Intranodal interaction with dendritic cells dynamically regulates surface expression of the co-stimulatory receptor CD226 protein on murine T cells

Dendritic cells (DCs) are the most potent antigen-presenting cells of the immune system. Depending on their maturation status, they prime T cells to induce adaptive immunity or tolerance. DCs express CD155, an immunoglobulin-like receptor binding CD226 present on T and natural killer (NK) cells. CD226 represents an important co-stimulator during T cell priming but also serves as an activating receptor on cytotoxic T and NK cells. Here, we report that cells of the T and NK cell lineage of CD155^−/− mice express markedly elevated protein levels of CD226 compared with wild type (WT). On heterozygous CD155^+/− T cells, CD226 up-regulation is half-maximal, implying an inverse gene-dosis effect. Moreover, CD226 up-regulation is independent of antigen-driven activation because it occurs already in thymocytes and naïve peripheral T cells. In vivo, neutralizing anti-CD155 antibody elicits up-regulation of CD226 on T cells demonstrating, that the observed modulation can be triggered by interrupting CD155-CD226 contacts. Adoptive transfers of WT or CD155^−/− T cells into CD155^−/− or WT recipients, respectively, revealed that CD226 modulation is accomplished in trans. Analysis of bone marrow chimeras showed that regulators in trans are of hematopoietic origin. We demonstrate that DCs are capable of manipulating CD226 levels on T cells in vivo but not in vitro, suggesting that the process of T cells actively scanning antigen-presenting DCs inside secondary lymphoid organs is required for CD226 modulation. Hence, a CD226 level divergent from WT may be exploited as a sensor to detect abnormal DC/T-cell cross-talk as illustrated for T cells in mice lacking CCR7.     

NKG2D and CD94 bind to multimeric α2,3-linked N-acetylneuraminic acid

Killer lectin-like receptors on natural killer cells mediate cytotoxicity through glycans on target cells including the sialyl Lewis X antigen (sLeX). We investigated whether NK group 2D (NKG2D) and CD94 can bind to sialylated N-linked glycans, using recombinant glutathione S-transferase-fused extracellular lectin-like domains of NKG2D (rNKG2Dlec) and CD94 (rCD94lec). Both rNKG2Dlec and rCD94lec bound to plates coated with high-sLeX-expressing transferrin secreted by HepG2 cells (HepTF). The binding of rNKG2Dlec and rCD94lec to HepTF was markedly suppressed by treatment of HepTF with neuraminidase and in the presence of N-acetylneuraminic acid. Moreover, rNKG2Dlec and rCD94lec bound to α2,3-sialylated human α₁-acid glycoprotein (AGP) but not to α2,6-sialylated AGP. Mutagenesis revealed that ¹⁵²Y of NKG2D and ¹⁴⁴F and ¹⁶⁰N of CD94 were critical for HepTF binding. This is the first report that NKG2D and CD94 bind to α2,3-sialylated but not to α2,6-sialylated multi-antennary N-glycans.     

Involvement of CD226+ NK cells in immunopathogenesis of systemic lupus erythematosus

Dysfunction of immune systems, including innate and adaptive immunity, is responsible for the immunopathogenesis of systemic lupus erythematosus (SLE). NK cells are a major part of the innate immune system, and diminished populations of NK cells have been reported in SLE patients. However, the mechanisms behind this decrease and the role of NK cells in SLE pathogenesis remain poorly understood. In this study, we found that a deficiency of NK cells, especially CD226(+) NK cells, is prominent in patients with active SLE. Meanwhile, expression of the CD226 ligands CD112 and CD155 on plasmacytoid dendritic cells is observed in SLE patients; thus, activation of CD226(+) NK cells may be induced by CD226-ligand interactions. Furthermore, IFN-α, which is mainly produced by plasmacytoid dendritic cells, can mediate the activation-induced cell death of NK cells. Therefore, these processes likely contribute to the loss of NK cells in patients with active SLE. Despite the impaired cytotoxicity of peripheral NK cells in human SLE patients and mouse SLE models, we provide evidence that CD226(+) NK cells infiltrate the kidneys of predisease MRL-lpr/lpr mice. Kidney-infiltrating NK cells displayed an activated phenotype and a marked ability to produce cytotoxic granules. These results suggest that, before apoptosis, activated NK cells can infiltrate tissues and, to some extent, mediate tissue injury by producing cytotoxic granules and immunoregulatory cytokines.     

