1⁺) OT-I cells (1:1) in mice and after 5 weeks, challenged the mice with OVA-expressing Listeria monocytogenes (L.m.-OVA) (Fig. 8a). Within a week post-infection, LEC-educated cells expanded from making up 0.5 ± 0.4% of all circulating CD8⁺ T cells to 11.2 ± 5.0%, and mDC-educated cells further expanded by at least two-fold (Fig. 8b, c). These trends were not observed in the dLNs (Supplementary Fig. 7a), due to the systemic route of L.m.-OVA infection.

**Fig. 8: LEC-educated CD8⁺ T cells protect against infection and give rise to a persistent secondary memory population.**

Importantly, LEC-educated cells acquired effector function to a similar extent as mDC-educated cells upon ex vivo re-stimulation, exhibiting comparable cytotoxic potential as demonstrated through the frequency of cells producing effector cytokines (IFNγ, TNFα, IL-2) and expressing CD107a, which signifies cytolytic granule release (Fig. 8d, e, Supplementary Fig. 7b). Interestingly, LEC-educated cells trended towards higher polyfunctionality than DC-educated cells, defined as double or triple positive for IFNγ, TNFα, and IL-2 (p = 0.07; Fig. 8f, Supplementary Fig. 7c).

Consistent with our previous observations, LEC-educated cells acquired a predominantly effector phenotype (CD44⁺CD62L⁻) by d8 post-infection (Fig. 8g, h), and most were found within the KLRG1⁺CD127⁻ subset. Remarkably, we also observed a significant T_CM-like (CD44⁺CD62L⁺) subset among LEC-educated cells, which led to a significantly larger T_CM/T_eff/EM ratio relative to mDC-educated cells (Fig. 8h). We additionally noted a significantly higher ratio of memory precursors (KLRG1⁻CD127⁺) to short-lived effectors within the LEC-educated population in both the spleen and LN (Supplementary Fig. 7d).

To directly compare the protective functions of LEC-educated versus mDC-educated cells, we injected mice with either LEC-educated or mDC-educated cells (Fig. 8a, bottom branch). In comparison to mice that received resting CD8⁺ T cells (Educ: Naïve), LEC-educated CD8⁺ T cells and mDC-educated T cells were similarly competent at controlling the bacterial load following infection with L.m.-OVA (Fig. 8i).

We also noted that LEC-educated T cells uniquely retained a population of T_CM-like phenotype following antigen re-encounter (Fig. 7d, n) or pathogenic infection (Fig. 8g, h), so we asked whether LEC-educated cells were preferentially capable of generating a secondary T_CM-like subset. We co-transferred LEC-/mDC-educated OT-I cells (Fig. 8a, top branch) and tracked the transferred cells for 2 weeks post-infection. While the cytotoxic effector fraction (CD44⁺GzB⁺) decreased over time in both groups, only the LEC-educated cells showed a concomitant and significant increase in the T_CM-like phenotype within 9 days post-infection (Fig. 8j, Supplementary Fig. 7e). Furthermore, 3 weeks post-infection, LEC-educated cells possessed a higher ratio of CD127⁺CD62L⁺ to CD127⁺CD62L⁻ cells in the spleen and LN (Fig. 8k). Additionally, we observed similar trends for the T_CM/T_eff/EM and the memory precursor-to-short-lived effector ratios in both organs (Supplementary Fig. 7f). Finally, the frequency of CD44⁺CD62L⁺CD122⁺ cells was significantly higher among LEC-educated cells in both the LN and spleen.

We followed up with an evaluation of LEC-educated T_CM responses to homeostatic signals in vitro (Supplementary Fig. 8a). Similar to conventional T_CM/T_SCM cells³³, LEC-educated cells underwent homeostatic turnover and displayed increased survival in response to IL-15 compared to mDC-educated cells (Supplementary Fig. 8b). Notably, LEC-educated cells retained their initial CD44⁺CD62L⁺ phenotype following IL-15-dependent proliferation (Supplementary Fig. 8c), suggesting self-renewal capacity.

Overall, our findings indicate that LEC-educated CD8⁺ T_CM/T_SCM cells can be reactivated to participate in protective immunity while also possessing the capacity for self-renewal, giving rise to a persistent secondary-memory population.

Requirements for LEC-educated CD8⁺ T_CM polyfunctionality

Finally, we sought to identify the signals that may lead to functional re-activation of LEC-educated CD8⁺ T cells. To evaluate the roles of TCR stimulation and co-stimulation, LEC-educated OT-I cells were treated with αCD3 with or without αCD28 (Fig. 9a). TCR ligation by αCD3 was sufficient to induce IFNγ production; but co-stimulation through αCD28 was required to induce polyfunctional cells (Fig. 9b).

**Fig. 9: Stringent requirements for reactivation of LEC-educated CD8⁺ T cells into functional effector cells.**

To ask if undesirable reactivation of LEC-educated cells could be triggered under homeostatic conditions, we examined their functional profile following reactivation with immature DCs (iDCs) in the presence of immunogenic or immunosuppressive signals. LEC-educated or mDC-educated OT-I cells were cultured together with iDCs (iDC-recall), with or without SIINFEKL, either in the presence of LPS + IFNα + IL-12 or IL-10 + TGFβ (Fig. 9c). LEC-educated cells only became functional effectors when reactivation took place under T_h1-polarizing conditions (LPS + IFNα + IL-12; Fig. 9d), but not in a suppressive setting (IL-10 + TGFβ), as demonstrated by their low production of IFNγ and GzB. Importantly, re-stimulation with iDCs in the absence of the cognate peptide failed to induce significant levels of IFNγ/GzB in LEC-educated cells under any of the conditions described above, while LPS + IFNα + IL-12 stimulation in the absence of cognate peptide was enough to induce both IFNγ and GzB in mDC-educated cells. Our observations suggest that LEC-educated CD8⁺ T_CM cells display antigen specificity and more stringent reactivation requirements compared to conventional DC-educated T_CM cells.

Discussion

This study illuminates a new role for LECs in immunomodulation: the generation of antigen-experienced CD8⁺ T cells, which emerge from antigen encounter as memory-like cells. The ability of non-hematopoietic APCs to drive memory formation has been previously reported⁵⁴. In one example, LSEC-educated T cells exhibit an antigen-experienced phenotype with T_CM-like characteristics¹⁵. Here, we explored whether by analogy it is feasible that LEC education of naïve CD8⁺ T cells can generate CD8⁺ T cells with multiple differentiation states, including those with memory-like functional phenotypes. Because LECs are known to be capable of scavenging and cross-presenting exogenous antigen¹⁰, we hypothesized that they may contribute to the generation of memory T cells.

Education of CD8⁺ T cells by LECs versus by professional APCs leads to their acquisition of vastly different phenotypes, which may be dependent on a variety of signals, including TCR affinity, duration of antigen exposure, cytokines, and the balance between co-stimulatory/inhibitory signals, which are all known to regulate the outcomes of T cell differentiation^55,56. The conditions of LEC education are characterized by low co-stimulation coupled with inhibitory PD-L1/PD-1 signaling^7,10, which are typically observed in transient inflammation and would favor memory rather than terminal effector differentiation. The cytokine milieu in LEC–OT-I co-cultures was consistent with these findings (Supplementary Fig. 2c): we failed to detect IL-12, a key regulator of effector T cell generation that antagonizes memory generation⁵⁷. Notably, we detected IL-7 and IL-15 expressions in LECs (Supplementary Fig. 2d,e), two cytokines known to promote the survival, preferential generation, and homeostatic proliferation of functional T_CM⁵⁸ and T_SCM³⁶ cells.

