Figure 1.

Troglitazone (TRO) and the PPARγ inactive Δ2-troglitazone (Δ2-TRO) reduce glioma cell viability and TGF-β₁ release. Δ2-TRO was synthesized as previously described in [11]. (a), (c) Concentration-dependent inhibition of glioma cell viability by TRO (a) or Δ2-TRO (c) in the indicated cell lines are given as mean ± SEM percentage relative to time- and solvent-matched controls. Cell viability assays (MTT assay, 96 hours) were performed as described earlier [12, 13]. Inhibitory concentrations IC₅₀ and IC₉₀, defined as concentrations shown to inhibit tumor cell viability by 50% or 90%, respectively, were determined by nonlinear regression data analysis: TRO: F98 (62 μM, 166 μM), SMA-560 (26 μM, 407 μM), U-87 MG (120 μM, 324 μM), and U-373 MG (123 μM, 331 μM); Δ2-TRO: F98 (46 μM, 95 μM), SMA-560 (23 μM, 93 μM), U-87 MG (78 μM, 132 μM), and U-373 MG (71 μM, 126 μM). Troglitazone and the PPARγ inactive Δ2-troglitazone reduce TGF-β₁ release at low micromolar doses: (b), (d) quantification of TGF-β ₁ release by F98, SMA-560, U87-MG, and U-373 MG glioma cell culture supernatants following TRO (b) or Δ2-TRO (d) treatment for 48 hours. TGF-β ₁ protein levels in glioma cell culture supernatants were determined as described in [9] using the mouse/rat/porcine/canine or the human quantikine TGF-β ₁ ELISA Kit (R&D Systems, Minneapolis, Minn, USA), respectively. Each experiment was repeated at least 3 times (n ≥ 3). Drug concentrations shown to inhibit TGF-β ₁ release by 50% or 90%, respectively, were determined by nonlinear regression data analysis: TRO: F98 (7 μM, 11 μM), SMA-560 (8 μM, 15 μM), U-87 MG (8 μM, 28 μM), and U-373 MG (10 μM, 30 μM); Δ2-TRO:F98 (3 μM, 5 μM), SMA-560 (3 μM, 8 μM), U-87 MG (4 μM, 14 μM), and U-373 MG (4 μM, 14 μM). Δ2-Troglitazone displays higher potencies than troglitazone. Using IC₉₀ concentrations of Δ2-TRO and equimolar concentrations of TRO, the PPARγ inactive Δ2-TRO displays a significantly stronger effect in both experimental paradigms (*** = P < .001, t-test) (e), (f).