Emerging roles of podoplanin in vascular development and homeostasis

Abstract

Podoplanin (PDPN) is a mucin-type O-glycoprotein expressed in diverse cell types, such as lymphatic endothelial cells (LECs) in the vascular system and fibroblastic reticular cells (FRCs) in lymph nodes. PDPN on LECs or FRCs activates CLEC-2 in platelets, triggering platelet activation and/or aggregation through downstream signaling events, including activation of Syk kinase. This mechanism is required to initiate and maintain separation of blood and lymphatic vessels and to stabilize high endothelial venule integrity within lymph nodes. In the vascular system, normal expression of PDPN at the LEC surface requires transcriptional activation of Pdpn by Prox1 and modification of PDPN with core 1-derived O-glycans. This review provides a comprehensive overview of the roles of PDPN in vascular development and lymphoid organ maintenance and discusses the mechanisms that regulate PDPN expression related to its function.

Keywords

podoplanin / CLEC-2 / Prox1 / O-glycosylation / lymphatic vascular development and maintenance / lymphoid organ homeostasis

Introduction

Podoplanin (PDPN), also known as gp38, aggrus, and T1α, is a ~38 kDa type-1 transmembrane sialomucin-type O-glycoprotein consisting of 172 and 162 amino acids in mice and humans, respectively. PDPN homologs exist in other mammals, such as rats, dogs, and hamsters, and is relatively well-conserved across species. PDPN has an extracellular domain, a single transmembrane domain, and a short cytoplasmic tail, which has been shown to interact with members of ERM (ezrin, radixin, moesin) proteins in epithelial cells to activate RhoA and promote cell trans-differentiation [ 1]. PDPN is widely expressed in type I alveolar cells, kidney podocytes, osteoblasts, fibroblastic reticular cells, several tumor cell types, and lymphatic endothelial cells in vascular endothelium [ 1− 5].

CLEC-2, the widely-recognized receptor for PDPN, is a 32−40 kDa type 2 transmembrane C-type lectin-like protein expressed extensively in various cells, including platelets, megakaryocytic cell lines [ 6], liver sinusoidal endothelial cells [ 7], and liver Kupffer cells [ 8]. CLEC-2 has one YXXL motif in its cytoplasmic tail, which resembles the immunoreceptor tyrosine-based activation motif and undergoes tyrosine phosphorylation by Src kinases upon activation [ 6, 9]. Tyrosine kinase Syk binds to the phosphorylated YXXL of CLEC-2 with its tandem Src homology 2 domains, resulting in Syk activation and downstream signaling events, including tyrosine phosphorylation of LAT, SLP-76, Vav1/3, Btk, and phospholipase Cγ2 [ 6, 9].

Recent studies have shown that PDPN in lymphatic endothelial cells or fibroblastic reticular cells activates platelet CLEC-2, which induces downstream signaling events, including activation of the Syk tyrosine kinase and, ultimately, platelet activation and/or aggregation [ 10, 11]. This mechanism is important for the separation of blood and lymphatic vessels and stability of high endothelial venule integrity within lymph nodes. In this review, the functions of PDPN in the development of the lymphatic vascular system and maintenance of lymph node homeostasis as well as the regulation of PDPN expression related to its function are discussed.

Lymphangiogenesis and lymphatic endothelial cell lineage identity

The lymphatic vascular system is one of two circulatory systems present in vertebrates; it is closely related to the blood vascular system in terms of origin, morphogenesis, and the regulatory molecules required for its development and growth. The lymphatic system develops in parallel and secondarily to the blood vascular system through lymphangiogenesis [ 12, 13]. As revealed by lineage tracing, lymphatic endothelial cells (LECs) are differentiated entirely from venous endothelial cells in a process controlled by the transcription factor prospero homeobox 1 (Prox1). Migrated LECs then form primary lymph sacs from which LECs germinate, proliferate, and migrate to give rise to the entire lymphatic vascular network. Vascular endothelial growth factor-3 (VEGF-3) is a critical regulator for lymphatic vessel growth, acting via its receptor, VEGFR-3, which is primarily found on LEC [ 12, 14] (Fig. 1).

