Abstract
We described previously the cDNA cloning of three functional rat histamine H3 receptor (rH3R) isoforms as well as the differential brain expression patterns of their corresponding mRNAs and signaling properties of the resulting rH3A, rH3B, and rH3C receptor isoforms (Mol Pharmacol 59:1–8). In the current report, we describe the cDNA cloning, mRNA localization in the rat central nervous system, and pharmacological characterization of three additional rH3R splice variants (rH3D, rH3E, and rH3F) that differ from the previously published isoforms in that they result from an additional alternative-splicing event. These new H3R isoforms lack the seventh transmembrane (TM) helix and contain an alternative, putatively extracellular, C terminus (6TM-rH3 isoforms). After heterologous expression in COS-7 cells, radioligand binding or functional responses upon the application of various H3R ligands could not be detected for the 6TM-rH3 isoforms. In contrast to the rH3A receptor (rH3AR), detection of the rH3D isoform using hemagglutinin antibodies revealed that the rH3D isoform remains mainly intracellular. The expression of the rH3D-F splice variants, however, modulates the cell surface expression-levels and subsequent functional responses of the 7TM H3R isoforms. Coexpression of the rH3AR and the rH3D isoforms resulted in the intracellular retention of the rH3AR and reduced rH3AR functionality. Finally, we show that in rat brain, the H3R mRNA expression levels are modulated upon treatment with the convulsant pentylenetetrazole, suggesting that the rH3R isoforms described herein thus represent a novel physiological mechanism for controlling the activity of the histaminergic system.
Histamine receptors are members of the superfamily of seven transmembrane domain (7TM) G-protein-coupled receptors (GPCRs). The histamine H3 receptor (H3R) was pharmacologically identified in 1983 and holds great promise as a target for the development of therapeutics for numerous disorders, including obesity, epilepsy, and such cognitive diseases as attention deficit hyperactivity disorder and Alzheimer's disease (see Bakker, 2004; Leurs et al., 2005 for reviews). The cloning of the H3R cDNA allowed for the subsequent cloning of related sequences, including a variety of H3R isoforms from different species (for review, see Bakker, 2004; Leurs et al., 2005).
Alternative splicing of pre-mRNA represents a widespread mechanism for increasing the variability of eukaryotic gene expression by generating structurally distinct isoforms from a single gene. Alterations in the expression of GPCR isoforms could be associated with disease (Schmauss et al., 1993). Although α1-AR adrenoceptors and dopamine receptors are prime examples of alternatively spliced GPCRs (Cogé et al., 1999; Kilpatrick et al., 1999; Hawrylyshyn et al., 2004), more and more GPCR splice variants are identified for other members of the GPCR superfamily. A variety of H3R isoforms from several species has been reported (for review, see Bakker, 2004; Leurs et al., 2005). In addition to the 445 amino acids containing rH3AR, two presumably nonfunctional truncated isoforms (reported as rH3T or rH3(nf1) and rH3(nf2)) (Drutel et al., 2001; Morisset et al., 2001) and three functional rH3R isoforms have been detected: rH3B, rH3C, and the rH3(410) receptor, generated by deletions in the third intracellular loop of the rH3R of 32, 48, and 35 amino acids, respectively. Because several H3R isoforms have been shown to possess specific pharmacological characteristics in terms of ligand-binding and initiation of signal-transductions events (Drutel et al., 2001), the H3R mRNA splicing can significantly affect cellular responses to histamine. A detailed understanding of the spectrum of H3R splice variants in different species is of importance not only for the understanding of histaminergic system, but also for future drug development efforts.
In the present study, we describe the identification of three additional 6TM-rH3 isoforms after an RT-PCR approach using rat brain cDNA. Although the mRNA for these 6TM-rH3 isoforms is detected in the rat brain, in attempting their characterization, we failed to detect radioligand binding using H3R specific radioligands as well as functional effects upon heterologous expression in COS-7 cells. Coexpression of 7TM-rH3Rs with the 6TM-rH3 isoforms, however, revealed that the 6TM-rH3 isoforms inhibit the cell surface trafficking and subsequent functional activity of the 7TM-rH3Rs. The regulation of the expression of the 6TM-rH3 isoforms may therefore represent a novel mechanism for the regulation of H3R functionality. To study possible in vivo functional relationships between 7TM-rH3R and 6TM-rH3 isoforms, relative expression levels were analyzed in a model of generalized tonic-clonic seizures induced by pentylenetetrazole (PTZ). Data from a study on kainic acid-induced status epilepticus indicates that systemic kainic acid induces a direct or indirect selective increase in H3R isoforms with a full third intracellular loop in areas that suffer rapid neuronal damage (Lintunen et al., 2005). No data are currently available regarding whether 7TM-rH3R and 6TM-rH3 isoforms are similarly regulated under pathophysiological conditions in the rat brain. The PTZ seizure model used in this study allows us to determine whether an H3R isoform-specific response occurs in a pathological setting. Our findings therefore uncover a new mechanism that may control the regulation of H3R activity in the brain.
