Francesca Novi

The impact of G‐protein‐coupled receptor hetero‐oligomerization on function and pharmacology

Although highly controversial just a few years ago, the idea that G‐protein‐coupled receptors (GPCRs) may undergo homo‐oligomerization or hetero‐oligomerization has recently gained considerable attention. The recognition that GPCRs may exhibit either dimeric or oligomeric structures is based on a number of different biochemical and biophysical approaches. Although much effort has been spent to demonstrate the mechanism(s) by which GPCRs interact with each other, the physiological relevance of this phenomenon remains elusive. An additional source of uncertainty stems from the realization that homo‐oligomerization and hetero‐oligomerization of GPCRs may affect receptor binding and activity in different ways, depending on the type of interacting receptors. In this brief review, the functional and pharmacological effects of the hetero‐oligomerization of GPCR on binding and cell signaling are critically analyzed.

Potent activation of dopamine D3/D2 heterodimers by the antiparkinsonian agents, S32504, pramipexole and ropinirole

Recombinant, human dopamine D3 and D2 receptors form functional heterodimers upon co‐expression in COS‐7 cells. Herein, actions of the antiparkinsonian agents, S32504, ropinirole and pramipexole, at D3/D2L heterodimers were compared to their effects at the respective monomers and at split, chimeric D3trunk/D2tail and D2trunk/D3tail receptors: the trunk incorporated transmembrane domains (TDs) I–V and the tail TDs VI and VII. In binding assays with the antagonist [3H]nemonapride, all agonists were potent ligands of D3 receptors showing, respectively, 100‐, 18‐ and 56‐fold lower affinity at D2L receptors, mimicking the selective D3 receptor antagonist, S33084 (100‐fold). At D3trunk/D2tail receptors, except for ropinirole, all drugs showed lower affinities than at D3 sites, whereas for D2trunk/D3tail receptors, affinities of all drugs were higher than at D2L sites. The proportion of high affinity binding sites recognized by S32504, pramipexole and ropinirole in membranes derived from cells co‐expressing D3 and D2L sites was higher than in an equivalent mixture of membranes from cells expressing D3 or D2L sites, consistent with the promotion of heterodimer formation. In contrast, the percentage of high and low affinity sites (biphasic isotherms) recognized by S33084 was identical. Functional actions were determined by co‐transfection of a chimeric adenylyl cyclase (AC)‐V/VI insensitive to D3 receptors. Accordingly, D3 receptor‐transfected cells were irresponsive whereas, in D2L receptor‐transfected cells, agonists suppressed forskolin‐stimulated cAMP production with modest potencies. In cells co‐transfected with D3 and D2L receptors, S32504, ropinirole and pramipexole potently suppressed AC‐V/VI with EC50s 33‐, 19‐ and 11‐fold lower than at D2L receptors, respectively. S32504 also suppressed AC‐V/VI activity at split D3trunk/D2tail and D2trunk/D3tail chimeras transfected into COS‐7 cells. In conclusion, antiparkinson agents behave as potent agonists at D3/D2‘heterodimers’, though any role in their actions in vivo remains to be demonstrated.

S33138 [N-[4-[2-[(3aS,9bR)-8-cyano-1,3a,4,9b-tetrahydro[1]-benzopyrano[3,4-c]pyrrol-2(3H)-yl)-ethyl]phenylacetamide], A Preferential Dopamine D3 versus D2 Receptor Antagonist and Potential Antipsychotic Agent: I. Receptor-Binding Profile and Functional Actions at G-Protein-Coupled Receptors