CD22 ligand binding regulates normal and malignant B lymphocyte survival in vivo

The CD22 extracellular domain regulates B lymphocyte function by interacting with alpha2,6-linked sialic acid-bearing ligands. To understand how CD22 ligand interactions affect B cell function in vivo, mouse anti-mouse CD22 mAbs were generated that inhibit CD22 ligand binding to varying degrees. Remarkably, mAbs which blocked CD22 ligand binding accelerated mature B cell turnover by 2- to 4-fold in blood, spleen, and lymph nodes. CD22 ligand-blocking mAbs also inhibited the survival of adoptively transferred normal (73-88%) and malignant (90%) B cells in vivo. Moreover, mAbs that bound CD22 ligand binding domains induced significant CD22 internalization, depleted marginal zone B cells (82-99%), and reduced mature recirculating B cell numbers by 75-85%. The CD22 mAb effects were independent of complement and FcRs, and the CD22 mAbs had minimal effects in CD22AA mice that express mutated CD22 that is not capable of ligand binding. These data demonstrate that inhibition of CD22 ligand binding can disrupt normal and malignant B cell survival in vivo and suggest a novel mechanism of action for therapeutics targeting CD22 ligand binding domains.     

Decreased expression of DNAM-1 on NK cells from acute myeloid leukemia patients

This study tested the hypothesis that the expression of CD112 and CD155 (DNAM-1 ligands) on leukemic blasts induces a decreased expression of the activating receptor DNAM-1 on natural killer (NK) cells from acute myeloid leukemia (AML) patients. DNAM-1 is a co-receptor involved in the activation of NK cell cytotoxicity after its interaction with its ligands CD112 and CD155 on target cells. Here we study the expression of DNAM-1 on NK cells and DNAM-1 ligands on blasts from AML patients stratified by age. The results demonstrate that NK cells from AML patients younger than 65 years have a reduced expression of DNAM-1 compared with age-matched controls. The analysis of DNAM-1 ligands showed a high expression of CD112 and CD155 on leukemic blasts. An inverse correlation between CD112 expression on leukemic blasts and DNAM-1 expression on NK cells was found. Furthermore, downregulation of DNAM-1 was induced on healthy donors' NK cells after in vitro culture with leukemic blasts expressing DNAM-1 ligands. In conclusion, these results support the hypothesis that receptor-ligand crosslinking downregulates DNAM-1 expression on NK cells from patients <65 years of age. Considering the relevance of DNAM-1 in NK recognition and killing of leukemic cells, the reduced expression of this receptor on NK cells from AML patients can represent an additional mechanism of tumor escape.     

Tumor-primed human natural killer cells lyse NK-resistant tumor targets: evidence of a two-stage process in resting NK cell activation

