Isolation and Characterization of Vasohibin-2 as a Homologue of VEGF-Inducible Endothelium-Derived Angiogenesis Inhibitor Vasohibin

【摘要】  Objective- We recently isolated vasohibin, a novel vascular endothelial growth factor (VEGF)-inducible endothelium-derived angiogenesis inhibitor. Our aim is to find DNA sequences homologous to vasohibin and determine their expression profile.

Methods and Results- By the search of DNA sequences in the database, we found one homologous gene and designated it vasohibin-2. Overall amino acid sequence homology between the 50%. Vasohibin-2 exhibited antiangiogenic activity. Vasohibin-2 expression in cultured endothelial cells was low and not inducible by the stimulation that induced vasohibin-1. However, the immunohistochemical analysis revealed that vasohibin-1 and -2 were diffusely expressed in endothelial cells in embryonic organs during mid-gestation. After that time point, vasohibin-1 and -2 became faint, but persisted to a certain extent in arterial endothelial cells from late gestation to neonate. Expression of vasohibin-1 and -2 could be augmented in vivo by local transfection with the VEGF gene in the embryonic brain or by cutaneous wounding in adult mice.

Conclusion- These results suggest that vasohibin-2, in combination with vasohibin-1, forms a novel family of angiogenesis inhibitors.

We found vasohibin-2 as a homologue of VEGF-inducible endothelium-derived angiogenesis inhibitor vasohibin. Vasohibin-2 exhibited antiangiogenic activity. Immunohistochemical analysis revealed that 2 vasohibins were ubiquitously expressed in endothelial cells in developing embryonic organs during mid-gestation. Vasohibin-2, in combination with vasohibin-1, forms a novel family of angiogenesis inhibitors.

【关键词】  angiogenesis inhibitor endothelial cells vascular development vasohibin

Introduction

Angiogenesis, the formation of new blood vessels from pre-existing vessels, is essential for normal development but is also involved in various pathophysiological conditions. 1 Although angiogenesis is thought to be controlled by the local balance between pro-angiogenic factors and angiogenesis inhibitors, 2 little is known about the negative feedback regulation of angiogenesis. We recently isolated a gene for a candidate negative feedback angiogenesis regulator that was preferentially expressed in endothelial cells (ECs), and designated vasohibin. 3 Vasohibin was isolated from a cDNA microarray analysis to search for vascular endothelial growth factor (VEGF)-inducible genes in human umbilical vein endothelial cells (HUVECs). 4

Generally, growth factors and cytokines belong to the families composed of homologous proteins. There are a number of examples of this among angiogenesis regulators, e.g., VEGF family members 5-11 that play essential roles in angiogenesis or lymphangiogenesis; 12 angiopoietin family members 13-16 that regulate vascular stability as agonists or antagonists of the tyrosine kinase with Ig and endothelial growth factor (EGF) homology domains (TIE)-2 receptor; fibroblast growth factor (FGF) family members, 17 a large group that plays important roles in angiogenesis as well as in other areas; and thrombospondin family members 18 that participate in various functions including inhibition of angiogenesis.

Based on the sequence of the prototype vasohibin (vasohibin-1), we searched the NCBI database for homologous DNA sequences. Here we report one vasohibin-1 paralogue in mammals with antiangiogenic activity, which we have designated vasohibin-2. A number of other angiogenesis inhibitors have already been reported, such as angiostatin, 19 endostatin, 20 16 kDa prolactin, 21 pigment epithelium-derived factor, 22 thrombospondins, 23 PF-4, 24 and IP-10. 25 In terms of origin, vasohibin-1 is intrinsic to ECs, whereas most of other angiogenesis inhibitors are extrinsic to ECs. Therefore, we examined whether vasohibin-2 shared this property with vasohibin-1. Our analysis revealed that both vasohibin-1 and -2 were expressed in ECs. Thus, we propose that vasohibin-2 forms part of an inhibitory angiogenesis regulation system in combination with vasohibin-1.

Materials and Methods

Cell Culture

HUVECs and human aortic endothelial cells were obtained from KURABO and were routinely cultured on type I collagen-coated dishes (IWAKI) in endothelial basal medium containing endothelial cell growth supplements and 10% fetal bovine serum.

