Scaffold-based pan-agonist design for the PPARα, PPARβ and PPARγ receptors

Abstract

As important members of nuclear receptor superfamily, Peroxisome proliferator-activated receptors (PPAR) play essential roles in regulating cellular differentiation, development, metabolism, and tumorigenesis of higher organisms. The PPAR receptors have 3 identified subtypes: PPARα, PPARβ and PPARγ, all of which have been treated as attractive targets for developing drugs to treat type 2 diabetes. Due to the undesirable side-effects, many PPAR agonists including PPARα/γ and PPARβ/γ dual agonists are stopped by US FDA in the clinical trials. An alternative strategy is to design novel pan-agonist that can simultaneously activate PPARα, PPARβ and PPARγ. Under such an idea, in the current study we adopted the core hopping algorithm and glide docking procedure to generate 7 novel compounds based on a typical PPAR pan-agonist LY465608. It was observed by the docking procedures and molecular dynamics simulations that the compounds generated by the core hopping and glide docking not only possessed the similar functions as the original LY465608 compound to activate PPARα, PPARβ and PPARγ receptors, but also had more favorable conformation for binding to the PPAR receptors. The additional absorption, distribution, metabolism and excretion (ADME) predictions showed that the 7 compounds (especially Cpd#1) hold high potential to be novel lead compounds for the PPAR pan-agonist. Our findings can provide a new strategy or useful insights for designing the effective pan-agonists against the type 2 diabetes.

Figure 1. The ligand-binding domains of PPARα, PPARβ and PPARγ receptors.

The ligand-binding domains for PPARα (red), PPARβ (blue) and PPARγ (yellow) receptors share some common features: 1) composed of 12 α-helices arranged in an antiparallel helix sandwich, and a 4-stranded antiparallel β sheet; 2) Y-shaped hydrophobic ligand binding pocket with a volume of ∼1300 cubic angstroms; and 3) a C-terminal helix (Helix 12 or AF2 helix) showing widely conformational variations in different crystals and playing essential roles in activation of PPAR receptors.

Figure 2. Diagrammatically showing the core hopping procedure.

The Original LY465608 structure contains 3 major components: 3 major components: a polar acidic head (Core A), a linker group (Core B) and a hydrophobic tial (Core C). Owe to forming significant hydrogen bonds with the ligand-binding domain, the Core A would be retained during the core hopping procedure. The 1^st core hopping operation was aimed at the Core B to generate 5 scaffolds named Fragment B₁ to B₅ respectively. The 2^nd core hopping operation was aimed at the Core C to generate 4 scaffolds, named Fragment C₁ to C₄, respectively. Thus, a total of 20 combinations were obtained.

Figure 3. Diagrammatically showing the favorable conformation obtained by docking Ly465608 and Cpd#1 into (A) PPARα, (B) PPARβ and (C) PPARγ, respectively.

The binding pocket in the current study is defined by those residues with its heavy atoms within a distance limitation of 5 Å from LY465608 or Cpd#1. The AF2 function domain is shown in red helix, and the hydrophobic surfaces of the ligand-binding domain are colored in green. The dotted lines show the hydrogen bonding interactions between the receptors and ligands.

Figure 4. The RMS deviations for the backbone structures of the apo, LY465608-bound and Cpd#1-bound states of PPARα, PPARβ and PPARγ receptors.

Both the fluctuations of total RMS deviations and final RMS deviations for all the simulations systems are no more than 1 Å during our molecular dynamics simulations, indicating that all the simulation systems are in the equilibrium states.