Using ligand-based virtual screening to allosterically stabilize the activated state of a GPCR

Abstract

G-protein coupled receptors play an essential role in many biological processes. Despite an increase in the number of solved X-ray crystal structures of G-protein coupled receptors, capturing a G-protein coupled receptor in its activated state for structural analysis has proven to be difficult. An unexplored paradigm is stabilization of one or more conformational states of a G-protein coupled receptor via binding a small molecule to the intracellular loops. A short tetrazole peptidomimetic based on the photoactivated state of rhodopsin-bound structure of Gt(alpha)(340-350) was previously designed and shown to stabilize the photoactivated state of rhodopsin, the G-protein coupled receptor involved in vision. A pharmacophore model derived from the designed tetrazole tetrapeptide was used for ligand-based virtual screening to enhance the possible discovery of novel scaffolds. Maybridge Hitfinder and National Cancer Institute diversity libraries were screened for compounds containing the pharmacophore. Forty-seven compounds resulted from virtually screening the Maybridge library, whereas no hits resulted with the National Cancer Institute library. Three of the 47 Maybridge compounds were found to stabilize the MII state. As these compounds did not inhibit binding of transducin to photoactivated state of rhodopsin, they were assumed to be allosteric ligands. These compounds are potentially useful for crystallographic studies where complexes with these compounds might capture rhodopsin in its activated conformational state.

Figure 1

a. Tetrazole peptidomimetic aligned with the first model from the TrNOE NMR structure of Gt_α(340-350) (PDB ID: 1AQG). The TrNOE structure is shown in gray, and the tetrazole peptidomimetic is shown in green. b. Pharmacophore model with spatial features and distance constraints. Rendered using Pymol (42).

Figure 2

The hit compounds from the initial screen. Compounds 3, 4, and 5 stabilized the MII state.

Figure 3

Experimental results of extra MII stabilization. The dark spectrum from the UV/Vis scan was subtracted from the spectrum taken after the rhodopsin was exposed to light. Compounds: A. 3 (100mM), B. 4 (25mM), C. 5 (50mM) are shown. The light gray trace shows the MII stabilization and MI depletion with no compound, and the dark trace shows MII stabilization and MI depletion when compound is present. In this test, there is a large difference between MII stabilization in the sample containing the compound versus the sample with no compound.

Figure 4

Dose-dependent stabilization of extra MII in the presence of specified compounds. The curves were fit using Kaleidagraph 4.0 with the maximal amount of extra MII being the saturating amount of extra MII for an individual compound and not the maximal MII from the peptide VLEDLKSCGLF. The EC50 values for compounds 1, 2, and 3 were 16.3±12.3 mM, 3.0±0.28 mM, and 11.8±6.7 mM, respectively.

Figure 5

Binding and Release Assay Results. A. Results of compounds found via ligand-based virtual screening L1. Positive Control; L2. Negative Control; L3. 1 (12.5 mM), L4. 2 (50 mM); L5. 3 (25 mM); L6. 4 (12.5 mM); L7. 5 (12.5 mM). B. Results from the high affinity peptide L1. Positive Control; L2. Negative Control; L3. 1 mM. All of the bands with compound added look similar to the positive control. If the compound inhibited transducin from binding to rhodopsin, the bands would have looked similar to the negative control. At these concentrations, the compounds do not inhibit transducin from binding to rhodopsin.