Initial differentiation of the LEC lineage begins from venous ECs [ 15]. Through yet-unknown mechanisms, a subset of venous endothelial cells starts to become committed to the lymphatic endothelial fate at embryonic day (E) E8.5 (Fig. 1). These cells express the transcription factor Sox18, which directly activates Prox1 by binding to its proximal promoter. Transcription factor Prox1 then serves as master regulator of LEC identity during embryonic development and postnatal remodeling [ 16− 19]. Besides Prox1, differentiating LECs at E9.5 also express lymphatic endothelial hyaluronan receptor-1 (Lyve-1) and VEGFR-3. Starting at E10.5, PDPN is expressed on the LEC surface. LECs, stimulated by VEGF-C secreted from adjacent tissue, migrate and proliferate to form primary lymph sacs around E11.5. During this step, the lymphatic and venous systems become separated via a process that involves PDPN on the LEC surface and its ligand CLEC-2 on the platelet surface [ 6, 10]. PDPN activates CLEC-2, thus activating the hematopoietic protein Syk/Slp-76 signaling pathway and platelet aggregation. Aggregated platelets “seal off” the lymph sacs from cardinal veins (Fig. 2) [ 10, 20]. Finally, lymphatic vessel maturation and remodeling occur in a stepwise manner, leading to the formation of the complete lymphatic network.

Transcriptional regulation of PDPN by Prox1 in lymphangiogenesis

Mouse Prox1 was named as such because of its homology to the Drosophila homeobox protein prospero [ 21]. During mouse embryogenesis, Prox1 is expressed in the developing central nervous system (CNS), eye lens, pancreas, liver, skeletal muscles, and heart. Mice with tamoxifen-induced deficiency of Prox1 (inducible Prox1^−/−) during the embryonic, postnatal, or adult stages exhibit interconnected blood and lymphatic vessels and show de-differentiation of LECs into BECs; reduced expression of LEC markers (VEGFR-3, PDPN, Lyve-1, etc.) and the appearance of BEC markers (endoglin, CD34, etc.) may also be notably observed [ 16, 17, 22, 23]. These findings indicate that LEC differentiation is tightly regulated and that the differentiated LEC phenotype is a plastic and dynamic condition that depends on constant Prox1 activity for maintenance. However, downstream target genes of Prox1, through which it controls the switch from BECs to LECs and maintains LEC identity, remain to be identified. In vitro, Prox1 interacts with transcription factor COUP-TFII to regulate expression of several lineage genes, including VEGFR-3, fibroblast growth factor receptor 3 (FGFR-3), and neuropilin 1 [ 23]. However, phenotypes of mice lacking VEGFR3, FGFR-3, or neuropilin 1 do not resemble the lymphatic defects in inducible Prox-1^−/− mice [ 24− 27]. A number of in vitro results suggest that forkhead transcription factor Foxc2 may be a downstream target of Prox1 in LECs [ 28]. However, Foxc2 was recently shown to be regulated by transcription factor NFATc1 and is not involved in LEC differentiation [ 29].

When Prox1 was adenovirally transduced into human dermal microvascular endothelial cells, expression of PDPN was upregulated. On the other hand, siRNA-mediated knockdown of Prox1 in primary human LECs caused the loss of PDPN [ 16, 17, 22]. However, siRNA-mediated knockdown of PDPN does not affect expression of Prox1 in LEC [ 30]. Mice with global deletion of PDPN (Pdpn^−/−) exhibit misconnected blood and lymphatic vessels that closely resemble those observed in inducible Prox1^−/− mice [ 31]. Both Prox-1 and PDPN are also constantly expressed in LECs naturally and appear to function in an LEC-autonomous manner. Thus, PDPN is hypothesized to function downstream of Prox1, and PDPN expression is regulated by Prox1 in LECs at the transcriptional level. Four putative binding elements for Prox1 in the 5′ upstream regulatory region of the Pdpn gene were first identified. Using chromatin immunoprecipitation assay, Prox1 was observed to directly bind to the 5′ regulatory sequence of the Pdpn gene in LECs. DNA pull down assay confirmed that Prox1 binds to the putative binding element. In addition, luciferase reporter assay indicated that Prox1 binds to the 5′ regulatory sequence of Pdpn to regulate Pdpn gene expression [ 32]. These results suggest that Prox1 regulates PDPN expression at the transcriptional level in the lymphatic vascular system. Thus, PDPN potentially serves as a major downstream target gene of Prox1 in the pathway essential for differentiation and maintenance of lymphatic identity.

However, Prox1 is not expressed in lymph node fibroblastic reticular cells or in many other cell types expressing PDPN. Therefore, alternative pathways must be involved in PDPN expression in tissue other than lymphatic vessels. For example, PDPN is regulated by the AP-1 transcription factor in skin cancers, osteosarcomas, and gliomas [ 33− 35]. A more detailed understanding of transcriptional regulation of PDPN in various cell types and their respective downstream signaling pathways is needed for elucidating the biological functions of PDPN.

