Gi protein activity is sensitised by type 5 adenylyl cyclase (AC5)

Markus Milde, Moritz Buenemann. University Marburg, School of Pharmacy, Pharmacology and Clinical Pharmacy, Karl-von-Frisch-Str. 1, 35043 Marburg, Germany

The second messenger cAMP is synthesised by adenylyl cyclases (AC). ACs themselves are regulated through receptors and G proteins. Here we investigated the dynamic interactions between the alpha-subunit of the type 1 inhibitory G protein (Gαi1) and human type 5 adenylyl cyclase (AC5). To allow detection, these proteins were labelled with CFP and YFP, respectively. YFP was fused to the N-terminus of AC5 using the 3-fragment Multisite Gateway Pro cloning system by Invitrogen. YFP was cloned into Gαi1 between amino acids 91 and 92 as described previously (Buenemann et al., PNAS 2003). CFP-Gαi1 was constructed accordingly. The interaction dynamics were recorded with high temporal resolution by means of FRET in HEK cells. The cells additionally expressed unlabelled α2A adrenoceptor and Gβγ subunits to allow the studying of agonist induced changes in the interaction. During the experiments the cells were superfused with either agonist/buffer-solutions in different concentrations or agonist-free buffer to wash out the agonist between applications and to allow the interaction to recover to baseline. Upon application of noradrenalin a developing FRET signal could be observed, which indicates the interaction between Gαi1-CFP and YFP-AC5. In contrast to previous results on the interaction kinetics of Gi proteins the interaction between Gαi1-CFP and YFP-AC5 recovered with a remarkably slow time course after the agonist was withdrawn. For further investigation we compared the interaction between Gαi1 and AC5 with the interaction between Gαi1-YFP and Gβγ-CFP (Gi-FRET). Furthermore we tested the acceleration of deactivation of Gi proteins and the dissociation of Gαi1-CFP from YFP-AC5 by overexpression of RGS4. The dissociation of Gαi1-CFP from YFP-AC5 after agonist withdrawal takes about 2 times longer than the deactivation of the Gi protein itself (half-life of 55.1±4.3 s and 29.3±3.7 s, respectively). While RGS4 accelerates the Gi protein deactivation it has hardly any effect on the dissociation of Gαi1-CFP from YFP-AC5 (half-life 15.8±1.7 s and 48.3±4.3 s, respectively). From this we hypothesised that AC5 possibly traps active G proteins which would affect the equilibrium in the G protein cycle and shift it towards a higher amount of active G proteins. Should this hypothesis be correct, the concentration-response curve of the interaction between Gαi1-CFP and YFP-AC5 should be shifted leftwards compared to the concentration-response of the G protein activation. Indeed, we found the concentration-response curve for the Gαi1-AC5 interaction to be shifted leftwards in comparison to the G protein activation concentration-response (Gαi1-AC5 interaction: log(EC50) = -8.52±0.13 log(M); Gi protein activation: log(EC50) = -8.05±0.22 log(M)). Effector-specific influence on the dynamics of the G protein cycle may represent a powerful mechanism to fine-tune receptor evoked responses. (Results are presented as mean with SEM of at least 10 individual experiments.)