FOXO1 promotes endothelial cell elongation and angiogenesis by up-regulating the phosphorylation of myosin light chain 2.

The forkhead box O1 (FOXO1) is an important transcription factor related to proliferation, metabolism, and homeostasis, while the major phenotype of FOXO1-null mice is abnormal vascular morphology, such as vessel enlargement and dilation. In in vitro mouse embryonic stem cell (ESC)-differentiation system, Foxo1-/- vascular endothelial cells (ECs) fail to elongate, and mimic the abnormalities of FOXO1-deficiency in vivo. Here, we identified the PPP1R14C gene as the FOXO1 target genes responsible for elongating using transcriptome analyses in ESC-derived ECs (ESC-ECs), and found that the FOXO1-PPP1R14C-myosin light chain 2 (MLC2) axis is required for EC elongation during angiogenesis. MLC2 is phosphorylated by MLC kinase (MLCK) and dephosphorylated by MLC phosphatase (MLCP). PPP1R14C is an inhibitor of PP1, the catalytic subunit of MLCP. The abnormal morphology of Foxo1-/- ESC-ECs was associated with low level of PPP1R14C and loss of MLC2 phosphorylation, which were reversed by PPP1R14C-introduction. Knockdown of either FOXO1 or PPP1R14C suppressed vascular cord formation and reduced MLC2 phosphorylation in human ECs (HUVECs). The mouse and human PPP1R14C locus possesses an enhancer element containing conserved FOXO1-binding motifs. In vivo chemical inhibition of MLC2 phosphorylation caused dilated vascular structures in mouse embryos. Furthermore, foxo1 or ppp1r14c-knockdown zebrafish exhibited vascular malformations, which were also restored by PPP1R14C-introduction. Mechanistically, FOXO1 suppressed MLCP activity by up-regulating PPP1R14C expression, thereby promoting MLC2 phosphorylation and EC elongation, which are necessary for vascular development. Given the importance of MLC2 phosphorylation in cell morphogenesis, this study may provide novel insights into the role of FOXO1 in control of angiogenesis.

Independent origins of fetal liver haematopoietic stem and progenitor cells.

Self-renewal and differentiation are tightly controlled to maintain haematopoietic stem cell (HSC) homeostasis in the adult bone marrow1,2. During fetal development, expansion of HSCs (self-renewal) and production of differentiated haematopoietic cells (differentiation) are both required to sustain the haematopoietic system for body growth3,4. However, it remains unclear how these two seemingly opposing tasks are accomplished within the short embryonic period. Here we used in vivo genetic tracing in mice to analyse the formation of HSCs and progenitors from intra-arterial haematopoietic clusters, which contain HSC precursors and express the transcription factor hepatic leukaemia factor (HLF). Through kinetic study, we observed the simultaneous formation of HSCs and defined progenitors-previously regarded as descendants of HSCs5-from the HLF+ precursor population, followed by prompt formation of the hierarchical haematopoietic population structure in the fetal liver in an HSC-independent manner. The transcription factor EVI1 is heterogeneously expressed within the precursor population, with EVI1hi cells being predominantly localized to intra-embryonic arteries and preferentially giving rise to HSCs. By genetically manipulating EVI1 expression, we were able to alter HSC and progenitor output from precursors in vivo. Using fate tracking, we also demonstrated that fetal HSCs are slowly used to produce short-term HSCs at late gestation. These data suggest that fetal HSCs minimally contribute to the generation of progenitors and functional blood cells before birth. Stem cell-independent pathways during development thus offer a rational strategy for the rapid and simultaneous growth of tissues and stem cell pools.

Hematopoietic Stem Cell-Independent Differentiation of Mast Cells From Mouse Intraembryonic VE-Cadherin+ Cells.