Materials and Methods
Materials. Immepip, clobenpropit, and thioperamide were synthesized at the department of Medicinal Chemistry at the Vrije Univeristeit Amsterdam. Gifts of pcDEF3 (Dr. J. Langer, Robert Wood Johnson Medical School, Piscataway, NJ), pTLN21CRE-Luc (Dr. W. Born, National Jewish Medical and Research Center, Denver, CO), and of the cDNAs encoding the PTX-insensitive mutant rat Gαi/o proteins Gαi1C351I, Gαi2C352I, Gαi3C351I, and GαoC351I (Dr. G. Milligan, University of Glasgow, Glasgow, UK), the cDNA encoding the FLAG-tagged rH3AR (Dr. F. Cogé, Institut de Recherches Servier, Croissy sur Seine, France), the cDNA encoding the human histamine H1 receptor (Dr. H. Fukui, University of Tokushima, Tokushima, Japan), the cDNA encoding the KSHV-GPCR ORF74 (Dr. T. Schwartz, University of Copenhagen, Copenhagen, Denmark), mianserin hydrochloride (Organon NV, The Netherlands), and the HA antibody and rhodamine-labeled secondary antibody (Dr. J. van Minnen, Vrije Universiteit, Amsterdam, The Netherlands), are greatly acknowledged. All other materials were from commercial suppliers.
Constructs. The reverse transcription and PCR amplification for cloning of the rH3D-F (6TM-rH3) isoform cDNAs were performed as described previously (Drutel et al., 2001). The full-length cDNAs were isolated with primers overlapping the rat H3R cDNA sequence. The forward sequence included a Kozak sequence (underlined) (5′-CCG CCA CCA TGG AGC GCG CGC CGC CCG ACG GGC TG-3′). The reverse sequence was based on cDNA for rat orphan GPCR (Genbank accession number AB015646) (5′-CTC TAC CCC ATA ACC ACC CAC C-3′). The use of these primers resulted in the amplification of at least three different products. After cloning in pCRII-TOPO, the cDNAs were sequenced on both DNA strands and subcloned in pcDNA3. The sequence of the identified rH3F isoform is identical to one of the sequences found in the GenBank database (accession number AB015646; GI: 6681587), the sequences of the rH3D and rH3E isoforms have been deposited in the GenBank database (accession numbers DQ112342 and DQ112343, respectively). The hydropathic profile of the H3D-F isoforms was analyzed using the TMHMM Server at the Center for Biological Sequence Analysis, Technical University of Denmark, DTU, Lyngby, Denmark (http://www.cbs.dtu.dk/services/TMHMM/).
Construction of rH3R-Gα Protein Fusion Constructs. Fusion proteins between the rat H3R and PTX-insensitive mutant rat Gα-proteins of the Gi/o class were created by PCR using Turbo Pfu to remove the translation initiation codon from the Gα-protein cDNA sequence and the stop codon from the rH3AR cDNA sequence.
HA-Tagging of the rH3R Isoforms. N-Terminal hemagglutinin (HA)-tagged expression constructs of the rH3AR and rH3D isoform were generated by PCR (5′-GCC ACC ATG GGC TAC CCA TAC GAC GTC CCA GAC TAC GCC GCG GAG CGC GCG CCG C-3′) and cloned into pcDNA 3.1 (Invitrogen, Leek, The Netherlands). Construct integrity was verified by sequence analysis.
Animals. The study was conducted in accordance with the European Convention (1986) guidelines and approved by the local committee for Animal Experiments and the Provincial State Office of Western Finland and the Animal Ethics Committee of Abo Akademi University. Male Sprague-Dawley rats (260–280 g) were given PTZ (50 mg/kg, i.p.). Animals were stunned with CO2 and killed by decapitation 6 h (PTZ, n = 3), 24 h (PTZ, n = 3), and 48 h (PTZ, n = 3; saline control, n = 3) after injection.
In Situ Hybridization Histochemistry. Probes were labeled with deoxy-[α-33P]ATP (PerkinElmer Life and Analytical Sciences, Boston, MA) at their 3′ ends using terminal deoxynucleotide transferase (Promega, Madison, WI), and subsequent in situ hybridization histochemistry was performed essentially as described previously (Drutel et al., 2001). The following oligo-probe sequence was used for detecting H3DEF isoform mRNAs: 5′-AAG TTT CCC GAG GCG CTC GAC ACA GTA ATC GGG GAT GCA GCG GCC-3′.
