Expression, binding, and signaling properties of CRF2(a) receptors endogenously expressed in human retinoblastoma Y79 cells: passage‐dependent regulation of functional receptors

Corticotropin-releasing factor (CRF) and its structurally related family members urocortin 1–3 (UCN1–3) potently modulate neuroendocrine, autonomic, and behavioral responses to stress by activating two CRF receptors: CRF1 and CRF2 (Dautzenberg and Hauger 2002; Bale and Vale 2004; Grigoriadis 2005; Hauger et al. 2006; Steckler and Dautzenberg 2006). Both receptor subtypes are highly homologous (~70%) and belong to the class B1 subfamily of G protein-coupled receptors (GPCRs) (Dautzenberg et al. 2001b; Harmar 2001). Three biologically active splice variants, CRF2(a-c), have been identified for the CRF2 receptor, whereas only one high affinity variant of the CRF1 receptor has been established to be a fully functional GPCR (Hauger et al. 2003a). Corticotropin-releasing factor type 1 and 2 receptors differ strongly in terms of their agonist and antagonist binding preferences. Binding and functional studies in cell lines recombinantly or endogenously expressing CRF1 receptors revealed a distinct ligand-selective profile: CRF, UCN1, and the non-mammalian CRF agonists fish urotensin I, and frog sauvagine bind with high affinity to the mammalian CRF1 receptor and stimulate cAMP and calcium signaling pathways (Donaldson et al. 1996; Dautzenberg et al. 1997, 2001a, 2004b). In contrast, UCN2 and UCN3 do not bind to or activate CRF1 receptors at physiologically relevant concentrations (Hsu and Hsueh 2001; Lewis et al. 2001; Reyes et al. 2001; Dautzenberg et al. 2004a,b; Grigoriadis 2005). Pharmacological characterization of the CRF2 receptor splice variants revealed no major differences between CRF2(a), CRF2(b), and CRF2(c) receptors (Donaldson et al. 1996; Kostich et al. 1998; Palchaudhuri et al. 1999; Dautzenberg et al. 2004b). However, the binding profiles of these three CRF2 receptors markedly diverge from the binding profile of the CRF1 receptor (Donaldson et al. 1996; Perrin et al. 1999; Dautzenberg et al. 2001b; Hsu and Hsueh 2001; Lewis et al. 2001; Reyes et al. 2001). Urotensin I, sauvagine, and UCN1–3 bind with up to 1000-fold higher affinities to the CRF2 receptor than species homologs of CRF (see Hauger et al. 2003a). In agreement with the binding data, a similar rank order of potency is typically observed when these five agonists are used to stimulate cAMP stimulatory G protein (Gs)-coupled cAMP signaling (Donaldson et al. 1996; Dautzenberg et al. 2001b; Hsu and Hsueh 2001; Lewis et al. 2001; Reyes et al. 2001) or phospholipase C-mediated transient mobilization of intracellular calcium stores (Dautzenberg et al. 2004a). Therefore, UCN2 and UCN3 are generally considered to represent endogenous ligands for mammalian CRF2 receptor variants, whereas UCN1 is thought to be an endogenous ligand for both CRF receptors. Pharmacological characterization of CRF1 and CRF2 receptors has mainly been completed using recombinant receptor expression systems (Perrin and Vale 2002; Grigoriadis 2005; Hauger et al. 2006). However, in a recombinant setting the imbalanced receptor-G protein stoichiometry may strongly influence the receptor signaling properties (see Kenakin 1997). Thus, confirming recombinant GPCR data in an endogenous cellular setting is of high importance. Another aspect of scientific interest is whether or not CRF1 and CRF2(a) receptors engage in crosstalk or are co-regulated in the CNS. Therefore, identification of cell lines endogenously expressing CRF2(a) receptors alone or together with CRF1 receptors is critical for gaining further insight into the regulation of CRF receptor signaling. A large number of brain-derived or neuroendocrine cell lines, including Y79 retinoblastoma, IMR-32 neuroblastoma, CATH.a cathecholaminergic, AtT-20 pituitary, PC12 pheochromocytoma, and small lung cell carcinoma NCI-H82 cells endogenously express CRF1 receptors (Vita et al. 1993; Dieterich and DeSouza 1996; Iredale et al. 1996; Hauger et al. 1997; Kiang et al. 1998; Dautzenberg and Hauger 2002; Dermitzaki et al. 2007). Despite the widespread expression of CRF2 receptors in the CNS and the periphery, few cell lines have been found to express endogenously one of the three CRF2 receptor isoforms (Kiang et al. 1998; Hsu and Hsueh 2001; Brar et al. 2004; Nemoto et al. 2005). The human pancreatic carcinoid BON cell line only expresses the CRF2 receptor but the splice variant has not been determined (von Mentzer et al. 2007). The rat aortic smooth muscle A7r5 cell line exclusively expresses the CRF2(b) receptor that couple to the Gs protein (Hsu and Hsueh 2001; Hoare et al. 2005). Although rodent CATH.a catecholaminergic cells express CRF2(a) receptors (Brar et al. 2004) and pheochromocytoma PC12 cells express CRF2(b) receptors (Dermitzaki et al. 2007), in addition to CRF1 receptors, molecular mechanisms regulating CRF receptor signaling have not been characterized in these two cell lines. However, we have extensively studied regulation of the CRF1 receptor endogenously expressed in human retinoblastoma Y79 cells and found that this cell line provides a valuable system for studying CRF receptor regulation in an endogenous setting (Hauger et al. 1997, 2003b; Dautzenberg et al. 2001a, 2002a). In the present study, we demonstrate that CRF2(a) receptors are endogenously expressed in Y79 cells, and CRF2(a) receptors can be up-regulated with increasing duration of cell culture (> 20 passages). We also established functional Gs-coupling and cAMP signaling when retinoblastoma CRF2(a) receptors are activated by their selective agonists, UCN2 and UCN3. CRF1 and CRF2 receptor signal transduction in Y79 cells could be functionally separated using selective ligands and antagonists. Finally, we provide the first evidence that CRF2(a) receptor function is rapidly regulated by a homologous desensitization mechanism.