Magnolol dimer-derived fragments as PPARγ-selective probes

Abstract

Partial agonists of the transcription factor PPARγ (peroxisome proliferator-activated receptor γ) have shown potential for the treatment of metabolic and inflammatory conditions and novel activators serve as valuable tool and lead compounds. Based on the natural product magnolol (I) and recent structural information of the ligand-target interaction we have previously developed magnolol dimer (II) which has been shown to have enhanced affinity towards PPARγ and improved selectivity over RXRα (retinoid X receptor α), PPARγ's heterodimerization partner. In this contribution we report the synthesis and evaluation of three fragments of the dimeric lead compound by structural simplifications. Sesqui magnolol A and B (III and IV) were found to exhibit comparable activities to magnolol dimer (II) and selectivity over RXRα persisted. Computational studies suggest a common pharmacophore of the distinctive biphenyl motifs. Truncated magnolol dimer (V) on the other hand does not share this feature and was found to act as an antagonist.

Fig. 1. Structure of PPARγ agonists magnolol (I) and magnolol dimer (II).

Scheme 1. Retrosynthetic analysis of target compounds III, IV and V.

Scheme 2. Synthesis of Wittig salt 9.

Scheme 3. Reported synthesis of building block 12.

Scheme 4. Wittig reaction and deprotection.

Fig. 2. Magnolol (I) and magnolol dimer-based derivatives – sesqui magnolol A (III) and B (IV), but not truncated magnolol dimer (V), induced the PPARγ-Gal4 transactivation of the Luciferase gene expression in a dose-dependent manner. Data points represent mean ± SEM. Concentrations were log-transformed and EC₅₀ and E_max values were obtained using the non-linear fitting of the dose–response data using the variable slope.

Fig. 3. Sesqui magnolol A (III), sesqui magnolol B (IV) and truncated magnolol dimer (V) do not activate the RXRα-Gal4 receptor, in contrast to magnolol (I) and a positive control 9-cis-retinoic acid. Bars show mean ± SEM. *p < 0.01 in comparison to DMSO, one-way ANOVA with Dunett's post hoc test.

Fig. 4. Two magnolol molecules (cyan, A left, B right) co-crystallized in the binding site of PPARγ and the best-ranked docking pose of IV (green) and their interactions with the binding site. Green and red arrows signify hydrogen bonds, yellow spheres mark hydrophobic contacts. Interactions with Ser342 and H₂O41 are predicted.

Fig. 5. Two magnolol molecules (cyan, A left, B right) co-crystallized in the binding site of PPARγ and the best-ranked docking pose of III (blue) and their interactions with the binding site. Interactions with Ser342 and H₂O41 are predicted.

Fig. 6. Two magnolol molecules (cyan, A left, B right) co-crystallized in the binding site of PPARγ and the best-ranked docking pose of V (grey) and their interactions with the binding site. Interactions with Ser342, Ser289, H₂O41 and H₂O35 are predicted.

Fig. 7. Truncated magnolol dimer (V) dose-dependently inhibits the pioglitazone-induced PPARγ-Gal4 mediated Luciferase gene expression. Bars show mean ± SEM. **p < 0.01, *p < 0.05 in comparison to pioglitazone, one-way ANOVA with Tukey's post hoc test.

Fig. 8. Betulinic acid (grey) co-crystallized in the binding site of PPARγ and the best-ranked docking pose of V (green). Interactions with Tyr473, Leu340 and three water molecules are predicted.