Lymphatic vessels in solid organ transplantation and immunobiology

2.1 Heart transplantation

The heart contains a robust LV network extending from the subepicardium to the myocardium and subendocardium. In humans, the most extensive lymphatic network extends from the base of the heart proximal to the pulmonary artery and drains into a cardiac LN in the right mediastinum.7 The primary function of cardiac lymphatics is to prevent myocardial edema through the drainage of interstitial fluid.

Previous reports have shown that the density of LYVE1⁺ and PROX1⁺ LVs significantly decreased one month posttransplant, and patients with at least one rejection episode of International Society for Heart and Lung Transplantation (ISHLT) grade IIIa or higher had significantly lower density of VEGF3 posttransplant.8 Efforts to target LVs have yielded new insight into our understanding of their role in heart transplantation. A role for VEGFR3 in immune cell trafficking from peripheral tissues to SLOs via lymphatic CCL21 production has been demonstrated.9 LVs in the allograft expressed VEGFR3, contained APCs, and produced the dendritic cell (DC) chemoattractant CCL21.9 Treatment of transplant recipients with adenoviral-expressed anti-VEGFR3 immunoglobulin (Ad.VEGFR3-Ig) significantly prolonged rat cardiac allograft survival. Ad.VEGFR3-Ig therapy did not affect allograft lymphangiogenesis, but rather, reduced CCL21 production and CD8⁺ effector T cell entry into allografts, concomitant with reduced DC recruitment and increased forkhead box P3 transcription factor (FOXP3) expression (in the spleen), suggesting that targeting VEGFR3 signaling in lymphatics perturbed their immunomodulatory function, rather than their growth. In a complementary approach, a neutralizing anti-VEGFR3 antibody reduced arteriosclerosis, the number of activated lymphatics expressing VEGFR3 and CCL21, and graft-infiltrating CD4⁺ T cells in chronically rejecting mouse cardiac allografts.10

Cardiac allograft ischemia-reperfusion injury (IRI) associated with transplantation also results in increased VEGF-C/VEGFR3 signaling, leading to increased cardiac lymphangiogenesis.10 This led to LV activation, increasing inflammation within cardiac allografts, whereas pharmacological inhibition of VEGF-C/VEGFR3 reduced both acute and chronic cardiac allograft rejection10 (Figure 2A). In line with the immunomodulatory role of LVs and lymphangiogenesis in heart transplantation, a recent report demonstrated that lymphatic flow is significantly increased in murine heart allografts.11 Of note, this increased lymphatic flow was correlated with an increase in the number of donor-derived cells in the mediastinal draining LNs and an increase in cardiac allograft vasculopathy (CAV), with increased graft infiltration by CD4⁺ T cells, CD8⁺ T cells, and CD68⁺ macrophages.11 Along these lines, targeted delivery of immunotherapeutics to the LNs prolonged cardiac allograft survival in a mouse model, suggesting the transport of immune cells (and antigens) via the lymphatics to the draining LNs exacerbates cardiac allograft rejection.12

Together, cardiac lymphatics remain an untapped avenue for therapeutic intervention in cardiovascular health and disease.7 Efforts primarily targeting VEGFR3 signaling pathways have demonstrated improvements in reverse cholesterol transport, coronary artery disease, myocardial infarction, and cardiac allograft survival (reviewed in Brakenhielm et al7). Overall, current evidence supports the notion that the role of LVs is deleterious in the context of acute and chronic rejection in cardiac transplantation.

2.2 Kidney transplantation

Renal lymphatics originate from the renal lobule and follow the main arteries or veins toward the hilum.13 The intrarenal lymphatic system originates from the superficial network, which lines the fibrous capsule. These LVs lack valves, which allows for bidirectional lymph flow, and drain either directly to the hilum, or connect to the deeper cortical lymphatic capillaries that run along the capsular arterial branches.13 Renal lymph is formed from the interstitial fluid in the cortex, contains capillary filtrate and fluid reabsorbed from the renal tubules, and is regulated by interstitial fluid volume and intrarenal venous pressure.13 The renal lymphatics play an important role in controlling the pressure, volume, and protein content of the interstitial fluid and also participate in regulating inflammation.

Lymphoceles, the accumulation of lymph without an apparent epithelial lining, has long been associated with renal transplantation. The incidence of lymphocele posttransplant has been reported to be between 0.03% and 33.9%; however, symptomatic ranges are lower (0.03%-26%).14-16 Kidney transplants undergoing acute rejection have a significant risk for the development of lymphoceles.17 During transplant recipient preparation of the recipient for implantation, the iliac artery and vein are exposed, and in this process, LNs surrounding these vessels are often damaged. In addition, surgical damage of LVs in the donor kidney occurs during procurement. Together, these may both contribute to the overall disruption of the renal lymphatic system, although the implications or long-term clinical consequences remain to be determined.