Image Analysis and Data Interpretation. Autoradiographic films were quantified by digitizing the film images using the MCID 5+ image analysis system (Imaging Research, St. Catherines, ON, Canada) and by measuring gray scale pixel values. The relative optic density was converted to integrated optic density based on a curve derived from 14C standards exposed to films. Gray scale values were determined by using a total of four sections for each animal.
Cell Culture and Transfection. African green monkey kidney COS-7 cells were maintained and transfected as described previously (Drutel et al., 2001; Bakker et al., 2004a). HEK 293 cells were cultures under similar conditions and transfected with cDNA encoding the rH3AR using the LipofectaminePlus method according to the manufacturer's protocols.
Reporter-Gene Assay. H3R isoform-mediated modulation of cAMP mediated gene transcription activity was measured using the luciferase reporter-gene plasmid pTLNC121–3 (2.5 μg/106 cells) containing 21 cAMP-responsive elements. Luminescence was assayed 48 h after incubation of transfected cells with ligands as described previously (Bakker et al., 2004b).
[35S]GTPγS Binding Assays. Transfected COS-7 cells were resuspended in 4°C binding buffer (20 mM HEPES, 3 μM GDP, 10 mM MgCl2, and 150 mM NaOH, pH 7.4). For measurement of agonist-stimulated GTPγS binding, 6 μg of the crude cell extract was incubated in binding buffer with ligands for 15 min at 30°C after which 0.1 to 0.2 nM [35S]GTPγS (1250 Ci/mmol; PerkinElmer Life and Analytical Sciences) was added to make a final total volume of 100 μl. Bound radioactivity was separated by filtration after 15 min through Whatman GF/C filters on a Brandel cell harvester using 4°C wash buffer (20 mM HEPES and 5 mM MgCl2, pH 7.4). Radioactivity retained on the filters was measured by liquid scintillation counting.
Receptor Binding Studies. Radioligand binding studies for the H1R, H2R, and H3R using [3H]mepyramine, [125I]aminopotentidine, and [Nα-methyl-3H]histamine, respectively, were performed as described previously (Bakker et al., 2004b). The H3R radioligand binding studies using [125I]iodophenpropit (125IPP) were carried out under the same experimental conditions as for [Nα-methyl-3H]histamine. CXCL8 was labeled with 125I using the Iodogen method (Pierce, Rockford, IL) and subsequently used in ORF74 radioligand binding studies as described previously (Smit et al., 2002).
Detection of Tagged rH3Rs. In the enzyme-linked immunosorbent assays (ELISA), a mouse anti-HA monoclonal primary antibody was used as primary antibody, and a goat anti-mouse–horseradish peroxidase conjugate as secondary antibody for the detection of tagged rH3Rs in transfected cells. The 3,3′-, 5,5′-tetramethylbenzidine liquid substrate system for ELISA was used as substrate and the optical density was measured at 450 nm using a Victor2 Wallac multilabel counter (PerkinElmer Life and Analytical Sciences). The same primary antibody was used for immunocytochemistry in conjunction with a secondary rhodamine labeled goat-anti-mouse antibody. Permeabilization of cells was achieved by an optional incubation of the cells for 5 min with 0.5% Nonidet P-40 in TBS before antibody application. For imaging, coverslips were mounted in 90% glycerol containing 0.02 M Tris-HCl, pH 8.0, 0.002% NaN3, and 2% 1,4-diazabicyclo-(2,2,2)-octane (Merck, Darmstadt, Germany).
Time-Resolved FRET. The time-resolved fluorescence resonance energy transfer (FRET) experiments were conducted essentially as described previously (Bakker et al., 2004a). Energy transfer was measured by exciting the Eu3+ at 320 nm and monitoring the allophycocyanin emission for 1 ms at 665 nm using a Novostar (BMG LABTECH GmbH, Offenburg, Germany) configured for time-resolved fluorescence after a 50-μs delay.
Cross-Linking and Immunoblotting of rH3A Receptors. Cells were harvested by centrifugation, and the resulting pellet was resuspended in 150 μl of cross-linking buffer (150 mM NaCl, 100 mM Na-HEPES, 5 mM EDTA, pH 7.5, and 5 mM DTT) to give a final protein concentration of approximately 0.5 mg/ml. The samples were incubated at room temperature with continual mixing for 12 min with either a 0.12 mM or a 0.25 mM concentration of the cell permeable cross-linker bis(sulfosuccinimidyl)suberate (BS3), after which the cross-linking mixture was removed by centrifugation and the resultant pellet was used for immunoblotting as described previously (Chazot et al., 2001). Immunoblots were probed either with anti-H3C 188–197Cys antibody (Shenton et al., 2005) at a 0.2 μg/ml, or with an anti-H3 329–358 antibody used at a final protein concentration of 1.5 μg/ml (Chazot et al., 2001).
