7TMRs (seven-transmembrane receptors) exist as conformational collections in which different conformations would lead to differential downstream behaviours such as receptor phosphorylation, G-protein activation and receptor internalization. In this context, a ligand may cause differential activation of some, but not all, of the signalling events, which are associated to a particular receptor, and it would lead to biased agonism. The aim of the present study was to investigate whether H2R (histamine H2 receptor) ligands, described as inverse agonists because of their negative efficacy at modulating adenylate cyclase, could display some positive efficacy concerning receptor desensitization, internalization or even signalling through an adenylate-cyclase-independent pathway. Our present findings indicate that treatment with H2R inverse agonists leads to receptor internalization in HEK (human embryonic kidney)-293T transfected cells, by a mechanism mediated by arrestin and dynamin, but independent of GRK2 (G-protein-coupled receptor kinase 2)-mediated phosphorylation. On the other hand, we prove that two of the H2R inverse agonists tested, ranitidine and tiotidine, also induce receptor desensitization. Finally, we show that these ligands are able to display positive efficacy towards the ERK1/2 (extracellular-signal-regulated kinase 1/2) pathway by a mechanism that involves Gβγ and PI3K (phosphoinositide 3-kinase)-mediated signalling in both transfected HEK-293T cells and human gastric adenocarcinoma cells. These results point to the aspect of pluridimensional efficacy at H2R as a phenomenon that could be extended to naïve cells, and challenge previous classification of pharmacologically relevant histaminergic ligands.
7TMRs (seven-transmembrane receptors) represent the largest family of cell-surface receptors. They mediate signalling across the plasma membrane of a wide variety of stimuli. The canonical linear sequence of 7TMR signalling begins when an extracellular stimulus binds and switches the receptor from an inactive state to an active state conformation which in turn leads to activation of its coupled heterotrimeric G-protein that dissociates from the receptor, and both the Gα subunit and the Gβγ dimer modulate the activity of an effector enzyme or ion channel. In the classical models, signalling by 7TMR is terminated by receptor phosphorylation, principally mediated by GRKs (G-protein-coupled receptor kinases); arrestin binding to the phosphorylated receptor, which leads to uncoupling from G-protein and consequent receptor desensitization; and finally internalization by clathrin-coated vesicles .
Since 7TMRs are key regulators of almost every known function of eukaryotic cells, they have emerged as the most commonly targeted receptors for human therapeutics. At present, they represent the target of approximately 27% of all FDA (U.S. Food and Drug Administration)-approved drugs [2,3]. It is often assumed that treatment with agonists leads to tachyphylaxis, but 7TMR desensitization is not considered when antagonists or inverse agonists are used. This fact mainly results from the traditional classification of ligands which considers that 7TMRs represent switches that alternate between ‘off’ and ‘on’ states. Therefore agonists and inverse agonists promote an active or inactive state of receptors, whereas antagonists block the cellular signalling coupled to the receptor. However, increasing evidence indicates 7TMRs exist as conformational collections where each conformation promotes different downstream effects such as receptor phosphorylation, G-protein activation or receptor internalization among others. In this context, ligand binding stabilizes the different conformations through a process known as conformational selection . A consequence is biased agonism, where a ligand is able to cause differential activation of some signalling events associated with a particular receptor, resulting in differential activation of specific signal transduction pathways . This functional selectivity suggests that it is the ligand–receptor complex which governs the ultimate downstream signalling event and not the receptor itself . Therefore it is likely that an antagonist or an inverse agonist stabilizes a particular conformation that fails to stimulate G-protein signalling, but results in desensitization, internalization or even G-protein-independent signalling. Since Galandrin and Bouvier proposed this pluridimensional characteristic of efficacy in 2006 , the term efficacy is understood in the context of the different behaviours modulated by a 7TMR . Biased agonism has been mainly studied regarding adrenergic receptors, but little is known about other ligands used therapeutically [9,10]. Cimetidine and ranitidine are used clinically to control gastric acid secretion and rank among the most widely prescribed and over-the-counter drugs in the world [11,12]. Therefore, in the present study, we sought to establish whether widely used H2R (histamine H2 receptor) ligands classified as inverse agonists for their negative efficacy in modulating adenylate cyclase display positive efficacy regarding receptor desensitization, internalization or adenylate-cyclase-independent signalling.
