Figure 1.
Identification of αHS that preferentially recognize HS epitopes of tumor vasculature. A–D
, Patient glioblastoma tumor sections were stained for HS using the αHS antibodies AO4B08 (A and D), EV3C3 (B) or HS4E4 (C) (green) and for vWF (A–C, endothelial marker; red) or integrin αvβ3 (D, marker of activated endothelium; red), and analyzed by fluorescence microscopy. Strong co-association between HS epitopes and tumor vasculature is indicated by merged images (A–D, right panels; yellow). Images shown are representative of three different tumors and at least five sections per tumor, and were obtained at 20× magnification. Scale bar, 50 µM. E, Upper panels: Specificity of αHS was assessed by AO4B08 staining of U-87 MG glioblastoma xenograft tumor sections following either no treatment (Ctrl) or pre-treatment with Heparinase III lyase (HS lyase) to digest HS. Lower panels: 3G10 antibody was used to detect HS stubs of HSPGs that remain after HS lyase digestion. Expecedly, 3G10 staining was not detectable in untreated tumor sections with intact HS (Ctrl, left panel); however, 3G10 staining of HS lyase treated specimens (right panel) shows a vessel like pattern that largely resembles the AO4B08 staining (upper left panel). Scale bar, 20 µM.
Table 1.
αHS used in this study and their epitope specificities.*
Figure 2.
Potent pro-angiogenic effects of αHS in primary human ECs. A–B
, HUVECs were grown in serum free medium in the presence of the indicated concentrations of either heparin (A) or the low molecular weight heparin, Tinzaparin (B). Cell proliferation was determined by measurement of incorporated [3H]thymidine over a period of 48 h. C, HUVECs were grown in serum free medium in the absence (Ctrl) or presence of AO4B08 (αHS, 3 µg/ml) for 72 h. Cell morphology was visualized by crystal violet staining, and captured using an inverted light microscope equipped with a digital camera. Left and mid panels show representative pictures of an elongated EC phenotype in αHS-treated as compared with control cells. Right panel shows quantitative analysis of the length of individual cells in three separate microscopic fields per well (n = 4), using Image J software. D, HUVECs were cultured on matrigel, i.e. a tumor-derived extracellular matrix, in serum free medium in the absence (Ctrl) or presence of AO4B08 (αHS; 3 µg/ml) and were allowed to form tubes for approximately 20 h. Left and mid panels show representative pictures of increased tube formation in αHS-treated as compared with control cells from at least three independent experiments. Right panel shows quantitative analysis of the number of intact tubes from three microscopic fields per well (n = 4) using Image J software. E and F, αHS stimulates EC proliferation: HUVECs were grown in serum free medium in the absence or presence of AO4B08 (αHS) at the indicated concentrations and assessed for proliferation by [3H]thymidine incorporation over a period of 3 h. F, HUVECs were subjected to cell cycle phase analysis after 6 h without (Ctrl) or with AO4B08 (αHS) treatment. Fraction of cells in the respective cell cycle phase was determined by flow cytometry following nuclear staining with propidium iodide. Data are presented as the average ± S.D. * Statistically different from untreated Ctrl, P<0.05.
Figure 3.
αHS counter-acts hypoxia and starvation induced cell-death in ECs. A
, HUVECs were grown for 72 h in full medium (10% FBS), serum free medium, or in serum free medium supplemented with AO4B08 (αHS; 3 µg/ml), as indicated, under hypoxic conditions (1% O2). Cell lysates were analyzed for total and cleaved caspase-3 and 7 as well as α-tubulin by immunoblotting. Upper panel shows representative immunoblots from three independent experiments. Lower panels show quantification of cleaved caspase/total caspase ratios from a representative experiment. B, HUVECs were grown in serum free medium in the absence (Ctrl) or presence of AO4B08 (αHS; 3 µg/ml) under hypoxic conditions for 72 h, and then stained with crystal violet. Left panel shows representative images of stained cells. Right panels shows quantitative analysis of viable cells. Data are presented as the average±S.D. *Statistically different from untreated Ctrl, P<0.05.
Figure 4.
αHS stimulates the proliferation of ECs and transformed cells in an HSPG-dependent manner.