The novel, potential antipsychotic, S33138 (N-[4-[2-[(3aS,9bR)-8-cyano-1,3a,4,9b-tetrahydro[1]benzopyrano[3,4-c]pyrrol-2(3H)-yl)-ethyl]phenylacetamide), displayed ∼25-fold higher affinity at human (h) dopamine D3 versus hD2L (long isoform) and hD2S (short isoform) receptors (pKi values, 8.7, 7.1, and 7.3, respectively). Conversely, haloperidol, clozapine, olanzapine, and risperidone displayed similar affinities for hD3, hD2L, and hD2S sites. In guanosine-5′-O-(3-[35S]thio)-triphosphate ([35S]-GTPγS) filtration assays, S33138 showed potent, pure, and competitive antagonist properties at hD3 receptors, displaying pKB and pA2 values of 8.9 and 8.7, respectively. Higher concentrations were required to block hD2L and hD2S receptors. Preferential antagonist properties of S33138 at hD3 versus hD2L receptors were underpinned in antibody capture/scintillation proximity assays (SPAs) of Gαi3 recruitment and in measures of extracellular-regulated kinase phosphorylation. In addition, in cells cotransfected with hD3 and hD2L receptors that assemble into heterodimers, S33138 blocked (pKB, 8.5) the inhibitory influence of quinpirole upon forskolin-stimulated cAMP formation. S33138 had low affinity for hD4 receptors (<5.0) but revealed weak antagonist activity at hD1 receptors (Gαs/SPA, pKB, 6.3) and hD5 sites (adenylyl cyclase, 6.5). Modest antagonist properties were also seen at human serotonin (5-HT)2A receptors (Gαq/SPA, pKB, 6.8, and inositol formation, 6.9) and at 5-HT7 receptors (adenylyl cyclase, pKB, 7.1). In addition, S33138 antagonized hα2C adrenoceptors ([35S]GTPγS, 7.2; Gαi3/SPA, 6.9; Gαo/SPA, 7.3, and extracellular-regulated-kinase, 7.1) but not hα2A or hα2B adrenoceptors (<5.0). Finally, in contrast to haloperidol, clozapine, olanzapine, and risperidone, S33138 displayed negligible affinities for multiple subtypes of α1-adrenoceptor, muscarinic, and histamine receptor. In conclusion, S33138 possesses a distinctive receptor-binding profile and behaves, in contrast to clinically available antipsychotics, as a preferential antagonist at hD3 versus hD2 receptors.

Partial agonist actions of aripiprazole and the candidate antipsychotics S33592, bifeprunox, N-desmethylclozapine and preclamol at dopamine D2L receptors are modified by co-transfection of D3 receptors: potential role of heterodimer formation

Aripiprazole and the candidate antipsychotics, S33592, bifeprunox, N‐desmethylclozapine (NDMC) and preclamol, are partial agonists at D2 receptors. Herein, we examined their actions at D2L and D3 receptors expressed separately or together in COS‐7 cells. In D2L receptor‐expressing cells co‐transfected with (D3 receptor‐insensitive) chimeric adenylate cyclase‐V/VI, drugs reduced forskolin‐stimulated cAMP production by ∼20% versus quinpirole (48%). Further, quinpirole‐induced inhibition was blunted by aripiprazole and S33592, confirming partial agonist properties. In cells co‐transfected with equal amounts of D2Land D3 receptors (1 : 1), efficacies of aripiprazole and S33592 were attenuated. Further, in cells co‐transfected with D2L and an excess of D3 receptors (1 : 3), aripiprazole and S33592 were completely inactive, and they abolished the actions of quinpirole. Likewise, bifeprunox, NDMC and preclamol lost agonist properties in cells co‐transfected with D2Land D3 receptors. Accordingly, at split D2trunk/D3tail and D3trunk/D2tail chimeras, agonist actions of quinpirole were blocked by aripiprazole and S33592 that, like bifeprunox, NDMC and preclamol, were inactive alone. Conversely, when a 12 amino acid sequence in the third intracellular loop of D3 receptors was replaced by the homologous sequence of D2L receptors, aripiprazole, S33592, bifeprunox, NDMC and preclamol inhibited cAMP formation by ∼20% versus quinpirole (42%). Moreover, at D2L receptor‐expressing cells co‐transfected with modified D3i3(D2) receptors, drugs behaved as partial agonists. To summarize, low efficacy agonist actions of aripiprazole, S33592, bifeprunox, NDMC and preclamol at D2L receptors are abrogated upon co‐expression of D3 receptors, probably due to physical association and weakened coupling efficacy. These findings have implications for the functional profiles of antipsychotics.

