PMC:4608092 / 22002-47359 JSONTXT

{"project":"2_test","denotations":[{"id":"26126547-8594589-55294633","span":{"begin":4834,"end":4836},"obj":"8594589"},{"id":"26126547-22615451-55294634","span":{"begin":4968,"end":4970},"obj":"22615451"},{"id":"26126547-12671680-55294635","span":{"begin":4971,"end":4973},"obj":"12671680"},{"id":"26126547-10072525-55294636","span":{"begin":5276,"end":5278},"obj":"10072525"},{"id":"26126547-23946885-55294637","span":{"begin":16740,"end":16742},"obj":"23946885"},{"id":"26126547-24658080-55294638","span":{"begin":16943,"end":16945},"obj":"24658080"},{"id":"26126547-23850618-55294639","span":{"begin":17263,"end":17265},"obj":"23850618"},{"id":"26126547-23850618-55294640","span":{"begin":17290,"end":17292},"obj":"23850618"},{"id":"26126547-23410753-55294641","span":{"begin":17293,"end":17295},"obj":"23410753"},{"id":"26126547-23850618-55294642","span":{"begin":17765,"end":17767},"obj":"23850618"},{"id":"26126547-21799892-55294643","span":{"begin":19575,"end":19576},"obj":"21799892"},{"id":"26126547-24658080-55294644","span":{"begin":20112,"end":20114},"obj":"24658080"},{"id":"26126547-23410753-55294645","span":{"begin":21885,"end":21887},"obj":"23410753"}],"text":"Results\n\nInduction of RNF213 in HeLa Cells and HUVECs\nWe investigated the induction of RNF213 in HeLa cells and HUVECs after treatment of various angiogenic factors (TGF-β, IL-1β, VEGF, and PDGF-BB) and antiangiogenic factors (IFN-α, IFN-β, and IFN-γ). After 24 hours of treatment, none of the factors, except for IFN-γ, changed RNF213 protein levels in HeLa cells (Figure3). However, IFN-γ increased RNF213 protein levels in a dose- and time-dependent manner (Figure3A and 3B) with corresponding increases in mRNA levels (Figure3C and 3D). We then investigated the effects of treatment with the same factors on HUVECs at the 24-hour time point. Whereas TGF-β, IL-1β, VEGF, and PDGF-BB did not change RNF213 protein levels (Figure4A), IFN-β and IFN-γ, but not IFN-α, increased protein levels in a dose-dependent manner (Figure5). IFN-β and IFN-γ also induced RNF213 in a time-dependent manner (Figure4B and 4C). We further examined mRNA levels of RNF213 after treatment with IFN-β and IFN-γ. Both factors increased mRNA levels, but IFN-β increased mRNA levels at a lower dose (0.1 ng/mL) and an earlier time point (3 hours) than IFN-γ (Figure4D through 4G).\nFigure 3 Screening the effects of angiogenic factors and antiangiogenic factors on RNF213 protein and mRNA expression in HeLa cells. A and B, HeLa cells were treated with various concentrations of angiogenic factors and antiangiogenic factors for 24 hours (A), or in the presence of 10 ng/mL of IFN-γ for the indicated times (B), and levels of RNF213 protein expression were examined by western blotting analysis using β-tubulin as a loading control. C and D, HeLa cells were treated with various concentrations of IFN-γ for 24 hours (C) or in the presence of 10 ng/mL of IFN-γ for the indicated times (D). RNA samples were analyzed by qPCR using PPIA as an internal control. Fold induction by IFN-γ was compared with activities of 0 ng/mL of IFN-γ. A column represents a mean of 3 independent experiments. Bars indicate SD. *P\u003c0.05, by Student t test compared with IFN-γ 0 ng/mL. IFN-γ indicates interferon γ; IL, interleukin; PDGF-BB, platelet-derived growth factor; qPCR, quantitative polymerase chain reaction; TGF-β, transforming growth factor β; VEGF, vascular endothelial growth factor.