De novo lymphangiogenesis has been reported in 61%-74% of renal protocol biopsies from kidney transplant patients, although there was no significant difference in LV density in acute cellular rejection versus those below the threshold for acute rejection or chronic allograft nephropathy.18 Sequential biopsies in human kidney transplants revealed that grafts with higher lymphatic densities within nodular areas of cellular infiltrates (resembling TLOs) had improved kidney functional outcomes compared to those with lower LV density 1 year posttransplant.18 Another study demonstrated kidney transplants with higher perivascular lymphatic density had a reduced proportion with progression of interstitial fibrosis/tubular atrophy grade from 3-12 months posttransplant and was associated with a reduced rate of estimated glomerular filtration rate decline after 12 months.19 In line with this, VEGF-C overexpression in mouse kidney allografts significantly attenuated acute rejection20 (Figure 2B). Overall, perivascular lymphatic density or lymphatic density in proximity to TLOs may be the best indicators of early outcomes in kidney transplants, in part, related to their ability to provide a route for immune cell emigration.

In chronically rejected human renal grafts, concomitant lymphangiogenesis and ectopic germinal center formation has been reported.18 Another report demonstrated that high LV density with nodular infiltrates is associated with graft loss,21 suggesting that in chronically rejecting kidney transplants, the inability to clear immune cell infiltrates may occur despite the high density of LVs, which may be a reflection of LV dysfunction rather than lack of (re)growth. In a rat kidney transplants, lymphangiogenesis is associated with chronic renal allograft injury, and the mechanistic target of rapamycin (mTOR) inhibitor sirolimus was shown to be a potent inhibitor of lymphangiogenesis.22 In rat kidney transplants immunosuppressed with cyclosporine, there was an 18-fold increase in the number of interstitial LVs with no change in the number of perivascular LVs.23 This increase in interstitial LVs was correlated with the extent of interstitial fibrosis,23 which mirrors previous reports of de novo tubulointerstitial lymphangiogenesis in failing human kidney transplants.24

Whereas some reports demonstrate lymphangiogenesis associated with acute and chronic rejection, recent experimental evidence demonstrates that enhancing lymphangiogenesis may improve acute transplant survival.20 One variable that may be considered in comparing differences between reports is the use of immunosuppressive agents. For example, chronic lymphedema has been reported in renal transplant recipients given sirolimus,25 whereas observations in many animal models did not utilize immunosuppressants.20 Another observation that can be inferred is that in renal allografts, LV density alone reflect outcomes, and further investigation of the proximity and interaction of LVs with inflammatory cell infiltrates may yield more conclusive insights. In a model of renal injury and fibrosis, lymphangiogenesis within the kidney and the corresponding draining renal LNs promoted intrarenal inflammation and fibrosis,26 whereas in another model, enhancing renal lymphatic expansion was shown to prevent hypertension through the reduction of renal immune cell accumulation.27 Clearly the role of lymphangiogenesis in regulating inflammation and immune cell transit is context dependent, and further kidney transplant studies are warranted to determine what conditions may promote a more tolerogenic environment.

2.3 Lung transplantation

The pulmonary lymphatic network is present throughout the pleura, in the perivascular connective tissue around the venules and arterioles, within the juxta-alveolar region of the interlobular septum, as well as within the connective tissue of the terminal and respiratory bronchioles.28 Within the lungs, forward lymph flow is dependent on changes in pressure associated with respiration, as lung collecting LVs lack supporting smooth muscle cells.29 In a mouse model of lung transplantation, donor LVs have been reported to begin sprouting toward the recipient within 3 days posttransplant, establishing functional connections by 14 days posttransplant.30 Of note, lymph flow in the pulmonary system flows in two opposite directions: in the periphery of the lung, lymph flows toward the pleura, whereas in other regions, lymph flows toward the hilum.28

In one study, 56 of 99 lung transplant recipients (57%) experienced noncardiogenic pulmonary edema (or pulmonary reimplantation response) in the first week posttransplant.31 Although the reimplantation response (referred to as primary graft dysfunction in more recent literature) is generally thought to be a consequence of IRI, it remains unclear whether lymphatic disruption contributes to this condition. Although edema is a hallmark of lymphatic dysfunction, and is to be expected considering the interruption of the lymphatic system during transplantation, the fact that many patients do not develop a reimplantation response suggests that lymphatic disruption alone is not sufficient to promote acute graft injury. Pulmonary edema in the context of a reimplantation generally resolves within a few days with supportive care, and it remains unclear whether effects persist as a result of this lymphatic perturbation.

Experimentally, pulmonary lymphatic flow and lymphatic drainage were assessed after lung transplantation in dogs and demonstrated a lack of visible drainage to the mediastinal LNs 3 days posttransplant in three of the four dogs investigated32 (Figure 2C). In lung allografts treated with immunosuppression, lymphatic drainage was reestablished as early as 2 weeks posttransplant.32 Acute rejection was associated with the cessation of lymphatic drainage from the transplanted lung to the mediastinal LNs.32 Recent experimental work in murine lung allografts showed that stimulating lymphangiogenesis alleviates rejection, in part, through improving the clearance of HA,33 whereas disruption of lymphatic structures and function in the transplanted lung lead to impaired lymphatic flow and perturbed graft function.29

Chronic airway inflammation predisposes to persistent lymphatic hyperplasia, which may lead to bronchial (lymph)edema and airway obstruction.34 Disruptions in the lymphatic system are also seen in a variety of pulmonary disorders, including increased lymphangiogenesis in pneumonia, regression of LVs in asthma, remodeling, and growth of LVs in idiopathic pulmonary fibrosis, among others (reviewed in Yao & McDonald34). It has been demonstrated that rapamycin can reverse VEGF-C-driven lymphatic abnormalities in the respiratory tract,35 suggesting that the use of mTOR-based immunosuppressants in the context of lung transplantation may contribute to modified lymphatic phenotypes. However, due to concerns with regard to bronchial anastomotic healing, rapamycin use early after lung transplantation may not be advisable.36

Overall, in apparent contrast to hearts, LVs in the lung may serve a beneficial role, in part, through facilitating pulmonary drainage and immune cell transport early after transplantation. An important consideration is that lung grafts, unlike hearts, are able to locally provide a suitable environment for the activation of alloimmune responses, and therefore, do not have a requirement for lymphatic flow to draining LNs to trigger rejection.37 As HA can act as a damage-associated molecular pattern (DAMP), promoting inflammation through activation of TLR4, leading to immune cell activation,38 clearance of this may be central to limiting early alloimmune responses. Interestingly, induction of BALT, a TLO that has been characterized by formation of high endothelial venules (HEVs) and LVs, has been observed in lung allografts that undergo chronic rejection, and has also been associated with a tolerant state.39, 40 Therefore, future studies will have to further elucidate how lymphangiogenesis shapes the fate of pulmonary allografts.

4.1 Calcineurin inhibitors

Calcineurin inhibitors are the primary mode of immunosuppression currently used. The calcineurin inhibitor tacrolimus (FK506) inhibits T cell activation through complexing with FKBP12 (FK506 binding protein 12), inhibiting calcineurin signaling, ultimately preventing the dephosphorylation of nuclear factor of activated T cells (NFAT) and the transcription of interleukin-2 (IL2).49 NFATc1 signaling has been shown to regulate LEC development, with inhibition resulting in irregular patterning of lymphatic sprouts, in part, through the reduction of PDPN, VEGFR3, and fibroblast growth factor receptor 3 (FGFR3) expression50 (Figure 3B). In particular, decreased calcineurin activity diminished VEGF-A-induced pulmonary lymphangiogenesis.50 NFATc1 signaling also directly interacts with forkhead box protein C2 (FOXC2) on enhancer elements to promote FOXC2-target gene expression, which is essential in multiple aspects of developmental and pathological lymphangiogenesis, including the regulation of lymphatic valve formation.51 In T cells, calcineurin/NFAT signaling has been shown to be required for the switch from catabolic to anabolic metabolism after T-cell receptor and co-stimulation.52 As LEC metabolism has recently been shown to be important in lymphangiogenesis and lymphatic function,53, 54 it would be of interest to investigate the metabolic consequences of tacrolimus on LECs.

Interestingly, application of topical tacrolimus improved LV function with increased contraction frequency and decreased dermal backflow and increased collateral LV formation in the context of secondary lymphedema.55 In this context, the effects were not due to increased expression of VEGF-C or VEGF-A but rather occurred through the suppression of transforming growth factor beta 1 (TGFβ1), interferon gamma (IFNγ), and IL4 and IL13 production.55

Although it is clear tacrolimus can have both direct and indirect effects on lymphangiogenesis and lymphatic function, what remains unclear are the context dependent effects of tacrolimus in solid organ transplantation and the organ-specific LEC and stromal cell responses. Additional investigation into the effects of calcineurin inhibition on LVs is warranted to determine whether tacrolimus may result in perturbation of lymphangiogenesis or LV function and whether counteracting these effects may improve transplant outcomes.

4.2 mTOR inhibitors

mTOR is a central signaling molecule, regulating signaling from insulin, growth factors, and amino acids but also acts as a sensor of nutrients, oxygen, and energy. Rapamycin inhibits mTOR signaling through interacting with FKBP12, resulting in the inhibition of mechanistic target of rapamycin complex 1 (mTORC1) and mTORC2 (Figure 3C). Inhibition of mTOR impairs lymphangiogenesis through inhibition of VEGF-A and VEGF-C signaling via p70S6 kinase.56 Further, rapamycin has been shown to inhibit LEC tube formation via downregulation of VEGFR3.57 In generalized lymphatic anomalies with phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha (PIK3CA) variants, mTOR inhibition with rapamycin prevented lymphatic hyperplasia and dysfunction in mice expressing an active form of PIK3CA (His1047Arg).58

In rat renal allografts, sirolimus inhibited lymphangiogenesis, leading to improved graft function and a reduction of chronic kidney allograft injury compared to use of the calcineurin inhibitor cyclosporine.22 Indeed, chronic lymphedema has been reported in renal transplant patients attributable to sirolimus in the same location where the arteriovenous fistula used for dialysis occurred.25 Further study of the effects of mTOR inhibitors on lymphangiogenesis and lymphatic function in the context of transplantation is warranted to determine the effects of mTOR inhibition in nonrenal transplants and whether use of mTOR inhibitors in place of calcineurin inhibitors alters effects on LVs and graft outcome.