Ligand-receptor interactions as controlled by wild-type and mutant Thr(370)Lys alpha2B-adrenoceptor-Galpha15 fusion proteins

Agonist-directed trafficking of signalling at serotonin 5-HT2A, 5-HT2B and 5-HT2C-VSV receptors mediated Gq/11 activation and calcium mobilisation in CHO cells

Several examples of agonist-directed trafficking of receptor signalling at 5-HT2A and 5-HT2C receptors have been reported that involve independent downstream transduction pathways. We now report the functional selectivity of a series of chemically diverse agonists at human (h)5-HT2A, h5-HT2B and h5-HT2C-VSV by examining two related responses, the upstream activation of Gq/11 proteins in comparison with its associated cascade of calcium mobilisation. At the h5-HT2A receptor, d-lysergic acid diethylamide (LSD) and the antiparkinsonian agents lisuride, bromocriptine and pergolide exhibit a higher potency for Gq/11 activation than calcium release in contrast with all the other tested ligands such as 5-HT, mCPP and BW723C86, that show an opposite preference of signalling pathway. Comparable observations are made at h5-HT2B and h5-HT2C-VSV receptors, suggesting a similar mechanism of functional selectivity for the three serotonin receptors. Interestingly, the non-hallucinogenic compound lisuride behaves as a partial agonist for both Gq/11 activation and calcium release at the three 5-HT2 receptors, in contrast with DOI, LSD, pergolide and bromocriptine, which are known to provoke hallucinations, and behave as more efficacious agonists. Hence, a functional selectivity for Gq/11 activation together with a threshold of efficacy at h5-HT2A (and possibly h5-HT2B and/or h5-HT2C-VSV) may contribute to hallucinogenic liability. Thus, our results extend the notion of agonist-directed trafficking of receptor signalling to all the 5-HT2-receptor family and indicate that measures of Gq/11 activation versus calcium release may be useful to identify more effective therapeutic drugs with limited side effects.

Alpha 2A-adrenoceptor-specific stimulation of [35S]GTP gamma S binding to membrane preparations of rat frontal cortex

Functional activation of alpha 2A adrenergic receptors in the crude membranes from rat frontal cortex was studied by a [35S]-guanosine 5'-O-(gamma-thiotriphosphate) ([35S]GTP gamma S) binding assay. alpha 2A agonists UK14304 and guanfacine decreased the ability of GDP to compete with [35S]GTP gamma S binding to the membranes and 0.1 mM GDP was found to be optimal for the following functional experiments. However, even after careful optimization of experimental conditions the specificity of ligands for rat alpha 2 adrenoceptors were not sufficient, as agonists as well as antagonists became activators of other signal transduction systems before achieving their maximal effect in the alpha 2A-adrenergic system. Only using compromising concentration of agonist (up to 1 microM UK14304) and antagonist (up to 1 microM RS79948) to inhibit agonist's effect, allowed us to filtrate out alpha 2A specific effect for characterization of signal transduction in rat frontal cortex membranes for the comparison efficacies of this system for different animals from behavioral experiments.

Ligand screening system using fusion proteins of G protein-coupled receptors with G protein alpha subunits

G protein-coupled receptors (GPCRs) constitute one of the largest families of genes in the human genome, and are the largest targets for drug development. Although a large number of GPCR genes have recently been identified, ligands have not yet been identified for many of them. Various assay systems have been employed to identify ligands for orphan GPCRs, but there is still no simple and general method to screen for ligands of such GPCRs, particularly of G(i)-coupled receptors. We have examined whether fusion proteins of GPCRs with G protein alpha subunit (Galpha) could be utilized for ligand screening and showed that the fusion proteins provide an effective method for the purpose. This article focuses on the followings: (1) characterization of GPCR genes and GPCRs, (2) identification of ligands for orphan GPCRs, (3) characterization of GPCR-Galpha fusion proteins, and (4) identification of ligands for orphan GPCRs using GPCR-Galpha fusion proteins.