Post-translational regulation of the PDPN by core 1 O-glycosylation

Mucin-type O-glycosylation is a prevalent post-translational modification found on membrane and secreted proteins [ 31, 36− 38]. Mucin-type O-glycosylation occurs in the Golgi apparatus via sequential reactions catalyzed by specific glycosyltransferases. The core of all mucin-type O-glycans is serine/threonine-linked N-acetylgalactosamine (GalNAc-Ser/Thr), also known as the Tn antigen, which is normally further modified to form distinct subtypes of O-glycans. The most common form of O-glycan, core 1-derived O-glycans, are synthesized by addition of galactose to the Tn antigen, a reaction catalyzed solely by the core 1 synthase, glycoprotein-N-acetylgalactosamine 3-β-galactosyltransferase (C1galt1, also named T-synthase) [ 37, 39]. Core 1 structure can be further branched to form an extended core 1 or core 2 structure or can be modified by addition of sialic acids. Such glycans are known as core 1-derived O-glycans (Fig. 3A and 3B) [ 37− 39]. Core 1 derived O-glycans are present on many membrane-bound or secreted proteins in most cell types, especially in epithelial cells and endothelial cells [ 38].

The extracellular domain of mouse PDPN contains 27 potential sites of O-glycosylation (Figs. 3 and 4). The molecular weight of unmodified PDPN is about 17 kDa; however, the molecular weight of PDPN isolated from different tissue ranges from 37 kDa to 41 kDa, consistent with extensive O-glycosylation on the serine and threonine residues of its protein backbone [ 5]. Results of in vitro lectin blotting assay further suggested that PDPN is modified primarily by sialylated core 1 O-glycans [ 40, 41].

Homozygous, ubiquitous deletion of the gene encoding core 1 synthase, C1galt1, is embryonically lethal in mice [ 38]. To determine the importance of endothelial O-glycans in vessel development, mice containing C1galt1 flanked by loxP sites, C1galt1^f/f, were crossed with Tie2-Cre transgenic mice, which express Cre recombinase in endothelial and hematopoietic precursors. The cross generated mice deficient in core 1 synthase and thus deficient in core 1 O-glycosylation, specifically in endothelial and hematopoietic cells (previously called EHC T-syn^−/−, now referred to as EHC C1galt1^−/−) [ 31]. EHC C1galt1^−/− mice exhibit neonatal lethality, disorganized lymphatic vessels misconnected with blood vessels, and mosaic expression of blood and lymphatic EC markers in LECs. In addition, PDPN levels are substantially reduced in EHC C1galt1^−/− lymphatic endothelium, and Pdpn^−/− mice exhibited abnormal and blood-filled lymphatic vessels that phenocopy those of EHC C1galt1^−/− mice [ 4, 31]. Inducible C1galt1^−/− mice (C1galt1^f/f, Rosa26-rtTA, tetO-Cre) were generated, and ubiquitous loss of C1galt1 in adult mice showed similar blood-lymph mixing phenotypes as EHC C1galt1^−/− mice [ 31]. Thus, continued expression of PDPN into adulthood is required to maintain proper lymphatic vascular architecture.

Recently, core 1 O-glycan-deficient or desialylated PDPN was observed to be targeted for proteolytic degradation by various proteases, such as metalloproteinases (MMP)-2/9, which are present in the lymph [ 42]. Indeed, core 1 O-glycan-deficient, but not wild-type, endothelial cells showed reduced PDPN surface levels after incubation with lymph. Furthermore, the lymph from mice with sepsis, which contained bacteria-derived sialidase, induced a decrease in PDPN levels even in wild-type endothelial cells. The MMP inhibitor GM6001 prevented decrease in PDPN surface levels in vitro and also prevented decrease in PDPN levels in LECs in EHC C1galt1^−/− mice and in mice with sepsis in vivo. In addition, core 1 O-glycan-deficient or desialylated PDPN reduced interaction between endothelial cells and platelets induced by shear stress [ 42]. Thus, sialylated core 1 O-glycans of PDPN promote platelet adhesion and protect PDPN from proteolytic degradation induced by MMPs in the lymph [ 42].

PDPN-CLEC-2 interaction in development and homeostasis of lymphatic vasculature

Interaction between PDPN and CLEC-2 is stabilized by multiple regions within human PDPN, including three platelet aggregation-stimulating (PLAG) domains, a region between amino acids 80 and 103 that contains potential O-glycosylation sites and a region between amino acids 103 to 128 [ 43]. Within the human PLAG domain, disialyl-core 1 at Thr52 is known to be critical for the association with CLEC-2 that triggers platelet aggregation [ 40, 41]. Thr34 of mouse PDPN corresponds to human PDPN Thr52, and substitution of this residue in mouse PDPN with alanine completely abolishes platelet aggregation in vitro [ 44]. Furthermore, Bianchi et al. [ 45] found that the Thr34Ala mutation in mouse PDPN-Fc reduced CLEC-2 binding and toxicity in vivo while maintaining anti-lymphangiogenic activity both in vitro and in vivo. However, whether O-glycans on other serine/threonine residues of PDPN also promote binding to CLEC-2 and the extent of core 1-derived O-glycosylation on PDPN required for platelet activation are not known.

