Afferent Lymphatic Transport and Peripheral Tissue Immunity

Abstract

Lymphatic vessels provide an anatomical framework for immune surveillance and adaptive immune responses. Though appreciated as the route for antigen and dendritic cell transport, peripheral lymphatic vessels are often not considered active players in immune surveillance. Lymphatic vessels, however, integrate contextual cues that directly regulate transport including changes in intrinsic pumping and capillary remodeling and express a dynamic repertoire of inflammatory chemokines and adhesion molecules that facilitates leukocyte egress out of inflamed tissue, and these mechanisms all together contribute to the course of peripheral tissue immunity. In this review, we focus on context-dependent mechanisms that regulate fluid and cellular transport out of peripheral non-lymphoid tissues to provide a framework for understanding the effects of afferent lymphatic transport on immune surveillance, peripheral tissue inflammation, and adaptive immunity.

The lymphatic system is composed of a hierarchy of lymphatic vessels that transport lymph unidirectionally to lymph nodes (LNs) and the specialized lymphatic endothelial cells (LECs) that compose both these vessels and LN lymphatic networks. Blunt ended lymphatic capillaries are composed of a single layer of LECs anchored to extracellular matrix (ECM) fibers and extend into peripheral tissues to facilitate uptake of fluid, cells, and lipids. Collecting vessels, differentiated from capillaries by the investment of lymphatic muscle cells (LMCs), a continuous basement membrane, and valves, contract to generate intrinsic propulsive forces that transport lymph to LNs and prevent retrograde flow. The structure and function of lymphatic vessels as it varies across organ sites in health and disease was recently covered in two excellent reviews (1, 2). Here we focus on the role of afferent lymphatic transport in peripheral tissue immune responses.

Lymphatic transport and interstitial fluid dynamics regulate the physiological context of peripheral tissues at steady state and in response to challenge (e.g. infection, wound healing, tumor formation). Peripheral tissues receive necessary nutrients and oxygen from vascular transudate, which drives a net excess of fluid within tissue interstitium, generating directional interstitial fluid flows (0.1–1.0 μm/s) toward lymphatic capillaries (3). At steady state, the sparse basement membrane and discontinuous intercellular junctions (termed button-like junctions) found in lymphatic capillaries allow passive, paracellular fluid transport to form lymph (4, 5). In addition to the movement of fluid, the distribution of signaling molecules, cytokines, and antibodies is biased towards lymphatic transport dependent upon oncotic pressure gradients that limit vascular reabsorption. Large particulates, such as exosomes, chylomicrons, and protein complexes, must enter lymphatic vessels to access the systemic circulation (6). Lymph formation is further supported by active endocytic and transcellular LEC transport (7) and, as a net result, lymph is composed of a unique repertoire of tissue-derived lipids, metabolites, soluble proteins, and antigens that reflects the immunological status of the tissue from which it originates (8). Upon arrival in LNs, lymph and its constituents orchestrate the rapid activation of adaptive immune responses where both the functional heterogeneity of LN LECs (9) and fluid transport (10) directly impact antigen distribution, presentation, and leukocyte interactions.

Here we provide an overview of afferent lymphatic transport during peripheral tissue immune responses. We discuss our current understanding of how lymphatic vessels move fluid and leukocytes out of resting and inflamed tissue, how lymphatic transport is affected by inflammatory context, and how it contributes to immune surveillance, adaptive immune activation, and resolution in peripheral non-lymphoid tissue.