We further corroborated that LEC-educated CD8⁺ T cells exhibit key functional characteristics of T_CM cells³¹. Upon secondary challenge, LEC-educated cells were capable of producing multiple effector cytokines (Fig. 7c, j), and showcased high proliferative and survival capacity even in response to delayed challenge (Fig. 7i, m). Most importantly, under the correct conditions, LEC-educated T_CM cells gave rise to potent cytotoxic lymphocytes that protected the host from pathogen challenge (Fig. 8b–i). Interestingly, LEC-educated cells also appeared to sustain a persistent secondary T_CM-like population (Fig. 8g, j, k, Supplementary Fig. 7e, f), showcasing their potential for self-renewal (Supplementary Fig. 8). While we did not confirm if LEC-educated cells exhibit multipotency and self-renewal at the level of individual cells⁵⁹ in similar fashion to canonical T_SCM³³ and T_CM⁶⁰ cells, LEC-educated cells appeared to be more differentiated than T_SCM cells based on CD44 expression, while simultaneously being distinct from effector-like mDC-educated cells (Fig. 4a). Therefore, LEC-educated cells may possess an intermediate differentiation state between T_SCM and T_CM cells, distinct from terminal effector cells^61,62,63.

We speculate that the functional significance of LEC-educated memory CD8⁺ T cells is their contribution to diversity of memory T cell states. As much as they have in common to conventional T_CM cells, they differ in important functional properties. Canonical T_CM cells are known to produce IFNγ as a bystander effect during inflammation despite the absence of cognate antigen^38,64. In contrast, like naïve cells, LEC-educated cells required all three signals—cognate antigen, co-stimulation, and T_h1-polarizing cytokines, to become reactivated (Fig. 9)⁶⁵. A similar phenomenon was observed for LSEC cross-primed CD8⁺ T cells¹⁵ and in other non-hematopoietic APCs¹⁴, where functional reactivation of antigen-experienced CD8⁺ T cells required multiple synergistic signals. Such prerequisites may safeguard against undesired reactions and hinder the onset of autoimmunity, given that LECs would normally collect endogenous antigens under homeostatic conditions. These key differences may also highlight the need for CD4⁺ T cell help, in the form of co-stimulation or accessory cytokines, to generate canonical T_CM cells, although LECs express low levels of co-stimulatory molecules, such as CD80 and CD86 and thereby are unlikely to stimulate optimal CD4⁺ T cell responses^3,6.

These more stringent reactivation requirements may be mistaken for tolerance induction by LECs, as reported in tumors and tumor-dLNs⁹. Based on our findings, efficient reactivation of LEC-educated cells and subsequent cytotoxic lymphocyte generation would be quite unlikely in such suppressive microenvironments due to the strong immunoregulatory factors present. After all, the tumor-draining microenvironment rather resembles reactivation in the presence of IL-10 + TGFβ, which completely failed to induce functional GzB-expressing effector T cells among LEC-educated cells (Fig. 9d). However, in the setting of chronic inflammation rich in inflammatory cytokines and danger signals, re-stimulation of LEC-educated cells would likely favor generation of cytotoxic effector T cells. Indeed, while controversial, lymphangiogenesis has been frequently associated with aggravation of inflammatory diseases and transplant rejection^66,67.

We envision that LEC-educated T_CM-like CD8⁺ cells are physiologically significant in at least two scenarios. First, they may be very important in first-line containment against systemic antigen spread. Both viral and bacterial pathogens employ different strategies to circumvent innate immunity, which results in serious pathology⁶⁸. Therefore, LEC priming may spare pathogen-specific T cells from elimination due to first non-immunogenic encounter, generating antigen-experienced cells that may be activated into functional effectors under the correct conditions. A similar role was recently suggested for LSEC-primed CD8⁺ T cells¹⁵. Second, LEC priming may initiate the generation of antigen-experienced CD8⁺ T_CM-like cells following infection, because freely draining antigens arrive in the LN much earlier than antigen-loaded migratory DCs from peripheral tissues¹. Consistent with this, in the first moments of a viral infection, virions were reported to drain to the LN and infect resident cells^54,69 at the same sub-anatomic locations where LECs are reportedly capable of antigen archiving⁷⁰. Given their constant contact with the antigen-rich pool of the lymph⁷¹, LECs can generate a population of antigen-experienced T cells, which may be readily expanded by activated professional APCs. In line with our hypothesis, stromal cells were able to induce naïve T cell recruitment and activation early during infection, even though their sustained expansion at later time points was markedly limited⁷². Of course, different pathogens engage different immunostimulatory pathways and produce different response kinetics, which directly affect the balance of antigen presentation by professional APCs versus by LECs, possibly explaining seemingly contradictory data with regard to LEC-induced T cell proliferation⁷⁰.

In conclusion, our findings demonstrate a new immunomodulatory role for LECs. Exogenous antigen cross-presentation by LECs induces an expandable pool of memory-like T cells, leading to effector T cell generation under inflammatory conditions to fight an ongoing infection, while preserving a subset of long-lived cells to combat future pathogen encounter. Our work reveals the multifaceted effects of antigen education by LECs, highlighting their dynamic roles in response to the local microenvironment for the regulation of immunity. These findings also help explain and reconcile the seemingly contradictory effects of lymphangiogenesis in tumors versus chronic inflammatory diseases or transplants^66,67. As we begin to better understand the impact of antigen presentation by LECs in immunomodulation and unravel the various signals implicated, a new challenge emerges: how can we harness this function to improve protection against pathogens and tumors, or to prevent autoimmunity and transplant rejection?

Methods

Reagents

Chemicals were purchased from Sigma-Aldrich (Buchs, CH) unless otherwise noted. The mature MHCI epitope, OVA_256-264 (SIINFEKL) peptide, was from GenScript (Piscataway, NJ, USA) and the gp100_25–33 peptide was from Anaspec (Fremont, CA, USA). The Endotoxin-free OVA was from Hyglos GmbH, (Bernried am Starnberger See, DE). Ultra-pure lipopolysaccharide (LPS, 01:11 B4) was from Invivogen (InvivoGen, San Diego, CA, USA). CpG-B 1826 oligonucleotide (CpG) was obtained from Microsynth (Balgach, CH). Recombinant murine GM-CSF and IL-15 were purchased from Peprotech (Oak Park, CA, USA), IL-10 from eBioscience (Vienna, Austria), IL-12, IFNα and recombinant human TGFβ1 from R&D systems (Minneapolis, MN, USA). Antibodies used in flow cytometry were from BD (Franklin Lakes, NJ, USA), eBioscience (Waltham, MA, USA), or BioLegend (San Diego, CA, USA) unless otherwise noted.