NK cells are defined as those cells that lyse tumor cells without priming. In this study, we show that the preincubation of resting human NK cells with the leukemia cell CTV-1 primes NK cells to lyse NK-resistant cell lines, primary leukemias, and solid tumors even when HLA-matched, allogeneic or autologous. The primed NK cells remained nonresponsive to HLA-C matched and mismatched normal mononuclear cells from multiple donors. CD69, a known NK trigger receptor, was shown to be the predominant trigger on the tumor-primed NK cells because lysis was blocked with the rCD69 protein. The lack of lytic activity against normal hemopoietic cells implied that the ligand for CD69 is tumor restricted, and this was confirmed by experiments using fluorochrome labeled rCD69. It has been recently shown that resting NK cells require prior stimulation with IL-2 before triggering by all known NK-triggering ligands. In this study, we show that a tumor cell can provide the NK priming signal independently of IL-2. These data provide evidence for two NK evasion strategies for tumor cells, namely the prevention of priming (type1 evasion) and failure to trigger (type 2 evasion). Most NK-resistant cell lines are type 1 and fail to prime resting NK cells but are lysed by IL-2-primed NK cells. In contrast, CTV-1 cells prime resting NK cells but fail to trigger (type 2), and coincubation with CTV-1 primes for triggering by type 1 NK-resistant tumor cells. These tumor-activated NK cells lyse a broad spectrum of tumor cells with a degree of specificity never previously reported.     

Identification and characterization of murine DNAM-1 (CD226) and its poliovirus receptor family ligands

The leukocyte adhesion molecule DNAM-1 (CD226) is a member of the immunoglobulin superfamily and constitutively expressed on the majority of CD4+ and CD8+ T lymphocytes, natural killer (NK) cells, monocytes/macrophages, and a subset of B lymphocytes. The poliovirus receptor (PVR; CD155) and its family member nectin 2 (CD112) have recently been identified as the ligands for DNAM-1. Interaction of DNAM-1 with the ligands induces NK cell- and CD8+ T cell-mediated cytotoxicity and cytokine secretion. However, in vivo function of the receptor-ligand interaction has remained unclear. Here, we identified murine DNAM-1 and PVR homologues that physically and functionally bind each other. We demonstrated that ligand binding of murine DNAM-1 induced a costimulatory signal in antigen-specific CD8+ T cells. These results should provide a useful animal model to explore a role of DNAM-1 in immune responses in vivo.     

Characterization of in vitro inhibition of human immunodeficiency virus by purified recombinant CD4.

The first step in infection of human T cells with human immunodeficiency virus (HIV) is binding of viral envelope glycoprotein gp120 to its cellular receptor, CD4. The specificity of this interaction has led to the development of soluble recombinant CD4 (rCD4) as a potential antiviral and therapeutic agent. We have previously shown that crude preparations of rCD4 can indeed block infection of T cells by HIV type 1 (HIV-1). Here we present a more detailed analysis of this antiviral activity, using HIV-1 infection of the T lymphoblastoid cell line H9 as a model. Purified preparations of rCD4 blocked infection in this system at nanomolar concentrations; combined with the known affinity of the CD4-gp120 interaction, this finding suggests that the inhibition is simply due to competition for gp120 binding. As predicted, rCD4 had comparable activity against all strains of HIV-1 tested and significant activity against HIV-2. Higher concentrations of rCD4 blocked infection even after the virus had been adsorbed to the cells. These findings imply that the processes of viral adsorption and penetration require different numbers of gp120-CD4 interactions. Recombinant CD4 was able to prevent the spread of HIV infection in mixtures of uninfected and previously infected cells. Our studies support the notion that rCD4 is a potent antiviral agent, effective against a broad range of HIV-1 isolates, and demonstrate the value of purified rCD4 as an experimental tool for studying the mechanism of virus entry into cells.     

Cutting edge: primary structure of the light chain of fusion regulatory protein-1/CD98/4F2 predicts a protein with multiple transmembrane domains that is almost identical to the amino acid transporter E16

The CD98 light chain (CD98LC) was copurified from HeLa S3 cells by an affinity chromatography using a mAb specific for the fusion regulatory protein-1 (FRP-1) which is identical to the CD98 heavy chain. On the basis of the N-terminal sequence (63 amino acids) of purified CD98LC polypeptide, we have cloned a PCR fragment (155 bp) from a HeLa S3 cDNA library and finally obtained a full cDNA clone encoding the CD98LC. Fluorescence in situ hybridization analysis using the cDNA assigned the CD98LC gene to the long arm of human chromosome 16 (16q24). The predicted amino acid sequence suggested that CD98LC is a protein with multiple transmembrane domains and is almost identical to the amino acid transporter E16. Resting monocytes and lymphocytes expressed CD98LC as analyzed by a newly isolated anti-CD98LC mAb, which showed cross-reactivity with insect Sf9 cells as well as with various mammalian cell lines.     