Hematopoietic stem cell (HSC)-independent hematopoiesis from hemogenic endothelial cells (HECs) in the mouse embryo has been recognized as a source of tissue-resident hematopoietic cells in adult mice. Connective tissue mast cells (MCs) have been reported to originate from VE-cadherin (VE-cad)-expressing HECs in the yolk sac and embryo proper (EP) by a VE-cad-Cre-mediated lineage-tracing analysis. However, it remains unclear whether MCs are generated via a conventional HSC-dependent hematopoietic differentiation pathway, or whether through a fast-track pathway bypassing the emergence of HSCs. Here, we investigated whether EP-derived VE-cad+ cells differentiate into MCs independently of HSCs. VE-cad+ cells isolated from the embryonic day (E) 9.5-10.5 EP robustly formed connective tissue-type MCs in a newly established co-culture system using PA6 stromal cells. In contrast, bone marrow (BM) reconstitution assays of cultured cells indicated that E9.5 VE-cad+ cells did not differentiate into transplantable HSCs in this culture condition. Lymphoid-biased HSCs with a limited self-renewal capacity were occasionally detected in some cultures of E10.5 VE-cad+ cells, while MC growth was constantly observed in all cultures examined. HSCs purified from adult BM required a more extended culture period to form MCs in the PA6 co-culture than the embryonic VE-cad+ cells. Furthermore, E9.5-E10.5 VE-cad+ cells contributed to tissue-resident MCs in postnatal mice when transplanted into the peritoneal cavity of newborn mice. These results suggest that EP-derived VE-cad+ cells generate MCs independently of HSC development in vitro and possess the potential of generating connective tissue MCs in vivo, although the exact differentiation program remains unsolved.

The canonical smooth muscle cell marker TAGLN is present in endothelial cells and is involved in angiogenesis.

Elongation of vascular endothelial cells (ECs) is an important process in angiogenesis; however, the molecular mechanisms remain unknown. The actin-crosslinking protein TAGLN (transgelin, also known as SM22 or SM22α) is abundantly expressed in smooth muscle cells (SMCs) and is widely used as a canonical marker for this cell type. In the course of studies using mouse embryonic stem cells (ESCs) carrying an Tagln promoter-driven fluorescence marker, we noticed activation of the Tagln promoter during EC elongation. Tagln promoter activation co-occurred with EC elongation in response to vascular endothelial growth factor A (VEGF-A). Inhibition of phosphoinositide 3-kinase (PI3K)-Akt signaling and mTORC1 also induced EC elongation and Tagln promoter activation. Human umbilical vein endothelial cells (HUVECs) elongated, activated the TAGLN promoter and increased TAGLN transcripts in an angiogenesis model. Genetic disruption of TAGLN augmented angiogenic behaviors of HUVECs, as did the disruption of TAGLN2 and TAGLN3 genes. Tagln expression was found in ECs in mouse embryos. Our results identify TAGLN as a putative regulator of angiogenesis whose expression is activated in elongating ECs. This finding provides insight into the cytoskeletal regulation of EC elongation and an improved understanding of the molecular mechanisms underlying the regulation of angiogenesis.

Bone morphogenetic protein 4 differently promotes distinct VE-cadherin+ precursor stages during the definitive hematopoietic development from embryonic stem cell-derived mesodermal cells.

Definitive hematopoietic cells develop from fetal liver kinase 1 (Flk1)+ mesodermal cells during the in vitro differentiation of mouse embryonic stem cells (ESCs). VE-cadherin+CD41-CD45-(V+41-45-) hemogenic endothelial cells (HECs) and VE-cadherin+CD41+CD45- (V+41+45-) cells mediate the definitive hematopoietic development from Flk1+ cells. Bone morphogenetic protein 4 (BMP4) is known to be essential for the formation of mesoderm. However, the role of BMP4 in differentiation of the VE-cadherin+ definitive hematopoietic precursors from the mesoderm has been elusive. We addressed this issue using a co-aggregation culture of ESC-derived Flk1+ cells with OP9 stromal cells. This culture method induced V+41-45- cells, V+41+45- cells, and CD45+ cells from Flk1+ cells. V+41+45- cells possessed potential for erythromyeloid and T-lymphoid differentiation. When Flk1+ cells were cultured in the presence of a high concentration of BMP4, the generation of V+41-45- cells was enhanced. The increase in V+41-45- cells led to the subsequent increase in V+41+45- and CD45+ cells. The addition of BMP4 also increased hematopoietic colony-forming cells of various lineages. Furthermore, BMP4 promoted the expansion of V+41+45- cells independently of the preceding V+41-45- cell stage. These results suggest that BMP4 has promotive effects on the differentiation of V+41-45- HECs from Flk1+ mesodermal cells and the subsequent proliferation of V+41+45- hematopoietic precursors. These findings may provide insights for establishing a culture system to induce definitive hematopoietic stem cells from ESCs.