In the present study, we show that ranitidine and tiotidine, but not cimetidine, induced H2R desensitization in transfected HEK (human embryonic kidney)-293T cells mediated by arrestin, clathrin and dynamin leading to receptor down-regulation. Furthermore, the three inverse agonists induced ERK1/2 (extracellular-signal-regulated kinase 1/2) activation not only in HEK-293T cells, but also in human gastric adenocarcinoma cells as histamine and the specific H2R agonist amthamine.
The present study shows that the H2R ligands classified as inverse agonists display positive efficacy regarding receptor desensitization and internalization as well as MAPK (mitogen-activated protein kinase) activation. Our findings may have relevant clinical implications given that some of these H2R ligands are used clinically in long-term treatment so they may explain therapeutic differences and side effects of histaminergic ligands.
Cell culture medium, antibiotics, IBMX (isobutylmethylxanthine), cAMP, BSA, cycloheximide, amthamine, cimetidine, ranitidine, tyrphostin AG1478 and pertussis toxin were obtained from Sigma Chemical Company. Tiotidine and LY294002 were from Tocris Cookson. [3H]cAMP and [3H]tiotidine were purchased from PerkinElmer Life Sciences. FBS was from Natocor. Other chemicals used were of analytical grade and were obtained from standard sources.
pcDNA3-β1arrestin (arrestin 2), pcDNA3-β2arrestin (arrestin 3), pcDNA3-HA-dynaminK44A, pcDNA3-β1-arrestin-(319–418) and pcDNA3-GRK2-K220R were gifts from Dr J. Benovic (Microbiology and Immunology Department, Kimmel Cancer Center, Thomas Jefferson University, Philadelphia, PA, U.S.A.). pEGFP-C2-Eps15 EH29 construct was a gift from Dr A. Benmerah (Departement de Maladies Infectieuses, Institut Cochin, Université Paris 5, Paris, France). GRK2, GRK3, GRK5 and GRK6 cDNAs were subcloned into the pCEFL vector (pCEFLGRK2, pCEFLGRK3, pCEFLGRK5 and pCEFLGRK6) as described previously . pCEFLHA-H2R was generated previously in our laboratory . pCEFL-Gα transducin was kindly provided by Dr S. Gutkind (Oral and Pharyngeal Cancer Branch, National Institutes of Health, Bethesda, MD, U.S.A.). The plasmid containing the PH (pleckstrin homology) domain of GRK2 was constructed by PCR amplification of bovine GRK2. To obtain the PH construct, the sequence coding for residues 553–651 was amplified and inserted into an EcoRI/Xba site of pCEFL-HA.
HEK-293T and AGS (human gastric cancer) cells were cultured in DMEM (Dulbecco's modified Eagle's medium) and F12K (Kaighn's modification of Ham's F12 medium) respectively, supplemented with 10% FBS and 5 μg/ml gentamicin at 37°C in a humidified atmosphere containing 5% CO2.
For transient transfection of HEK-293T cells, cells were grown to 80–90% confluence. cDNA constructs were transfected into cells using Lipofectamine™ 2000 (Invitrogen). The transfection protocol was optimized as recommended by the supplier. Assays were performed 48 h after transfection, and the expression of the constructs was confirmed by immunoblotting using specific antibodies.
For concentration-response assays, cells were incubated for 3 min in basal culture medium supplemented with 1 mM IBMX at 37°C, followed by 9 min of exposure to different concentrations of ligands.
For desensitization assays, cells were pre-treated with 10 μM H2R ligands in the absence of IBMX for different periods of time as shown in the Figures. Cells were then washed and resuspended in fresh medium containing 1 mM IBMX, incubated for 3 min, and exposed to 10 μM amthamine or 100 μM histamine for 9 min to determine whether the system was able to generate a cAMP response.
In all experiments, the reaction was stopped by the addition of ethanol followed by centrifugation at 2000 g for 5 min. The ethanol phase was then dried, and the residue was resuspended in 50 mM Tris/HCl (pH 7.4) and 0.1% BSA. cAMP content was determined by competition of [3H]cAMP for PKA (protein kinase A), as described previously .