The expression,regulation and adhesion function of a novel CD molecule,CD226, on human endothelial cells

CD226 is a 67 kDa type I transmembrane glycoprotein mainly expressed on activated T cells, NK cells and platelets, and involved in the differentiation of cytotoxic T lymphocytes (CTL) and NK, as well as platelet activation and aggregation. Here we found that the expression of CD226 protein and CD226mRNA were very weak in resting HUVEC and ECV304 cells, whereas high level expression could be observed when these cells were stimulated. The binding activities between activated endothelial cells and activated Jurkat cells could be partly blocked by CD226/Ig fusion protein. Similarly, CD226/Ig could also partly block the adhesion between activated endothelial cells and some leukocytes or colo205 cells. These data provided the evidence that activated endothelial cells could express high level of CD226, and CD226 was involved in the endothelial cells' adhesion. The above findings suggested that CD226 is a novel inducible adhesion molecule on human endothelial cells.     

The Role of T Cell Costimulation via DNAM-1 in Kidney Transplantation

DNAX accessory protein-1 (DNAM-1, CD226) is a co-stimulatory and adhesion molecule expressed mainly by natural killer cells and T cells. DNAM-1 and its two ligands CD112 and CD155 are important in graft-versus-host disease, but their role in solid organ transplantation is largely unknown. We investigated the relevance of this pathway in a mouse kidney transplantation model. CD112 and CD155 are constitutively expressed on renal tubular cells and strongly upregulated in acutely rejected renal allografts. In vitro DNAM-1 blockade during allogeneic priming reduced the allospecific T cell response but not the allospecific cytotoxicity against renal tubular epithelial cells. Accordingly, absence of DNAM-1 in recipient mice or absence of CD112 or CD155 in the kidney allograft did not significantly influence renal function and severity of rejection after transplantation, but led to a higher incidence of infarcts in CD112 and CD155 deficient kidney allografts. Thus, DNAM-1 blockade is not effective in preventing transplant rejection. Despite of being highly expressed, CD112 and CD155 do not appear to play a major immunogenic role in kidney transplantation. Considering the high incidence of renal infarcts in CD112 and CD155 deficient grafts, blocking these molecules might be detrimental.     

Using carboxyfluorescein diacetate succinimidyl ester to monitor human NK cell division: analysis of the effect of activating and inhibitory class I MHC receptors

Human NK cells labelled intracellularly with the fluorescent dye 5- and 6-carboxyfluorescein diacetate succinimidyl ester (CFSE) were used to assess the effect of ligating class I MHC receptors on NK cell division. The NK cell lines used in these studies expressed a selection of the killer immunoglobulin-like receptors CD158b and CD158a and the CD94/NKG2 family of C-type lectin receptors. The NK cells were cultured in medium containing recombinant (r)IL-2 and receptors were ligated using plastic bound mAb or using soluble murine IgG mAb and FcRII+ gamma-irradiated murine P815 cells. The results obtained show that ligating class I MHC-activating receptors in either culture system stimulates NK cells to divide. Quantitative analysis of cell division reveals that a substantial loss of NK progenitor cells occurs when NK cell-activating receptors are ligated using plastic bound mAb, consistent with concomitant activation-induced cell death. By contrast, progenitor cell loss is prevented when activating receptors are ligated using soluble mAb and P815 cells, suggesting a role for cellular costimulation in cell survival. When inhibitory receptors are coligated with activating receptors using soluble mAb and P815 cells, NK cell division is inhibited. These results demonstrate the potential importance of the activating and inhibitory class I MHC receptors in regulating NK cell proliferation.     

