EFEMP1 suppresses malignant glioma growth and exerts its action within the tumor extracellular compartment

Abstract

Purpose: There are conflicting reports regarding the function of EFEMP1 in different cancer types. In this study, we sought to evaluate the role of EFEMP1 in malignant glioma biology.

Experimental design: Real-time qRT-PCR was used to quantify EFEMP1 expression in 95 glioblastoma multiforme (GBM). Human high-grade glioma cell lines and primary cultures were engineered to express ectopic EFEMP1, a small hairpin RNA of EFEMP1, or treated with exogenous recombinant EFEMP1 protein. Following treatment, growth was assayed both in vitro and in vivo (subcutaneous (s.c.) and intracranial (i.c.) xenograft model systems).

Results: Cox regression revealed that EFEMP1 is a favorable prognostic marker for patients with GBM. Over-expression of EFEMP1 eliminated tumor development and suppressed angiogenesis, cell proliferation, and VEGFA expression, while the converse was true with knock-down of endogenous EFEMP1 expression. The EFEMP1 suppression of tumor onset time was nearly restored by ectopic VEGFA expression; however, overall tumor growth rate remained suppressed. This suggested that inhibition of angiogenesis was only partly responsible for EFEMP1's impact on glioma development. In glioma cells that were treated by exogenous EFEMP1 protein or over-expressed endogenous EFEMP1, the EGFR level was reduced and AKT signaling activity attenuated. Mixing of EFEMP1 protein with cells prior to s.c. and i.c. implantations or injection of the protein around the established s.c. xenografts, both significantly suppressed tumorigenicity.

Conclusions: Overall, our data reveals that EEFEMP1 suppresses glioma growth in vivo, both by modulating the tumor extracellular microenvironment and by altering critical intracellular oncogenic signaling pathways.

Figure 1

EFEMP1 over-expression suppressed U251HF cell growth in vivo not in vitro. A, FLAG immunoblot of conditioned-medium (CM) proteins from cell cultures of U251HF (p) and stable transfectants of FLAG-tagged EFEMP1. B & C, real-time qRT-PCR quantification of EFEMP1 mRNA levels in U251HF transfected with FLAG-tagged or untagged EFEMP1 constructs, normalized to GAPDH, and compared with untransfected cell arbitrarily set to unity. D, MTT detection of cell in vitro growth speed. E, Kaplan-Meier survival curves for mice after i.c. implantation with U251HF and its EFEMP1 transfectants.

Figure 2

EFEMP1 suppression of tumorigenicity, in vivo cell proliferation, and VEGFA-induced angiogenesis. A, weights of s.c. xenografts dissected 24 days after implantation with 5 × 10⁶ cells of U251HF and its EFEMP1 transfectants. B, total (THb) hemoglobin concentration and tissue oxygen saturation (S_tO₂) of s.c. xenografts in live mice by modulated imaging before tumor dissections. C-D, immunofluorescence of tumor frozen-sections, with CD31 antibody detecting blood vessel density and Ki67 antibody detecting proliferation index after DAPI-counterstaining of the nuclei. E, real-time qRT-PCR quantification of gene expressions in s.c. tumors above, with mean and SD for 6-9 tumors, normalized to ACTB. F, tumor growth curve based on tumor volume measurement after s.c. implantation.

Figure 3

EFEMP1 knock-down promoted tumorigenicity and VEGFA production by two glioma primary cultures representing tumor mass cell populations. A, growth curves of s.c. xenograft from implantation of 1 × 10⁶ cells of adherent primary cultures of GBM (GBM16B) and anaplastic oligodendroglioma (OG2B), infected with lentivirus of empty vector or shRNA of EFEMP1 (shEFEMP1). B, weights of above tumors dissected 28 days after implantation. C, immunoblotting of protein extracts from above s.c. xenografts with EFEMP1 and Actin antibodies (upper panel), and quantification of VEGFA protein level in tumor by an VEGF-165 immunometric assay in duplicates (bottom panel).

Figure 4

EFEMP1 suppressed EGFR-AKT signaling. Whole cell lysate of three human glioma cell lines were immunoblotted with EGFR, pAKT and total AKT antibodies after treatments with EFEMP1 protein or EGF (B) for various time periods and in different conditions. (C), protein array analysis of cell lysates of EFEMP1-transfected and un-transfected U251HF following a 48-hour serum starvation, showing relative levels of changes on phosphorylated proteins normalized to the level total proteins.

Figure 5

Human recombinant EFEMP1 protein is able to suppress glioma xenograft growth. A, weight of s.c. xenografts dissected 28 days after implantation of U251HF cells (1 × 10⁶) mixed with EFEMP1 protein (20 ng) or vehicle prior implantation, or peri-tumoral injection with EFEMP1 (10 ng × 2 sides) or vehicle 1 week after implantation and a repeat of EFEMP1 treatment in the following week. Mean (SD) are based on 4 tumors with treatments per group. B, Kaplan-Meier survival curves for mice after i.c. implantation with U251HF cells (1 × 10⁵) mixed with 4 ng EFEMP1 protein or same volume of vehicle. C, T2-weighted images of mice from vehicle and EFEMP1-treatment groups 32 days after implantation. D, s.c. xenograft growth curve after implantation with the same amount of GBM16B expressing shEFEMP1, mixed with the same volume of EFEMP1 or vehicle prior implantation. Symbols (error bars) on the curves represent averages (SDs) of tumor volumes based on longitudinal data from 8 or 9 tumors per group that was log-transformed to stabilize variances for the repeated-measures ANOVA. P < 0.001 for the test of any separation between curves among all three groups, and P = 0.046 for the test of any difference in growth rates among all three groups.