Morphology regulation in vascular endothelial cells.

Morphological change in endothelial cells is an initial and crucial step in the process of establishing a functional vascular network. Following or associated with differentiation and proliferation, endothelial cells elongate and assemble into linear cord-like vessels, subsequently forming a perfusable vascular tube. In vivo and in vitro studies have begun to outline the underlying genetic and signaling mechanisms behind endothelial cell morphology regulation. This review focuses on the transcription factors and signaling pathways regulating endothelial cell behavior, involved in morphology, during vascular development.

Dual inhibition of mTORC1 and mTORC2 perturbs cytoskeletal organization and impairs endothelial cell elongation.

Elongation of endothelial cells is an important process in vascular formation and is expected to be a therapeutic target for inhibiting tumor angiogenesis. We have previously demonstrated that inhibition of mTORC1 and mTORC2 impaired endothelial cell elongation, although the mechanism has not been well defined. In this study, we analyzed the effects of the mTORC1-specific inhibitor everolimus and the mTORC1/mTORC2 dual inhibitor KU0063794 on the cytoskeletal organization and morphology of endothelial cell lines. While both inhibitors equally inhibited cell proliferation, KU0063794 specifically caused abnormal accumulation of F-actin and disordered distribution of microtubules, thereby markedly impairing endothelial cell elongation and tube formation. The effects of KU0063794 were phenocopied by paclitaxel treatment, suggesting that KU0063794 might impair endothelial cell morphology through over-stabilization of microtubules. Although mTORC1 is a key signaling molecule in cell proliferation and has been considered a target for preventing angiogenesis, mTORC1 inhibitors have not been sufficient to suppress angiogenesis. Our results suggest that mTORC1/mTORC2 dual inhibition is more effective for anti-angiogenic therapy, as it impairs not only endothelial cell proliferation, but also endothelial cell elongation.

CXCR4 Signaling Negatively Modulates the Bipotential State of Hemogenic Endothelial Cells Derived from Embryonic Stem Cells by Attenuating the Endothelial Potential.

Hemogenic endothelial cells (HECs) are considered to be the origin of hematopoietic stem cells (HSCs). HECs have been identified in differentiating mouse embryonic stem cells (ESCs) as VE-cadherin+ cells with both hematopoietic and endothelial potential in single cells. Although the bipotential state of HECs is a key to cell fate decision toward HSCs, the molecular basis of the regulation of the bipotential state has not been well understood. Here, we report that the CD41+ fraction of CD45- CD31+ VE-cadherin+ endothelial cells (ECs) from mouse ESCs encompasses an enriched HEC population. The CD41+ ECs expressed Runx1, Tal1, Etv2, and Sox17, and contained progenitors for both ECs and hematopoietic cells (HCs) at a high frequency. Clonal analyses of cell differentiation confirmed that one out of five HC progenitors in the CD41+ ECs possessed the bipotential state that led also to EC colony formation. A phenotypically identical cell population was found in mouse embryos, although the potential was more biased to hematopoietic fate with rare bipotential progenitors. ESC-derived bipotential HECs were further enriched in the CD41+ CXCR4+ subpopulation. Stimulation with CXCL12 during the generation of VE-cadherin+ CXCR4+ cells attenuated the EC colony-forming ability, thereby resulted in a decrease of bipotential progenitors in the CD41+ CXCR4+ subpopulation. Our results suggest that CXCL12/CXCR4 signaling negatively modulates the bipotential state of HECs independently of the hematopoietic fate. Identification of signaling molecules controlling the bipotential state is crucial to modulate the HEC differentiation and to induce HSCs from ESCs. Stem Cells 2016;34:2814-2824.