A ligand for CD5 is CD5

Recognition by scavenger receptor cysteine-rich domains on membrane proteins regulates innate and adaptive immune responses. Two receptors expressed primarily on T cells, CD5 and CD6, are linked genetically and are structurally similar, both containing three scavenger receptor cysteine-rich domains in their extracellular regions. A specific cell surface interaction for CD5 has been difficult to define at the molecular level because of the susceptibility of CD5 protein to denaturation. By using soluble CD5 purified at neutral pH to preserve biological activity, we show that CD5 mediates species-specific homophilic interactions. CD5 domain 1 only is involved in the interaction. CD5 mAbs that have functional effects in humans, rats, and mice block homophilic binding. Ag-specific responses by mouse T cells in vitro were increased when engagement of human CD5 domain 1 was inhibited by mutation or by IgG or Fab fragment from a CD5 mAb. This showed that homophilic binding results in productive engagement. Enhancement of polyclonal immune responses of rat lymph node cells by a Fab fragment from a CD5 mAb shown to block homophilic interactions provided evidence that the extracellular region of CD5 regulates inhibition in normal cells. These biochemical and in vitro functional assays provide evidence that the extracellular region of CD5 regulates immunity through species-specific homophilic interactions.     

The murine NK receptor 2B4 (CD244) exhibits inhibitory function independent of signaling lymphocytic activation molecule-associated protein expression

2B4 (CD244) is a receptor belonging to the CD2-signaling lymphocytic activation molecule family and is found on all murine NK cells and a subset of NKT and CD8+ T cells. Murine 2B4 is expressed as two isoforms (2B4 short and 2B4 long) that arise by alternative splicing. They differ only in their cytoplasmic domains and exhibit opposing function when expressed in the RNK-16 cell line. The ligand for 2B4, CD48, is expressed on all hemopoietic cells. Previous studies have shown that treatment of NK cells with a 2B4 mAb results in increased cytotoxicity and IFN-gamma production. In this report, we used CD48+/- variants of the P815 tumor cell line and 2B4 knockout mice to show that engagement of 2B4 by its counterreceptor, CD48, expressed on target cells leads to an inhibition in NK cytotoxicity. The addition of 2B4 or CD48 mAb relieves this inhibition resulting in enhanced target cell lysis. This 2B4-mediated inhibition acts independently of signaling lymphocytic activation molecule-associated protein expression. Imaging studies show that 2B4 preferentially accumulates at the interface between NK and target cells during nonlytic events also indicative of an inhibitory receptor. This predominant inhibitory function of murine 2B4 correlates with increased 2B4 long isoform level expression over 2B4 short.     

T11/CD2 activation of cloned human natural killer cells results in increased conjugate formation and exocytosis of cytolytic granules

