For certain class C GPCRs, such as the GABAb receptor and taste receptors, the evidence for heterodimerization is overwhelming. However, there is no reported heterodimerization among different metabotropic glutamate receptors, despite several closely related receptor subtypes, coexpressed in the same neurons [51, 98]. Ability to form class C dimers is clearly a tightly regulated phenomenon. What is then to be made of FRET studies which indicate numerous interacting combinations of different class A GPCRs  supported by more recent techniques detecting heterodimers using bimolecular fluorescence complementation [142, 143]? Often, these investigations suggest that closely related GPCRs, such as the P1/P2- adrenoceptors and CCR2/CCR5 chemokine receptors, form homo- and heterodimers with equal efficiency [95, 144] . If the same range of heterodimers existed in physiology, would the original (and successful) definitions of GPCR subtypes using in vitro pharmacology have been possible?
To stringently test the relevance of GPCR heterodimers, strict criteria should accompany their acceptance as a new pharmacological entity. In the first instance, identification needs to be supported in native cells, in addition to heterologous systems. For example, co-immunoprecipitation and coopera-
tivity in ligand binding and function have indicated |/S or k/8 opioid, D-/D2 dopamine, and A-/A2A adenosine receptor heterodimers -n vivo [145-148]. Second, many heterodimers are indicated by novel patterns of signaling or regulation when the different GPCRs are expressed together. Thus, D1 /D2 receptor coexpression results in Gq/11 activation, compared to Gs or Gi coupling for the individual dopamine receptors . In addition, coexpression of |/S opioid receptors produces a unique pattern of arrestin-mediated ERK signaling . Moreover, trafficking of orexin-1 receptors is altered by coexpression of the CB1 cannabinoid receptor, and can then be modulated by cannabinoid ligands  - These experiments demonstrate close interaction between GPCR subtypes, with physiological relevance and the potential for novel regulation by drugs. Yet, the studies are limited in that they do not distinguish between a new pharmacologically distinct heterodimer and cross-talk between closely complexed receptors in the same microdomains.
A cornerstone for the proof of heterodimerization should therefore be binding experiments, to show that communication exists between different orthosteric binding sites. An important thermodynamic consideration is that any effects should be reciprocal. If the binding of a selective ligand to one protomer affects selective ligand binding at the second, then the converse should also be the case. Surprisingly, this is manifestly true for some heterodi-mer combinations (e.g., CCR5/CCR2; ) but not for others. For example, an entirely new pharmacology is generated for kappa and delta opioid receptor heterodimers, with eliminated affinity for the selective agonists at either parent receptor  - In such cases, the integrity of the individual protomer orthosteric sites cannot be maintained, and domain swapping between monomers might generate pharmacologically distinct binding sites in the heterodi-mer. It is then imperative to show that this does not result from GPCR overexpression, but is physiologically relevant, as appears to be case for native k/S opioid heterodimers -148]- For heterodimers formed by two protomers which retain their individual characteristics, the acceleration of dissociation kinetics provides the strongest validation of communicating binding sites [114, 130] - For CCR5/CCR2 receptor heterodimers, as for the respective homodi-mers, there is strong negative cooperativity, which suggests only one ligand occupies the heterodimer  - Here, cross-talk between the binding of different ligands can certainly generate a new structure activity relationship, and the presentation of a different dimer cytoplasmic interface has potential (though unproven) to interact uniquely with different effectors. If such careful definitions of GPCR heterodimers are repeated more widely, there is certainly scope for new avenues in drug discovery in the future.
Was this article helpful?