The murine epidermis is widely used as a model to study gene function and drug action. Unfortunately, expression of many genes and treatment with some drugs results in blistering, ulceration, tumours or other severe phenotypes. Studies typically require hundreds of mice. We have developed a new approach, clonal sampling, that has the potential to substantially reduce and refine such experiments, whilst also decreasing their duration. In addition, this approach provides a tool to determine whether organotypic human epidermal cultures might be able to replace mice in some applications. To achieve these goals, we will use an inducible cre recombinase transgenic mouse line that allows a representative sample of epidermal progenitor cells to be genetically labelled in adult mice. The fate of cohorts of labelled cells will then be tracked in vivo by wholemount confocal microscopy of tail epidermis from animals culled at a series of time points. Each mouse yields data on the fate of hundreds of clones at single cell resolution; only 3 mice are needed at each time point to generate data of sufficient quality for quantitative analysis. By using statistical physics, clone fate data are used to construct a quantitative model that both describes and predicts cell behaviour. We have demonstrated that in normal epidermis, cell behaviour conforms to a remarkably simple model which successfully predicts the effects of a 6 month experiment using topical retinoid treatment. We will combine expression of a reporter gene with mutant Notch pathway genes that alter progenitor cell fate and result in marked phenotypes when expressed widely in the epidermis. By collecting clone fate data from only 9 mice per experiment we will be able to determine how each mutant alters cell fate. By targeting only 0.2% of epidermal cells we avoid the risk of causing tissue breakdown with consequent distress to the animal. Secondly, we will build on our success in predicting the effects of retinoids by validating the ability of clonal sampling to analyse the effects of presenilin inhibitors and a cyclin dependent kinase inhibitor. Again, only 18 mice are required to quantitate the effects of each drug. Finally, we have been able to apply similar biophysical modelling to human epidermal cultures. We will test the same genes and drugs used above in cell culture to determine whether cell cultures can replace animal usage in these studies.
Doupé DP, Alcolea MP, Roshan A, et al (2012). A single progenitor population switches behavior to maintain and repair esophageal epithelium. Science 337 (6098): 1091‒1093. doi:10.1126/science.1218835
Klein AM, Brash D, Jones PH, Simons BD (2010). Stochastic fate of p53-mutant epidermal progenitor cells is tilted toward proliferation by UV B during preneoplasia. Proc Natl Acad Sci USA 107: 270‒275. doi:10.1073/pnas.0909738107
Jones PH (2010). Stem cell fate in proliferating tissues: equal odds in a game of chance. Dev Cell 19: 489‒490. doi:10.1016/j.devcel.2010.10.001
Jones PH, Doupé DP, Klein AM, Simons BD (2010). The ordered architecture of murine ear epidermis is maintained by progenitor cells with random fate. Dev Cell 18 (2): 317‒323. doi:10.1016/j.devcel.2009.12.016