Mice

The following mice strains were used in this study at age 6–12 weeks unless noted otherwise. C57BL/6 (CD45.2 or CD45.1 Ly5) wild-type mice and OT-I CD45.2 transgenic mice, C57BL/6-Tg(TcraTcrb)1100Mjb/J, were purchased from Harlan Laboratories (Gannat, FR). Transgenic OT-I CD45.1/2 mice and OT-3 mice were obtained from Dietmar Zehn. Pmel transgenic mice (B6.Cg-Thy1a/Cy Tg(TcraTcrb)8Rest/J), were a gift from Pedro Romero (Ludwig Institute for Cancer Research, Université de Lausanne). β₂m^−/− mice (lacking MHC class I) were a gift from Hugh Robson MacDonald. Constitutively OVA-expressing mice (C57BL/6-Tg(CAG-OVA)916Jen/J) were a gift from Anita Chong. Animals were housed in pathogen-free facilities. All procedures and studies involving mice were approved by the Cantonal Veterinary Committee of Vaud, Switzerland (VD2518, VD2981, and VD2992) or the University of Chicago IACUC (protocol # 72414, 72578).

Primary cell isolation

To obtain primary LN LECs and LN FRCs, LNs (axillary, cervical, brachial, inguinal, and popliteal) were digested with 1.3 Wünsch Units/mL Liberase DH and 200 Kunitz/mL DNase I (Roche, Basel, CH) to obtain a single cell suspension and cultured as described⁷³. Cells were cultured for 5 days until confluent, detached by Accutase (Biological Industries, Lucerna-Chem AG, Lucerne CH), and stained with DAPI or PI to exclude dead cells, as well as mAbs against gp38 (clone 8.1.1), CD31 (clone 390), and CD45 (clone 30-F10) and FACS sorted (FACS Aria II, BD, Franklin Lakes, NJ, USA) into the following subpopulations: FRCs (CD45⁻gp38⁺CD31⁻) LECs (CD45⁻gp38⁺CD31⁺), BECs (CD45⁻gp38⁻CD31⁺), and DN (CD45⁻gp38⁻CD31⁻) as described (Supplementary Fig. 9b)⁷⁴. BMDCs were harvested from C57Bl/6 mice, differentiated in GM-CSF as described⁷⁵, and used at day 7 of culture. When indicated, BMDCs were matured (mDCs) by culturing in the presence of antigen (see section below on generation of ex vivo educated CD8⁺ T cells) plus LPS (10 ng/mL) or CpG-B (0.1 µM) for another 12 h.

For some experiments (Fig. 2), dLNs of i.d. vaccination sites (axillary, brachial, inguinal, and popliteal) were digested in digestion medium (DMEM containing 2% FBS and 2 mM CaCl₂) supplemented with 1 mg/mL collagenase IV and 40 µg/mL DNase I for 30 min, followed by subsequent digestion in digestion medium supplemented with 3.5 mg/mL collagenase D and 40 µg/mL for another 30 min as described elsewhere⁷⁶. The resulting cell suspensions were filtered through 70 µm cell strainers, treated with CD16/CD32 antibody (clone 93) to block nonspecific antibody binding to FcγRs, and then subsequently stained using DAPI, CD45, gp38, and CD31 as indicated above, as well as F4/80 (clone CI:A3-1) and CD11c (clone N418). In addition to LECs and FRCs as indicated earlier, DCs (CD45⁺CD11c⁺F4/80⁻) and macrophages (Mφ; CD45⁺CD11c⁻F4/80⁺) were also collected.

Preparation and implantation of LEC spheroids

Primary murine LECs were expanded from WT C57Bl/6 or constitutively OVA-expressing mice, and purified via FACS as described above. After sorting, LECs were rested and further expanded on collagen-precoated tissue culture-treated polystyrene Petri dishes, in growth medium (αMEM supplemented with 10% FBS, 100 U/mL penicillin, and 100 μg/mL streptomycin). LECs were then processed to form spheroids in hanging drop culture as described by Korff⁷⁷. In brief, LECs were detached from plates using Accutase, and then resuspended at 1.0 × 10⁵ cells/mL in growth medium. Sterile methylcellulose (3 mg/mL in αMEM) was mixed into the cell suspension to a final concentration of 0.6 mg/mL. Hanging drops of 20 μL/droplets were seeded onto the lids of Petri dishes containing PBS, and cultured for at least 3 days until spheroids were visible. Spheroids were collected in 50 mL conical tubes with serum-free IMDM and centrifuged at 20 × g for 5 min at 4 °C, and washed twice to remove excess serum and antibiotics.

Recipient mice were anesthetized under isoflurane (4% induction, 1.5% maintenance) and then laid on their sides. A single 5 mm incision was made on the ventral side of the ear, nearly parallel to the base of the ear, cutting through the cartilage area but without damaging the dorsal dermis. Using a spatula (~4 mm wide by ~20 mm long), an implantation pocket was gently created between the cartilage and the dorsal dermis. Bovine thrombin (~0.01 U in < 5 μL) was applied to the implantation pocket. At this point, spheroids were mixed with bovine fibrinogen (final concentration of 10 mg/mL in IMDM) and delivered intradermally (i.d.) by pipet into the pocket. After allowing 5 min for gelation, the incision was sealed by applying Histoacryl surgical glue (B. Braun, Melsungen, Germany).

Synthesis of peptide-conjugated nanoparticles (NPs)

To explore the mechanisms of cross-presentation poly(propylene sulfide)-NPs with ~30 nm diameter were synthesized and characterized as described²⁹. The long peptide containing the mature MHCI epitope SIINFEKL-Cys-OVA_250-264 (COVA_250–264) was synthesized in-house and activated with a 2-pyridylthiol as previously described²⁹. Core sulfhydryl groups on NPs were reacted with the activated peptide and purified on a Sepharose CL6B column (Sigma-Aldrich). Endotoxin levels of antigens were routinely assessed by a colorimetric assay based on the HEK-Blue™ TLR4 cell line (InvivoGen, San Diego, CA, USA) according to the manufacturer’s protocol using a standard curve generated from the E-Toxate™ endotoxin standard (Sigma).

Generation of ex vivo LN LEC/mDC-educated CD8+ T cells

CD8⁺ T cells from healthy, adult OT-I, OT-3, mice, or Pmel mice were negatively selected from total splenocytes using the EasySep Mouse CD8⁺ T Cell Isolation Kit from Stemcell Technologies (Vancouver, BC, Canada). In some experiments, the resulting CD8⁺ T cells were further purified by FACS based on naïve (CD44⁻CD62L⁺) or VM (CD44⁺) phenotypes. Subsequently, CD8⁺ T-cells were directly co-cultured with LN LECs or mDCs in the presence of either 1 nM NPssCOVA_250-264, 1 µM full-length OVA protein, 1 nM SIINFEKL peptide, or 1 μM of gp100 peptide as indicated (human gp100_25-33; KVPRNQDWL, which differs from the murine sequence by only two amino acids and is functional in reports using murine cells⁷⁸). In some experiments, LN LECs or mDCs were pulsed with antigen, then washed with PBS prior to the addition of T cells. More specifically, 10⁴ LECs or DCs were co-cultured with naïve CD8⁺ T cells in 96-well plates for 72 h for OT-I and for 6 days for Pmel CD8⁺ T cells at a ratio of 1:10 APC:T cells in 200 μL of co-culture media (IMDM with 10% FBS 100 U/mL penicillin, and 100 μg/mL streptomycin—all from Life Technologies, Carlsbad, CA, USA). Supernatants were harvested and frozen for cytokine analysis. Cells were then processed and stained for immunological markers to be analyzed by flow cytometry (Supplementary Fig. 9c). Cellular proliferation was monitored by CFSE dilution and apoptosis was determined by Annexin V staining (BioVision, Milpitas, CA, USA). T cell proliferation was determined by assessing CFSE intensity using the proliferation algorithm in FlowJo 9.4.11.

