Upon stimulation with angiogenic factors, vascular endothelial cells (ECs) secrete a negative-feedback regulator of angiogenesis, vasohibin-1 (VASH1). Because VASH1 lacks a classical signal sequence, it is not clear how ECs secrete VASH1. We isolated a small vasohibin-binding protein (SVBP) composed of 66 amino acids. The level of Svbp mRNA was relatively high in the bone marrow, spleen and testes of mice. In cultured ECs, Vash1 mRNA was induced by VEGF, and Svbp mRNA was expressed constitutively. The interaction between VASH1 and SVBP was confirmed using the BIAcore system and immunoprecipitation analysis. Immunocytochemical analysis revealed that SVBP colocalized with VASH1 in ECs. In polarized epithelial cells, SVBP accumulated on the apical side, whereas VASH1 was present throughout the cells and partially colocalized with SVBP. Transfection of SVBP enhanced VASH1 secretion, whereas knockdown of endogenous SVBP markedly reduced VASH1 secretion. SVBP increased the solubility of VASH1 protein in detergent solution and inhibited the ubiquitylation of VASH1 protein. Moreover, co-transfection of SVBP significantly augmented the inhibitory effect of VASH1 on EC migration. These results indicate that SVBP acts as a secretory chaperone for VASH1 and contributes to the anti-angiogenic activity of VASH1.
Angiogenesis is essential for physiological events, such as development, reproduction, and wound healing. It is also involved in various pathological processes, including tumor growth and ischemic retinopathy (Carmeliet, 2003). Angiogenesis is tightly regulated by stimulators, such as vascular endothelial growth factor (VEGF) and fibroblast growth factor-2 (FGF-2); angiogenesis inhibitors, such as angiostatin, endostatin and thrombospondin-1; and factors that modulate vascular maturation, such as angiopoietin-1 and transforming growth factor-β (Kerbel, 2008; Sato, 2006). Recently, ECs were shown to produce factors that regulate angiogenesis in an autocrine manner. The expression of Delta-like 4 and its cognate receptor Notch1 are detected mainly in tip cells and stalk cells, respectively, and contribute to their regulation during sprouting angiogenesis (Hellstrom et al., 2007; Leslie et al., 2007; Siekmann and Lawson, 2007).
We previously isolated the angiogenesis inhibitor, vasohibin-1 (VASH1), which is produced and secreted by vascular ECs upon stimulation by VEGF and FGF-2 (Hosaka et al., 2009; Kimura et al., 2009; Nasu et al., 2009; Sato et al., 2009; Sato and Sonoda, 2007; Shibuya et al., 2006; Shimizu et al., 2005; Tamaki et al., 2009; Wakusawa et al., 2008; Watanabe et al., 2004; Yamashita et al., 2006; Yoshinaga et al., 2008). VASH1 regulates EC proliferation and migration in an autocrine manner by acting as a negative-feedback regulator of angiogenesis. Purified VASH1 protein has been shown to inhibit the migration and network formation of human umbilical vein endothelial cells (HUVECs) in vitro. The administration of exogenous VASH1 strongly inhibits pathological and physiological angiogenesis without any significant side effects (Hosaka et al., 2009; Kimura et al., 2009; Shen et al., 2006; Watanabe et al., 2004; Yamashita et al., 2006). Most recently, we showed that VASH1 exhibits anti-lymphangiogenic activity and inhibits lymph node metastasis (Heishi et al., 2010). We also isolated VASH2, a homologue of VASH1 that displays more than 50% amino acid similarity to VASH1 (Kimura et al., 2009; Sato and Sonoda, 2007; Shibuya et al., 2006).
VASH1 protein is post-translationally processed into several truncated forms in ECs (Sato and Sonoda, 2007; Sonoda et al., 2006). Each VASH1 protein has at least two proteolytic cleavage sites in each terminus. Biochemical and functional analyses of VASH1 have revealed that some basic residues at the C-terminus are important for heparin binding and anti-angiogenic activity.