Analytical Methods. All data shown are expressed as mean ± S.E.M. Data from radioligand binding assays and functional assays data were evaluated by a nonlinear, least-squares curve-fitting procedure using Prism (GraphPad Software, Inc., San Diego, CA).
Results
Cloning of cDNAs Encoding Additional Rat H3R Isoforms
The existence of additional H3R isoforms was investigated by RT-PCR analysis of rat whole-brain total RNA using specific primers 1 and 2 (see Materials and Methods), and revealed the existence of three previously uncharacterized full-length cDNAs with putative corresponding proteins of 497 (rH3D), 465 (rH3E), and 449 (rH3F) amino acids. The amino acid sequence of the rH3F isoform corresponds to one of the sequences found in the GenBank database (accession number AB015646; GI: 6681587). These three new rH3 isoforms correspond in a large part to the sequences of rat histamine H3R isoforms A, B, and C (rH3AR, rH3BR, and rH3CR, respectively) and exhibit exactly the same differences in length for the third intracellular loop (Fig. 1A). This insertion in the third intracellular loop in the rH3D, rH3E, and rH3F isoforms seems to be created by a retention/deletion system already described for the third intracellular loop (Cogé et al., 2001; Drutel et al., 2001; Morisset et al., 2001). However, the rH3D, rH3E, and rH3F isoforms differ from the rH3AR, rH3BR, and rH3CR in their C-terminal region, in which the last C-terminal 53 amino acids, which correspond to the seventh transmembrane-domain and carboxyl terminus of the rH3AR, rH3BR, and rH3CR proteins, are replaced by a sequence of 105 amino acids that do not share any homology with the last 53 C-terminal amino acids of the rH3AR, rH3BR, and rH3CR. Sequence analysis of the rH3R gene indicates that the alternative C-terminal domain that is found in the rH3D-F isoforms is due to a change in the open reading frame upon alternative splicing using previously unidentified exon/intron junctions present within the rH3R gene (see Fig. 1B). Analysis of the hydropathic profile of the rH3D, rH3E, and rH3F isoforms does not reveal a clear putative seven transmembrane domain (Fig. 2, A and B). Therefore, the rH3D, rH3E, and rH3F isoforms are predicted to possess only six transmembrane domains 6TM-rH3 isoforms) and an extracellular C-teriminal domain (Fig. 2, C and D).
Heterologous Expression of Epitope-Tagged rH3R Isoforms. We have successfully characterized the rH3AR, rH3BR, and rH3CR using COS-7 cells heterologously expressing these receptors (Drutel et al., 2001). We therefore used the same approach in this study to characterize the three additional 6TM-rH3 isoforms described herein. To evaluate the cell surface expression of the 6TM-rH3 isoforms, we generated the cDNAs coding for the N-terminally HA-tagged rH3AR (HA-rH3AR) and the N-terminally HA-tagged rH3D (HA-rH3D) isoform by PCR. Although we detected clear immunological evidence for the cell surface expression of the HA-rH3AR with the use of an ELISA assay using anti HA-antibodies on intact cells, we can hardly detect the HA-rH3D isoforms on the cell surface on intact cells (Fig. 2E). We observe, however, a clear immunofluorescent signal upon permeabilization of cells expressing the HA-rH3D isoform (Fig. 2E), indicating successful synthesis of the HA-tagged rH3D isoform protein and retention of the HA-tagged rH3D isoform inside the cell. There is an apparent difference in detection of the HA-rH3D isoform versus the HA-rH3AR, which might indicate differences in, for example, the rate of synthesis, the inherent stability, or the rate of degradation of the H3 isoforms, but we have not pursued this issue further. To evaluate the plasma membrane localization of the HA-rH3AR and the HA-rH3D isoform, we subsequently performed immunocytochemistry studies using rhodamine labeled anti-HA antibodies. Plasma membrane localization of the rhodamine-derived fluorescence was easily detected using cells expressing HA-rH3ARs in intact cells (Fig. 2F, top left) and only a limited intracellular fluorescence was observed in Nonidet P40-permeabilized cells (Fig. 2F, top right). In contrast, corroborating the findings obtained by the ELISA studies, using cells transfected with cDNA encoding the HA-rH3D isoform, appreciable rhodamine-derived fluorescence is detected only in permeabilized cells (Fig. 2F, bottom right) and not on intact cells (Fig. 2F, bottom left). Moreover, no plasma membrane localization of the rhodamine-derived fluorescence was detected using permeabilized HA-rH3D isoform-expressing cells. These data indicate an intracellular localization for the HA-rH3D isoform.
Are the 6TM-rH3 Isoforms Functional H3Rs?