In vitro reactivation

LEC/mDC-educated OT-I CD8⁺ T cells were harvested on day 3 of co-culture, washed at least twice with basal medium and counted. 10⁵ LEC-educated or mDC-educated CD8⁺ T cells were subsequently co-cultured with fresh 10⁴ mDCs loaded with or without the SIINFEKL peptide (1 nM) for 24 h in 200 μL of co-culture media (mDC-recall). Supernatants were harvested and frozen for cytokine analysis. Cells were then processed and stained for immunological markers to be analyzed by flow cytometry. The levels of intracellular cytokines were determined after 3 h reactivation and 2 h of Brefeldin A (BfA) treatment. When indicated, LEC-educated or mDC-educated CD8⁺ T cells were reactivated by non-LPS-treated DCs (immature DCs, iDCs) loaded with or without the SIINFEKL peptide (1 nM) for 24 h in the presence or absence of the following cytokines: rmIL-12 (5 ng/mL), rmIFNα (2 kU/mL), rmIL-10 (60 U/mL), rhTGFβ1 (500 U/mL). In some experiments, reactivation of LEC-educated or mDC-educated CD8⁺ T cells was also conducted with plate-bound anti-CD3 antibody (145.2C11) at 5 μg/mL and/or soluble anti-CD28 (37.51) at 2 μg/mL (aCD3/CD28, eBioscience). The same procedures, described above, were performed to analyze the phenotype and function of the reactivated cells.

Antigenic challenge

To determine whether exogenous antigen uptake by LNSCs can drive CD8⁺ T cell proliferation in vivo, we i.d. administered 50 μg of endotoxin-free OVA in all four limbs at a volume of 10 μL/limb, or saline as a control. For one study (Fig. 2), a total dose of 100 µg of endotoxin-free OVA was administered i.d. in six locations, including all four footpads (10 µL per limb) and two locations along the spinal axis of the mouse in the dorsal skin (30 µL per spot). To test the capacity of LN LEC/mDC-educated T cells to respond to secondary antigen encounter in vivo, we subcutaneously administered 50 μg of endotoxin-free OVA plus 100 μg of ultra-pure LPS.

Infection with OVA-expressing L.m

To assess the functional potential of LN LEC/mDC-educated CD8⁺ T cell upon pathogenic challenge, mice were infected i.v. with L.m.-OVA⁷⁹ (10³ colony forming units, cfu) acquired from log phase of growth in brain heart infusion (BHI) (Sigma-Aldrich) medium. The mice were challenged at least 5 weeks following adoptive transfer of LN LEC/mDC-educated CD8⁺ T cells. The mice were sacrificed either 8 days after challenge to analyze T cell responses or, 3 weeks after challenge to assess the formation of secondary memory. For the analysis of bacterial load in the spleen, mice were infected with 10⁴ cfu L.m.-OVA at least 9 weeks following adoptive transfer of LN LEC/mDC-educated CD8⁺ T cells and sacrificed 3 days after. Spleens were collected, homogenized, and resuspended in 10 mL sterile PBS. Cell suspensions were then plated on BHI agar plates and incubated overnight at 37 ^oC. The following day, colony-forming units were counted and the amounts of L.m.-OVA were calculated with respect to the relative dilutions.

Ex vivo killing assay

To assess the specific killing capacity of reactivated LEC-educated and mDC-educated CD8⁺ T cells, the cells were cultured, following in vivo antigen re-encounter, with a 1:1 mixture of CFSE^high-labeled (5 μΜ) SIINFEKL-loaded and CFSE^low-labeled (0.1 μΜ) mismatched-peptide-loaded DCs (mixed target cells) at different effector over target cell ratios. We kept the number of mixed target cells constant at 5 × 10⁴ cells. Mixed target cells alone served as a control for non-specific death. After 12–16 h, the ratio of the surviving CFSE^high and CFSE^low cell populations was assessed by flow cytometry to calculate the percentage of specific killing as follows: % specific killing: 100−[100*(CFSE^high/CFSE^low)_sample/(CFSE^high/CFSE^low)_control].

Flow cytometry

Cells were washed and stained with a cocktail of surface antibodies in staining buffer, containing HBSS (Life Technologies) supplemented with 0.5% bovine serum albumin. Cell viability was determined by propidium iodide incorporation in staining buffer after surface antibody staining or with live/dead fixable cell viability reagent (Life Technologies) in PBS before surface antibody staining. For blocking non-specific binding of immunoglobulin to Fc receptors, cells were also pre-incubated with purified anti-CD16/CD32 antibody (clone 93) for 10 min on ice prior to immunostaining. For intracellular/intranuclear staining, cells were fixed and permeabilized with the Foxp3/Transcription Factor Fixation/Permeabilization kit (eBioscience) according to the manufacturer’s instructions. Cells were stained in permeabilization buffer with a cocktail of monoclonal antibodies. For phospho-specific staining, cells were fixed with IC fixation buffer (eBioscience) and subsequently incubated with ice-cold 100% methanol for membrane permeabilization. Cells were stained with a cocktail of surface and phospho-flow antibodies in staining buffer. The following anti-mouse antibodies were used: CD62L (MEL-14), CD44 (IM7), KLRG1 (2F1/KLRG1), PD-1 (RMP1-30), CXCR3 (CXCR3-173), CD43 (1B11), CD27 (LG.3A10), CD122 (TM-b1), CCR7 (4B12), CD127 (A7R34), T-bet (eBio4B10), Eomes (Dan11mag), LFA-1 (H155-78), Sca-1 (D7), Bcl-2 (BCL/10C4), CD45.1 (F20), CD45.2 (104), CD8 (53-6.7), CD3e (145-2C11), IFNγ (XMG1.2), IL-2 (JES6-5H4), TNFα (MP6-XT22), CD107 (1D4B), Granzyme-B (NGZB), ki-67 (SolA15), pS6^S235/236 (D57.2.2E), pmTOR^S2448 (O21-404), pAkt^S473 (D9E). Phospho-flow antibodies were from Cell Signaling Technology (Danvers, MA, USA). Finally, cells were resuspended in staining buffer and analyzed by flow cytometry (CyAn ADP Flow Cytometer, DAKO or BD LSRFortessa 5-laser SORP). Data analysis was performed using FlowJo (v9.4, Tree Star Inc., Ashland, OR, USA). Identification of OT-I cells from flow cytometry plots generally followed the gating strategies in Supplementary. Fig. 9a, c.

Quantitative real-time RT–PCR

Total RNA was isolated with the RNeasy mini plus Kit (QIAGEN, Hombrechtikon, CH) and reverse transcription was carried out with the iScript cDNA Synthesis Kit (Bio-Rad Laboratories AG, Cressier, CH) according to the manufacturer’s instructions. Quantitative RT-PCR was performed using SYBR Green mix and a LightCycler^® 96 instrument (both Roche). Primer sequences were as follows: IL-7, 5′-TCCTCCACTGATCCTTGTTC-3′ and 5′-CTTCAACTTGCGAGCAGCAC-3′; IL-15, 5′-GTGACTTTCATCCCAGTTGC-3′ and 5′-TTCCTTGCAGCCAGATTCTG-3′. For isolation from cells in culture, cells were pelleted and directly used. For RNA isolation from control and inflamed skin-dLNs of C57Bl/6 mice, draining LN were isolated from mice 5 days after i.d. injection with either PBS or 15 μg of CpG-B in all four limbs at a volume of 10 μL/limb.