To elucidate the biological functions of VASH1 in detail, clarification of how newly synthesized VASH1 is secreted from ECs and acts on ECs is necessary. However, VASH1 protein does not contain a classical signal sequence typical of secreted proteins (Hegde and Kang, 2008). Also, it does not colocalize with the endoplasmic reticulum (ER) marker calnexin, suggesting that VASH1 might be transported and released into the extracellular space via an unconventional secretary pathway (Watanabe et al., 2004). In the present study, we characterized a VASH1 binding partner, which we designated small vasohibin-binding protein (SVBP), and determined a unique chaperone-like function for regulating VASH1 secretion.
Isolation of the VASH1 binding factor
We performed yeast two-hybrid analysis to isolate potential VASH1 binding partner(s). Four strongly positive clones were obtained, each encoding hypothetical protein LOC374969 cDNA. This hypothetical protein is also registered as coiled-coil domain containing 23 (CCDC23; GenBank accession number NP_955374). Because the protein is composed of only 66 amino acids, we named it small vasohibin-binding protein (SVBP). SVBP is highly conserved among mammals, and the amino acids corresponding to residues 32–52 are predicted to form a coiled-coil structure (Fig. 1A). Similarly to VASH1 and VASH2, SVBP does not have any classical signal sequences in the N-terminal region. We prepared recombinant human VASH1 (Watanabe et al., 2004), human VASH2 (Shibuya et al., 2006) and human SVBP proteins (Fig. 1B) and examined their direct interaction using the BIAcore system. SVBP bound strongly to VASH1 (KD=3.1×10−8 M; Fig. 1C). We found that SVBP also bound to the VASH1 homologue, VASH2 (KD =8.7×10−8 M; Fig. 1D).
Expression of Svbp mRNA in various organs and ECs
Because SVBP is a hypothetical protein, we performed Northern blot analysis to determine whether Svbp mRNA was actually expressed. The mouse EC line MS1 expressed a possible band of Svbp mRNA, and the intensity of this band was reduced by three different specific siRNAs (Fig. 2A). We also detected the expression of Svbp mRNA in various mouse organs (Fig. 2B). Svbp mRNA expression was especially high in the bone marrow, spleen and testes. We then evaluated changes in the expression of VASH1 and SVBP in ECs upon VEGF stimulation. As shown in Fig. 2C, a low level of Vash1 mRNA was expressed under basal conditions, and it was induced by VEGF stimulation, as previously reported (Shimizu et al., 2005; Watanabe et al., 2004). By contrast, a high level of Svbp mRNA was expressed, even under basal conditions. These results, together with the steady state expression of SVBP in various organs, suggest that Svbp mRNA is expressed constitutively.
Interaction of SVBP and VASH1 in cultured ECs
We prepared mouse monoclonal antibodies against SVBP peptides (Asp2–Lys13). One of antibodies was able to detect recombinant SVBP protein by western blotting (Fig. 1B). Using this antibody, we tried to detect SVBP protein transiently expressed in MS1 cells. As shown in Fig. 3A, SVBP was detected in both the conditioned medium (CM) and cell lysate. The amount of SVBP in the CM was gradually increased in a time-dependent manner, implying that SVBP is constantly secreted in the medium. To characterize the interaction between VASH1 and SVBP, we performed co-immunoprecipitation followed by western blotting. Endogenous VASH1 was co-immunoprecipitated from both cell lysate and CM by the anti-SVBP antibody (Fig. 3B). We previously reported that the full-length VASH1 protein (44 kDa) is post-translationally truncated, resulting in at least three smaller forms (42 kDa, 36 kDa and 27 kDa) (Sonoda et al., 2006). Western blot analysis revealed that the 42 kDa protein was mainly detected in the cell lysate, whereas the 36 kDa and 27 kDa forms were detected in the CM, and all forms were co-immunoprecipitated with anti-SVBP antibody (Fig. 3B). We obtained similar results using MS1 cells transiently transfected with VASH1 and SVBP expression vectors (data not shown). Using the BIAcore system, we confirmed that SVBP bound to the 36 kDa (77–365) and 27 kDa (77–318) VASH1 proteins (Fig. 3C,D). We previously reported that proteolytic cleavage near Arg29 and Arg76 of VASH1 generates the two major forms of VASH1 (42 kDa and 36 kDa, respectively), and the subsequent proteolytic cleavage removing most of the basic region at the C-terminus generates the 27 kDa protein (Sonoda et al., 2006; Watanabe et al., 2004). These results suggest that de novo produced VASH1 and SVBP interact within cells and are secreted together, and this interaction might persist during the post-translational processing of VASH1.