Radioligand Binding Studies. To evaluate whether the identified mRNA species for the 6TM-rH3 isoforms code for functional H3Rs, we transfected COS-7 cells with the cDNA encoding either the rH3AR or one of the 6TM-rH3 isoforms and evaluated corresponding membrane preparations for their ability to bind either the inverse H3R agonist radioligand 125IPP or the H3R agonist radioligand [Nα-methyl-3H]histamine. Cell homogenates derived from cells expressing the rH3AR bind 125IPP with high affinity (KD = 2.1 nM, Bmax = 2.5 pmol/mg of protein) and exhibits the expected affinities for IPP, immepip, and thioperamide (pKb for 125IPP, 8.2 ± 0.1; pKi values for immepip and thioperamide, 6.5 ± 0.2 and 7.7 ± 0.1, respectively). In contrast, we failed to detect specific 125IPP-binding to cells transfected with cDNAs coding for any of the 6TM-rH3 isoforms (Fig. 3A). Likewise, we did not detect [Nα-methyl-3H]histamine binding to membranes of cells transfected with cDNAs encoding the 6TM-rH3 isoforms (data not shown).
Functional Assays. The H3R couples to members of the Gi/o family of G-proteins to inhibit adenylyl cyclase activity and subsequently inhibits the formation of intracellular cAMP (Leurs et al., 2005). Cotransfection of COS-7 cells with the rH3AR encoding cDNA together with pTLN121CRE, a cAMP-responsive element-binding protein (CREB)-responsive firefly-luciferase reporter gene, allowed us to monitor H3R-induced modulation of forskolin-induced reporter-gene expression. In concert with the Gi/o-coupled nature of the H3R (see Leurs et al., 2005), treatment of cotransfected cells with varying concentrations of the H3R agonists immepip, R(α)-methylhistamine, and histamine results in the dose-dependent inhibition of 10 mM forskolin-induced firefly-luciferase expression by approximately 60% with EC50 values of approximately 2, 7, and 43 nM, respectively (Fig. 3B), in rH3AR-expressing cells. Under these assay conditions, we do not observe constitutive activity for the rH3AR because the inverse H3R agonist thioperamide (Leurs et al., 2005) is without effect on the forskolin-induced luciferase expression (Fig. 3C). Although 1 mM R(α)-methylhistamine potently induces signaling via the rH3AR, under the same assay conditions, R(α)-methylhistamine is without effect on the forskolin-induced firefly-luciferase expression in cells cotransfected with cDNAs encoding the reporter gene and either of the 6TM-rH3 isoforms (Fig. 3D).
Is There a Role for the 6TM-rH3 Isoforms?
Detection of Epitope-Tagged rH3R Isoforms in Coexpression Studies. The immunological and immunocytochemistry data obtained with the N-terminally HA-tagged HA-rH3D isoform indicates poor plasma membrane expression of the rH3D isoform (Fig. 2, E and F). In addition, in cells transfected with cDNA encoding either of the 6TM-rH3 isoforms we have been unable to detect 1) specific 125IPP or [Nα-methyl-3H]histamine binding (Fig. 3A) and 2) modulation of forskolin-induced transcriptional events that are otherwise modulated by rH3AR activation under the same conditions. In recent years, accumulating evidence suggests that some GPCRs [e.g., the GABAB receptors (White et al., 1998) and odorant receptors (Hague et al., 2004a)] require particular protein-protein interactions to allow proper plasma membrane expression. It seems that the 6TM-rH3 isoforms are retained intracellularly; we postulated, therefore, that the additional expression of GPCRs might aid the cell surface expression of the 6TM-rH3 isoforms. Conversely, the 6TM-rH3 isoforms may possibly retain the coexpressed GPCRs intracellularly by acting as dominant-negative isoforms.
Do the 6TM-rH3 Isoforms Interfere with Cell Surface Expression of 7TM-rH3Rs? We cotransfected COS-7 cells with the cDNAs coding for the HA-rH3AR and the rH3D isoform. Similar to the effects of coexpression of alternative splice variants of human α1A-adrenoceptors (Cogé et al., 1999), we found that the coexpression of the rH3D isoform reduced the expression of the HA-rH3AR at the plasma membrane. This phenomenon is observed using either an ELISA assay on intact cells or by applying immunocytochemistry techniques using a rhodamine labeled anti-HA antibody (Fig. 4, A and B, respectively).
Evaluation of Ligand-Binding Sites upon Coexpression of 7TM-H3R and 6TM-rH3 Isoforms. To evaluate whether the loss of HA-rH3AR-derived immunofluorescence at the cell surface upon coexpression of the rH3D isoform also results in a loss of ligand binding sites for H3R ligands we performed radioligand binding assays. As shown in Fig. 3A, 125IPP binding sites are detected upon the expression of 7TM-rH3Rs (Drutel et al., 2001), such as the rH3AR, whereas expression of the 6TM-rH3 isoforms does not result in the formation of 125IPP binding sites.
