In the presence of GTP, the catalytic GDP/GTP exchange function of rhodopsin leads to dissociation of the nucleotide-free MII'Gt complex and dissociation of the Gt holoprotein into the activated Gta subunit and the G|3y dimer as a quasi-irreversible step because of the subsequent GTP hydrolysis step. The fast rate of MII-Gt dissociation requires millimolar GTP concentrations [12]. Mil stabilization is abolished when GDP is bound [158,159]. Thus, either GDP or GTP can dissociate the nucleotide free MII-Gt complex. The scheme implies that Gt does not bind with significant affinity to MI or Mill, in agreement with other studies [77,79,160], Gt binds non-cooperatively to Mil with 1:1 stoichiometry, and shows a bell-shaped pH-dependence with a maximum at pH 7.6 [35]. For an in-depth analysis of the dissociation constant of the MII-Gt complex see [12]. Transient formation of extra Mil [161] provides a direct measure of the Gt activation kinetics and activation energy (see [73]). The stabilization of Mil is also observed with C-terminal peptides of Gt [Gta(340-350) and Gty(60-71)farnesyl] and arrestin but not with RK (see below).

3.6.2 The rhodopsin- Gt interface

Early approaches employing biochemical techniques, including proteolysis and chemical modifications of rhodopsin and Gt, allowed the first insight into the main structural elements (for a review see [7]). Major progress was made with the development of heterologous expression systems for these proteins, allowing the investigation of mutant proteins by biochemical and biophysical means. Binding sites at Gt

Another approach is to probe the interface by synthetic peptides from rhod-opsin and Gt. Synthetic peptides derived from Gta, including the C-terminal stretch [Gta(340-350] and the one spanning residues 311-323, compete with Gt to form extra Mil [162,163]. Interaction of these stretches with rhodopsin was confirmed by mutational analysis [164,165], Besides sites in the Ga subunit, the C-terminal domain of Gty(60-71)farnesyl could be identified as directly stabilizing Mil [166,167], Both the C-terminal peptide sequence and the farnesyl moiety are specific determinants for the R*-Gt interaction [166], The C-terminal tail of the Gty-subunit is masked in the Gt(3y dimer and becomes exposed on collisional coupling of the holoprotein to the receptor (see [168]). The crystal structure of Gt [169] revealed that both C-terminal regions of Gta and Gty, although only partly resolved in the structure, are localized to a common surface of the Gt holoprotein (see [167,170]). By transferred nuclear Overhauser effect spectroscopy, a helical turn conformation followed by an open reverse turn was determined for the Gta(340-350) peptide in the rhodopsin-bound state [171]. Mil stabilization, with analogs of this peptide, provided information about individual residues contributing to the interaction with R* [172]. Binding of the peptides from the C-termini of Gta and Gty was also confirmed by a photoregeneration assay [27]. FTIR difference spectroscopy was extended to the R*Gt interaction. Specific protonation changes are induced when Gt or C-terminal peptides of Gta and Gty bind to rhodopsin and shift the MI/MII equilibrium [88,139,173,174], Binding sites at rhodopsin

Loop structures of rhodopsin were shown to interact with Gt by competition with synthetic peptides and mutagenesis approaches. Only peptides from the second and third loop, and the cytoplasmic helix VIII (Figure 2), competed for Gt-dependent stabilization of Mil [175]. Interaction of these loops with Gt was confirmed by nucleotide exchange catalysis of mutant rhodopsins [176-179], flash photolysis [27,176], light scattering [180] and fluorescence techniques [179]. The lack of peptide inhibition and the overall minor effect of point mutations in the first cytoplasmic loop suggested that the residues connecting helices I and II are not involved directly in recognition by Gt [181], Sites at either the second or the third cytoplasmic loop, in conjunction with the fourth loop, appear to be sufficient to maintain the empty-site R*-Gt interaction. A more complex interaction pattern may be required to allow for the fast release of GDP from the Gt-rhodopsin collisional complex. GTP binding to the empty Gta site, and the subsequent release of Gta#GTP from rhodopsin, requires intact structures of both the second and the third loop [176,180]. It is not known, however, whether only loop structures contribute to the interaction surface. Light-induced exposure of the cytoplasmic end of transmembrane helix VII, as probed by specific binding of a monoclonal antibody recognizing this epitope, may argue for additional binding sites closer to the hydrophobic core of rhodopsin [182], The fourth loop follows helix VII and is restricted by two palmitoylated cysteines anchoring it to the disk membrane. According to the X-ray structure, this loop adopts a cationic amphipathic helical conformation (helix VIII) with Lys311 and Arg314 facing the cytoplasm. The cytoplasmic helix VIII is attached to helix VII by a short linker (VIII in Figure 2). Interaction of this loop with Gt has been discussed previously (see [179]). Based on fluorescence studies with a corresponding loop peptide the interaction of this loop was specified for the (3y-subunit [183,184], Further studies have confirmed interaction of the fourth loop with Gt [27, 178,179] and suggested that the N-terminal part of this loop is involved in binding of the C-terminus of Gta and that this loop plays a structural role in binding of G(3y [27,179], Only minor effects on Gt activation by R* were seen when the palmitoyl modifications at Cys322 and Cys323 were removed, either using high concentratons of NH2OH in the dark [185] or mutational substitution of the palmitoyl-anchoring cysteines by serines [179,186],

3.6.3 Conclusions

Although most binding sites on both rhodopsin and Gt are identified, a mutual assignment of the binding sites is not possible on the data available. Controversial assignments could be explained by the formation of receptor dimers (see [187]). Also, the hydrophobicity and charge pattern derived from the available ground state structures of both proteins can only give vague hints about how docking of the two proteins could occur. Critical residues, e.g. the ERY sequence, are buried in the structure. From site-directed spin labelling studies (see [188]) and other work it is known that the receptor undergoes large structural changes upon receptor activation, thereby exposing buried binding sites (see [7,187]). Similar changes are thought to occur in the G-protein upon receptor binding. In the absence of a clarifying crystal structure of the R*«Gt-complex, new approaches were devised to obtain insight into the R-Gt interface. The nuclear Overhauser effect between site-directed 19F-labels on the cytoplasmic receptor surface can be used to measure light-induced conformational changes [189]. In another approach, site-directed Cys mutagenesis was used to introduce crosslinkers at specific sites of the rhodopsin surface. Covalent crosslinking of the R*»Gt-complex followed by trypsin degradation and mass spectrometric analysis can reveal the Gt-sequences binding to the labelled sites on rhodopsin [190,191].

Different models were proposed for temporal aspects and the mode of interaction. Based on kinetic studies a "sequential fit" mechanism involving a sequence of microscopic recognition via interacting C-termini of Gt and conformational interlocking was proposed [167]. By mutagenesis studies, the a5 helix was identified as a functional microdomain, affecting nucleotide release [192], and the unit consisting of Gta's C-terminus/a5 helix/|36/a5 loop was suggested to constitute a dominant channel for transmission of the GPCR-induced conformational change leading to G-protein activation [193], It appears that the receptor uses the G-protein's heterodimer as a lever, tilting it to pull open the guanine nucleotide binding pocket of Ga [194],

Was this article helpful?

0 0

Post a comment