Synovial tissues and immunohistochemistry
Arthritic synovial tissues from knee joints of the patients with RA (n = 8) and normal synovium from hip joints of the patients with femoral neck fracture (normal control; n = 3) were obtained by surgery at Keio University Hospital and used for the experiments. They were fixed with formalin and paraffin-embedded specimens were subjected to immunohistochemistry. The primary antibodies against VE-cadherin (LifeSpan Biosciences, Seattle, WA), CD31 (DAKO, Glostrup, Denmark) or CD34 (DAKO) and biotinylated secondary antibodies were used. After incubation with streptavidin-biotin-complex, the color was developed with 3,3’-diaminobenzidin tetrahydrochlorides. Non-immune rabbit IgG instead of the primary antibody were incubated with tissue sections as a negative control. To identify the extracellular matrix-associated vascular channels, tissue sections were proceed to periodic acid-Schiff (PAS) staining. Number of PAS-positive vessels containing red blood cells and lined with CD31/CD34-positive cells, CD31/CD34-negative cells or the mixture in RA and normal synovial tissues was counted under microscope. For double-immunofluorescent staining for VE-cadherin and osteoblast (OB)-cadherin (Invitrogen, Camarillo, CA), Opal 4-color Kit (PerkinElmer, Waltham, MA) was used and observed by a confocal laser-scanning microscope (LSM 700; Carl Zeiss, Oberkochen, Germany). Informed consent was obtained from the patients for the experimental use of the samples. The study protocol that compiled with the principles outlined in the Declaration of Helsinki was approved by the Ethical Committee Review Boards at Keio University and Nippon Dental University.
Cell Cultures
Rheumatoid synovial fibroblast-like cells (RSFLs) of the human were purchased from Health Science Research Resources Bank (Osaka, Japan; RSFL1, HT32251238; RSFL2, HT51213801; RSFL3, HT47768187; and RSFL4, HT42487497) or isolated from RA synovium (RSFL5-7) [13, 14]. RSFLs between 5–8 passages were used to neglect contamination of different cell-types and phenotypic alterations of the cells according to a previous study [15]. They were maintained in DMEM containing 10% fetal bovine serum (FBS) and passaged at 7.6 x 103 cells/cm2 when they reached at subconfluent states. Human umbilical vein endothelial cells (HUVEC) and GFP-labeled HUVEC (GFP-HUVEC; Angio-Proteomie, Boston, MA) were cultured in Quick Coating Solution (Angio-Proteomie)-coating culture dishes in EGM-2 MV medium (Lonza, Basel, Switzerland) containing 5% FBS. For stimulation of RSFLs, they were cultured in 1% FBS or FBS-free medium containing 0.2% lactalbumin hydrolysate for 24 h, and then treated with VEGF165 (80 ng/ml; R&D Systems, Minneapolis MN) and/or tumor necrosis factor-α (TNF-α; 10 ng/ml; AbD Serotec, Oxford, UK) for 72 h. For the experiment under the hypoxic condition, RSFLs were cultured in 1% O2 for 72 h using Bionix hypoxic culture kit (Sugiyama-gen, Tokyo, Japan) or treated with 150 µM deferoxamine (Sigma-Aldrich, St. Louis, MO) for 48 h. To analyze VEGF receptor (VEGFR)-dependency of VE-cadherin expression, RSFLs cultured for 24 h in FBS-free medium were treated with placental growth factor 1 (PROSPEC, Ness-Ziona, Israel) or SU1498 (1 µM; Calbiochem, San Diego, CA) for 72 h. For knockdown experiments, RSFLs were transfected with short interfering RNAs (siRNAs) against neuropilin 1 (siNRP1; 75 nM, sc-36038; Santa Cruz Biotechnology, Santa Cruz, CA), VE-cadherin (siVE-CDH; 75 nM, L-003641) or a negative control siRNA (siCtrl: 100 nM D-001810-10-05; Dharmacon, Chicago, IL). Experimental protocols were reviewed and approved by the institutional review boards of Nippon Dental University and Keio University.
Immunocytochemistry
RSFLs and HUVEC were cultured on Lab-Teck Chamber II (ThermoFisher Scientific, Waltham, MA) until 24 h after reaching the confluency to mature cell-cell contacts, and fixed in 3.7% paraformaldehyde for 15 min at 23˚C. After treatment with 0.1% Triton X-100 in PBS for 7 min, the cells were reacted with antibodies to VE-cadherin or epithelial (E)-cadherin (clone HECD-1; R&D Systems), followed by incubation with Alexa Fluor 546 anti-rabbit IgG (Invitrogen). They were observed by a confocal laser-scanning microscope.