Immune Surveillance

Through the constitutive transport of lymph, memory lymphocytes, and antigen presenting cells (APCs), the afferent lymphatic vasculature provides the anatomic framework for immune surveillance in peripheral, non-lymphoid tissues (Figure 1), defined here as the initial detection of pathogenic or inflammatory insult. The observation that afferent lymph harbors a significant population of migratory leukocytes distinct from those found in blood and efferent lymph was made decades ago following cannulation of ovine afferent lymphatic vessels (11, 12), and similar findings have been reported in mice (13, 14) and humans (15, 16). These early studies profiled afferent lymph across tissues and demonstrated a relatively constant proportion of mature lymphocytes (80–90%), myeloid cells (5–20%), and various granulocytes and plasma cells in sheep (11) and mice (17). Lymph-borne lymphocytes at steady state are predominantly of a memory phenotype and mostly CD4⁺ (including regulatory T cells from skin (18)), though also include B cells, CD8⁺ and γδ T cells (19). In contrast, naïve T cells are largely excluded from non-lymphoid peripheral tissue draining lymph, with the exception of gut (19, 20), where Peyer’s patches likely provide a source of naïve T cells that enter afferent lymph. Importantly, the recirculation of memory lymphocytes provides the opportunity for rapid response to secondary challenge, whereas migratory APCs amplify memory and activate de novo responses in LNs to drive protective adaptive immune responses.

Figure 1. The lymphatic anatomy of immune surveillance.

Under basal conditions, immune surveillance is regulated by constitutive lymph transport and the trafficking of lymphocytes (naive, memory and regulatory) and antigen presenting cells between peripheral tissue, lymphoid, and systemic compartments. Vascular transudate generates directional interstitial fluid flows in the direction of draining, low pressure lymphatic capillaries, where open, discontinuous button-like junctions (dotted red lines) facilitate passive uptake and permit integrin-independent cell transmigration. Oncotic gradients and interstitial fluid flow (0.1–1 μm/s) enrich afferent lymph for tissue-derived factors, antigens, large particulates, and protein complexes that reflect the immunological status of the tissue from which it drains. Collecting lymphatic vessels propel fluid towards draining lymph nodes through intrinsic pumping mechanisms regulated by invested LMCs, and continuous zipper-like intercellular junctions (solid red lines) restrict vessel permeability. Leukocytes migrating through afferent lymph are predominantly of memory and regulatory phenotypes, distinct from both efferent lymph and blood where naive T cells also recirculate in search of cognate antigen loaded on migratory DCs and other antigen presenting cells. Resident memory T cells are defined by their exclusion from afferent lymph and long-term retention in peripheral, non-lymphoid tissue at basal conditions. Dendritic cells (DC); naive T cells (T_N); memory T cells (T_Mem); resident memory T cells (T_RM); regulatory T cells (T_REG); lymphatic muscle cell (LMC).

Lymph Transport to Lymph Nodes

In response to peripheral tissue challenge, pro-inflammatory mediators activate the vascular endothelium, increasing permeability and fluid influx that drives tissue swelling. Accumulation of interstitial fluid (termed edema) is itself a critical feature of host defense as it directly promotes effector molecule (e.g. immunoglobulins and complement) accumulation, and thereby facilitates local inflammation, pathogen control, and adaptive immunity (21). Multiple overlapping signals regulate vascular permeability (reviewed (22)). Mast cells, for example, produce histamine that activates type I histamine (H₁) receptor leading to Src-mediated tyrosine phosphorylation of adherens junction molecules, VE-cadherin and β-catenin (23). Disruption of tight and adherens junctions disrupts the size exclusion properties of post-capillary venules allowing for the influx of larger macromolecules. The importance of vascular permeability is highlighted in mouse models of Herpes Simplex Virus Type 2 (HSV-2) and Vesicular Stomatitis Virus (VSV) infection, where production of IFNγ by memory CD4⁺ T cells increases vascular permeability, which is necessary for antibody access to infected neuronal tissue and viral control (24). While increased vascular permeability enhances tissue access, the extent to which fluid and effector molecules accumulate in the interstitial space also depends on local lymphatic transport.

