Lymphangiogenesis, Myeloid Cells and Inflammation

Lianping Xing; Rui-Cheng Ji


Expert Rev Clin Immunol. 2008;4(5):599-613. 

In This Article

Lymphatic-associated Inflammatory Disorders

The cornea is normally devoid of blood and lymphatic vessels; however, inflammation induces lymphangiogenesis in the corneal stroma, usually accompanied by hemangiogenesis.[102] For this reason, corneal inflammation is the most commonly used animal model to investigate the involvement of various factors in lymphangiogenesis.[35,103,104,105,106,107] Several reports indicate that lymphangiogenesis in the inflamed cornea is related to the VEGF-C/VEGFR-3 signaling pathway. The expression of VEGF-C/VEGFR-3 mRNA was dramatically upregulated 3days after the injury but gradually decreased thereafter.[108] The source of VEGF-C probably comes from bone marrow-derived CD11b+ myeloid cells that are recruited to the inflamed corneal stroma. Immunophenotypic analysis using combinations of cell surface markers demonstrates that these CD11b+ cells are composed of immature DCs characterized by CD45+CD11b+CD11c+ and, mostly, MHC classII-CD80-CD86-[104] and CD45+CD11b+CD11c- macrophages.

The involvement of macrophages suggests that therapeutic prevention of macrophage activation might prolong corneal graft survival. Furthermore, the blockade of VEGFR-3 signaling may suppress trafficking of corneal DCs to draining LNs and inhibit induction of delayed-type hypersensitivity and rejection of corneal transplants.[109] The expression of VEGF-C/VEGFR-3 on tissue DCs and macrophages implicates a novel potential relationship between leukocyte trafficking, immune cells and lymphangiogenic signals in the eye. Lymphangiogenesis may decrease the success rate of keratoplasty in the vascularized cornea by accelerating antigen recognition and the graft rejection reaction. The inhibition of newly formed lymphatics will, at least in part, improve the outcome of keratoplasty.

Wound healing involves a complex interplay among cells, growth factors and cytokines. The cascade of events that leads to wound healing begins with clotting and the recruitment of inflammatory cells, such as neutrophils and macrophages, mitosis of wound cells, synthesis of extracellular matrix components, angiogenesis and lymphangiogenesis.

The cutaneous healing model is convenient and effective for studying lymphangiogenesis and its evolution with pathological changes. VEGFR-3+ lymphatic vessels are observed in the granulation tissue from day 5 in wounds made in the dorsal skin of pigs, similar to in chronic inflammatory wounds of humans. The newly formed lymphatic vessels are distinct from the PAL-E/laminin/vWF+ vessels, are fewer in number and appear to sprout from pre-existing LECs at the wound edge.[102] This observation has been demonstrated in mouse skin wounds using 5'-Nase cerium and VEGFR-3 immunohistochemical staining. The development of lymphatic vasculature changes at different stages, from the lymphatic-like structures to the newly formed lymphatic vessels with an extremely thin and indented wall. LECs are eventually featured by typical intercellular junctions, which are stained positively with anti-VEGFR-3 antibody and 5'-Nase cerium but are negative for VEGF-C.[103] In parallel with angiogenesis, VEGFR-3+ lymphatic vessels disappear after day 14, suggesting that the lymphangiogenic process is transient in healing wounds.[102]

Transgenic mice overexpressing VEGF-A in the skin under the keratin-14 promoter have more extensive lymphatic vessel formation in response to acute ultraviolet B (UVB) irradiation.[110,111] VEGF-A-induced lymphatic vessels are greatly enlarged with incompetent valves, sluggish flow and delayed lymph clearance, indicating that these vessels do not function properly. This impairment can be prevented by the systemic blockade of VEGF-A signaling.[112] In contrast to the beneficial effect of anti-VEGF-A, the systemic blockade of VEGF-C signaling via a VEGFR-3 neutralizing antibody delays the recovery phase of UVB irradiation-induced inflammation and edema. Anti-VEGFR-3 leads to enlargement of LYVE-1+ lymphatic vessels at day 9 after UVB irradiation, the time when lymphatic vessels have collapsed and returned to a baseline level in IgG treatment control mice.[113] As VEGF-A does not bind to VEGFR-3 and the number of CD31+ blood vessels is similar between treated and control mice, the effect observed in anti-VEGFR-3 treatment is likely to be specifically due to inhibition of lymphatic function. Thus, promotion of lymphatic function might represent a new strategy to reduce or prevent UVB-induced skin damage.[113]

