Fig. 1 .  NCAM expression by TEC. Representative cytofluorimetric analysis showing the expression of the polysialilated embryonic NCAM in two different cell lines of renal TEC (A and B) and in the Kaposi's cell line KS-IMM (C), but not in normal renal endothelial cells obtained from glomeruli (D). (E) Western Blot analysis of TEC (1), HMEC (2), renal endothelial cells (3) and KS-IMM (4). Dark area is the isotypic control. Each experiment was performed at least five times with comparable results.

Fig. 2 .  NCAM expression in tumor vessels of renal carcinomas. (A–B) Representative indirect immunofluorescence of a renal clear cell carcinoma stained with rabbit anti-NCAM IgG (A) or with non-immune rabbit IgG as control (B). In panel A, a positive staining was observed on tumor cells and on endothelial cells (arrows) lining a vessel (original magnification: ×400). (C–D) Representative scanning immunoelectron microscopy of a vessel of renal clear cell carcinoma stained with rabbit anti-NCAM IgG (C) or with non-immune rabbit IgG as control (D). In panel C, immunogold labeling (arrows) of the endothelial cell surface facing the lumen of the vessel (Original magnification: ×2500).

Fig. 3 .  NCAM expression by TEC and HMEC transfected with PAX2 antisense and sense. (A) Representative gel showing PAX2 mRNA expression by TEC (lane 1) and by HMEC transfected with PAX2 sense (HMEC-S; lane 2). Lanes 3 and 4 show the absence of PAX2 mRNA expression in HMEC and TEC transfected with PAX2 antisense (TEC-AS). The vector in which PAX2 cDNA was cloned was used as positive control (+Ctr). RT-PCR was performed with the same primers used to clone the gene. (B) Representative Western Blot analysis showing PAX2 expression by HMEC-S and TEC but not by HMEC and TEC-AS. Beta-actin was used as loading control. (C–D) Representative cytofluorimetric analysis of NCAM expression by TEC, TEC-AS, HMEC and HMEC-S. Three experiments were performed with similar results.

Fig. 4 .  Binding of NCAM agonistic C3d peptide to TEC. Representative cytofluorimetric analysis of the binding of biotinilated C3d NCAM peptide to TEC or HMEC. Binding was revealed using PE-conjugated streptavidin. (A) Dose-dependent binding to TEC of C3d at 1 μM (dark line), 10 nM (grey line), 2 nM (dotted line) and 1 nM (pointed line). The dark area is streptavidin alone. (B) Binding to TEC of C3d (white area, 1 μM) as compared to the inactive modified C3d-ala (dark area, 1 μM). (C–D) Positive binding of C3d (grey line, 1 μM) was almost completely abrogated by cell preincubation with 7.5 U/ml heparin (dark line) in HMEC (C) but not in TEC (D). Three experiments were performed with similar results. The dark area is streptavidin alone.

Fig. 5 .  NCAM stimulation by NCAM agonistic C3d peptide favors TEC organization in capillary-like structures and apoptosis resistance. (A) TEC (5.0 × 104 cells/well) were plated on growth factor reduced Matrigel in DMEM + 5%BSA, stimulated with 3 μM C3d, C3d-ala, C3d in the presence of 7.5 U/ml heparin or vehicle alone, and observed at different times. Data are expressed as the mean ± SEM of the capillary-like surface evaluated by the computer analysis system in five different fields at ×20 magnification in duplicate wells of four different experiments. (B) Cell motility was analyzed during tube formation by time-lapse analysis at different times (see Material and methods). Results are expressed as mean ± SEM of three separate experiments. ANOVA with Dunnet's multicomparison test was performed among the different stimuli versus vehicle alone (*P < 0.05). (C–F) Representative micrograph of the network of tubes formed by TEC after 6 h. Increase in cell organization as compared to unstimulated cells (C) was observed when TEC were stimulated by 3 μM C3d (D) or C3d plus 7.5 U/ml heparin (F), but not by 3 μM C3d-ala (E). Magnification: ×150. (G) Apoptosis of TEC incubated for 48 h with vincristine (100 ng/ml), as evaluated by TUNEL assay. Reduction of apoptosis was observed by cell stimulation with 3 μM C3d, but not with 3 μM C3d-ala. Results are expressed as mean ± SEM of five separate experiments. ANOVA with Dunnet's multicomparison test was performed among vincristine plus C3d or C3d-ala and vincristine alone (*P < 0.05).

