Cultured endothelial cells display endogenous activation of the canonical Wnt signaling pathway and express multiple ligands, receptors, and secreted modulators of Wnt signaling

Authors


Abstract

A growing body of evidence implicates Wnt signaling in the control of angiogenesis. To better understand the role of the Wnt/β-catenin pathway in endothelial cells (EC), we examined endogenous signaling activity and signaling component expression in vascular cells. We observed stabilization of cytosolic β-catenin and activation of a T-cell factor (TCF) -luciferase promoter, hallmarks of canonical Wnt signaling activity, in cultured EC. This activity was increased in subconfluent EC, which are known to display characteristics of angiogenic EC, compared with confluent EC, which have a more differentiated phenotype. Endogenous TCF activity was inhibited by transfection with a secreted inhibitor of canonical Wnt signaling. A systematic analysis of Wnt, Fzd, SFRP, and Dkk gene expression in human EC (cultured and freshly isolated), smooth muscle cells (cultured), and aorta demonstrated that numerous Wnt signaling components are expressed by vascular cells. We conclude that Wnt signaling components are expressed and active in cultured EC. Developmental Dynamics 235:3110–3120, 2006. © 2006 Wiley-Liss, Inc.

INTRODUCTION

Angiogenesis, the formation of new blood vessels from a pre-existing vascular network, is a critical process in embryonic development and contributes to pathologies such as cancer and diabetic retinopathy. Angiogenic vessel growth involves proliferation and migration of endothelial cells (EC) to form a new vessel sprout, followed by lumen formation and recruitment of support cells, including pericytes or smooth muscle cells (Darland and D'Amore,2001). Many signaling pathways have been implicated in the control of angiogenesis, and recent evidence indicates that Wnt signaling, too, is important for vessel growth (Goodwin and D'Amore,2002).

There are multiple pathways by which Wnt ligands can signal, but the Wnt/β-catenin, or canonical, pathway is best understood and has been most clearly implicated in promoting cell proliferation and differentiation (Reya and Clevers,2005). Nineteen Wnt ligands, 10 Fzd receptors, and 2 low density lipoprotein (LDL) receptor-related protein (LRP) coreceptors have been identified in mammals, as have numerous secreted modulators of the secreted Frizzled-related protein (SFRP), Dickkopf (Dkk), and Wnt-inhibitory factor families (Miller,2002); many of these components participate in the Wnt/β-catenin pathway. In the absence of a canonical Wnt signal, cytosolic β-catenin is rapidly phosphorylated, ubiquitinated, and degraded. Wnt-Fzd binding in the presence of LRP5 or LRP6 results in inactivation of the β-catenin phosphorylation complex, so that β-catenin accumulates in the cytosol. β-catenin then translocates to the nucleus, binds transcription factors of the T-cell factor (TCF) family and promotes transcription of target genes (Miller,2002). Wnt/β-catenin signaling affects proliferation, differentiation, and survival of a variety of cell types, including endothelial progenitor cells (Choi et al.,2004).

β-catenin stabilization and TCF activity have been observed in blood vessels in vivo during both developmental and pathologic angiogenesis (Blankesteijn et al.,2000; Eberhart et al.,2000; Yano et al.,2000a,b; Eberhart and Argani,2001). In mice with lacZ expression regulated by TCF, β-galactosidase activity is observed in angiogenic vessels of the developing nervous system (Maretto et al.,2003), and vascular defects have been observed in mice deficient for Wnt2 (Monkley et al.,1996), Wnt7b (Shu et al.,2002), Fzd4 (Hsieh et al.,2005), and Fzd5 (Ishikawa et al.,2001).

Many in vitro systems are commonly used to model angiogenic EC behavior. While Wnt/β-catenin signaling clearly participates in angiogenesis in vivo, its role in endogenous EC activity has not been fully explored. We first assessed Wnt/β-catenin signaling activity in confluent vs. subconfluent EC and determined whether extracellular Wnt signals were responsible for observed activity. We next identified potential mediators of Wnt signaling in vascular cells. Certain Wnts and Fzd have been identified in cultured EC (Wright et al.,1999; Cheng et al.,2003; Masckauchan et al.,2005), but a comprehensive analysis of the membrane-associated and secreted signaling components has yet to be reported. We, therefore, conducted a systematic analysis of expression of all known Wnts, Fzd, and secreted modulators in human vascular cells, namely umbilical vein EC (HUVEC; freshly isolated and cultured), dermal EC (HMVEC; neonatal and adult), aortic smooth muscle cells (HASMC), and aortic tissue.