\nFigure 4 Screening effects of angiogenic factors and antiangiogenic factors on RNF213 protein and mRNA expression in HUVECs. A through C, HUVECs were treated with various concentrations of angiogenic factors for 24 hours (A), or in the presence of 10 ng/mL of IFN-β (B) or IFN-γ (C) for the indicated times, and levels of RNF213 protein expression were examined by western blotting analysis using β-tubulin as a loading control. D through G, HUVECs were treated with various concentrations of IFN-β (D) or IFN-γ (F) for 24 hours or in the presence of 10 ng/mL of IFN-β (E) or IFN-γ (G) for the indicated times. RNA samples were analyzed by qPCR using PPIA as an internal control. Fold induction by IFN-β or IFN-γ was compared with activities of 0 ng/mL of IFN-β or IFN-γ. A column represents a mean of 3 independent experiments. Bars indicate SD. *P\u003c0.05, by Student t test compared with 0 ng/mL IFN-β or IFN-γ. HUVECs indicates human umbilical vein endothelial cells; IFN, interferon; IL, interleukin; PDGF-BB, platelet-derived growth factor; qPCR, quantitative polymerase chain reaction; TGF-β, transforming growth factor β; VEGF, vascular endothelial growth factor.\nFigure 5 Effects of IFNs on RNF213 protein expression in HUVECs. HUVECs were treated with various concentrations of IFNs for 24 hours, and RNF213 protein expression was examined by western blotting analysis using β-tubulin as a loading control. Representative western blotting findings are shown in upper panel. A column represents a mean of 3 independent experiments (lower panel). Bars indicate SD. *P\u003c0.05, by Student t test compared with 0 ng/mL IFNs. HUVECs indicates human umbilical vein endothelial cells; IFNs, interferons. These findings indicated that IFN-β increased RNF213 expression levels in HUVECs in a vascular EC-specific manner, whereas IFN-γ did not. The other factors had no effect on expression levels of RNF213. Upregulation of RNF213 protein was preceded by an increase in RNF213 mRNA levels, which suggested that IFN-β increased RNF213 expression at the transcriptional level. Because IFN-β induced RNF213 in HUVECs, we chose IFN-β for further characterization because of its tissue specificity.\n\nInduction of RNF213 by IFN-β Is Mediated by STATx\nWe examined the putative binding sites for transcriptional factors in the promoter region up to 3 kb from the transcriptional start site of the RNF213 gene. We performed a computer search for potential regulatory elements in this region using MatInspector V2.2 at the TRANSFAC website (http://www.gene-regulation.com/pub/databases.html).24 We found a single STATx-binding site at the −514 position (Figure6A). STAT1 is a signaling molecule in the IFN-β-signaling pathway.25,26 Therefore, we conducted promoter assays using a fusion plasmid containing the 3-kb RNF213 promoter region and a luciferase reporter gene. The promoter significantly increased luciferase activity after treatment with IFN-β (Figure6B), whereas disruption of the STATx-binding site by missense mutagenesis27 abrogated promoter activity (Figure6C). Therefore, we conclude that IFN-β upregulates RNF213 through the STATx-binding site in its promoter region.\nFigure 6 Activation of RNF213 promoter activity in HUVECs after IFN-β treatment. A, Constructs for the RNF213 gene promoter-luciferase fusion plasmids and the sequences of the STATx-binding site (potential STATx1-binding site; WT or mutant). Mutated nucleotides are underlined. B, Dose-response effects of IFN-β on RNF213 gene promoter activity. Each column with a bar represents mean±SD of 3 independent experiments. Blue columns represent pGL4.14 vector. Red columns represent RNF213 WT promoter pGL4.14 vector. Fold represents relative mean luciferase activity of RNF213 WT promoter pGL4.14 to PGL4.14. *P\u003c0.05, luciferase activity at 0.1, 1, or 10 ng/mL of IFN-β were compared with 0 ng/mL of IFN-β in cells transfected with RNF213 WT promoter pGL4.14 (red) by Student t test. #P\u003c0.05 luciferase activity in cells transfected with RNF213 WT promoter pGL4.14 (red) was compared with cells transfected with pGL4.14 (blue) at the same IFN-β dose by Student t test. C, Mutation