Human Urotensin-II (hU-II) is a cyclic 11-amino acid peptide that plays a role in cardiovascular homeostasis. Its receptor is a member of the class A of G-protein-coupled receptors, called GPR14. In recent years, several nonpeptide ligands have been reported in the literature. Most were identified by high-throughput screening and optimized by medicinal chemistry methods. Other nonpeptide ligands were discovered starting from the 3D structure of hU-II or other ligands. They were identified by a virtual screening approach based on a 3D pharmacophore or by structural similarity with others cyclic peptides. In this review, nonpeptide agonists and antagonists are presented in relation to structure–activity relationships.
Urotensin-II (U-II) is a cyclic peptide originally isolated from the urophysis, the hormone storage and secretion organ of the caudal neurosecretory system, of the goby fish (Gillichthys mirabilis) [69]. Isoforms of this peptide were subsequently found in other species, including the mouse, rat, frog, monkey, pig, and human [20], [21], [22], [40], [41], [65]. U-II was recently identified as a ligand for the class A G-protein-coupled receptor (GPCR), GPR14, through screening of a small number of known and suspected GPCR ligands [4], [58]. It is now referred to as the UT receptor and it shares some similarities with the somatostatin receptor/subtype 4 as well as some opioid receptors [61]. The UT receptor is expressed in the motor neurons of the spinal cord and mainly in cardiovascular tissue, where it plays a role in cardiovascular homeostasis [4], [60]. Very recently, high-affinity UT receptors were identified in rat cortical astrocytes [16].
Human urotensin-II (hU-II) is a cyclic peptide composed of 11 amino acids (Glu-Thr-Pro-Asp-Cys-Phe-Trp-Lys-Tyr-Cys-Val) [22] generated by proteolytic cleavage of a precursor prohormone. The cyclic hexapeptide located in the C-terminal region is fully conserved among species and is required for biological activity. The sequence Trp-Lys-Tyr and the cyclic structure both appear to be essential for the molecule's binding affinity [44], [45], [47], [50], [67]. In contrast, the N-terminal region is highly variable in sequence and of minor functional importance. In fact, a recent report showed that an octapeptide analogue of U-II named UII-related peptide (URP), Ala-Cys-Phe-Trp-Lys-Tyr-Cys-Val being the minimal-active fragment, has high affinity for the UT receptor [17].
Human U-II appears to be the most potent vasoconstrictor known [37], but biological data reveal variability in its action, including tissue- and species-dependent effects [38]. However, despite the heterogeneity of vascular activities, a recent study showed that deletion of the UT receptor gene in knockout mice led to a selective loss of U-II contractile activity in isolated aorta [8], thus confirming the key role of the UT/U-II complex in vascular homeostasis. Moreover, new data suggest that increased levels of hU-II may play a crucial role in the development of carotid atherosclerosis in hypertensive patients [75]. Several groups have also implicated U-II in other dysfunctions such as renal disease [63], [79] and heart failure [39], [71], and these dysfunctions in rats can be attenuated by a specific UT receptor antagonist [10], [11], [73].
Identification of nonpeptide agonists and antagonists of hU-II is of great importance for the development of novel therapeutic strategies for cardiovascular pathologies. To gain a better understanding of the interactions between ligand and receptor, several pharmaceutical companies and research groups have tried to develop highly potent ligands. Historically, different approaches exist for the discovery of new nonpeptide ligands of GPCR receptors. The main approach consists of high-throughput screening (HTS) of chemical libraries, which typically yields several candidates or “hits”, which then need to be optimized by considering structure–activity relationships (SARs). Medicinal chemistry methods are used to modify the hydrophobic, steric, and electronic properties of the ligands to optimize their affinity and selectivity. The other approaches begin with knowledge of the 3D structure of the natural ligand and/or others ligands of the receptor. If the main sequence of the peptide has a precise secondary structure, historic studies of nonpeptide analogs which mimic that structure could be the key to defining new ligands [80]. Data for other peptides with the same secondary structure can also be a powerful tool to find new structures. New approaches such as virtual screening based on 3D pharmacophores defined from the key residues of the peptide or from the structure of ligands are also very promising. The definition of a precise GPCR structure is an important challenge presently. Most of the current models are based on the 3D structure of the bovine rhodopsin receptor [68]. The docking approach using this model is still very ambiguous but several groups have demonstrated its application to drug design [6], [42], [43], [49], [81].