With respect to CLEC-2, a region between amino acids 87 and 147 has been shown to be critical for interaction with human PDPN, and the three-dimensional structure of the region between amino acids 58 and 87 also assists in binding [ 43]. Recently, crystal structures of the C-type lectin-like domain from human CLEC-2 in complex with O-glycosylated peptides derived from human PDPN revealed two binding sites in CLEC-2. Distinct from the canonical PLAG binding site, a negatively charged Glu-Asp doublet in PDPN accepts the positively charged patch formed by Arg107, Arg152, and Arg157 in CLEC-2. This unique binding mode sheds light on how PDPN-CLEC-2 interaction achieves specificity and helps explain the structural requirements for function [ 46].

Mice with a conditional deletion of CLEC-2, Syk, SLP-76 and PLCγ2 in hematopoietic cells phenocopy Pdpn^−/− and inducible Prox1^−/− mice were shown to have misconnected blood and lymphatic vessels [ 6, 47, 48]. Because mice lacking nearly all platelets have normal lymphatic vasculature, it was thought that platelets were not required for this phenotype [ 49]. However, recent studies have demonstrated that CLEC-2 expression and downstream signaling through SLP-76 are required specifically in platelets [ 10, 50]. Interaction between CLEC-2 in platelets and PDPN on LECs induces platelet aggregation and prevents blood from flowing into new lymphatic vessels budding from the cardinal vein (Fig. 2). Inhibiting platelet aggregation, for instance by injecting a PDPN-blocking antibody, is sufficient to disrupt lymphatic development [ 51]. These data support the current model, in which PDPN is upregulated in Prox-1⁺Lyve-1⁺ LECs during budding of the lymph sac from the cardinal vein and interacts with CLEC-2 on platelets. This interaction, in turn, activates downstream signaling in platelets and triggers platelet aggregation. Platelet aggregation is required for a complete separation of the budding lymphatic vessels from the developing blood vessels. Additionally, recent studies demonstrate that following initial separation of blood and lymphatic vessels, the O-glycan-dependent PDPN-CLEC-2 interaction maintains functional separation and prevents retrograde flow of blood into lymphatics by sustaining lymphovenous valve integrity [ 52]. Therefore, PDPN-CLEC-2 interaction plays an essential role in lymphangiogenesis during development and maintenance of lymphatic vessel integrity after birth.

PDPN in the development or maintenance of lymphoid organ

PDPN expression was previously undetected in the spleens of mice which lack lymphocytes despite the presence of fibroblastic reticular cells (FRCs), as indicated by VCAM-1 and ER-TR7 staining [ 53, 54]. More recently, Peters et al. [ 55] observed that Pdpn^−/− adult mice lack peripheral lymph nodes and exhibit only lymph node remnants. These mice also exhibit defects in the formation of normal lymphoid follicles and germinal centers in spleen. Our recent studies also revealed that mice with postnatal, inducible, or conditional deletion of Pdpn in FRCs exhibit spontaneous bleeding in the mucosal lymph nodes and present significantly reduced levels of VE-cadherin, which is essential for overall vascular integrity [ 11]. These characteristics were also observed in mice with platelet-specific deficiency of CLEC-2 [ 11]. Further investigation of mice with postnatal PDPN deficiency revealed that the high endothelial venules (HEVs) within the lymph nodes exhibit decreased vascular integrity. As a result, labeled red blood cells injected intravenously leaked from HEVs in PDPN-deficient mice [ 11]. Platelets are able to access the perivascular space around HEVs in a lymphocyte transmigration-dependent manner. Once at the extra-lumenal surface of HEVs, CLEC-2 in platelets is activated by extravascular PDPN on FRCs. PDPN activation of platelet CLEC-2 induces release of sphingosine 1-phosphate (S1P), which is a bioactive lipid that in turn binds to its receptor S1PR1 on HEVs. Activation of S1PR1 increases HEV integrity by promoting VE-cadherin reassembly at the endothelial adherens junctions and subsequently facilitates effective sealing of the functional leakage caused by migrating lymphocytes [ 11]. The present study highlights the critical role of PDPN in maintaining HEV integrity through interactions with platelet CLEC-2. Local S1P release after PDPN-CLEC-2-mediated platelet activation is essential for HEV integrity.

Conclusions

PDPN is crucial for the development and maintenance of the lymphatic vascular system and lymphoid organ homeostasis. Prox1 induces transcription of Pdpn in LECs, and robust expression of PDPN at the LEC surface requires post-translational modification of the core protein structure with core 1-derived O-glycans by C1galt1 (Fig. 5). PDPN in LECs or FRCs activates CLEC-2 in platelets and ultimately triggers platelet activation and/or aggregation. This mechanism is vital to separate and maintain intact lymphatic vessels and to stabilize HEV integrity within lymph nodes.