Saturation binding experiments were carried out by incubating the cells for 40 min with increasing concentrations of [3H]tiotidine, ranging from 0.4 to 240 nM in the absence or presence of 1 μM unlabelled tiotidine. The incubation was stopped by dilution with 3 ml of ice-cold 50 mM Tris/HCl (pH 7.4) and the bound fraction was collected in 200 μl of ethanol. Experiments on intact cells were carried out at 4°C to avoid ligand internalization. The kinetic studies performed with 2 nM [3H]tiotidine at 4°C showed that equilibrium was reached at 30 min and persisted for 4 h (results not shown).
Receptor internalization and recovery
HEK-293T cells were incubated for different times with 10 μM inverse agonists and the number of receptor sites was analysed by radioligand-binding assay. The recovery of binding sites was evaluated by saturation binding assays at 60 min after washing the cells exposed previously to 10 μM inverse agonists for 90 min. In assays performed with 50 μM cycloheximide, the inhibitor was added 30 min before ligand treatment.
Western blot assays
For Western blot assays, cells were lysed in 50 mM Tris/HCl (pH 6.8), 2% SDS, 100 mM 2-mercaptoethanol, 10% (v/v) glycerol and 0.05% Bromophenol Blue and sonicated to shear DNA. Total cell lysates were resolved by SDS/PAGE (12% gel), blotted and incubated with primary antibodies against dynamin, GRK2, GRK3, GRK5, GRK6, ERK1/2, pERK, actin (Santa Cruz Biotechnology), arrestin (BD Biosciences Pharmingen) or GFP (Invitrogen), followed by horseradish-peroxidase-conjugated anti-rabbit or anti-mouse (Santa Cruz Biotechnology) and developed by ECL (GE Healthcare) following the manufacturer's instructions. Films were scanned and quantified using Scion Image® software (NIH).
Statistical analysis was performed from at least three independent experiments. Binding data, sigmoidal dose-response, desensitization fittings and comparison of best fit values according to extra-sum of squares F test were performed with GraphPad Prism 5.00 for Windows. One-way ANOVA followed by the Dunnett's post-test was performed using GraphPad InStat version 3.01. Specific binding was calculated by subtraction of non-specific binding from total binding. Statistics of densitometric Western blot analysis were carried out by one-way ANOVA or Student's t test followed by Dunnett's or Tukey's multiple comparison post-test performed with GraphPad Prism 5.00 for Windows.
H2R inverse-agonist-induced H2R internalization
The effect of cimetidine, ranitidine and tiotidine on cAMP accumulation was evaluated in concentration-response assays performed in transiently H2R-transfected HEK-293T cells. As reported previously [15,16], all ligands inhibited amthamine-induced cAMP accumulation and reduced cAMP levels in a concentration-dependent fashion, an effect not blocked by pertussis toxin pre-treatment (Figure 1). These results confirm the negative efficacy of the ligands regarding cAMP regulation.
Negative efficacy of cimetidine, ranitidine and tiotidine
We next evaluated H2R internalization in transfected HEK-293T cells as cell-surface receptor loss following cimetidine, ranitidine or tiotidine exposure. In saturation binding assays, we found that, despite the absence of positive signalling through the Gs/adenylate cyclase/cAMP pathway in the presence of the ligands, treatment with ranitidine and tiotidine led to significant H2R internalization (Figure 2A). The extent and rate of H2R internalization was similar to that of amthamine-promoted endocytosis (Figure 2B). On the other hand, cimetidine-treated cells displayed considerably slower kinetics, showing a modest reduction in H2R membrane sites which was significant following 2 h of ligand exposure (Figure 2).
Inverse-agonist-induced H2R internalization
We have reported previously that amthamine treatment evokes H2R internalization into clathrin-coated vesicles by GRK2 activation as well as arrestin 3, dynamin and Eps15 . In order to study the machinery involved in inverse-agonist-induced H2R internalization, cells were co-transfected with the H2R and dominant-negative mutants of arrestin (DNArr), dynamin (DNDyn), Eps15 (EH29) or GRK2 (DNGRK2). Saturation binding assays were performed after 90 min of exposure to ranitidine or tiotidine, given that at this time significant H2R internalization was achieved. The proper expression of these constructs in HEK-293T co-transfected cells was confirmed by Western blot analysis (Figure 3A).