In response to increased interstitial fluid load and inflammatory mediators, lymphatic vessels can adapt their intrinsic pumping activity to both increase and decrease transport dependent on context. In lymphatic vessels of the guinea pig mesentery histamine increases the frequency of collecting lymphatic constriction by activating LMC H₁ receptors (25). In contrast, histamine may also act through H₁ receptors expressed by LECs to cause nitric oxide (NO)-dependent relaxation (26, 27), and stimulation of LMC H₂ induces relaxation that may reduce fluid flows at least in a subset of vessels of the mesentery (25). Imposed fluid flow through ex vivo mesenteric lymphatic vessel explants also induces NO production and subsequently reduces, rather than elevates, collecting lymphatic contraction and pumping (28). In vivo, lymph stasis, initiated by a surgical model of persistent mouse tail lymphedema, drives the accumulation of Th2 CD4⁺ T cells (29) that inhibit collecting lymphatic pumping through local activation of macrophage inducible nitric oxide synthase (iNOS) (30) and exacerbate local pathology. While increased intrinsic pumping would presumably counteract vascular permeability and reduce edema, a decrease in pumping may limit transport to LNs. To this point, during oxazolone treatment of skin, a rapid, transient reduction in collecting lymphatic vessel contraction, dependent upon the overproduction of NO by CD11b⁺Gr1⁺ myeloid cells, transiently decreases antigen-loaded dendritic cell (DC) accumulation in LNs and impairs induction of experimental autoimmune encephalomyelitis (31). Therefore, it appears as though local inflammatory mediators act in coordination with lymphatic transport to affect tissue physiology and the kinetics of antigen presentation in LNs. How inflammatory context (e.g. infection vs allergy) may differentially affect collector contraction within various tissue sites, remains an important question moving forward.

The contextual plasticity of collecting lymphatic vessels, described above, is progressively lost with age (32) and may be disrupted by infection. Cutaneous infection with methicillin-resistant Staphylococcus aureus (MRSA) results in an acute reduction in lymphatic vessel contractility and lymph flow that persists long after MRSA infection and inflammation are resolved due to a permanent loss of LMCs (33). In the gut, Yersinia pseudotuberculosis infection induces persistent local inflammation driven by lymphatic leakage in mesenteric adipose tissue and impaired DC migration to LNs (34). Consistent with these preclinical findings, patients with lymphedema exhibit increased risk for skin and soft tissue infections and conversely, lymphedema can arise as a consequence of infection (35, 36), suggesting the possibility of regional changes in lymphatic transport that have long-lasting effects on local immune surveillance.

In addition to inflammation and flow-mediated changes to collecting lymphatic contraction, recent evidence indicates that lymphatic capillaries are responsive to inflammatory context and may remodel their interendothelial junctions to regulate paracellular transport. Lymphatic capillary junctions, normally discontinuous (buttons), become continuous (zippering) during Mycoplasma Pulmonis (M. pulmonis) infection of the trachea and lose the loose, interendothelial flaps presumed to facilitate passive fluid and cellular uptake (37). The intestinal lacteal, a specialized lymphatic capillary of the small intestine, exhibits a similar zippering phenotype following antibiotic depletion of commensal microbiota (38) or increased bioavailability of vascular endothelial growth factor (VEGF-A) (39), in both cases leading to impaired dietary lipid absorption. During cutaneous Vaccinia virus (VACV) infection, a rapid reduction in lymphatic transport is dependent upon type I IFN signaling, and associated with elongation of lymphatic capillary blunt ends and viral retention within skin (40). Peripheral lymphatic capillaries are therefore sensitive to changing context and may determine differential transport of fluid and soluble macromolecules that impacts immune surveillance. How lymphatic capillary zippering is regulated across diverse peripheral tissues and its physiological relevance in vivo, however, remains to be carefully dissected in a cell-specific manner.

Dendritic Cell Trafficking via Afferent Lymphatic Vessels

While lymph flow can deliver soluble antigen to LNs in minutes (41–44), the data discussed above suggest the possibility that under certain inflammatory contexts, lymphatic vessel-intrinsic mechanisms reduce fluid transport and thereby may impact the kinetics of antigen presentation in LNs. Interestingly, though fluid flow is reduced following VACV infection by scarification, DCs continue to migrate across inflamed lymphatic vessels and prime CD8⁺ T cells in draining LNs (40). Similarly, while subcutaneous injection of HSV-1 leads to rapid virion transport to LNs and antigen presentation by resident DCs, CD8⁺ T cell activation following vaginal administration depends specifically on migratory DCs (45). Therefore, how lymphatic vessels actively regulate the specific transport of DCs, in concert with fluid transport, likely determines the kinetic of adaptive immune priming in LNs.

