PC12 cells were a gift from Drs. M. Sendtner and S. Wiese (Department of Neurology, University of Wuerzburg, Germany). Cow pulmonary artery endothelial (CPAE) cells were purchased from ATCC (CCL-209). PC12 cells were cultured in DMEM with glutamax-I (Gibco) supplemented with 10% horse serum, 5% fetal bovine serum, 100 U/ml penicillin G, and 100 μg/ml streptomycin (Gibco) in 5% CO₂ at 37°C. For differentiation experiments, PC12 cells were plated on poly-L-ornithine coated tissue culture dishes and allowed to adhere over night (o/n). After one wash with serum-free DMEM, the cells were differentiated in serum-free DMEM containing 50 ng/ml human recombinant NGF (PAN Biotech) for 3 days 18. Although Fc-endostatin dimer application induced the formation of multicellular PC12 aggregates on Matrigel 8, Matrigel was not chosen for the current study since it is an extracellular matrix preparation generally used for endothelial tube formation assays.

For construction and expression of AP fusion proteins see 19. PC12 cells were either grown to 80% confluence or differentiated in 6-well plates, and AP staining was performed as described in 20. Staining was monitored with a Nikon Eclipse TE2000-U inverted microscope and documented using the Spot Insight QE Color imaging software (Visitron). Quantitative measurement of AP fusion protein binding to proliferating PC12 cells was carried out as described previously 21.

PC12 cells plated on poly-L-ornithine coated glass coverslips were fixed in phosphate-buffered saline (PBS) containing 4% paraformaldehyde at room temperature for 20 min, washed three times in prewarmed tris-buffered saline (TBS) for 5 min, and incubated in blocking buffer (TBS with 10% goat serum) at room temperature for 1 h. After washing once with prewarmed TBS, the cells were incubated with anti-Flt-1 (sc-316), anti-Flk-1 (sc-504) or anti-Neuropilin (sc-5541) antibodies (2 μg/ml, Santa Cruz Biotechnology) in blocking buffer at 4°C o/n. Unbound primary antibodies were removed by washing, and the cells were incubated in 20 mM ammonium chloride solution for 30 min to reduce autofluorescence. Cells were stained for 1 h at 37°C with a secondary Cy3-conjugated goat anti-rabbit antibody (1:500, Dianova) in blocking buffer. After washing, the coverslips were mounted in Kaiser's glycerol gelatine (Merck). Fluorescent preparations of proliferating PC12 cells were documented at 600-fold magnification (Nikon Eclipse TE2000-U). Image acquisition of differentiated PC12 cells was performed with a Zeiss Axiophot at 1000-fold magnification.

PC12 cell proliferation upon stimulation with 50 and 100 ng/ml VEGF₁₆₅ (R&D Systems) was assayed after 24 h, 48 h, and 72 h using the CellTiter 96^® AQueous One Solution Cell Proliferation Assay (Promega). These assays were performed using two different cell densities (2 × 10⁴ cells/cm² and 6.7 × 10⁴ cells/cm²) and full serum as well as serum-deprived conditions (0.1% and 0.4% horse serum). In addition, increasing VEGF₁₆₅ concentrations from 0.2 to 400 ng/ml were added to PC12 cells cultured in 5% fetal bovine serum, and [³H]thymidine incorporation was measured 72 hours after onset of stimulation. The fluorometric CaspACE™ Assay System and western blot analyses with an anti-cleaved caspase-3 antibody (1:1000, Cell Signaling, #9664) were used to monitor apoptosis of differentiated PC12 cells (see below). All experiments were performed in triplicate.

