Are the Three Models Mutually Exclusive

The previous description focused on the differences between the models; yet, it is obvious that none of the model suffices to explain all experimental observations. Some of the shortcomings of the lever-arm model and the gear-shift model have been pointed out above (see Section 4.1.2.). The explanatory power of the C-terminal latch hypothesis also has its limitations: for instance, it ignores a role for GPy in opening the guanine nucleotide exit pathway and therefore does not take into account the evidence suggesting that GPy may prime Ga by inducing a preactivated conformation -37] . The wasp venom mastoparan is the prototypical receptomimetic peptide; mastoparan has been reported to affect the circular dichroism spectrum of Gai sistent with a reduced helical content. This observation has been interpreted as evidence for mastoparan-lnduced melting of helix a5 [38]. However, this interpretation is incompatible with the currently proposed version of the C-terminal latch model, that is, rigid body motion of helix a5. Mastoparan was generally accepted to engage the C -terminus of Ga because its action was blocked by pertussis toxin pretreatment [39]. Pertussis toxin ADP-ribosylates a cysteine residue four amino acids removed from the C-terminus on Ga subunits of the Gi/Go/Gt family. Pertussis toxin also abrogates the action on Ga of D2N [29] and of receptomimetic lipoamines [40]. Binding of D2N and mastoparan is mutually exclusive [29], and the action of D2N is blocked by an antibody to the Ga C-terminus [21]. This has led to the proposal that cationic receptomimetic compounds, including D2N, bind to and exert their action via the C-terminal latch of Ga [21]. More recently, the position of D2N was examined in a crystal comprising Gai-1 and a second activating peptide KB-752 [41]. Like D2N, KB-752 has some GEF activity and, when combined, the action of the two peptides is potentiated. Those residues of D2N, which were traceable in the crystal structure, did not contact the C-terminus but rather were found in an adjacent position, namely in a cleft between the helix a4 and the p6 strand. Similarly, KB-752 was found in the pocket close to the switch 2 region, a position normally occluded by GPy [41] - It is not clear, however, why this crystal did not reveal evidence for a GDP-free form of Ga. This shortcoming may be remedied by studying D2N and KB-752 bound to the A326S-mutant of Gai-1 [36]- Nevertheless, and at the very least, the complex of Gai-1, D2N, and KB-752 highlights that the lever-arm and C-terminal latch models are not mutually exclusive; it is also conceivable that some aspects of the gear-shift model represent a better approximation of reality than the other two models.

Finally, it is evident that the bulk of the observations have been made with the pair Gai-1 (or Gat) and rhodopsin, given the lack of a better alternative, and the resulting insights are often extrapolated to all receptors and G proteins. This generalization is likely to be an oversimplification as receptors differ widely in the sequence of their intracellular loops. Moreover, a comparison of the structure of rhodopsin and of the p2-adrenergic receptor reveals substantial differences. For example, in the inactive rhodopsin, there is an ionic lock between the ERY motif at the (cytoplasmic) bottom of TM3 and an arginine residue at the bottom of TM6 (E247 in rhodopsin) which clamps rhodopsin in the ground state [42]. In contrast, in the p2-adrenergic receptor, the (homologous DRY motif-based) lock was seen in an open position, although the receptor was also trapped in its inactive conformation by the antagonist carazol (which acts as an inverse agonist like the 11-cis retinal chromophore in rhodopsin) [43, 44]. Thus, it is not surprising that the relative importance of contact sites on Gai-1 differ for individual receptors [45].

0 0

Post a comment