Mechanisms involved in H2R inverse-agonist-induced internalization
DNDyn, DNArr or EH29 abolished H2R internalization suggesting that the process is dependent on dynamin and arrestin, as well as the correct assembly of clathrin-coated vesicles through Eps15. Surprisingly, DNGRK2 expression failed to induce ranitidine- or tiotidine-evoked H2R internalization, revealing that GRK2-mediated H2R phosphorylation is not necessary for inverse-agonist-promoted H2R internalization (Figure 3B).
Following endocytosis, 7TMRs may be either recycled to the plasma membrane or sorted for lysosomal degradation. To determine the fate of H2R sites after internalization, we evaluated the recovery of membrane H2R sites in the presence or absence of the well-characterized protein synthesis inhibitor cycloheximide. Cells were exposed to ranitidine and tiotidine for 90 min to assess H2R sites after cell washing and incubation for 60 min in fresh medium. The removal of the stimulus led to a rapid recovery of the number of H2R-binding sites that was abolished by cycloheximide (Figure 4). Conversely, cycloheximide treatment did not completely dampen the recovery of internalized H2Rs after amthamine treatment (Figure 4, inset). These findings indicate that the presence of H2R sites in the plasma membrane following the removal of the inverse agonists represents de novo H2R protein synthesis and not H2R recycling , suggesting that ranitidine and tiotidine induce H2R down-regulation.
H2R internalization and recovery in the presence of cycloheximide
H2R inverse agonists induce receptor desensitization
In order to evaluate whether these ligands displayed H2R desensitization, H2R-transfected HEK-293T cells were exposed to 10 μM amthamine, cimetidine, ranitidine or tiotidine for different times. After carefully washing, cells were re-challenged with the agonist, and the cAMP response was evaluated. The cAMP response evoked by amthamine in cells exposed previously to the inverse agonists was 41±7% for ranitidine and 36±9% for tiotidine, whereas cimetidine did not modify the H2R response to the agonist. The extent of receptor desensitization appeared to be ligand-dependent, since amthamine induced a 90% H2R desensitization (Figure 5A). It is worth noting that similar results were obtained in response to the endogenous ligand histamine (Figure 5B).
Inverse agonists promoted desensitization
As GRK2 and GRK3 were described previously to mediate H2R homologous desensitization , we next evaluated whether GRKs were involved in inverse-agonist-induced H2R desensitization. For that purpose, HEK-293T cells were co-transfected with the H2R and each of the most ubiquitous members of the GRK family of proteins (GRK2, GRK3, GRK5 and GRK6). The overexpression of GRKs was evaluated by Western blot analysis (Figure 6A) and had no significant effect on inverse agonist-mediated H2R desensitization (Figure 6B). Our findings reveal that the GRK family is not involved in the mechanism by which H2Rs lose their ability to respond to amthamine following inverse agonist exposure.
Characterization of H2R desensitization promoted by inverse agonists
In an attempt to determine whether H2R internalization mediated inverse-agonist-induced H2R desensitization, desensitization assays with HEK-293T cells co-transfected with dominant-negative mutants shown to block H2R endocytosis (Figure 2B), as well as with wild-type arrestin 2 or arrestin 3 were carried out. Neither dominant-negative mutants nor arrestin overexpression had a drastic effect on H2R desensitization (Figure 6C), suggesting that internalization is not responsible for inverse-agonist-induced H2R desensitization, and ruling out uncoupling from G-protein by arrestin.
Modulation of ERK1/2 by H2R inverse agonists
Diverse 7TMRs activate MAPK pathways through G-protein-dependent or -independent mechanisms involving Gβγ activity, 7TMR internalization, arrestin and/or dynamin recruitment or even EGFR (epidermal growth factor receptor) transactivation . Therefore we explored whether H2R ligands with negative efficacy at modulating the adenylate cyclase pathway displayed positive efficacy concerning MAPK modulation. As shown in Figures 7(A) and 7(B), pERK levels were increased by all ligands, being the maximum activation observed at 5 min following ligand treatment. Given that amthamine and histamine increase cAMP levels and cimetidine, ranitidine and tiotidine diminish them (Figure 7C), these findings show that, regardless of the efficacy towards the adenylate cyclase pathway, all ligands displayed positive efficacy with respect to ERK1/2 modulation, thus behaving as agonists concerning this signalling pathway.