The transfection of 0.25 mg/106 cells of cDNA coding for the rH3AR resulted in the expression of 2.5 pmol/mg of protein 125IPP binding sites. Coexpression of the rH3AR together with the rH3D isoform, however, resulted in an rH3D-isoform gene-dosage–dependent reduction of rH3AR-derived 125IPP binding sites (Fig. 5A); the remaining 125IPP binding sites exhibit a pharmacological indistinguishable from that of the rH3AR (pKb for 125IPP, 8.1 ± 0.1; pKi values for immepip and thioperamide, 6.9 ± 0.2 and 7.6 ± 0.1, respectively). The coexpression of the rH3E or rH3F isoform together with the rH3AR resulted in a similar reduction of 125IPP binding sites (Fig. 5, B and C, respectively), indicating that the 6TM-rH3 isoforms interfere with the expression of the rH3AR. The maximal inhibition of 125IPP binding sites (as evaluated by a transfection of cells with an rH3AR/6TM-rH3 isoform cDNA ratio of 1:10) is ∼50 to 75% (Fig. 5, A–C). To evaluate whether the observed inhibition of 125IPP binding-site expression is specific for the 6TM-rH3 isoforms, we performed additional experiments in which we cotransfected cells with the same amount of rH3AR cDNA in combination with various amounts of cDNA encoding the human histamine H1 receptor (hH1R) (Fig. 5D). Although the transfection of COS-7 cells with the hH1R cDNA resulted in the formation of binding-sites for the H1R radioligand [3H]mepyramine with pharmacological characteristics of the hH1R (data not shown), no specific 125IPP binding was detected in hH1R-expressing cells (Fig. 5D). Cotransfection of cells with cDNAs encoding both the rH3AR and the hH1R, however, did not influence the formation of 125IPP binding sites.
Similar to the findings with the hH1R, expression of the rat H2R (rH2R) or the viral chemokine receptor from Karposi's sarcoma herpes virus KSHV-GPCR (also known as ORF74) in COS-7 cells resulted in the formation of binding sites for the H2R radioligand 125IAPT and 125I-CXCL8, a radioligand for ORF74, respectively, but not in the formation of 125IPP binding sites. Coexpression of the rH3AR with either the rH2R or ORF74, however, did not affect the formation of rH3AR-derived 125IPP binding sites (Fig. 5E).
We also evaluated the effects of the coexpression of the rH3BR and rH3CR with the rHE and rHF isoforms, respectively, on the formation of 125IPP binding sites. Similar to the expression of the rH3AR, expression of the rH3BRorrH3CRin COS-7 cells results in the formation of 125IPP binding sites (Drutel et al., 2001). Coexpression of the rH3BR or rH3CR together with either the rH3E or rH3F isoform resulted in an rH3E isoform and rH3F isoform gene-dosage dependent reduction of rH3BR-derived (Fig. 5F) and rH3CR-derived (Fig. 5G) 125IPP binding sites, respectively. The maximal inhibition of 125IPP binding sites, as evaluated by a transfection of cells with an rH3BR or rH3CR-to-6TM-rH3 isoform cDNA ratio of 1:10 is ∼50 to 75% (Fig. 5, F and G), similar to our findings on the coexpression of the rH3AR with the 6TM-rH3 isoforms (Fig. 5, A–C).
Although the coexpression of the rH3AR with the rH3D isoform inhibits the formation of 125IPP binding sites, the remaining 125IPP binding sites exhibit unchanged pharmacological characteristics of the rH3AR, as evidenced by its unchanged affinity for IPP (Fig. 5H). Taken together, these radioligand-binding data clearly demonstrate that the expression of the 6TM-rH3 isoforms selectively interferes with the expression of the 7TM-rH3Rs.
Evaluation of H3R Ligand-Induced [35S]GTPγS Binding upon Coexpression of 7TM-H3R and 6TM-rH3 Isoforms
Creation of rH3AR Gα Fusion Proteins. Because we failed to detect H3R-agonist mediated [35S]GTPγS binding to activated Gα proteins in cell membranes derived from 6TM-rH3 isoform expressing cells (data not shown), we chose to evaluate H3R-agonist induced [35S]GTPγS binding to assess the effects of the coexpression of 6TM-rH3 isoforms on the functionality of 7TM-rH3Rs. To assess the effects of the coexpression of 6TM-rH3 isoforms on the functionality of 7TM-rH3Rs, we chose to evaluate H3R-agonist induced [35S]GTPγS binding to activated Gα proteins. To increase the sensitivity of this assay (Milligan, 2000), we created fusion proteins consisting of the rH3AR fused to one of the PTX-insensitive mutant rat Gαi/o proteins: Gαi1C351I, Gαi2C352I, Gαi3C351I, or GαoC351I (creating rH3AR-Gαo1C351I, rH3AR-Gαi1C351I, rH3AR-Gαi2C352I, and rH3AR-Gαi3C351I, respectively) by PCR according to Materials and Methods.
