Macromolecular Complexes of GPCRs

GPCR signaling is achieved via interaction with several classes of molecules, including G proteins, arrestin, and receptor activity-modifying proteins (RAMPs) [106]. In addition, many GPCRs may form homodimers and/or heterodimers, though the functional significance of GPCR oligomerization remains unclear [107-109] . Formation of GPCR macromolecular complexes has been studied in situ by atomic force microscopy (AFM) [110, 111], as well as by indirect approaches, such as disulfide cross - linking, site - directed muta-genesis, and spectroscopy [112-115]. These experimental data provide a starting point for the computational modeling of GPCR functional complexes.

A semiempirical model of rhodopsin oligomerization based on AFM-derived geometry was proposed that includes dimer interface contacts formed by TM4 and TM5, as well as residues in ICL2 and in the C-terminal region [111] . Also, an attempt to build a complete model of the rhodopsin signaling complex was performed by combining Gt heterotrimer structure with the predicted functional tetramer of rhodopsin [ 105] , though the stoichiometry and helical arrangement of this model are questionable (e.g., an axial rotation of TM6 by 90°). Fanelli and coworkers generated putative models of dopamine (D2)/adenosine (A2a) GPCR heterodimers [102] and lutropin receptor homodimers [116], by rigid body docking of predicted monomer models. The initial docking models were clustered, filtered, and ranked according to an empirical index defined by membrane topology, described by the normalized tilt angle and z-offset for the receptor [117]. The homodimer models generated by the method suggest a binding interface formed by TMs 4, 5, and 6 without involvement of loop regions, while heterodimer modeling could not reliably distinguish between two alternative configurations. Other studies from Fanelli and coworkers combined mutation data with structural modeling to suggest a mechanism of G protein binding and activation for several GPCRs. The common feature of these models, that is, binding of the C-terminus of Gaqa to the conserved D(E)RY motif, accessible only in the "open" conformation of the activated receptor [103, 104, 118, 119], has been recently confirmed in highresolution crystal structure of the Ops*-Ga-CT complex [98].

While these models deviate significantly in the details of intermolecular contacts and conformations of protein subunits, they provide tangible hypotheses for stoichiometry, orientation, and binding interfaces in GPCR signaling complexes useful for the design of further experimental tests. Additional spectroscopy, NMR, cryo- electron microscopy, and crystallography experiments are needed to resolve ambiguous structural and functional details of these complexes. More accurate methods for conformational modeling of membrane proteins - 120, 121] - which combine energy-based global optimization techniques with EM and other experimental restraints, may augment this analysis and improve its accuracy.

0 0

Post a comment