|Home | About | Journals | Submit | Contact Us | Français|
Previously, we reported that a predominant action of an IGF-1R-targeted antibody was through inhibiting tumor-derived VEGF, and indirectly, angiogenesis. Here we examined the direct anti-angiogenic activity of the IGF-1R-targeted antibody SCH717454 that inhibits ligand-receptor binding, and the mechanism by which tumors circumvent its anti-angiogenic activity. Inhibition of ligand stimulated activation of IGF-1R, insulin receptor (IN-R) or downstream signaling (phosphorylation of Akt [Ser473]) was determined by receptor-specific immunoprecipitation and immunoblotting. Inhibition of angiogenesis was determined by proliferation and tube formation using human umbilical vein endothelial cells (HUVECs) in vitro and in Matrigel plugs implanted in mice. SCH717454 blocked IGF-1 but not IGF-2 stimulated phosphorylation of Akt in sarcoma cells. Immunoprecipitation using anti-IGF-1R and anti-IN-R antibodies revealed that SCH717454 equally blocked IGF-1 and IGF-2 stimulated IGF-1R phosphorylation, but not IGF-2 stimulated phosphorylation of IN-R. SCH717454 completely blocked VEGF-stimulated proliferation and tube formation of HUVECs, but exogenous IGF-2 and insulin circumvented these inhibitory effects. Co-culture of HUVECs with IGF-2-secreting tumor cells completely abrogated SCH717454 inhibition of VEGF-stimulated HUVEC tube formation. In mice SCH717454 inhibited angiogenesis in VEGF-infused Matrigel plugs, but had no inhibitory activity when plugs contained both VEGF+IGF-2. These results reveal for the first time, a role for IGF-1R signaling in VEGF-mediated angiogenesis in vitro and indicate direct anti-angiogenic activity of SCH717454. Both in vitro and in vivo IGF-2 circumvented these effects through IN-R signaling. Many childhood cancers secrete IGF-2, suggesting that tumor-derived IGF-2 in the microenvironment maintains angiogenesis in the presence of IGF-1R-targeted antibodies allowing tumor progression.
Many childhood cancers (including rhabdomyosarcoma, osteosarcoma, Ewing sarcoma, neuroblastoma, medulloblastoma and Wilms tumor) show the presence of both active Type-1 insulin-like growth factor receptor (IGF-1R) and the autocrine production of its ligands IGF-1 and or IGF-2 (1). IGF-1 and -2 and IGF-1R regulate all aspects of the malignant phenotype (2) with IGF-1R being activated by its ligands and also indirectly by steroid hormones (3). Activated IGF-IR is capable of phosphorylating other tyrosine-containing substrates of which the insulin receptor substrates (IRS-I-4) link the receptor to a cascade of enzyme activations via PI3K-Akt-mTOR and RAF-MAPK systems (4).
Deregulated insulin-like growth factor signaling through the IGF-1R thus potentially offers an important molecular target for pediatric cancer therapeutics development. For example, the alveolar subtype of rhabdomyosarcoma is associated with t(2;13)(q35;q14) and t(1;13)(q36;q14) which generate PAX3-FKHR or PAX7-FKHR chimeric transcription factors that enhance transcription of target genes including IGF-1R (5). For the embryonal subtype of rhabdomyosarcoma, the IGF-2 gene, which normally shows monoallelic expression as a result of silencing of the maternal allele through imprinting, shows biallelic transcription (6). This loss of imprinting at the IGF-2 locus may be a primary genetic event for embryonal rhabdomyosarcoma. IGF-1R is a potent mediator of autocrine growth in Ewing sarcoma (7, 8). Cases of Ewing sarcoma with the Type-1 EWS-FLI1 chimeric transcription factor are associated with an improved prognosis and with lower IGF-1R expression compared to cases with non-Type 1 translocations (9). EWS-FLI1 silencing leads to increased levels of insulin-like growth factor binding protein-3 gene (IGFBP-3), a major regulator of IGF-1 (10). Additionally, IGF-1 is a mitogen for osteosarcoma, neuroblastoma, brain tumors (including glioblastoma, astrocytoma, medulloblastoma), Wilms tumor, and hepatocellular carcinoma (11–17). The role of the IGF-1 axis in acute lymphoblastic leukemia is less clearly defined (18).