RESULTS

β-Catenin Accumulates in Sparse EC More Than in Confluent EC

Confluent EC in culture display many characteristics of mature endothelial monolayers in vivo, whereas subconfluent EC are reminiscent of angiogenic EC (Pauly et al.,1992). We, therefore, examined β-catenin localization in both confluent and sparse EC populations. Immunohistochemistry of bovine aortic EC (BAEC) revealed that β-catenin was more clearly localized to the cytoplasm and nucleus in sparse cells, whereas in confluent cells β-catenin was primarily localized to the intercellular junctions (Fig. 1A). Immunoblotting of cytosolic cell fractions demonstrated that cytosolic β-catenin levels in sparse BAEC were significantly higher than in confluent cells (Fig. 1B). This pattern of β-catenin is consistent with increased activation of the Wnt/β-catenin pathway in subconfluent cells compared with confluent EC.

Figure 1.

Comparison of β-catenin localization in sparse or migrating vs. confluent endothelial cells. A: β-catenin was immunolocalized in bovine aortic endothelial cells (BAEC) cultured at low density or to confluence. B: Cytosolic fractions of BAEC cultured at low density or to confluence were immunoblotted for β-catenin and α-tubulin. Densitometric analysis of β-catenin and α-tubulin expression in cytosolic fractions of BAEC was conducted for three independent experiments; mean densitometric units are shown. Error bars indicate standard deviation. *P < 0.05. C: Confluent BAEC monolayers were scraped with a pipet tip and cultured for 24 hr in growth medium. Immunohistochemical analysis of β-catenin location was conducted and the cells photographed at a site distant from the scrape wound (left panel) and at the site of injury (right panel). Scale bars = 50 μm in A, 100 μm in C.

BAEC Migrating Into a Scrape Wound Display Cytosolic and Nuclear β-Catenin

To further examine the differences in β-catenin accumulation between subconfluent and confluent EC populations, we conducted a scrape wound assay. In this model, a region of a confluent EC monolayer is removed, and over the following days cells from intact areas of the monolayer migrate and proliferate to fill the scrape wound (Herman,1993). One day after wounding, β-catenin was clearly observed in the cytoplasm and nucleus of cells migrating into the scraped area (Fig. 1C, right panel). In BAEC located distant from the wound site, β-catenin was localized primarily at the intercellular junctions (Fig. 1C, left panel). These observations suggest that the Wnt/β-catenin pathway is activated not only in sparsely plated, proliferating cells, but also in EC that are migrating.

Cultured EC Display Endogenous TCF Activity That Is Inhibited by SFRP1

We next determined whether the β-catenin accumulation observed in cultured EC correlates with an increase in TCF activity. The TOPflash reporter construct, in which expression of the firefly luciferase gene is regulated by six tandem TCF binding sites upstream of a minimal TK promoter (Korinek et al.,1997), was used to measure TCF activity. We transfected subconfluent BAEC with a reporter for activity of TCF, a transcription factor downstream of β-catenin in the canonical Wnt signaling pathway, and either a control plasmid or a plasmid encoding SFRP1, a Wnt-binding inhibitor of canonical signaling. TCF activity was indeed detected in cultured EC, and this endogenous activity was reduced by 61% in the presence of SFRP1 (P < 0.0001, Fig. 2A). An immortalized porcine aortic EC line (PAEC) was also transfected with these constructs, with similar results (Fig. 2B). These findings indicate that EC in culture constitutively signal by means of the Wnt/β-catenin pathway, and that this signaling activity can be antagonized by a Wnt-binding protein. Similar results were observed for bovine lung microvascular EC and bovine aortic smooth muscle cells (BASMC; data not shown), indicating that endogenous Wnt signaling activity occurs in a variety of cultured vascular cells.

Figure 2.

Effect of secreted Frizzled-related protein-1 (SFRP1) on endogenous T-cell factor (TCF) activity of cultured endothelial cells. A,B: Bovine aortic endothelial cells (BAEC; A) and porcine aortic endothelial cells (PAEC; B) were transfected with TOPflash, a TCF-responsive firefly luciferase reporter plasmid; pRL-TK, a Renilla luciferase plasmid used for normalization; and either pCDNA3.1 or pCDNA-SFRP1. Luciferase activity was analyzed 24 hr posttransfection. Firefly luciferase readings were divided by Renilla luciferase readings to normalize for transfection efficiency. This experiment was conducted three times with similar results; mean TOPflash/RL ratios for a representative are shown. Error bars indicate standard deviation. *P < 0.001, **P < 0.0001.