Ca2+ responses in Chinese hamster ovary-K1 cells demonstrate an atypical pattern of ligand-induced 5-HT1A receptor activation

Little experimental evidence has been reported for diverse signaling via 5-hydroxytryptamine (5-HT)1A receptors despite the fact that agonists seem to be more efficacious at dorsal raphe somatodendritic 5-HT1A autoreceptors than at postsynaptic 5-HT1A receptors. The present study investigated Ca2+ responses in Chinese hamster ovary (CHO)-K1 cells expressing a human 5-HT1A receptor by 5-HT, prototypical 5-HT1A agonists, N-(3-chloro-4-fluorobenzoyl)-4-fluoro-4-[(5-methyl-6-; methylaminopyridin-2-yl)-methylaminomethyl]-piperidine (F 14679), and especially N-(3-chloro-4-fluorobenzoyl)-4-fluoro-4-[(5-methylpyridin-2-yl)-; methylaminomethyl]piperidine (F 13640) as representative ligands of a new chemical class (methylamino-pyridine) that combines both high efficacy and selectivity for 5-HT1A receptors. 5-HT (pEC50 = 6.70 +/- 0.02) induced a pertussis toxin-sensitive, transient high-magnitude Ca2+ response. High-magnitude Ca2+ responses (Emax, percentage versus 5-HT) were also found with F 13640 (107 +/- 4), 5-carboxamidotryptamine (100 +/- 3), and F 14679 (87 +/- 3). In contrast, the prototypical 5-HT1A receptor agonists buspirone, ipsapirone, and 8-(hydroxy-2-(di-n-propylamino)tetralin, and also flesinoxan and eptapirone, were virtually inactive (< or =5). This atypical pattern of 5-HT1A receptor activation contrasts with the broad spectrum of the ligands' partial agonist properties as observed by measuring guanosine 5'-O-(3-[35 S]thio)triphosphate ([35S]GTPgammaS) binding responses with membranes of either CHO-K1 or C6-glial cells stably expressing a human 5-HT1A receptor. Remarkably, differences between ligands that seem small in the [35S]GTPgammaS binding assay translate into huge differences in the magnitude of Ca2+ responses. Therefore, some of these 5-HT1A ligands (i.e., F 13640) may in a selective way induce responses that may be not at all be achieved with other ligands (i.e., buspirone). In conclusion, the pharmacology of 5-HT1A receptor ligands seems to be codetermined by the effector pathway.

Mutation in a protein kinase C phosphorylation site of the 5-HT1A receptor preferentially attenuates Ca2+ responses to partial as opposed to higher-efficacy 5-HT1A agonists

The Thr(149)Ala mutation in a putative protein kinase C phosphorylation site of the 5-HT(1A) receptor's second intracellular loop has been shown to affect the closing of Ca(2+) channels and Ca(2+) mobilisation without interfering with the inhibitory cAMP pathway (Mol Pharmacol 52 (1997) 164). Here, the Ca(2+) responses for a series of 5-HT(1A) agonists were compared between the wild-type (wt) and mutant Thr(149)Ala 5-HT(1A) receptor as part of a fusion protein containing a G(alpha)(15) protein. Neither the mutation nor the fusion process modified the [(3)H]WAY 100635-based ligand binding profile of the fusion proteins as compared to the wt 5-HT(1A) receptor protein. Whereas at the wt 5-HT(1A) receptor, 5-HT induced a Ca(2+) response in CHO-K1 cells via endogenous G(i/o) proteins, the Ca(2+) response to 5-HT at the mutant Thr(149)Ala 5-HT(1A) receptor was fully dependent on either the co-expression or the fusion to a recombinant G(alpha)(15) protein. Buspirone, flesinoxan and 8-OH-DPAT produced a graded partial response (26 to 62%) at the wt 5-HT(1A):G(alpha)(15) fusion protein; F 13640, 5-CT and F 14679 behaved as higher-efficacy agonists with maximal Ca(2+) responses similar to 5-HT. The maximal Ca(2+) responses at the mutant Thr(149)Ala 5-HT(1A):G(alpha)(15) fusion protein were significantly attenuated for flesinoxan and 8-OH-DPAT (-45 and -36%, respectively); the response to the other 5-HT agonists was not significantly affected. A similar effect was observed upon treatment with phorbol 12-myristate 13-acetate at the Thr(149)Ala 5-HT(1A):G(alpha)(15) fusion protein. In conclusion, the amplitude of the Ca(2+) responses induced by partial, but not that to fuller 5-HT(1A) receptor agonists, is affected by the Thr(149)Ala mutation of the 5-HT(1A):G(alpha)(15) fusion protein.