analysis of RNF213 gene promoter activity in HUVECs. HUVECs were transiently transfected with the RNF213 WT promoter pGL4.14 plasmid or RNF213 STATx mutation promoter pGL4.14 plasmid, and luciferase activities were measured after 24 hours of IFN-β treatment. Fold inductions by 10 ng/mL of IFN-β (red) were compared with the activities of IFN-β 0 ng/mL (blue). Each column with a bar represents mean±SD of 3 independent experiments. *P\u003c0.05, by Student t test compared with 0 ng/mL of IFN-β with the same promoter. #P\u003c0.05 by Student t test compared between RNF213 WT promoter and mutation promoter at 10 ng/mL of IFN-β. HUVECs indicates human umbilical vein endothelial cells; IFN, interferon; WT, wild type.\n\nRole of RNF213 in Antiangiogenic Activity of IFN-β\nFirst, we evaluated effects of treatment with IFN-β on HUVEC proliferation, and found that IFN-β did not increase proliferation (Figure7). Angiogenic activity was then evaluated by tube formation and migration assays. We confirmed that IFN-β lowered tube formation and inhibited migration (Figures8 and 9) without affecting proliferation rates (Figure7). Because IFN-β induced RNF213, we hypothesized that antiangiogenic activity of IFN-β is mediated by RNF213. To test this hypothesis, we first depleted STAT1 by siRNA. Antiangiogenic activity of IFN-β, except for branching, was normalized by depletion of STAT1 and phosphorylated STAT1 by siRNA transfection (Figure10A and 10B). Notably, siRNA also downregulated RNF213 (Figure10A). This downregulation was likely mediated by activation of the promoter. We then depleted RNF213 protein levels by siRNA transfection. Depletion of RNF213 restored tube formation and migration (Figures1 and 2). These rescue experiments by siRNA transfection demonstrated that RNF213 was involved in antiangiogenic activity of IFN-β in ECs.\nFigure 7 Effects of IFN-β on HUVEC proliferation as evaluated by trypan blue dye exclusion tests. At 1 day after HUVEC seeding, 0 or 10 ng/mL of IFN-β was added. Data with bars represent mean±SD (n=3). No significant difference (P\u003c0.05) between 0 ng/mL of IFN-β (blue line) and 10 ng/mL of IFN-β (red line) was observed at each time point according to Student t test. HUVEC indicates human umbilical vein endothelial cell; IFN, interferon.\nFigure 8 Antiangiogenic activity of IFN-β in HUVECs. A, Tube formation assays in HUVECs after 20 hours of culture with IFN-β on Matrigel. A concentration of 0 ng/mL of IFN-β was used as a positive control (100%). Scale bars indicate 100 μm. Representative images are shown in left panel. The tubal areas were quantified by imaging analysis (right panel). Data with bars represent mean±SD (n=3). *P\u003c0.05 according to Student t test compared with 0 ng/mL of IFN-β. B, Migration assays for HUVECs after treatment with IFN-β (1 ng/mL). A concentration of 0 ng/mL of IFN-β was used as a control. Scale bars indicate 100 μm. Representative images are shown in left panel. Re-endothelialized areas were quantified by imaging analysis (right panel). Data with bars represent mean±SD (n=3). *P\u003c0.05 according to Student t test compared with 0 ng/mL of IFN-β. HUVECs indicates human umbilical vein endothelial cells; IFN, interferon.\nFigure 9 Tube formation by HUVECs after 20 hours of culture on Matrigel with IFN-β. IFN-β (0 ng/mL) served as the positive control (100%). Total tube length and number of branches were quantified by imaging analysis. Data with bars represent mean±SD (n=3). *P\u003c0.05 according to Student t test compared with 0 ng/mL of IFN-β. HUVECs indicates human umbilical vein endothelial cells; IFN, interferon.