BAEC Express Wnts and Fzd Both In Vivo and In Vitro

A reverse transcriptase-polymerase chain reaction (RT-PCR) strategy was used to identify the ligands and receptors that might contribute to endogenous Wnt/β-catenin signaling activity. Because bovine sequences for Wnt and Fzd genes have not been reported, degenerate primers that recognize multiple members of the Wnt and Frizzled family (Tanaka et al.,1998; Wright et al.,1999) were used to clone fragments of the bovine genes. Analysis of mRNA isolated both from freshly isolated and cultured BAEC demonstrated that Wnt2b and Fzd4 are expressed by EC freshly scraped from the bovine aorta and that Wnt2b, Fzd3, Fzd4, and Fzd6 are expressed by BAEC in vitro (Fig. 3). Because EC in vivo are apposed by pericytes or smooth muscle cells, gene expression in BASMC was also examined. These cells expressed Wnt2b, Fzd1, and Fzd4 (Fig. 3). Both Wnts and their receptors are, therefore, expressed by bovine aortic cells and by EC in vivo as well as in vitro.

Figure 3.

Wnt and Fzd expression by bovine aortic cells. A: cDNA from freshly scraped bovine aortic endothelial cells (BAEC), cultured BAEC, and cultured bovine aortic smooth muscle cells (BASMC) was amplified using degenerate PCR primers specific for conserved regions of Wnt and Fzd genes. B: Amplifed PCR products were excised from the gel, cloned into the TOPO-TA vector, and sequenced. Wnt and Fzd genes were identified by conducting a BLAST search for homologous sequences.

Because some degree of chance was involved with selecting clones that expressed different genes, it is likely that bovine vascular cells express more Wnt and Fzd genes than were identified in this study. To more systematically identify genes of the Wnt signaling pathway expressed by vascular cells, we next turned to an RT-PCR approach using primers designed to match specific genes.

Wnt Genes Are Expressed by Human Vascular Cells

Because sequences are published for all known human Wnt signaling genes, samples of human vascular cells and tissues were used to conduct the systematic RT-PCR analysis of expressed Wnt signaling components. Nine vascular mRNA samples were analyzed: freshly isolated umbilical vein EC (HUVEC), three different isolates of cultured HUVEC, microvascular EC from adult dermis and from neonatal dermis, aortic SMC, and intact aorta. Human fetal brain mRNA served as a positive control for nearly all Wnt signaling components. Of the nineteen human Wnts identified to date, sixteen were identified in at least one of the vascular samples (Fig. 4; Table 1). Wnt2b, -3, -4, and -5a were expressed by all vascular samples. Wnt7b and Wnt16 were expressed by the intact aorta only, and Wnt2 was expressed by both HASMC and the aorta. Wnt5b, -6, -7a, -8a, -10a, -10b, -11, -14, and -15 were expressed by two to six vascular cell populations each. Expression of Wnt1-3a and -8b was not observed for any of the samples tested. All vascular cell types expressed at least five Wnts, including Wnts that are known to signal by means of the Wnt/β-catenin pathway.

Figure 4.

Expression of Wnt genes by human vascular cells. cDNA from freshly isolated human umbilical vein cells (HUVEC), cultured HUVEC from three isolates, microvascular endothelial cells derived from adult dermis (HMVECad) and neonatal dermis (HMVECnd), cultured aortic smooth muscle cells (HASMC), and whole aorta was amplified using polymerase chain reaction primers specific for individual Wnt genes. Human fetal brain cDNA was used as a positive control. *, the positive control for Wnt1 was a plasmid containing Wnt1 sequence.

Table 1. Wnt Expression by Human Vascular Cellsa
GeneHUVEC freshHUVEC 1HUVEC 2HUVEC 3HMVEC adHMVEC ndHASMCAorta, whole
  • a

    ad, adult dermis; HUVEC, umbilical vein endothelial cells; HMVEC, microvascular endothelial cells; HASMC, aortic smooth muscle cells; nd, neonatal dermis.