Endogenous RGS proteins facilitate dopamine D(2S) receptor coupling to G(alphao) proteins and Ca2+ responses in CHO-K1 cells

The role of RGS proteins on dopaminergic D2S receptor (D2SR) signalling was investigated in Chinese hamster ovary (CHO)-K1 cells, using recombinant RGS protein- and PTX-insensitive G alphao proteins. Dopamine-mediated [35S]GTPgammaS binding was attenuated by more than 60% in CHO-K1 D2SR cells coexpressing a RGS protein- and PTX-insensitive G(alphao)Gly184Ser:Cys351Ile protein versus cells coexpressing a similar amount of PTX-insensitive G alphaoCys351Ile protein. Dopamine-agonist-mediated Ca2+ responses were dependent on the coexpression with a G alphao Cys351Ile protein and were fully abolished upon coexpression with a G alphaoGly184Ser:Cys351Ile protein. These results suggest that interactions between the G alphao protein and RGS proteins are involved in efficient D2SR signalling.

Evidence for protean agonism of RX 831003 at alpha 2A-adrenoceptors by co-expression with different G alpha protein subunits

Intrinsic properties of alpha(2) AR ligands were investigated by measuring two distinct signalling pathways via the alpha(2A) AR protein in CHO-K1 cells: (i) a Ca(2+) response mediated by a promiscuous G(alpha 15) protein; and (ii) a pertussis toxin-resistant [(35)S]GTP gamma S binding response mediated by a G(alpha o)Cys(351)Ile protein. The dexefaroxan analogue RX 831003 was virtually without intrinsic activity at the wt alpha(2A) AR via a G(alpha 15) protein, but induced a partial positive Ca(2+) response [pEC(50): 7.79 (0.17), E(max): 38+/-1% vs (-)-adrenaline] at the mutant Thr(373L)ys alpha(2A) AR. RX 831003 displayed a similar potency (pIC(50): 7.68 (0.21) for both the wt (E(max): -18+/-4%) and Thr(373)Lys alpha(2A) AR (E(max): -19+/-4%) inhibition of basal [(35)S]GTP gamma S binding via a G(alpha o)Cys(351)Ile protein. These data indicate that the alpha(2) AR ligand RX 831003 behaves as a protean agonist at the alpha(2A) AR and that its activity is highly dependent on the co-expressed G(alpha) protein subunit.

Sign-reversal during persistent activation in mu-opioid signal transduction

A concept of signal transduction in biological systems specifies that any instantaneous input is appreciated by its departure from the moving average of past activity. The concept provides an adequate account of the occurrence of both the one-directional (e.g. analgesic) effects induced by opioid receptor activation, and of the contra-directional (e.g. hyperalgesic) effects that can be observed when activation is discontinued. Following this transduction concept, the numerical simulations reported here revealed, remarkably, that under some parametric conditions, the input's effect may reverse even as input is maintained at a constant magnitude. In in vitro conditions that are proximal to the signal transduction that occurs when an opioid agonist binds to the G-protein coupled opioid receptor, the effects of opioid receptor activation were monitored by measuring time-dependent Ca(2+) responses in CHO-K1 cells transfected with a mu-opioid receptor and G(alpha 15) protein. The results indicate morphine to produce an initial increase in intracellular Ca(2+) concentration followed by a decrease below basal level. The occurrence of a sign-reversal was confirmed in native conditions of receptor-to-G protein coupling; the continuous in vivo infusion over a 2-week period of 0.31 mg rat(-1)day(-1) of fentanyl initially caused an increase of the mechanical threshold to induce a pain response (i.e. analgesia) that was followed by a decrease (i.e. hyperalgesia). The findings indicate that with opioid signaling systems, transduction mechanisms operate that may cause the sign of the effect to reverse not only when activation is discontinued but also whilst it is maintained at a constant magnitude.

