Death domain mutagenesis of KILLER/DR5 reveals residues critical for apoptotic signaling

The Fas/tumor necrosis factor (TNF)/TRAIL receptors signal death through a cytoplasmic death domain (DD) containing six alpha -helices with positively charged helix 2 interacting with negatively charged helix 3 of another DD, DD mutation occurs in head/neck and lung cancer (TRAIL receptor KILLER/DR5 ) and in Epr mice (Fas). We examined the apoptotic potential of known KILLE R DR5 lung tumor-derived mutants (n = 6) and DD mutants (n = 18) generated based on conservation with DR4, Fas, Fas-associated death domain (FADD), an d tumor necrosis factor receptor 1 (TNFR1), With the exception of Arg-330 r equired in Fas or FADD for aggregation or for TNFR1 cytotoxicity, surprisin gly major loss-of-function KILLER/DR5 alleles (W325A, L334A (Epr-like), I33 9A, and W360A) contained hydrophobic residues. Loss-of function of I339A (h ighly conserved) has not been reported in DDs. Charged residue mutagenesis revealed the following points. 1) E326A, conserved in DR4, is dispensable f or death; the homologous residue is positively charged in Fas, TNFR1, and F ADD and is critical for DD interactions, 2) K331A, D336A, E338A, K340A, K34 3A, and D351A have partial loss-of-function suggesting multiple charges sta bilize receptor-adapter interactions. Analysis of the tumor-derived KILLER/ DR5 mutants revealed the following. 1) L334F has partial loss-of-function v ersus L334A, whereas E338K has major loss-of-function versus E338A, example s where alanine and tumor-specific substitutions have divergent phenotypes. 2) Unexpectedly, S324F, E326K, K386N, and D407Y have no loss-of-function w ith tumor-specific or alanine substitutions. Loss-of-function KILLER/DR5 mu tants were deficient in recruitment of FADD and caspase 8 to TRAIL death-in ducing signaling complexes. The results reveal determinants within KILLER/D R5 for death signaling and drug design.