Wnt1
Wnt2+++
Wnt2b++++++++
Wnt3++++++++
Wnt3a
Wnt4++++++++
Wnt5a++++++++
Wnt5b+++++
Wnt6*++
Wnt7a+++++
Wnt7b+
Wnt8a++
Wnt8b
Wnt10a++
Wnt10b++
Wnt11+++++
Wnt14++++++
Wnt15++
Wnt16+

Fzd and LRP Genes Are Expressed by Human Vascular Cells

All vascular cells that were examined expressed Fzd genes (Fig. 5; Table 2). Whereas Fzd6 and Fzd8 were identified in every sample, Fzd1, -2, -3, -4, -5, -7, -9, and -10 were expressed in some, but not all, vascular cell populations. Freshly isolated HUVEC, one isolate of cultured HUVEC, HMVEC, HASMC, and whole aorta expressed at least 6 of the 10 Fzd genes; the other two HUVEC isolates expressed 2 and 3 Fzd genes, respectively. For canonical signaling to take place, either LRP5 or LRP6 must also be expressed. Both of these genes were expressed by the dermal microvascular EC, aortic SMC, whole aorta, and freshly isolated HUVEC. LRP5 and/or LRP6 were expressed by two of the three cultured HUVEC populations. These results confirm that not only the ligands, but also the necessary receptors for Wnt signaling are expressed by vascular cells.

Figure 5.

Expression of Fzd and LRP genes by human vascular cells. cDNA from freshly isolated human umbilical vein cells (HUVEC), cultured HUVEC from three isolates, microvascular endothelial cells derived from adult dermis (HMVECad) and neonatal dermis (HMVECnd), cultured aortic smooth muscle cells (HASMC), and whole aorta was amplified using polymerase chain reaction primers specific for individual Fzd and LRP genes. Human fetal brain cDNA was used as a positive control.

Table 2. Fzd and LRP Expression by Human Vascular Cells
GeneHUVEC freshHUVEC 1HUVEC 2HUVEC 3HMVEC adHMVEC ndHASMCAorta, whole
  1. aPresence or absence of Fzd and LRP bands observed in Figure 6 are indicted with + and −, respectively. HUVEC, umbilical vein endothelial cells; HMVECad, human microvascular endothelial derived from adult dermis; HMVECnd, human microvascular endothelial derived from neonatal dermis; HASMC, aortic smooth muscle cells.

Fzd1++++++
Fzd2++++++
Fzd3++
Fzd4++++
Fzd5++++++
Fzd6++++++++
Fzd7+++++
Fzd8++++++++
Fzd9++++
Fzd10+++
LRP5+++++++
LRP6++++++
Figure 6.

Expression of SFRP, WIF, and Dkk genes by human vascular cells. cDNA from freshly isolated human umbilical vein cells (HUVEC), cultured HUVEC from three isolates, microvascular endothelial cells derived from adult dermis (HMVECad) and neonatal dermis (HMVECnd), cultured aortic smooth muscle cells (HASMC), and whole aorta was amplified using polymerase chain reaction primers specific for individual SFRP, WIF, and Dkk genes. Human fetal brain cDNA was used as a positive control.

Secreted Regulators of Wnt Signaling Are Expressed by Human Vascular Cells

Of the known secreted regulators of Wnt signaling, only WIF-1 was not expressed by any of the vascular cell types analyzed (Fig. 6; Table 3). Dkk1 and Dkk3 were expressed by all cell populations, SFRP2 and SFRP4 were expressed by SMC and aorta only, and Dkk4 was expressed only by the intact aorta. All populations except the intact aorta expressed SFRP1. SFRP3 and Dkk2 were both expressed by the intact aorta, freshly isolated HUVEC, and by one cultured EC population; and SFRP5 was expressed by five of the nine cell types analyzed. All vascular cells expressed at least three secreted regulators of Wnt signaling, and the intact aorta expressed all but SFRP1 and WIF-1. These data suggest that Wnt signaling may be modulated by various secreted components, both in vitro and in vivo.

Table 3. SFRP, WIF, and Dkk Expression by Human Vascular Cellsa
GeneHUVEC freshHUVEC 1HUVEC 2HUVEC 3HMVEC adHMVEC ndHASM CAorta whole
  • a

    Presence or absence of SFRP, WIF, and Dkk bands observed in Figure 6 are indicted with + and −, respectively. SFRP, secreted Frizzled-related protein; HUVEC, umbilical vein endothelial cells; HMVECad, human microvascular endothelial derived from adult dermis; HMVECnd, human microvascular endothelial derived from neonatal dermis; HASMC, aortic smooth muscle cells.