Colocalization of SVBP and VASH1
Next, we transiently transfected human VASH1 and SVBP cDNAs into MS1 cells and performed immunocytochemical analysis. The intracellular localization of VASH1 and SVBP was determined by immunostaining with VASH1- or SVBP-specific antibodies. SVBP partly colocalized with VASH1 within the cells (Fig. 4A). To further prove that the interaction between SVBP and VASH1 occurs inside the cell, we performed a Duolink in situ proximity ligation assay (PLA). This assay can detect protein–protein interactions using a combination of specific antibodies against them (Yamazaki et al., 2009). As shown in Fig. 4B, positive signals (red) were clearly observed by the treatment with both anti-VASH1 antibody and anti-Flag antibody (panel d), whereas there were no positive signals by the treatment with other combinations (panels a–c), indicating that SVBP and VASH1 do interact within a cell. Because ECs overlap each other in their peripheral cellular region when they are confluent and form a thin cellular sheet, it is very difficult to analyze the polarity as well as the subcellular localization of proteins within cells. Therefore, we used a polarized epithelial cell line (MDCK cells) as a model to observe the subcellular localization of VASH1 and SVBP in relation to cell polarity (Sabath et al., 2008). We confirmed that the tight junction marker ZO-1 was present at the apical side of MDCK cells (Fig. 4C). When VASH1 and SVBP cDNAs were co-transfected into MDCK cells, SVBP localized mainly on the apical side. This apical localization coincided with ZO-1. By contrast, VASH1 was present throughout the cell and partially colocalized with SVBP on the apical side (Fig. 4D). Positive signals of PLA were observed in the apical side of MDCK cells co-transfected with VASH1 and SVBP expression vectors (Fig. 4E), indicating that SVBP and VASH1 binds and colocalizes in the apical side of cells.
SVBP enhances VASH1 solubility and secretion
A previous report showed that VASH1 does not contain any classical signal sequence for translocation to the ER and does not colocalize with the ER marker calnexin (Watanabe et al., 2004). The above evidence that SVBP interacts with VASH1 both intracellularly and extracellularly prompted us to evaluate whether SVBP might affect VASH1 secretion. We transfected equal amounts of VASH1 expression vector and increasing amounts of SVBP expression vector into MS1 cells and found that SVBP enhanced the secretion of VASH1 in a dose-dependent manner (Fig. 5A). This increased secretion of VASH1 by SVBP was confirmed by ELISA, which measured the concentration of the VASH1–SVBP complex (Fig. 5B). Moreover, we confirmed that Brefeldin A, an inhibitor of ER- and Golgi-dependent secretory transport (Nickel, 2007; Seelenmeyer et al., 2003), does not affect the secretion of VASH1 (Fig. 5B), suggesting that VASH1 might be released from ECs via an unconventional secretion pathway (Nickel and Rabouille, 2009). We also observed that SVBP enhanced the secretion of VASH2, which also lacks a signal sequence (Fig. 5C). Furthermore, the interaction between VASH1 and SVBP dramatically changed the ability to extract VASH1 protein into 1% Triton X-100 cell lysis buffer (Fig. 5A). SVBP increased the amount of VASH1 in the soluble fraction, which coincided with a decrease of VASH1 in the insoluble fraction. The solubility of β-actin, an internal control, was not affected by SVBP. When endogenous SVBP was knocked down by siRNA, we observed the opposite results (Fig. 5D). SVBP expression did not affect Vash1 mRNA levels (data not shown), and Svbp siRNA specifically decreased endogenous Svbp mRNA levels with no effect on the levels of Vash1 mRNA (Fig. 5E). Also, Svbp siRNA decreased the ability to extract VASH1 protein into 1% Triton X-100 cell lysis buffer and decreased the secretion of VASH1 into the medium. These results indicate that SVBP has an important role in VASH1 secretion, probably by increasing the solubility of the VASH1 protein.