The T11 (CD2) antigen has been found to be an alternate pathway for antigen-independent activation of resting T cells. T11 triggering also results in activation of NK cells and enhancement of their cytolytic function. The present studies were carried out to further define the mechanisms whereby cytotoxicity is enhanced after T11 activation. A series of clonal human NK cell lines were analyzed after incubation with monoclonal anti-T112 and anti-T113 antibodies specific for different epitopes of the CD2 protein. Anti-T112/3 triggering resulted in increased cytotoxicity against a variety of target cells. Similar results were obtained with F(ab')2 fragments of anti-T112/3, indicating that this effect was not mediated through binding of FcR. The induction of cytotoxicity was found to be associated with increased formation of effector cell-target cell conjugates and with release of secretory granule-localized 35S-labeled proteoglycans. Both enhanced conjugate formation and cytotoxicity could be blocked by anti-lymphocyte function-associated antigen (LFA-1) mAb. Ultrastructural analysis of NK cells after T11 activation demonstrated increased adherence of effector cells to targets and other NK cells as well as a directional reorientation of cytoplasm and intracellular granules toward the area of contact between cells. Discharge of granules occurred into pockets bounded by closely apposed plasma membranes. In the presence of anti-LFA-1 and anti-T112/3, the close apposition and formation of pockets between effector cells and target cells did not occur but the cells exocytosed their intracellular granules. T11 activation of NK cloned cells also resulted in the formation of the homotypic conjugates and autocytotoxicity. As seen with resistant allogeneic targets, autocytotoxicity was mediated by F(ab')2 fragments of T112/3 antibodies and could be blocked by anti-LFA-1 antibody. Ultrastructural analysis of NK cloned cells after T11 activation confirmed the presence of homotypic conjugates with reorientation of effector cells toward one another and discharge of cytolytic granules into pockets formed between NK cloned cells. Taken together, these results indicate that T11-induced cytolytic function of NK cells is, in part, mediated through increased binding of effector cells and targets and that enhanced conjugate formation is at least in part mediated by the LFA-1 antigen. In addition, T11 activation results in the triggering of the cytolytic mechanism of NK cells and the exocytosis of cytolytic granules and their constituents.     

Functional requirement for SAP in 2B4-mediated activation of human natural killer cells as revealed by the X-linked lymphoproliferative syndrome

X-linked lymphoproliferative syndrome (XLP) is an immunodeficiency characterized by life-threatening infectious mononucleosis and EBV-induced B cell lymphoma. The gene mutated in XLP encodes SLAM (signaling lymphocytic activation molecule-associated protein)-associated protein (SAP), a small SH2 domain-containing protein. SAP associates with 2B4 and SLAM, activating receptors expressed by NK and T cells, and prevents recruitment of SH2 domain-containing protein tyrosine phosphatase-2 SHP-2) to the cytoplasmic domains of these receptors. The phenotype of XLP may therefore result from perturbed signaling through SAP-associating receptors. We have addressed the functional consequence of SAP deficiency on 2B4-mediated NK cell activation. Ligating 2B4 on normal human NK cells with anti-2B4 mAb or interaction with transfectants bearing the 2B4 ligand CD48 induced NK cell cytotoxicity. In contrast, ligation of 2B4 on NK cells from a SAP-deficient XLP patient failed to initiate cytotoxicity. Despite this, CD2 or CD16-induced cytotoxicity of SAP-deficient NK cells was similar to that of normal NK cells. Thus, selective impairment of 2B4-mediated NK cell activation may contribute to the immunopathology of XLP.     

Human Immunodeficiency Virus Type 1 (HIV-1) Infection and Transcytosis Activity of a HIV-1 Susceptible Clone from HeLa Cell

In order to clarify the transmission process of human immunodeficiency virus type 1 (HIV-1) through the epithelial cell barrier, HeLa cells susceptible and non-susceptible to HIV-1 were cloned and designated as P6 HeLa and N7 HeLa cells, respectively. P6 HeLa cells could be infected with the LAI strain of HIV-1 and mediated HIV-1 transcytosis. In contrast, N7 HeLa cells exhibited neither HIV-1 infection nor transcytosis. CD4 and galactosylceramide as the receptors for HIV-1 were not detected on P6 HeLa cells, although an anti-CD4 monoclonal antibody (mAb) blocked HIV-1 infection. Since HIV-1-infected P6 HeLa cells exhibited no fusion and survived, we speculated that the P6 HeLa cells expressed molecules other than CD4 which facilitated HIV-1 infection. Two mAbs (A-14 ITK and C57 a9-9) which inhibited the HIV-1 infection of P6 HeLa cells were generated. Each mAb recognized distinct molecule(s) as shown by Western blotting. Transcytosis by the P6 HeLa cells was inhibited by C57 a9-9 but not by A-14 ITK or anti-CD4 mAb. Both infection and transcytosis may be responsible for HIV-1 transmission through epithelial cells in a complex manner. Although infection and transcytosis occurred via different mechanisms, the molecule(s) recognized by C57 a9-9 mAb may be associated with both processes.     

