When stress signals activate p53, the cell has a choice between pausing the cell cycle and repairing any DNA damage, or — if the damage is too great — sentencing the cell to death by apoptosis. Yardena Samuels-Lev and colleagues now describe a family of proteins, the ASPPs (for apoptosis-stimulating protein of p53), that might influence this decision.

The ASPP family was found in a database search for homologues of a previously identified p53 binding protein, 53BP2. The authors discovered two sequences: one, which they have named ASPP2, is a long isoform of 53BP2; the other, ASPP1, is a novel protein. The residues known to interact with p53 from the previously solved crystal structure of the p53–53BP2 complex are conserved in both ASPPs, and immunoprecipitations confirmed that both interact with p53.

Does binding to ASPPs alter p53 function? Cotransfection of ASPP genes and TP53 into a TP53−/− cell line caused a marked increase in the proportion of cells that underwent apoptosis compared with cells transfected with TP53 alone. 53BP2 also interacts with the anti-apoptotic protein BCL2, so could the pro-apoptotic effect of ASPP be due to an inhibitory effect on BCL2? Apparently not, because ASPP had no effect on apoptosis induced by E2F1 or BAX expression. Furthermore, antisense oligonucleotides that block the production of ASPP1 and ASPP2 had no effect on BAX-mediated apoptosis.

So, how do ASPPs aid in this cellular euthanasia? Chromatin immunoprecipitation revealed that ASPP2 expression caused an eightfold increase in the amount of p53 bound to the BAX promoter, but it had no effect on the amount of p53 bound to the WAF1 promoter — the gene through which p53 induces cell-cycle arrest. ASPPs had even more dramatic effects on p53-mediated transactivation of the pro-apoptotic genes BAX and PIG3 , increasing their expression 20–30-fold, but had no effect on transactivation of p53 target genes with other functions, such as MDM2 , cyclin G and WAF1.

The pro-apoptotic effects of ASPPs required their amino termini, indicating that 53BP2 — the N-terminally truncated form of ASPP2 — might be a dominant-negative mutant of ASPP2. In support of this, expression of 53BP2 reduced the ability of p53 to cause apoptosis. This raises the possibility that binding of ASPP's N terminus to another protein is involved in boosting p53's ability to transactivate pro-apoptotic target genes.

The discovery of ASPPs might also clear up a mystery concerning why two oncogenic p53 mutations, 181L and 181C, can transactivate WAF1 but are very inefficient at stimulating apoptosis. These mutants coimmunoprecipitated poorly with ASPP1, although they seemed to interact well with ASPP2.

Might ASPP inactivation offer a selective advantage to tumours that have wild-type TP53? In a panel of 58 breast tumours with matched normal tissue, most of the TP53 wild-type tumours had reduced expression of either ASPP1 or (less commonly) ASPP2, whereas most of those with mutant TP53 had normal ASPP expression levels.

Cleopatra had an accomplice — the asp — to help her die, and now, it seems, p53 also uses an ASPP to nudge cells towards death. The discovery of this family explains why the promoters of many of p53's target genes, particularly the pro-apoptotic ones, have weak p53 binding sites. By modulating p53's ability to bind at these promoters, the cell can control the decision to live or die. The precise mechanism by which ASPPs boost p53's killing power, and the tantalizing possibility of sensitizing tumours to chemotherapy or radiotherapy by reactivating ASPP expression, are exciting avenues for future research.