3,5,6,7,8,3',4'-Heptamethoxyflavone reduces interleukin-4 production in the spleen cells of mice.

In our previous studies, we reported anti-inflammatory functions of 3,5,6,7,8,3',4'-heptamethoxyflavone (HMF), which is a polymethoxyflavone rich in various citrus fruits. Here, we investigated the immunomodulatory function of HMF in mice. HMF administration (50 mg/kg, i.p., 2 times/week) tended to reduce the production of antigen-specific IgE induced by ovalbumin in combination with aluminum hydroxide gel. Fluorescence-activated cell sorting analysis revealed the reduction of interleukin-4(+)CD4(+) spleen cells and sustained presence of interferon-γ(+)CD4(+) spleen cells in mice administered HMF, whereas the ratio of CD4(+)CD8(-) versus CD4(-)CD8(+) spleen cells was not affected. Interleukin-4 release from CD3/CD28-stimulated spleen cells of mice administered HMF was reduced, whereas interferon-γ release was not affected. These results suggest that HMF has an immunomodulatory function via reduced interleukin-4 expression.

Inhibition of the PI3K-Akt and mTORC1 signaling pathways promotes the elongation of vascular endothelial cells.

Endothelial cell morphology needs to be properly regulated during angiogenesis. Vascular endothelial growth factor (VEGF) induces endothelial cell elongation, which promotes sprouting of pre-existing vessels. However, therapeutic angiogenesis using VEGF has been hampered by side effects such as elevated vascular permeability. Here, we attempted to induce endothelial cell elongation without an overdose of VEGF. By screening a library of chemical inhibitors, we identified phosphatidylinositol 3-kinase (PI3K)-Akt pathway inhibitors and mammalian target of rapamycin complex 1 (mTORC1) inhibitors as potent inducers of endothelial cell elongation. The elongation required VEGF at a low concentration, which was insufficient to elicit the same effect by itself. The elongation also depended on Foxo1, a transcription factor indispensable for angiogenesis. Interestingly, the Foxo1 dependency of the elongation was overridden by inhibition of mTORC1, but not by PI3K-Akt, under stimulation by a high concentration of VEGF. Dual inhibition of mTORC1 and mTORC2 failed to induce cell elongation, revealing mTORC2 as a positive regulator of elongation. Our findings suggest that the PI3K-Akt-Foxo1 and mTORC1-mTORC2 pathways differentially regulate endothelial cell elongation, depending on the microenvironmental levels of VEGF.

Activin A in combination with OP9 cells facilitates development of Flk-1(+) PDGFRα(-) and Flk-1(+) PDGFRα(+) hematopoietic mesodermal cells from murine embryonic stem cells.

Lateral mesoderm-derived hemogenic endothelial cells are known to originate the definitive hematopoietic lineage in mouse embryogenesis. The developmental process of the definitive hematopoietic lineage can be recapitulated by inducing differentiation of mouse embryonic stem (ES) cells in a co-culture system with OP9 stromal cells. However, the signaling molecules that can modulate the development of the definitive hematopoietic lineage in the OP9 co-culture system have yet to be identified. Here we report that activin A enhanced the hematopoietic potential of endothelial cells derived from ES cells in the OP9 co-culture system. Activin A in combination with OP9 cells augmented development of Flk-1(+) PDGFRα(+) early mesodermal cells and Flk-1(+) PDGFRα(-) lateral mesodermal cells from ES cells. These Flk-1(+) mesodermal cells further differentiated into CD41(+) endothelial cells, which preferentially possessed high hematopoietic potential. Furthermore, Flk-1(+) PDGFRα(+) cells but not Flk-1(+) PDGFRα(-) cells produced hematopoietic progenitors with a bimodal pattern when cultured as an aggregate with OP9 cells. Our results suggest that activin A in combination with OP9 cells facilitates differentiation of ES cells to Flk-1(+) mesodermal cells, which encompass various precursors that separately contribute to the development of hematopoietic lineages.