SVBP prevents ubiquitylation of VASH1
The solubility of some proteins in detergent solution is closely concerned with protein folding regulated by interactions with molecular chaperones (Dou et al., 2003; Muchowski et al., 2000). In general, misfolded or unfolded proteins are either refolded by molecular chaperones or degraded immediately by the ubiquitin proteasome system (UPS) (Hishiya and Takayama, 2008; Muchowski et al., 2000), which is known as an intrinsic protein quality control. We hypothesized that SVBP might contribute to the quality control of VASH1 protein in ECs. Therefore, we examined first whether the UPS affects VASH1 stability in ECs. To detect the total amount of VASH1 protein, we prepared cell lysate using 1% SDS cell lysis buffer, which was different to levels in the experiments performed in Fig. 5. As shown in Fig. 6A,B, the total amount of VASH1 protein was reduced in the presence of Myc–ubiquitin-B (UBB) compared with VASH1 alone. By contrast, this reduction was recovered in a dose-dependent manner by co-transfection with SVBP, implying that the VASH1 protein is stable when associated with SVBP. In the presence of a proteasome inhibitor, MG-132, polyubiquitylated VASH1 proteins were detected as many ladder bands of a higher molecular mass than the 42 kDa form (Fig. 6C). Co-transfection with SVBP dramatically reduced these polyubiquitylated VASH1 proteins. These results suggest that the interaction between SVBP and VASH1 enhances the stability of VASH1 protein by preventing ubiquitylation.
SVBP accelerates VASH1 function
Finally, we examined whether SVBP affects the anti-angiogenic activity of VASH1. The co-expression of SVBP and VASH1 significantly inhibited VEGF-inducible EC migration (Fig. 7A). Stimulation with VEGF (20 ng/ml) increased MS1 migration 2.2-fold relative to control cells. MS1 cells transfected with VASH1 no longer responded to VEGF stimulation. Co-transfection with VASH1 and SVBP downregulated the basal level of EC migration compared with other samples, regardless of VEGF stimulation. We then examined whether this anti-angiogenic activity might be derived from secreted VASH1. The CM from MDCK cells transfected with VASH1 and/or SVBP expression vector was added to HMVECs. As shown in Fig. 7B, VEGF-inducible migration of HMVECs was inhibited by CM containing the VASH1–SVBP complex in a concentration-dependent manner. These results indicate that SVBP regulates the secretion of VASH from ECs and contributes to the anti-angiogenic activity of VASH1.
Secretory proteins generally contain a typical signal sequence composed of successive hydrophobic amino acids that is usually located in the N-terminal region and removed by signal peptidase after the protein passes through the ER membrane (Hegde and Kang, 2008). VASH1 has no such classical secretion signal sequence (Watanabe et al., 2004), and does not colocalize with the ER marker calnexin in ECs (Watanabe et al., 2004). These observations led us to hypothesize that de novo produced VASH1 might be secreted via an unconventional secretory pathway. Here, we identified SVBP as a novel VASH1 binding partner. The association between endogenous SVBP and VASH1 proteins in cell lysate and CM of ECs appeared crucial for VASH1 secretion, because knockdown of SVBP significantly impaired VASH1 secretion. We also showed that SVBP bound to VASH2 and enhanced its secretion from ECs. The expression of VASH1 mRNA is induced by VEGF or FGF-2, mediated by protein kinase C-δ downstream of VEGF receptor-2 (Shimizu et al., 2005). However, we observed a constitutive expression of Svbp mRNA in ECs under basal conditions and in various normal mouse organs. Thus, the scenario of VASH1 secretion might be as follows: SVBP is prepared in advance and accumulates under or on the cell surface. Once VASH1 has been induced, it binds SVBP, which facilitates the secretion of VASH1 (Fig. 8). The molecular mechanism by which VASH1 exhibits anti-angiogenic activity is not fully understood. Co-transfection of VASH1 and SVBP to ECs decreased the basal migration of ECs, suggesting that anti-angiogenic activity of VASH1 might not simply be caused by the blockade of VEGF-mediated signaling in ECs. We predict a putative receptor for VASH1, which transduces specific signals for anti-angiogenesis in ECs. If we could isolate such a receptor, it should help us to characterize the molecular mechanism of VASH activity.
