D2/D3 dopamine receptor heterodimers exhibit unique functional properties.

Evidence for heterodimerization has recently been provided for dopamine D(1) and adenosine A(1) receptors as well as for dopamine D(2) and somatostatin SSTR(5) receptors. In this paper, we have studied the possibility that D(2) and D(3) receptors interact functionally by forming receptor heterodimers. Initially, we split the two receptors at the level of the third cytoplasmic loop into two fragments. The first, containing transmembrane domains (TM) I to V and the N-terminal part of the third cytoplasmic loop, was named D(2trunk) or D(3trunk), and the second, containing the C-terminal part of the third cytoplasmic loop, TMVI and TMVII, and the C-terminal tail, was named D(2tail) or D(3tail). Then we defined the pharmacological profiles of the homologous (D(2trunk)/D(2tail) and D(3trunk)/D(3tail)) as well as of the heterologous (D(2trunk)/D(3tail) and D(3trunk)/D(2tail)) cotransfected receptor fragments. The pharmacological profile of the cross-cotransfected fragments was different from that of the native D(2) or D(3) receptors. In most cases, the D(3trunk)/D(2tail) was the one with the highest affinity for most agonists and antagonists. Moreover, we observed that all of these receptor fragments reduced the expression of the wild type dopamine D(2) and D(3) receptors, suggesting that D(2) and D(3) receptors can form complexes with these fragments and that these complexes bind [(3)H]nemonapride less efficiently or are not correctly targeted to the membrane. In a second set of experiments, we tested the ability of the split and the wild type receptors to inhibit adenylyl cyclase (AC) types V and VI. All of the native and split receptors inhibited AC-V and AC-VI, with the exception of D(3), which was unable to inhibit AC-VI. We therefore studied the ability of D(2) and D(3) to interact functionally with one another to inhibit AC-VI. We found that with D(2) alone, R-(+)-7-hydroxydypropylaminotetralin hydrobromide inhibited AC-VI with an IC(50) of 2.05 +/- 0.15 nm, while in the presence of D(2) and D(3) it inhibited AC-VI with an IC(50) of 0.083 +/- 0.011 nm. Similar results were obtained with a chimeric cyclase made from AC-V and AC-VI. Coimmunoprecipitation experiments indicate that D(2) and D(3) receptors are capable of physical interaction.