The migration of mature DCs to LNs is regulated at each step by the lymphatic vasculature, which provides directional cues to increase the probability that antigen-loaded DCs reach their cognate T lymphocytes in LNs. While interstitial fluid flows are slow (0.1–1μm/s) (46), they are sufficient to bias interstitial gradients (47, 48), influence matrix remodeling and orientation (49, 50), and activate adhesive properties and chemokine secretion in draining lymphatic vessels (51). These mechanisms together collaborate to facilitate directional DC migration towards draining lymphatic capillaries in peripheral, non-lymphoid tissue. DC migration is importantly largely, though not exclusively, dependent upon the homeostatic chemokine CCL21, which is expressed by LECs, required for CCR7-dependent homing (52), and elevated by inflammatory cytokines, such as TNF-ɑ (53, 54), and inflammatory fluid flows (48, 50, 54). DCs increase CCL21 production by LECs during transmigration that supports subsequent waves of DC migration (55–57). Interestingly, the cytoplasmic tail of programmed death-ligand 1 (PD-L1) on conventional type 1 (cDC1) and type 2 (cDC2) dermal DCs reinforces G-coupled protein receptor signaling and enhances CCR7-directed LN migration via CCL21 gradients (58), suggesting there are multiple, intrinsic and extrinsic mechanisms that regulate the lymphatic migration of DCs. Importantly, CCR7-dependent DC migration is required for antigen presentation to naive CD8⁺ T cells from skin during HSV-1 infection and following DC adoptive transfer (45, 53, 59, 60); for presentation of intestinal epithelial cell-derived antigens from lamina propria in mesenteric LNs (61); and presentation of tumor antigen by cross-presenting cDC1s in tumor-draining LNs (62, 63).

Interestingly, however, during allergic sensitization, cDC2s can migrate to LNs in a CCR7-independent manner (64), suggesting that mechanisms of lymphatic homing and migration may be cell type and state specific. Given that LECs activate specific secretion profiles in response to inflammatory cytokines (type 1, 2, 17, etc) (65, 66) and toll-like receptor stimulation (67), it is possible that a dynamic LEC chemokine repertoire may preferentially support the tissue exit of specific DC subtypes in a context-dependent manner. To this point, dermal LECs secrete CXCL12 in response to cutaneous irritants such as 2,4-dinitrofluorobenzene, which drives CXCR4⁺ DC exit including langerin^+/−, CD11c⁺, and CD11b⁺ subsets (68), and CX3CL1 stimulated by oxazolone treatment recruits CX3CR1⁺CD11c⁺ dermal DCs (54). The signaling sphingolipid, sphingosine-1-phosphate (S1P), is required for lymphocyte egress from lymphoid tissues (69), but may also play a role in mediating DC egress from peripheral tissues such as skin and lung (70, 71). Lymph S1P is maintained by LECs (72, 73) and migratory DCs show a tissue-specific requirement for the S1P receptors. Dermal DCs seem to rely only on S1P₁ to migrate to draining LNs following topical application of FITC in organic solvents, whereas both S1P₁ and S1P₃ are required for the accumulation of CD103⁺ DCs in mesenteric LNs draining LPS-inflamed gut (74).

In addition to producing chemotactic gradients, peripheral LECs express several atypical chemokine receptors (ACKRs), including ACKR2 and ACKR4, that scavenge inflammatory chemokines to shape interstitial gradients, which may amplify directional cues. ACKR4, expressed by dermal LECs and keratinocytes, scavenges excess CCL19, another CCR7-binding chemokine, to improve CCL21-dependent homing of CCR7⁺ Langerhan cells to lymphatic vessels following cutaneous terephthalic acid application (75); and ACKR2 (also known as D6 and expressed by dermal LECs (76, 77)) scavenges CCL2 and may promote the migration of DCs by preventing the perilymphatic accumulation of CCR2⁺ inflammatory myeloid cells (78). Inflammatory chemokine scavenging by the afferent lymphatic endothelium may also prevent persistent LN activation. Indeed, mice lacking ACKR2 display exacerbated inflammatory responses following subcutaneous administration of complete Freund’s adjuvant (CFA) (79).