Reciprocal mutations of highly conserved residues in transmembrane helices 2 and 7 of the alpha(2A)-adrenoceptor restore agonist activation of G(i1)alpha

Fusion proteins were constructed between the alpha(2A)-adrenoceptor and the alpha-subunit of the G-protein G(i1). Mutation of the highly conserved Asp(79) in transmembrane (TM) helix 2 of the receptor to Asn reduced the capacity of agonists to activate G(i1)alpha by 95% without altering [3H]antagonist or agonist ligand-binding affinity. A reciprocal mutation in TM helix 7 (Asn(422)Asp) was without effect on signalling effectiveness. Combination of these two mutations overcame the effect of the Asp(79)Asp mutation. By examining alterations in this helix 2-helix 7 microdomain, we further demonstrate the utility of receptor-G-protein fusion proteins to quantitate mutational effects on receptor-G-protein interactions and information transfer.

Differential signalling of both wild-type and Thr(343)Arg dopamine D(2short) receptor by partial agonists in a G-protein-dependent manner

G-protein activation and Ca(2+) responses by the wild-type D(2short) receptor and a mutation Thr(343)Arg, in the distal BBXXB motif of its third intracellular loop, were investigated in CHO-K1 cells in terms of ligand:receptor:G-protein interactions. No evidence was obtained for constitutive, agonist-independent receptor activation, but differences in the ligand-mediated activation profiles of both the wild-type and mutant Thr(343)Arg D(2short) receptor were observed. Most of the partial agonists, but not bromocriptine, displayed an enhanced response at the mutant D(2short) receptor, suggesting that the mutation brings the receptor in a more active state. This enhancement was apparent both at the level of G-protein activation ([35S]GTPgammaS binding) and at the effector (Ca(2+) response) and occurred with different G(alpha)-proteins. Partial agonists were also found to act differently via the wild-type D(2short) receptor depending on the involved G(alpha)-protein. Compared with higher efficacy agonists, partial agonists displayed Ca(2+) responses with slower and dissimilar kinetic properties. Lisuride and in particular bromocriptine produced a more potent response in the co-presence of a G(alphao) protein instead of a chimeric G(alphaq/o)- or a promiscuous G(alpha15)-protein. S(+)-propylnorapomorphine showed a similar partial response irrespective of the combined G(alpha)-protein. Bromerguride and (+)-UH 232 induced weak (16 to 21% versus dopamine) intrinsic activity in the co-presence of a G(alphaq/o)-protein in contrast to their silent properties with a G(alpha15)- or a G(alphao)Cys(351)Ile-protein. In conclusion, the present data strongly suggest that multiple activation binding sites are involved with these ligands at the D(2short) receptor, and that their activation may be unravelled by either the mutation or co-expressed G(alpha)-proteins being investigated.