Reverse Transcriptase-Polymerase Chain Reaction Analysis and Cloning of Human Vasohibin-2 cDNA

Total RNA was extracted from HUVECs by an acid guanidinium-phenol-chloroform method using ISOGEN (Nippon Gene, Toyama, Japan). First-strand cDNA was generated from total RNA using a 1st Strand cDNA Synthesis Kit (Roche Diagnostics, Mannheim, Germany). Polymerase chain reaction (PCR) was performed in DNA thermal cycler (TAKARA, Tokyo, Japan) with 5 U Taq DNA polymerase per 100 µL reaction mixture. For cloning and quantification of human vasohibin-2, the primer pair 5'-GCAGCCATGAAGCCAACCGT-3' (sense, Exon 1) and 5'-CTCTTGCTGGCCGGTATGGG-3' (antisense, Exon 8) was synthesized. For quantification of human vasohibin-1, primer pairs for human vasohibin-1 (sense; 5'-AGATCCCCATACCGAGTGTG-3' and antisense; 5'-GGGCCTCTTTGGTCATTTCC-3'), and human ß-actin control (sense; 5'-ACAATGAGCTGCGTGTGGCT-3' and antisense; 5'-TCTCCTTAAT GTCACGCACGA-3') were synthesized. PCR products were cloned into PCR2.1 TOPO vector (Invitrogen, Carlsbad, Calif) and directly sequencing using a DNA sequencing kit (Applied Biosystems, Foster City, Calif) and an ABI Prism 310 DNA sequencer (Applied Biosystems).

Preparation of Recombinant Vasohibin-1 and Vasohibin-2 Proteins

Recombinant vasohibin-1 protein was described previously. 3 For the preparation of vasohibin-2, we cloned the cDNA of an N-terminal-truncated form of vasohibin-2 (66-355) into the expression vector p3xFLAG-CMV14 (Sigma, St. Louis, Mo). The FLAG-tagged vasohibin-2 protein was expressed using the Bac-To-Bac Baculovirus expression system (Invitrogen). We purified the protein using an anti-FLAG M2 affinity column (Sigma).

Mouse Corneal Micropocket Assay

The mouse corneal micropocket assay were performed as described previously. 3,26 Briefly, 0.3 µg hydron pellets (IFN Sciences, New Brunswick, NJ) containing vehicle or 80 ng FGF-2 were implanted into the corneas of male BALB/c mice under pentobarbital anesthesia. A total of 5 ng vasohibin-1 or vasohibin-2 protein was added to the pellets. After 7 days, the corneal vessels were photographed, and neovascular area (mm 2 ) was determined using the software package National Institutes of Health (NIH) Image.

Network Formation by HUVECs

Ice-cold growth factor reduced Matrigel was poured into a 35-mm dish and allowed to gel for 1 hour at 37°C. Thereafter, HUVECs (5 x 10 4 ) were plated onto Matrigel and incubated with the vasohibin-2 (10 nM) or vehicle (20 mmol/L Tris/HCl, 0.1% NP40, 0.1 mol/L Glycine, pH 7.0) in medium199 (M-199) containing 5% fetal bovine serum. After 9-hour incubation, cells were photographed.

Preparation of Antibodies

Anti-human vasohibin-2 monoclonal antibody (mAb) was established using a human vasohibin-2 peptide (Ser206-Lys218) and keyhole limpet hemocyanin conjugate as described previously. 3 The resultant mAb 5E3 (IgG1, subclass ) was purified by protein A-agarose affinity chromatography (MAPS II kit, Bio-Rad Laboratories Inc, Hercules, Calif).

Anti-mouse vasohibin-1 and -2 polyclonal antibodies (Abs) were prepared using Cys-mouse vasohibin-1 peptide (Gly296-Arg309) or Cys-mouse vasohibin-2 peptide (Leu230-Arg242) conjugated with maleimide-activated keyhole limpet hemocyanin. Rabbits were immunized subcutaneously with mouse vasohibin-1 or mouse vasohibin-2 peptide conjugates. The first injection consisted of 200 µg of conjugate in an equal volume of Freund?s complete adjuvant. The rabbits were boosted with 100 µg conjugate in equal volumes of incomplete Freund?s adjuvant.

Northern Blot Analysis

Human embryonic message HUNTER was obtained from GenoTechnology (St. Louis, Mo). Northern blotting was performed as described previously. 3 Briefly, the membranes were hybridized with 32 P-labeled human vasohibin-2 cDNA probe in hybridization solution at 42°C for 24 hours. After hybridization, membranes were washed once with 0.1% SDS in 2 x standard saline citrate (SSC) at room temperature and then 3 times with 0.1% SDS in 0.2 x SSC at 65°C. Autoradiography was performed using an imaging plate and the results analyzed by a FLA2000 Image Analyzer (Fuji Film, Tokyo, Japan).