Impaired wound healing is a common complication of diabetes, which is primarily due to impaired inflammatory cell function and decreased secretion of cytokines/growth factors along with a prolonged inflammatory phase.[114] Both LYVE-1+ lymphatic vessels and CD31+ blood vessels are significantly reduced in corneal wound healing in diabetic mice. The overexpression of VEGF-C improves the healing of wounds by enhanced angiogenesis and lymphangiogenesis. The blockage of VEGF-C signaling with a VEGFR-3 inhibitor severely delays the wound closure.[115]

Macrophages may play an important role in lymphangiogenesis in the skin wound healing process because numerous F4/80 and LYVE-1 or podoplanin double-positive cells are observed in full-thickness wounds. Glucose treatment of control macrophages decreases the expression levels of VEGF-C and VEGFR3. IL-1 increases VEGF-C and VEGFR-3 mRNA in macrophages and the application of IL-1-treated diabetic macrophages to diabetic mice induces lymphatic vessel formation and accelerated wound healing.[57] However, the origin of the cells involved in the formation of lymphatic vessels in wounds, the path of their trafficking and regulation, and their relationship to F4/80 or CD11b+ blood endothelial progenitor cells has not been investigated.

These studies indicate strongly that adequate functional lymphatic vasculature is beneficial for wound healing in basal or diseased conditions, such as diabetics, raising a possibility that improvement of lymphatic functions by overexpressing VEGF-C and/or modifying macrophages could be a promising new therapy for would healing. Thus, lymphatic vessels in wounds function to maintain normal tissue pressure by draining the protein-rich lymph from the interstitial space, as well as by facilitating the delivery of cells that mediate the immune response.[116]

In human kidney transplants undergoing rejection, the proliferating podoplanin-positive lymphatic vessels are accompanied by mononuclear infiltrates and contain the entire repertoire of T and B lymphocytes to provide specific cellular and humoral alloantigenic immune responses during the inflammatory process.[47] The density of lymphatic vessels is significantly higher in areas with cellular infiltrates in biopsies from patients who received a kidney transplant. Graft function at 1year after transplantation is better in patients who have lymphatic vessels in their infiltrates compared with those whose cellular infiltrates do not contain lymphatic vessels, implying that lymphangiogenesis might have an impact on the pathogenicity of these cellular infiltrates.[117] The massive lymphangiogenesis in the transplant parenchyma may be driven by VEGF-C-expressing macrophages, providing exit routes for lymphocytes and macrophages.[56] Furthermore, the newly formed lymphatic vasculature network may also affect the survival of the transplant by organizing the perivascular lymphocytes into immunologically active follicular structures around the graft.[117]

Lymphatic complications are common side effects of mammalian targeting of rapamycin inhibitor-based immunosuppression in kidney transplantation. Rapamycin impairs the recovery of lymphatic flow across surgical incisions, resulting in prolonged wound edema in a murine skin-flap model, and potently inhibits VEGF-C-mediated LEC proliferation and migration.[118] This finding possibly warns against the utilization of a rapamycin inhibitor immediately after tissue injury because lymphangiogenesis may contribute to the export of rejection infiltrates and the maintenance of potentially detrimental alloreactive immune responses in transplants. Manipulation of lymphangiogenesis and lymphatic function will present novel chemotherapeutic or gene therapy opportunities in this field.

Chronic inflammation in the respiratory tract is associated with dramatic architectural changes in the walls and vasculature of the airways. Increased lymphangiogenesis was first recognized in a mouse model of chronic respiratory tract infection induced by tracheal administration of Mycoplasma pulmonis.[48] Angiogenesis starts 2days after infection while lymphangiogenesis begins from day 7 and is associated with increased size of local draining LNs. Anti-VEGFR-3 treatment prevents lymphangiogenesis in the lung and the enlargement of draining LNs, but increases the mucosal edema. Interestingly, newly formed lymphatic vessels persist after antibiotic treatment when inflammation declines. These findings indicate that lymphangiogenesis in both the infection area and the draining LNs is driven by the VEGF-C/-D-mediated signaling pathway, that VEGFR-3 inhibition may result in severe exacerbation of mucosal edema[28] and that remodeling of lymphatic vessels differs from the blood vessels, which regress when inflammation stops. Similar to other inflammatory models, macrophages are considered a source of VEGF-C. A recent study reported that blood endothelial cells from patients with cystic fibrosis, a common genetic disorder characterized by severe lung inflammation and fibrosis, produce high levels of VEGF-A, VEGF-C and basic FGF, suggesting that other cell types may also stimulate angiogenesis and lymphangiogenesis in inflammation.[119]