To date, no patients with mutations in PDPN or CLEC-2 have yet been described. Increased expression of PDPN is commonly found in many types of tumors, especially in squamous carcinomas [ 2, 56− 58]. Considering that platelets interact with tumor cells, interactions between CLEC-2 and PDPN in tumors are also likely involved in tumor progression and metastasis [ 59, 60]. However, the manner in which PDPN contributes to tumor development remains unclear. Further studies are necessary to define the role of PDPN in human diseases.

A number of unanswered questions regarding the role of PDPN in vascular development and maintenance persist. For example, does PDPN activate other signaling pathways in addition to the one mediated by CLEC-2? If so, are additional receptors available for PDPN? Does PDPN itself transduce signals upon binding to CLEC-2? Addressing these questions will lead to new insights into the novel roles of PDPN in physiological and pathological conditions.

References

[1] Martín-Villar E, Megías D, Castel S, Yurrita MM, Vilaró S, Quintanilla M. Podoplanin binds ERM proteins to activate RhoA and promote epithelial-mesenchymal transition. J Cell Sci 2006; 119(Pt 21): 4541–4553

[2] Breiteneder-Geleff S, Matsui K, Soleiman A, Meraner P, Poczewski H, Kalt R, Schaffner G, Kerjaschki D. Podoplanin, novel 43-kd membrane protein of glomerular epithelial cells, is down-regulated in puromycin nephrosis. Am J Pathol 1997; 151(4): 1141–1152

[3] Ramirez MI, Millien G, Hinds A, Cao Y, Seldin DC, Williams MC. T1α a lung type I cell differentiation gene, is required for normal lung cell proliferation and alveolus formation at birth. Dev Biol 2003; 256(1): 61–72

[4] Schacht V, Ramirez MI, Hong YK, Hirakawa S, Feng D, Harvey N, Williams M, Dvorak AM, Dvorak HF, Oliver G, Detmar M. T1α/podoplanin deficiency disrupts normal lymphatic vasculature formation and causes lymphedema. EMBO J 2003; 22(14): 3546–3556

[5] Schietinger A, Philip M, Yoshida BA, Azadi P, Liu H, Meredith SC, Schreiber H. A mutant chaperone converts a wild-type protein into a tumor-specific antigen. Science 2006; 314(5797): 304–308

[6] Suzuki-Inoue K, Kato Y, Inoue O, Kaneko MK, Mishima K, Yatomi Y, Yamazaki Y, Narimatsu H, Ozaki Y. Involvement of the snake toxin receptor CLEC-2, in podoplanin-mediated platelet activation, by cancer cells. J Biol Chem 2007; 282(36): 25993–26001

[7] Chaipan C, Soilleux EJ, Simpson P, Hofmann H, Gramberg T, Marzi A, Geier M, Stewart EA, Eisemann J, Steinkasserer A, Suzuki-Inoue K, Fuller GL, Pearce AC, Watson SP, Hoxie JA, Baribaud F, Pöhlmann S. DC-SIGN and CLEC-2 mediate human immunodeficiency virus type 1 capture by platelets. J Virol 2006; 80(18): 8951–8960

[8] Tang T, Li L, Tang J, Li Y, Lin WY, Martin F, Grant D, Solloway M, Parker L, Ye W, Forrest W, Ghilardi N, Oravecz T, Platt KA, Rice DS, Hansen GM, Abuin A, Eberhart DE, Godowski P, Holt KH, Peterson A, Zambrowicz BP, de Sauvage FJ. A mouse knockout library for secreted and transmembrane proteins. Nat Biotechnol 2010; 28(7): 749–755

[9] Fuller GL, Williams JA, Tomlinson MG, Eble JA, Hanna SL, Pöhlmann S, Suzuki-Inoue K, Ozaki Y, Watson SP, Pearce AC. The C-type lectin receptors CLEC-2 and Dectin-1, but not DC-SIGN, signal via a novel YXXL-dependent signaling cascade. J Biol Chem 2007; 282(17): 12397–12409

[10] Bertozzi CC, Schmaier AA, Mericko P, Hess PR, Zou Z, Chen M, Chen CY, Xu B, Lu MM, Zhou D, Sebzda E, Santore MT, Merianos DJ, Stadtfeld M, Flake AW, Graf T, Skoda R, Maltzman JS, Koretzky GA, Kahn ML. Platelets regulate lymphatic vascular development through CLEC-2-SLP-76 signaling. Blood 2010; 116(4): 661–670