Immunohistochemical Analysis

Human tissue specimens were obtained after obtaining informed consent and ethical committee approval from Tohoku University. Specimens were fixed in 10% formalin, embedded in paraffin, and cut into 3-µm-thick sections. Sections were dewaxed in xylene, rehydrated in a graded ethanol series (100%, 90%, 80%, and 70%), and bleached in 0.3% H 2 O 2 /methanol. Sections were then incubated in 0.1% Trypsin/0.05 mol/L Tris buffer (pH 7.6) for 30 minutes at 37°C for PECAM-1 staining or incubated in citrate buffer (pH 6.0) for 5 minutes at 120°C for vasohibin-1 and vasohibin-2 staining. Sections were blocked with 1% bovine serum albumin (BSA)/PBS for 30 minutes at room temperature. Primary antibody reactions were performed overnight at 4°C at a dilution of 1:20 for anti-human PECAM-1 mAb (Dako Cytomation, Kyoto, Japan), and 1:400 for anti-human vasohibin-1 mAb and anti-human vasohibin-2 mAb. For all specimens, secondary antibody reactions were performed using biotinylated goat anti-mouse Ab (Vector Laboratories Inc., Burlingame, Calif) at a dilution of 1:500 for 30 minutes at room temperature. Vectastatin ABC kit (Vector Laboratories Inc) was used according to the manufacturer ? s instructions. Sections were visualized using 3,3'-Diaminobenzidine (DAB) (Sigma)/0.03%H 2 O 2 in 0.05 mol/L Tris buffer (pH7.6) and counterstained with hematoxylin.

All the animal studies were approved by the committee for animal study in our institute. Mouse tissues were fixed in 4% paraformaldehyde/PBS overnight at 4°C, dehydrated in a graded ethanol series (70%, 80%, 90%, and 100%) and xylene, and embedded in paraffin. Sections were dewaxed, rehydrated, and bleached as described. For PECAM-1 and vasohibin-2 staining, sections were incubated in Proteinase K (Dako Cytomation) at room temperature for 5 minutes. For vasohibin-1 and von Willebrand factor (vWF) staining, sections were incubated in citrate buffer (pH 6.0) for 5 minutes at 120°C. Primary antibody reactions were performed at a dilution of 1:20 for rat anti-mouse PECAM-1 mAb (Research Diagnosis Inc), 1:400 for rabbit anti-mouse vasohibin-1 Ab and rabbit anti-mouse vasohibin-2 Ab, or 1:100 for rabbit anti-human vWF Ab (Dako Cycomation) overnight at 4°C. Secondary antibody reactions were performed at a dilution of 1:500 for PECAM-1 (biotinylated goat anti-rat Ab, Vector Laboratories Inc), or 1:500 for vasohibin-1, vasohibin-2, and vWF (biotinylated goat anti-rabbit Ab, Vector Laboratories Inc) at room temperature for 30 minutes.

Quantitative Real-Time Reverse-Transcriptase PCR

Total RNA was extracted using ISOGEN. First-strand cDNA was generated using a first-strand cDNA synthesis kit for reverse-transcriptase (RT)-PCR. Quantitative real-time RT-PCR was performed using a Light Cycler System (Roche Diagnostics Corp). PCR conditions consisted of an initial denaturation step at 95°C for 10 minutes, followed by 40 cycles of 15 seconds at 95°C, 5 seconds at an annealing temperature (as noted), and 15 seconds at 72°C. The primer pairs used were: mouse ß-actin forward 5-TCGTGCGTGACATCAAAGAG and reverse 5-TGGACAGTGAGGCCAGGATG; mouse vasohibin-1 forward 5-AGCACAGAGAGATGAGGAAC and reverse 5-CGTCGTCGGCTGGAAAGTAG (annealing temperature 68°C); mouse vasohibin-2 forward 5'-CGGCCCGTGAGCCTCGCCAC and reverse 5' CATGGACAACCTGTAGTTTG (annealing temperature 68°C); and mouse VEGF forward 5' TCTGCTCTCTTGGGTGCACT and reverse 5' TTCCGGTGAGAGGTCTGGTT (annealing temperature 66°C). Each mRNA level was measured as a fluorescent signal corrected according to the signal for ß-actin.

