How does ERK1/2-dependent phosphorylation target BimEL to the proteasome?

The death of a cell sounds like a rather catastrophic event and so is perhaps most easily associated with disease. However, a special form of cell death, called apoptosis, is a perfectly normal part of our embryological development during which excess, unwanted cells are removed in a carefully controlled fashion. For example, the cells that form the webs between our fingers when we are in the womb are removed by apoptosis. Apoptosis is also important for the removal of diseased or damaged cells such as those with potentially cancer-causing gene mutations. Indeed, defects in this process of removing damaged cells can contribute to the development of cancer and auto-immunity. The fate of a cell, whether to die or not, is determined by the fine balance of pro-death and pro-survival proteins inside the cell. If pro-survival proteins accumulate to excess then cells can accumulate, resulting in developmental abnormalities or cancer. If pro-death proteins accumulate then too many cells die and this can again cause problems during development but can also contribute to diseases such as Alzheimer's dementia. These studies tell us that the abundance of pro-death or pro-survival proteins represents a key point of control. One such pro-death protein, called Bim, appears to be relatively important because genetically engineered mice that lack Bim have a hyper-active immune system (similar to auto-immunity) and can develop some forms of leukaemia, a type of a cancer of the blood cells. In both cases this is because cells don't die when they should do. The abundance of the Bim protein is an important point of control and there are mechanisms in place to make sure that Bim does not accumulate at the wrong time or place. In particular, survival signals inside the cell modify the Bim protein by attaching specific signals or 'flags' to it, thereby directing it for destruction. In this way the Bim protein is broken down into its constituent amino acids for recycling and the cell is protected from death. In this proposal, we want to understand the mechanism by which the Bim protein is 'flagged' for destruction. A specific enzyme allows the attachment of several copies of a small molecule called ubiquitin to Bim; this ubiquitin molecule acts as the 'flag', directing the Bim protein for destruction and protecting the cell from death. We want to identify the enzyme responsible for attachment of the ubiquitin flags to Bim for two reasons. First, because it is interesting in its own right as it will teach us much about how 'normal' cell death is controlled. Second, because Bim is an important pro-death molecule, the enzyme responsible for flagging Bim for destruction may contribute to cancer or auto-immunity, and so may teach us more about these diseases.

BimEL, a pro-apoptotic, BH3-only protein, must associate with pro-survival Bcl-2 proteins to kill cells. ERK1/2-dependent phosphorylation of BimEL targets the protein for ubiquitination and proteasomal degradation and is cyto-protective. The identity of the E3 ligase responsible for BimEL ubiquitination is not known, but one could postulate that it would bind in a phospho-specific manner to phosphorylated BimEL (Model 1). Our recent data has shown that the initial effect of ERK1/2 phosphorylation is to cause BimEL to separate from pro-survival proteins; this precedes, and is independent of, BimEL proteasomal degradation. This suggests a distinct model in which BimEL phosphorylation serves only to dissociate BimEL from pro-survival proteins, with the E3 ligase then binding in a non-phosphospecific fashion to a site exposed in unbound BimEL (Model 2). These two models predict E3 ligases with different properties whose identification will require different strategies. In this study we will identify the relationship between BimEL dissociation and degradation (Model 1 versus Model 2) and identify the minimal sequences of BimEL required for ERK1/2-dependent degradation (the BimEL degron) using site-directed mutagenesis, truncation analysis and biochemical assays in cells. This information will be used to identify the BimEL E3 ligase by a combination of strategies including testing of candidate proteins, affinity purification and proteomics, facilitated by substrate trapping ubiquitin mutants and, if appropriate, yeast 2-hybrid screening, should our analysis suggest that phosphorylation is not required for degradation after dissociation. Finally, we will validate the role of the E3 ligase in regulating BimEL ubiquitination in in vitro reconstitution assays and in vivo. If the E3 ligase is specific for BimEL then knock-down of its expression should increase in BimEL and induce Bim-dependent apoptosis; this will be tested in wild type and Bim-/- fibroblasts.