CD4 immunoadhesin, but not recombinant soluble CD4, blocks syncytium formation by human immunodeficiency virus type 2-infected lymphoid cells.

Recombinant soluble CD4 (rCD4) has been shown to be an effective inhibitor of human immunodeficiency virus type 1 (HIV-1) and HIV-2 infection of lymphoid cells in vitro. In this report, we characterized the effects of rCD4, the V1V2 fragment of CD4, and the immunoadhesin CD4-immunoglobulin G on syncytium formation between lymphoid cells infected by HIV-1 or HIV-2 and uninfected cells. All three molecules blocked HIV-1-mediated syncytium formation, but only CD4-immunoglobulin G blocked HIV-2-mediated syncytium formation. rCD4 and the V1V2 fragment of CD4 enhanced HIV-2-mediated syncytium formation. These results suggest that the process of cell fusion is significantly different between HIV-1- and HIV-2-infected cells.     

HIV-infected humans, but not chimpanzees, have circulating cytotoxic T lymphocytes that lyse uninfected CD4+ cells

It has been suggested that autoimmune phenomena contribute to the depletion of CD4+ T cells and the development of AIDS in HIV-1 infected humans based, in part, on observations that some HIV-1-infected humans have autoantibodies reactive with Ag expressed on uninfected CD4+ cells. In this study, 11 of 14 asymptomatic HIV-1-infected homosexuals and hemophiliacs, but none of 17 uninfected homosexuals or heterosexuals, were found to have cytotoxic lymphocytes in blood that can lyse uninfected CD4+ T cells from humans and chimpanzees but not human B lymphoblastoid cells or mouse T cells. The cytotoxic PBL were concluded to be CTL rather than NK cells, with the phenotype being CD3+, TCR-1 alpha beta+, CD8+, CD4-, CD16- based on findings that PBL-mediated lysis of uninfected CD4+ cells was 1) blocked by a mAb to CD3, which inhibits CTL but not NK activity; 2) diminished by treatment of PBL with a mAb to CD8 and C, but not by treatment with mAb to CD4 or CD16 and C; and 3) blocked by mAb WT31 directed against the TCR-1 alpha beta. In contrast, PBL from HIV-1-infected chimpanzees, which to date have not developed AIDS, lacked detectable CTL lytic for uninfected CD4+ cells.     

The Epidermal Growth Factor-like Domain of CD93 Is a Potent Angiogenic Factor

Human CD93, an epidermal growth factor (EGF)-like domain containing transmembrane protein, is predominantly expressed in the vascular endothelium. Studies have shown that AA4, the homolog of CD93 in mice, may mediate cell migration and angiogenesis in endothelial cells. Soluble CD93 has been detected in the plasma of healthy individuals. However, the role of soluble CD93 in the endothelium remains unclear. Recombinant soluble CD93 proteins with EGF-like domains (rCD93D123, with domains 1, 2, and 3; and rCD93D23, with domains 2 and 3) were generated to determine their functions in angiogenesis. We found that rCD93D23 was more potent than rCD93D123 in stimulating the proliferation and migration of human umbilical vein endothelial cells (HUVECs). Production of matrix-metalloproteinase 2 increased after the HUVECs were treated with rCD93D23. Further, in a tube formation assay, rCD93D23 induced cell differentiation of HUVECs through phosphoinositide 3-kinase/Akt/endothelial nitric oxide synthase and extracellular signal-regulated kinases-1/2 signaling. Moreover, rCD93D23 promoted blood vessel formation in a Matrigel-plug assay and an oxygen-induced retinopathy model in vivo. Our findings suggest that the soluble EGF-like domain containing CD93 protein is a novel angiogenic factor acting on the endothelium.