VEGF Plasmid Construction

Total RNA was extracted from OP9, 27 a murine bone marrow-derived stromal cell line, using ISOGEN. First-strand cDNA was generated from total RNA using a 1st Strand cDNA Synthesis Kit for RT-PCR. Primers were designed to amplify full-length mouse VEGF 164 (sense ATGAACTTTCTGCTCTCTCTTG, antisense: TCACCGCCTTGGCTTGTCAC). PCR was performed using the Advantage 2 PCR system (BD Biosciences Clontech Palo Alto, Calif). PCR products were subcloned into the PCR2.1 TOPO vector (Invitrogen, Carlsbad, Calif), and then inserted into the pCAX expression vector containing the chicken ß-actin promoter and CMV-IE enhancer 28 (a generous gift from Dr N. Osumi, Tohoku University Postgraduate School of Medicine).

Mouse Whole Embryo Culture

Mouse whole embryo culture and electroporation were performed according to the method described by Osumi and Inoue. 29 Briefly, time-pregnant Institute of Cancer Research mice at embryonic day 10.5 (E10.5) were euthanized. Embryos were freed from the uterine wall, deciduas, and Reichert?s membrane. The yolk sac and chorioallantoic placenta were left intact. Embryos were then transferred into culture bottles that contained 2.5 mL 100% rat serum (Charles River, Yokohama, Japan), 2 mg/mL glucose, and penicillin/streptomycin solution (Invitrogen) at a dilution of 1:400. Culture bottles were set in a whole embryo culture system (Model RKI10-0310, Ikemoto Co, Tokyo, Japan) that rotated the culture bottles at 20 rpm in 95% O 2. After an initial 2-hour culture, embryos were transferred to a chamber-type electrode, and 0.5 µL plasmid solution (at 10 µg/µL) was injected into the brain using a microcapillary tube (Narishige Scientific Instrument Laboratory., Tokyo, Japan). Electroporation was performed using an electroporator CUY21 (Tokiwa Science, Tokyo, Japan) at 70 V voltage, 50 ms duration, for 5 times with a 1-second interval. Embryos were then cultured for 20 hours.

Mouse Wound Healing Model

The back of the mouse was shaved and disinfected with 70% ethanol. A 6-mm-diameter circle was drawn on the skin of the mid-dorsal region and a full-thickness wound created by excision of the area with curved scissors. After 5 days, mice were euthanized and the skin obtained to determine vasohibin-1 and -2 expression.

Calculations and Statistical Analysis

The statistical significance of differences was evaluated by unpaired analysis of variance (ANOVA), and probability values were calculated by Student t test. P <0.05 was considered statistically significant.

Results

Isolation of Vasohibin Homologue

Searching the NCBI database for DNA sequences homologous to human vasohibin-1 revealed the gene FLJ12505 and related cDNAs. FLJ12505 was composed of 8 exons and 7 introns located on human chromosome 1q32.3 ( Figure 1 A). We designated this gene as vasohibin-2. The related cDNAs appeared to be alternative splicing forms of FLJ12505 (transcript 1: AK022567, transcript 2: BC051856 ) ( Figure 1 A). Transcripts 1 and 2 differed by the lack of 200 bp in the mid region of exon 2 in transcript 1, and the lack of exon 4 and exon 5 in transcript 2. For transcript 1, the missing 200 bp of exon 2 resulted in an altered translation start codon, such that its open reading frame started 200 bp downstream from that of transcript 2 ( Figure 1 A).

Figure 1. Human vasohibin-2 gene and its alternatively spliced forms. A, Two alternative splicing forms were identified in the NCBI database (transcript 1 and transcript 2). The structure of transcript 3 identified in HUVECs is shown at the bottom. White triangles indicate start codons and closed triangles indicate stop codons. Arrows indicate primers. F, forward; R, reverse. B, Expression of vasohibin-2 in HUVECs was shown by RT-PCR. Arrow indicates a transcript of 1300 bp.

Because vasohibin-1 was isolated as a VEGF-inducible gene in HUVECs, 3,4 we investigated whether vasohibin-2 transcripts were also present in HUVECs. We designed PCR primer pairs that amplify both transcripts 1 and 2 ( Figure 1 A). Theoretically, transcripts 1 and 2 would amplify 1196 and 1189 bp RT-PCR products, respectively. However, RT-PCR revealed only a single transcript of 1300 bp in HUVECs ( Figure 1 B). Direct sequencing showed that this RT-PCR product was composed of 8 exons including a complete exon 2 ( Figure 1 A). We entered this cDNA sequence into the NCBI database as AY834202. Of the 3 possible alternative splicing forms of vasohibin-2, transcript 3 exhibited the highest similarity to vasohibin-1. The deduced amino acid sequence of transcript 3 compared with that of vasohibin-1 with asterisks indicating identical amino acids (Figure I, available online at http://atvb.ahajournals.org). Based on the human vasohibin-2 cDNA sequence, we identified a mouse orthologue of vasohibin-2 located on chromosome 1H6. The human and mouse vasohibin family members and the percent identity between the sequences are summarized (Table I, available online at http://atvb.ahajournals.org).