[11] Herzog BH, Fu J, Wilson SJ, Hess PR, Sen A, McDaniel JM, Pan Y, Sheng M, Yago T, Silasi-Mansat R, McGee S, May F, Nieswandt B, Morris AJ, Lupu F, Coughlin SR, McEver RP, Chen H, Kahn ML, Xia L. Podoplanin maintains high endothelial venule integrity by interacting with platelet CLEC-2. Nature 2013; 502(7469): 105–109

[12] Alitalo K, Tammela T, Petrova TV. Lymphangiogenesis in development and human disease. Nature 2005; 438(7070): 946–953

[13] Oliver G, Srinivasan RS. Lymphatic vasculature development: current concepts. Ann N Y Acad Sci 2008; 1131(1): 75–81

[14] Lohela M, Bry M, Tammela T, Alitalo K. VEGFs and receptors involved in angiogenesis versus lymphangiogenesis. Curr Opin Cell Biol 2009; 21(2): 154–165

[15] Srinivasan RS, Dillard ME, Lagutin OV, Lin FJ, Tsai S, Tsai MJ, Samokhvalov IM, Oliver G. Lineage tracing demonstrates the venous origin of the mammalian lymphatic vasculature. Genes Dev 2007; 21(19): 2422–2432

[16] Hong YK, Harvey N, Noh YH, Schacht V, Hirakawa S, Detmar M, Oliver G. Prox1 is a master control gene in the program specifying lymphatic endothelial cell fate. Dev Dyn 2002; 225(3): 351–357

[17] Johnson NC, Dillard ME, Baluk P, McDonald DM, Harvey NL, Frase SL, Oliver G. Lymphatic endothelial cell identity is reversible and its maintenance requires Prox1 activity. Genes Dev 2008; 22(23): 3282–3291

[18] Wigle JT, Oliver G. Prox1 function is required for the development of the murine lymphatic system. Cell 1999; 98(6): 769–778

[19] Wigle JT, Harvey N, Detmar M, Lagutina I, Grosveld G, Gunn MD, Jackson DG, Oliver G. An essential role for Prox1 in the induction of the lymphatic endothelial cell phenotype. EMBO J 2002; 21(7): 1505–1513

[20] Tammela T, Alitalo K. Lymphangiogenesis: molecular mechanisms and future promise. Cell 2010; 140(4): 460–476

[21] Oliver G, Sosa-Pineda B, Geisendorf S, Spana EP, Doe CQ, Gruss P. Prox 1, a prospero-related homeobox gene expressed during mouse development. Mech Dev 1993; 44(1): 3–16

[22] Hong YK, Foreman K, Shin JW, Hirakawa S, Curry CL, Sage DR, Libermann T, Dezube BJ, Fingeroth JD, Detmar M. Lymphatic reprogramming of blood vascular endothelium by Kaposi sarcoma-associated herpesvirus. Nat Genet 2004; 36(7): 683–685

[23] Lee S, Kang J, Yoo J, Ganesan SK, Cook SC, Aguilar B, Ramu S, Lee J, Hong YK. Prox1 physically and functionally interacts with COUP-TFII to specify lymphatic endothelial cell fate. Blood 2009; 113(8): 1856–1859

[24] Dumont DJ, Jussila L, Taipale J, Lymboussaki A, Mustonen T, Pajusola K, Breitman M, Alitalo K. Cardiovascular failure in mouse embryos deficient in VEGF receptor-3. Science 1998; 282(5390): 946–949

[25] Colvin JS, Bohne BA, Harding GW, McEwen DG, Ornitz DM. Skeletal overgrowth and deafness in mice lacking fibroblast growth factor receptor 3. Nat Genet 1996; 12(4): 390–397

[26] Gu C, Rodriguez ER, Reimert DV, Shu T, Fritzsch B, Richards LJ, Kolodkin AL, Ginty DD. Neuropilin-1 conveys semaphorin and VEGF signaling during neural and cardiovascular development. Dev Cell 2003; 5(1): 45–57

[27] Kitsukawa T, Shimizu M, Sanbo M, Hirata T, Taniguchi M, Bekku Y, Yagi T, Fujisawa H. Neuropilin-semaphorin III/D-mediated chemorepulsive signals play a crucial role in peripheral nerve projection in mice. Neuron 1997; 19(5): 995–1005

[28] Petrova TV, Karpanen T, Norrmén C, Mellor R, Tamakoshi T, Finegold D, Ferrell R, Kerjaschki D, Mortimer P, Ylä-Herttuala S, Miura N, Alitalo K. Defective valves and abnormal mural cell recruitment underlie lymphatic vascular failure in lymphedema distichiasis. Nat Med 2004; 10(9): 974–981