Immunoblot
Immunoblot
Total proteins of RSFLs were prepared by homogenization in lysis buffer (50 mM HEPES, 150 mM NaCl, 1 mM EDTA, 1 mM dithiothreitol, 0.5% Nonidet P-40) containing proteinase inhibitors (Complete mini proteinase inhibitor cocktail, Roche, Mannheim, Germany) on ice. For isolation of membrane and cytoplasmic/nuclear fractions, RSFLs were treated with ProteoExtract Native Membrane Extraction Kit (Calbiochem) according to the company’s instruction. To analyze the VE-cadherin ectodomain shedding, proteins were precipitated from the culture media with 3.3% trichloroacetic acid and the samples were resolved by SDS-polyacrylamide gels. Proteins electrotransferred onto PVDF membranes were subjected to immunoblot using antibodies against CD31, CD34, neural (N)-cadherin (Invitrogen), OB-cadherin, VEGF (clone A-20; Santa Cruz Biotechnology), VEGFR1, VEGFR2, NRP1 (Cell Signaling), GAPDH (Ambion, Austin, TX) or β-actin (Sigma-Aldrich). For immunoblot of VE-cadherin, antibodies specific to the COOH-terminal region (clone D87F2; Cell Signaling, Danvers, MA: and sc-6458; Santa Cruz Biotechnology) or the NH2-terminal region of VE-cadherin (sc-52751, Santa Cruz Biotechnology) were used. After incubation with secondary antibodies, the bands were detected by ECL Select (GE Healthcare, Buckinghamshire, UK) and visualized using EZ-Capture (ATTO, Tokyo, Japan). Intensity of the bands (n = 4) standardized by that of β-actin bands were quantified using ImageJ software (NIH, Bethesda, MD; http://imagej.nih.gov/ij).
Reverse Transcription (Rt)-pcr
Total RNAs isolated from RSFLs were reverse transcribed into cDNA using SuperScript II Reverse Transcriptase (Invitrogen) with oligo-dT primer and applied for PCR amplification. VEGF121 (452 bp), VEGF165 (584 bp) and VEGF189 (656 bp) were detected by RT-PCR using the primers (forward, 5’-CGAAACCATGAACTTTCTGCTGTC-3’; reverse, 5’-TCACCGCCTCGGCTTGTCACAT − 3’). For the detection of VEGF206, a specific primer set (forward, 5’-GAAGGAGGAGGGCAGAATCATCACG-3’; reverse, 5’-TCCAGGGCATTAGACAGCAGCG-3’) was used.
Cell Adhesion Assay
RSFLs adhesion on culture plates was monitored using RTCA-DP xCELLigence (Roche), which can monitor the adhesion by measuring electronic impedance at the cell-sensor electrode interface integrated on the bottom of culture E-plate [16, 17]. Each plate was pre-coated with 20 µg/ml of VE-cadherin or E-cadherin Fc-chimeras (R&D Systems) for 3 h at 37˚C and then incubated with DMEM containing 0.25% FBS for 1 h at 37˚C. RSFLs were seeded on a plate at a density of 1.7x103 cells/plate, and installed on the analyzer. Cell adhesion was monitored every 2 min for 5 h.
Tube Formation Assay
RSFLs (2.5x105 cells/cm2) labeled with CellTracker Red CMTPX (ThermoFisher Scientific) were cultured in serum-free condition for 24 h on dishes coated with Matrigel (µ slide angiogenesis; ibidi, Nippon Genetics, Tokyo, Japan) and observed by a phase contrast microscope (LSM700, Carl Zeiss). For inhibition of branching morphology, RSFLs were cultured on the gels for 24 h in culture medium containing anti-VE-cadherin antibody (10 µg/ml; BD biosciences, Franklin Lakes, NJ), anti-E-cadherin antibody (10 µg/ml; R&D Systems), non-immune rat IgG (10 µg/ml; Vector Laboratories, Burlingame, CA), VE-cadherin Fc-chimera (20 µg/ml) or E-cadherin Fc-chimera (20 µg/ml), and treated with siVE-CDH (75 nM) or siCtrl (100 nM). The branching point numbers were quantitatively evaluated according to the previous study [18, 19]. To determine tube-like formation by RSFLs in the presence of HUVEC, RSFLs labeled with CellTracker Red CMTPX and GFP-HUVEC were co-cultured on basement membrane matrices for 24 h and monitored by a confocal laser-scanning microscope. The data of tube-like structures were analyzed using Zen software (Carl Zeiss).
Statistical analysis
Data obtained by the tube assay were shown as mean ± SD. Results between the two independent groups were determined by the student’s t-test or Mann-Whitney U test. Comparisons among more than three groups were analyzed by Tukey-Kramer’s HSD or Steel-Dwass test. P values less than 0.05 were considered significant.