TRAIL (TNF-Related Apoptosis Inducing Ligand), a cytokine belonging to the TNF family, triggers cell death by apoptosis and has gained considerable interest in oncology due to its selectivity for tumor cells. Clinical trials based on recombinant TRAIL or agonistic antibodies targeting TRAIL receptors are now ongoing. TRAIL signal transduction is however relatively complex and remains poorly understood due to the large number of receptors to which TRAIL can interact with. A better understanding of its regulation should help design efficient TRAIL-based therapeutic approaches to treat cancer.
Death domain containing receptors such as TNF-R1, Fas (CD95) or TRAIL (TNF-Related Apoptosis Inducing Ligand or APO2L) agonistic receptors, can trigger cell death by apoptosis (Micheau & Tschopp, 2003;Igney and Krammer, 2002). The use of TRAIL in oncology has gained considerable interest in the past decade due its potent antitumor activity and selectivity (Ashkenazi and Herbst, 2008). Contrary to TNF or Fas ligand which are generaly toxic in vivo, TRAIL administration appears to be safe, leading thus to its evaluation in phase I and II clinical trials , associated or not with chemotherapy (Ashkenazi et al., 1999; Kelley et al., 2001; Lawrence et al., 2001) (EORTC 2004) .
Preliminary clinical trials (Phase I or II) evaluating TRAIL or TRAIL derivative efficacy, alone or combined with chemotherapy have, so far, ended most of the time in reserved, even disappointing results (Micheau et al., 2013).
TRAIL binds 2 agonistic receptors TRAIL-R1 and TRAIL-R2 (DR4 and DR5) and 3 antagonistic receptors TRAIL-R3, TRAIL-R4 and OPG (DcR1, DcR2 and OPG). Engagement of TRAIL-R1 and/or TRAIL-R2 by TRAIL or agonistic antibodies targeting TRAIL-R1 or TRAIL-R2 induce the formation of the DISC (Death Signaling Inducing complex) including the adaptor protein FADD and initiator caspases like caspase-8 (Figure A) . TRAIL triggers apoptosis in most tumor cells both in vitro and in vivo (Mitsiades et al., 2001) (Naka et al., 2002) , in a p53 independant manner (Almasan and Ashkenazi, 2003).
The impact of TRAIL antagonistic receptors on tumor cell sensitivity remains however a key question in oncology since their regulatory activity is still a matter of controversy (Mérino et al., 2007). TRAIL-R3 (Degli-Esposti et al., 1997b ; MacFarlane et al., 1997 ; Pan et al., 1997a ; Sheridan et al., 1997) is a glycosyl-phosphatidylinositol (GPI)-anchored membrane protein receptor devoid of intracellular domain. TRAIL-R4 (Degli-Esposti et al., 1997a) on the other hand is a transmemnbrane protein that harbors a truncated death domain. We have shown recently that TRAIL-R3 and TRAIL-R4 differentially inhibit TRAIL-induced cell death (Mérino et al, 2006). While TRAIL-R3 acts as a real decoy receptor competing with agonistic receptors for TRAIL binding, (Figure B) TRAIL-R4 rather plays a regulatory fonction as it appears to be co-recruited with TRAIL-R2 within the TRAIL DISC where it impairs caspase-8 activation (Figure C) .
Combining TRAIL with chemotherapy is certainly a promissing approach to treat patients suffering from cancer (Ashkenazi and Herbst, 2008). However little is known on the physiological nor on the physiopathological functions of TRAIL antagonistic receptors (Mérino et al., 2007). Increasing evidence indicate that their importance may be underestimated. They have been found to be expressed and fonctionnal in primary tumors such as AMLs or Lung carcinomas (Riccioni et al., 2005; Aydin et al., 2007). We have recently shown that some chemotherapeutic compounds such as oxaliplatin may impair chemotherapy mediated TRAIL-induced sensitization in p53 wild-type colon cancer cell lines due to the regulation of TRAIL-R3 expression (Toscano et al., 2008), while in p53 mutated cell lines, oxaliplatin is able to enhance TRAIL-induced apoptosis through JNK-mediated phosphorylation of Bcl-xL (El Fajoui et al, 2011). Moreover, c-FLIP, a caspase-8 inhibitor, is also largely expressed in primary tumor cells, and like TRAIL-R4 is able to compromise TRAIL-induced cell death (Shirley et al. 2013). Importantly, however, other conventional chemotherapeutic compounds, including 5FU, can restore senstitivity to TRAIL-induced cell death, in cells overexpressing TRAIL-R4 or c-FLIP, or cells exhibiting a deficiency in Bax, a Bcl-2 member allowing activation of the mitochondrial apoptotic machinery upon TRAIL stimulation. However when two checkpoints are acting simultaneously, chemotherapy can no longer restore TRAIL-induce cell death in tumor cells (Morizot et al. 2011). The analysis of the level of expression of these inhibitors woud thus probably be an excellent asset to monitor TRAIL therapeutic efficiency in ongoing or future clinical trials. Alternatively, designing novel selective agonists targeting TRAIL-R1 or TRAIL-R2 (Pavet et al, 2010) could be usefull to avoid resistance induced by antagonistic receptors including TRAIL-R4.
Our aim is to understand the molecular mechanisms regulating TRAIL signalling, in order to optimize TRAIL-based antitumor therapies.