In vivo bioluminescence imaging of murine xenograft cancer models with a red-shifted thermostable luciferase.

PURPOSE: Conventional in vivo bioluminescence imaging using wild-type green-emitting luciferase is limited by absorption and scattering of the bioluminescent signal through tissues. Imaging methods using a red-shifted thermostable luciferase from Photinus pyralis were optimized to improve the sensitivity and image resolution. In vivo bioluminescence imaging performance of red- and green-emitting luciferases were compared in two different xenograft mouse models for cancer.

METHODS: Human hepatoblastoma cell line (HepG2) and human acute monocytic leukemia cell line (Thp1) cells were genetically engineered using retroviral vector technology to stably express the red-shifted or the wild-type green luciferase. A xenograft model of liver cancer was established by subcutaneous injection of the HepG2-engineered cells in the flank regions of mice, and a leukemia model was generated by intravenous injection of the engineered Thp1 cells. The cancer progression was monitored with an ultrasensitive charge-coupled device camera. The relative intensities of the green- and red-emitting luciferases were measured, and the resulting spatial resolutions of the images were compared. Imaging was performed with both intact and scarified live animals to quantify the absorption effects of the skin and deep tissue.

RESULTS: The red-emitting luciferase was found to emit a bioluminescence signal with improved transmission properties compared to the green-emitting luciferase. By imaging the HepG2 models, which contained tumors just beneath the skin, before and after scarification, the percentage of light absorbed by the skin was calculated. The green bioluminescent signal was 75 +/- 8% absorbed by the skin, whereas the red signal was only 20 +/- 6% absorbed. The Thp1 model, which contains cancer cells within the bones, was likewise imaged before and after scarification to calculate the percentage of light absorbed by all tissue under the skin. This tissue was responsible for 90 +/- 5% absorption of the green signal, but only 65 +/- 6% absorption of the red signal.

CONCLUSION: Two different bioluminescent mouse cancer models demonstrate the utility of a new red-shifted thermostable luciferase, Ppy RE-TS, that improved the in vivo imaging performance when compared with wild-type P. Pyralis luciferase. While wild-type luciferase is currently a popular reporter for in vivo imaging methods, this study demonstrates the potential of red-emitting firefly luciferase mutants to enhance the performance of bioluminescence imaging experiments.