RNA sequencing

RNA sequencing was conducted with naïve or LEC/mDC-educated CD8⁺ T cells, coming from 3 OT-I mice, at three different time points during co-culture (day 1, 2, 3), making for a total of 21 samples across seven different conditions. Total RNA was extracted with the RNeasy mini Plus Kit (Qiagen, Hombrechtikon, CH). The Agilent RNA ScreenTape assay was performed in combination with the 4200 TapeStation system (Agilent Technologies) for the determination of quality and quantity of the RNA samples. Additionally, concentration of the RNA samples was determined with Qubit and NanoDrop (both ThermoFisher Scientific, Waltham, MA, USA). Sequencing libraries were generated with a TruSeq Stranded mRNA Prep Kit (Illumina, San Diego, USA) according to manufacturer’s protocols. The libraries were quantified by Qubit DNA HS and profile analysis was performed by Fragment Analyzer (Advanced Analytical). Seven samples were then pooled for each sequencing run (Illumina NextSeq 500) with 400 million reads (nominally ~55 million reads/sample) of 75 bps in single-end mode.

Transcriptional profile analysis

Raw-sequencing files were first checked for sequence quality and potential contamination using the FastQC program (Babraham Bioinformatics, Cambridge, United Kingdom). From here, read mapping and read counting was performed on the HTSStation platform⁸⁰, which is an integrated modular platform using the Bowtie2 algorithm to map reads to a reference mouse genome⁸¹ and the (Bioconductor) Limma.R package⁸² to perform differential expression statistical analyses⁸³ and generation of adjusted p-values values via one-way ANOVA, adjusting for the false-discovery rate using the Benjamini–Hochberg method.

To visualize global changes in gene expression as a function of T cell education and time, we applied PCA on all 45,707 detected genes, and hierarchical clustering on a pre-filtered subset of these genes with a median rpkm >5, and with adjusted p < 0.01 versus expression on naïve OT-I cells. This subset of 16,752 genes essentially represented substantially expressed genes that were differentially regulated in naïve or in antigen-experienced CD8⁺ T cells. The rpkms of these 16,752 genes were submitted to the GENE-E platform (https://software.broadinstitute.org/GENE-E/index.html) for hierarchical clustering, using Pearson correlation coefficients among the genes to potentially identify modules of genes that appear co-regulated or induced uniquely due to LEC- or mDC education.

To compare data expression in LEC-educated cells with memory/effector cell gene signatures, the genes with the greatest differences in expression (all of those with an adjusted p value of <0.05) and a change in expression of >1.5-fold between day 3 LEC/mDC-educated cells) were selected. To detect genes most differentially regulated in LEC-educated cells on day 3, we selected genes that fulfilled the following conditions: (1) an adjusted p < 0.01 between any pair of comparisons among naïve OT-I cells and LEC-educated or mDC-educated OT-I on day 3, and (2) |log₂FC| ≥ 2.89-fold difference in expression between LEC-educated versus mDC-educated OT-I cells. These criteria led to a final panel of ~200 genes, which represent the genes that are differentially regulated in antigen-experienced T cells, and furthermore, most differentially modulated by the nature of the APC used (Supplementary Table 1). These ~200 genes were mapped onto biological functions and pathways using Gene Ontology Enrichment Analysis (Gene Ontology Consortium—http://geneontology.org/). Furthermore, we used the Gene Ontology Atlas⁴⁷ to build lists of genes related to specific biological processes, and used R and Excel to map these genes on the volcano plots to better visualize gene expression patterns between naïve OT-I cells, or LEC-educated or mDC-educated OT-I cells.

Statistical analysis

Statistical analysis was performed using one-way or two-way analysis of variance (ANOVA) or unpaired, two-tailed Student’s t-test followed by Bonferroni post-test with Prism 5 software (Graphpad, San Diego, CA, USA) unless otherwise stated. If the groups to be compared had significantly different variances (p < 0.05 by F test) then Welch’s correction was performed. In all the animal studies, mice were randomly assigned to the different groups and treated in a random order. The investigators were not blinded to group allocation during the experiment or when assessing the outcome. The number of animals chosen per experiment was based on previous publications using similar models and addressing similar questions, as well as on previous studies performed in our lab when statistical significance was actually achieved. To account for data reproducibility, the experiments were performed at least twice. For in vitro experiments, samples were randomly assigned to positions in multi-well plates and the processing order was random. The investigators were not blinded to sample identity. Sample sizes were chosen as the minimum number of mice to sort enough cells to perform three to four technical replicates per group per assay. Results are shown as mean ± standard deviation, unless noted otherwise, with significance indicated as *p ≤ 0.05, **p ≤ 0.01, and ***p ≤ 0.001. When Box-and-Whiskers plots are used, the line in the middle of the box indicates the median and the whiskers go down to the smallest and up to the largest value.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Data availability

The RNA sequencing data have been deposited to the NIH Gene Expression Omnibus under accession number GSE88830. Source data are provided as a Source Data file. All other data are included in the supplemental information or available from the authors upon reasonable request.