ERK phosphorylation in response to H2R ligands
Arrestin and/or dynamin recruitment by diverse 7TMRs leads to MAPK activation , so we evaluated their role in ERK1/2 modulation. HEK-293T cells were co-transfected with H2R and DNDyn or DNArr and stimulated with H2R ligands. Results showed that arrestin did not mediate ERK1/2 modulation by any of the ligands assayed. On the other hand, only amthamine-induced ERK activation was mediated by dynamin (Figure 8). These findings show that, although the other ligands induced ERK activation, the mechanism involved is different from that of amthamine. It is worth noting that AG1478, the pharmacological inhibitor of EGFR, had no effect on ERK1/2 activation by these ligands (results not shown), ruling out the possibility that EGFR transactivation might be responsible for ERK1/2 activation as described previously for other 7TMRs [20,21]. Gβγ signalling was found previously to stimulate the PI3K (phosphoinositide 3-kinase) signalling pathway, leading to downstream activation of ERK1/2 . In order to evaluate the potential implication of Gβγ in MAPK activation by H2R ligands, ERK1/2 modulation was evaluated in HEK-293T cells co-transfected with H2R and Gαt (Gα transducin), a widely used scavenger of Gβγ [23,24]. In the presence of Gαt, cimetidine-, ranitidine- and tiotidine-induced ERK1/2 phosphorylation was significantly inhibited (Figure 9A). These results were confirmed by the PH domain of GRK2, another well-described scavenger of Gβγ [25,26] (Figure 9B). Moreover, LY294002, an inhibitor of PI3K, also blocked cimetidine-, ranitidine- and tiotidine-induced ERK1/2 phosphorylation (Figure 9C). In addition, the increase in pERK levels following amthamine treatment was not dampened by Gαt, confirming the central role of dynamin on amthamine-induced ERK1/2 activation (Figure 9D). Overall, these results show that ligands described previously as H2R inverse agonists regarding the Gs/adenylate cyclase/cAMP pathway display positive efficacy towards ERK1/2 through a pathway involving the Gβγ dimer and PI3K activation.
Characterization of inverse-agonist-mediated ERK phospho-rylation
Gβγ involvement in inverse-agonist-mediated ERK phosphorylation
cAMP accumulation and MAPK activation by H2R ligands in human gastric adenocarcinoma AGS cells
AGS cells not only endogenously express H2Rs, but also represent a relevant model concerning H2R histaminergic ligands and their clinical use. Regarding cAMP accumulation, histamine and amthamine displayed positive efficacy, whereas cimetidine, ranitidine and tiotidine did not significantly reduce cAMP basal levels (Figure 10A). Although constitutive activity of H2Rs was consistently reported, its modulation is difficult to achieve in naïve systems . Interestingly, positive efficacy of cimetidine, ranitidine and tiotidine towards ERK1/2 modulation was also found in AGS cells where the highest response was observed at 10 min (Figure 10). Moreover, cimetidine-induced phosphorylation of ERK1/2 was similar to that stimulated by amthamine (Figures 10B and 10C). These findings show clearly that H2R ligands classically classified as antagonist or inverse agonists on the basis of cAMP modulation may induce ERK1/2 phosphorylation through an adenylate-cyclase-independent pathway not only in overexpression models, but also in AGS cells that endogenously express H2Rs.
Biased inverse agonism in human gastric AGS cells
In the context of pluridimensional efficacy where receptors exhibit diverse behaviours and ligands display different efficacies depending on the readout chosen as the receptor response, the major finding of the present study was the identification of different efficacies displayed by H2R ligands towards receptor desensitization/internalization, Gαs/adenylate cyclase and Gβγ/ERK1/2 pathways.
H2R ligands were traditionally classified according to their ability to modulate cAMP levels. Thus cimetidine, ranitidine and tiotidine were originally considered to be neutral antagonists, but when receptor constitutive activity became evident, they were reclassified as inverse agonists, with negative efficacy [15,16]. However, the diversity of effectors considered in the present study challenges this simple scheme, making it difficult to categorize these ligands according to unique efficacy terms.