Real-time analysis of dopamine: antagonist interactions at recombinant human D2long receptor upon modulation of its activation state

1. Antipsychotic drugs may mediate their therapeutic effects not only by preventing the binding of dopamine but also by decreasing the propensity of the dopamine receptor to assume an active R* state. Ligand-mediated activation and blockade of the recombinant human D(2long) receptor was investigated in CHO-K1 cells upon modulation of its R* state. 2. Both the Ala(371)Lys (A371K) and Thr(372)Arg (T372R) D2long receptor mutants could be activated in a ligand-dependent manner via a chimeric G(alphaq/o) protein, and more efficaciously so than with the promiscuous G(alpha15) protein. 3. Dopamine and partial agonists (E(max): lisuride >> (+)-UH 232 approximately bromerguride) displayed dissimilar Ca(2+) kinetic properties at wild-type and mutant receptors. A371K and T372R D2long receptor mutants demonstrated an attenuated and enhanced maximal response to these partial agonists, respectively. 4. Dopamine antagonists were unable to block the transient high-magnitude Ca(2+) phase at the wild-type D2long receptor upon simultaneous exposure to antagonist and dopamine, while full blockade of the low-magnitude Ca(2+) phase did occur at a later time (onset-time: haloperidol < bromerguride < (+)-butaclamol). A similar, though more efficacious, antagonist profile was also found at the A371K mutant receptor. Conversely, the blockade of the low-magnitude Ca(2+) phase was attenuated (haloperidol) or almost absent [(+)-butaclamol and bromerguride] at the T372R mutant receptor. 5. In conclusion, mutagenesis of the Ala(371) and Thr(372) positions affects in an opposite way the ligand-dependent activation and blockade of the D2long receptor. The observed attenuation of dopamine-mediated Ca(2+) signal generation with different decay-times may underlie distinct properties of the dopaminergic ligands.

Analytical pharmacology of G protein-coupled receptors by stoichiometric expression of the receptor and G(alpha) protein subunits

The description of a new family of recombinant proteins, which are constructed by the covalent fusion of the cDNA encoding a G protein-coupled receptor with that of a G(alpha) protein subunit, has recently been introduced as an original strategy to explore receptor pharmacology under defined experimental conditions. As such, a controlled 1:1 stoichiometry of receptor and G(alpha) protein expression can be achieved, as well as a forced spatial proximity to each other. Fusion proteins have been revealed as active at the receptor ligand binding level and functional at the G(alpha) protein and effector level. Insights on analytical pharmacological data are discussed for wild-type and mutant receptors interacting with a given G(alpha) protein subunit and different subtypes of either wild-type or mutant G(alpha) proteins activated by a single receptor subtype. A possible alteration of the receptor:G(alpha) protein selectivity may occur due either to the spatial proximity of both protein partners or to a constraint receptor state unable to accommodate to different G(alpha) protein states. Coactivation of endogenous G(alpha) proteins in host cells expressing a fusion protein has also been observed, but depends mainly on the coupling efficiency of the receptor and G(alpha) protein engaged in the fusion process. The ligand's apparent intrinsic activity has been shown to be either enhanced, attenuated, or unmodified when the functional responses of a fusion protein are compared to the coexpression of both fusion protein partners.

Diverse signalling by 5-hydroxytryptamine (5-HT) receptors

Fourteen different receptor subtypes might be regarded as a diversity that is sufficient to accommodate the wide-ranging physiological roles of 5-hydroxytryptamine (5-HT). However, it is becoming clear that, for 5-HT as for other neurotransmitters, the concept of a receptor as a gatekeeper for a specific cellular process or event is too restrictive. Multiple receptor-mediated biochemical cascades can be activated in cells in response to an agonist by a number of mechanisms. Whereas it is well established that different agonists do not necessarily elicit the same magnitude of response, they probably also select between various possible signal transduction pathways. Receptor signalling may be diverse via a single receptor subtype as a consequence of specific agonist-receptor-G protein interactions. 5-HT receptors are even more heterogeneous when one considers that the amino acid sequence of these receptor subtypes may vary from individual to individual, and that there is an increasing number of receptor isoforms due to alternative splicing and RNA editing of 5-HT receptor transcripts. Activation, in particular constitutive, agonist-independent activation, of some of these receptor isoforms has been reported to be altered. This implies that ligands with similar binding affinities may display different pharmacological properties (partial agonist, antagonist, or inverse agonist) versus these receptor isoforms, depending on their activation state. Therefore, intervention with receptor ligands to modify hampered neurotransmission pathways is a difficult task, and one needs to consider the growing evidence of diversity in G protein-coupled receptor signalling.