The role of IGF-1R signaling in the pathogenesis of childhood cancer, and its role in preventing apoptosis induced by a multitude of cellular stresses including cytotoxic drugs, radiation and hypoxia (19) indicate that targeting this pathway may have considerable utility for therapy of these rare cancers. As dysregulated IGF-I signaling is common to several adult malignancies, targeting IGF-IR has become a major focus for therapeutics development (20, 21). Currently there are both small molecule drugs and fully human or humanized antibodies directed at the IGF-1R. At least six fully human or humanized antibodies have entered adult phase-I to -III clinical trials. These agents show specificity for IGF-IR although they may inhibit chimeric receptors formed through heterodimerization with the insulin receptor.
In preclinical cancer models antibody mediated down regulation of IGF-1R significantly retards growth of many tumors (22), and induces regressions when combined with cytotoxic agents (20). The prototypical anti-IGF-1R antibody, α-IR3, was shown to mediate down regulation of IGF-IR, significantly retarded growth of several cell lines in vitro, and retarded the growth of rhabdomyosarcoma xenografts (23). Several classes of small molecule inhibitors of IGF-1R have been described (24–27). In contrast to the antibodies, small molecule ATP-competitive IGF-1R inhibitors do not induce downregulation of the receptor, an effect that may be important for the activity of anti-IGF-1R directed antibody treatments (28).
Previously it was shown by the Pediatric Preclinical Testing Program (PPTP) that SCH717454, a human antibody that binds IGF-1R to block ligand binding, induced regressions in several sarcoma histotypes, notably osteosarcoma and Ewing sarcoma (29), tumors that secrete predominantly IGF-1. Of interest, another IGF-1R-targeted antibody, figitumumab (CP751871) caused rapid downregulation of IGF-1R in several sarcoma xenografts that was associated with a dramatic decline in tumor-associated VEGF (30), suggesting that a significant effect on tumor growth may be due to anti-angiogenic effects of blocking IGF-1R signaling. Further, SCH717454 was shown to reduce tumor microvessel density (21) indicating anti-angiogenic activity. The IGF/IGF-1R axis has also been implicated in regulating angiogenesis as several reports have shown decreased angiogenesis by insulin-like growth factor-binding proteins (IGFBPs) -3 and -5 that counter-balance the effects of IGFs (31). In contrast, theIGFBP-3-induced endothelial cell motility and migration may suggest a direct rolefor this binding protein in promoting angiogenesis (32, 33). Despite the promise of efficacy in tumor models, antibodies that target IGF-1R and block ligand binding have rather limited antitumor activity against pediatric solid tumor xenografts (29, 34), causing significant retardation of tumor growth, but relatively few objective regressions as single agents. Early clinical results seem to parallel the preclinical data with some examples of disease stabilization but an objective response rate ~10% (35–37) in Ewing sarcoma. The poor response rate has been related to alternative signaling through the IN-R when cells are treated with antibodies that block ligand binding to IGF-1R (38, 39).
Whereas the role of IGF-1R-targeted antibodies has focused on the role of IGFs regulating tumor cell proliferation, the role of IGF-1 receptor-mediated signaling in tumor angiogenesis has not been studied. Here we have examined the activity of SCH717454 in blocking IGF-1 and IGF-2 stimulated signaling in sarcoma cells and vascular endothelial cells in vitro, and its effects on angiogenesis in vivo. Our results point to the critical role of IGF-1R mediated signaling in VEGF-driven angiogenesis.
