br Conflict of interest br

Conflict of interest
Introduction Endothelin receptors are members of the superfamily of G-protein-coupled receptors (GPCRs) and central regulators of vascular tension and other physiological functions in higher eukaryotes [1]. The human endothelin system comprises the highly homologous endothelin-A (ETA) and endothelin-B (ETB) receptors, as well as the three related natural 21-amino-acid peptide agonists endothelin (ET)-1, ET-2 and ET-3. The two receptors recognize the agonists ET-1 and ET-2 with similar high affinity, but they have a marked differential selectivity for ET-3 and numerous pharmaceutically relevant artificial agonists. This variation in their ligand portfolio in combination with complex 5-Formyl-UTP patterns of the receptors and ligands in various tissues is supposed to be a major determinant for their sometimes divergent effects on human physiology [2], [3]. Unbalanced ET-1 levels affect vasoconstriction and cause severe diseases such as pulmonary arterial hypertension [4]. Activating ETB as vasodilator by ETB-selective agonists [5] or the blockade of vasoconstriction caused by ETA [6] are currently discussed as future strategies for medical applications. ETB displays selectivity for the ET-1 analog ET-3 for the related sarafotoxin 6c from the venom of the snake black mamba and for artificial agonists such as the linear peptides [Ala1,3,11,15]-endothelin-1 (4Ala-ET-1) and Suc-[Glu9,Ala11,15]-endothelin-18–21 (IRL1620) (Fig. 1a). In contrast, ETA is rather selective to small chemical compounds such as bosentan and sitaxentan [8]. A multiple molecular interaction network of the ETB/ET-1 complex was recently revealed by crystallization [7], and a conformational dynamics has been observed in ligand-bound and ligand-free forms [7], [9]. Similar to other GPCRs such as opioid receptors or the neurotensin receptor, ET-1 docks into the ETB orthosteric pocket mainly formed by transmembrane domains and covered by a conserved signature β-hairpin motif in the extracellular loop 2 (ECL2) of peptide binding GPCRs [10]. However, the molecular mechanisms of ligand discrimination between ETB and ETA are unknown. Despite diverse physiological functions and ligand portfolios, GPCRs have common mechanisms to trigger downstream signaling by G-protein activation [11]. These mechanisms are determined by the more conserved membrane-integrated seven-transmembrane helix (TM1–7) bundle and the intracellular domains (Fig. 1b) [12], [13]. The initial ligand recognition is modulated by the extracellular domains, and understanding the molecular details of the ETB ligand selectivity could provide valuable inputs for directed drug design [14]. The ETA and ETB receptors share an overall 57% sequence identity and variations in between the N-terminal domain and the extracellular loops ECL1–3 are most likely relevant for ligand discrimination (Fig. S1). We have constructed chimeric ETB receptors by swapping structural units of the N-terminal domain and of the ECL1–3 loops of ETA with corresponding units of ETB (Fig. S2). The chimeras were analyzed for their affinity to the common non-selective agonist ET-1 as well as to a selection of ETB-selective agonists. All GPCR derivatives were cell-free synthesized in the presence of preformed empty nanodiscs (NDs) assembled with the lipid 1,2-dielaidoyl-sn-glycero-3-phospho-(1′-rac-glycerol) (DEPG) that has previously been shown to provide a suitable environment for the membrane integration and folding of ETB [15], [16]. The strategy of co-translational insertion of the GPCRs into defined membranes ensured that GPCR folding and ligand interaction is not affected by detergent contacts, as all studies could be performed in native-like membrane environments [17]. Previous modeling studies as well as more recent crystal structures of ETB proposed several models of ETB/ligand interaction [7], [9], [18]. Our studies indicate the second β-strand and a short linker region in ECL2 as well as the N-terminus of ETB (Fig. S1) as essential selectivity filters for several agonists such as IRL1620, sarafotoxin 6c and 4Ala-ET-1. Interactions of the ligands with the N-terminal domain of ETB further support selectivity at different extents and are most important for IRL1620 and 4Ala-ET-1. In contrast, the tight binding of ET-3 requires synergic interactions with both ECL2-B2 and the N-terminal domain, thus indicating a more complex recognition and interaction network for this ETB-selective agonist. We further present evidence that the overall topology of the common agonist ET-1 and constraints within its N-terminal domain are further discrimination parameters for ETB and ETA. Our data agree with other selectivity models of peptide binding GPCRs, including the human δ-opioid receptor [19], the human orexin receptors [20] and the neuropeptide Y receptors [21], [22]. Several engineered ETB derivatives were generated that have lost the binding of ETB specific agonists, while high affinity to the agonist ET-1 common to ETB and ETA remained conserved. The documented strategy could therefore open alternative options to address ligand selectivity of GPCRs and to analyze GPCR/drug interactions on the molecular level.