At steady state, the permissive nature of lymphatic capillary interendothelial “button” junctions allows for integrin-independent DC transendothelial migration across the lymphatic endothelium (5, 80). However, the lymphatic capillary zippering observed in some infection and inflammatory contexts (37), may presumably remove these passive portals and thus increase the requirements for DC transmigration (5). To support DC transendothelial migration, inflammatory cytokines including TNF-ɑ and IL-1β (81–83), and transmural fluid flows (51) induce expression of cellular adhesion molecules including ICAM-1, VCAM, and selectins. Though shear stresses are much lower in lymphatic capillaries as compared to blood, active binding through cell adhesion molecules and selectins is still required for DC entry into inflamed dermal lymphatic capillaries (51). Both antibody blockade and genetic ablation of ICAM-1, VCAM or their respective integrin ligands (LFA-1 and CD11b) significantly reduces dermal DC migration to LNs in several models of skin inflammation (81, 84, 85) and inhibits adaptive immune responses (86). The hyaluronan (HA) binding lymphatic vessel endothelial receptor 1 (LYVE-1), expressed by lymphatic capillaries, facilitates HA-dependent DC docking and transendothelial migration under inflammatory contexts and LYVE1-deficient animals exhibit reduced CD8⁺ T cell priming following vaccination (87). Owing to the low shear stresses in lymphatic capillaries, intraluminal DCs exhibit a slow, semi-directional, ICAM-1 dependent crawling behavior (88). This proceeds until DCs reach collecting vessels (89) where elevated shear stresses are hypothesized to downregulate adhesion molecules to permit flow-induced transport to LNs (90). While the immunological consequences of intralymphatic crawling are unknown, it may suggest that the lumen of lymphatic capillaries is a specialized niche wherein prolonged interactions between DCs, lymphocytes (91), and LECs occur that may have functional consequences for immune responses (83).

Peripheral Tissue Inflammation and Leukocyte Egress

The number of leukocytes present in afferent lymph increases during acute and chronic inflammation (13, 92, 93), indicating a dynamic process of tissue exit that may contribute to peripheral tissue inflammation. Like fluid transport, leukocyte egress is likely critical for the removal of pro-inflammatory signals that would otherwise exacerbate tissue inflammation and immunopathology. Multiple studies indicate that the induction of lymphangiogenesis by the lymphangiogenic growth factor, VEGF-C, in chronically inflamed tissues (e.g. dextran sulfate sodium-induced colitis, psoriasis-like cutaneous inflammation, and tumors) increases fluid and leukocyte transport out of tissue and thereby reduces ongoing inflammatory processes (94–96). Similarly, during cutaneous Leishmania major infection, VEGF-A-induced lymphangiogenesis restricts lesion formation without affecting parasite burden by reducing inflammatory lesion size (97).

While it is clear that multiple cell types are capable of accessing afferent lymph, the mechanisms that direct the egress of diverse leukocyte subtypes have only been examined in a handful of tissue and inflammatory contexts. Still, these data indicate interesting cell type and tissue state specificity. For example, neutrophils use CCR7 to egress acutely inflamed skin (CFA) (98) and cremaster muscle (99), but CXCR4 and CD11b to reach LNs following cutaneous Staphylococcus aureus infection, where they boost T cell proliferation in draining LNs (100). ICAM-1 and its cognate ligands are required for neutrophil transendothelial migration in several contexts including CFA-induced inflammation and Mycobacterium bovis infection of skin (101, 102). B cells also exhibit elevated rates of egress from chronically CFA-inflamed skin (103) via both CCR7-dependent and independent mechanisms (104); and entrance of monocytes and macrophages into afferent lymph (105–108) is inhibited by integrin ɑ1β1 in concanavalin A inflamed skin (109).