As a general matter of drug design of nonpeptide ligands, peptides are usually associated with precise, folded conformations of the backbone which are difficult to imitate by cyclic or constrained systems in organic chemistry. The conformation, size (compared to nonpeptide ligands), and physicochemical properties of peptides generate specific and multiple interactions with the receptor, involving both lateral chains and the backbone. So, it is a tremendous challenge to design nonpeptide ligands and this is even true for the design of agonists for which blind HTS gives better results. Moreover, nonpeptide ligands tend to bind differently and sometimes at a different site compared to the native ligand. However, the situation is different for hU-II, which is associated with a turn structure. Like some others ligands with this secondary structure, the definition and the 3D conformations of the main residues have given some important clues for the design of nonpeptide structures [80]. Indeed, actual data demonstrated that hU-II and its nonpeptide ligands share common features. Intermolecular interactions involving these few key points explain, in most cases, the first micromolar affinities recorded for the hits. Unfortunately, structural clues concerning their optimization are largely unknown. Molecular modeling of GPCRs should give some insight into this problem. However, in the literature, interest in the GPCR model for optimization of the hits, and particularly for nonpeptide ligands, has not materialized (see Blakeney et al. [9] for a review on nonpeptidic ligands).
Check access to the full text by signing in through your organization.
The first nonpeptide agonist 1 (see Fig. 1) named AC-7954 with a pEC 50 of 6.5, was identified by Croston et al. [23] in a multiplexed R-SAT assay [12], [13], [14], [74] to screen a library of 180,000 small diverse organic molecules. Testing of the (+)-1 and (−)-1 enantiopure revealed that the UT receptor activity and the high selectivity reside primarily in the (+)-1 (pEC 50 of 6.6). Following classical optimization with modulation of the hydrophobic properties of the initial ligand, Lehmann et
NMR studies of hU-II were carried out in different solvents and showed several discrepancies related to secondary structure and the positions of the three key residues (Trp, Lys, and Tyr). Flohr et al. [44] determined the first 3D structure in aqueous solution by NMR spectroscopy and molecular dynamics. A β-turn, associated with the cyclic part, was observed with a distance between Trp and Tyr of 12.2 Å. Carotenuto et al. [15] determined the 3D structure (NMR) in an SDS-micelles environment,
Identifying GPCR nonpeptide ligands by receptor structure-based in silico screening is more challenging, particularly for agonists, because GPCRs are expected to undergo significant conformational changes upon activation [51]. The high-resolution crystal structure of bovine rhodopsin [68] has paved the way for structure-based design of GPCR ligands, and virtual screening of chemical libraries may be applied to GPCR homology models based on the rhodopsin template. Five recent reports have
Nonpeptide agonists and antagonists of the UT receptor have recently been identified using a suite of biochemical and computational methods. These complementary approaches hold great interest and promise for the development of novel therapeutics for the treatment of cardiovascular pathologies. For this drug design, the main approaches consist of HTS studies followed by medicinal chemistry methods for optimization of the hits. SAR studies of each drug series have demonstrated clear relationships
We thank the CRIHAN (Centre de ressources informatiques de Haute Normandie) and the European Community (FEDER) for the molecular modeling software. This work was supported by a grant from the French Ministry of Research and Technology. We sincerely thank Dr. Hubert Vaudry for the invitation to write this article for this special issue of Peptides.