In the present study, we have shown that not only amthamine, but also ranitidine and tiotidine, induced significant receptor desensitization and internalization (Figures 2 and 3). It is worth noting that attenuation mechanisms of 7TMR signalling have originally been described as adaptive processes evoked by agonists to prevent receptor overstimulation, but, despite the lack of positive cAMP response evoked by these ligands, they induced substantial internalization and desensitization of H2Rs.
Receptor desensitization triggered by ranitidine and tiotidine was independent of arrestin 2, arrestin 3 and the GRKs, whereas receptor internalization was mediated by arrestin, dynamin and clathrin, but not by GRK2 phosphorylation. Moreover, inverse-agonist-induced H2R internalization appeared to mediate receptor down-regulation rather than recycling, as observed previously for the agonist amthamine . These findings indicate that the inverse-agonist-induced receptor desensitization/internalization protein partner profile and cellular fate once the receptor is internalized is strikingly different from that observed when the processes are triggered by agonists. Although ligand efficacies are shared, the mechanisms involved differ, indicating that the receptor partners engaged in a certain pathway are strongly dependent on the ligand.
Furthermore, an additional level of selectivity was revealed when the signalling cascades leading to ERK1/2 activation were examined. Again, although all ligands evaluated displayed positive efficacy concerning ERK1/2 phosphorylation, the underlying mechanisms differed. Amthamine-stimulated ERK1/2 phosphorylation was mediated by dynamin (Figure 8), as described previously for dimaprit, another H2R agonist , whereas cimetidine, ranitidine and tiotidine led to an increase in pERK levels by a mechanism independent of dynamin, arrestin, H2R internalization or even EGFR transactivation, but mediated by Gβγ.
Intriguingly, tiotidine, ranitidine and cimetidine seemed to inactivate Gαs, diminishing cAMP levels, but stimulating the Gβγ-dependent pERK pathway. Considering the paradigm of heterotrimeric G-protein activation, an apparent discrepancy in the activation of the Gβγ dimer seems to exist, whereas the Gαs subunit remains inactive. However, the way in which G-proteins are activated and propagate the signal has been challenged. Originally, Gα was thought to be the only G-protein subunit able to govern the direct interaction with effector molecules, but Gβγ has signal transduction properties of its own . Therefore subunit dissociation produces two signal transduction molecules when heterotrimeric G-proteins are activated. However, the mechanism by which divergent signalling pathways within the cell are controlled by an unique receptor remains controversial. Thus subunit dissociation is a critical event of the signal transduction mechanisms and it is essential that the subunit dissociation hypothesis be unequivocally established. Although this hypothesis is generally accepted [29,30], there is evidence suggesting that G-protein activation occurs without dissociation and that subunit dissociation occurs without activation . Our findings indicate that G-protein heterotrimers do not need to be fully activated and that Gβγ pathway activation does not necessarily imply Gαs activity.
Overall, these results indicate that different ligands can lead to recruitment of distinct subsets of signalling effectors to activate a single pathway. This is reminiscent of other cases where distinct effectors were selectively engaged by different ligand–receptor pairs to stimulate a common downstream signalling integrator .
Nowadays, 7TMRs are thought as allosteric machines where a punctual modification in the free energy of the receptor is transmitted to the rest of the protein affecting the receptor response to the cytosolic signalling proteins. MD modelling of proteins predicts that numerous protein conformations can exist and that receptors exist in ‘ensembles’ of multiple conformations [33,34]. Under these circumstances, ligand binding alters the receptor ensemble formation, causing the stabilization of different receptor conformations with different properties and behaviours. The machinery involved in inverse-agonist-stimulated internalization was independent of GRK2-induced phosphorylation. Therefore ranitidine and tiotidine may induce or stabilize a conformation that exposes a receptor domain which recruits arrestin and the machinery involved in H2R internalization. In consequence, GRK2 phosphorylation may not be necessary to induce inverse-agonist-mediated H2R internalization. This idea suggests that the conformation induced by the H2R inverse agonist is similar to that induced by GRK2-mediated phosphorylation. Similar results concerning phosphorylation-independent antagonist-stimulated endocytosis of GPCRs were reported previously for cholecystokinin receptor . Moreover, Haribabu et al.  reported that a C-terminal truncation mutant of human CXCR4 (CXC chemokine receptor 4) could also internalize upon agonist challenge, although this mutant receptor showed no phosphorylation. In addition, Zhang et al.  demonstrated that agonist-stimulated δ-opioid receptor internalization includes both receptor-phosphorylation-dependent and -phosphorylation-independent mechanisms and both were mediated by clathrin and β-arrestins.
