Selection of in vivo retrovirally transduced hepatocytes leads to efficient and predictable mouse liver repopulation

Stable liver gene transfer is generally limited by the low efficiency of commonly used vectors. One way to circumvent this difficulty is to confer a selective advantage on transduced hepatocytes, allowing them to progressively repopulate the liver. We have used a strategy for in vivo selection and controlled amplification of a small percentage of retrovirally engineered hepatocytes. Using a bicistronic retroviral vector encoding a reporter gene and Bcl2, which confers on liver cells a survival advantage against the Fas apoptotic pathway, we demonstrate that 1.5% of initially transduced hepatocytes repopulate up to 85% of the liver after 10 injections of a Fas agonist antibody. Moreover, we show that the kinetics of liver repopulation is highly predictable. This system provides a general means of expanding at will engineered hepatocytes in vivo and offering the possibility to obtain a genetically modified liver expressing a gene of interest in a desired proportion of hepatocytes.