T cell Egress from Inflamed Peripheral Tissue

The abundance of T cells in afferent lymph and their critical role in immune surveillance and adaptive immunity has encouraged more studies examining mechanisms of T cell exit from inflamed tissues, particularly from skin. While it appears that multiple mechanisms may regulate CD4⁺ and CD8⁺ T cell egress via lymphatic vessels (Figure 2), as with DCs, the predominant pathway studied to date is CCL21-CCR7, which is required for egress at both steady state (110) and during acute inflammation in lung (111, 112) and skin (104). CCR7-deficient CD4⁺ Th1 cells dramatically accumulate in skin and exacerbate pathology associated with delayed-type hypersensitivity responses (113) and transgenic expression of CCR7 in CD4⁺ T cells is sufficient to enhance egress from lung following antigen challenge (111). Therefore, T cell egress through lymphatic vessels may provide an important control point to balance protective immunity with damaging immunopathology. The mechanisms that regulate T cell egress, however, seem to vary by context. CD4⁺ and CD8⁺ T cell exit from chronically inflamed skin and tumors appears to be CCR7-independent (13, 114), however, the signals that direct exit in this context remain unknown (115). In addition to chemokines, T cells use multiple adhesion molecules to facilitate transendothelial migration including common lymphatic endothelial and vascular endothelial receptor-1 (116)), ICAM-1 (84) and macrophage mannose receptor-1, which facilitates interactions with lymphocyte CD44 (117); though when, where, and for which subsets each of these are used remains incompletely understood. In addition to ɑβ T cells, γδ T cells are found in afferent lymph draining resting (19, 118), inflamed (119), and malignant (114, 120) tissues. Important surveyors of peripheral non-lymphoid tissue, γδ T cells are presumed to exit tissue via CCR7-independent mechanisms owing to a lack of CCR7 surface expression in bovine (118) and ovine lymph (119); here high expression of E-selectin and CCR6-dependent chemotaxis towards CCL20 may indicate alternative mechanisms of lymphatic homing (119). These mechanisms, however, have not been tested functionally in vivo.

Figure 2. T cell egress from peripheral tissues via lymphatic capillaries.

Peripheral lymphatic capillaries mediate the directional homing and transmigration of T lymphocytes from steady state and inflamed peripheral tissues. (1) Under basal conditions, the trafficking of memory and regulatory lymphocytes is largely dependent on the homeostatic chemokine, CCL21, constitutively expressed by peripheral lymphatic capillaries. (2) Inflammatory cytokines, such as TNFɑ and IL-1β, activate regional blood and lymphatic vessels, increasing vascular permeability, interstitial fluid flow, and inducing lymphatic endothelial production of a diverse repertoire of chemokines and adhesion molecules that presumably direct the egress of T cells. Activated CD4⁺ and CD8⁺ effector T cells are recruited to inflamed peripheral tissues where recognition of cognate antigen, downregulates lymphatic homing receptors (S1PRs and CCRs), supports tissue retention, and ultimately long-term residence as T_RM. IFNγ produced by antigen-specific CD8⁺ T cell responses further activates the lymphatic vasculature to induce expression of MHCII and PD-L1, which may act to negatively regulate ongoing cytotoxic immunity. Active mechanisms increase the overall abundance of lymph-borne lymphocytes. (3) S1P acts both on CD4⁺ T cells, which use S1P₁ and S1P₄ to enter afferent lymph, and lymphatic endothelial cells where S1P₂ regulates VCAM expression and transendothelial migration. Additionally, during acute inflammatory processes in skin, CCR7 and ICAM-1 is required for T cell egress via lymphatic vessels, while CCR7-dependence is lost in chronic inflammation. (4) In grafts, T_REG cell migration to lymph nodes is necessary to limit alloimmune responses and transmigrating T_REG condition the lymphatic endothelium to express higher levels of VCAM through activation of LTβR that supports subsequent egress of inflammatory lymphocytes (CD4⁺ and CD8⁺ T cells) to resolve peripheral tissue inflammation. (5) Lymphatic capillaries also directly regulate peripheral tissue inflammation by scavenging inflammatory chemokines through expression of decoy receptors (ACKR2/D6) limiting their effect in peripheral tissue and transport to lymph nodes. Memory T cells (T_Mem); effector T cells (T_Eff); resident memory T cells (T_RM); regulatory T cells (T_REG); antigen presenting cell (APC).

