TROY (TNFRSF19) promotes glioblastoma survival signaling and therapeutic resistance

Abstract

Of the features that characterize glioblastoma, arguably none is more clinically relevant than the propensity of malignant glioma cells to aggressively invade into the surrounding normal brain tissue. These invasive cells render complete resection impossible, confer significant resistance to chemo- and radiation-therapy, and virtually assure tumor recurrence. Expression of TROY (TNFRSF19), a member of the TNF receptor superfamily, inversely correlates with patient survival and stimulates glioblastoma cell migration and invasion in vitro. In this study, we report that TROY is overexpressed in glioblastoma tumor specimens and TROY mRNA expression is increased in the invasive cell population in vivo. In addition, inappropriate expression of TROY in mouse astrocytes in vivo using glial-specific gene transfer in transgenic mice induces astrocyte migration within the brain, validating the importance of the TROY signaling cascade in glioblastoma cell migration and invasion. Knockdown of TROY expression in primary glioblastoma xenografts significantly prolonged survival in vivo. Moreover, TROY expression significantly increased resistance of glioblastoma cells to both IR- and TMZ-induced apoptosis via activation of Akt and NF-κB. Inhibition of either Akt or NF-κB activity suppressed the survival benefits of TROY signaling in response to TMZ treatment. These findings position aberrant expression and/or signaling by TROY as a contributor to the dispersion of glioblastoma cells and therapeutic resistance.

Implications: Targeting of TROY may increase tumor vulnerability and improve therapeutic response in glioblastoma. Mol Cancer Res; 11(8); 865-74. ©2013 AACR.

Figure 1

TROY expression is increased in GBM. A, Immunoblot analysis of TROY in GBM tumor samples (–16) and in non-neoplastic brain tissue samples (NB, 17–19). IHC analysis of TROY expression in non-neoplasic brain (B), GBM biopsy sample from the tumor core (C) and GBM biopsy sample from the tumor edge (D). Samples are representative of 29 matched biopsy samples on a tissue microarray.

Figure 2

Aberrant TROY expression increases migration in situ. DF-1 fibroblasts producing RCAS-AP (A,D), an equal mixture of DF-1 cells producing RCAS-AP and DF-1 cells producing RCAS-TROY (B,E) or an equal mixture of DF-1 cells producing RCAS-AP and DF-1 cells producing RCAS-HGF (C,F ) were injected into the frontal lobe of newborn G-tva transgenic mice. 10 weeks after injection, mice were sacrificed and brain sections were analyzed for AP activity. Distal migration from the injection site (asterisk) is indicated by positive AP staining (black arrows).

Figure 3

TROY expression increases therapeutic resistance. A, The proliferation of T98G cells expressing TROY relative to control transduced T98G cells expressing copGFP alone was determined at the indicated time points by Alamar Blue assay. B, T98G glioma cells (Ctrl) or T98G cells expressing TROY were irradiated with 2 Gy – 6 Gy and survival was determined after 12 days by clonogenic assay (**p < 0.05). C, T98G glioma cells or T98G cells expressing TROY were treated with vehicle (DMSO) or 250 μM TMZ for 48 hours. The percentage of cellular apoptosis was determined by annexin V staining followed by flow cytometry (**, p<0.001). D, U118 glioma cells transfected with non-targeting siRNA or siRNA targeting TROY were treated with DMSO or 250 μM TMZ for 48 hours. The percent cell viability was measured by Alamar Blue assay and normalized to the control siRNA untreated with TMZ (*, p <0.01, **, p<0.001). Data represents the mean and S.D. from three independent experiments with each experiment conducted in triplicate.

Figure 4

TROY expression increases survival signaling. A, T98G glioma cells and T98G cells expressing TROY or B, SNB19 glioma cells and SNB19 cells expressing TROY were serum starved for 24 hours, the cells lysed, and the lysates immunoblotted with the indicated antibodies. C, T98G cells or T98G cells expressing TROY were treated with 20 μM LY29402 (upper panel) or 20 μM BAY117082 (lower panel) for 30 min. Whole cell lysates were immunoblotted with the indicated antibodies.

Figure 5

TROY induced resistance to TMZ is dependent upon Akt and NF-κB. A, T98G glioma cells expressing TROY or control T98G cells were serum deprived for 24 hr then treated with 250 μM TMZ for 48 hr. TMZ induced apoptosis was assayed by immunoblot analysis of cell lysates with an antibody to cleaved PARP. B, T98G glioma cells expressing TROY were left untreated, treated with 250 μM TMZ for 48 hr, or treated with 250 μM TMZ for 48 hr in combination with 20 μM LY294002 or 20 μM BAY-11-709. Apoptosis was assayed by immunoblot analysis of cleaved PARP.

Figure 6

Requirement for the MPD in TROY signaling. A, SNB19 glioma cells were transduced with an inducible homodimerization chimera encoding the WT TROY cytoplasmic domain (left) or a chimera encoding the TROY cytoplasmic domain with a deletion of the MPD (right). Serum starved cells were treated with the synthetic dimerizer AP20817 (50 nM) for varying time points, lysed, and immunoblotted with the indicated antibodies. B, T98G glioma cells expressing full length TROY (WT), TROY containing a deletion of the MPD (ΔMPD), or GFP (ctrl), were serum starved, lysed, and immunoblotted with the indicated antibodies. C, Radial cell migration assay T98G glioma cells expressing GFP (ctrl), T98G cells expressing wild type TROY (WT), or T98G expressing TROY containing a deletion of the MPD (ΔMPD). Migration was assayed over 24 hours on 10 μg/ml laminin (*, p < 0.05). D, T98G cells expressing GFP (ctrl), wild type TROY (WT), or the TROY ΔMPD variant were subjected to irradiation (2 Gy) and the surviving fraction was measured by colony formation 12 days later (**, p < 0.05). E, Cell lysates of T98G cells expressing GFP (ctrl), wild type TROY (WT), or the TROY ΔMPD variant were immunoblotted with the indicated antibodies.

Figure 7

Knockdown of TROY expression increases xenograft survival. A, Kaplan-Meier survival curves of athymic nude mice with intracranial xenografts of GBM10 transduced with a control non-targeting shRNA or a shRNA targeting TROY. Curves show a significant survival benefit for mice with xenografts with knockdown of TROY expression (p = 0.004). B, lysates of transduced GBM10 cells used for intracranial xenografts were immunoblotted 48 hours after injection with the indicated antibodies.