Characterization of rH3AR. Gα fusion proteins—the four different rH3AR fusion proteins rH3AR-Gαo1C351I, rH3AR-Gαi1C351I, rH3AR-Gαi2C352I, and rH3AR-Gαi3C351I—were subsequently characterized by 125IPP binding assays upon their heterologous expression in COS-7 cells. Based on these studies (data not shown), we decided to continue our experiments using the rH3AR-Gαo1C351I fusion protein as this fusion protein exhibited a pKb value for 125IPP of 8.4 ± 0.2 that corresponds to the obtained pKb value of 125IPP for the wild-type rH3AR of 8.2 ± 0.1. In addition, the affinities of the wild-type rH3AR and the rH3AR-Gαo1C351I fusion protein for the H3R agonist immepip and the inverse H3R agonist thioperamide are similar (6.5 ± 0.2 and 7.7 ± 0.1 versus 6.9 ± 0.1 and 7.9 ± 0.1, respectively).
Subsequently, we compared the capability of the rH3AR-Gαo1C351I fusion protein to mediate the inhibition of 10 mM forskolin-induced activation of cAMP response element-mediated gene transcription in COS-7 cells. We found the rH3AR-Gαo1C351I fusion protein to potently inhibit the forskolin-induced response upon activation with H3R agonists. In concert with our findings on the wild-type rH3AR, treatment of cells cotransfected with cDNAs encoding the rH3AR-Gαo1C351I fusion protein and the cAMP response element-reporter gene with varying concentrations of the H3R agonists immepip, R(α)-methylhistamine, and histamine results in the dose-dependent inhibition of 10 mM forskolin-induced firefly luciferase expression by approximately 40% with EC50 values of approximately 4, 30, and 76 nM, respectively (Fig. 6A). These data indicate that the rH3AR-Gαo1C351I fusion protein is fully functional and shows an rH3AR pharmacology.
Coexpression of rH3AR-Gαo1C351I Fusion Proteins and the rH3D Isoform. Consistent with our findings on the coexpression of 7TM-rH3ARs with the 6TM-rH3 isoforms (see Fig. 5), coexpression of the rH3AR-Gαo1C351I fusion protein together with the rH3D isoform, results in an rH3D-isoform gene-dosage dependent reduction of rH3AR-Gαo1C351I fusion protein-derived 125IPP binding sites (Fig. 6B), and the remaining 125IPP binding sites exhibit an rH3AR-Gαo1C351I-like pharmacological profile (pKb for 125IPP, 8.5 ± 0.1; pKi values for immepip and thioperamide, 7.0 ± 0.2 and 7.6 ± 0.1, respectively). The maximal inhibition of rH3AR-Gαo1C351I-derived 125IPP binding sites, as evaluated by a transfection of cells with an rH3AR-Gαo1/rH3D isoform cDNA ratio of 1:10 is ∼65% (Fig. 6B).
We subsequently assessed the influence of coexpression of the rH3D isoform on the [35S]GTPγS binding induced by agonist-mediated activation of coexpressed rH3AR-Gαo1C351I fusion proteins. The H3R agonist immepip (1 mM) resulted in a robust stimulation of [35S]GTPγS binding in rH3AR-Gαo1C351I expressing cells that was inhibited by coincubation with a 1 mM concentration of the inverse H3R agonist thioperamide (Fig. 6C). Under the assay conditions used, we could not detect significant thioperamide-mediated inhibition of basal rH3AR-Gαo1C351I mediated [35S]GTPγS binding, indicating that we could not detect constitutive rH3AR-Gαo1C351I activity. The 1 mM immepip-induced [35S]GTPγS binding was inhibited by 70% by coexpression of the rH3AR-Gαo1C351I fusion protein with the rH3D isoform (Fig. 6C). The 6TM-rH3 isoforms themselves did not mediate changes in [35S]GTPγS binding upon incubation with H3R ligands (data not shown).
Can rH3Rs Form Homo-Oligomers? In view of the emerging concept of GPCR dimerization that is now well documented in literature (for review, see, for example, Pfleger and Eidne, 2005), we speculated that the 6TM-rH3 isoforms might interfere with the cell surface expression of 7TM-rH3 receptors through dimerization.