Partial to complete antagonism by putative antagonists at the wild-type alpha(2C)-adrenoceptor based on kinetic analyses of agonist:antagonist interactions

1. Activation of the recombinant human alpha(2C)-adrenoceptor (alpha(2C) AR) by (-)-adrenaline in CHO-K1 cells transiently co-expressing a chimeric G(alpha q/i1) protein induced a rapid, transient Ca(2+) response with a high-magnitude followed by a low-magnitude phase which continued throughout the recorded time period (15 min). 2. Activation of the alpha(2C) AR by various alpha(2) AR agonists revealed the following rank order of high-magnitude Ca(2+) response [E(max) (%) versus 10 microM (-)-adrenaline]: UK 14304 (102+/-4)=talipexole (101+/-3)=(-)-adrenaline (100)=d-medetomidine (98+/-1)>oxymetazoline (81+/-4) reverse similarclonidine (75+/-5). 3. The methoxy- (RX 821002) and ethoxy-derivatives (RX 811059) of idazoxan and the dexefaroxan analogue atipamezole were fully effective as antagonists of both the high- and the low-magnitude Ca(2+) response. However, though acting as full antagonists of the high-magnitude response, the further putative alpha(2) AR antagonists idazoxan (27%), SKF 86466 (29%) and dexefaroxan (59%) reversed the low-magnitude response only partially. 4. In conclusion, kinetic analyses of agonist : antagonist interactions at the alpha(2C) AR demonstrate a wide spectrum of partial to complete antagonism of the low-magnitude Ca(2+) response for structurally related alpha(2) AR ligands.

Disparate ligand-mediated Ca(2+) responses by wild-type, mutant Ser(200)Ala and Ser(204)Ala alpha(2A)-adrenoceptor: G(alpha15) fusion proteins: evidence for multiple ligand-activation binding sites

Ligand : receptor interactions were analysed at wt, mutant Ser(200)Ala and Ser(204)Ala alpha(2A) ARs by measuring Ca(2+) responses in CHO-K1 cells either by co-expression with a G(alpha15) protein or at a receptor : G(alpha15) protein stoichiometry of 1.0 using fusion proteins. The magnitude of the UK 14304-mediated Ca(2+) response as elicited by a G(alpha15) protein was largest with both mutant Ser(200)Ala and Ser(204)Ala alpha(2A)ARs compared to the wt alpha(2A) AR in the co-expression and fusion protein experiments. The activation profiles of the wt and both mutant alpha(2A) ARs as analysed by a series of alpha(2) AR agonists differed. d-Medetomidine and clonidine appeared most efficacious at the Ser(204)Ala alpha(2A) AR, whereas oxymetazoline was also partially active at the Ser(200)Ala alpha(2A) AR. Talipexole was silent at both mutant alpha(2A) ARs. The intrinsic activity of (-)-adrenaline was either absent or partial at the Ser(204)Ala and Ser(200)Ala alpha(2A) AR, respectively. This latter observation is related to its lower binding affinity for both mutant alpha(2A) ARs. Ligands characterized as antagonists at wt and Ser(200)Ala alpha(2A) ARs demonstrated either no intrinsic activity (i.e., RX 811059) or positive efficacy with a different rank order of maximal response at the Ser(204)Ala alpha(2A) AR (atipamezole=SKF 86466=idazoxan>dexefaroxan) than Asp(79)Asn alpha(2A) AR (atipamezole>idazoxan approximately SKF 86466>dexefaroxan) and Thr(373)Lys alpha(2A) AR (SKF 86466>atipamezole approximately idazoxan>dexefaroxan). These effects were only observed in the co-expression experiments at concentrations in line with their binding affinities. In conclusion, these Ca(2+) data suggest that multiple activation binding sites exist for these ligands at the alpha(2A) AR, and that their activation may be affected in different ways by the mutations being investigated.