SFRP1+++++++
SFRP2++
SFRP3+++
SFRP4++
SFRP5+++++
WIF-1
Dkk1++++++++
Dkk2+++
Dkk3++++++++
Dkk4+

DISCUSSION

The localization of β-catenin to the intercellular junctions of confluent EC and the cytosol and nucleus of sparse or migrating EC is consistent with in vivo findings. β-catenin is rarely, if ever, observed in the cytoplasm of EC in normal adult vessels, although it is localized to this compartment in angiogenic EC of the developing embryo (Eberhart et al.,2000; Eberhart and Argani,2001), ischemic myocardium (Blankesteijn et al.,2000) and tumors (Eberhart et al.,2000; Yano et al.,2000a,b). A similar correlation between proliferation and β-catenin stabilization has been observed for colonocytes (Sellin et al.,2001) and bronchial epithelial cells (Steel et al.,2005). This β-catenin stabilization and corresponding TCF activity may regulate vascular cell proliferation, survival, and/or differentiation in vitro.

Our systematic RT-PCR analysis of Wnt signaling components expressed by human vascular cells demonstrated that most Wnts and all Fzd and LRP were expressed by at least one vascular cell type. Of the expressed Wnt genes, Wnt2b, -4, and -5a are of particular interest, as they were expressed by all vascular cell populations examined and have functions consistent with angiogenic processes, namely regulation of proliferation, branching, and tubulogenesis (Stark et al.,1994; Brisken et al.,2000; Lin et al.,2001; Li et al.,2002; Kubo et al.,2003). Expression of Fzd4 and Fzd5 was not surprising, given that these receptors are required for normal vascularization of the retina (Xu et al.,2004), corpus luteum (Hsieh et al.,2005), and yolk sac and placenta (Ishikawa et al.,2001). Mutations in Fzd4 and LRP5 have also been associated with the human retinal vascular disease FEVR (Robitaille et al.,2002; Kondo et al.,2003; Toomes et al.,2004). It remains to be determined which Wnt-Fzd binding partners have physiological or pathological roles in the vasculature.

SFRP1 and Dkk1, two known inhibitors of canonical Wnt signaling, were both expressed by all cultured EC populations. In light of the activation of canonical signaling observed in cultured EC, the relative levels of inhibitors compared with Wnt or Fzd levels may be regulating activity of the Wnt pathways. The finding that SFRP1 is not expressed by cells of the intact human aorta is surprising, given that this gene is expressed in the bovine aorta in vivo (Duplàa et al.,1999). Because inhibitors of the canonical pathway may have redundant function, however, it is possible that other SFRP or Dkk proteins replace SFRP1 in the human aorta.

Although the differences in gene expression patterns between the three cultured HUVEC samples were initially surprising, in fact variability in gene expression between different human samples is frequently observed for Wnt and Fzd genes. Isolates of normal human pancreas, colon, and mammary tissues from different individuals display alternate Wnt and Fzd gene expression profiles (Huguet et al.,1994; Dimitriadis et al.,2001; Heller et al.,2004). This variability may also explain certain discrepancies between studies of HUVEC gene expression. For example, Fzd3 is expressed by cultured HUVEC in some studies (Wright et al.,1999; Cheng et al.,2003), but not in others (this study and Masckauchan et al.,2005). This variation in gene expression is likely due to functional redundancy among many Wnt and Fzd proteins. Other results from our study confirmed and expanded on previous studies in which Wnt signaling component gene expression was examined (Wright et al.,1999; Mao et al.,2000; van Gijn et al.,2001; Cheng et al.,2003; Masckauchan et al.,2005).

Wnt signaling plays important roles in physiological and pathological angiogenesis. It will be interesting to determine how the expressed signaling components contribute to vascular cell proliferation, migration, survival, and differentiation. Further studies will also be needed to identify the cells that produce and respond to these factors in the vasculature. A better understanding of these processes becomes ever more important as Wnt agonists are considered as proangiogenic agents and Wnt antagonists are designed for anticancer treatment.

EXPERIMENTAL PROCEDURES

Cell Isolation and Culture

BAEC were isolated as previously described (Gimbrone,1976) and used between passages 2 and 12. BASMC were isolated as described (Ross,1971) and used at passage 5. PAEC, an immortalized EC line, were generously provided by Dr. Shay Soker (Children's Hospital, Boston, MA). BAEC, BASMC, and PAEC were maintained in Dulbecco's modified Eagle's medium (DMEM; JRH Biosciences) supplemented with 10% calf serum (CS; Hyclone Laboratories), 233.6 μg/ml glutamine, 80 units/ml penicillin, and 80 μg/ml streptomycin (GPS; Irvine Scientific). HUVEC, both freshly isolated and cultured, were generously provided by Dr. Michael Gimbrone (Brigham and Women's Hospital, Boston, MA). HUVEC were maintained in gelatin-coated tissue culture flasks in medium containing M199/20%fetal calf serum (FCS)/GPS supplemented with 20 ng/ml basic fibroblast growth factor (R&D Systems) and 100 μg/ml heparin (Sigma) and were used between passages 2 and 6. Human microvascular EC derived from adult dermis (HMVECad; provided by Dr. Kameran Lashkari, Schepens Eye Research Institute, Boston, MA) and neonatal dermis (HMVECnd, Cascade Biologics) were maintained in Medium 131 supplemented with Microvascular Growth Supplement (Cascade Biologics) and used at passages 3 and 5. HASMC were generously provided by Dr. Peter Libby (Brigham and Women's Hospital), cultured in DMEM/10%FCS/GPS, and used at passage 5.

