To determine if the FGF21’s anti-diabetic effects necessitate the presence of adipose tissue, we reconstituted adipose mass by implanting WAT from WT mice into Tg mice, and assessed the efficacy of rmuFGF21 in these animals. In pilot experiments (Figure S4) we determined that small (∼100 mg) pieces of WAT successfully grafted, and we could increase adipose mass by approximately 2 g (Figure S4A). We found that although the MRI apparatus detected the WAT following transplant, it was ineffective at monitoring the WAT transplants over time (>2–3 weeks, possibly caused by delipidation of WAT or additional cellular infiltrate) despite the fact that upon necropsy the fat pads were still intact and approximately the same size as their pre-implantation size. Nonetheless we considered the transplantation methodology successful due to the fact that leptin levels increased (Figure S4B), glucose and insulin levels had normalized (Figure S4C–D), and the steatosis in Tg mice had been reversed (Figure S4E vs. S4F). We observed that every transplanted fat pad was viable and had low grade inflammatory cell infiltrate (Figure S4G). Based on these data we investigated
