Critical signaling and ligand interaction mechanisms of the dopamine D1 receptor: Insight into effective pharmacotherapy

Open Access
Lee, Sangmin
Graduate Program:
Doctor of Philosophy
Document Type:
Date of Defense:
June 05, 2014
Committee Members:
  • Richard Bernard Mailman, Dissertation Advisor
  • Richard Bernard Mailman, Committee Chair
  • Kent Eugene Vrana, Committee Member
  • John Ellis, Committee Member
  • Thomas E Spratt, Committee Member
  • Dopamine D1 receptor
  • SKF-83959
  • funtional selectivity
  • ergolines
  • rotigotine
  • structure-based drug design
Dopamine D1 receptor full agonists have been efficacious in Parkinson’s disease (PD) animal models and PD patients. SKF-83959 is reported to be a functionally selective dopamine D1 receptor ligand with high bias for D1-mediated phospholipase C (PLC) versus D1-coupled adenylate cyclase (AC) signaling. The signaling bias of SKF-83959 is commonly accepted and proposed to explain D1-mediated behavioral activity in PD animal models, but there is substantial (although not all unanimous) literature that failed to account for SKF-83959-mediated PLC activation. Thus, we decided to conduct an in-depth pharmacological characterization of SKF-83959. Contrary to common assumptions, SKF-83959 is a partial agonist (not an antagonist) at AC in vitro and ex vivo. In addition, it shows partial agonistic activity for β-arrestin activation. SKF 83959 failed to show D1-mediated PLC signaling in a cellular expression system. We conclude that SKF-83959 is not a highly-biased functionally selective D1 ligand, and that its reported behavioral effects can be explained solely by its partial D1 agonism for canonical signaling pathway(s). Current dopamine D1 receptor full agonists have poor pharmacokinetic properties due to their intrinsic catechol moiety, and it is important to determine how novel non-catechol D1 ligands might be designed. To provide a scientific platform for structure-based drug design, we investigated the molecular interactions of the D1 receptor with several ergolines that have significant D1 activity and oral bioavailability, but not a catechol moiety. I focused on the conserved amino acids of the D1 receptor (T3.37, S5.42, S5.43, S5.46, F6.51, and F6.52) that are known to play a critical role in ligand interactions and/or receptor activation. Mutations to alanine (T3.37A, S5.42A, S5.43A, S5.46A, F6.51A, and F6.52A) on the D1 receptor were basically used to examine the role of the conserved amino acids in ligand interactions. A T3.37A mutation greatly decreased the D1 affinity and efficacy of the ergolines. However, a hydrogen bond-conservative T3.37S mutation markedly restored the loss of D1 affinity and efficacy suggesting the possible role of a hydrogen bond provided by T3.37. Unexpectedly, a S5.42A mutation increased the D1 affinity and efficacy for D1-mediated AC activation suggesting that this mutation may induce a favorable D1 receptor conformation for the ergolines. Although a S5.43A mutation failed to decrease the affinity of the ergolines consistently, a S5.46A mutation significantly decreased the affinity of the ergolines but to a small degree. Both the S5.43A and S5.46A mutations showed no significant effects on D1 efficacy of the ergolines. S5.42A/S5.46A and S5.43A/S5.46A double mutations elicited equal or greater effects than those of the single mutations. An F6.51A mutation dramatically decreased the D1 affinity of the ergolines, and an F6.52A mutation showed smaller, but significant decreases than the F6.51A mutation. The F6.51A mutation greatly decreased the ergoline efficacy for AC activation, but an aromatic-ring conserved F6.51W mutation markedly restored the D1 affinity and efficacy. This suggests the critical role of the hydrophobic and aromatic interactions provided by F6.51. Docking simulations illustrated that B-ring nitrogen of the ergoline agonists is located close to T3.37 and S5.46. In addition, the B-ring and the D-ring of the ergoline backbone are located close to F6.52 and F6.51, respectively. Rotigotine is another non-catechol drug that has reasonable D1 receptor efficacy. S5.42A and S5.43A mutations greatly decreased the D1 affinity and efficacy of rotigotine, whereas a S5.46A mutation failed to make changes suggesting that S5.42 and S5.43 may provide hydrogen bonds for rotigotine. An F6.51A mutation decreased the D1 affinity and efficacy of rotigotine to a far greater extent than an F6.52A mutation indicating that hydrophobic and aromatic interactions of F6.51 are particularly important. Mutagenesis results with 5-OH DPAT and 7-OH DPAT also supported the interaction between the thiophene group of rotigotine and F6.51. The simulations showed that the hydroxyl group is located close to S5.42 and S5.43 and that the thiophene group interacts closely with F6.51. In conclusion, we report that PLC activation by SKF-83959 is not a D1-mediated response, and that it is highly likely non-specific effects occurring at supra-pharmacological concentrations. Using AC activation (not PLC activation) as a functional end-point of D1 receptor signaling, we investigated the molecular interactions between the D1 receptor and non-catechol ligands (the ergolines and rotigotine). This study provides molecular mechanisms for the critical signaling and ligand interactions of the D1 receptor that may help design novel non-catechol D1 agonists.