Medium 200, fetal bovine serum (FBS) and Alamar Blue (AB) were purchased from Invitrogen (Carlsbad, CA). Low serum growth supplement (LSGS) was obtained from Cascade Biologics Inc (Portland OR). Endothelial Tube formation assay kits were from Cell Biolabs, Inc. (San Diego, CA). Growth factor–reduced Matrigel for in vivo experiments and precoated Matrigel inserts for invasion assays were purchased from BD Biosciences (Palo Alto, CA). SCH 717454 was provided by Schering-Plough Research Institute and was diluted in 20 mM sodium acetate buffer (pH5) containing 150 mM sodium chloride. VEGF was purchased from R&D Systems Inc (Minneapolis, MN).
Human umbilical vein endothelial cells (HUVEC) were obtained from the American Type Culture Collection (ATCC). All experiments were done using endothelial cells between passages 3 and 8. HUVECs were maintained in endothelial cell growth medium M200 (Invitrogen) in high glucose supplemented medium with 15% FBS, endothelial cell growth supplements (LSGS Medium, Cascade Biologics), and 2 mM glutamine at 37°C with 5% CO2. All cells were maintained as sub confluent cultures and split 1:3, 24 hr before use.
The cell lines used in this study were developed by this laboratory, and the identity of lines used are confirmed routinely by comparison of short tandem repeat (STR) assays and compared to our original STR profiles established on early passage lines developed in this laboratory. The samples were run on a 3730 Genetic Analyzer from Applied Biosystems and analyzed with Genemapper 3.2 software also by Applied Biosystems. All lines were analyzed for 8 markers (AMEL (gender), CSF1PO, D13S317, D16S539, D5S818, D7S820, TH01, and TPOX) within the last six months. All sarcoma cell lines were cultured in RPMI 1640 supplemented with 10% FBS.
Cell lysis, protein extraction and immunoblotting were as described previously (30, 40). We used primary antibodies to glyceraldehyde-3-phosphate dehydrogenase (GAPDH), ribosomal protein S6 (rpS6), phospho-rpS6 (Ser235/236), AKT, phospho-AKT (Ser473), IGF-1R, and phospho-IGF-1R (Tyr1131), insulin receptor (IN-R), and phospho-IN-R (Tyr1146) (Cell Signaling). Non-specific IgG1 was purchased from Binding Site (Birmingham, U.K.). Immunoreactive bands were visualized by using SuperSignal Chemiluminiscence substrate (Pierce, Rockford, IL) and Biomax MR and XAR film (Eastman Kodak Co., Rochester, NY). Immunoprecipitations were performed by adding either 2μg of IGF-1R antibody (Santa Cruz biotechnology Inc., CA, USA) or 1:50 dilution of IN-R antibody (Cell Signaling) to 500μg of protein sample extracted from different cell lines EW-8, Rh18 and Rh30. Samples were rotated overnight at 4°C, after which time 20μL of protein A/G agarose beads (Santa Cruz Biotechnology) were added and the samples again rotated for 3 hr in 4°C. Beads were washed with cell lysis buffer (Cell Signaling) four times and 30μL of 3x LDS sample buffer (Invitrogen) was added to the immunoprecipitates and heated at 70°C for 10 min. Fifteen μL of total sample was resolved on a 4–12% SDS-polyacrylamide gel. Proteins were transferred to a PVDF membrane and immunodetection was performed with specific primary antibodies.
HUVECs were seeded on 6-well plates at a density of approximately 1×105 cells/well in M200 medium. Cells were treated with 5 or 10 μg/ml of SCH717454 one day after seeding. After 2 days, AB was added directly into culture media at a final concentration of 10% and the plates were incubated at 37°C. Optical density was measured spectrophotometrically at 540 and 630 nm 3–4 h after adding AB. As a negative control, AB was added to medium without cells.
