Knocking in is the new knocking out

In the beginning there was the transgenic mouse; then came the gene knockout mouse; then the regulatable transgenic mouse; and now we have the knock-in, knockout mouse. Gerard Evan and colleagues have made a p53ER^TAM mouse in which the endogenous Trp53 gene has been replaced by one that encodes a p53 fusion protein, the function of which is solely dependent on the presence of an artificial ligand, 4-hydroxytamoxifen (4-OHT).

Why go to the extensive trouble of constructing such a mouse? Is the wealth of data from existing models not enough for addressing the complex issues of tumorigenesis? Sadly not, because as our mouse models have evolved, so has the complexity of the questions we wish to address.

Take p53 for example. The p53 pathway is disrupted frequently in human tumours, a reflection perhaps of the fact that p53 integrates a myriad of signals generated by cells in response to stress and pathological stimuli. Initially, researchers complacently thought that p53, as a transcription factor, would act through the regulation of specific, readily identifiable gene networks. However, it is now abundantly evident that p53 is regulated, and elicits many of its affects, subtly at the protein level. So if you could make a mouse in which p53 remains under the control of its endogenous gene promotor, where the protein is made at normal physiological levels in all cells, but that protein is non-functional until an exogenous source of ligand is provided, would such a mouse really provide new answers to our new questions? The p53ER^TAM knock-in mouse (Trp53^K1/K1) is viable and expresses a regulatable p53 protein and has already begun to shed new light on p53 function.

The Trp53^K1/K1 mice develop normally in the absence of 4-OHT but go on to develop lymphoid tumours with similar kinetics to standard p53 knockout animals, demonstrating that p53ER^TAM is essentially non-functional in the absence of 4-OHT. In the presence of 4-OHT, p53ER^TAM is functional, but, crucially, is not activated. On exposure of either Trp53^K1/K1 mouse embryo fibroblasts (MEFs) or adult mice to DNA damage in the presence of 4-OHT, p53 functions as expected — p53 target genes are expressed and apoptosis is induced. Moreover, on withdrawal of 4-OHT, p53 function is negated, indicating that p53 functionality in these animals requires the constant presence of 4-OHT.

With these ground rules established, the authors have started to examine the temporal regulation of p53 in response to deregulated oncogene expression and DNA damage. RAS, when expressed in cultured MEFs, induces replicative senescence due to p53 activation. In Trp53^K1/K1 MEFs that express activated HRAS, RAS deregulation persistently generates a p53 activation signal and p53 function is required to maintain replicative senescence in most cells. However, in response to DNA damage, such as γ-radiation, the signal to activate p53 is transient. Restoration of p53 function triggers apoptosis in irradiated Trp53^K1/K1 mice only 48–72 hours after radiation exposure. This indicates that such damage is efficiently repaired and questions the requirement of p53 in the resolution of radiation-induced damage.

These initial data are already beginning to probe our previous ideas about p53 biology and functionality during tumourigenesis. By raising more questions in the future, we might at last be in a position where we can effectively manipulate this fascinating and frustrating tumour-suppressor gene.