Antiangiogenic antitumor activities of IGFBP-3 are mediated by IGF-independent suppression of Erk1/2 activation and Egr-1-mediated transcriptional events

Abstract

Most antiangiogenic therapies currently being evaluated in clinical trials target the vascular endothelial growth factor pathway; however, the tumor vasculature can acquire resistance to vascular endothelial growth factor-targeted therapy by shifting to other angiogenesis mechanisms. Insulin-like growth factor binding protein-3 (IGFBP-3) has been reported to suppress tumor growth and angiogenesis by both IGF-dependent and IGF-independent mechanisms; however, understanding of its IGF-independent mechanisms is limited. We observed that IGFBP-3 blocked tumor angiogenesis and growth in non-small cell lung cancer and head and neck squamous cell carcinoma. Conditioned media from an IGFBP-3-treated non-small cell lung cancer cell line displayed a significantly decreased capacity to induce HUVEC proliferation and aortic sprouting. In cancer cells, IGFBP-3 directly interacted with Erk1/2, leading to inactivation of Erk1/2 and Elk-1, and suppressed transcription of early growth response protein 1 and its target genes, basic fibroblast growth factor and platelet-derived growth factor. These data suggest that IGF-independent Erk1/2 inactivation and decreased IGFBP-3-induced Egr-1 expression block the autocrine and paracrine loops of angiogenic factors in vascular endothelial and cancer cells. Together, these findings provide a molecular framework of IGFBP-3's IGF-independent antiangiogenic antitumor activities. Future studies are needed for development of IGFBP-3 as a new line of antiangiogengic cancer drug.

Figure 1

IGFBP-3 suppresses tumor growth and angiogenesis in NSCLC xenografts and vascular endothelial cells. (A) H1299 xenograft tumor growth (left) 10 days after injection with IGFBP-3–expressing adenoviruses (Ad-BP-3) or empty viruses (Ad-EV). Tumor growth is expressed as the mean ± SEM. An immunohistochemical analysis of CD31 (right) was performed in xenograft tissues, and the number of CD31-immunoreactive vessels per high-power field was counted. The results represent the mean calculated from 5 mice (bars, SDs). *P < .05 compared with the control group. Representative CD31 immunostaining in H1299 xenograft tissues is included. (B) Matrigel plug assay with A549 cells. Gross observed results of blood vessels are expressed as the mean of 5 tumors ± SEM, **P < .01. (C-D) Effect of CM from indicated NSCLC cell lines that had been infected with Ad-EV or Ad-BP-3 (C-D) or treated with rBP-3 (E-F) on HUVEC proliferation (C,E), chick aortic sprouting (D), and HUVEC tube formation (F). The results represent the means (bars, SDs) of 5 identical wells. *P < .05; **P < .01; ***P < .001.

Figure 2

IGFBP-3 down-regulates bFGF transcription. (A) Western blot analysis of IGFBP-3 and bFGF expression in NSCLC and HNSCC cells 2 days after infection with Ad-EV or Ad-BP-3. (B) Reduced bFGF levels in the CM from rBP-3–treated UMSCC38 cells. N.S. indicates nonspecific bands. (C) Semiquantitative RT-PCR analysis of bFGF expression in H460 cells infected with Ad-EV or Ad-BP-3 (left) or transfected with scrambled (Con) or IGFBP-3 (BP-3) siRNA (right). (D) Luciferase assay to determine the effect of IGFBP-3 on bFGF promoter activity in H460 cells transiently transfected with bFGF-Luc in association with Ad-BP-3 or Ad-EV infection at the indicated doses (left), rBP-3 treatment (middle), or scrambled (Con) or IGFBP-3 (BP-3) siRNA cotransfection (right). The results represent the means (bars, SDs) of triplicate results. *P < .05; **P < .01; ***P < .001.

Figure 3

IGFBP-3 down-regulates bFGF expression promoter activity by regulating Egr-1 transcription. (A) Western blot (top) and semiquantitative RT-PCR (bottom) analyses of IGFBP-3's effect on Egr-1 and Sp-1 protein and mRNA expression in H460 cells that had been infected with Ad-BP-3 or Ad-EV for 2 days. (B) Luciferase assay to determine Egr-1's effect on bFGF expression. H460 cells transiently transfected with bFGF-Luc and pEgr-1, pBP-3, or both; *P < .05, **P < .01. (C-D) RT-PCR analysis of PDGFa and PDGFb mRNA expression in H460 cells (C) and Western blot analysis of PDGF protein expression in H1299 and A549 cells (D) after infection with Ad-EV or Ad-BP-3 or treatment with rBP-3. (E) Immunohistochemical analysis of Egr-1, bFGF, and PDGF expression in H1299 xenografts 10 days after injection with Ad-EV or Ad-BP-3.