Determining c-Myb protein levels can isolate functional hematopoietic stem cell subtypes.

The transcription factor c-Myb was originally identified as a transforming oncoprotein encoded by two avian leukemia viruses. Subsequently, through the generation of mouse models that affect its expression, c-Myb has been shown to be a key regulator of hematopoiesis, including having critical roles in hematopoietic stem cells (HSCs). The precise function of c-Myb in HSCs although remains unclear. We have generated a novel c-myb allele in mice that allows direct observation of c-Myb protein levels in single cells. Using this reporter line we demonstrate that subtypes of HSCs can be isolated based upon their respective c-Myb protein expression levels. HSCs expressing low levels of c-Myb protein (c-Myb(low) HSC) appear to represent the most immature, dormant HSCs and they are a predominant component of HSCs that retain bromodeoxyuridine labeling. Hematopoietic stress, induced by 5-fluorouracil ablation, revealed that in this circumstance c-Myb-expressing cells become critical for multilineage repopulation. The discrimination of HSC subpopulations based on c-Myb protein levels is not reflected in the levels of c-myb mRNA, there being no more than a 1.3-fold difference comparing c-Myb(low) and c-Myb(high) HSCs. This illustrates how essential it is to include protein studies when aiming to understand the regulatory networks that control stem cell behavior.

Redefining the in vivo origin of metanephric nephron progenitors enables generation of complex kidney structures from pluripotent stem cells.

Recapitulating three-dimensional (3D) structures of complex organs, such as the kidney, from pluripotent stem cells (PSCs) is a major challenge. Here, we define the developmental origins of the metanephric mesenchyme (MM), which generates most kidney components. Unexpectedly, we find that posteriorly located T(+) MM precursors are developmentally distinct from Osr1(+) ureteric bud progenitors during the postgastrulation stage, and we identify phasic Wnt stimulation and stage-specific growth factor addition as molecular cues that promote their development into the MM. We then use this information to derive MM from PSCs. These progenitors reconstitute the 3D structures of the kidney in vitro, including glomeruli with podocytes and renal tubules with proximal and distal regions and clear lumina. Furthermore, the glomeruli are efficiently vascularized upon transplantation. Thus, by reevaluating the developmental origins of metanephric progenitors, we have provided key insights into kidney specification in vivo and taken important steps toward kidney organogenesis in vitro.

Molecular pathways governing development of vascular endothelial cells from ES/iPS cells.

Assembly of complex vascular networks occurs in numerous biological systems through morphogenetic processes such as vasculogenesis, angiogenesis and vascular remodeling. Pluripotent stem cells such as embryonic stem (ES) and induced pluripotent stem (iPS) cells can differentiate into any cell type, including endothelial cells (ECs), and have been extensively used as in vitro models to analyze molecular mechanisms underlying EC generation and differentiation. The emergence of these promising new approaches suggests that ECs could be used in clinical therapy. Much evidence suggests that ES/iPS cell differentiation into ECs in vitro mimics the in vivo vascular morphogenic process. Through sequential steps of maturation, ECs derived from ES/iPS cells can be further differentiated into arterial, venous, capillary and lymphatic ECs, as well as smooth muscle cells. Here, we review EC development from ES/iPS cells with special attention to molecular pathways functioning in EC specification.

Regulated expression and role of c-Myb in the cardiovascular-directed differentiation of mouse embryonic stem cells.

