In vitro sensitivity testing of minimally passaged and uncultured gliomas with TRAIL and/or chemotherapy drugs

Abstract

TRAIL/Apo-2L has shown promise as an anti-glioma drug, based on investigations of TRAIL sensitivity in established glioma cell lines, but it is not known how accurately TRAIL signalling pathways of glioma cells in vivo are reproduced in these cell lines in vitro. To replicate as closely as possible the in vivo behaviour of malignant glioma cells, 17 early passage glioma cell lines and 5 freshly resected gliomas were exposed to TRAIL-based agents and/or chemotherapeutic drugs. Normal human hepatocytes and astrocytes and established glioma cell lines were also tested. Cross-linked TRAIL, but not soluble TRAIL, killed both normal cell types and cells from three tumours. Cells from only one glioma were killed by soluble TRAIL, although only inefficiently. High concentrations of cisplatin were lethal to glioma cells, hepatocytes and astrocytes. Isolated combinations of TRAIL and chemotherapy drugs were more toxic to particular gliomas than normal cells, but no combination was generally selective for glioma cells. This study highlights the widespread resistance of glioma cells to TRAIL-based agents, but suggests that a minority of high-grade glioma patients may benefit from particular combinations of TRAIL and chemotherapy drugs. In vitro sensitivity assays may help identify effective drug combinations for individual glioma patients.

Figure 1

In vitro responses of glioma cells, astrocytes and hepatocytes to TRAIL. (A) Cells from the indicated early passage or established glioma cell lines, ex vivo gliomas, normal astrocytes or normal hepatocytes were incubated in vitro with TRAIL or with anti-DR5 antibody. Black triangles indicate high and low drug concentrations, when applicable (see Table 2 and the Materials and Methods section). Survival was assayed using the CellTiter Glo kit and depicted using ‘bubble’ graphs. The areas of the circles denote net survival following each treatment, relative to untreated cells (set at 100%, left column). Small circles indicate efficient killing, large circles reflect survival and/or proliferation, as illustrated in the graphical legend. Glioma assays were performed in duplicate (data are represented by circles). Four replicates were performed for hepatocytes and eight replicates for astrocytes. For astrocyte and hepatocyte data, grey circles depicting average survival are overlaid upon black circles indicating average survival plus standard error. (B) DEVDase activity in D2247 and D2302 cells was monitored 6 h following treatment with the specified TRAIL formulations, anti-DR5 antibody or normal media.

Figure 2

In vitro sensitivity of three early passage glioma cell lines and one ex vivo tumour to TRAIL in combination with chemotherapy drugs. Cells from the early passage lines D2235 (A), D2247 (B) and D2248 (C) and the freshly resected glioma RMH020 (D) were incubated in vitro with the stated formulations of TRAIL or anti-DR5 antibody, alone or together with the listed chemotherapy drugs as described in the Materials and Methods section and Table 2. The resulting survival was assayed and graphed as described in the legend to Figure 1.

Figure 3

In vitro sensitivity of glioma cells to TRAIL in combination with chemotherapy drugs. Cells from 17 early passage glioma cell lines (A) and 5 uncultured gliomas (B) were incubated in vitro with the stated formulations of TRAIL or anti-DR5 antibody, alone or together with the listed chemotherapy drugs. The resulting survival was assayed and graphed as described in the legend to Figure 1. Grey circles depicting average survival are overlaid upon black circles indicating average survival plus standard error.

Figure 4

Propidium iodide uptake assay of glioma cell sensitivity to combination treatments. Cells from the indicated early passage glioma cell lines or ex vivo gliomas were incubated in vitro with soluble TRAIL at 100 ng ml⁻¹ (+) or 1000 ng ml⁻¹ (++) alone or together with temozolomide (13.7 μg ml⁻¹) or cisplatin (54 μg ml⁻¹) for 48 h. Flow cytometry measurement of propidium iodide exclusion was used to quantitate the proportion of surviving cells. The areas of the circles denote survival following each treatment. Small circles indicate efficient killing, large circles reflect survival.

Figure 5

In vitro sensitivity of astrocytes and hepatocytes to TRAIL in combination with chemotherapy drugs. Normal human hepatocytes (A) and astrocytes (B) were incubated in vitro with the stated formulations of TRAIL or anti-DR5 antibody, alone or together with the listed chemotherapy drugs. The resulting survival was assayed and graphed as described in the legend to Figure 1. (A) Quadruplicate assays were performed to examine hepatocyte survival following incubation with each TRAIL formulation alone. Grey circles depicting average survival are overlaid upon black circles indicating average survival plus standard error. Responses to combination treatment were assayed in duplicate; circles depict each result. (B) Eight replicates were performed to investigate astrocyte survival following incubation with each TRAIL formulation alone. Combination treatments were tested in quadruplicate. Grey circles depicting average survival are overlaid upon black circles indicating average survival plus standard error.

Figure 6

Treatments selectively toxic in vitro to glioma cells relative to normal cells. Coloured circles indicate gliomas killed in vitro by each treatment at least 3 times (A, C) or 10 times (B, D) more efficiently than the most sensitive astrocyte replicate (A, B) or hepatocyte replicate (C, D).

Figure 7

Immunoblot analyses of candidate TRAIL signalling regulators. Immunoblotting was performed on lysates from the indicated glioma early passage cell lines or 293T cells transiently transfected with expression plasmids encoding the various apoptotic pathway components. (A–D) Anti-DR5 (A), anti-caspase-8 (B), anti-FADD (C) and anti-XIAP (D) signals were quantitated using a chemidoc instrument and plotted relative to the 293T transfectant lysates (set at 100%). A nonspecific band detected by the anti-DR5 antibody is indicated by an asterisk. Long exposures using autoradiograph film revealed some signals too weak to be detected by chemidoc (+). Illustrative immunoblots are inset within each graph. (E) Autoradiography was used to assay cFLIP_L and p53 expressions in early passage lines and 293T cells transfected with the cFLIP_L expression plasmid (cFLIP_L, GAPDH immunoblots) or empty vector (p53 immunoblot). Irrelevant lanes separating the 293T transfectant signals from those of the early passage lines have been removed.