After blocking in 5% non-fat milk and 0.05% Tween-20 in PBS, blots were incubated with rabbit antibodies to total Egr-1 (1:1000 dilution, Santa Cruz Biotechnologies, Santa Cruz, CA, USA), Mcl-1, Bax (1:1000, Cell Signaling Technology, Danvers, MA, USA) or mouse EPZ-5676 FDA monoclonal antibodies to c-FLIP (1:500, Alexis Pharmaceuticals, Axxora UK Ltd., Nottingham, UK), Bcl-XL (clone H-5, 1:200, Santa Cruz Biotechnologies), Bcl-2 (clone 100, 1:1000, Santa Cruz Biotechnologies) or X-linked inhibitor of apoptosis protein (XIAP; 1:2000, Assay Designs, Ann Arbon, MI, USA). For detection, appropriate horseradish peroxidase-conjugated goat secondary antibodies were used (Thermo Fisher Scientific, Rockford, IL, USA). Protein bands were visualised with SuperSignal West Pico Chemiluminescent Substrate (Pierce) on X-ray film (Agfa, Morstel, Belgium).

Transfections and plasmids Dominant-negative Egr-1 construct (EBGN-EGR-1) expresses a truncated version of murine Egr-1 lacking the transactivational domain and containing only the zinc-finger DNA-binding site (amino acids 322�C533) fused to GST. The empty vector, EBGN, contains a nuclear-expressed GST (Al-Sarraj et al, 2005); both these vectors are a kind gift from Professor G Thiel (University of Saarland Medical Center, Homburg, Germany). pEBS14luc, an Egr-1 reporter construct, contains four copies of Egr-1 response element of the Egr-1 gene promoter in the pGL3-promoter vector (also a gift from Professor G Thiel, University of Saarland Medical Center) (Al-Sarraj et al, 2005).

To normalise for transfection efficiency, a constitutive Renilla luciferase expressing plasmid was used (pRL-CMV, Promega Corporation, Madison, WI, USA). For transfection, HCT15 cells (2 �� 106) were pelleted and resuspended in transfection solution V (Lonza Group Ltd., Basel, Switzerland) containing 2.5��g of plasmid unless otherwise stated. Transfection was performed by nucleofection using program T13 according to the manufacturer’s protocol (Amaxa). GFP plasmid (2.5��g) was used to determine transfection efficiency, which was 48��7%. Control cells were subjected to the same transfection condition without any plasmids. At 24h after transfection, cells were resuspended in media and seeded for Annexin V and protein assays. Similarly, stable transfection of Bcl-2 or empty vector (Neo) was carried out in HCT15 cells using the same transfection protocol (a kind gift from Dr Peter Daniel, University of Berlin, Berlin, Germany).

Pools of stable clones were selected with 1��M of G418. siRNA transfection was carried out by the same nucleofection protocol as for plasmids using 50�C75nM siRNA. The following c-FLIP sequences were targeted: c-FLIPS/L1: 5��-GGAGCAGGGACAAGTTACA-3��, c-FLIPS/L2: 5��-GCAAGGAGAAGAGTTTCTT-3��, c-FLIPS/L3: Cilengitide 5��-GAGGTAAGCTGTCTGTCGG-3�� (Nakajima et al, 2008), c-FLIPS1: 5��-CACCCTATGCCCATTGTCC-3��, cFLIPS2: 5��-CATGGAACTGCCTCTACTT-3�� (Zhang et al, 2004; Longley et al, 2006).