Figure 4

IGFBP-3 inhibits Egr-1 expression independently IGF-1. (A-B) The wild-type 1.2-kb Egr-1 promoter reporter construct (Egr1-A-Luc) was transiently transfected, with or without pBP-3 or pBP-3-ggg, into NSCLC H460 cells. (A) Cells were stimulated by cotransfection of plasmids containing mutants of K-Ras (V12) or H-Ras (V12) or by exposure to hypoxia (1% O₂) or γ-radiation (8 Gy). (B) Cells were stimulated by IGF-1 (50 ng/mL) or FBS (10% and 30%) for 24 hours or cotransfected with plasmids expressing CA MEK. (C) R⁻ (IGF-1R null mouse fibroblasts) and R⁺ (R⁻ cells transfected with IGF-1R) cell lines were cotransfected with Egr1-A-Luc and pBP-3 or pIGFBP-3-ggg and then stimulated by IGF-1 (50 ng/mL) or FBS (10%) for 24 hours. The data are the mean ± SD from 3 independent experiments, with 4 replicates per experiment. *P < .05, **P < .01. (D) In vitro evaluation of the antiangiogenic potential of IGFBP-3. pBP-3–transfected H460 cells show less stimulatory activity for HUVEC proliferation than untransfected H460 cells in a coculture assay system. The values are the mean ± SD from 2 separate experiments, with 3 replicates per experiment. *P < .05, **P < .01, ***P < .001.

Figure 5

IGFBP-3 inhibits Egr-1 transcription by inactivating Erk-Elk1 and Elk1 binding to SRE sites in the Egr-1 promoter. (A-B) IGFBP-3's effects on Egr-1 promoter activity. H460 cells were transiently cotransfected with Egr-1 Luc constructs (A) or an Egr-1 Luc construct carrying mutations in the 3 5′-SRE sites (SRE 3, 4, and 5; B), along with an empty vector (pEV) or pBP-3. The important genetic elements in the Egr-1 regulatory region are shown, including the SRE sites, CRE sites, GC, and TATA boxes. The data are the mean ± SD from 3 independent experiments, with 4 replications per experiment. *P < .05, **P < .01, ***P < .001 compared with pEV-transfected cells. (C) IGFBP-3 reduces in vivo binding of Elk-1 to the Egr-1 promoter. H460 cells, treated with rBP-3 or bovine serum albumin or infected with Ad-BP-3 or Ad-EV, cross-linked and immunoprecipitated with antibodies specific for IGFBP-3, Elk-1, SRF, or a normal serum control antibody (immunoglobulin). The second primer denotes PCR samples using a pair of negative control primers corresponding to the exon 1 sequence of the Egr-1 gene. (D-E) Western blot analysis for the indicated proteins in H460 cells (D) and R⁻ and R⁺ cells (E) treated with the indicated concentrations of rBP-3 for 2 days and stimulated with IGF-1 for 15 minutes.

Figure 6

IGFBP-3 binds to and inactivates Erk. (A,C-D) H1299 cells were pretreated with rBP-3 for 3 hours (A) or transiently transfected with pEV or pBP-3 (C), or with pEV-Flag or pBP-3-Flag (D) for 2 days. Lysates from H1299 cells were used for immunoprecipitation with rabbit anti-Erk1/2 antibody. Erk1/2, Flag, IGFBP-3, and p38α were detected by Western blotting. Whole cell lysates (WCLs) were subjected to a Western blot assay for phospho-Erk and phospho-p38 to determine specific inhibition of Erk by IGFBP-3. EV: empty vector. pBP-3-Flag: pCMV6-IGFBP-3-Flag. (B) Ni-NTA bead-bound his-tagged rBP-3 was incubated with H1299 whole cell lysate (150, 300, or 450 μg) for 5 hours before being washed and subjected to Western blot analysis for Erk1/2, p38α, and IGFBP-3. (E) H1299 cells were incubated with rBP-3 for 3 hours and stimulated with 10% FBS; samples were removed at different time points, as shown. The cell lysates were used for immunoprecipitation with rabbit anti-Erk antibody. Western blotting was used to detect IGFBP-3 and Erk from immunoprecipitation or whole cell lysates.

Figure 7

Translocalization and colocalization of rBP-3 with Erk in H1299 cells. (A) H1299 cells on cover slips were treated with rhIGFBP-3 (10 μg/mL) and fixed at 0, 10, 30, and 60 minutes after being washed 3 times in PBS. Nuclei (blue) were stained with Hoechst 33342 (1 μg/mL). Erk1/2 (green) was stained with mouse anti-Erk antibody and secondary antibodies conjugated with Alexa Fluor 488. IGFBP-3 (red) was stained with rabbit anti-IGFBP-3 antibody and secondary antibodies conjugated with Alexa Fluor 568. For the negative control, cells treated with rhIGFBP-3 for 60 minutes were stained with secondary but not primary antibodies. Neg Con: negative control. Colocalization map: schematic plot of colocalization; white dots represent the colocalization of Erk and IGFBP-3. The empty arrowhead indicates IGFBP-3 accumulation near the cell membrane. (B) Schematic model of IGFBP-3's angiogenesis inhibition. IGFBP-3 inhibits tumor angiogenesis by IGF-dependent and -independent mechanisms. In the IGF-independent mechanism, IGFBP-3 directly binds to and inactivates Erk1/2, prevents Elk-1 activation and binding between activated Elk-1 and Egr-1 promoter, and inhibits expression of Egr-1 and its target genes, including bFGF and PDGF, resulting in suppression of angiogenesis and tumor growth.