Take one molecule — tumour-necrosis factor (TNF)-α — that binds two receptors. The first of these receptors (TNF-RI) contains an intracellular death domain and induces apoptosis. The second, TNF-RII, contains no such death domain, and associates with anti-apoptotic molecules such as the TNF-R-associated factors (TRAFs) and 'inhibitor of apoptosis' (IAP) proteins. But signalling by the TNF-RII pathway potentiates the pro-apoptotic effects of TNF-RI-mediated signalling. How can this be? A report by Jonathan Ashwell and colleagues in Nature now shows how components of the TNF-RII pathway might interact to promote cell death.

Previous work had shown that signalling through TNF-RII results in decreased levels of the anti-apoptotic molecule TRAF2. The authors explored the mechanism behind this decrease by comparing levels of TRAF2 in response to TNF-α signalling in Jurkat T cells (which contain TNF-RI but very little TNF-RII) with those in 4E3 cells (Jurkat T cells transfected with TNF-RII). They observed decreased levels of TRAF2 in the 4E3 cells, but not in the Jurkat T cells.

Ashwell and co-workers repeated this experiment in the presence of a proteasome inhibitor, and this time saw no decrease in TRAF2 in the 4E3 cells. As proteins are targeted for proteasome-mediated digestion by the addition of polyubiquitin chains, the authors then looked for the presence of ubiquitylated TRAF2. Using anti-TRAF2 antibodies, they were able to immunoprecipitate a large amount of polyubiquitylated material in response to TNF-α.

So, what might target TRAF2 for degradation? Enter the IAPs, which belong to TNF-RII signalling complexes and have E3 ubiquitin ligase activity. To test their involvement, beads that were coated with glutathione S-transferase (GST)–IAP fusion proteins were incubated with in vitro translated TRAF2. The TRAF2 was pulled down by GST–c-IAP1 and GST–c-IAP2, but not by the other IAPs tested. However, only c-IAP1 caused ubiquitylation of TRAF2, as tested in vitro and also by co-transfection experiments. So c-IAP1 seems to cause the proteasome-mediated degradation of TRAF2.

Finally, Ashwell and colleagues used a mutant form of c-IAP1, which has no E3 activity, to study the functional importance of the c-IAP1-mediated ubiquitylation of TRAF2. As expected, they found that TNF-α-induced degradation of TRAF2 was prevented by overexpression of the c-IAP1 mutant — an event that also substantially reduced apoptosis. These findings are therefore very satisfying — not only do they explain how TNF-RII-mediated signalling can potentiate the pro-apoptotic TNF-RI pathway, but they also identify a physiological function for c-IAP1.