[29] Norrmén C, Ivanov KI, Cheng J, Zangger N, Delorenzi M, Jaquet M, Miura N, Puolakkainen P, Horsley V, Hu J, Augustin HG, Ylä-Herttuala S, Alitalo K, Petrova TV. FOXC2 controls formation and maturation of lymphatic collecting vessels through cooperation with NFATc1. J Cell Biol 2009; 185(3): 439–457

[30] Navarro A, Perez RE, Rezaiekhaligh M, Mabry SM, Ekekezie II. T1α/podoplanin is essential for capillary morphogenesis in lymphatic endothelial cells. Am J Physiol Lung Cell Mol Physiol 2008; 295(4): L543–L551

[31] Fu J, Gerhardt H, McDaniel JM, Xia B, Liu X, Ivanciu L, Ny A, Hermans K, Silasi-Mansat R, McGee S, Nye E, Ju T, Ramirez MI, Carmeliet P, Cummings RD, Lupu F, Xia L. Endothelial cell O-glycan deficiency causes blood/lymphatic misconnections and consequent fatty liver disease in mice. J Clin Invest 2008; 118(11): 3725–3737

[32] Pan Y, Wang WD, Yago T. Transcriptional regulation of podoplanin expression by Prox1 in lymphatic endothelial cells. Microvasc Res 2014; 94: 96–102

[33] Durchdewald M, Guinea-Viniegra J, Haag D, Riehl A, Lichter P, Hahn M, Wagner EF, Angel P, Hess J. Podoplanin is a novel fos target gene in skin carcinogenesis. Cancer Res 2008; 68(17): 6877–6883

[34] Kunita A, Kashima TG, Ohazama A, Grigoriadis AE, Fukayama M. Podoplanin is regulated by AP-1 and promotes platelet aggregation and cell migration in osteosarcoma. Am J Pathol 2011; 179(2): 1041–1049

[35] Peterziel H, Müller J, Danner A, Barbus S, Liu HK, Radlwimmer B, Pietsch T, Lichter P, Schütz G, Hess J, Angel P. Expression of podoplanin in human astrocytic brain tumors is controlled by the PI3K-AKT-AP-1 signaling pathway and promoter methylation. Neuro-oncol 2012; 14(4): 426–439

[36] An G, Wei B, Xia B, McDaniel JM, Ju T, Cummings RD, Braun J, Xia L. Increased susceptibility to colitis and colorectal tumors in mice lacking core 3-derived O-glycans. J Exp Med 2007; 204(6): 1417–1429

[37] Ju T, Brewer K, D’Souza A, Cummings RD, Canfield WM. Cloning and expression of human core 1 β1,3-galactosyltransferase. J Biol Chem 2002; 277(1): 178–186

[38] Xia L, Ju T, Westmuckett A, An G, Ivanciu L, McDaniel JM, Lupu F, Cummings RD, McEver RP. Defective angiogenesis and fatal embryonic hemorrhage in mice lacking core 1-derived O-glycans. J Cell Biol 2004; 164(3): 451–459

[39] Ju T, Cummings RD, Canfield WM. Purification, characterization, and subunit structure of rat core 1 β1,3-galactosyltransferase. J Biol Chem 2002; 277(1): 169–177

[40] Kaneko M, Kato Y, Kunita A, Fujita N, Tsuruo T, Osawa M. Functional sialylated O-glycan to platelet aggregation on Aggrus (T1α/Podoplanin) molecules expressed in Chinese hamster ovary cells. J Biol Chem 2004; 279(37): 38838–38843

[41] Kaneko MK, Kato Y, Kameyama A, Ito H, Kuno A, Hirabayashi J, Kubota T, Amano K, Chiba Y, Hasegawa Y, Sasagawa I, Mishima K, Narimatsu H. Functional glycosylation of human podoplanin: glycan structure of platelet aggregation-inducing factor. FEBS Lett 2007; 581(2): 331–336

[42] Pan Y, Yago T, Fu J, Herzog B, McDaniel JM, Mehta-D’Souza P, Cai X, Ruan C, McEver RP, West C, Dai K, Chen H, Xia L. Podoplanin requires sialylated O-glycans for stable expression on lymphatic endothelial cells and for interaction with platelets. Blood 2014; 124(24): 3656–3665

[43] Kato Y, Kaneko MK, Kunita A, Ito H, Kameyama A, Ogasawara S, Matsuura N, Hasegawa Y, Suzuki-Inoue K, Inoue O, Ozaki Y, Narimatsu H. Molecular analysis of the pathophysiological binding of the platelet aggregation-inducing factor podoplanin to the C-type lectin-like receptor CLEC-2. Cancer Sci 2008; 99(1): 54–61

[44] Kato Y, Fujita N, Kunita A, Sato S, Kaneko M, Osawa M, Tsuruo T. Molecular identification of Aggrus/T1α as a platelet aggregation-inducing factor expressed in colorectal tumors. J Biol Chem 2003; 278(51): 51599–51605