T-cell-dependent humoral immunity to a persistent airway infection may be a novel pathway to induce inflammatory lymphangiogenesis and chronic airway pathology. A potential relationship between the inflammatory reaction and lymphangiogenesis has been indicated in the rat remnant kidney that is an established model of renal tubulointerstitial fibrosis and progresses to end-stage renal failure. Fibrotic tubulointerstitial regions have massive LEC proliferation and relatively high levels of VEGF-C mRNA.[120] As LECs secrete chemokines that attract DCs, it is possible that increased lymphatics could enhance the immunologic surveillance of remnant kidneys. In cirrhotic livers, collective expression changes observed within some functional groups of genes have suggested that endothelial cells may contribute to lymphangiogenesis, enhancement of fibrogenesis and inflammatory processes and changes in cell-cell interactions with upregulation of adherens junction proteins.[109] In the orbital fat of patients with mucormycosis and panendophthalmitis, significant lymphatic proliferation was present within granulation tissue associated with the acute inflammation, which is normally devoid of lymphatic channels.[121]

Thus, an expanded lymphatic network provides conduits for extravasated fluid from leaky blood vessels and activated antigen-presenting cells to local LNs. When lymphangiogenesis is impaired or inhibited, airway inflammation may lead to bronchial lymphedema and exaggerate airflow obstruction. Transfering immune serum to B-cell-deficient mice demonstrates that inflammatory cells recruited to the infected airways are activated by the humoral response and this activation is correlated with the induction of genes for remodeling growth factors, such as VEGF-D.[122]

Joint lesion in inflammatory arthritis is driven by the inflammatory synovial tissue or 'pannus', a hyperplastic, locally invasive tissue composed of various cell types, including synovial cells and inflammatory cells. These cells produce a vast array of inflammatory mediators, including cytokines and chemokines that destroy the extracellular matrix in the joint by direct and indirect mechanisms. The pannus is extremely vascular, providing portals of entry for effector cells to enter the joint from the circulation and mediate joint destruction through autocrine and paracrine mechanisms. As a result of increased angiogenesis, edema develops in the pannus due to the accumulation of fluid that has leaked from newly formed blood vessels. Using anti-LYVE-1, podoplanin or other antibodies to identify LECs in joint sections from patients with RA and OA, an increased lymphatic vasculature network is reported in both RA and OA samples, but RA specimens have significantly more lymphatic vessels.[51,123]

We observed a marked increase in the number and size of lymphatic vessels in the pannus of joints of TNF-Tg arthritic mice compared with those of normal joints. CD11b+/Gr-1-/lo osteoclast precursors from joints of these mice express podoplanin protein on their surface.[55] Many VEGF-C+ cells are observed in pannus tissues. VEGF-C protein levels are high (0.01-0.8 ng/ml) in the synovial fluid of RA patients and this is positively correlated with TNF and IL-1 levels in the same samples.[124] Similar to the lymphatic vessel histology, synovial fluid of OA patients also has increased VEGF-C protein but at a lower level. Sources of VEGF-C-producing cells in individuals with arthritis include synovial cells, macrophages, DCs and osteoclasts and their precursors. In invitro cultures, TNF stimulates VEGF-C expression twofold in synovial cells[124] and three- to fourfold in macrophages and osteoclast precursors.[125]

We found that the canonical NF-κB signal mediates TNF- and RANKL-induced VEGF-C expression through the κB binding element in the VEGF-C promoter.[75] These cytokines promote the binding of NF-κB p65 and p50 proteins directly to the intact VEGF-C promoter in osteoclast precursors. NF-κB deletion or inhibiton blocks cytokine-induced VEGF-C expression. NF-κB-mediated VEGF-C expression is observed in human breast cancer cells.[126] These findings suggest that NF-κB may play an important role in inflammation-induced lymphangiogenesis, which may also apply in other cell types,such as DCs. DCs express RANKL receptor on their cell surface and RANKL may stimulate DCs to produce VEGF-C through a similar mechanism to that which occurs in osteoclast precursors.[49] Apart from the involvement of cytokine-induced VEGF-C production, the precise role of NF-κB in inflammatory lymphangiogenesis has not been investigated in detail. Similarly, little is known regarding the role of NF-κB in LEC biology, except in I-κB promoter-driven LacZ mice; lymphatic vessels are also stained positive for LacZ staining.[127] Given the fact that no global edema is reported in mice deficientin any single NF-κB member, including p65, p105/050, p100/p52, RelB and c-Rel,[128,129,130,131] more studies are required to elucidate the role of NF-κB in physiological and pathological lymphangiogenesis.