HUVECs were grown in M200 containing LSGS until 40–50% confluent (41). Cells were washed with PBS, trypsinized for 5–10 min, collected with 0.2% FBS and centrifuged at 300g for 5 min. Cells were then resuspended with 0.2% FBS and counted using a Beckman Coulter Z2. A volume of 400 μl of this mix containing 5 × 105 cells was placed on to Boyden Chambers (8 μm pore) inserts with and without SCH717454 antibody in 24 well plates with 500 μl of M200. VEGF (R and D Systems) in 1% BSA was added to a final VEGF concentration of 10 ng/ml to the lower chambers as a chemo-attractant. Cells were pretreated with antibody for 30 min in suspension, then placed in the chambers and incubated at 37°C 5% CO2 for 18–24 hrs. Migration was determined using the procedures of (42). Experiments were performed in duplicate on multiple occasions as described in the figure legends.
The Matrigel invasion assays were carried out using Matrigel-precoated inserts (BD Bioscience) following the manufacturer’s instructions. Six hundred μl of M200 medium with or without VEGF(50 ng/ml, R&D Systems) was placed in the lower wells. Proliferating HUVECs (4 × 105 cells/ml) were pretreated with 5 or 10 μg/ml of SCH717454 and 100 μl of cell suspension was loaded into each of the upper wells. The chambers were incubated for 18–20 hr at 37°C. After incubation, the inserts were removed, and the non-invading cells on the upper surface were removed with a cotton swab. The cells on the lower surface of the membrane were fixed in 100% methanol for 15 minutes, air-dried, and stained with Diff-Quik stain for 2 min. Assays were performed in triplicate for each treatment group and the results were expressed as migrated cells per field for each condition (43).
The Endothelial Tube Formation Assay (CBA200, Cell Biolabs Inc., San Diego, CA, USA) was used in addition to the HUVEC proliferation assay (44). The ECM gel was thawed at 4°C. Cell culture plates (96-well) were bottom-coatedwith a thin layer of ECM gel (50 μl/well), which was leftto polymerize at 37°C for 60 min. HUVECs (2–3×104) stimulated with VEGF in 150 μl medium were added to each well on the solidified ECM gel. Culture medium was addedto each well in the presence or absence of SCH717454. The plates were incubated at 37°C for 12–18 hr and the endothelial tubes were observed using a fluorescent microscope after staining with Calcein AM. Three microscopefields were selected at random and photographed.. Each experiment was performed at leastthree times.
Alternatively, HUVECs were grown as described in the bottom well of 24mm TranswellR with 0.4μm Pore Polyester Membrane Insert, (Corning, Life Sciences), and stimulated with PBS or VEGF with or without SCH717454 (10 μg/ml). Sarcoma cells (1 × 106) were placed in the upper chamber in serum-free medium. After incubation for 24 – 48 hr endothelial tube formation was assessed as described above.
To further characterize anti-angiogenetic properties of SCH717454 in vivo, we performed murine Matrigel plug experiments (45). PBS was used as a negative control, and VEGF (100ng/mL) as a positive control. Alternatively plugs containing VEGF (100 ng/ml) and IGF-2 (50 ng/ml) were implanted. Matrigel was injected subcutaneously into CB17SC scid−/− female mice (Taconic, Germantown, NY), forming semi-solid plugs. Animals received treatment of SCH717454 (20 mg/kg) I.P. immediately after the Matrigel injection and on day 3. On day 7, plugs were excised under anesthesia, fixed in PBS-buffered 10% formalin containing 0.25% glutaraldehyde, and were processed for H & E staining. Vascular identity of the infiltrating cells was established with CD34 immunostaining. Neovascularization (“hotspots”) were chosen for analysis. Eight hotspots were identified for each Matrigel or tumor section. The ImagePro Plus analysis software (National Institutes of Health, Bethesda, MD) was used to quantify the percentage of area occupied by the vessel-like structures in each field. The mean ± SE from each group were compared. The negative control was obtained by tissue staining with secondary antibody only.