[45] Bianchi R, Fischer E, Yuen D, Ernst E, Ochsenbein AM, Chen L, Otto VI, Detmar M. Mutation of threonine 34 in mouse podoplanin-Fc reduces CLEC-2 binding and toxicity in vivo while retaining antilymphangiogenic activity. J Biol Chem 2014; 289(30): 21016–21027

[46] Nagae M, Morita-Matsumoto K, Kato M, Kaneko MK, Kato Y, Yamaguchi Y. A platform of C-type lectin-like receptor CLEC-2 for binding O-glycosylated podoplanin and nonglycosylated rhodocytin. Structure 2014; 22(12): 1711–1721

[47] Abtahian F, Guerriero A, Sebzda E, Lu MM, Zhou R, Mocsai A, Myers EE, Huang B, Jackson DG, Ferrari VA, Tybulewicz V, Lowell CA, Lepore JJ, Koretzky GA, Kahn ML. Regulation of blood and lymphatic vascular separation by signaling proteins SLP-76 and Syk. Science 2003; 299(5604): 247–251

[48] Sebzda E, Hibbard C, Sweeney S, Abtahian F, Bezman N, Clemens G, Maltzman JS, Cheng L, Liu F, Turner M, Tybulewicz V, Koretzky GA, Kahn ML. Syk and Slp-76 mutant mice reveal a cell-autonomous hematopoietic cell contribution to vascular development. Dev Cell 2006; 11(3): 349–361

[49] Shivdasani RA, Rosenblatt MF, Zucker-Franklin D, Jackson CW, Hunt P, Saris CJ, Orkin SH. Transcription factor NF-E2 is required for platelet formation independent of the actions of thrombopoietin/MGDF in megakaryocyte development. Cell 1995; 81(5): 695–704

[50] Osada M, Inoue O, Ding G, Shirai T, Ichise H, Hirayama K, Takano K, Yatomi Y, Hirashima M, Fujii H, Suzuki-Inoue K, Ozaki Y. Platelet activation receptor CLEC-2 regulates blood/lymphatic vessel separation by inhibiting proliferation, migration, and tube formation of lymphatic endothelial cells. J Biol Chem 2012; 287(26): 22241–22252

[51] Uhrin P, Zaujec J, Breuss JM, Olcaydu D, Chrenek P, Stockinger H, Fuertbauer E, Moser M, Haiko P, Fässler R, Alitalo K, Binder BR, Kerjaschki D. Novel function for blood platelets and podoplanin in developmental separation of blood and lymphatic circulation. Blood 2010; 115(19): 3997–4005

[52] Hess PR, Rawnsley DR, Jakus Z, Yang Y, Sweet DT, Fu J, Herzog B, Lu M, Nieswandt B, Oliver G, Makinen T, Xia L, Kahn ML. Platelets mediate lymphovenous hemostasis to maintain blood-lymphatic separation throughout life. J Clin Invest 2014; 124(1): 273–284

[53] Ngo VN, Cornall RJ, Cyster JG. Splenic T zone development is B cell dependent. J Exp Med 2001; 194(11): 1649–1660

[54] Bekiaris V, Withers D, Glanville SH, McConnell FM, Parnell SM, Kim MY, Gaspal FM, Jenkinson E, Sweet C, Anderson G, Lane PJ. Role of CD30 in B/T segregation in the spleen. J Immunol 2007; 179(11): 7535–7543

[55] Peters A, Pitcher LA, Sullivan JM, Mitsdoerffer M, Acton SE, Franz B, Wucherpfennig K, Turley S, Carroll MC, Sobel RA, Bettelli E, Kuchroo VK. Th17 cells induce ectopic lymphoid follicles in central nervous system tissue inflammation. Immunity 2011; 35(6): 986–996

[56] Ji RC. Lymphatic endothelial cells, tumor lymphangiogenesis and metastasis: new insights into intratumoral and peritumoral lymphatics. Cancer Metastasis Rev 2006; 25(4): 677–694

[57] Swartz MA, Lund AW. Lymphatic and interstitial flow in the tumour microenvironment: linking mechanobiology with immunity. Nat Rev Cancer 2012; 12(3): 210–219

[58] Ordóñez NG. Podoplanin: a novel diagnostic immunohistochemical marker. Adv Anat Pathol 2006; 13(2): 83–88

[59] Lowe KL, Navarro-Nunez L, Watson SP. Platelet CLEC-2 and podoplanin in cancer metastasis. Thromb Res 2012; 129(Suppl 1): S30–S37

[60] Dang Q, Liu J, Li J, Sun Y. Podoplanin: a novel regulator of tumor invasion and metastasis. Med Oncol 2014; 31(9): 24

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg