*3.2. Dose-Response Experiments*

Figure 2 shows the representative time courses of NADH and FAD fluorescence for an experiment with increasing doses of rosiglitazone (2, 10 and 50 μM) first in the absence and then, after a 30 min washout period, in the presence of the PPARγ antagonist GW9662 (10 μM). These

dose-response experiments in the absence and presence of the PPARγ antagonist GW9662 revealed a PPARγ-independent increase in mitochondrial oxidation by rosiglitazone as evidenced by a dose-dependent decrease in NADH autofluorescence (Figures 2 and 3A) and a dose-dependent increase in FAD autofluorescence (Figures 2 and 3B). Neither of these was attenuated or abolished by GW9662 at a dose previously used to abolish endogenous and exogenous cardioprotection in the same [11] and in other models [41,42]. In order to test for a group- rather than a single drug-effect, the same dose-response curve was repeated with 2, 10, and 50 μM pioglitazone with essentially the same results (Figure 3).

**Figure 2.** Shows representative time courses of NADH (closed blue circles) and FAD (open green diamonds) autofluorescence for an experiment with rosiglitazone (Rosi) in increasing doses from 0 to 50 μM (5 min each) first in the absence and then, after a 30 min washout period, in the presence of the PPARγ antagonist GW9662 (GW, 10 μM).

**Figure 3.** Shows a concentration-dependent decrease in autofluorescence of reduced NADH (Panel (**A**)) and a concomitant concentration-dependent increase in autofluorescence of oxidized FAD (Panel (**B**)) with administration of the thiazolidinediones (TZDs) rosiglitazone (Rosi, open dark-blue square) and pioglitazone (Pio, closed red square) compared to % baseline (bl). The PPARγ antagonist GW9662 (GW, 10 μM) did not alter the effect of rosiglitazone (open light-blue triangle) or pioglitazone (closed pink triangle) on NADH (Panel (**A**)) or FAD (Panel (**B**)) at any of the concentrations. Statistics: ANOVA followed by SNK post-hoc test with *P* < 0.05 (two-tailed) vs. \* 0 μM TZD; n = 6 per group. Please note that the slight horizontal offset of curves for any given concentration is for visual purposes only.
