Death receptor 5 (DR5) is of particular interest in cancer drug discovery since its activation selectively induces apoptosis in cancer cells while sparing many normal cells. Recombinant Apo2L/TRAIL, the natural ligand for DR4 and DR5, has progressed to cancer clinical trials either as a single agent or in combination with other chemotherapeutic drugs. As an alternative, DR5-specific agonist antibodies may avoid the limitations of the ligand, namely binding to decoy receptors such as DcR1, DcR2, and OPG. A major hurdle for the clinical developments of DR5 agonist antibodies has been the difficulty in identifying biomarkers predictive of efficacy in patient subgroups. The analysis of data from 200 cancer cell lines indicated that no obvious genetic mutations, including those in RAS and RAF, were correlative of sensitivity to LBY135, a DR5 agonist antibody. To find clues for this problem, we have employed a strategy aimed at the identification of genes or pathways that modify DR5-mediated apoptosis after genetic perturbation. To achieve that, pooled shRNA- based DR5 sensitization/rescue screenings were conducted in seven cancer cell lines across various lineages, and Solexa-based deep sequencing of the integrated hairpins was used to deconvolute the shRNA compositions. A set of common sensitizers and rescuers were revealed, many of them were known pathway components, such as shRNAs targeting caspase 8, caspase 3, and DR5 itself being the top rescuers, indicating the robustness of the screens. In addition to those sets of common genes, interestingly, we also found that inhibition of BRAF-MEK-ERK had cell-context-dependent phenotypes on DR5 mediated apoptosis. Specifically, inhibition of BRAF-MEK-ERK pathway sensitized Miapaca2 cells to DR5 mediated apoptosis, but rescued Colo205 cells from it. The opposite phenotypes correlated well with different kinetics of caspase 8 and caspase 9 activity, but could not be explained by changes in DR5 expression level. Pathway dissection revealed that cFLIP and cIAP1, two of the endogenous apoptosis inhibitors, mediated cross-talk between DR5 mediated apoptosis and the BRAF-MEK-ERK pathway. In Miapaca2 cells, combination of U0126 (MEK inhibitor) and DR5 antibody accelerated the degradation of cFLIP and cIAP1 proteins, while in colo205, both cFLIP and cIAP1 mRNAs were upregulated. Consistently, inhibition of cFLIP or cIAP1 by shRNAs or chemical inhibition eliminated the rescue phenotype in Colo205 cells, and further sensitized Miapaca2 to DR5-mediated apoptosis. Microarray analysis of Miapaca2 and Colo205 cells treated with MEK inhibitor revealed genes that were downregulated in both cell lines, including DUSP6, ETV5 and ERG1, which are the canonical downstream targets of ERK; in addition, 180 genes were also identified as being differentially regulated by MEK inhibition between the two cell lines. A set of transcription factors that could be responsible for regulating those genes were predicted by gene set enrichment analysis (GSEA), followed by individual examination using shRNA to confirm their involvement. Among many transcription factor tested, only FOXO3 and SP1 were found to be directly involved in the cross-talk between the two pathways. Together, our screen on DR5 revealed a cell-context dependent cross-talk between DR5-mediated apoptosis and the BRAF-MEK-ERK pathway. cFLIP and cIAP1 are likely to play central roles in this cross-talk. Our results also suggest that inhibition of BRAF-MEK-ERK is a double-edged sword, which can lead to sensitizing or antagonizing effect on DR5 agonist antibody. While this raised concerns for combination therapy targeting both pathways in clinical, our data also suggested that a triple combination with clAP inhibitors could be a solution for the antagonism.

Citation Information: Clin Cancer Res 2010;16(14 Suppl):A54.