In addition to chemokines, S1P may play an important role in T lymphocyte egress from tissue (121). Effector and memory CD4⁺ T cells respond to S1P gradients in vitro and require S1P₁ and S1P₄ for migration into lymphatic vessels following co-injection with LPS (122). Interestingly, S1P₁ and S1P₄ appear to exhibit non-redundant roles and coordinate with S1P₂ on lymphatic vessels to activate adhesive sites for transendothelial migration (122). This highlights the important fact that S1P also regulates endothelial cell homeostasis (123, 124), which should be considered when manipulating S1P signaling in vivo. Consistent with the hypothesis that S1P regulates T cell exit from tissue, S1P receptors are transcriptionally repressed in tissue resident memory T cells (T_RM) (125, 126) and enforced S1pr1 expression reduces resident memory formation by CD8⁺ T cells (126). S1P receptor stimulation and desensitization through application of the small molecule FTY720, is sufficient to reduce CD4⁺ T cell egress to LNs (121), and rescues CD69^−/− retention in skin following HSV-2 infection (127). FTY720-treated CD4⁺ T cells notably arrest their migration at the basal surface of lymphatic monolayers in vitro and capillaries in vivo (121), perhaps suggesting that chemokine receptors (e.g. CCR7) generate directional gradients whereas S1P signaling acts in coordination with cell adhesion molecules to facilitate transendothelial migration. Interestingly, antigen recognition in peripheral, non-lymphoid tissue boosts tissue retention (112, 113) and the formation of tissue resident memory (128). Consistent with this, influenza virus-specific T cells are largely CCR7^−/− in lung (111, 129) and TCR stimulation is sufficient to modulate surface expression of CCR7 (130). Taken together, the antigen-dependence of T cell egress might contribute to the focusing of the interstitial T cell repertoire through combined S1P and CCR7-dependent signals to promote the recirculation of bystander T cells while limiting overt pathology. How local antigen, inflammatory context, and lymphatic-derived signals may over time regulate interstitial T cell repertoires and contraction to memory is an interesting area for continued study.

Interestingly, peripheral lymphatic vessels are responsive to cytokines derived from infiltrating and transmigrating T lymphocytes and subsequently adapt their surface expression and function. Regulatory T cells (T_REG), which migrate from skin during steady state, show increased migration during CHS inflammation and migratory T^REG exhibit more potent suppressive activity when compared to LN resident T_REG cells (18). Similarly in allograft models, T_REG cell egress suppresses alloimmune responses and improves graft survival (131). Interestingly, T_REG cells activate lymphotoxin β receptor (LTβR) signaling in LECs to specifically facilitate their VCAM-dependent egress from grafted tissue (132), and, in so doing, also condition LECs to increase expression of CCL21 that supports subsequent effector CD4⁺ and CD8⁺ T cell migration (133). This resolves the allograft response by both removing excessive inflammatory cells (e.g. CD4⁺ and CD8⁺ T cells, neutrophils, and macrophages) and inhibiting activation in LN. Furthermore, IFNγR signaling in lymphatic capillaries seems to limit effector T cell responses in both infected and malignant skin (134). IFNγ drives expression of both MHCII and PD-L1 on peripheral lymphatic capillaries (134), which may indicate the potential for direct signaling between T lymphocytes and inflamed capillaries, consistent also with prior observations that peripheral lymphatic vessels can scavenge and cross-present tumor-derived antigens (94). How the crosstalk between accumulating lymphocytes and lymphatic vessels contributes to ongoing inflammatory processes remains to be systematically dissected across tissue sites, but these studies indicate that it may have significant implications for local pathology and disease control.

