Most GPCRs are cell surface proteins that can signal through G proteins in a monomeric state. These proteins can however influence each other through possible allosteric interaction within dimeric or even oligomeric entities. The discovery of the mandatory heterodimeric nature of the GABAB receptor, a class C GPCR, stimulated research on the allosteric interaction between GPCRs. Whereas GABAB agonists bind to the GB1 subunit, the GB2 subunit is solely responsible for G protein activation. Such an asymmetric activation process was then reported for other class C GPCRs, including the dimeric and heterodimeric mGluRs but also for many class A GPCRs homo and heterodimers including the LTB4, OT, D2, 5HT, Mu-a2A receptors.
As mandatory dimers, the class C GPCRs represent an excellent model to elucidate the structural basis for this asymmetrical activation process. We then first identified the dimer interfaces of the Class C mGlu and GABAB receptors in their inactive and active state and identified TM4 and TM5 as the interface of the inactive state, while TM6 is indeed the main component of the active interface. Indeed, cross-linking TM4 or TM5 between subunits prevents their activation by agonists, while cross-linking TM6 generates constitutively active receptors. These findings were largely confirmed by the cryo-EM structures of either the mGlu5, GABAB or CaSR in their antagonists and agonist+PAM stabilized states. Indeed, these structures revealed a close contact between the extracellular tip of TM6, leaving enough space for the possible outward movement of at least one TM6 within a dimer for a G protein activation by one subunit. Such a hypothesis could largely explain that only one subunit within such dimers could be active at a time.
The observation that the GABAB PAMs bind at the active GABAB dimer interface, to the intracellular side of TM6 shades some strong doubts about this idea, as the presence of the PAM made impossible any outward shift of either one of the TM6. This was confirmed by the cryo-E structure of the active GABAB heterodimer bound to the active Gi protein. Indeed, no outward shift of either TM6 was observed, revealing a mainly symmetric structure of the 7TM dimer. However, the structure revealed that, steric hindrance between the active nucleotide free G proteins largely explain why both subunits cannot bind a G protein at a time. Accordingly, the possible asymmetry within class C GPCR dimers may rather come from the action of the G protein on only one subunit, rather than from a direct allosteric interaction between the 7TMs. Such a process may be limited to class C GPCRs as the structural basis for the activation of the 7TMs from the other GPCR classes is different.
 

暂无