Fig. 6 .  Role of FGFR1 in the angiogenic activity of NCAM. (A) Representative Western Blot analysis of immunoprecipitates (IP) of cell lysates obtained after stimulation of TEC with vehicle, C3d (3 μM) or C3d-ala (3 μM) for 5 and 15 min. Immunoprecipitation was performed with the anti-NCAM Ab and the blots (IB) performed with anti-FGFR1, antiphosphorylated FGFR1 (P-FGFR) or anti-VEGFR2 Abs. (B) Representative Western Blot analysis of immunoprecipitates (IP) of cell lysates obtained after stimulation of TEC with vehicle, C3d (3 μM) or C3d-ala (3 μM) for 15 min. Immunoprecipitation was performed with the anti-FGFR1 Ab and the blot (IB) performed with the anti-NCAM Ab. Three experiments were performed with similar results. (C) Capillary-like formation by TEC (5.0 × 10⁴ cells/well) plated on growth factor reduced Matrigel after stimulation with 3 μM C3d, VEGF (10 ng/ml), bFGF (10 ng/ml) or vehicle alone, in the presence or absence of anti-FGFR1 Abs (see Material and methods). Data are expressed as the mean ± SEM of the capillary-like surface evaluated by the computer analysis system in five different fields at ×20 magnification in duplicate wells of three different experiments. ANOVA with Newmann–Keuls multicomparison test was performed: vehicle versus stimuli (§P < 0.05) and anti-FGFR treated cells plus stimuli versus stimuli alone (*P < 0.05).

Fig. 7 .  Effect of VEGF on the expression of NCAM and PAX2 by normal endothelial cells during formation of capillary-like structures. (A–C) Representative cytofluorimetric analysis of NCAM expression by HMEC during formation of capillary-like structures induced by 10 ng/ml VEGF on Matrigel. NCAM expression was negative before stimulation (A), was acquired 24 h after stimulation (B), to disappear after 48 h (C). Dark area is the isotypic control. Similar results were obtained using renal endothelial cells derived from human glomeruli. Each experiment was performed at least five times with comparable results. (D–G) Representative micrographs showing the immunofluorescence staining for PAX2 on capillary-like structures spontaneously formed by TEC or formed by HMEC stimulated with VEGF (10 ng/ml) after 24 h. Nuclear expression of PAX2 was observed in TEC (D) but not in HMEC (F). Panels E and G are phase contrast micrographs of the same field observed by fluorescence in panels D and F, respectively. Magnification: ×250. Three experiments were performed with similar results.

Fig. 8 .  Transient NCAM expression during stem cell differentiation into endothelial cells. (A–D) Representative cytofluorimetric analysis of NCAM expression in CD133+ renal stem cells cultured in endothelial differentiating medium. Undifferentiated cells were negative (A), but they acquired NCAM after 5–10 days (B and C) of culture and lost NCAM expression after 15 days (D). (E–F) Representative cytofluorimetric analysis of NCAM expression in peripheral blood mononuclear cells cultured in endothelial differentiating medium to obtain EPC. Undifferentiated cells were negative (E), but they acquired NCAM after 5 days of culture (F). Dark area is the isotypic control. (G) Percentage of CD133+ renal stem cells cultured in endothelial differentiating medium expressing the endothelial markers Muc18, CD105, KDR (VR2), VE-cadherin (VE-cad) and NCAM. Data are the mean ± SEM of five different cell preparations.

Fig. 9 .  Effect of C3d-saporin on TEC apoptosis. TEC (dark column) and HMEC (grey column) were stimulated with 10 μM saporin alone or with 10 μM C3d-saporin for 30 min at 37°C, extensively washed and incubated for 48 h in DMEM + 5% FCS. Apoptosis was determined by TUNEL. A significant increase in apoptosis was induced by C3d on TEC but not HMEC. Results are expressed as mean ± SEM of five separate experiments. ANOVA with Dunnet's multicomparison test was performed between C3d-saporin and saporin alone (*P < 0.05).