The H2R ligands tested in the present study displayed distinct efficacy profiles towards adenylate cyclase and MAPKs, demonstrating the existence of functional selectivity of H2R ligands. Whereas amthamine behaves as full agonist concerning both adenylate cyclase and MAPK signalling, cimetidine, ranitidine and tiotidine displayed negative and positive efficacy for these pathways. Our results show that, whereas amthamine-induced receptor conformation triggers the classical pathways involving Gαs activation and GRK2–βarrestin–dynamin–clathrin recruitment for receptor desensitization/internalization leading to ERK1/2 activation and receptor recycling to the cell membrane, ranitidine and tiotidine may stabilize a conformation that exposes a receptor domain which activates the Gβγ/ERK1/2 pathway, recruits arrestin and the machinery involved in H2R desensitization and internalization without GRK2 phosphorylation and changing the receptor fate from recycling to degradation. Conversely, cimetidine induced a conformation engaged in Gβγ/ERK1/2 activation, but not in receptor desensitization/internalization. Overall, our results show that efficacy is not linear (i.e. ligands do not facilitate all behaviours of receptors), but rather is collateral, whereby only a subset of receptor behaviours can be activated or even inactivated. There are many possible receptor-based efficacies, and ligands may have various subsets of efficacies. Similarly, it has been shown previously that β2-adrenergic, V2-vasopressin, serotonin 5-HT2C and δ-opioid receptor ligands can act as inverse agonists on the adenylate cyclase pathway, but as agonists for MAPK signalling [38–41].
The observations made in AGS cells, which endogenously express H2R, suggest that the pluridimensionality of signalling efficacies may be extended to naïve cells making our findings pharmacologically relevant. Recent data provide evidence of the existence of ligand-specific H2R conformations that explain the differences among these ligands’ affinities, potencies and efficacies observed in neutrophils and eosinophils . Nevertheless, additional studies are needed to assess the extent to which the effector-dependent signalling efficacies are detected in normal and pathophysiological conditions.
Since cimetidine and ranitidine profiles as biased ligands were described in in vitro studies, the therapeutic impact of this phenomenon in vivo cannot be inferred. These ligands are currently marketed and used to treat duodenal ulcers and prevent their recurrence. They are also used to treat gastric ulcers and Zollinger–Ellison disease . However, these ligands are used because they inhibit H2R-coupled cAMP response, but, although both induce ERK activation like histamine, only ranitidine promotes the depletion of receptors from the membrane surface. Therefore the outcome of a long-term treatment with these ligands would not be alike.
Although the signalling bias shown in the present study runs counter to classic concepts of ligand efficacy, those aspects of drug receptor mechanisms has become firmly established and has been demonstrated for many receptors . In the same way, the findings of the present study stress the relevance of studying further the different efficacies of a ligand that appears to induce an inactive state of the receptor if accurate evaluation and understanding of its pharmacological behaviour is intended.
dominant-negative mutant of arrestin
dominant-negative mutant of dynamin
dominant-negative mutant of GRK2
epidermal growth factor receptor
dominant-negative mutant of Eps15
extracellular-signal-regulated kinase 1/2
G-protein-coupled receptor kinase
human embryonic kidney
histamine H2 receptor
mitogen-activated protein kinase
Natalia Alonso carried out the experiments, analysed data and assisted the writing of the paper. Federico Monczor and Emiliana Echeverría contributed to perform experiments and analysed data. Carlos Davio and Carina Shayo contributed to analyse data and to design the experiments. Natalia Fernández planned the research, conducted some experiments and wrote the paper. All authors participated in editing the paper in its final form before submission.
We thank Dr Liliana Bianciotti, from the University of Buenos Aires, for a critical reading and correction of the paper before submission.
This work was supported by the Universidad de Buenos Aires [grant numbers UBACyT 20020090300054 and UBACyT 20020100100601], the Agencia Nacional de Promoción Científica y Tecnológica [grant numbers PICT 2010-1545, PICT 2010-0237, PICT 2010-1642 and PICT 2010-1571], and the Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET) [grant number PIP0344].