Antiangiogenic Activity of Vasohibin-2

We performed a mouse corneal micropocket assay to examine whether vasohibin-2 exhibited antiangiogenic activity. Vasohibin-2 protein inhibited angiogenesis in the cornea ( Figure 2 A). Like vasohibin-1, vasohibin-2 protein alone had no effect on the cornea (data not shown). Quantitative analysis revealed that the antiangiogenic activity of vasohibin-2 was significant and almost equivalent to that of vasohibin-1 ( Figure 2 B). To examine whether vasohibin-2 acts directly on ECs, we plated HUVECs on Matrigel and treated them with vasohibin-2. Vasohibin-2 apparently inhibited network formation by HUVECs ( Figure 2 C).

Figure 2. Antiangiogenic activity of vasohibin-2. A, A mouse corneal micropocket assay was performed as described in Materials and Methods. B, Quantification of vascular area (mm 2 ) was performed as described in Materials and Methods. Mean and standard deviation (SD) shown (n=4). C, Network formation by HUVECs. HUVECs were plated on Matrigel in the absence or presence of vasohibin-2 (10 nmol).

Expression of Vasohibin-1 and Vasohibin-2

We analyzed vasohibin-2 expression. RT-PCR analysis showed that basal expression of vasohibin-2 was very low in both HUVECs and human aortic endothelial cells, and was not inducible by stimulators that induced vasohibin-1 ( Figure 3 A). We then characterized the mRNA expression in vivo. We have previously shown that vasohibin-1 expression is abundant in human embryonic tissues. 3 Northern blotting revealed vasohibin-2 mRNA expression in various human embryonic organs at 6 to 12 embryonic weeks ( Figure 3 B). Thus, the expression of vasohibin-2 mRNA was evident in human embryonic tissues.

Figure 3. Expression of vasohibin-2 mRNA in vitro and in vivo. A, HUVECs or human aortic endothelial cells cultured on type I collagen-coated dishes were starved in 1% fetal bovine serum /M-199 for 12 hours, then stimulated with 1 nmol/L VEGF 165, 2 nmol/L FGF-2, or 50 nmol/L phorbol 12-myristate 13-acetate (PMA) for 24 hours. Thereafter, vasohibin expression was analyzed by RT-PCR. B, Northern blot analysis of human embryonic tissues was performed as described in Materials and Methods. C, Total RNA was extracted from whole embryos at the indicated embryonic periods and analyzed by quantitative RT-PCR. mRNA levels of vasohibin-1 and vasohibin-2 were standardized according to ß-actin expression level, and mean relative expression and SD are shown (n=3).

We compared mRNA levels of vasohibin-1 and -2 in mice embryos at various gestational periods by real-time RT-PCR. When standardized with respect to ß-actin, expression of vasohibin-2 was twice that of vasohibin-1 ( Figure 3 C), with both showing peak levels at E14.5. In contrast to vasohibin-1 and -2, VEGF expression reached maximum levels by E14.5 but showed sustained expression to E18.5 (data not shown).

Vasohibin-1 and Vasohibin-2 Protein Localization In Vivo

We next investigated the localization of vasohibin-1 and -2 proteins in vivo. We confirmed the specificity of our mAbs and revealed that both proteins were selectively expressed in PECAM-1-positive ECs (Figure II, available online at http://atvb.ahajournals.org). We then analyzed various human embryonic organs at different gestational stages. Both vasohibin-1 and -2 proteins were present in vessels from 20-week human embryonic organs ( Figure 4A and 4 C). Whereas vasohibin-1 staining became noticeably faint, vasohibin-2 staining was retained, especially in ECs from large vessels in neonate ( Figure 4B and 4 D). However, careful observation revealed that vasohibin-1 was also present in arterial ECs in neonatal samples (Figure III, available online at http://atvb.ahajournals.org).

Figure 4. Localization of vasohibin-1 and -2 proteins in human embryonic tissues. Immunostaining of human 20-week embryonic lung (A, vasohibin-1; C, vasohibin-2) or neonatal lung (B, vasohibin-1; D, vasohibin-2). Scale bar, 100 µm.

We further performed immunohistochemical analysis of mouse embryonic tissues. We confirmed the specificity of our anti-mouse vasohibin-1 and vosohibin-2 polyclonal Abs and found that mouse vasohibin-1 and -2 proteins were present in ECs (Figure IV, available online at http://atvb.ahajournals.org). Vasohibin-1 and -2 proteins were diffusely present in ECs at mid-gestation (E10.5) ( Figure 5A to 5 C) but became faint thereafter (E16.5) except in arterial ECs (Figure V, available online at http://atvb.ahajournals.org).