Molecular chaperones comprise several highly conserved families, including heat shock proteins (HSPs), and contribute to post-translational quality control of proteins (Wickner et al., 1999). Although the native structure of a protein is determined principally by its amino acid sequence, the process of folding in vivo often requires the assistance of molecular chaperones, which are often required to maintain the proper conformation of proteins under changing environmental conditions. Misfolded or unfolded proteins exhibit insolubility in detergent solution. Some molecular chaperones can help such proteins to complete the correct folding and recover solubility. For instance, HSP70 and HSP90 are reported to function as molecular chaperones for microtubule-associated protein Tau or endothelial nitric oxide synthase and increase their solubility in detergent solution (Dou et al., 2003; Jiang et al., 2003; Petrucelli et al., 2004). Because SVBP non-covalently binds VASH1 and increases its solubility in 1% Triton X-100 (a nonionic detergent) cell lysis buffer, SVBP might have a chaperone-like role for VASH1. Molecular chaperones, including HSPs, generally accumulate on membranes (Horvath et al., 2008), and our data show that SVBP accumulates on the apical plasma membrane of polarized epithelial cells, supporting a chaperone-like function of SVBP.
Leaderless secretory proteins (e.g. FGF-2, interleukin-1β and galectins) are released from a variety of cell types via unconventional secretory pathways independently of the ER and Golgi (Calderwood et al., 2007; Hughes, 1999; Keller et al., 2008; Nickel, 2005; Nickel and Rabouille, 2009; Piotrowicz et al., 1997; Torrado et al., 2009). Four potential mechanisms of unconventional protein export have been proposed: export across the plasma membrane, export by secretory lysosomes, export through the release of exosomes derived from multivesicular bodies, and export mediated by plasma membrane shedding of microvesicles (Nickel, 2005; Nickel and Rabouille, 2009). Piotrowicz and colleagues previously reported that the molecular chaperone HSP27 binds to FGF-2 in ECs, which facilitates FGF-2 secretion from ECs via an unconventional secretory pathway (Piotrowicz et al., 1997). Importantly, a recent report showed that correctly folded FGF-2 might be required for translocation across the plasma membrane (Torrado et al., 2009). Interestingly, HSPs themselves lack a secretion signal sequence but are released from cells and function as extracellular ligands, giving rise intracellular signaling (Calderwood et al., 2007). However, the detailed mechanism by which molecular chaperones modulate the secretion of leaderless secretory proteins remains obscure. Further study is required to determine whether SVBP is involved in the secretion of other leaderless proteins such as FGF-2.
Recently, the level of Ccdc23 (also known as SVBP) mRNA was shown to be significantly higher in adenomatous polyposis coli than in normal colorectal mucosa (Gaspar et al., 2008), but the expression of VASH1 appears to be restricted to ECs. Thus, the expression of Svbp mRNA might not always associate with the expression of Vash1 mRNA. SVBP might have additional role(s) beyond that of a VASH1 binding partner. This possibility is currently under investigation.
Materials and Methods
Yeast two-hybrid screen
We performed a yeast two-hybrid screen to isolate VASH1 binding partners as previously described (Colland et al., 2004). Full-length human VASH1 (amino acids 1–365) and the human placental random-primed cDNA library were used as bait and prey, respectively. We screened 1.3×108 independent clones. The prey fragments of the positive clones were amplified by polymerase chain reaction (PCR) and sequenced. The resulting sequences were used to identify the corresponding gene in the GenBank database.
Human SVBP cDNA (CCDC23; GenBank accession number NM_199342) was cloned into the NotI and XbaI sites of the p3xFLAG-CMV14 plasmid vector (Sigma) to generate p3xFLAG-SVBP, in which SVBP cDNA was attached to a triple repeat of the FLAG tag sequence at the 3′ end. The DNA fragment containing the SVBP cDNA joined with 3xFLAG was cloned into the SmaI site of baculovirus transfer vector pYNG (Katakura). Human ubiquitin B cDNA (UBB; GenBank accession number NM_018955) was cloned into the EcoRI site of the pCMV Myc plasmid vector (Clontech) to generate pCMV Myc-UBB, in which UBB cDNA was attached to a Myc sequence at the 5′ start site.