Immunohistochemistry

BAEC were plated at various densities on four-well chamber slides (Lab-Tek). Two days after plating, the cells were fixed with 4% paraformaldehyde, then blocked with phosphate-buffered saline (PBS) containing 3% donkey serum, 3% bovine serum albumin, and 0.2% Triton X-100. Primary antibody specific for β-catenin (Sigma) was used at a 1:2,000 dilution in the blocking solution; the cy3-conjugated secondary antibody (Jackson Immunochemicals) was used at a 1:500 dilution. Mouse IgG (Sigma) was used in place of primary antibody as a negative control. Cells were photographed with a digital camera attached to a Nikon Axiophot microscope. This experiment was conducted three times with similar results.

Cell Fractionation and Immunoblotting

For extraction of cytosolic proteins, BAEC were lysed in nuclear extraction buffer (Nuclei EZ Lysis Buffer, Sigma) supplemented with 2 μg/ml aprotinin, 5 μg/ml leupeptin, 10 μg/ml phenylmethyl sulfonyl fluoride, and 10 mM sodium fluoride (Sigma). Confluent monolayers were further disrupted with 20 strokes of a Dounce homogenizer. Lysis was confirmed by visualization of cells stained with trypan blue. Nuclei and unlysed cells were removed by centrifugation for 10 min at 500 × g; the supernatant was then separated into cytosolic and membrane-associated fractions by centrifugation for 30 min at 115,000 × g. Ten micrograms of cytosolic protein was electrophoresed through 10% polyacrylamide gels, using the prelabeled Benchmark Protein Ladder (Invitrogen) for molecular weight comparison, and transferred to nitrocellulose filters. Filters were blocked with 5% nonfat powdered milk, 3% bovine serum albumin, 0.2% Nonidet P-40 in PBS; all antibodies were diluted in blocking solution. Primary antibodies for β-catenin (1:2,000, Sigma) and α-tubulin (1:500, Oncogene) were followed by incubation with horseradish peroxidase–conjugated secondary antibodies (Amersham). Immunoreactive bands were visualized by using ECL plus (Amersham) and detected on Hyperfilm (Amersham). Blots were scanned into Adobe Photoshop, and densitometric analyses were conducted using NIH Image software (available at http://rsb.info.nih.gov/nih-image/). Results of three experiments were combined and the mean values, standard deviations, and statistical significance calculated using Statview (SAS Institute). Statistical significance was defined as P < 0.05, as determined by analysis of variance in combination with Fischer's protected least significant difference.

Monolayer Scrape Wound Assay

BAEC were grown to confluence in four-well plates (Nunc). The monolayer was scraped with a 1,000 μl pipet tip, generating a cleared area approximately 2 mm wide. Detached cells were removed, and the remaining cells were cultured in DMEM/10%CS/GPS for 24 hr before fixation and immunostaining as described above. Cells were photographed with a Nikon inverted microscope attached to a digital camera. This experiment was conducted three times with similar results.

Plasmids, Transient Transfections, and Luciferase Assays

pCDNA-SFRP1 was generated by cloning the full-length coding sequence of the bovine SFRP1 gene (Duplàa et al.,1999) into the EcoRI and HindIII sites of pCDNA3.1 (Invitrogen). TOPflash, a reporter plasmid containing six TCF-binding motifs upstream of a minimal TK promoter and the firefly luciferase gene, was purchased from Upstate Biotechnologies; pRL-TK, containing Renilla luciferase under the control of the minimal TK promoter, was purchased from Promega.