Figure 5. Localization of vasohibin-1 and -2 proteins in mouse early embryonic tissues. A, PECAM-1 immunostaining performed on E10.5 mouse embryonic brain. B, Vasohibin-1 immunostaining performed on the same sample. C, Vasohibin-2 immunostaining performed on the same sample. Scale bar 100 µm. Arrows indicate positively stained vessels.

Inducible Expression of Vasohibin-1 and Vasohibin-2 In Vivo

Whereas vasohibin-2 expression was very low and not inducible in vitro, vasohibin-2 was clearly expressed in ECs of developing embryonic organs. We therefore examined the inducibility of vasohibin-2 in mice using a whole mouse embryo culture system. Genes for VEGF or GFP were transfected into mouse embryonic brains. Twenty hours after the transfection, neovessels were evident in brains transfected with VEGF but not with GFP ( Figure 6A and 6 B). We determined mRNA levels of vasohibin-1 and -2 in the brain by real-time RT-PCR. Results showed the increase of vasohibin-1 and -2 mRNA in VEGF-transfected brain ( Figure 6 C). We also applied a postnatal mouse skin wound healing model. Granulation became evident at 5 days after injury, at which point the mRNA expression of both vasohibin-1 and -2 was augmented (Figure VI, available online at http://atvb.ahajournals.org). Immunohistochemical analyses revealed that vasohibin-1 and -2 proteins were present in ECs of the granulation tissue (Figure VI).

Figure 6. Inducible expression of vasohibin-1 and -2 in mouse embryo. Gene transfection and mouse embryonic culture were performed as described in Materials and Methods. A, Representative image of a GFP expression vector-transfected mouse embryo. B, Representative image of a VEGF expression vector-transfected mouse embryo. Arrowheads indicate neovascularization. C, Total RNA was extracted from mouse embryonic brains, and mRNA expression of vasohibin-1 and -2 was compared between GFP- and VEGF-transfected mouse embryos (n=3). White bars, GFP-transfected mouse embryos; blue bar, vasohibin-1 mRNA in VEGF-transfected mouse embryos; purple bar, vasohibin-2 mRNA in VEGF-transfected mouse embryos. mRNA levels were standardized to that of ß-actin, and mean relative expression level and SD are shown. (n=3).

Discussion

Here we identified a gene that showed significant homology to the vasohibin-1 gene and termed it vasohibin-2. Whereas alternative splicing variants of this gene have already been entered into the database, our present study identified a novel transcript with 8 complete exons. This transcript, termed transcript 3, was the only vasohibin-2 transcript detected in normal tissues by our hands (data not shown), and was also the transcript that showed the highest homology to vasohibin-1. Thus, although the significance of other transcripts remains unclear, we concluded that transcript 3 represented the principal isoform of vasohibin-2.

Similar to vasohibin-1, 3 we could not identify any classical signal sequences or known functional motifs in the primary structure of vasohibin-2. Nevertheless, vasohibin-2 protein was present in the culture medium when COS7 cells were transfected with this gene (data not shown). Similar to vasohibin-1, 3 vasohibin-2 protein inhibited network formation by HUVECs ( Figure 2 C), but it did not exhibit any effect on the migration of human smooth muscle cells (data not shown). Moreover, vasohibin-2 protein inhibited angiogenesis with a potency equivalent to that of vasohibin-1 ( Figure 2 B). We therefore propose that vasohibin-1 and -2 form a novel family of endogenous angiogenesis inhibitors. It is not known whether or not vasohibin-1 and -2 share the common signaling pathway for angiogenesis inhibition. Identification of the receptors for vasohibin-1 and -2 in the future should answer this question.

Although vasohibin-2 expression was very low and was not inducible in vitro, vasohibin-2 was clearly expressed in vivo. The reason for this discrepancy remains unclear. It is possible that as yet unknown factor(s) present in vivo are required to induce vasohibin-2 in ECs. Nonetheless, the spatial and temporal expression pattern of vasohibin-2 in vivo was akin to that of vasohibin-1. Indeed, vasohibin-2 expression increased when angiogenesis was induced in embryonic brain or postnatal skin in mice. These results suggest the involvement of vasohibin-1 and -2 in the control of angiogenesis.