The Paired Activation of the Two Components of the Muscarinic M3 Receptor Dimer Is Required for Induction of ERK1/2 Phosphorylation

Muscarinic M3 receptors stimulate ERK1/2, the mitogen-activated protein kinase pathway. A mutant of the muscarinic M3 receptor in which most of the third intracellular (i3) loop had been deleted (M3-short) completely lost the ability to stimulate the ERK1/2 phosphorylation in COS-7 cells. This loss was evident despite the fact that the receptor was able to couple efficiently to the phospholipase C second messenger pathway. In co-transfected cells, M3-short greatly reduced the ability of M3 to activate ERK1/2. In another set of experiments we tested the ability of a mutant M3/M2(16aa) receptor, in which the first 16 amino acids of the i3 loop of the M3 receptor were replaced with the corresponding segment of the muscarinic M2 receptor to stimulate ERK1/2 phosphorylation. This mutant is not coupled to Gαq, but it is weakly coupled to Gαi. Despite its coupling modification this receptor was able to stimulate ERK1/2 phosphorylation. Again, M3-short greatly reduced the ability of M3/M2(16aa) to activate ERK1/2 in co-transfected cells. Similar results were obtained in stable-transfected Chinese hamster ovary (CHO) cells lines. In CHO M3 cells carbachol induced a biphasic increase of ERK1/2 phosphorylation; a first increase at doses as low as 0.1 μm and a second increase starting from 10 μm. In CHO M3-short and in double-transfected CHO M3/M3-short cells we observed only the lower doses increase of ERK1/2 phosphorylation; no further increase was observed up to 1 mm carbachol. This suggests that in double-transfected CHO cells M3-short prevents the effect of the higher doses of carbachol on the M3 receptor. In a final experiment we tested the ability of co-transfected chimeric α2/M3 and M3/α2 receptors to activate the ERK1/2 pathway. When given alone, carbachol and, to a lesser extent, clonidine, stimulated the coupling of the co-transfected chimeric receptors to the phospholipase C second messenger pathway, but they were unable to stimulate ERK1/2 phosphorylation. On the contrary, a strong stimulation of ERK1/2 phosphorylation was observed when the two agonists were given together despite the fact that the overall increase in phosphatidylinositol hydrolysis was not dissimilar from that observed in cells treated with carbachol alone. Our data suggest that the activation of the ERK1/2 pathway requires the coincident activation of the two components of a receptor dimer.

Receptor crosstalk: haloperidol treatment enhances A2A adenosine receptor functioning in a transfected cell model

A2A adenosine receptors are considered an excellent target for drug development in several neurological and psychiatric disorders. It is noteworthy that the responses evoked by A2A adenosine receptors are regulated by D2 dopamine receptor ligands. These two receptors are co-expressed at the level of the basal ganglia and interact to form functional heterodimers. In this context, possible changes in A2A adenosine receptor functional responses caused by the chronic blockade/activation of D2 dopamine receptors should be considered to optimise the therapeutic effectiveness of dopaminergic agents and to reduce any possible side effects. In the present paper, we investigated the regulation of A2A adenosine receptors induced by antipsychotic drugs, commonly acting as D2 dopamine receptor antagonists, in a cellular model co-expressing both A2A and D2 receptors. Our data suggest that the treatment of cells with the classical antipsychotic haloperidol increased both the affinity and responsiveness of the A2A receptor and also affected the degree of A2A–D2 receptor heterodimerisation. In contrast, an atypical antipsychotic, clozapine, had no effect on A2A adenosine receptor parameters, suggesting that the two classes of drugs have different effects on adenosine–dopamine receptor interaction. Modifications to A2A adenosine receptors may play a significant role in determining cerebral adenosine effects during the chronic administration of antipsychotics in psychiatric diseases and may account for the efficacy of A2A adenosine receptor ligands in pathologies associated with dopaminergic system dysfunction.

Functional rescue of the inactive splice variant of the dopamine D3 receptor D3nf.

