Agonist ligand discrimination by the two orexin receptors depends on the expression system.

Despite the recent successes in producing orexin receptor subtype-selective antagonists, these are not commonly available, and therefore, agonist ligands are regularly used to ascribe cell and tissue responses to OX(1) or OX(2) receptors. In the current study, we have compared the native "subtype-selective" agonist, orexin-B, and its reputedly enhanced synthetic variant, Ala(11), d-Leu(15)-orexin-B, in two different recombinant cell lines. Ca2+ elevation was used as readout, and the two "selective" ligands were compared to the subtype-non-selective orexin-A, as is customary with these ligands. In transiently transfected HEK-293 cells, orexin-B showed 9-fold selectivity for the OX(2) receptor and Ala(11), d-Leu(15)-orexin-B 23-fold selectivity, when the potency ratios of ligands were compared between OX(1) and OX(2). In stable CHO-K1 cells, the corresponding values were only 2.6- and 14-fold, respectively. In addition to being low, the selectivity of the ligands was also variable, as indicated by the comparison of the two cell lines. For instance, the relative potency of Ala(11), d-Leu(15)-orexin-B at OX(2) in CHO cells was only 2.3-fold higher than its relative potency at OX(1) in HEK-293 cells; this indicates that Ala(11), d-Leu(15)-orexin-B does not show high enough selectivity for OX(2) to be useful for determination of receptor subtype expression. Comparison of the potencies of orexin-A and -B with respect to a number of published responses in OX(1)-expressing CHO cells, demonstrates that these show great variation: i.e., orexin-A is 1.6-18-fold more potent than orexin-B, depending on the response assessed. These data together suggest that orexin receptor ligands show signal trafficking, which makes agonist-based pharmacology unreliable.