Inflammatory Lymphangiogenesis

Importantly, the mechanisms discussed above may all be influenced by the progressive lymphatic vessel remodeling observed in inflamed tissues. Tissue-infiltrating macrophages secrete VEGF-C and drive local lymphangiogenesis in tumors (135, 136) and experimental inflammatory models in gut and skin (95, 96, 137). Under chronic conditions, lymphatic vessels are often dilated and leaky (95, 138), which can ultimately lead to poor tissue clearance (33, 139) and thus exacerbate local pathology. Importantly, whether once expanded, lymphatic networks contract as inflammation resolves remains poorly understood. In LNs, LECs undergo extensive inflammatory lymphangiogenesis in part driven by B cell-derived VEGF-A (140) and contract post-inflammation in an IFN-dependent manner (141, 142). However, the extent to which lymphangiogenic vessels in non-lymphoid peripheral tissues contract post-inflammation appears both tissue-specific and dependent on the inflammatory insult itself. Airway lymphatic networks remain expanded for several months after M. pulmonis clearance and are resistant to dexamethasone treatment (143); and transgenic VEGF-C expression leads to persistent durable lymphatic hyperplasia in skin (144). In contrast, lymphangiogenic corneal networks completely regress following sterile injury (145, 146), and surprisingly exhibit accelerated lymphangiogenic responses following repeated challenge (147), perhaps suggesting a type of inflammatory memory (148). How the immunological experience of a peripheral tissue may lay the groundwork for future regional immune responses through long-lasting changes in lymphatic morphology and transport, particularly in barrier tissues, will continue to be an exciting area of study.

Conclusions

As this field grows, it becomes increasingly clear that lymphatic vessels and their associated transport properties directly contribute to immune homeostasis and disease through multiple overlapping and context-dependent mechanisms. As the exquisite phenotypic and functional heterogeneity of lymphatic vessels is revealed, both within and across tissues, we are in critical need of functional studies that dissect the relative contribution of distinct LEC subsets and pathways to immunity with anatomic resolution in vivo. These mechanisms will inform therapeutic strategies to leverage lymphatic biology for immune benefit in patients across disease type.

There is already growing interest in identifying strategies that boost immune surveillance or alter peripheral tissue inflammation through the manipulation of resident lymphatic vessels. This is particularly true in the context of cancer where it was first shown in melanoma (149) and subsequently in glioblastoma (150) that overexpression of VEGF-C improves tumor-immune surveillance and response to immunotherapy. Similar approaches in the context of the central nervous system have raised the exciting prospect that boosting lymphatic transport through VEGF-C delivery may improve neurodegeneration (151, 152) and traumatic brain injury (153). VEGF-C, however, has pleiotropic effects on lymphatic and blood vessels, as well as recruited myeloid cells, and it remains unclear what the relative contribution of each of these players may be to the therapeutic effects observed. Furthermore, at least in a vascularized tissue, rapid and potent immunity can be achieved in the absence of VEGF-C driven lymphangiogenesis (40) and DC trafficking proceeds even in the face of severe lymphatic hypoplasia (154). Therefore, therapeutic approaches to boost immune surveillance or reduce chronic inflammation may not need to necessarily focus on hyperproliferative vessel growth. Importantly, the data we discuss above paints a less binary picture of lymphatic transport, where its plasticity and transient responsiveness to inflammatory context is required for both the activation and resolution of inflammatory insult. Given the vast array of mechanisms through which lymphatic vessels regulate the inflammatory state of the peripheral tissues that they drain, a nuanced approach to manipulating their function is likely warranted. While there is clearly a therapeutic opportunity to target lymphatic vessels for immune benefit across disease states, there is also much to understand about normal and inflamed lymphatic vessel function and its multifaceted roles in immunity to inform these approaches.