RATIONALE: c-myb null (knockout) embryonic stem cells (ESC) can differentiate into cardiomyocytes but not contractile smooth muscle cells (SMC) in embryoid bodies (EB). OBJECTIVE: To define the role of c-Myb in SMC differentiation from ESC. METHODS AND RESULTS: In wild-type (WT) EB, high c-Myb levels on days 0-2 of differentiation undergo ubiquitin-mediated proteosomal degradation on days 2.5-3, resurging on days 4-6, without changing c-myb mRNA levels. Activin-A and bone morphogenetic protein 4-induced cardiovascular progenitors were isolated by FACS for expression of vascular endothelial growth factor receptor (VEGFR)2 and platelet-derived growth factor receptor (PDGFR)α. By day 3.75, hematopoesis-capable VEGFR2+ cells were fewer, whereas cardiomyocyte-directed VEGFR2+/PDGFRα+ cells did not differ in abundance in knockout versus WT EB. Importantly, highest and lowest levels of c-Myb were observed in VEGFR2+ and VEGFR2+/PDGFRα+ cells, respectively. Proteosome inhibitor MG132 and lentiviruses enabling inducible expression or knockdown of c-myb were used to regulate c-Myb in WT and knockout EB. These experiments showed that c-Myb promotes expression of VEGFR2 over PDGFRα, with chromatin immunopreciptation and promoter-reporter assays defining specific c-Myb-responsive binding sites in the VEGFR2 promoter. Next, FACS-sorted VEGFR2+ cells expressed highest and lowest levels of SMC- and fibroblast-specific markers, respectively, at days 7-14 after retinoic acid (RA) as compared with VEGFR2+/PDGFRα+ cells. By contrast, VEGFR2+/PDGFRα+ cells cultured without RA beat spontaneously, like cardiomyocytes between days 7 and 14, and expressed cardiac troponin. Notably, RA was required to more fully differentiate SMC from VEGFR2+ cells and completely blocked differentiation of cardiomyocytes from VEGFR2+/PDGFRα+ cells. CONCLUSIONS: c-Myb is tightly regulated by proteosomal degradation during cardiovascular-directed differentiation of ESC, expanding early-stage VEGFR2+ progenitors capable of RA-responsive SMC formation.

Variation in hematopoietic potential of induced pluripotent stem cell lines.

Induced pluripotent stem (iPS) cells were originally generated from somatic cells by ectopic expression of four transcription factor genes: Oct3/4, Sox2, Klf4 and c-Myc. Currently, iPS cell lines differ in tissue origin, the combination of factors used to construct them, the method of gene delivery and expression of pluripotency markers. Thus to evaluate iPS cells for haematotherapy, the hematopoietic potential among iPS lines should be compared. Here, we compare differentiation capacity of six iPS lines into mesodermal cells and hematopoietic cells (HCs) through embryoid body (EB) formation. We show that the mouse embryonic fibroblast (MEF)-derived iPS lines 20D17 and 178B5 resemble CCE ES cells in terms of morphology in culture, number and size of EBs and differentiation capacity into mesodermal cells compared to iPS cells derived from adults, although all iPS lines could form EBs. The number of mesodermal cells differentiated from MEF-derived iPS cell lines showed a 3.9-407-fold increase compared to that from iPS lines derived from adults. Furthermore, 178B5 iPS cells generated Ter119(+) erythroid cells (3.35%) efficiently in culture. We conclude that hematopoietic potential differs among the six lines and that MEF-derived 20D17 and 178B5 iPS cells generate HCs more efficiently than adult-derived iPS cells.

Hematopoiesis from pluripotent stem cell lines.

