1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

The mechanisms regulating the growth and differentiation of hepatic sinusoidal endothelial cells (HSECs) are not well defined. Because Wnt signaling has become increasingly important in developmental processes such as vascular and hepatic differentiation, we analyzed HSEC-specific Wnt signaling in detail. Using highly pure HSECs isolated by a newly developed protocol selecting against nonsinusoidal hepatic endothelial cells, we comparatively screened the multiple components of the Wnt pathway for differential expression in HSECs and lung microvascular endothelial cells (LMECs) via reverse-transcription polymerase chain reaction (RT-PCR). As confirmed via quantitative RT-PCR and northern and western blotting experiments, Wnt2 (and less so Wnt transporter wls/evi) and Wnt coreceptor Ryk were overexpressed by HSECs, whereas Wnt inhibitory factor (WIF) was strongly overexpressed by LMECs. Exogenous Wnt2 superinduced proliferation of HSECs (P < 0.05). The Wnt inhibitor secreted frizzled-related protein 1 (sFRP1) (P < 0.005) and transfection of HSECs with Wnt2 small interfering RNA (siRNA) reduced proliferation of HSECs. These effects were rescued by exogenous Wnt2. Tube formation of HSECs on matrigel was strongly inhibited by Wnt inhibitors sFRP1 and WIF (P < 0.0005). Wnt signaling in HSECs activated the canonical pathway inducing nuclear translocation of β-catenin. GST (glutathione transferase) pull-down and co-immunoprecipitation assays showed Fzd4 to be a novel Wnt2 receptor in HSECs. Gene profiling identified vascular endothelial growth factor receptor-2 (VEGFR-2) as a target of Wnt2 signaling in HSECs. Inhibition of Wnt signaling down-regulated VEGFR-2 messenger RNA and protein. Wnt2 siRNA knock-down confirmed Wnt2 specificity of VEGFR-2 regulation in HSECs. Conclusion: Wnt2 is an autocrine growth and differentiation factor specific for HSECs that synergizes with the VEGF signaling pathway to exert its effects. (HEPATOLOGY 2008;47:1018–1031.)

Hepatic sinusoidal endothelial cells (HSECs) are uniquely differentiated microvascular endothelial cells that exert highly specialized functions compared with endothelial cells from other organs. HSECs display a special morphology with numerous fenestrations, with intracellular gaps and an incomplete basal lamina. Functionally, HSECs are highly active “professional” endocytes clearing the portal blood from macromolecular waste molecules.1 Among the HSEC-specific scavenger receptors involved in these processes, stabilin-2, identified by us, is the major hyaluronan receptor of the liver, binding other ligands such as advanced glycation endproduct (AGE) or collagen peptides.2 The stabilin-2 homologue -1 also functions as an endocytic receptor.3 Stabilin-1 and stabilin-2 use the classical constitutive clathrin-mediated endocytic pathway in HSECs; in addition, stabilin-1 fulfills a second role as an intracellular cargo carrier.4 In HSECs, clathrin-coated vesicles and stabilin-1/2+ early endosomes travel along microtubules organized in a special netlike way, linking HSEC morphology to its functions.5

In contrast to the mechanisms active in professional endocytosis by HSECs, little is known about the processes that regulate proliferation and differentiation of HSECs. Vascular endothelial growth factor (VEGF) is important in HSEC differentiation6 and protects hepatocytes from toxic damage via induced secretion of hepatocyte growth factor by HSECs. Recently, the Wnt signaling pathway has gained importance in both liver differentiation and disease as well as in vascular development.7, 8 Wright et al.9 were the first to show induction of endothelial proliferation upon Wnt1 overexpression. However, Wnt1-induced proliferation of endothelial cells has become a controversial issue.10 In the meantime, several other Wnt ligands have been demonstrated to be expressed in various endothelia, including Wnt-2, Wnt-5a, Wnt-7a, and Wnt-10b. Wnt2-deficient mice display vascular abnormalities, including a defective placental vasculature.11 Of the known murine frizzled (Fzd) receptors for Wnt, Fzd1, Fzd2, Fzd3, and Fzd5 have been shown to be expressed in vascular wall cells. Fzd5-deficient mice display vascular abnormalities and are embryonic lethal,12 whereas Fzd4−/− mice show a defective vasculature in the corpora lutea.13

In this study, we investigated the existence and functionality of a specific Wnt signaling pathway in highly purified HSECs compared with blood microvascular endothelial cells from lung microvascular endothelial cells (LMECs). We present evidence that HSECs are Wnt2-producing cells and require Wnt2 for proliferation and morphogenesis, suggesting an autocrine signaling mechanism. We used gene profiling to search for genes affected by the inhibition of Wnt2 signaling. VEGF receptor-2 (VEGFR-2) was found to be a Wnt2 target gene down-regulated by inhibition of Wnt signaling, indicating cross-activation of the VEGF signaling cascade by the Wnt pathway.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Reagents and Antibodies.

Collagen I (rat tail) was obtained from Sigma (Steinheim, Germany), secreted frizzled-related protein 1 (sFRP1) and Wnt inhibitory factor-1 (WIF1) were obtained from R&D Systems Inc. (Minneapolis, MN). Collagenase P was obtained from Roche (Mannheim, Germany). Alexa-Fluor-488 AcLDL, Alexa-Fluor-555 Tranferrin, TOTO-3 iodide, and the pcDNA3 plasmid were obtained from Invitrogen (Karlsruhe, Germany). AGE–bovine serum albumin (BSA) was obtained from MBL (Woburn, MA) and was conjugated to fluorescein isothiocyanate using a FluoroTag-FITC conjugation kit from Sigma. Liver extracellular matrix substrate was generated as described by Sellaro et al.14 Kinase insert domain receptor (KDR)/VEGFR-2 antibody as well as control and Wnt2 small interfering RNA (siRNA) were obtained from Santa Cruz Biotechnology (Santa Cruz, CA) and were used according to the manufacturer's instructions. Rabbit anti-rat polyclonal antibody against Wnt2 was custom-made by Peptide Specialty Laboratories (Heidelberg, Germany) using the immunizing peptide NKRFKKPTKNDLVYC. Affinity purification of the antiserum was performed according to the manufacturer's instructions. Stabilin-1 (1.26) and stabilin-2 (1.3; 3.1) monoclonal antibodies were generated via immunization of stabilin-1 and stabilin-2 knockout mice.15 All peroxidase and fluorescent-labeled secondary antibodies were obtained from Jackson IR Laboratories (West Grove, PA). Human umbilical vein endothelial cells and lung lymphatic human microvascular endothelial cells were obtained from Cambrex (Verviers, Belgium). bEnd3 cells were purchased from LGC Promochem (Wesel, Germany).


Sprague-Dawley rats were purchased from Janvier (Le Genest-St-Isle, France) and received humane care according to the guidelines of the National Institutes of Health. Animal experiments were approved by the animal ethics committee in Baden-Württemberg (Regierungspräsidium Karlsruhe AZ:35-9185.82A-30/02).

Isolation and Purification of HSECs and LMECs.