BAEC and PAEC were plated into 12-well tissue culture dishes and grown to 60% confluence. Transfection medium for each well contained 0.4 μg TOPflash, 0.02 μg pRL-TK, 0.4 μg of either pCDNA3.1 or pCDNA-SFRP1, and 4.8 μl of FuGENE 6 (Roche) in DMEM/5%CS. Cells in triplicate wells were transfected for each condition. Transfection efficiency, assessed by transfecting cells with 0.8 μg/well pCDNA3.1-lacZ (Invitrogen) and subsequent staining with a β-galactosidase staining kit (Stratagene), was typically observed to be 20–50% for BAEC and 40–80% for PAEC. Luciferase expression was detected using the Dual Luciferase Reporter Assay System (Promega) with a Turner Luminometer (Turner Designs). Readings for firefly luciferase and Renilla luciferase were recorded, and the ratio of firefly to Renilla luciferase activity was determined. Assays were performed three times with similar results; a representative experiment is shown. Mean value, standard deviation, and statistical significance were determined with analysis of variance in combination with Fischer's protected least significant difference (Statview, SAS Institute).

RT-PCR

Total RNA was isolated using RNAzol (TelTest) and DNase-treated using the DNAfree kit (Ambion) following the manufacturers' protocols. Samples of total RNA from human aorta and human fetal brain were purchased from Ambion and Stratagene, respectively. RNA was reverse-transcribed using 1 μg of RNA per 10-μl reaction with the Prostar First-Strand RT-PCR kit (Stratagene). As a negative control, reactions lacking reverse transcriptase were included for each RNA sample. RT-PCR analysis using primers specific for β-actin was conducted for each reverse transcriptase-containing sample to assess cDNA integrity, and for the negative control reactions to confirm that the RNA was not contaminated by genomic DNA.

Degenerate primers used for identification of Wnt genes in bovine samples were 5′-GGGGAATTCCARGARTGYAARTGYCAT-3′ and 5′-AAA- ATCTAGARCARCACCARTGRAA-3′ (Wright et al.,1999). The two primer pairs used for identification of Fzd genes were 5′-ATCGGAATTCTAYCCNGARCGNCCNAT-3′ and 5′-ATCGAAGCTTNGCNGCNAGRAACCA-3′ (Wright et al.,1999), and 5′- TAYCCNGARCGNCCNATYAT-3′ and 5′-AGAGTNAGDATNACCCACCA-3′ (Tanaka et al.,1998). Reactions (25 μl total) were composed of 1× PCR buffer, 4 mM MgCl2, 2.5 mM dNTPs, 4 μM each primer, 5 U Taq polymerase (all from Roche), and 2 μl of cDNA. PCR conditions were 10 min at 95°C; 40 cycles of 30 sec at 94°C, 30 sec at 55°C, and 90 sec at 72°C; followed by 10 min at 72°C. Products were electrophoresed through agarose gels, extracted from the gels using a Qiaquick gel extraction kit (Qiagen), and cloned using the TOPO TA Cloning Kit for Sequencing (Invitrogen). Inserts from at least 20 clones were sequenced, and the corresponding genes were identified using BLAST (www.ncbi.nlm.nih.gov/BLAST/). Partial cDNA sequences for bovine Fzd1, Fzd3, Fzd4, Fzd6, and Wnt2b were submitted to GenBank; accession numbers are AY624568, AY624569, AY624570, AY624571, and AY624572, respectively.

Specific primers for human Wnt, Fzd, LRP, sFRP, Dkk, and WIF genes are described in Table 4. PCR reactions (25 μl total) consisted of 1× PCR MasterMix (Promega), 4 μM each primer, and 1 μl of cDNA. PCR conditions were 5 min at 95°C; 40 cycles of 30 sec at 94°C, 30 sec at 60°C, and 45 sec at 72°C; followed by 10 min at 72°C. An annealing temperature of 55°C was used for primers specific for Wnt5b, Wnt6, Wnt8a, Wnt8b, Wnt10a, Wnt14, Wnt15, Wnt16, sFRP5, WIF-1, Dkk1, Dkk2, Dkk3, and Dkk4. For primer pairs of my design, reaction products were cloned and sequenced as described above to confirm identity of the amplified products.

Table 4. Primers Used for RT-PCR of Human Samplesa
GeneSize (bp)Forward (5′-3′)Reverse (5′-3′)Reference
  • a

    RT-PCR, polymerase chain reaction.