RNA was isolated using RNAeasy Minikit (Cat No 74104) (Qiagen). RT-PCR for IN-R isoforms was carried out using Super Script III First-Strand Kit (Cat No 18080-051) (Invitrogen) using oligonucleotide primers: F o r w a r d 5′-ACGACCCCAAATCACAGAACC-3′ a n d R e v e r s e 5′-TCGCCGATAACTCACTTCATACAG-′ to human IN-R. PCR amplification was carried out for 30 cycles of 30s at 96°C, 30s at 60°C and 1.5 min at 72°C using a DNA thermal cycler (Mastercycler gradient, Eppendorf). After electrophoresis, the PCR products (968bp) were purified by using QIA quick gel Extraction Kit (Cat-28706, Qiagen) and sequenced using forward primers for the determination for the full length IN-R gene and variant A that lacks exon 11.
Cells were plated at 5×105/mL in serum-free medium. After 24 hr conditioned medium was used for quantitation of IGF-1 and IGF-2 secretion by sandwich enzyme immunoassay technique (IGF-1: R&D Systems, Minneapolis, MN; IGF-2: DSL, Webster, TX). The medium was diluted (1:10) with supplied sample buffers prior to IGF-2 assay procedure. Briefly, standards, samples and controls were pipetted into the wells of a 96-well microplate precoated with a monoclonal IGF-1, or IGF-2 specific antibodies. After washing, an enzyme-linked polyclonal antibody specific for IGF-1 or IGF-2 was added to the wells. Following a wash, a substrate solution was added to the wells and the intensity of color developed in proportion to the amount of IGF-1/2 bound in the initial step was measured using a microplate reader set to 450 nm (MRX Revelation Absorbance Reader, Dynex Technologies). All measurements were done in duplicate, normalized to nuclei counts from the attached cells in each well at the end of drug exposure, and plotted as picograms of IGF-1/-2 per 106 cells.
Significance of correlations was done using GraphPad Prism software (GraphPad Software). Unpaired t tests were used for all analyses assuming Gaussian populations with a 95% confidence interval. Data are presented as mean ± SE. Differences were analyzed with the Student t test, and significance was set at P less than 0.05.
To understand whether SCH717454 equally inhibited phosphorylation of Akt induced by IGF-1 and IGF-2, sarcoma cells were serum starved overnight, then incubated with or without antibody (10 μg/ml) for 5 – 60 min, and stimulated with IGF-1 or IGF-2 for 5 min. As shown in Figure 1A, IGF-1 stimulated phosphorylation of Akt (Ser473) in control (no antibody), but stimulation was rapidly suppressed by SCH717454 in each tumor cell line, with almost complete abrogation of phospho-Akt in EW-8 cells. In contrast under the same conditions SCH717454 did not suppress IGF-2-induced phosphorylation of Akt in EW-8 cells, whereas it did slightly reduce IGF-2-induced Akt phosphorylation in Rh18, Rh30 and JR1 cells. Isotype-matched IgG1 failed to inhibit ligand stimulated Akt phosphorylation (Supplemental Figure 1A). The concentrations of IGF-1 (10 ng/ml) and IGF-2 (50 ng/ml) were used as they gave equivalent induction of Akt phosphorylation (Supplemental Figure 1B). We next examined the ability of longer exposures to antibody treatment (1–24 hr) to block phosphorylation of Akt by IGFs. Again, the antibody suppressed IGF-1-induced phosphorylation. However, even at 24 hr the effect on IGF-2 signaling was only partial in EW-8 and Rh30 cells, whereas it was more marked in Rh36 and JR1 cells (Supplemental Figure 1C).