Rat HSECs were isolated as described previously.5 Resultant cell isolates were plated on collagen I–coated dishes or subjected for further cell purification. Primary HSECs were cultured using normal growth media (1/1 (v/v) mixture of EBM-2 (CC-3156; Cambrex, Walkersville, MD) and Williams' E (Invitrogen, Karlsruhe, Germany)), containing EGM SingleQuots (CC-4133), 0.2% BSA, and 10 ng/mL hepatocyte growth factor at 37°C in a humidified incubator (5% CO2). Highly pure rat HSECs devoid of Kupffer cell and stellate cell contamination were generated using an N-terminal stabilin-2 monoclonal antibody and MACS technology. A MACS system (Miltenyi Biotec) was used according to the manufacturer's instructions. Briefly, HSECs were washed, resuspended in binding buffer [phosphate-buffered saline (PBS) (pH 7.2), 0.5% BSA, and 2 mM ethylene diamine tetraacetic acid] and incubated at 4°C with Stab2 (3.1) antibody for 15 minutes. Cells were washed with binding buffer, centrifuged at 300g for 10 minutes, resuspended and incubated with goat anti-mouse immunoglobulin G Microbeads at 4°C for 15-20 minutes. Labeled cells were loaded onto a column installed in a magnetic field. The column was rinsed with buffer, and negative cells were passed through. Trapped cells were eluted after the removal of the column from the magnet. For higher purity, Stab2 positive fraction was reloaded onto a new column again. The column was rinsed with buffer; positive fraction was eluted as described above.

LMECs were purified from type 2 collagenase (Worthington, Lakewood, NJ) digested cell extracts using the same protocol in combination with biotinylated PECAM1 antibody (BD Bioscience) and Streptavidin Microbeads.

Fluorescence-Activated Cell Sorting.

For fluorescence-activated cell sorting (FACS) analysis, 5 × 106 cells were washed with PBS, fixed with 4% paraformaldehyde on ice for 30 minutes, washed, and resuspended in 100 μL of PBS/1% BSA containing diluted antibody (1/100) or corresponding isotype control. After 45 minutes of incubation on ice, the cells were washed twice with 1% BSA/PBS and resuspended in 400 μL of the same buffer. For cell permeabilization, 0.5% Saponin was added to the 1% BSA/PBS solution. Stained cells were analyzed with FACS-Calibur (BD Biosciences, Heidelberg, Germany). Results were evaluated with WinMDI software. PECAM1-PE, CD11b/c-PE, rat endothelial cell antigen antibody, and isotype controls were obtained from BD Biosciences.

Proliferation Assay.

Mitogenic activity was determined by measuring the incorporation of tritium-labeled thymidine (H3-thymidine) into DNA of proliferating cells. Primary cultures of HSECs were incubated in the presence of 0.185 MBq/mL H3-thymidine (specific activity, 365 MBq/mmol; Amersham, Giles, UK) for 16-18 hours, and incorporation of radioactivity was measured in a 1900CA liquid-scintillation analyzer (Packard) using whole cell lysates.

Tube Formation Assay (Matrigel Assay).

HSECs were seeded on growth factor-reduced Matrigel in normal growth medium with or without 5 μg/mL sFRP1 and 2.5 Mg/mL WIF1. Capillary-like tube formation was quantified by counting numbers of junctions/enclosed circles in 3 randomly chosen optical fields using light microscopy after 16 hours.

Western Blot Analysis.

Whole cell lysates were generated by scraping cells into an ice-cold RIPA-P buffer [150 mM NaCl, 1% NP40, 0.5% sodium desoxycholate, 0.1% sodium dodecyl sulfate, 50 mM Tris/HCL (pH 8.0), 10 mM NaF, 1 mM Na3OV4] supplemented with a complete protease inhibitor cocktail (Roche) and performing 2-3 freeze–thaw cycles. Protein content was determined using the DC Assay (BioRad, Munich, Germany). Immunoprecipitation of antibodies and isotype controls (2 μg) was performed using Immunopure Protein G (PIERCE, Bonn, Germany) according to the manufacturer's instructions. Nuclear cell extracts were generated using a CelLytic-NuCLEAR-Extraction Kit (Sigma) according to the manufacturer's instructions. In brief, cells were allowed to swell in hypotonic buffer and were disrupted, the cytoplasmic fraction was removed, and the nuclear proteins were released from the nuclei via a high-salt buffer. Protein samples (50-100 μg of protein) were subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis, and western blot analysis was performed as described previously16 using KDR/VEGFR-2 (1/200) or Wnt2 (1/1000) antibodies.


HSECs were plated on collagen I–coated cover slips prior fixation with 4% paraformaldehyde for 15 minutes, washed with PBS, and incubated with blocking solution (2% fetal bovine serum/PBS) for 30 minutes. After permeabilization, cells were incubated with antibodies to Stab-1/2 (1/200), LYVE-1 (1/1000), and Wnt2 (1/100). The antigen was detected with a Cy2/3-conjugated secondary antibody (1/250). Specimens were mounted in Vectashield (Vector) and analyzed via confocal microscopy.

Construction of the Plasmid pcDNA3-mWnt2.

The pCDNA3-mWnt2 (6.5 kb) plasmid was constructed by inserting an XhoI-XbaI fragment containing the mWnt2 complementary DNA (cDNA) (IRAVp968B0961, 1.1 kb, RZPD, Berlin, Germany) into the XhoI-XbaI site of the pcDNA3 plasmid (5.4 kb). Polymerase chain reaction (PCR) primers for generating an XhoI-XbaI fragment containing the mWnt2 cDNA were as follows: XhoI-mWnt2 forward primer, 5′-AAACTCGAGATGAACGTCCCTCTCGGTG- GAATC-3′; mWnt2-XbaI backward primer, 3′-AAATCTAGATCATGTAGGCGTCGCCCAGTC- GGC-5′.

Conditioned Media.

Wnt2-expressing cells were generated via transfection of CHO-K1 cells with pcDNA3-mWnt2 using SuperFect reagent (Qiagen, Hilden, Germany). Control CHO-K1 cells were transfected with the empty pcDNA3 vector. Clones were selected using neomycin and were screened for mWnt2 expression. Stable transfected clones were cultured in F12/HAM media (10% fetal bovine serum) until confluence. Media was replaced and cells were cultured in the presence of 0.5% fetal bovine serum for 24 hours before collection of media.

RNA Isolation and cDNA Synthesis.

For RNA isolation, the cells were lysed directly in plastic Petri dishes and the RNA was isolated using RNeasy Mini kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. Total RNA (2 μg) was treated with 2 U RNase-free DNase (Ambion, TX) and used for reverse transcription with Superscript II reverse transcriptase (Invitrogen, Karlsruhe, Germany) using oligo-dT primers according to the manufacturer's instructions. We used 0.5 μL of cDNA for PCR analysis.

Reverse-Transcription PCR and Real-Time Reverse-Transcription PCR Analysis.

Analysis was performed using the oligonucleotide primers listed in Table 1. The PCR program consisted of initial denaturation at 95°C for 30 seconds, annealing at 58°C for 40 seconds, and extension at 72°C for 30 seconds for 25-30 cycles. Specificity of all PCR reactions was tested via parallel reactions using water instead of cDNA (not shown). The PCR products were subjected to 1.5% agarose gel electrophoresis and visualized via ethidium bromide. Real-time PCR analysis was performed using CyBRGreen PCR Master Mix (Applied Biosystems, Darmstadt, Germany) under standard conditions. The experiments were performed with an ABI PRISM 7000 sequence detection system (Applied Biosystems). Expression levels of analyzed genes were normalized to glyceraldehyde 3-phosphate dehydrogenase messenger RNA (mRNA) expression.

Table 1. Oligonucleotides Used for RT-PCR
GenePrimer SequenceGenePrimer Sequence
  1. Specific primers were synthesized based on available sequences for each Wnt and Fzd family member. Primer design was performed with the program Primer 3 ( Cross-reaction of primers with the genes was excluded by comparison of the sequence of interest with a database (Blast 2.2, National Center for Biotechnology Information, Bethesda, MD), and all primers were intron-spanning. Sequences for Fzd2 and Fzd 7 were partially overlapping. PCR products were 300-400 bp in size.