We previously found that the dopamine D3 receptor can be split at the third cytoplasmic loop into two fragments (D3trunk and D3tail), and that the mixture of the two fragments retains the binding and functional activity of the wild type receptor. The dopamine D3 receptor gene gives rise to several inactive receptor splice variants, one of which is the D3nf. Since this gene variant very closely resembles our D3trunk fragment, in this study we investigated if the transfection of D3nf with D3tail could result in the rescue of a functional dopamine receptor. Our experiments showed that D3tail can indeed rescue the activity of D3nf, and that the pharmacological profile of this split D3nf/D3tail receptor is identical to that of the wild type D3 receptor.

Apomorphine, dopamine and phenylethylamine reduce the proportion of phosphorylated insulin receptor substrate 1.

We tested the ability of dopamine, apomorphine, phenylethylamine and pergolide to inhibit the proliferation of fetal calf serum-stimulated human breast cancer (MCF)-7 cells. While the first three compounds were able to block the proliferation of MCF-7 cells, pergolide failed to do so (up to 100 microM). The inhibitory effect of dopamine, apomorphine and phenylethylamine was also evident in serum-starved insulin-stimulated MCF-7 cells. Apomorphine also inhibited the proliferation of the human oestrogen receptor-negative breast cancer (MDA-MB231) and prostate carcinoma (LNCaP) cell lines. In a second set of experiments, we measured the ability of dopamine, apomorphine, phenylethylamine and pergolide to inhibit the phosphorylation (or increase the dephosphorylation) of the insulin receptor substrate (IRS)-1, a major intracellular substrate of the insulin-like growth factor (IGF)-1 receptor. Dopamine, apomorphine and phenylethylamine all reduced to zero the level of phosphorylated IRS-1 with potencies ranging between 0.01 and 1 microM. Finally, we found that fibroblasts from IRS-1 null (-/-) mice were less sensitive to the anti-proliferative effect of apomorphine compared to fibroblasts from wild type-mice, suggesting that the inhibition of IRS-1 phosphorylation by apomorphine is an important aspect of the activity of this compound.

Deoxamuscaroneoxime derivatives as useful muscarinic agonists to explore the muscarinic subsite: Demox, a modulator of orthosteric and allosteric sites at cardiac muscarinic M2 receptors

A series of muscarinic agonists, straight chained, branched, cyclic alkyl and aromatic derivatives of the oxime 1 (demox) was designed with the aim of investigating their activity on muscarinic receptor subtypes. Effects on M1 receptor were assessed functionally by a microphysiometer apparatus, while M2, M3, and M4 receptor potency and affinity were studied on isolated preparations of guinea pig heart, ileum, and lung, respectively. The results suggest that the substitution of a hydrogen with a long side-chain or bulky group generally induces a decrease in potency at M1 and M3 subtypes, while a general increase in this parameter is obtained at M2 subtype. Among the agonists 2-18, compound 4 behaves as a full agonist with a preference for M3 subtype. Moreover, compound 12 is inactive at M1 and M4 receptors while it displays a full agonist activity at M2 and M3 subtypes. Since demox displays a variable response on cardiac M2 receptors regulating heart force, an in-depth inquiry of the functional behaviour of this compound was carried out at M2 receptors. In presence of 10(-11) and 10(-10) M demox, the binding of [3H]-NMS was increased by approximately 30% as a consequence of an increase of the association of [3H]-NMS to membranes; this effect was not observed in presence of a higher concentration of [3H]-NMS. Higher concentrations of demox decreased the binding of [3H]-NMS to heart atrial membranes but significantly retarded the dissociation of this radioligand. Our results suggest that demox may interact with orthosteric and allosteric sites of atrial M2 muscarinic receptor.

Polyamines May Modulate Both G Protein-Coupled Receptors and G Proteins

Polyamines pipertramine (PIP) and pipertramine amide (PIPAM), unlike prototype methoctramine, displayed a potent inotropic effect in guinea pig atria. While the inotropic effect of PIP may be rationalized, as previously suggested, through a direct activation of a G i / o protein, PIPAM effects were possibly the result of multiple interactions. Besides a direct G i / o protein activation, PIPAM activated also muscarinic M2 receptors by interacting most likely with an allosteric site. Binding experiments in CHO cells, expressing human cloned muscarinic M 2 receptors, confirmed that both compounds were effective muscarinic ligands.