Evaluation of molecular interactions using the BIAcore system
Recombinant human SVBP-FLAG protein was expressed in the silkworm, Bombyx mori, as previously reported (Takahashi et al., 2007). Briefly, the constructed transfer vector was mixed with cysteine-proteinase-deleted baculoviral genomic DNA and then co-transfected in BmN cells of B. mori larvae. The resulting recombinant baculovirus was infected in B. mori pupae. Harvested pupae were lysed and the SVBP-FLAG protein purified using an anti-FLAG M2 affinity column (Sigma). Purified recombinant full-length VASH1, its truncated forms and VASH2 proteins were prepared as previously described (Shibuya et al., 2006; Sonoda et al., 2006; Watanabe et al., 2004). The interaction of VASH1 and VASH2 with SVBP was analyzed using a BIAcore 3000 (BIAcore AB). SVBP (20.6 μg/ml) in 10 mM sodium acetate (pH 4.5) was immobilized on a CM5 sensor chip using the amine-coupling method according to the manufacturer's protocol. VASH1, its truncated forms, or VASH2 in 10 mM HEPES (pH 7.4), 0.15 M NaCl, 3 mM EDTA, 0.005% Surfactant P20 (pH 7.4) at concentrations of 62.5 nM to 1 μM was passed over the surface of the sensor chip at a flow rate of 20 μl/minute. The interaction was monitored as the change in surface plasmon resonance response at 25°C. After 2 minutes of monitoring, the same buffer was introduced to the sensor chip in place of the VASH1 solution to start the dissociation. The sensor surface was regenerated with 10 mM glycine (pH 2.0) at the end of each experiment. Both the association rate constant (Ka) and the dissociation rate constant (Kd) were calculated according to the BIAevaluation software. The dissociation constant (KD) was determined as Kd/Ka.
The murine EC line MS1 was cultured in α minimal essential medium (αMEM; Wako Pure Chemical) supplemented with 10% fetal bovine serum (FBS; JRH Biosciences) (Arbiser et al., 1997). Polarized epithelial Madin-Darby canine kidney (MDCK) cells were cultured in Dulbecco's modified Eagle medium (Wako Pure Chemical) supplemented with 10% FBS (Martin-Belmonte et al., 2001). Human umbilical vein endothelial cells (HUVECs) and human microvascular endothelial cells (HMVECs) were purchased from KURABO and cultured on type I-C collagen-coated dishes (Asahi Techno Glass) in endothelial basal medium containing 5% FBS and EC growth supplements (Cambrex Bio Science Walkersville).
Total RNA was extracted from cultured cells using the RNeasy Mini Kit (Qiagen). RT-PCR was carried out using the First Strand cDNA Synthesis Kit (Roche Diagnostics) according to the manufacturer's instructions. Briefly, total RNA was reverse transcribed for 50 minutes at 42°C using oligo(dT) primers. PCR was performed using sets of primers specific for the target genes. Thermal cycler conditions were 20–30 cycles of 94°C for 15 seconds for denaturing, 56°C for 30 seconds for annealing and 72°C for 45 seconds for extension. PCR products were separated on a 1.5% agarose gel and visualized under ultraviolet by ethidium bromide staining. The primer pairs were as follows: mouse Ccdc23, sense (5′-ATGGATCCACCTGCCCGGAA-3′) and antisense (5′-TCACTCCCCAGGCGGCTGCA-3′); mouse Vash1, sense (5′-ATGTGGAAGGCATGTGGCCAAG-3′) and antisense (5′-CACCCGGATCTGGTACCCACT-3′).