Wnt1 fwgggtttctgctacgttgctactFzd1 fwaagttcttcctgtgctccatgt
Wnt1bwggaggtgattgcgaagataaacFzd1 bwctcccccagaaagtgatagttg
Wnt2 fwccctgatgaatcttcacaacaaFzd2 fwgggcactaagaaagaaggctgt
Wnt2 bwtctcccacaacacataacttcgFzd2 bwcgatgaacaggtacacgaagag
Wnt2b fwgatgccaaagagaagaggcttaFzd4 fwagctgcagttcttcctttgttc
Wnt2b bwcagccttgtccaagacacagtaFzd4 bwtttcacccagatgtactgatcg
Wnt3a fwtgaatttggaggaatggtctctFzd5 fwctgaggttctgtgcatggatta
Wnt3a bwtgggcaccttgaagtatgtgtaFzd5 bwgttccatgtcaatgaggaaggt
Wnt3 fwagtcatttccaacctcaagtggFzd7 fwcctaccctactgctccctacct
Wnt3 bwcaggctgtcatctatggtggtaFzd7 bwagataatgggtcgctctggata
Wnt4 fwattgaggagtgccaataccagtFzd9 fwggttttgactctcacctggttc
Wnt4bwctctcggacgtctacaaaggacFzd9 bwatggaaaagactccgattttca
Wnt5a fwtcatgaacttgcacaacaatga  
Wnt5a bwccgtcttaaactggtcatagccWIF1 fwtgtcatgaattctgaaggcaac
Wnt5b fwtctctgtgacttgctcttctgcWIF1 bwggcaaatacattttcctgggta
Wnt5b bwatgatgaacatctcgggtctctbCat fwgcaaggtagagtgctgaaggtt
Wnt6 fwaatgtcagttccagttccgtttbCat Bwgtctcctaacctctctccagca
Wnt6 bwtgctgtgcatccataaagagtcGSK3b fwcaagccaaactttgtgactcag
Wnt7a fwcgaaccctcatgaacttacacaGSK3b bwtatcaggatccagcaagaggtt
Wnt7a bwgggtcctcttcacagtaattggAxin1 fwgagcctgtcaacccctactatg
Wnt7b fwatgtaagtgtcacggagtgtcgAxin1 bwtccaacttttcttcagcctctc
Wnt7b bwggtgtactggtgcgtgttgtagDvl1 fwcagagtacctcctctcggctaa
Wnt8d fwcaggacttccatggttctaaggDvl1 bwtcgttgctcatgttctcaaagt
Wnt8b bwtccctcagggtagtcttgacatRyk fwgctgggattccaagtagacaac
Wnt14 (Wnt9a) fwagtggacttccacaacaacctcRyk bwaacccccacactgatgtaaaac
Wnt14 (Wnt9a) bwggccacaacaaatactctcacaWls fwtgggatttccatgacctttatc
Wnt14b (Wnt9b) fwctccctttctcgagtctaccaaWls bwatgatgaaagccatagccagtt
Wnt14b (Wnt9b) bwtttcctttgactccatgaggtt  
Wnt10a fwatcttcagcagaggttttcgag  
Wnt10a bwgagtccagaaagtccttggaga  
Wnt10b fwcataaccgcaattctggagttt  
Wnt10b bwcacttacacacgttgacccact  
Wnt11 fwtccctggaaacgaagtgtaagt  
Wnt11 bwccagtggtacttgcagtgacat  

Statistical Analysis.

Paired or unpaired 2-tailed t tests were performed using GraphPad InStat3 software depending on effective matching of the analyzed data. The standard deviation (SD) is indicated by error bars. Significance was assumed for P values <0.05.

cDNA Microarray Analysis and Statistical Procedures.

Total RNA samples were quality assayed using an Agilent 2100 Bioanalyzer (Palo Alto, CA). Hybridization probe generation (3-5 μg of total RNA) and GeneChip Rat Genome 230 2.0 array processing were performed according to the standard protocols available from Affymetrix (Santa Clara, CA) as described previously.17 Raw microarray data from Affymetrix CEL files were normalized, and probes (3 for each probe set) from 1 probe set are summarized using the median polish function resulting in 1 value per probe set, which is scaled to be on a log2 scale.18 Analysis of variance was performed using SAS software package MicroArray-Solutions version 1.3 (SAS Institute, Heidelberg, Germany). Loglinear mixed models were fitted for values of perfect matches. Gene annotation was obtained through the Affymetrix NetAffx web site (


  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Improved Purification of HSECs.

HSECs are routinely purified via ex vivo perfusion of rat liver followed by adhesion steps or CD31-based selection resulting in a mixture of HSECs and nonsinusoidal hepatic endothelial cells.19 In order to analyze HSECs specifically, we designed an isolation protocol generating highly purified cultures of HSECs. Following ex vivo perfusion of rat liver, HSEC cultures depleted of contaminating cell types were established using MACS-technology in combination with monoclonal antibodies 1.3 and 3.1 against the extracellular N-terminal part of stabilin-2. The purity of these HSEC preparations was routinely >95% as analyzed by expression of stabilin-2 (Fig. 1A) as well as of stabilin-1 and LYVE-1 (not shown). As expected, PECAM-1 was also expressed in the majority of HSECs, confirming its inability to select HSECs from nonsinusoidal hepatic endothelial cells. The absence of Kupffer cells from the stabilin-2+ HSEC isolates was demonstrated by lack of CD11b/c expression (Fig. 1A), and depletion of stellate cells was shown by the absence of desmin (not shown). Stabilin-2+ HSECs plated on collagen-coated plastic dishes showed the cobblestone pattern typical of all blood vascular endothelial cells; however, they were characterized by numerous fenestrations detectable within the cells by a lack of staining of surface or cytoskeletal structures (Fig. 1B, indicated by arrows) that are typical for HSEC. Furthermore, HSECs exhibited the typical netlike organization of microtubules along which stabilin-1– and stabilin-2–positve endosomes were identified.5 In addition, stabilin-2+ HSECs were shown to be endocytically active in vitro as revealed by rapid internalization of ligands such as AGE-BSA, AcLDL, and transferrin (Fig. 1C). Because HSECs are known to dedifferentiate upon culture with certain extracellular matrix substrates,14 we compared the phenotype of HSECs cultured on collagen I compared with liver extracellular matrix (Fig. 1D) after a 24-hour culture. However, we could not detect significant changes in the expression of the marker proteins stabilin-1/2 as well as LYVE-1.

thumbnail image

Figure 1. Hepatic sinusoidal endothelial cells (HSECs) isolated and purified using an improved protocol are highly pure, exhibit the characteristic morphology, and exert typical functions in endocytosis. (A) HSECs were selected by an anti–stabilin-2 monoclonal antibody (1.3) using the MACS technology. Isolated cells were subjected to FACS analysis using anti–stabilin-2 monoclonal antibody (3.1) and anti-CD31 monoclonal antibody. CD11b/c expression was used to quantify the degree of possible macrophages/Kupffer cell contamination. FACS data are representative for at least 3 independent experiments. Experiments with similar results are shown. (B) HSECs show the characteristic endothelial morphology upon culture (left). Bar = 100 μm. Immunofluorescent analysis of phenotypic marker expression in HSECs was performed using confocal microscopy. Cells were plated on collagen I–coated cover slips and subjected to immunocytochemistry for stabilin-1 (green)/β-tubulin (red) (middle) and stabilin-2 (green)/LYVE-1 (red) (right) after culture for 3 hours. Arrows show numerous fenestrations. Bar = 16 μm. (C) Ligand internalization of fluorescein isothiocyanate–labeled AGE-BSA (left), Alexa-Fluor-488–labeled acetylated LDL (middle), and Alexa-Fluor-555 transferrin (right) 15 minutes after addition of ligands demonstrated the high endocytic capacity of HSECs (blue; TOTO-3 iodide). Scale bar, 8 μm. (D) HSECs were plated on collagen I–coated (D1, D2) or liver extracellular matrix–coated (D3, D4) cover slips and subjected to immunocytochemistry for stabilin-1 (green)/β-tubulin (red) (D1, D3) and stabilin-2 (green)/LYVE-1 (red) (D2, D4) after culture for 24 hours. Bar = 8 μm. (E) Primary lung microvascular endothelial cells (LMEC) were purified using biotinylated CD31 antibody and the MACS technology (Streptavidin MicroBeads). Purity of the isolated cells was determined via FACS analysis of rat endothelial cell antigen (RECA) and PECAM-1.