\nFigure 10 Effects of STAT1 depletion on RNF213 expression and antiangiogenic activities of IFN-β in HUVECs. A, Western blot analyses of RNF213 and STAT1 protein expressions and STAT1 protein phosphorylation (at Ser727; p-STAT1) in HUVECs treated with IFN-β for 24 hours after control or STAT1 siRNA transfection. β-tubulin served as the loading control. Representative western blot experiments are shown in the upper panel. A column with a bar (lower panel) represents mean±SD (n=3). *P\u003c0.05 according to Student t test compared with 0 ng/mL of IFN-β with the same siRNA treatment. #P\u003c0.05, cells treated with 10 ng/mL of IFN-β were compared between STAT1 siRNA treatment and no siRNA treatment using Student t test. †P\u003c0.05, STAT1 siRNA treatment was compared with control siRNA at 10 ng/mL of IFN-β using Student t test. B, Tube formation assays for HUVECs cultured with IFN-β on Matrigel after control or STAT1 siRNA transfection. Treatment with 0 ng/mL of IFN-β after control siRNA transfection was evaluated as a positive control (100%). Scale bars indicate 100 μm. Representative images are shown in upper panel. The tube area, total tube length, and number of branches were quantified by imaging analysis (lower panel). A column with a bar represents mean±SD (n=3). *P\u003c0.05, according to Student t test comparing 0 ng/mL of IFN-β within the same siRNA treatment paradigm. HUVECs indicates human umbilical vein endothelial cells; IFN, interferon; p-STAT1, phosphorylated signal transducer and activator of transduction 1.\nFigure 11 Effects of RNF213 depletion on antiangiogenic activity of IFN-β in HUVECs. A, Western blot analysis of RNF213 protein expression in HUVECs treated with IFN-β for 24 hours after control or RNF213 siRNA transfection. β-tubulin served as the loading control. Representative western blotting results are shown in left panel. A column with a bar (right panel) represents mean±SD (n=3). *P\u003c0.05, according to Student t test comparing 0 ng/mL of IFN-β within the same siRNA treatment paradigm. #P\u003c0.05, cells exposed to 10 ng/mL of IFN-β were compared between RNF213 siRNA treatment and no treatment using Student t test. †P\u003c0.05, using Student t test comparing control siRNA at 10 ng/mL of IFN-β. B, Tube formation assays for HUVECs cultured with IFN-β on Matrigel after control or RNF213 siRNA transfection. Treatment with 0 ng/mL of IFN-β after control siRNA transfection was used as a positive control (100%). Scale bars indicate 100 μm. Representative images are shown in left panel. Tube area was quantified by imaging analysis (right panel). A column with a bar represents mean±SD (n=3). *P\u003c0.05, according to Student t test comparing 0 ng/mL of IFN-β within the same siRNA treatment paradigm. C, Migration assays for HUVECs treated with IFN-β (1 ng/mL) after control or RNF213 siRNA transfection. Treatment with 0 ng/mL of IFN-β after control siRNA transfection served as the control. Scale bars indicate 100 μm. Representative images are shown in left panel. Re-endothelialized areas were quantified by imaging analysis (right panel). A column with a bar represents mean±SD (n=3). *P\u003c0.05, according to Student t test comparing 0 ng/mL of IFN-β within the same siRNA treatment paradigm. HUVECs indicates human umbilical vein endothelial cells; IFN, interferon.\nFigure 12 Tube formation assays for HUVECs after 20 hours of culture on Matrigel with IFN-β after control or RNF213 siRNA transfection. Treatment with 0 ng/mL of IFN-β after control siRNA transfection served as the positive control (100%). Total tube length and number of branches were quantified by imaging analysis. A column with a bar represents means±SD (n=3). *P\u003c0.05, by Student t test compared with 0 ng/mL of IFN-β within the same siRNA treatment. HUVECs indicates human umbilical vein endothelial cells; IFN, interferon; iPSECs, induced pluripotent stem cell-derived vascular endothelial cells.