Binding of prostaglandins to human PPARgamma: tool assessment and new natural ligands.

The peroxisome proliferator-activated receptors (PPAR) form a family of nuclear receptors with a wide variety of biological roles from adipogenesis to carcinogenesis. More ligands (agonist and antagonist) are needed to explore the multiple functions of PPAR, particularly PPARgamma. In order to complete such ligand screening, a binding test should be assessed versus the classical transactivation reporter gene assay. In the present work, the full-length human PPARgamma protein as well as its ligand binding domain portion were expressed in Escherichia coli. Bacterial membrane preparations expressing those constructs were characterized using a classical binding competition assay [3H]rosiglitazone as the radioligand. When the receptor preparations were soluble, binding had to be measured with a new alternative method. The systems were assessed using a series of reference PPAR (alpha, beta and gamma) ligands. The full-length human PPARgamma fused to glutathione-S-transferase, expressed in E. coli and tested as a bacterial membrane-bound protein led to the most accurate results when compared to the literature. Furthermore, in an attempt to complete the panel of natural PPARgamma ligands, 29 commercially available prostaglandins were screened in the binding assay. Prostaglandins H(1) and H(2) were found to be modest ligands, however as potent as 15Delta(12-14 )prostaglandin J(2). These results were confirmed in the classical transactivation assay. The fact that these three prostaglandins were equally potent, suggests new pathways of PPARgamma-linked gene activation.

Structure-activity relationship studies of melanin-concentrating hormone (MCH)-related peptide ligands at SLC-1, the human MCH receptor.

Melanin-concentrating hormone (MCH) is a cyclic nonadecapeptide involved in the regulation of feeding behavior, which acts through a G protein-coupled receptor (SLC-1) inhibiting adenylcyclase activity. In this study, 57 analogues of MCH were investigated on the recently cloned human MCH receptor stably expressed in HEK293 cells, on both the inhibition of forskolin-stimulated cAMP production and guanosine-5'-O-(3-[(35)S]thiotriphosphate ([(35)S]- GTPgammaS) binding. The dodecapeptide MCH-(6-17) (MCH ring between Cys(7) and Cys(16), with a single extra amino acid at the N terminus (Arg(6)) and at the C terminus (Trp(17))) was found to be the minimal sequence required for a full and potent agonistic response on cAMP formation and [(35)S]- GTPgammaS binding. We Ala-scanned this dodecapeptide and found that only 3 of 8 amino acids of the ring, namely Met(8), Arg(11), and Tyr(13), were essential to elicit full and potent responses in both tests. Deletions inside the ring led either to inactivity or to poor antagonists with potencies in the micromolar range. Cys(7) and Cys(16) were substituted by Asp and Lys or one of their analogues, in an attempt to replace the disulfide bridge by an amide bond. However, those modifications were deleterious for agonistic activity. In [(35)S]- GTPgammaS binding, these compounds behaved as weak antagonists (K(B) 1-4 microm). Finally, substitution in MCH-(6-17) of 6 out of 12 amino acids by non-natural residues and concomitant replacement of the disulfide bond by an amide bond led to three compounds with potent antagonistic properties (K(B) = 0.1-0.2 microm). Exploitation of these structure-activity relationships should open the way to the design of short and stable MCH peptide antagonists.