Download figure to PowerPoint

Because stabilin-2 is not expressed by continuous endothelial cells, the isolation protocol for LMECs that were used as a control population for HSECs slightly differed. LMECs were isolated from rat lung after ex vivo perfusion using a biotinylated PECAM-1 antibody and Streptavidin MicroBeads (Fig. 1E). Purity of isolates was determined using rat endothelial cell antigen and was routinely around 95%.

Differential Expression of Wnt Signaling Components in HSECs and LMECs.

In order to identify components of the Wnt signaling pathway selectively expressed in HSECs, real-time reverse-transcription PCR (RT-PCR) was used to screen quantitatively for expression of Wnt family members and Fzd receptors in HSECs compared with LMECs. Total RNA was isolated from highly pure cell preparations directly after the ex vivo isolation and analyzed via real-time RT-PCR for Wnt and Fzd family members as well as Wnt signaling components with specific primers (Fig. 2A -D). Wnt2 (P < 0.05), and to a lesser extent the intracellular Wnt transporter Wntless (Wls)/evenness interrupted (evi) (P < 0.05) supporting Wnt secretion, and Wnt coreceptor Ryk were overexpressed by HSECs, whereas the Wnt-binding Wnt inhibitor WIF1 (P < 0.0005) and to a lesser extent Wnt3a, Wnt5a, Wnt7a, and Wnt11 were overexpressed by LMEC. Wnt4, Wnt5b, Wnt6, Wnt7b, Wnt10a/b, Fzd2, Fzd4, Fzd5, Fzd7, Fzd9, as well as GSK3β, β-catenin, Dvl1, and Axin1 were similary expressed in both endothelial cell types. Differential expression of Wnt2 mRNA levels were confirmed via northern blot analysis (Fig. 3A) and protein level was confirmed via western blot analysis using the custom-made affinity-purified Wnt2 antiserum (Fig. 3B). Further confirmation of lineage-specific Wnt2 expression in HSECs was performed via RT-PCR analysis of Wnt2 mRNA levels in various endothelial cell types (Fig. 3C-D). In lung lymphatic human microvascular endothelial cells, Wnt2 mRNA was only detectable in very low levels, whereas no expression of Wnt2 was detectable in human umbilical vein endothelial cells (Fig. 3C) or murine brain microvascular endothelial cells (Fig. 3D). These data suggest a specific role of Wnt2 in the biology of HSECs.

thumbnail image

Figure 2. Analysis of differential expression of components of the Wnt signaling pathway in HSECs and LMECs. Real-time RT-PCR quantification of (A) Wnt family members, (B) Fzd receptors, (C) Wnt signaling components Ryk and Wls, and (D) WIF-1 mRNA expression in HSECs (gray bars) compared with LMECs (black bars) using specific primers that were synthesized based on available sequences for each Wnt and Fzd family member. Resulting expression levels of were normalized via division through the mean expression value of the reference gene (Actin) and are shown as relative quantification units. Note the break in y axis scaling (A, D). Data are presented as the mean ± standard deviation from 3 independent experiments measured 3 times each. *P < 0.05. **P < 0.0005.

Download figure to PowerPoint

thumbnail image

Figure 3. Confirmation of differential Wnt2 expression in HSECs compared with LMECs and continuous endothelial cell of different vascular beds. (A) Northern blot analysis of Wnt2 expression in HSECs and LMECs confirmed increased mRNA levels in HSECs. 18S RNA served as a loading control. (B) Wnt2 was detected via western blot analysis using the custom-made affinity-purified Wnt2 antiserum and chemiluminescence. Equal protein amounts (200 μg, whole cell lysate) were loaded. β-Actin was included as a loading control. (C) RT-PCR analysis of hWnt2 and Fzd4, Fzd5, Fzd9 expression in lung lymphatic human microvascular endothelial cells (HMVEC-LLy) and human umbilical vein endothelial cells (HUVEC). (D) RT-PCR analysis of mWnt2 expression in pcDNA3 (empty vector; control) and pcDNA3-mWnt2 transfected bovine brain microvascular endothelial cells (bend-3).

Download figure to PowerPoint

Wnt2 Regulates Proliferation and Tube Fformation of HSECs.

We next investigated the effects of Wnt2 on the proliferation and differentiation of cultured HSECs. To analyze the effect of Wnt2 on proliferation, primary cultures of HSECs were incubated in normal growth media, control-conditioned media, or Wnt2-conditioned media with or without sFRP1 in the presence of H3-thymidine. The incorporated radioactivity reflecting the proliferation rate was quantified in whole cell lysates and normalized to the proliferation rate of HSECs in normal growth media (Fig. 4A). In the presence of Wnt2-conditioned media, proliferation of HSECs was increased by approximately 40%; proliferation induced by exogenous Wnt2 was inhibited by adding sFRP1. Moreover, culture of HSECs in the presence of increased concentrations of sFRP1 was able to reduce proliferation of the cells to approximately 70% of the proliferation in normal growth media without inhibitor (Fig. 4B).

thumbnail image

Figure 4. Wnt2-induced proliferation of HSECs. (A) Primary cultures of HSECs were incubated in normal growth media (NGM), control-conditioned media (Control-CM) or Wnt2-conditioned media (Wnt2-CM) with or without 3 μg/mL sFRP1 in the presence of H3-thymidine (0.185 MBq/mL) for 16-18 hours. (B) HSECs were incubated with H3-thymidine in normal growth media with or without sFRP1 (5 μg/mL). Incorporation of radioactivity was measured in whole cell lysates using a 1900CA liquid scintillation analyzer (Packard). Proliferation was normalized to HSECs in normal growth media; the percentage of untreated, normal growth media was set at 100%. Data are presented as the mean ± standard deviation from 3 independent experiments performed in duplicate. *,**Significantly different from control-conditioned media or untreated, normal growth media.

Download figure to PowerPoint

To evaluate the effects of endogenous Wnt on HSEC differentiation, we analyzed HSEC tube formation on growth factor–reduced Matrigel. HSECs were seeded on Matrigel in normal growth medium and in the presence of the Wnt inhibitors sFRP1 and WIF1 (Fig. 5). Tube formation was quantified using light microscopy by counting the number of junctions of branching tubes as well as the number of circles fully enclosed by capillary-like structures 16 hours after seeding. Addition of Wnt inhibitors resulted in a significant impairment of cord formation (P < 0.0005). Structures that did form despite Wnt inhibition were poorly formed or discontinuous.

thumbnail image

Figure 5. Inhibition of Wnt2 impairs endothelial tube formation of HSECs in vitro. HSECs were seeded on growth factor–reduced Matrigel in normal growth medium (untreated; left panel) in the presence of 5 μg/mL sFRP1 (middle panel) or 2.5 μg/mL WIF-1 (right panel). Number of junctions (black bars) and number of enclosed circles of capillary-like structures were quantified by light microscopy after 16 hours. Representative micrographs and statistical summary are shown. Data are presented as the mean ± standard deviation from 4 independent experiments. **Significantly different from untreated control. Bar = 100 μm.