\n\nEffects of IFN-β on Angiogenic Activity of iPSECs\nWe then investigated effects of IFN-β treatment on tube formation by iPSECs derived from 2 control subjects with the WT RNF213 genotype and 2 MMD patients (homozygous for RNF213 R4810K). Treatment with IFN-β was accompanied by upregulation of RNF213 mRNA expression (Figure3A). Notably, IFN-β treatment lowered tube formation in iPSECs in both of the control subjects and the MMD patients (Figures3B and 4). These data support the notion that IFN-β treatment suppresses angiogenic activity in ECs with induction of RNF213 in humans. In the next step, we investigated the molecular mechanisms of antiangiogenic effects of RNF213.\nFigure 13 Effects of IFN-β on RNF213 protein expression and antiangiogenic activity in iPSECs from control subjects (GG genotype) and MMD patients (AA genotype). A, RNF213 mRNA levels in iPSECs treated with IFN-β for 24 hours. Fold induction attributed to IFN-β was compared with activity at 0 ng/mL of IFN-β. A column with a bar represents mean±SD of 2 controls (blue) or 2 patients (red) in 3 independent experiments. *P\u003c0.05, according to Student t test comparing 0 ng/mL of IFN-β in control (blue) iPSECs at 0.1, 1, or 10 ng/mL. #P\u003c0.05, according to Student t test comparing 0 ng/mL of IFN-β in patient (red) iPSECs at 0.1, 1, or 10 ng/mL. B, Representative photomicrographs (right panel) and quantified tubal area (left panel) of HUVECs and iPSECs treated with IFN-β. Scale bars indicate 100 μm. HUVECs without IFN-β treatment served as the positive control (100%). A column with a bar represents mean±SD of 2 controls or 2 patients in 3 independent experiments. *P\u003c0.05 according to Student t test comparing 0 ng/mL of IFN-β within HUVECs, control iPSECs, or patient iPSECs. HUVECs indicates human umbilical vein endothelial cells; IFN, interferon; iPSECs, induced pluripotent stem cell-derived vascular endothelial cells; MMD, moyamoya disease.\nFigure 14 Tube formation assays for iPSECs after 20 hours of culture on Matrigel with IFN-β. HUVECs without IFN-β treatment served as the control (100%). Total tube length and number of branches were quantified by imaging analysis. A column with a bar represents mean±SD of 2 controls (blue) or 2 patients (red) in 3 independent experiments. *P\u003c0.05, according to Student t test comparing 0 ng/mL of IFN-β within HUVECs, control iPSECs, or patient iPSECs. HUVECs indicates human umbilical vein endothelial cells; IFN, interferon; iPSECs, induced pluripotent stem cell-derived vascular endothelial cells.\n\nLoss of Function of the Walker B Motif in the First AAA+ of RNF213 Lowers Angiogenic Activity Whereas a Deletion Mutation of the First AAA+ Does Not Lower Angiogenic Activity\nMutation of glutamic acid of the Walker B motif (DExxbox) (WEQ) causes loss of function of ATPase hydrolysis.28 The Walker B motif in the first AAA+ of RNF213 is shown to stabilize hexamer formation, whereas deletion of the first AAA+, which results in loss of ATPase activity, does not initiate oligomerization.13 Therefore, we hypothesized that RNF213 WEQ mutation stabilizes oligomers and can cause deleterious effects in ECs by capture in the oligomeric state. We also hypothesized that RNF213 deletion of AAA+, which does not allow formation of oligomers, does not result in any deleterious effects as silencing RNF213 in vitro14 and in Rnf213 null mice.14,17 Based on these hypotheses, we investigated the effects of expression of the vectors RNF213 WT, RNF213 R4810K, RNF213 WEQ, and RNF213 deletion of AAA+ on tube formation and migration of HUVECs (Figure5A). RNF213 WEQ decreased tube formation and migration, similar to RNF213 R4810K (Figure5). In contrast, expression of RNF213 deletion of AAA+ did not lower tube formation or migration, which is in accord with a finding that silencing RNF213 did not impair angiogenesis.14 Expression of RNF213 WT did not decrease angiogenic activity. Therefore, we consider that RNF213 trapped in the oligomeric state may lead to low angiogenic activity.