References

Hirosue, S. & Dubrot, J. Modes of antigen presentation by lymph node stromal cells and their immunological implications. Front. Immunol. 6, 446 (2015).
Article PubMed PubMed Central CAS Google Scholar
Malhotra, D., Fletcher, A. L. & Turley, S. J. Stromal and hematopoietic cells in secondary lymphoid organs: partners in immunity. Immunol. Rev. 251, 160–176 (2013).
Article PubMed PubMed Central CAS Google Scholar
Card, C. M., Yu, S. S. & Swartz, M. A. Emerging roles of lymphatic endothelium in regulating adaptive immunity. J. Clin. Invest. 124, 943–952 (2014).
Article CAS PubMed PubMed Central Google Scholar
Turley, S. J., Fletcher, A. L. & Elpek, K. G. The stromal and haematopoietic antigen-presenting cells that reside in secondary lymphoid organs. Nat. Rev. Immunol. 10, 813–825 (2010).
Article CAS PubMed Google Scholar
Podgrabinska, S. & Skobe, M. Role of lymphatic vasculature in regional and distant metastases. Microvasc. Res. 95, 46–52 (2014).
Article PubMed PubMed Central Google Scholar
Dubrot, J. et al. Lymph node stromal cells acquire peptide-MHCII complexes from dendritic cells and induce antigen-specific CD4⁺ T cell tolerance. J. Exp. Med. 211, 1153–1166 (2014).
Article CAS PubMed PubMed Central Google Scholar
Tewalt, E. F. et al. Lymphatic endothelial cells induce tolerance via PD-L1 and lack of costimulation leading to high-level PD-1 expression on CD8 T cells. Blood 120, 4772–4782 (2012).
Article CAS PubMed PubMed Central Google Scholar
Cohen, J. N. et al. Lymph node-resident lymphatic endothelial cells mediate peripheral tolerance via Aire-independent direct antigen presentation. J. Exp. Med. 207, 681–688 (2010).
Article CAS PubMed PubMed Central Google Scholar
Lund, A. W. et al. VEGF-C promotes immune tolerance in B16 melanomas and cross-presentation of tumor antigen by lymph node lymphatics. Cell Rep. 1, 191–199 (2012).
Article CAS PubMed Google Scholar
Hirosue, S. et al. Steady-state antigen scavenging, cross-presentation, and CD8+ T cell priming: a new role for lymphatic endothelial cells. J. Immunol. 192, 5002–5011 (2014).
Article CAS PubMed PubMed Central Google Scholar
Mueller, D. L. Mechanisms maintaining peripheral tolerance. Nat. Immunol. 11, 21–27 (2010).
Article CAS PubMed Google Scholar
Redmond, W. L. & Sherman, L. A. Peripheral tolerance of CD8 T lymphocytes. Immunity 22, 275–284 (2005).
Article CAS PubMed Google Scholar
Schietinger, A., Delrow, J. J., Basom, R. S., Blattman, J. N. & Greenberg, P. D. Rescued tolerant CD8 T cells are preprogrammed to reestablish the tolerant state. Science 335, 723–727 (2012).
Article ADS CAS PubMed PubMed Central Google Scholar
Vezys, V., Olson, S. & Lefrançois, L. Expression of intestine-specific antigen reveals novel pathways of CD8 T cell tolerance induction. Immunity 12, 505–514 (2000).
Article CAS PubMed Google Scholar
Böttcher, J. P. et al. Liver-primed memory T cells generated under noninflammatory conditions provide anti-infectious immunity. Cell Rep. 3, 779–795 (2013).
Article PubMed CAS Google Scholar
Ahlers, J. D. & Belyakov, I. M. Memories that last forever: strategies for optimizing vaccine T-cell memory. Blood 115, 1678–1689 (2010).
Article CAS PubMed PubMed Central Google Scholar
Kurtulus, S., Tripathi, P. & Hildeman, D. A. Protecting and rescuing the effectors: roles of differentiation and survival in the control of memory T cell development. Front. Immunol. 3, 404 (2012).
PubMed Google Scholar
Harty, J. T. & Badovinac, V. P. Shaping and reshaping CD8+ T-cell memory. Nat. Rev. Immunol. 8, 107–119 (2008).
Article CAS PubMed Google Scholar
Joshi, N. S. & Kaech, S. M. Effector CD8 T cell development: a balancing act between memory cell potential and terminal differentiation. J. Immunol. 180, 1309–1315 (2008).
Article CAS PubMed Google Scholar
Fletcher, A. L. et al. Lymph node fibroblastic reticular cells directly present peripheral tissue antigen under steady-state and inflammatory conditions. J. Exp. Med. 207, 689–697 (2010).
Article CAS PubMed PubMed Central Google Scholar
Merad, M. et al. Langerhans cells renew in the skin throughout life under steady-state conditions. Nat. Immunol. 3, 1135–1141 (2002).
Article CAS PubMed PubMed Central Google Scholar
Stoitzner, P. et al. Langerhans cells cross-present antigen derived from skin. Proc. Natl Acad. Sci. USA 103, 7783–7788 (2006).
Article ADS CAS PubMed PubMed Central Google Scholar
Flacher, V. et al. Murine Langerin+ dermal dendritic cells prime CD8+ T cells while Langerhans cells induce cross-tolerance. EMBO Mol. Med. 6, 1191–1204 (2014).
Article CAS PubMed PubMed Central Google Scholar
Roozendaal, R. et al. Conduits mediate transport of low-molecular-weight antigen to lymph node follicles. Immunity 30, 264–276 (2009).
Article CAS PubMed PubMed Central Google Scholar
Gerner, M. Y., Torabi-Parizi, P. & Germain, R. N. Strategically localized dendritic cells promote rapid T cell responses to lymph-borne particulate antigens. Immunity 42, 172–185 (2015).
Article CAS PubMed Google Scholar
Itano, A. A. et al. Distinct dendritic cell populations sequentially present antigen to CD4 T cells and stimulate different aspects of cell-mediated immunity. Immunity 19, 47–57 (2003).
Article CAS PubMed Google Scholar
Onder, L. et al. IL-7-producing stromal cells are critical for lymph node remodeling. Blood 120, 4675–4683 (2012).
Article CAS PubMed PubMed Central Google Scholar
Miller, C. N. et al. IL-7 production in murine lymphatic endothelial cells and induction in the setting of peripheral lymphopenia. Int. Immunol. 25, 471–483 (2013).
Article CAS PubMed PubMed Central Google Scholar
Hirosue, S., Kourtis, I. C., van der Vlies, A. J., Hubbell, J. A. & Swartz, M. A. Antigen delivery to dendritic cells by poly(propylene sulfide) nanoparticles with disulfide conjugated peptides: cross-presentation and T cell activation. Vaccine 28, 7897–7906 (2010).
Article CAS PubMed Google Scholar
Obar, J. J. & Lefrançois, L. Memory CD8+ T cell differentiation. Ann. N. Y. Acad. Sci. 1183, 251–266 (2010).
Article ADS CAS PubMed PubMed Central Google Scholar
Wherry, E. J. et al. Lineage relationship and protective immunity of memory CD8 T cell subsets. Nat. Immunol. 4, 225–234 (2003).
Article CAS PubMed Google Scholar
Kaech, S. M. & Cui, W. Transcriptional control of effector and memory CD8+ T cell differentiation. Nat. Rev. Immunol. 12, 749–761 (2012).
Article CAS PubMed PubMed Central Google Scholar
Gattinoni, L. et al. A human memory T cell subset with stem cell-like properties. Nat. Med. 17, 1290–1297 (2011).
Article CAS PubMed PubMed Central Google Scholar
Zhang, Y., Joe, G., Hexner, E., Zhu, J. & Emerson, S. G. Host-reactive CD8+ memory stem cells in graft-versus-host disease. Nat. Med. 11, 1299–1305 (2005).
Article CAS PubMed Google Scholar
Farber, D. L., Yudanin, N. A. & Restifo, N. P. Human memory T cells: generation, compartmentalization and homeostasis. Nat. Rev. Immunol. 14, 24–35 (2014).
Article CAS PubMed Google Scholar
Cieri, N. et al. IL-7 and IL-15 instruct the generation of human memory stem T cells from naive precursors. Blood 121, 573–584 (2013).
Article CAS PubMed Google Scholar
Enouz, S., Carrié, L., Merkler, D., Bevan, M. J. & Zehn, D. Autoreactive T cells bypass negative selection and respond to self-antigen stimulation during infection. J. Exp. Med. 209, 1769–1779 (2012).
Article CAS PubMed PubMed Central Google Scholar
Haluszczak, C. et al. The antigen-specific CD8+ T cell repertoire in unimmunized mice includes memory phenotype cells bearing markers of homeostatic expansion. J. Exp. Med. 206, 435–448 (2009).
Article CAS PubMed PubMed Central Google Scholar
Böttcher, J. & Knolle, P. A. Global transcriptional characterization of CD8+ T cell memory. Semin. Immunol. 27, 4–9 (2015).
Article PubMed CAS Google Scholar
Crotty, S., Johnston, R. J. & Schoenberger, S. P. Effectors and memories: Bcl-6 and Blimp-1 in T and B lymphocyte differentiation. Nat. Immunol. 11, 114–120 (2010).
Article CAS PubMed PubMed Central Google Scholar
Wirth, T. C. et al. Repetitive antigen stimulation induces stepwise transcriptome diversification but preserves a core signature of memory CD8(+) T cell differentiation. Immunity 33, 128–140 (2010).