Download figure to PowerPoint

In order to confirm that Wnt2 is the decisive Wnt ligand in these HSEC-related processes and that it functions as an autocrine signaling molecule in HSECs, we evaluated the effects of Wnt2 kockdown in our cells (Fig. 6). Equal numbers of HSECs were seeded on collagen I–coated cell culture dishes and were transfected with nonsilencing (control) or Wnt2 siRNA for 7 hours and successively cultured in normal growth media (Fig. 6A-B), control-conditioned media (Fig. 6C), or Wnt2-conditioned media (Fig. 6D) for 48 hours. Wnt2 knockdown resulted in reduced densities of the cells (Fig. 6B-C), and more detached cells could be detected in the medium (not shown). Addition of exogenous Wnt2 (Fig. 6D) was able to rescue the phenotype of Wnt2 siRNA-transfected HSECs. For quantification, cells were cultured in the presence of H3-thymidine for the indicated treatments and proliferation was assessed (Fig. 6). Wnt2 knockdown resulted in reduced proliferation of cells, and addition of exogenous Wnt2 was able to restore reduced proliferation of Wnt2 siRNA-silenced HSECs. Thus, Wnt2 supports proliferation and endothelial phenotype of HSECs.

thumbnail image

Figure 6. Wnt2-conditioned media rescue the phenotype of Wnt2 siRNA-transfected HSECs. HSECs were transfected with (A) nonsilencing (control) or (B-D) Wnt2 siRNA for 7 hours and successively cultured in (A, B) normal growth media (NGM), (C) control-conditioned media (Control-CM), or (D) Wnt2-conditioned media (Wnt2-CM) for 48 hours. Representative micrographs are shown. Bar = 100 μm. Proliferation was quantified by culturing the cells of the indicated treatments in the presence of H3-thymidine (0.185 MBq/mL) for 16-18 hours. Incorporation of radioactivity was measured in whole cell lysates using a 1900CA liquid scintillation analyzer (Packard). Results are representative of 4 independent experiments performed in duplicate. Experiments with similar results are shown.

Download figure to PowerPoint

Identification of Wnt2 Receptors in HSECs.

Next, we were interested in signaling pathways that mediate the biological effects of Wnt2 in HSECs. To characterize the general pathway of Wnt signaling in HSECs, cytosolic and nuclear levels of β-catenin were analyzed via western blotting after Wnt signaling inhibition (Fig. 7A). HSECs treated with the Wnt inhibitors sFRP1 and WIF1 showed reduced levels of β-catenin in the nuclear fraction, demonstrating that Wnt signaling in HSECs uses the canonical Wnt signaling pathway mediated by β-catenin. Because expression of the known Wnt2 receptor Fzd9 and the putative Wnt2 receptor Fzd5 was rather low in highly pure HSEC preparations compared with Fzd4 (which has not yet been shown to function as a Wnt2 receptor), we further analyzed Wnt2 receptors in HSECs via co-immunoprecipitation (Fig. 7B). Wnt2 was added to HSEC lysates as a recombinant protein with a GST tag; subsequently, Fzd receptors Fzd4, Fzd5, and Fzd9 were immunoprecipitated by commercially available anti-Fzd4, anti-Fzd5, and anti-Fzd9 antibodies. Immmunoprecipitates were separated via sodium dodecyl sulfate–polyacrylamide gel electrophoresis, and western blots were developed using anti-GST antibody. Wnt2-Fzd4 co-immunoprecipitates gave the strongest band compared with Wnt2-Fzd5 and Wnt2-Fzd9 co-immunoprecipitates, indicating that Fzd4 is indeed a Wnt2 receptor in HSECs.

thumbnail image

Figure 7. In HSECs, Wnt2 binds to Fzd4 as its novel receptor and activates the canonical Wnt signaling pathway. (A) Western blot analysis of β-catenin (bCat) levels in HSECs after generating cytosolic and nuclear fractions. Cells were cultured with or without 5 μg/mL sFRP1/2.5 μg/mL WIF-1 for 14 hours, homogenized in hypotonic buffer, and subjected for nuclear fractioning. Nuclear proteins were extracted using high salt conditions and dialysed against PBS. Equal protein amounts (50 μg) were loaded on the gels. Representative blots of 2 independent experiments with similar results are shown. (B) Co-immunoprecipitation of hWnt2-GST by Fzd in lysates of HSECs. Cell lysates (50 μg of total protein) were subjected to immunoprecipitation for Fzd (2 μg of antibody; Fzd4, goat anti-Fzd4; Fzd5, goat anti-Fzd5; Fzd9, goat anti-Fzd9) or 2 μg of goat immunoglobulin G (gt IgG; isotype control) in the presence of 0.1 μg recombinant hWnt2-GST (recombinant protein with GST tag), separated by gel electrophoresis, and transferred to polyvinylidene fluoride (PVDF) membranes. The resulting blot was immunoblotted with antibody to GST. Representative blots of 3 independent experiments with similar results are shown. (C) Immunofluorescent analysis of cultured HSECs using confocal microscopy. Cells were plated on collagen I–coated cover slips and stained with the custom-made affinity-purified Wnt2 antiserum (Wnt2), normal rabbit immunoglobulin G (isotype control), or preimmune serum. Arrows show cell surface localizations of Wnt2 (blue; TOTO-3 iodide). (D) FACS analysis of Wnt2 expression in HSECs. After cell permeabilization and subsequent staining using the affinity-purified Wnt2 antiserum (Wnt2 ap), 90% of the cells were positive for Wnt2 compared with the isotype control (rabbit immunoglobulin G; black line, left panel). Surface staining for Wnt2 shows that Wnt2 can be detected at the cell surface (92% positive cells; right panel). Experiments were repeated 3 times independently.

Download figure to PowerPoint

Because Wnt2 receptor interactions require that Wnt2 is secreted or at least may access its receptors on the surface of HSECs, we analyzed the cell surface association of Wnt2 in HSECs. Fluorescent staining and FACS analysis of Wnt2 expression in the cells demonstrated cell surface association of Wnt2 (Fig. 7C-D). Wnt2 can be detected in certain regions at the cell surface (Fig. 7C) and is more prominent in perinuclear regions and Golgi-like structures.

Identification of Wnt2 Downstream Target Genes in HSECs via Gene Profiling.

Because inhibition of Wnt2 impairs important functions of HSECs such as proliferation and morphogenesis, we performed cDNA microarray analysis using Affymetrix DNA chips in order to analyze Wnt2 target genes (Table 2). Total RNA was isolated from untreated (control) or sFRP1/WIF1-treated HSECs, and cDNA microarray analysis was performed. Data exploration and interpretation of the results was performed according to Chu et al.18 Gene expression in HSECs affected by Wnt inhibition is summarized in Fig. 8 using a Bonferroni cutoff of 3.7 representing a P value of 0.0002 (horizontal dashed line). Upon sFRP1/WIF1 treatment, gene expression of 183 genes was significantly altered: 56 genes were up-regulated and 147 genes were down-regulated compared with control HSECs. When focusing on the genes with the strongest change in expression (i.e., genes with a fold change <0.75 and >1.25), 23 genes were up-regulated and 32 genes were down-regulated in HSECs after Wnt inhibition. The newly identified Wnt target genes included genes for receptors (CXCR4, KDR/VEGFR-2, Ptprn), genes involved in proliferation (Ier3, Esm1, Akap7), genes involved in regulation of transcription (Hmgn2, Churc1), factors involved in transport (Slc17a1, Vamp1, Sncaip, Nostrin), and genes involved in adhesion processes (Jam2, Icam2, Clec1-4a).