Wnt1309caacagcagtggccgatggtggcggcctgcctcgttgttgtgaag(Vider et al.,1996)
Wnt2326gtcatgaaccaggatggcacatgtgtgcacatccagagcttc(Vider et al.,1996)
Wnt2b320aagatggtgccaacttcaccgctgccttcttgggggctttgc(Sen et al.,2000)
Wnt3223gagagcctccccgtccacagctgccaggagtgtattcgcatc(Katoh,2002)
Wnt3a326caggaactacgtggagatcatgccatcccaccaaactcgatgtc(Katoh,2002)
Wnt4252tgccactgaggtggagccactcagccagctccacctgcgc(Dimitriadis et al.,2001)
Wnt5a216gacctggtctacatcgaccccgcagcaccagtggaacttgca(Vider et al.,1996)
Wnt5b250tgaaggagaagtacgacagcctcttgaactggttgtagcc(Heller et al.,2004)
Wnt6205ttatggaccctaccagcatatgtcctgttgcaggatg(Heller et al.,2004)
Wnt7a437gccgttcacgtggagcctgtgcgtgcagcatcctgccagggagcccgcagct(Li et al.,2001)
Wnt7b442gattcggccgctggaactgctctggcccacctcgcggaacttagOur design
Wnt8a180ctggtcagtgaacaatttccgtagcacttctcagcctgtt(Heller et al.,2004)
Wnt8b286gtcttttcacctgtgtcctcaggctgcagtttctagtcag(Heller et al.,2004)
Wnt10a152ctgttcttcctactgctgctacacacacctccatctgc(Heller et al.,2004)
Wnt10b200gcaccacagcgccatcctcaagggggtctcgctcacagaagtcagga(Sen et al.,2000)
Wnt11255ctggaaatgaagtgtaagtgctgtgtcccgtgggagcccacc(Lako et al.,1998)
Wnt14162acaagtatgagacggcactcagaagctaggcgagtcatc(Heller et al.,2004)
Wnt15150tgaaactgcgctatgactcgtgagtcctccatgtacacc(Heller et al.,2004)
Wnt16300gagagatggaactgcatgatgatggggaaatctaggaact(Heller et al.,2004)
Fzd1397gaactttcctccaacttcatgggcatttccattttacagaccgg(Kirikoshi et al.,2001)
Fzd2333ggtgagccagcactgcaagagcctaaaagtgaaatggtttcgatcg(Kirikoshi et al.,2001)
Fzd3456gctgtactcacagttaacatggctaaaatacccttgctagattt(Kirikoshi et al.,2001)
Fzd4401tgccttttcagggcaaagtgacaggaagagatttatggaatg(Kirikoshi et al.,2001)
Fzd5248tacccagcctgtcgctaaacaaaaccgtccaaagataaactgc(Kirikoshi et al.,2001)
Fzd6421tggcctgaggagcttgaatgtgactatcgcccagcaaaaatccaatgaOur design
Fzd7295gtttggatgaaaagatttcaggcgaccactgcttgacaagcacac(Kirikoshi et al.,2001)
Fzd8269acagtgttgattgctattagcatggtgaaatctgtgtatctgactgc(Kirikoshi et al.,2001)
Fzd9264ccctagagacagctgactagcagcgggggtttattccagtcacagc(Kirikoshi et al.,2001)
Fzd10170acacgtccaacgccagcatgacgagtcatgttgtagccgatg(Kirikoshi et al.,2001)
LRP5667gacatctacagccggacactgcacaagtcagcaggttctgcagg(Qiang et al.,2003)
LRP6736gattatccagaaggcatggcagcaatcaccatgcggttgatggc(Qiang et al.,2003)
SFRP1513ggaccggcccatctaccctccttgtttttcttgtcccacttgOur design
SFRP2499ctcgctgctgctgctcttcggcttcacatacctttggag(Schumann et al.,2000)
SFRP3431atggtctgcggcagcccggctgtcgtacactggcagctc(Schumann et al.,2000)
SFRP4574gttcctctccatcctagtgggctgagatacgttgccaaag(Schumann et al.,2000)
SFRP5201ctactggagggtgttttcacctttcccttaccctctcct(Heller et al.,2004)
WIF-1218cacctggattctatggagtgacagaggtctccctggtaac(Heller et al.,2004)
Dkk1202caggattgtgttgtgctagatgacaagtgtgaagcctaga(Heller et al.,2004)
Dkk2164ctcaactccatcaagtcctctacctcccaacttcacactc(Heller et al.,2004)
Dkk3215gaggttgaggaactgatggccagtctggttgttggttat(Heller et al.,2004)
Dkk4233gtcctggacttcaacaacatgttgcatcttccatcgtagt(Heller et al.,2004)

Acknowledgements

We thank Robyn Loureiro for technical advice and critical reading of the manuscript.

Ancillary