Immunohistochemical analysis confirmed that both vasohibin-1 and -2 proteins were localized to ECs. Interestingly, immunostaining of vasohibin-2 was more intense than that of vasohibin-1. Although the expression levels of different proteins cannot be directly compared by immunostaining, results from real-time RT-PCR in mice suggested that vasohibin-2 expression was higher than that of vasohibin-1. The precise quantification of the two proteins awaits future analysis.

Whereas high levels of VEGF persisted from E14.5 to E18.5 in mice embryos, vasohibin-1 and -2 levels peaked at E14.5 and declined thereafter. However, vasohibin-1 and -2 proteins persisted to some extent in arterial ECs in late gestation and through to the neonate. It will be interesting to determine whether the vasohibin proteins in arterial ECs exhibit any functions beyond antiangiogenesis.

Acknowledgments

This work was supported by the 21st Century COE Program Special Research Grant "The Center for Innovative Therapeutic Development for Common Diseases" and by a grant-in-aid for Scientific Research on Priority Areas of the Japanese Ministry of Education, Science, Sports, and Culture from the Ministry of Education, Science, Sports, and Culture of Japan.

【参考文献】
  Carmeliet P. Angiogenesis in health and disease. Nat Med. 2003; 9: 653-660.

Iruela-Arispe ML, Dvorak HF. Angiogenesis: a dynamic balance of stimulators and inhibitors. Thromb Haemost. 1997; 78: 672-677.

Watanabe K, Hasegawa Y, Yamashita H, Shimizu K, Ding Y, Abe M, Ohta H, Imagawa K, Hojo K, Maki H, Sonoda H, Sato Y. Vasohibin as an endothelium-derived negative feedback regulator of angiogenesis. J Clin Invest. 2004; 114: 898-907.

Abe M, Sato Y. cDNA microarray analysis of the gene expression profile of VEGF-activated human umbilical vein endothelial cells. Angiogenesis. 2001; 4: 289-298.

Ferrara N, Davis-Smyth T. The biology of vascular endothelial growth factor. Endocr Rev. 1997; 18: 4-25.

Neufeld G, Cohen T, Gengrinovitch S, Poltorak Z. Vascular endothelial growth factor (VEGF) and its receptors. Faseb J. 1999; 13: 9-22.

Makinen T, Olofsson B, Karpanen T, Hellman U, Soker S, Klagsbrun M, Eriksson U, Alitalo K. Differential binding of vascular endothelial growth factor B splice and proteolytic isoforms to neuropilin-1. J Biol Chem. 1999; 274: 21217-21222.

Silvestre JS, Tamarat R, Ebrahimian TG, Le-Roux A, Clergue M, Emmanuel F, Duriez M, Schwartz B, Branellec D, Levy BI. Vascular endothelial growth factor-B promotes in vivo angiogenesis. Circ Res. 2003; 93: 114-123.

Karkkainen MJ, Haiko P, Sainio K, Partanen J, Taipale J, Petrova TV, Jeltsch M, Jackson DG, Talikka M, Rauvala H, Betsholtz C, Alitalo K. Vascular endothelial growth factor C is required for sprouting of the first lymphatic vessels from embryonic veins. Nat Immunol. 2004; 5: 74-80.

Rissanen TT, Markkanen JE, Gruchala M, Heikura T, Puranen A, Kettunen MI, Kholova I, Kauppinen RA, Achen MG, Stacker SA, Alitalo K, Yla-Herttuala S. VEGF-D is the strongest angiogenic and lymphangiogenic effector among VEGFs delivered into skeletal muscle via adenoviruses. Circ Res. 2003; 92: 1098-1106.

Pipp F, Heil M, Issbrucker K, Ziegelhoeffer T, Martin S, van den Heuvel J, Weich H, Fernandez B, Golomb G, Carmeliet P, Schaper W, Clauss M. VEGFR-1-selective VEGF homologue PlGF is arteriogenic: evidence for a monocyte-mediated mechanism. Circ Res. 2003; 92: 378-385.

Makinen T, Veikkola T, Mustjoki S, Karpanen T, Catimel B, Nice EC, Wise L, Mercer A, Kowalski H, Kerjaschki D, Stacker SA, Achen MG, Alitalo K. Isolated lymphatic endothelial cells transduce growth, survival and migratory signals via the VEGF-C/D receptor VEGFR-3. Embo J. 2001; 20: 4762-4773.

Davis S, Aldrich TH, Jones PF, Acheson A, Compton DL, Jain V, Ryan TE, Bruno J, Radziejewski C, Maisonpierre PC, Yancopoulos GD. Isolation of angiopoietin-1, a ligand for the TIE2 receptor, by secretion-trap expression cloning. Cell. 1996; 87: 1161-1169.