Table 2. mRNA Profiles of HSECs Compared with Wnt Inhibitor-Treated HSECs
Gene*Gene TitleUniGene IDFold ChangeFunction
  • *

    Genes listed downward from Jam2 are repressed genes.

  • Fold induction of genes in control versus sFRP1/WIF1-treated HSECs (estimate of control versus sFRP1/WIF1-treated HSECs = log2 [fold change]). Genes with a fold change of 0.75 ≥ fold change ≥. 1.25 are presented.

Cxcr4Chemokine (C-X-C motif) receptor 4Rn.444312.10C-X-C chemokine receptor activity
Slc17a1Solute carrier family 17 (sodium phosphate), member 1Rn.111502.04Transport, ion transport
MawbpMAWD binding proteinRn.24171.97Biosynthesis, catalytic activity
RGD1307506, predictedSimilar to RIKEN cDNA 2310016C16Rn.1062641.76Response to oxidative stress
Ier3Immediate early response 3Rn.236381.74Regulation of T cell proliferation
Transcribed locusSarcolipin, predicted; similar to tumor necrosis factor, alpha-induced protein 8Rn.1478291.62Unknown
Esm1Endothelial cell-specific molecule 1Rn.63761.60Regulation of cell growth
Vamp1Vesicle-associated membrane protein 1Rn.319771.59Transport, vesicle-mediated transport
LOC680097/LOC6848Similar to germinal histone H4 geneRn.441201.55Unknown
Sncaip, predictedSynuclein, alpha interacting protein (synphilin)Rn.1511111.52Regulation of neurotransmitter secretion
Casp3Caspase 3, apoptosis-related cysteine proteaseRn.105621.41Apoptosis, cyclin-dependent protein kinase inhibitor activity
Trim36, predictedTripartite motif protein 36Rn.1379581.41Acrosome reaction
PtprnProtein tyrosine phosphatase, receptor type, NRn.110971.35Protein tyrosine phosphatase activity
Transcribed locusStrongly similar to XP_346949., hypothetical proteinRn.11301.34Unknown
Atoh7_predictedAtonal homolog 7 (Drosophila) (predicted)Rn.1335991.33Circadian rhythm
Transcribed locusHypothetical protein LOC305291; v-kit Hardy-Zuckerman 4 feline sarcoma viral oncogene homologRn.1636081.31Unknown
Transcribed locusVomeronasal 1 receptor, f5; similar to zinc finger protein 679Rn.433911.30Unknown
Transcribed locusStrongly similar to XP_342770.2, osteoclast inhibitory lectinRn.1672571.29Unknown
Rutbc3RUN and TBC1 domain containing 3Rn.669961.29 
Transcribed locusWeakly similar to NP_062289.1; DNA binding protein with his-thr domain (Mus musculus)Rn.125291.29DNA-binding
Transcribed locusHypothetical protein LOC362235; similar to ribosomal protein L1Rn.1585031.28Unknown
Pik3r1Phosphatidylinositol 3-kinase, regulatory subunit, polypeptide 1Rn.105991.27Protein amino acid phosphorylation
F3Coagulation factor IIIRn.99801.25Blood coagulation
Jam2Junction adhesion molecule 2Rn.64730.76Cell adhesion
Tarsl1Threonyl-tRNA synthetase-like 1Rn.441100.76Aminoacyl-tRNA ligase activity
RGD1308371, predictedHypothetical LOC293114 (predicted)Rn.38260.76Unknown
Pycr2Pyrroline-5-carboxylate reductase family, member 2Rn.1372660.75Proline biosynthesis
NostrinNitric oxide synthase traffickerRn.393830.75Transport
Lactb2Lactamase, beta 2Rn.12630.75Metabolism
Aldh6a1Aldehyde dehydrogenase family 6, subfamily A1Rn.20980.74Pyrimidine nucleotide metabolism
Akap7A kinase (PRKA) anchor protein 7Rn.1248020.74Protein kinase A binding
RGD1560873, predictedSimilar to RIKEN cDNA E230015L20 gene0.74Unknown
RGD1565002, predictedSimilar to dehydrogenase; reductase SDR family member 7 precursorRn.277710.74Metabolism
Clec14aC-type lectin domain family 14, member aRn.39620.74Sugar binding
Transcribed locusv-maf musculoaponeurotic fibrosarcoma oncogene family; protein G (avian); pyrroline-5-carboxylate reductase 1Rn.125300.74Metabolism
LOC289485Similar to BMP-2 inducible kinase0.74Protein amino acid phosphorylation
RGD1310224Similar to RIKEN cDNA 1810022C23Rn.31940.73Unknown
Transcribed locusSimilar to solute carrier family 35, member F30.73Transport, ion transport
Mccc2Methylcrotonoyl-coenzyme A carboxylase 2 (beta)Rn.336350.73Metabolism, ligase activity
LOC683319Similar to Pterin-4-alpha-carbinolamine dehydratase 2 (PHS 2); (DcoH-like protein DCoHm)0.71Metabolism
BckdhbBranched chain keto acid dehydrogenase E1, beta polypeptideRn.156230.71Amino acid catabolism
RGD1310681, predictedSimilar to RIKEN cDNA 6720467C03Rn.926010.69Unknown
Transcribed locusSimilar to glycerophosphodiester phosphodiesterase; domain containing 4/ myosin VIIARn.1536190.68Metabolism
Itm2bIntegral membrane protein 2BRn.1073350.67Apoptosis, induction of apoptosis
Churc1, predictedChurchill domain containing 1Rn.26960.67Transcriptional activator activity
C1qtnf1C1q and tumor necrosis factor related protein 1Rn.538800.67Metabolism
AsnsAsparagine synthetaseRn.111720.66Amino acid metabolism
Psat1Phosphoserine aminotransferase 1Rn.1008130.65Metabolism, L-serine biosynthesis
Scd2Stearoyl-coenzyme A desaturase 2Rn.835950.64Superoxide metabolism
KdrKinase insert domain protein receptorRn.888690.63Protein serine/threonine kinase activity
Icam2Intercellular adhesion molecule 2Rn.1622060.63Cell adhesion, protein binding
RGD1305677Similar to RIKEN cDNA 1810020E01Rn.1437680.62Unknown
Asah3l, predictedN-acylsphingosine amidohydrolase 3-like (predicted)Rn.324440.60Ceramide metabolism
Hmgn2High mobility group nucleosomal binding domain 2Rn.35170.58Regulation of transcription
Transcribed locusStrongly similar to XP_580072.1; similar to alkaline ceramidase 2Rn.63870.57Unknown
thumbnail image

Figure 8. Identification of Wnt target genes in HSECs by gene expression profiling. cDNA microarray analysis was performed using Affymetrix DNA chips, and data interpretation was performed according to Chu et al.18 Results are presented as a volcano plot. The x axis is the fold change of genes in control versus sFRP1 and WIF1-treated HSECs. The fold change representing the fold induction of genes in control versus sFRP1/WIF1-treated HSECs was calculated from the difference in least square mean estimates of the log 2 base scale intensity measures of the 2 treatments being contrasted [estimate of treatment = log2 (fold change)]. For each probe set, 3 Affymetrix DNA chips were used. The y axis is the significance as the negative logarithm of the P value associated with the difference. The horizontal line indicates the testing P value equal to 0.0002 using a Bonferroni cutoff of 3.7, and the vertical lines indicate a 1.25-fold change of fold changes (0.75 ≥ fold change ≥ 1.25). Genes typically involved in proliferation and known to be endothelial-related or angiogenesis-related are plotted by name.