Total RNA was isolated from cultured cells or mouse organs using QIAzol reagent (Qiagen), and mRNA was isolated from total RNA samples using Oligotex-dT30 Super (Takara) according to the manufacturers' instructions. Each mRNA sample (600 ng) was separated by 1% agarose-formaldehyde gel electrophoresis and transferred to a positively charged nylon membrane (Roche Diagnostics). Membranes were hybridized with antisense SVBP- or β-actin-specific RNA probes labeled with digoxigenin (DIG) using the SP6/T7 DIG RNA Labeling Kit (Roche Diagnostics). Hybridization was performed for 2 hours at 68°C using PerfectHyb reagent (Toyobo) containing 0.2 ng/ml DIG-labeled antisense RNA probe, After hybridization, the membranes were washed twice with 2× SSC, 0.1% sodium dodecyl sulfate (SDS) at 68°C for 15 minutes. The final wash was performed twice using 0.1× SSC, 0.1% SDS at 68°C for 15 minutes. The mRNA bands were detected using the Northern Blot Starter Kit (Roche Diagnostics), and the results were analyzed using a LAS-4000 (Fuji Photo Film).
Small interfering RNAs (siRNAs)
Three sets of specific siRNAs targeting mouse Svbp mRNA (mouse CCDC23; GenBank accession number NM_024462) were designed and synthesized by Invitrogen. The target sequences were: Svbp siRNA A, 5′-GGGUCAGAGCUAACCAAGAAGCAUU-3′; Svbp siRNA B, 5′-AAGAAGCAUUCAGAAGCCAAACCAU-3′; Svbp siRNA C, 5′-GAGAUCUAUGCUCUCAACAGAGUCA-3′. Stealth RNAi Negative Control Low GC Duplex (Invitrogen) was used as a negative control. Cells were cultured to 60–70% confluency and transfected with Svbp siRNAs using Lipofectamine RNAiMAX (Invitrogen) according to the manufacturer's instructions. The medium was changed after 3 hours of transfection and the cells cultured for an additional 24–48 hours.
Preparation of anti-SVBP monoclonal antibodies
A/J mice (Japan SLC) were immunized three times with Cys-SVBP peptide (Asp2–Lys13 of the CCDC23 protein) conjugated to keyhole limpet hemocyanin. Spleen cells from the immunized mice were fused with the myeloma cell line P3U1. Culture supernatant from each hybridoma was screened as previously described (Ohta et al., 1999; Watanabe et al., 2004). Positive hybridomas were cloned using the limiting-dilution technique. The resulting monoclonal antibodies (mAbs) were purified from ascites using protein-A–Sepharose beads (GE Healthcare Bio-Sciences).
Cells were lysed in 1% SDS cell lysis buffer consisting of 10 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1% SDS supplemented with 0.5 mM phenylmethylsulfonyl fluoride and Complete EDTA-free Protease Inhibitor Cocktail (Roche Diagnostics). The CM was concentrated 50-fold using an Amicon Ultra-15 (10,000 MWCO; Millipore). Equal amounts of protein from cell lysates and concentrated CM were separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto polyvinylidene fluoride (PVDF) membranes (Bio-Rad). Blotting was performed according to standard procedures. The primary antibodies were anti-human VASH1 mAb (Watanabe et al., 2004), anti-human VASH2 mAb (Shibuya et al., 2006), anti-human SVBP mAb, anti-FLAG M2 mAb (Sigma), and anti-Myc mAb (Clontech). Normalization was performed using anti-β-actin mAb (Sigma). Immunoreactive protein bands were detected using ECL Western Blotting Detection Reagents (GE Healthcare) or Immobilon Western HRP Substrate (Millipore) using an LAS-4000 (Fuji Photo Film). Densitometric quantification was performed using Multi Gauge Ver 3.0 software (Fuji Photo Film).
Extraction of soluble and insoluble VASH1 proteins from cultured cells
Cells were lysed in ice-cold 1% Triton X-100 cell lysis buffer consisting of 10 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1% sodium deoxycholate, 0.1% SDS, and 1% Triton X-100. Cell extracts were placed on ice for 1 hour and then centrifuged at 15,000 g for 15 minutes to separate the supernatant (soluble fraction) and pellet (insoluble fraction). The insoluble fraction was washed three times with ice-cold 1% Triton X-100 cell lysis buffer and re-extracted with sample buffer for SDS-PAGE. Western blotting was carried out as described above.