The differential effect of SCH717454 inhibiting ligand stimulated IGF-1R signaling could be due to the antibody blocking IGF-1 binding to the receptor more effectively than IGF-2, or a consequence of IGF-2 signaling through the IN-R, thus circumventing antibody blockade of IGF-1R. To determine between these alternatives, we examined whether SCH717454 equally inhibited IGF-1 and IGF-2 stimulation of IGF-1R phosphorylation and whether maintained signaling in the presence of SCH717454 was through activation of IN-R. EW-8 cells were serum-starved for 24 hr, then incubated a further 24 hr with or without SCH717454 (10 μg/ml), then stimulated for 5 min with either IGF-1 or IGF-2, or not stimulated. Because the antibody recognizing Y1131 of IGF-1R also cross-reacts with Y1146 of the IN-R, cell lysates were prepared and immunoprecipitated using antibody specific for IGF-1R, IN-R or using a non-specific IgG1. SCH717545 equally inhibited IGF-1 and IGF-2 stimulated phosphorylation of IGF-1R in EW-8 cells (Figure 1B), and in the other sarcoma cell lines (Supplemental Figure 2), confirming that this antibody prevents binding of both IGF-1 and IGF-2 to the IGF-1 receptor. IGF-1 also stimulated phosphorylation of IN-R, probably through IGF-1R/IN-R dimers, as this was completely blocked by SCH717454. However, IGF-2 potently stimulated IN-R phosphorylation in SCH717454 treated EW-8 cells, and increased phosphorylation slightly in Rh18 and Rh30 cells (Supplemental Figure 2). Thus, SCH717454 effectively inhibited IGF-1- and IGF-2-induced phosphorylation of IGF-1R in each cell line. In contrast, IGF-2 still induced phosphorylation of IN-R in the presence of SCH717454, particularly in EW-8 cells, suggesting that signaling through IN-R circumvents inhibition of IGF-1R. The insulin receptor ‘A’ isoform, lacking exon 11, has high affinity for IGF-2 (46). RT-PCR for IN-R isoforms showed that all sarcoma cell lines and human umbilical vein endothelial cells (HUVECs) expressed predominantly the A-variant, although the full length IN-R transcript was also detected in the EW-8 cell line (Supplemental Figure 3A). The ability for SCH717454 to block IGF-2 stimulated phosphorylation appears to correlate with the relative expression of IGF-1R and IN-R in these cell lines (Supplemental Figure 3B).
To test if receptor-binding antibodies had direct effects we examined whether SCH717454 inhibited proliferation of VEGF stimulated HUVECs. Cells were stimulated with PBS (control) or VEGF in the absence or presence of SCH717454 and cell number determined by Alamar Blue staining after 2 days. As shown in Figure 2A, VEGF stimulated proliferation compared to PBS, and SCH717454 inhibited VEGF-stimulated proliferation in a concentration-dependent manner with >80% inhibition at 10 μg/ml. Further assays demonstrated similar effects of SCH717454 on migration and invasion through Matrigel coated membranes (Figure 2A). SCH717454 also dramatically inhibited VEGF-stimulated tube formation of HUVECs in vitro. Cells were grown under serum-deficient conditions and stimulated with VEGF or PBS for 30 min before exposure to antibody. SCH717454 at 5 or 10 μg/ml dramatically inhibited formation of capillary-like structures (Figure 2B), whereas control isotype-matched IgG1 had no effect (Supplemental Figure 4). These results indicate that signaling through IGF-1R is necessary for VEGF-stimulated proliferation and tube formation of these endothelial cells. As IGF-2 can stimulate Akt phosphorylation in the presence of SCH717454 (Figure 1A), we tested whether exogenous ligands could circumvent the effect of SCH717454 in blocking VEGF-induced proliferation of HUVECs. Cells were grown under serum-depleted conditions with PBS (control), or VEGF without and with SCH717454, or with VEGF, antibody and exogenous IGFs or insulin. As shown in Figure 3, VEGF stimulated tube formation, and this was completely inhibited by SCH717454. Exogenous IGF-1 did not significantly overcome the effect of SCH717454, whereas IGF-2, and to a lesser extent exogenous insulin, circumvented the block.