Download figure to PowerPoint

Wnt2 Acts in HSECs to Cross-stimulate the VEGF Pathway via Induction of KDR/VEGFR-2.

Expression levels of genes known to be involved in proliferation, endothelial cell morphogenesis, and angiogenesis (CXCR4, MAWDbp, Ier3, Esm1, Jam2, KDR, Icam2, Hmgn2) were analyzed via real-time RT-PCR to confirm the results of the microarray analysis (Fig. 9A). In line with the microarray results, sFRP1/WIF1 treatment markedly induced CXCR4, MAWDbp, Ier3, and Esm1 expression, whereas Jam2, KDR/VEGFR-2, Icam2, and Hmgn2 expression was strongly down-regulated upon Wnt inhibition. Moreover, HSEC marker genes remain unaffected (not shown). Of the newly identified Wnt target genes, our attention was attracted by KDR/VEGFR-2, the major VEGF receptor. In addition to down-regulation of KDR/VEGFR-2 by Wnt signaling inhibition, stimulation of HSECs by exogenous Wnt2 resulted in increased KDR/VEGFR-2 mRNA expression levels. Furthermore, down-regulation of KDR/VEGFR-2 mRNA expression resulted in a reduction of KDR/VEGFR-2 protein expression. Western blot analysis demonstrated a decrease of KDR/VEGFR-2 protein 14-16 hours after Wnt signaling inhibition by sFRP1/WIF1 (Fig. 9C, left panel), as well as in Wnt2-siRNA transfected HSECs (right panel) 24 hours after transfection.

thumbnail image

Figure 9. Altered KDR/VEGFR-2 expression in HSECs upon Wnt2 inhibition by sFRP1/WIF1 and addition of exogenous Wnt2. (A) Cells were cultured in normal growth media without (untreated) or with 5 μg/mL sFRP1and 2.5 μg/mL WIF-1 for 14 hours. Total RNA was harvested and cDNA was synthesized. To confirm the results of the microarray analysis mRNA expression levels of genes known to be involved in proliferation, endothelial cell morphogenesis and angiogenesis (induced genes: CXCR4, MAWDbp, IER3, ESM1; repressed genes: JAM2, KDR, ICAM2, HMGN2) were analyzed quantitatively in untreated HSECs (black bars) or sFRP1/WIF1-treated HSECs (gray bars) via real-time RT-PCR. Resulting expression levels were normalized via division through the mean expression value of the reference gene (Actin) and are shown as relative quantification units. Note the break in y axis scaling. Data are presented as the mean ± standard deviation of 3 independent experiments performed 3 times each. *P < 0.05. **P < 0.0005. (B) HSECs were cultured in control-conditioned or Wnt2-conditioned media for 14 hours. Increased expression of KDR/VEGFR-2 mRNA levels was quantified using real-time PCR. Resulting expression levels of KDR/VEGFR-2 are shown as relative quantification. Data are presented as the mean ± standard deviation of 3 independent experiments performed 4 times each. (C) KDR/VEGFR-2 protein expression was detected via western blot analysis using an anti-KDR antibody and chemiluminescence. Cells were cultured with or without 5 μg/mL sFRP1 and 2.5 μg/mL WIF1 for 14 hours (left panel). HSECs were transfected with nonsilencing (control) or Wnt2 siRNA and harvested 24 hours after transfection (right panel). Equal protein amounts (50 μg, whole cell lysate) were loaded. β-Actin was included as a loading control. Representative blots of 3 (inhibition experiments) and 2 (siRNA transfections) independent experiments with similar results are shown.

Download figure to PowerPoint


  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Recent studies have suggested an important role of Wnt/β-catenin signaling in liver development, regeneration, and cancer and in vascular development.8, 20 Here, we identify Wnt2 as an important autocrine growth and differentiation factor specific for HSECs not produced by other microvascular endothelial cells, including LMECs. In contrast to HSECs, LMECs do not express Wnt2; rather, they differentially overexpress the Wnt inhibitor WIF, indicating general down-modulation of Wnt signaling in these cells. Identification of Wnt2 as overexpressed by HSECs is in line with the study of Zeng et al.,7 who analyzed Wnt ligands and Fzd receptors in mouse HSECs; however, we compared HSECs with other types of microvascular endothelia. In contrast to Zeng et al., we found a slightly different expression pattern for Wnt ligands and Fzd receptors in highly purified rat HSECs. These differences are most likely due to nonsinusoidal contaminants in murine HSECs, which are difficult to isolate in sufficient numbers; moreover, isolation of murine HSECs via selective adherence on collagen-coated or fibronectin-coated culture dishes requires culturing the cells for some time, which may induce changes in expression patterns. We overcame this problem by designing a novel protocol for the isolation of HSECs using the HSEC-specific surface proteins stabilin-1 and stabilin-2.

In order to act as an autocrine growth and differentiation factor for HSECs, Wnt2 must be secreted or at least gain access to its receptors on the cell surface. We showed that Wnt2 is indeed expressed on the cell surface of HSECs; however, we were unable to detect Wnt2 secretion into the cell supernatant of cultured HSECs, even after heparin treatment (not shown). Cell surface retention during secretion is observed with many secreted proteins. Specifically, cell surface association of Wnt proteins is well known.21 Differential expression in HSECs versus LMECs of the multipass transmembrane protein Wls/Evi, which is required for the secretion of Wnt proteins,22 supports the notion that HSECs are able to correctly process and secrete Wnt2. The present study shows that Wnt2 acts in an autocrine manner in HSECs, inducing proliferation and in vitro tube formation. Autocrine stimulation of HSEC proliferation by Wnt ligands is underscored by the finding that proliferation of HSECs was not only enhanced by addition of exogenous Wnt2 but was also reduced upon addition of the extracellular Wnt antagonists sFRP1 and WIF1. Autocrine stimulation of HSEC proliferation by Wnt2 was further corroborated by demonstrating that HSEC proliferation was specifically inhibited by an siRNA knockdown of Wnt2 that could be reversed via addition of exogenous Wnt2.

We present evidence that Wnt signaling in HSECs engages the canonical pathway. In general, identification of a bona fide receptor for a certain Wnt protein is hampered by lack of functionally active, purified Wnt proteins, the large number of Wnt and Fzd proteins, and the promiscuity of Wnt–Fzd receptor–ligand interactions. We used purified recombinant GST-tagged Wnt2 in combination with specific anti-Fzd antibodies to demonstrate that Fzd4, besides Fzd5 and Fzd9, may be an important receptor for Wnt2 in HSECs. Our results are in line with other recent studies. Fzd9 has been shown to function in Wnt/β-catenin signaling in mammalian cells and to act as a receptor for Wnt2.23 Ectopic expression studies demonstrated functional Fzd5–Wnt2 interactions, and mice that are deficient for Fzd5 show defects in yolk sac and placental vascularization that overlap with those seen in Wnt2−/− mice.11, 12 Interaction of Fzd4 and Wnt2 has been suggested indirectly by studies in rat granulosa cells.24 In contrast, we have demonstrated here direct binding of Wnt2 to Frz4 derived from HSECs.