The cell lysate or CM was incubated with anti-human VASH1 antibody, anti-human SVBP antibody, or normal mouse IgG (Santa Cruz Biotechnology) at 4°C overnight on a rotating mixer. Protein-A–Sepharose beads were then added and the samples incubated for another 2 hours at 4°C. The beads were washed three times with phosphate-buffered saline (PBS) containing 0.05% Tween 20. Immunoprecipitated proteins were eluted into sample buffer for SDS-PAGE, and western blotting was performed as described above.
Transfected MS1 or MDCK cells were fixed in 4% paraformaldehyde, permeabilized with 0.1% Triton-X100, and stained with VASH1, SVBP, FLAG (Sigma), ZO-1 (Sanko Junyaku) antibodies, or Phalloidin-FITC (Sigma). Nuclei were visualized using TO-PRO®3 (Invitrogen). All incubations were performed at 4°C in PBS containing 1% bovine serum albumin. Images were captured using a Fluoview FV1000 confocal microscope system (Olympus).
Duolink in situ proximity ligation assay (PLA)
The Duolink in situ PLA kit was purchased from Olink Bioscience. Transfected MS1 cells and MDCK cells were immunoreacted with the combination of anti-human VASH1 mAb (mouse), anti-FLAG plyclonal Ab (rabbit), normal mouse IgG, and normal rabbit IgG as described above. According to the manufacturer's protocol, a pair of anti-mouse MINUS and anti-rabbit PLUS was used to generate positive fluorescence signals indicating that two PLA probes bound in close proximity (<40 nm). Nuclei counterstaining using TO-PRO®3 and the images were captured as described above.
ELISA for VASH1-SVBP complex
We established a highly sensitive ELISA system that could quantify the VASH1–SVBP complex. A pair of specific monoclonal antibody against SVBP and VASH1 was used to coat a 96-well plate and for HRP labeling, respectively. MS1 cells and MDCK cells were transiently co-transfected with a combination of human VASH1 and SVBP expression vectors for 6 hours, then were washed with serum free α-MEM, and cultured in 0.1% FBS in α-MEM for 24 hours. The CMs were centrifuged at 5000 g for 15 minutes to remove cell debris. Subsequently, the supernatant was subjected to ELISA. The detailed procedure for the measurements carried out is described elsewhere (Heishi et al., 2010).
Endothelial cell migration
The migratory activity of ECs was measured by modified Boyden chamber assay in two different procedures (Kobayashi et al., 2006; Watanabe et al., 2004). First, transfected MS1 cells were plated on the upper chambers (inserts) of the Boyden chamber (8.0 μm pore size, Corning). The lower chamber was filled with 0.2% FBS in α-MEM with or without 20 ng/ml recombinant human VEGF (Sigma). After incubation for 18 hours, cells that migrated to the lower surface of the membrane were fixed with 4% formaldehyde, stained with crystal violet (Sigma), and counted in a blinded manner. The relative number of migrating cells was calculated; the mean number of migrating cells transfected with empty vector (mock) alone was set equal to 1.0.
Another procedure was prepared to evaluate anti-angiogenic activity of the VASH1–SVBP complex secreted from cells. MDCK cells were transiently co-transfected with a combination of human VASH1 and SVBP expression vectors for 6 hours, washed with serum free α-MEM and then cultured in 0.05% FBS in α-MEM for 24 hours. CM was concentrated 10-fold using an Amicon Ultra-15. HMVECs, preincubated in 0.5% FBS in Medium-199 (Invitrogen) for 16 hours, were suspended in concentrated CMs and then plated on the upper chambers of the Boyden chamber. The lower chamber was filled with 0.5% FBS in Medium-199 with 20 ng/ml recombinant human VEGF. HMVECs were allowed to migrate under VEGF stimulation for 4 hours. The number of cells that migrated across the filter was counted in nine fields per insert in a blinded manner.
Data are expressed as mean ± s.d. Significance was assessed by one-way analysis of variance (ANOVA) followed by Sheffe's t-test.
This study was supported by the Grant-in-Aid for Scientific Research on Priority Areas (17014006) and the Grant-in-Aid for Young Scientists (19790509) from the Ministry of Education, Culture, Sports, Science, and Technology of Japan.