Article CAS PubMed PubMed Central Google Scholar
Tolksdorf, F. et al. The PDL1-inducible GTPase Arl4d controls T effector function by limiting IL-2 production. Sci. Rep. 8, 16123 (2018).
Article ADS PubMed PubMed Central CAS Google Scholar
Scholz, G. et al. Modulation of mTOR signalling triggers the formation of stem cell-like memory T cells. EBioMedicine 4, 50–61 (2016).
Article PubMed PubMed Central Google Scholar
Choi, U. Y. et al. Tripartite motif-containing protein 30 modulates TCR-activated proliferation and effector functions in CD4+ T cells. PLoS One 9, e95805 (2014).
Article ADS PubMed PubMed Central CAS Google Scholar
Bai, A., Hu, H., Yeung, M. & Chen, J. Kruppel-like factor 2 controls T cell trafficking by activating L-selectin (CD62L) and sphingosine-1-phosphate receptor 1 transcription. J. Immunol. 178, 7632–7639 (2007).
Article CAS PubMed Google Scholar
Best, J. A. et al. Transcriptional insights into the CD8(+) T cell response to infection and memory T cell formation. Nat. Immunol. 14, 404–412 (2013).
Article CAS PubMed PubMed Central Google Scholar
The Gene Ontology Consortium. Gene Ontology Consortium: going forward. Nucleic Acids Res. 43, D1049–D1056 (2015).
Article CAS Google Scholar
Chang, J. T., Wherry, E. J. & Goldrath, A. W. Molecular regulation of effector and memory T cell differentiation. Nat. Immunol. 15, 1104–1115 (2014).
Article CAS PubMed PubMed Central Google Scholar
Araki, K. et al. mTOR regulates memory CD8 T-cell differentiation. Nature 460, 108–112 (2009).
Article ADS CAS PubMed PubMed Central Google Scholar
Pollizzi, K. N. et al. mTORC1 and mTORC2 selectively regulate CD8⁺ T cell differentiation. J. Clin. Invest. 125, 2090–2108 (2015).
Article PubMed PubMed Central Google Scholar
von Andrian, U. H. & Mempel, T. R. Homing and cellular traffic in lymph nodes. Nat. Rev. Immunol. 3, 867–878 (2003).
Article CAS Google Scholar
Kastenmuller, W. et al. Peripheral prepositioning and local CXCL9 chemokine-mediated guidance orchestrate rapid memory CD8+ T cell responses in the lymph node. Immunity 38, 502–513 (2013).
Article CAS PubMed PubMed Central Google Scholar
Sung, J. H. et al. Chemokine guidance of central memory T cells is critical for antiviral recall responses in lymph nodes. Cell 150, 1249–1263 (2012).
Article CAS PubMed PubMed Central Google Scholar
Bassett, J. D. et al. CD8+ T-cell expansion and maintenance after recombinant adenovirus immunization rely upon cooperation between hematopoietic and nonhematopoietic antigen-presenting cells. Blood 117, 1146–1155 (2011).
Article CAS PubMed Google Scholar
Williams, M. A. & Bevan, M. J. Effector and memory CTL differentiation. Annu. Rev. Immunol. 25, 171–192 (2007).
Article CAS PubMed Google Scholar
Zehn, D., Roepke, S., Weakly, K., Bevan, M. J. & Prlic, M. Inflammation and TCR signal strength determine the breadth of the T cell response in a bim-dependent manner. J. Immunol. 192, 200–205 (2014).
Article CAS PubMed Google Scholar
Cui, W., Joshi, N. S., Jiang, A. & Kaech, S. M. Effects of Signal 3 during CD8 T cell priming: Bystander production of IL-12 enhances effector T cell expansion but promotes terminal differentiation. Vaccine 27, 2177–2187 (2009).
Article CAS PubMed PubMed Central Google Scholar
Kaneko, S. et al. IL-7 and IL-15 allow the generation of suicide gene-modified alloreactive self-renewing central memory human T lymphocytes. Blood 113, 1006–1015 (2009).
Article CAS PubMed Google Scholar
Buchholz, V. R., Gräf, P. & Busch, D. H. The smallest unit: effector and memory CD8(+) T cell differentiation on the single cell level. Front. Immunol. 4, 31 (2013).
Article PubMed PubMed Central CAS Google Scholar
Graef, P. et al. Serial transfer of single-cell-derived immunocompetence reveals stemness of CD8+ central memory T cells. Immunity 41, 116–126 (2014).
Article CAS PubMed Google Scholar
Ahmed, R., Bevan, M. J., Reiner, S. L. & Fearon, D. T. The precursors of memory: models and controversies. Nat. Rev. Immunol. 9, 662–668 (2009).
Article CAS PubMed Google Scholar
Lefrançois, L. & Marzo, A. L. The descent of memory T-cell subsets. Nat. Rev. Immunol. 6, 618–623 (2006).
Article PubMed CAS Google Scholar
Gattinoni, L., Klebanoff, C. A. & Restifo, N. P. Paths to stemness: building the ultimate antitumour T cell. Nat. Rev. Cancer 12, 671–684 (2012).
Article CAS PubMed PubMed Central Google Scholar
Kambayashi, T., Assarsson, E., Lukacher, A. E., Ljunggren, H.-G. & Jensen, P. E. Memory CD8+ T cells provide an early source of IFN-gamma. J. Immunol. 170, 2399–2408 (2003).
Article CAS PubMed Google Scholar
Boesteanu, A. C. & Katsikis, P. D. Memory T cells need CD28 costimulation to remember. Semin. Immunol. 21, 69–77 (2009).
Article CAS PubMed PubMed Central Google Scholar
Swartz, M. A. Inflammatory lymphangiogenesis in postpartum breast tissue remodeling. J. Clin. Invest. 124, 3704–3707 (2014).
Article CAS PubMed PubMed Central Google Scholar
Angeli, V. & Randolph, G. J. Inflammation, lymphatic function, and dendritic cell migration. Lymphatic Res. Biol. 4, 217–228 (2006).
Article CAS Google Scholar
Finlay, B. B. & McFadden, G. Anti-immunology: evasion of the host immune system by bacterial and viral pathogens. Cell 124, 767–782 (2006).
Article CAS PubMed Google Scholar
Hickman, H. D. et al. Direct priming of antiviral CD8+ T cells in the peripheral interfollicular region of lymph nodes. Nat. Immunol. 9, 155–165 (2008).
Article CAS PubMed Google Scholar
Tamburini, B. A., Burchill, M. A. & Kedl, R. M. Antigen capture and archiving by lymphatic endothelial cells following vaccination or viral infection. Nat. Commun. 5, 3989 (2014).
Clement, C. C., Rotzschke, O. & Santambrogio, L. The lymph as a pool of self-antigens. Trends Immunol. 32, 6–11 (2011).
Article CAS PubMed Google Scholar
Thomas, S., Kolumam, G. A. & Murali-Krishna, K. Antigen presentation by nonhemopoietic cells amplifies clonal expansion of effector CD8 T cells in a pathogen-specific manner. J. Immunol. 178, 5802–5811 (2007).
Article CAS PubMed Google Scholar
Fletcher, A. L. et al. Reproducible isolation of lymph node stromal cells reveals site-dependent differences in fibroblastic reticular cells. Front. Immunol. 2, 1–15 (2011).
Article Google Scholar
Link, A. et al. Fibroblastic reticular cells in lymph nodes regulate the homeostasis of naive T cells. Nat. Immunol. 8, 1255–1265 (2007).
Article CAS PubMed Google Scholar
Lutz, M. B. et al. An advanced culture method for generating large quantities of highly pure dendritic cells from mouse bone marrow. J. Immunol. Methods 223, 77–92 (1999).
Article CAS PubMed Google Scholar
Broggi, M., Schmaler, M. & Lagarde, N. Isolation of murine lymph node stromal cells. J. Vis. Exp. 19, e51803 (2014).
Google Scholar
Korff T. Three-Dimensional Spheroid Culture of Endothelial Cells. In Methods in Endothelial Cell Biology (ed. Augustin, H. G.) Ch. 6 (Springer, Berlin, 2004).
Chapter Google Scholar
Overwijk, W. W. et al. gp100/pmel 17 is a murine tumor rejection antigen: induction of “self-”reactive, tumoricidal T cells using high-affinity, altered peptide ligand. J. Exp. Med. 188, 277–286 (1998).
Article CAS PubMed PubMed Central Google Scholar
Pope, C. et al. Organ-specific regulation of the CD8 T cell response to Listeria monocytogenes infection. J. Immunol. 166, 3402–3409 (2001).
Article CAS PubMed Google Scholar
David, F. P. A. et al. HTSstation: a web application and open-access libraries for high-throughput sequencing data analysis. PLoS One 9, e85879 (2014).
Article ADS PubMed PubMed Central CAS Google Scholar
Langmead, B., Trapnell, C., Pop, M. & Salzberg, S. L. Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biol. 10, R25 (2009).
Article PubMed PubMed Central CAS Google Scholar
Ritchie, M. E. et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 43, e47–e47 (2015).
Article PubMed PubMed Central CAS Google Scholar
Anders, S. & Huber, W. Differential expression analysis for sequence count data. Genome Biol. 11, R106 (2010).
Article CAS PubMed PubMed Central Google Scholar