Target identification for Wnt signals in HSECs by gene profiling revealed that KDR/VEGFR-2 is regulated by Wnt2. Inhibition of Wnt2 signaling in HSECs by sFRP1/WIF1 as well as by siRNA knockdown of Wnt2 resulted in down-regulation of KDR/VEGFR-2 expression; in contrast, Wnt2 positively regulated KDR/VEGFR-2 expression in HSECs. Other studies support these findings in different biological contexts, indicating that VEGF is a target of the Wnt signaling pathway in early colonic neoplasia25 and that regulation of Quek1/VEGFR-2 expression in quail somite development occurs via the cooperative actions of BMP4, FGF8, and Wnt-1/3a.26

In conclusion, the present study shows that Wnt2 is an autocrine growth and differentiation factor specific for HSECs that synergizes with the VEGF signaling pathway to exert its effects.


  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

The authors thank Xiaolei Yu for help in analyzing cDNA microarray data.


  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  • 1
    Seternes T, Sorensen K, Smedsrod B. Scavenger endothelial cells of vertebrates: a nonperipheral leukocyte system for high-capacity elimination of waste macromolecules. Proc Natl Acad Sci 2002; 99: 75947597.
  • 2
    Politz O, Gratchev A, McCourt PA, Schledzewski K, Guillot P, Johansson S, et al. Stabilin-1 and -2 constitute a novel family of fasciclin-like hyaluronan receptor homologues. Biochem J 2002; 362: 155164.
  • 3
    Kzhyshkowska J, Workman G, Cardo-Vila M, Arap W, Pasqualini R, Gratchev A, et al. Novel function of alternatively activated macrophages: stabilin-1-mediated clearance of SPARC. J Immunol 2006; 176: 58253832.
  • 4
    Kzhyshkowska J, Mamidi S, Gratchev A, Kremmer E, Schmuttermaier C, Krusell L, et al. Novel stabilin-1 interacting chitinase-like protein (SI-CLP) is up-regulated in alternatively activated macrophages and secreted via lysosomal pathway. Blood 2006; 107: 32213228.
  • 5
    Falkowska-Hansen B, Falkowski M, Metharom P, Krunic D, Goerdt S. Clathrin-coated vesicles form a unique net-like structure in liver sinusoidal endothelial cells by assembling along undisrupted microtubules. Exp Cell Res 2007; 313: 17451757.
  • 6
    Carpenter B, Lin Y, Stoll S, Raffai RL, McCuskey R, Wang R. VEGF is crucial for the hepatic vascular development required for lipoprotein uptake. Development 2005; 132: 32933303.
  • 7
    Zeng G, Awan F, Otruba W, Muller P, Apte U, Tan X, et al. Wnt'er in liver: expression of Wnt and frizzled genes in mouse. HEPATOLOGY 2007; 45: 195204.
  • 8
    Masckauchan TN, Kitajewski J. Wnt/Frizzled signaling in the vasculature: new angiogenic factors in sight. Physiology (Bethesda) 2006; 21: 181188.
  • 9
    Wright M, Aikawa M, Szeto W, Papkoff J. Identification of a Wnt-responsive signal transduction pathway in primary endothelial cells. Biochem Biophys Res Commun 1999; 263: 384388.
  • 10
    Cheng CW, Smith SK, Charnock-Jones DS. Wnt-1 signaling inhibits human umbilical vein endothelial cell proliferation and alters cell morphology. Exp Cell Res 2003; 291: 415425.
  • 11
    Monkley SJ, Delaney SJ, Pennisi DJ, Christiansen JH, Wainwright BJ. Targeted disruption of the Wnt2 gene results in placentation defects. Development 1996; 122: 33433353.
  • 12
    Ishikawa T, Tamai Y, Zorn AM, Yoshida H, Seldin MF, Nishikawa S, et al. Mouse Wnt receptor gene Fzd5 is essential for yolk sac and placental angiogenesis. Development 2001; 128: 2533.
  • 13
    Hsieh M, Boerboom D, Shimada M, Lo Y, Parlow AF, Luhmann UF, et al. Mice null for Frizzled4 (Fzd4-/-) are infertile and exhibit impaired corpora lutea formation and function. Biol Reprod 2005; 73: 11351146.
  • 14
    Sellaro TL, Ravindra AK, Stolz DB, Badylak SF. Maintenance of hepatic sinusoidal endothelial cell phenotype in vitro using organ-specific extracellular matrix scaffolds. Tissue Eng 2007; 13: 23012310.
  • 15
    Schledzewski K, Falkowski M, Moldenhauer G, Metharom P, Kzhyshkowska J, Ganss R, et al. Endothelium-specific hyaluronan receptor LYVE-1 is expressed by stabilin-1+, F4/80+, CD11b+ macrophages in malignant tumours and wound healing tissue in-vivo and in bone marrow cultures in-vitro: implications for the assessment of lymphangiogenesis. J Pathol 2006; 209: 6777.
  • 16
    Klein D, Schmandt T, Muth-Kohne E, Perez-Bouza A, Segschneider M, Gieselmann V, et al. Embryonic stem cell-based reduction of central nervous system sulfatide storage in an animal model of metachromatic leukodystrophy. Gene Ther 2006; 13: 16861695.
  • 17
    Zheng C, Li L, Haak M, Brors B, Frank O, Giehl M, et al. Gene expression profiling of CD34+ cells identifies a molecular signature of chronic myeloid leukemia blast crisis. Leukemia 2006; 20: 10281034.
  • 18
    Chu TM, Weir B, Wolfinger R. A systematic statistical linear modeling approach to oligonucleotide array experiments. Math Biosci 2002; 176: 3551.
  • 19
    DeLeve LD, Wang X, McCuskey MK, McCuskey RS. Rat liver endothelial cells isolated by anti-CD31 immunomagnetic separation lack fenestrae and sieve plates. Am J Physiol Gastrointest Liver Physiol 2006; 291: G1187G1189.
  • 20
    Ober EA, Verkade H, Field HA, Stainier DY. Mesodermal Wnt2b signalling positively regulates liver specification. Nature 2006; 442: 688691.
  • 21
    Smolich BD, McMahon JA, McMahon AP, Papkoff J. Wnt family proteins are secreted and associated with the cell surface. Mol Biol Cell 1993; 4: 12671275.
  • 22
    Bartscherer K, Pelte N, Ingelfinger D, Boutros M. Secretion of Wnt ligands requires Evi, a conserved transmembrane protein. Cell 2006; 125: 523533.
  • 23
    Karasawa T, Yokokura H, Kitajewski J, Lombroso P. Frizzled-9 is activated by Wnt-2 and functions in Wnt/beta-catenin signaling. J Biol Chem 2002; 277: 3747937486.
  • 24
    Ricken A, Lochhead P, Kontogiannea M, Farookhi R. Wnt signaling in the ovary: identification and compartmentalized expression of Wnt-2, Wnt-2b, and frizzled-4 mRNAs. Endocrinology 2002; 143: 27412749.
  • 25
    Zhang X, Gaspard JP, Chung DC. Regulation of vascular endothelial growth factor by the Wnt and K-ras pathways in colonic neoplasia. Cancer Res 2001; 61: 60506054.
  • 26
    Nimmagadda S, Geetha-Loganathan P, Scaal M, Christ B, Huang R. FGFs, Wnts and BMPs mediate induction of VEGFR-2 (Quek-1) expression during avian somite development. Dev Biol 2007; 305: 421429.