Many sarcoma cell lines secrete IGFs. Rhabdomyosarcomas secrete predominantly IGF-2 (47) whereas Ewing sarcoma cells secrete IGF-1 (8). From the results presented, one would anticipate that cells secreting IGF-2 could maintain angiogenesis in the presence of SCH717454, whereas those secreting predominantly IGF-1 would not. To test this hypothesis we determined the relative secretion of IGFs in panels of rhabdomyosarcoma and Ewing sarcoma cell lines, Figure 4A. Rhabdomyosarcoma lines secreted predominantly IGF-2, whereas Ewing sarcoma secreted IGF-1, with lower levels of IGF-2. To test whether IGF-2 secreting cells circumvent the antibody-mediated block on HUVEC cells, whereas IGF-1 secreting cells would not, we chose Rh30 rhabdomyosarcoma cells (that secrete intermediate levels of IGF-2) and two Ewing sarcoma cell lines (ES-1, EW-8) that secrete high levels of IGF-1. HUVECs were placed on Matrigel in the lower chamber of transwell plates and were stimulated with VEGF, or VEGF in the presence of SCH717454. Sarcoma cells (1 × 106 cells) were placed in the upper chamber, and proliferation and tube formation of HUVEC cells was determined after 20 – 30 hr. Rh30 cells that secrete IGF-2 completely reversed the antibody-induced inhibition of tube formation, Figure 4B. In contrast, EW-8 and ES-1 cells that secrete IGF1 and very low levels of IGF-2 failed to overcome the SCH717454 block on HUVECs even with 30 hr exposure, Figure 4B. These data support the contention that cells secreting IGF-2 into the tumor microenvironment may maintain angiogenesis in the presence of SCH717454 in vivo.
To directly test the anti-angiogenic activity of SCH717454 in vivo, mice were implanted subcutaneously with Matrigel plugs infused with PBS or VEGF. Mice were treated with SCH717454 (20 mg/kg I.P.) immediately after implantation of the plug and again after 3 days. Plugs were excised at day 7 and angiogenesis quantified as described in Materials and Methods. VEGF increased the number of vessels detected in Matrigel plugs by >10-fold over that in PBS infused (control) plugs. SCH717454 reduced vessel formation by ~75% (Figure 5A). Thus, in the mouse SCH717454 is a potent anti-angiogenic agent. However, adult mice have low circulating IGF-2 (48, 49), consequently, the effect of antibodies that block IGF-1R may be greater in the murine system than in humans, if circulating IGF-2 achieves adequate levels in tumor tissue. To test this indirectly, the experiments were repeated as above, with the addition of mice receiving Matrigel plugs infused with both VEGF and IGF-2. As shown in Figure 5B, inclusion of IGF-2 into the Matrigel plug completely abrogated the anti-angiogenic activity of SCH717454 in this model. In contrast, angiogenesis into Matrigel plugs where IGF-1 was included with VEGF was completely blocked by antibody treatment, Figure 5C.
Unlike humans, rodent tissue and plasma IGF-2 levels fall rapidly after birth (48, 49), and adult mice have low circulating levels of this growth factor. Potentially, this lack of circulating IGF-2 could impact by enhancing the antitumor activity of IGF-1R-targeted antibodies in tumor xenograft models or syngeneic tumor models. To test whether IGF-2 could overcome the block on IGF-1R signaling by SCH717454, we examined the phosphorylation of Akt induced by IGF-1 or IGF-2 in the presence of antibody. SCH717454 was effective in blocking IGF-1 stimulated Akt phosphorylation, whereas in many sarcoma cell lines it was less effective against IGF-2 stimulation. IGF-2 can signal both through IGF-1R, the insulin receptor A variant and through heterodimers of these receptors (50). IGF-2 potently activates the IN-R ‘A’ variant lacking exon 11 that is expressed by each of the sarcoma cell lines examined. Consequently, we examined the phosphorylation of IGF-1R and IN-R following ligand stimulation. SCH717454 equally suppressed IGF-1R phosphorylation stimulated by IGF-1 and IGF-2, demonstrating that this antibody blocks binding of both ligands. Although the results for cell lines varied, there was clear evidence that IGF-2 induced robust phosphorylation of IN-R in EW-8 Ewing sarcoma cells, and to a lesser extent in Rh18 and Rh30 rhabdomyosarcoma cells. These results suggest that in cells expressing IGF-2 and IN-R, this autocrine loop may abrogate the direct tumor cell anti-proliferative activity of IGF-1R-targeted antibodies, and is consistent with other reports (38, 39).
We were interested whether SCH717454 exerted a direct effect on angiogenesis using HUVECs in vitro. In this model system SCH717454 almost completely suppressed proliferation, tube formation and invasive properties of these endothelial cells induced by VEGF. Further, SCH717454 partially suppressed VEGF-stimulated migration. IGF-1 has been shown to promote angiogenesis, possibly through stimulation of VEGF (51), and is involved in vessel remodeling in the brain (52). Our data clearly indicate that signaling via the IGF-1R is essential for VEGF-stimulated processes that mimic angiogenesis in vitro, and suggest that IGF-1R signaling acts downstream of VEGF. Importantly, IGF-2 and insulin, but not IGF-1, could overcome SCH717454-induced blockade of IGF-1R, restoring normal VEGF-induced proliferation and tube formation.
Our data predict that tumor cells that secrete adequate IGF-2 may overcome the anti-angiogenic effects of IGF-1R-targeted antibodies, and that absence of circulating IGF-2 in the mouse may lead to over-prediction of the antitumor activity of these antibodies. To test the former prediction we examined the ability of cells expressing IGF-2 or IGF-1 to maintain VEGF-stimulated HUVEC proliferation and tube formation in the presence of SCH717454. Rh30 cells that express relatively low IGF-2 compared to other rhabdomyosarcoma cell lines, and Ewing sarcoma lines ES1 and EW-8 (high IGF-1) were selected to test the prediction. In these co-culture assays, SCH717454 completely blocked VEGF-stimulated HUVEC proliferation and tube formation. Co-culture with Rh30 cells completely reversed this inhibition, whereas co-culture with ES-1 and EW-8 cells had no effect in abrogating the effect of SCH717454. These results clearly support the idea that cells expressing even moderate levels of IGF-2 can maintain angiogenesis in the presence of SCH717454, whereas low levels of IGF-2 expressed by ES-1 and EW-8 cells appear to be inadequate to overcome IGF-1R inhibition. In contrast secretion of high levels of IGF-1 was not able to circumvent the antibody-mediated block on angiogenesis in vitro. Thus, one potential biomarker for effective angiogenesis inhibition by IGF-1R-targeted antibodies may be the relative IGF-1/IGF-2 expression in tumors.
A second prediction, that SCH717454 would have direct anti-angiogenic effects in vivo that could be overcome in the presence of IGF-2, was tested directly using a VEGF-stimulated Matrigel angiogenesis assay in mice. VEGF stimulated angiogenesis in this in vivo model was largely abrogated by treatment with SCH717454 confirming the anti-angiogenic activity of systemically administered antibody. In contrast, angiogensis was not inhibited in Matrigel plugs containing both VEGF and IGF-2. Thus, as in vitro, IGF-2, but not IGF-1, circumvented the anti-angiogenic effect of SCH717454 in vivo. This suggests that combination of IGF-1R-targeted antibodies with ligand binding antibodies with high affinity for binding IGF-2 (53), may effectively block IGF-driven tumor angiogenesis.
These studies were supported by UPHS awards CA77776 and CA23099 from the National Cancer Institute. SCH717454 was generously provided by Dr. Yan Wang (Schering-Plough Research Institute). HKB is a recipient of a Pelotonia Fellowship (The Ohio State University).
Conflict of Interest The authors know of no actual or perceived conflicts.