Interesting, anti-TNF treatment of TNF-Tg mice inhibits the joint synovitis and reduces the lumen diameters of lymphatic vessels, but the number of lymphatic vessels does not change.[94,125] This observation is consistent with that seen in chronic airway inflammation, in which antibiotic treatment reduces inflammation but does not restore an increased lymphatic vasculature network.[48] Another study compared synovial lymphangiogenesis in synovial specimens from patients with RA and spondyloarthritis before and after anti-TNF treatment. Interestingly, TNF blockade reduces joint inflammation but increases the numbers of lymphatic vessels and, probably, VEGF-C-expressing cells.[132] The mechanism responsible for increased lymphangiogenesis after inflammation resolution is not currently clear. These preclinical and clinical data suggest that inflammation is a prerequisite for the initial formation of new lymphatic vessels in synovial tissues. These vessels drain excessive fluid to the local draining LNs, resulting in dilation of the vessel lumen. It is likely that the resolution of inflammation does not affect the presence of the vessels but significantly reduces the amount of lymph drained, leaving vessels with narrow lumens. Another possibility is that some factors produced during the inflammatory process limit inflammation-induced lymphangiogenesis, resulting in insufficient lymphatic drainage, as show in Figure1. If this is the case, improvement of lymphatic function, such as via intra-articular administration of VEGF-C, may be a new therapy in the treatment of chronic inflammatory arthritis.

Hashimoto's thyroiditis is a condition caused by inflammation of the thyroid gland. It is an autoimmune disease associated with the infiltration of lymphocytes into the thyroid gland and the formation of LN-like structures. Lymphatic vessels are present within these infiltrates. Mice that overexpress the chemokine CCL21 in the thyroid develop similar lymphoid infiltrates and lymphatic vessels in their thyroid gland. Mature T and Bcells are required for lymphatic vessel formation. Deletion of lymphotoxin-β receptor or lymphotoxin-α abrogated development of lymphatic vessels in the inflamed areas in the thyroid but did not affect the development of neighboring lymphatics, implicating the involvement of lymphotoxin β-receptor signaling in this process.[133]

Tissue damage by mechanical injury and inflammation is followed by a reaction in the regional lymphoid tissue, lymphatics and LNs. Lymphoscintigraphic studies demonstrated that closed fractures of a lower limb cause a reaction of the local lymphoid tissue. There is dilation of the lymphatics draining the site of the fracture and enlargement of inguinal LNs. These changes persist even after clinical healing of the fracture. In the long-lasting nonhealing fractures, the lymphoscintigraphic pictures show impaired lymphatic drainage and the disappearance of LNs.[134] A combination of the lymphoscintigraphic images and immunohistochemical observations found that normal fracture healing with immune cell infiltrates and foci of ossification is accompanied by dilated lymphatic vessels and enlarged LNs. A prolonged nonhealing fracture lacking a cellular reaction proceeds with decreased LN size, indicating the existence of a functional axis between the bone wound and surrounding soft tissue and the local lymphatic (immune) system. Fast healing is regulated by an influx of LN regulatory cells into the wound; whereas, prolonged healing causes gradual exhaustion of the regional LN functional elements and reciprocal impairment in sending regulatory cells to the fracture gap.[134]

No lymphatic vessels are detected in normal bones or the bone marrow cavity using immunostaining with anti-LYVE-1 and podoplanin antibodies.[75,135] However, in the bone-implant interface of patients with aseptically loosened prosthetic joints, relatively few podoplanin-positive lymphatic vessels are identified, which localize in the pseudocapsule region.[136] Furthermore, wear particles and wear particle-containing macrophages are found in regional LNs draining arthroplasty tissues. Thus, the wear particles shed at the bone-implant interface are transported to draining LNs via lymphatics from the pseudocapsule, rather than directly from the pseudomembrane.[137] Currently, the functional importance by which wear particles are transported from arthroplasty tissues to LNs is uncertain. The relative absence of a lymphatic clearance mechanism may contribute to the accumulation of wear particles at the bone-implant interface and promote periprosthetic osteolysis through the stimulation of osteoclast formation and activity.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: