Evaluating reporter genes of different luciferases for optimized in vivo bioluminescence imaging of transplanted neural stem cells in the brain.

Bioluminescence imaging (BLI) has become the method of choice for optical tracking of cells in small laboratory animals. However, the use of luciferases from different species, depending on different substrates and emitting at distinct wavelengths, has not been optimized for sensitive neuroimaging. In order to identify the most suitable luciferase, this quantitative study compared the luciferases Luc2, CBG99, PpyRE9 and hRluc. Human embryonic kidney (HEK-293) cells and mouse neural stem cells were transduced by lentiviral vector-mediated transfer to express one of the four luciferases, together with copGFP. A T2A peptide linker promoted stoichiometric expression between both imaging reporters and the comparison of cell populations upon flow cytometry. Cell dilution series were used to determine highest BLI sensitivity in vitro for Luc2. However, Coelenterazine h-dependent hRluc signals clearly exceeded d-luciferin-dependent BLI in vitro. For the quantitative in vivo analysis, cells were transplanted into mouse brain and BLI was performed including the recording of emission kinetics and spectral characteristics. Differences in light kinetics were observed for d-luciferin vs Coelenterazine h. The emission spectra of Luc2 and PpyRE9 remained almost unchanged, while the emission spectrum of CBG99 became biphasic. Most importantly, photon emission decreased in the order of Luc2, CBG99, PpyRE9 to hRluc. The feasibility of combining different luciferases for dual color and dual substrate neuroimaging was tested and discussed. This investigation provides the first complete quantitative comparison of different luciferases expressed by neural stem cells. It results in a clear recommendation of Luc2 as the best luciferase selection for in vivo neuroimaging.