Molecular determinants of Bilobalide, Ginkgolide and Picrotoxin binding at 5-HT3 receptors

Andrew Thompson, Sarah Lummis

Cambridge University, Cambridge, UK

Flavonoids and the terpene lactones are regarded as the two main active components extracted from the leaves, roots and bark of the G. biloba tree. Among these are bilobalide (BB) and ginkgolides A (GA) and B (GB). Picrotoxin is extracted from the seeds of A. cocculus and consists of two components, picrotin (PTN) and picrotoxinin (PXN). All of these compounds are plant derived, share structural similarities and are non-competitive inhibitors of Gly and GABAA receptors (R) (Ivicet al., 2003). GlyR and GABAAR are members of the Cys-loop ligand-gated ion-channel family. Here we study the effects of these compounds on another member of this family, the 5-HT3R.

Human 5-HT3A (accession number: P46098) subunit cDNA was cloned into pGEMHE for Xenopus oocyte expression (Liman et al., 1992). cRNA was in vitro transcribed from linearised plasmid cDNA template using the mMessage mMachine T7 Transcription kit (Ambion, Austin, Texas, USA). Oocytes were injected with 5 ng cRNA, and currents recorded 1 - 4 days post-injection. Mutants were generated using QuikChange® site-directed mutagenesis (Stratagene, La Jolla, California, USA). Competition binding was estimated using standard procedures with transfected HEK293 cells (Thompson et al, 2007).

Wild type 5-HT3R were unaffected by GA and PTN up to a maximum concentration of 25 mM. pIC50 values for BB (3.33 ± 0.03), GB (3.14 ± 0.05) and PXN (4.97 ± 0.12) were up to 40-fold higher than previously reported at GlyR and GABAAR (Ivicet al., 2003). To identify potential binding sites in the 5-HT3R, substitutions were made at 8 channel lining residues (N-4’, E-1’, S2’, T6’, L7’ L9’, S12’, I16’, D20’). EC50 values were unaltered (< 5-fold change) by the mutations, indicating that they were well tolerated. Significant changes in the efficacy of BB, GB and PXN were seen at all but the 20’ substitution, but the effects were most dramatic at the 2’, 6’ and 12’ positions. For all compounds, inhibition was abolished by a T6’S mutation, and potency was increased by a S2’A mutation. A S12’A substitution abolished BB and GB inhibition, but only increased the PXN IC50 value 6-fold. Competition assays showed that [3H]granisetron binding was unaltered by all three compounds.

BB, GB and PXN are non-competitive inhibitors of the 5-HT3R. Consistent with other members of the Cys-loop family, binding of these compounds is altered by mutations at the 6’ residue (Hawthorne et al., 2006). S2’ is located on the next turn of the channel-lining α-helix. Reducing the size of the side chain in the S2’A mutant increases antagonist potency. The smaller Ala residue may allow the ligands to adopt a more energetically favourable position lower in the channel. S12’A abolishes BB and GB inhibition and may be a secondary binding site, or inhibit movement of the compounds as they descend into the channel.

Ivic, L. et al 2003. J Biol Chem. 278:49279-49285.

Liman, E. R. et al., (1992) Neuron, 9, 861-871.

Hawthawne, R. et al., (2006) J Neurochem. 98, 395-407.

Thompson A. J. et al., (2007) Br. J. Pharmacol., 151, 666-677.