Computational analysis of the extracellular domain of the Ca2+-sensing receptor: An alternate model for the Ca2+ sensing region

Abstract

The extracellular Ca2+ sensing receptor (CaSR) belongs to Class C G-protein-coupled receptors (GPCRs) which include receptors for amino acids, γ-aminobutyric acid and glutamate neurotransmitters. CaSR has been described as having an extended sequence containing a Ca2+ binding pocket within an extracellular amino (N)-terminal domain, called a Venus Fly Trap (VFT) module. CaSR is thought to consist of three domains: 1) a Ca2+−sensory domain, 2) a region containing 7 transmembrane (TM) helices, and 3) a carboxy (C)-terminal tail. We find that SPOCTOPUS (a combination of hidden Markov models and artificial neural networks) predicts that Homo sapiens CaSR contains two additional TM helices (190D – G210; 262S–E282), with the second TM helix containing a pore-lining region (265K – I280). This predicts that the putative Ca2+ sensory domain is within an extracellular loop, N-terminal to the highly conserved heptahelical bundle. This loop contains both the cysteine-rich domain (537V – C598) and a 14 residue “linker” sequence (599I – F612) thought to support signal transmission to the heptahelical bundle. Thus domain 1 may contain a 189 residue N-terminal extracellular region followed successively by TM-1, a short intracellular loop, TM-2 and a 329 residue extracellular loop; rather than the proposed 620 residue VFT module based on crystallography of the N-terminal region of mGluR1. Since the topologies of the two proteins differ, the published CaSR VFT model is questionable. CaSR also contains multiple caveolin-binding motifs and cholesterol-binding (CRAC/CARC) domains, facilitating localization to plasma membrane lipid rafts. Ion sensing may involve combination of pore-lining regions from CaSR dimers and CaSR-bound caveolins to form ion channels capable of monitoring ionized Ca2+ levels.