PC12 cells were grown to 50% confluence in 10 cm tissue culture dishes and NGF-differentiated for 72 h in serum-free DMEM. For induction of apoptosis, cells were washed three times with serum-free DMEM and incubated for 7.5 h under serum-deprived conditions in NGF-free DMEM. After addition of exogenous recombinant human VEGF₁₆₅ (R&D Systems) for the indicated time frames, PC12 cells were washed once with ice-cold PBS containing 100 μM sodium orthovanadate, followed by centrifugation at 5000 rpm for 5 min at 4°C. The cells were lysed in 200 μl of ice-cold lysis-buffer (HEPES, pH 7.8, 150 mM KOAc, 50 mM ß-glycerolphosphate, 25 mM NaF, 10 mM MgCl₂, 5 mM EGTA, 1 mM EDTA, 10% glycerol, 1% Triton X-100, 0,05% (v/v) ß-mercaptoethanol, 1 μg/ml aprotinin, 6 μg/ml chymostatin, 1 μg/ml leupeptin, 1 μg/ml pepstatin A, 1 mM PMSF, 1 mM sodium orthovanadate). Supernatants were collected after centrifugation at 14 000 rpm for 10 min at 4°C. Standardized samples containing 50 μg of whole protein (Bradford assay) were separated using 10–20% gradient gels. Proteins were wetblotted onto nitrocellulose and probed with anti-phospho-ERK1/2 (1:2000, Sigma, M8159) or anti-phospho-Akt-Ser473 (1:1000, Cell Signaling, #9271) antibodies. The blots were stripped and reprobed with antibodies detecting the respective non-phosphorylated proteins (anti-ERK1, 1:1000, Santa Cruz Biotechnology, sc-94; anti-Akt, 1:1000, Cell Signaling, #9272). Horseradish peroxidase-conjugated secondary antibodies (1:2000, Dianova) were visualized by enhanced chemiluminescence detection (Western Lightning™ Chemiluminescence Reagent System, PerkinElmer). Experiments were performed in triplicate.

To determine the expression profile of binding partners for murine VEGF₁₆₄, endostatin and the collagen XV endostatin homologue on neuronal and endothelial cell lines, PC12 and CPAE cells were incubated with AP fusion proteins. The AP-mVEGF₁₆₄ affinity probe strongly stained proliferating and differentiated PC12 cells (Fig. 1a, b). In contrast, AP-mVEGF₁₁₀ which lacks the C-terminal heparan sulfate binding domain, the endostatin domain of collagen XVIII (AP-mESXVIII), and the collagen XV endostatin homologue (AP-mESXV) did not bind to PC12 cells. Both AP-mVEGF₁₆₄ and AP-mESXVIII labelled CPAE cells while AP-mVEGF₁₁₀, AP-mESXV and control AP did not (Fig. 1c). Quantitative measurement of AP fusion protein binding to proliferating PC12 cells confirmed the above results (Fig. 1d). In agreement with lack of AP-mESXVIII and AP-mESXV binding to PC12 cells, recombinant human endostatins 22 did neither promote nor inhibit PC12 cell differentiation (data not shown). Thus, PC12 cells are not a useful model for understanding endostatin effects on cells.

Consistent with AP staining, it was shown by immunofluorescence that PC12 cells express the high-affinity receptor tyrosine kinases VEGF receptor 1 and 2 (VEGFR-1, VEGFR-2) as well as the low-affinity receptor neuropilin-1 (Fig. 2). These receptors are expressed on the cell surface of proliferating (Fig. 2a–c) and differentiated (Fig. 2e) PC12 cells and seem to possess a clustered morphology reminiscent of activated tyrosine kinases. Comparable results were obtained with antibodies from Santa Cruz Biotechnology and Dianova (data not shown).

While the addition of VEGF₁₆₅ had no effect on PC12 cell proliferation and neurite formation, a consistent but non-significant reduction of caspase-3 activity became apparent when VEGF₁₆₅ was administered to differentiated apoptotic PC12 cells (data not shown). To analyze VEGF signaling in PC12 cell survival, the activation of extracellular signal-regulated kinases ERK1/2 and Akt was examined by Western blot analyses using phospho-specific antibodies and their respective non phosphorylated counterparts. Control cultures demonstrated that ERK1/2 (p44/p42 MAPK) and Akt activation are sustained for 79.5 h in the presence of NGF (Fig. 3, lane 1). Removal of NGF for 7.5 h after 72 hours of differentiation led to a significant reduction of ERK1/2 and Akt phosphorylation (Fig. 3, lane 2). Exogenous addition of 100 ng/ml VEGF₁₆₅ to NGF-deprived PC12 cells resulted in transient ERK1/2 activity within 7–10 min which decreased almost to control levels 20 min after stimulation (Fig. 3, lanes 3, 4, and data not shown). Higher concentrations of recombinant VEGF (200 ng/ml) did not increase or prolong ERK1/2 activity (data not shown). Similar activation kinetics were observed for phosphorylation of Akt at serine 473 (Fig. 3). For comparison of signaling mechanisms, apoptotic PC12 cells were also NGF-stimulated. NGF-stimulation induced a much more pronounced and sustained activation of ERK1/2 and Akt which was even more prominent than in NGF treated non-apoptotic control cells (Fig. 3, lanes 5, 6).