\nFigure 15 Angiogenic activity of RNF213 WT, and R4810K, WEQ, and ΔAAA mutants. A, RNF213 protein expression in HUVECs transiently expressing the RNF213 mutant. HUVECs transfected with the FLAG-RNF213 expression vector (RNF213 WT, RNF213 R4810K, RNF213 WEQ, and RNF213 ΔAAA) were immunoblotted using anti-FLAG antibodies. Empty vector (“vector” in the figure)-transfected HUVECs served as the control. Representative western blotting findings are shown. Similar results were obtained in 3 independent experiments. B, Tube formation assays for HUVECs transiently expressing RNF213 WT, RNF213 R4810K, RNF213 WEQ, and RNF213 ΔAAA. The vector served as the positive control (100%). Scale bars indicate 100 μm. Representative images are shown in left panel. Tube area was quantified by imaging analysis (right panel). A column with a bar represents mean±SD (n=3). *P\u003c0.05, according to Student t test compared with the vector. C, Migration assays for HUVECs transiently expressing RNF213 WT, RNF213 R4810K, RNF213 WEQ, and RNF213 ΔAAA. The vector served as the control. Scale bars indicate 100 μm. Representative images are shown in left panel. Re-endothelialized areas were quantified by imaging analysis (right panel). A column with a bar represents mean±SD (n=3). *P\u003c0.05 according to Student t test compared with the vector. HUVECs indicates human umbilical vein endothelial cells; LPF, low-pass filter; WEQ, Walker B motif; WT, wild-type. We then investigated the effects of RNF213 R4810K on ATPase activity (Figure6). Although we detected ATPase activity in Walker motif in a recombinant fragment containing amino acids from 2319 to 2613,8 we had never determined ATPase activity with an entire 5207 amino acids. In the current study, we determined ATPase activity with the entire RNF213 proteins of WT, R4810K, and deletion AAA+. Surprisingly, RNF213 R4810K resulted in a loss of ATPase activity. Under the present experimental conditions using excessive detergent washing, RNF213 protein may not be able to maintain an oligomeric form. Therefore, we postulate that the results may represent monomeric ATPase activity. RNF213 R4810K can form oligomers, similar to RNF213 WT,13 suggesting that R4810K may stabilize oligomeric states of RNF213 by inhibiting ATP hydrolysis, thereby inhibiting ATPase activity.\nFigure 16 ATPase activity of RNF213 WT, R4810K, and ΔAAA mutants. A, Lysates from EGFP-RNF213–transfected HEK293 cells were IP with anti-GFP agarose. A total volume of 15 μL was subjected to SDS-PAGE followed by GelCode staining. RNF213 proteins fused with EGFP were detected in EGFP-RNF213–transfected cells (arrow heads, “EGFP-RNF213”). Representative SDS-PAGE images are shown. Similar results were obtained in 3 independent experiments. B, ATPase activity of immunoprecipitated extracts was assayed for ATPase activity. Indicated volumes (μL) of IP products were combined with buffer to yield a total volume of 50 μL for ATPase reaction for 30 minutes at room temperature (see details in the text). Phosphate release was measured using the Phosphate Sensor as ATPase activity. Relative activity was calculated based on average activity of EGFP at 0 μL, which was equal to 1. Data with bars represent mean±SD (n=3). *P\u003c0.05 according to Student t test compared with WT at 2 volumes. EGFP indicates enhanced green fluorescent proteins detected in EGFP-transfected cells (arrow head); IgG HC, IgG heavy chain; IgG LC, IgG light chain; IP, immunoprecipitated; Well, sample wells of the gel; WT, wild type.