A, HUVECs were grown in serum free medium in the absence (white bars) or the presence of AO4B08 (black bars; 1.25 µg/ml) and with the indicated concentration of heparin for 3 h and assessed for cell proliferation by the [3H]thymidine incorporation assay. *Statistically different from untreated control, P<0.05; # Statistically different from AO4B08-treated, P<0.05. Control cells were incubated without antibody. B, HUVECs were grown in serum free medium in the absence (white bars) or the presence of AO4B08 (black bars; 1.25 µg/ml) without (Ctrl) or with CS-6, CS4, or CS-0 (10 µg/ml), as indicated. Cell proliferation was determined by the [3H]thymidine incorporation assay. *Statistically different from untreated control, P<0.05. C, HUVECs were grown in serum free medium in the absence (white bars) or the presence of heparin (black bars; 10 µg/ml) without or with the different αHS antibody clones (1.25 µg/ml), as indicated. Cell proliferation was determined by the [3H]thymidine incorporation assay. *Statistically different from untreated control, P<0.05; # Statistically different from αHS-treated, P<0.05. D, Wild type (CHO-K1) cells and PG- deficient (PgsA-745) CHO cells were grown in serum free medium in the absence or the presence of different αHS antibody clones (1.25 µg/ml), as indicated. Cell proliferation was determined by the [3H]thymidine incorporation assay. *Statistically different from untreated control, P<0.05; # Statistically different from αHS-treated CHO-K1 cells, P<0.05. E and F, Human umbilical artery ECs (HUAECs; E) and U-87 MG glioblastoma cells (F) were grown in serum free medium in the presence of AO4B08 (αHS) at the indicated concentrations and proliferation was determined by the [3H]thymidine incorporation assay. *Statistically different from untreated control, P<0.05.
Figure 5.
αHS-induced proliferation of ECs and transformed cells depends on p38 MAPK signalling activation. A,
HUVECs were incubated in serum free medium with or without αHS (AO4B08; 1.5 µg/ml) for 5, 15, 45, and 60 min, and phosphorylated receptor tyrosine kinase proteins were determined on pooled cell lysates by human anti-phospho receptor tyrosine kinase antibody array analysis as described under Methods. Shown are representative immunoblots from two independent experiments. Indicated areas show FGFR, column 1: FGFR1; 2: FGFR2; 3: FGFR3; 4: FGFR4. PDGFR, column 1: PDGFRα; 2: PDGFRβ; 3: SCFR; 4: Flt3. VEGFR, column 1: VEGFR1; 2: VEGFR2; 3: VEGFR3. B, HUVECs were incubated in serum free medium with or without αHS (EV3C3; 1.5 µg/ml) for 5, 15, 45, and 60 min, and phosphorylated kinase proteins were determined on pooled cell lysates by human anti-phospho kinase antibody array analysis as described under Methods. Left panels show representative immunoblots from three independent experiments. Right panel shows quantification of p-p38 MAPK, p-CREB, and p-ERK1/2 levels at the various conditions, presented as relative intensities vs. array reference (red box). C, HUVECs were either untreated (Ctrl) or incubated with αHS (AO4B08;1.25 µg/ml) for the indicated time periods followed by immunoblotting for phospho-p38 MAPK, total p38 MAPK and α-tubulin. Left panel shows representative immunoblots of three separate experiments. Right panel shows quantification of relative phospho-p38 MAPK levels (P-p38/α-tubulin) from a representative experiment. D, Same experiment as in (C) was performed with mouse lung microvascular ECs, showing comparable induction of p38 MAPK phosphorylation by αHS as in HUVECs. E–F, αHS-induced proliferation depends on p38 MAPK. Proliferation was determined by [3H]thymidine incorporation in mouse ECs (E) and U-87 MG glioblastoma cells (F) following treatment with the p38 MAPK inhibitor SB203580 (1 µM) and/or αHS (AO4B08; 1.25 µg/ml) for 3 h, as indicated. Basal cell proliferation was not significantly affected by p38 MAPK inhibition whereas αHS induced proliferation was almost completely reversed. * Statistically different from untreated control, P<0.05; # Statistically different from αHS-treated, P<0.05.