Embryonic stem cells (ESCs) can differentiate into various types of hematopoietic cells (HPCs) when placed in an appropriate environment. Various methods for the differentiation of ESCs into specific HPC lineages have been developed using mouse ESCs. These ESC-differentiation methods have been utilized also as an in vitro model to investigate hematopoiesis in embryos and they provided critical perceptions into it. These methods have been adapted for use with human ESCs, which have the possibility of being employed in regenerative medicine; further improvement of these methods may lead to the efficient production of HPCs for use in transfusions. The generation of transplantable hematopoietic stem cells is a medical goal that is still difficult to achieve. Recently, induced pluripotent stem (iPS) cells have been established from differentiated cells. Thereby, iPS cells have expanded further possibilities of the use of pluripotent stem cell lines in clinical application. Indeed, iPS cells have been established from cells with disease genes and those which have undergone reprogramming and targeting have generated phenotypically normal HPCs. Here, we mainly summarize the recent progress in research on hematopoiesis conducted with ESCs and iPS cells.

Foxo1 is essential for in vitro vascular formation from embryonic stem cells.

The forkhead transcription factors regulate the correct organization of vascular system. One of them, Foxo1 is an important physiological regulator of endothelial cell morphology in response to VEGF, while underlying mechanisms are largely unknown. In order to elucidate the cellular function of Foxo1, we used a three-dimensional culture system for the differentiation of Flk1-expressing mesodermal precursor cells derived from ES cells to cord forming endothelial cells and associating vascular smooth muscle cells. While Foxo1(+/+) endothelial cells organized into long vessel-like structures associated with smooth muscle cells, Foxo1(-/-) endothelial cells could form only short sprouts. Foxo1(-/-) endothelial cells have punctate accumulation of filamentous actin, thick circumferential bundles of microtubules with small spikes at the tip of cells, and no interaction with smooth muscle cells. Our results suggest the involvement of Foxo1 in cytoskeletal remodeling of endothelial cells and recruitment of smooth muscle cells during vascular development.

Different roles of Foxo1 and Foxo3 in the control of endothelial cell morphology.

Foxo1, a member of the Foxo subfamily of forkhead box transcription factors, is known to be essential for progression of normal vascular development in the mouse embryos. In the cultures of endothelial cells derived from embryonic stem cells, Foxo1-deficient endothelial cells exhibit an abnormal morphological response to vascular endothelial growth factor-A (VEGF-A), which is characterized by a lack of cell elongation, yet the molecular mechanisms governing endothelial cell morphology under angiogenic stimulation remain unknown. Here, we report that transforming growth actor-beta also induces endothelial cell elongation in collaboration with Foxo1 and VEGF-A. Furthermore, tetracycline-regulated induction of Foxo3, another member of the Foxo subfamily, into Foxo1-null endothelial cells failed to restore abnormal morphological response to VEGF-A at an early differentiation stage. In contrast, Foxo1 and Foxo3 exerted the same function at a late differentiation stage, i.e. enhancement of VEGF responsiveness and promotion of cell elongation. Our results provide evidence that endothelial cell morphology is regulated by several mechanisms in which Foxo1 and Foxo3 express distinct functional properties depending on differentiation stages.

Thrombin induces rapid disassembly of claudin-5 from the tight junction of endothelial cells.

The cell-to-cell junction of endothelial cells (ECs) regulates the fence function of the vascular system. Previously we showed that ECs derived from embryonic stem cells (i.e., EECs) develop to form stable endothelial sheets in monolayer cultures. Immunohistochemical analysis revealed that these EECs formed intercellular junctions with the help of vascular endothelial cadherin (VECD) and claudin-5. In this study, we investigated the response of EC sheets to stimuli that are known to increase vascular permeability. While vascular endothelial growth factor A and histamine disrupted the EC junction by enhancing contraction of EECs, thrombin affected specifically the localization of claudin-5 at this junction. We could not detect any significant effect of thrombin on the localization of VECD. Concerning thrombin receptors, EECs expressed protease-activated receptor 1 (PAR1) but not PAR4. Consistent with this expression pattern, PAR1 agonists eliminated claudin-5 as effectively as thrombin itself. This is the first report to show that claudin-5 can be disassembled from the EC junction in a signal-dependent manner and to suggest that claudin-5 mobilization is a cause of PAR1-induced increase in vascular permeability.