\n\nEffects of Expression of the RNF213 R4810K Ortholog Rnf213 R4757K on Cerebral Angiogenesis In Vivo\nOur in vitro data strongly suggest that RNF213 R4810K lowers angiogenic activity in ECs when it is induced by environmental stimuli, such as IFNs. Therefore, we investigated the effects of ablation or upregulation of RNF213 R4810K in vivo using various genetically modified mice. These mouse strains involved ablation of Rnf213 (KO)17 and Tg mice, which overexpresses Rnf213 (R4757K or WT) in ECs (EC: EC-Mut-Tg and EC-WT-Tg) or SMCs (SMC: SMC-Mu-Tg) (Figure2). Tissue-specific upregulation of RNF213 was confirmed in Tg mice in ECs and SMCs (Figure7). To induce cerebral angiogenesis, mice at 3 weeks of age were exposed to hypoxia (8% O2 for 2 weeks). At the end of exposure, MRA was conducted in 3 mice for each strain. Exposure to hypoxia failed to induce angiogenesis in EC-Mut-Tg mice, whereas upregulation of Rnf213 WT in ECs, upregulation of Rnf213 R4757K in SMCs, and null Rnf213 or WT significantly induced angiogenesis (Figure8). However, we could not find any stenotic lesions, moyamoya vessels, or lesions indicative of cerebral infarction in any mice with different genotypes (Figure9).\nFigure 17 Tissue-specific upregulation of RNF213 in EC-Tg and SMC-Tg mice. Protein extraction from purified ECs from lungs of EC-Mut Tg and WT mice (left panel) and protein extraction from aorta of SMC-Mut Tg and WT mice (right panel) were immunoblotted. β-tubulin served as the loading control. Similar results were obtained from 3 independent experiments. ECs indicates endothelial cells; SMC, smooth muscle cells; Tg, transgenic; WT, wild type.\nFigure 18 Suppressive effect of Rnf213 mutant upregulation in ECs on angiogenesis in vivo. A, Representative images of GLUT-1-stained sections of cerebral cortex of EC-Mut Tg, EC-WT Tg, SMC-Mut Tg, KO, and WT mice under conditions of normoxia (N) and hypoxia (H). B, Quantified result of cerebral microvessels (left panel). A column with a bar represents mean±SD of the number of cerebral microvessels/mm2 from 6 mice per group. In the hypoxia condition, there was a significant difference in the number of cerebral microvessels among 5 genotypes using the nonparametric method, Kruskal–Wallis 1-way ANOVA (P=0.036), but not in the normoxia condition (P=0.41). *P\u003c0.05 according to Mann–Whitney U test compared with normoxia condition. Two-way ANOVA method was conducted for microvessel formation between genotypes and treatment with interaction term. Results and parameter estimates are described in table (right panel). Regression models are described as (microvessels)=(intercept)+α×[Genotype: EC-Mut-Tg]+β×[Genotype: EC-WT-Tg]+γ×[Genotype: SMC-Mut-Tg]+δ×[Genotype: KO]+ε×[Treatment]+ζ×[interaction: EC-Mut-Tg]+η×[interaction: EC-WT-Tg]+θ×[interaction: SMC-Mut-Tg]+ι×[interaction: KO]. Treatment (hypoxia) significantly induced the number of microvessels/mm2 (P\u003c0.001), whereas genotype did not (P=0.09). Interaction between genotype and treatment was significant (P=0.02). The coefficient on EC-Mut-Tg (ζ) was negative, suggesting that this genotype did not increase microvessels (P=0.002). #P\u003c0.05 according to 2-way ANOVA. ECs indicates endothelial cells; GLUT-1, glucose transporter; KO, knockout; SMC, smooth muscle cells; Tg, transgenic; WT, wild type.\nFigure 19 Representative MRI image of brain of EC-Mut Tg, EC-WT Tg, SMC-Mut Tg, KO, and WT mice with hypoxia. MRA (upper panel) represents MRA images. No stenotic lesions and moyamoya vessels were detected in brain. T2 (lower panel) represents T2-weighted images. No infarction was detected in brains. Absence of stenotic lesion, moyamoya vessel, and infarction was also confirmed in other 2 mice in each genotype. EC indicates endothelial cell; KO, knockout; MRA, magnetic resonance angiography; MRI, magnetic resonance imaging; SMC, smooth muscle cells; Tg, transgenic; WT, wild type.\n\nDi"}