Imaging gene expression in human mesenchymal stem cells: From small to large animals

Purpose: To evaluate the feasibility of reporter gene imaging in implanted human mesenchymal stem cells (MSCs) in porcine myocardium by using clinical positron emission tomography (PET)-computed tomography (CT) scanning. Materials and Methods: Animal protocols were approved by the Institutional Administrative Panel on Laboratory Animal Care. Transduction of human MSCs by using different doses of adenovirus that contained a cytomegalovirus (CMV) promoter driving the mutant herpes simplex virus type 1 thymidine kinase reporter gene (Ad-CMV-HSV1-sr39tk) was characterized in a cell culture. A total of 2.25 × 10⁶ transduced (n = 5) and control nontransduced (n = 5) human MSCs were injected into the myocardium of 10 rats, and reporter gene expression in human MSCs was visualized with micro-PET by using the radiotracer 9-(4-[fluorine 18]-fluoro-3-hydroxymethylbutyl)-guanine (FHBG). Different numbers of transduced human MSCs suspended in either phosphate-buffered saline (PBS) (n = 4) or matrigel (n = 5) were injected into the myocardium of nine swine, and gene expression was visualized with a clinical PET-CT. For analysis of cell culture experiments, linear regression analyses combined with a t test were performed. To test differences in radiotracer uptake between injected and remote myocardium in both rats and swine, one-sided paired Wilcoxon tests were performed. In swine experiments, a linear regression of radiotracer uptake ratio on the number of injected transduced human MSCs was performed. Results: In cell culture, there was a viral dose-dependent increase of gene expression and FHBG accumulation in human MSCs. Human MSC viability was 96.7% (multiplicity of infection, 250). Cardiac FHBG uptake in rats was significantly elevated (P < .0001) after human MSC injection (0.0054% injected dose [ID]/g ± 0.0007 [standard deviation]) compared with that in the remote myocardium (0.0003% ID/g ± 0.0001). In swine, myocardial radiotracer uptake was not elevated after injection of up to 100 × 10⁶ human MSCs (PBS group). In the matrigel group, signal-to-background ratio increased to 1.87 after injection of 100 × 10⁶ human MSCs and positively correlated (R² = 0.97, P < .001) with the number of administered human MSCs. Conclusion: Reporter gene imaging in human MSCs can be translated to large animals. The study highlights the importance of co-administering a "scaffold" for increasing intramyocardial retention of human MSCs.