Maisonpierre PC, Suri C, Jones PF, Bartunkova S, Wiegand SJ, Radziejewski C, Compton D, McClain J, Aldrich TH, Papadopoulos N, Daly TJ, Davis S, Sato TN, Yancopoulos GD. Angiopoietin-2, a natural antagonist for Tie2 that disrupts in vivo angiogenesis. Science. 1997; 277: 55-60.

Hasegawa Y, Abe M, Yamazaki T, Niizeki O, Shiiba K, Sasaki I, Sato Y. Transcriptional regulation of human angiopoietin-2 by transcription factor Ets-1. Biochem Biophys Res Commun. 2004; 316: 52-58.

Lee HJ, Cho CH, Hwang SJ, Choi HH, Kim KT, Ahn SY, Kim JH, Oh JL, Lee GM, Koh GY. Biological characterization of angiopoietin-3 and angiopoietin-4. Faseb J. 2004; 18: 1200-1208.

Fahmy RG, Dass CR, Sun LQ, Chesterman CN, Khachigian LM. Transcription factor Egr-1 supports FGF-dependent angiogenesis during neovascularization and tumor growth. Nat Med. 2003; 9: 1026-1032.

Lawler J. The functions of thrombospondin-1 and-2. Curr Opin Cell Biol. 2000; 12: 634-640.

O?Reilly MS, Holmgren L, Shing Y, Chen C, Rosenthal RA, Moses M, Lane WS, Cao Y, Sage EH, Folkman J. Angiostatin: a novel angiogenesis inhibitor that mediates the suppression of metastases by a Lewis lung carcinoma. Cell. 1994; 79: 315-328.

O?Reilly MS, Boehm T, Shing Y, Fukai N, Vasios G, Lane WS, Flynn E, Birkhead JR, Olsen BR, Folkman J. Endostatin: an endogenous inhibitor of angiogenesis and tumor growth. Cell. 1997; 88: 277-285.

Clapp C, Martial JA, Guzman RC, Rentier-Delure F, Weiner RI. The 16-kilodalton N-terminal fragment of human prolactin is a potent inhibitor of angiogenesis. Endocrinology. 1993; 133: 1292-1299.

Dawson DW, Volpert OV, Gillis P, Crawford SE, Xu H, Benedict W, Bouck NP. Pigment epithelium-derived factor: a potent inhibitor of angiogenesis. Science. 1999; 285: 245-248.

Jimenez B, Volpert OV, Crawford SE, Febbraio M, Silverstein RL, Bouck N. Signals leading to apoptosis-dependent inhibition of neovascularization by thrombospondin-1. Nat Med. 2000; 6: 41-48.

Maione TE, Gray GS, Petro J, Hunt AJ, Donner AL, Bauer SI, Carson HF, Sharpe RJ. Inhibition of angiogenesis by recombinant human platelet factor-4 and related peptides. Science. 1990; 247: 77-79.

Angiolillo AL, Sgadari C, Taub DD, Liao F, Farber JM, Maheshwari S, Kleinman HK, Reaman GH, Tosato G. Human interferon-inducible protein 10 is a potent inhibitor of angiogenesis in vivo. J Exp Med. 1995; 182: 155-162.

Ogawa S, Yoshida S, Ono M, Onoue H, Ito Y, Ishibashi T, Inomata H, Kuwano M. Induction of macrophage inflammatory protein-1alpha and vascular endothelial growth factor during inflammatory neovascularization in the mouse cornea. Angiogenesis. 1999; 3: 327-334.

Nakagawa T, Abe M, Yamazaki T, Miyashita H, Niwa H, Kokubun S, Sato Y. HEX acts as a negative regulator of angiogenesis by modulating the expression of angiogenesis-related gene in endothelial cells in vitro. Arterioscler Thromb Vasc Biol. 2003; 23: 231-237.

Takahashi M, Osumi N. Pax6 regulates specification of ventral neurone subtypes in the hindbrain by establishing progenitor domains. Development. 2002; 129: 1327-1338.

Osumi N, Inoue T. Gene transfer into cultured mammalian embryos by electroporation. Methods. 2001; 24: 35-42.

作者单位：Department of Vascular Biology (T.S., K.W., H.Y., K.S., H.M., M.A., Y.S.), Institute of Development, Aging, and Cancer, Tohoku University, Sendai, Japan; the Department of Pathology (T.M.), Tohoku University Hospital, Sendai, Japan; Diagnostics Science Division (H.O., H.S.), Shionogi & Co, Ltd,