Download references

Acknowledgements

The authors are grateful to Yue Wang, Yassin Ben Saida, Xavier Quaglia-Thermes, Giacomo O. Diaceri, Vasiliki Panagiotou, S. Oberle, Cristina Cartoni, E. Anastasiou, Elena Cabello, the EPFL Flow Cytometry Core Facility (Miguel Garcia, Loïc Tauzin, Valérie Glutz) and the University of Chicago Cytometry and Antibody Technology Core (David Leclerc, Mandel Davis, Mike Olson, and Robert Ladd) for technical support and assistance, as well as Maria-Luisa Alegre, Bana Jabri, Albert Bendelac, Amanda W. Lund, Catherine M. Card and Eleonora Simeoni for helpful discussions and suggestions. This work was funded by grants from the Swiss Cancer Research (KFS-3312-08-2013 to M.A.S.), the European Research Council (AdG-323053 to M.A.S.), and the National Institutes of Health (R01 CA219304-01A1 to M.A.S.).

Author information

These authors contributed equally: Efthymia Vokali, Shann S. Yu.

Authors and Affiliations

Institute of Bioengineering, École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland
Efthymia Vokali, Shann S. Yu, Sachiko Hirosue, Marcela Rinçon-Restrepo, Patricia Corthésy-Henrioud, Witold W. Kilarski & Melody A. Swartz
Pritzker School of Molecular Engineering, University of Chicago, Chicago, IL, USA
Shann S. Yu, Witold W. Kilarski & Melody A. Swartz
Department of Pathology and Immunology, Faculty of Medicine, University of Geneva, Geneva, Switzerland
Fernanda V. Duraes & Stéphanie Hugues
Swiss Vaccine Research Institute, Epalinges, Switzerland
Stefanie Scherer & Dietmar Zehn
Division of Immunology, Transplantation and Infectious Diseases, San Raffaele Scientific Institute, Milan, Italy
Anna Mondino
Ben May Department of Cancer Research, University of Chicago, Chicago, IL, USA
Melody A. Swartz

Authors

Efthymia Vokali
View author publications
You can also search for this author in PubMed Google Scholar
Shann S. Yu
View author publications
You can also search for this author in PubMed Google Scholar
Sachiko Hirosue
View author publications
You can also search for this author in PubMed Google Scholar
Marcela Rinçon-Restrepo
View author publications
You can also search for this author in PubMed Google Scholar
Fernanda V. Duraes
View author publications
You can also search for this author in PubMed Google Scholar
Stefanie Scherer
View author publications
You can also search for this author in PubMed Google Scholar
Patricia Corthésy-Henrioud
View author publications
You can also search for this author in PubMed Google Scholar
Witold W. Kilarski
View author publications
You can also search for this author in PubMed Google Scholar
Anna Mondino
View author publications
You can also search for this author in PubMed Google Scholar
Dietmar Zehn
View author publications
You can also search for this author in PubMed Google Scholar
Stéphanie Hugues
View author publications
You can also search for this author in PubMed Google Scholar
Melody A. Swartz
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

Conceptualization: E.V., S. Hirosue, S. Hugues, M.A.S. Methodology: E.V., S.S.Y., S. Hirosue, M.R.R., F.d.V.D., S.S., W.W.K., A.M., D.Z., S. Hugues, M.A.S. Investigation: E.V., S.S.Y., S. Hirosue, P.C.H., M.A.S. Validation: S.S.Y., S. Hirosue, A.M., D.Z., S. Hugues, M.A.S. Analysis: E.V., S.S.Y., P.C.-H., M.A.S. Resources: D.Z., S. Hugues, M.A.S. Supervision: M.A.S. First draft: E.V. Revisions: E.V., S.S.Y., M.A.S. Visualization: E.V., S.S.Y. Project administration: M.A.S. Supervision: M.A.S. Funding acquisition: M.A.S.

Corresponding author

Correspondence to Melody A. Swartz.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Peer review information Nature Communications thanks the anonymous reviewer(s) for their contribution to the peer review of this work.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Reporting Summary

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Vokali, E., Yu, S.S., Hirosue, S. et al. Lymphatic endothelial cells prime naïve CD8⁺ T cells into memory cells under steady-state conditions. Nat Commun 11, 538 (2020). https://doi.org/10.1038/s41467-019-14127-9

Download citation

Received: 12 December 2018
Accepted: 19 December 2019
Published: 27 January 2020
DOI: https://doi.org/10.1038/s41467-019-14127-9

This article is cited by

Lymphatic vessels in the age of cancer immunotherapy
- Triantafyllia Karakousi
- Tenny Mudianto
- Amanda W. Lund
Nature Reviews Cancer (2024)
Combined PD-L1/TGFβ blockade allows expansion and differentiation of stem cell-like CD8 T cells in immune excluded tumors
- Alessandra Castiglioni
- Yagai Yang
- Sören Müller
Nature Communications (2023)
Trafficking and retention of protein antigens across systems and immune cell types
- Thu A. Doan
- Tadg Forward
- Beth A. Jirón Tamburini
Cellular and Molecular Life Sciences (2022)
A short review on lymphatic endothelial cell heterogeneity
- Masayuki Miyasaka
Inflammation and Regeneration (2021)

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.