We have described the generation of anti-rH3C 268–277Cys antibodies (Shenton et al., 2005), which, based on immunoblotting, selectively recognize both the rH3AR and rH3CR isoform, but not the rH3BR isoform expressed in HEK 293 cells. Wild-type rH3AR-expressing cells were collected and subjected to cross-linking with varying amounts of the cell permeable cross-linker bis(sulfosuccinimidyl)suberate (BS3) before resolving the samples using SDS-PAGE. Immunoblotting with anti-H3C 268–277Cys antibody yielded three major protein species (Mr 90,000, 135,000, and ∼200,000, respectively), corresponding well to putative dimeric, trimeric and tetrameric rH3AR oligomers, respectively (Fig. 7, lanes 1–3). Performing similar experiments using native rat brain tissue, the major species observed is a coincident Mr 90,000 species (Fig. 7, lane 4). It is noteworthy that a recombinant putative monomeric Mr 47,000 species was clearly observed, which was barely detectable in the native forebrain preparation. Higher cross-linker concentrations yielded >200,000 species for both recombinant and native H3R preparations (Shenton et al., 2005).
We have successfully used the time-resolved FRET (tr-FRET) fluorescence (665-nm emission after excitation at 320 nm) for the detection of hH1R dimerization using epitope-tagged hH1Rs and fluorescently labeled antibodies recognizing the N-terminally epitope-tagged receptors (Bakker et al., 2004a). We have used this approach to confirm the formation of oligomerization of rH3Rs.
tr-FRET fluorescence results obtained with the different samples are shown in Fig. 7B. A clear specific tr-FRET signal is observed using live cells expressing the HA-rH3ARs. The data are presented as the tr-FRET that is observed using HA-rH3AR-expressing cells that have been incubated with both anti-HA-Eu3+ and anti-HA-allophycocyanin antibodies versus the tr-FRET that is observed using a mix of two populations of HA-rH3AR-expressing cells that before mixing were independently incubated with either of the two antibodies. The increased tr-FRET signal can only be explained by the resonance energy transfer from anti-HA-Eu3+ antibodies bound to HA-rH3ARs to anti-HA-allophycocyanin antibodies bound to HA-rH3ARs, indicative of the formation of rH3AR multimers in living cells.
Modulation of rH3AD and rH3DEF Isoform-Specific mRNAs in Rat Brain after Delivery of a Systemic Convulsant. We have successfully used specific oligonucleotide probes to characterize the 7TM-rH3R mRNA expression in the rat brain (Drutel et al., 2001). To evaluate the CNS expression of the 6TM-rH3 isoforms, we have designed domain specific probes. We have used one probe specific for the C terminus present in the 6TM-rH3 isoforms, and, for comparison, we have also performed studies using a oligonucleotide probe specific for the (full-length) third intracellular loop of the rH3AR, which is also present in the rH3D isoform (but not any of the other rH3 isoforms identified to date; Fig. 8A).
Significant increases in mRNA expression levels of H3R isoforms with full-length third intracellular loop (detected using probe H3AD) were observed in layers II-VIb of cortex (48 h after injection), caudate putamen (48 h after injection), piriform cortex (48 h after injection), and CA1 region of the hippocampus (24 h after injection) (Fig. 8B) after PTZ. Figure 8C illustrates mRNA expression levels and differences for H3A and H3D isoforms in representative sections from control and 48 h after injection animals.
In contrast, decreases in mRNA expression pattern of 6TM-rH3 isoforms (detected using probe H3DEF) were observed in layers II-VIb of cortex (6 h after injection), piriform cortex (6, 24, and 48 after injection), CA1 region of the hippocampus (24 h after injection), and CA3 region of the hippocampus (24 h after injection) (Fig. 8D). Figure 8E shows mRNA expression patterns and differences for 6TM-rH3 isoforms in representative sections from control animals and animals 24 h after injection.
Discussion
Our search for additional alternative splice variants of the rH3R resulted in the identification of three mRNAs coding for heretofore uncharacterized rH3 isoforms. In contrast to the known functional H3Rs, sequence analysis of these newly identified isoforms reveals that an additional splicing event occurs within a region corresponding to TM6, resulting in isoforms with an alternative C-terminal domain and isoforms that are predicted to possess only 6 TMs. The mRNAs encoding these 6TM-rH3 isoforms are expressed in the rat brain with a distribution pattern that is similar to the previously identified functional 7TM-rH3Rs (Drutel et al., 2001). Although there may be differences in the rate of synthesis, the inherent stability, and the rate of degradation of the various H3 isoforms, resulting in differences in their expression levels, extensive analysis of the 6TM-rH3 isoforms through heterologous expression studies, however, failed to identify any ligand-binding and functional signaling events modulated by these 6TM-rH3 isoforms. The 6TM-rH3 isoforms seem to be localized intracellularly, and succeeding studies revealed the ability of the 6TM-rH3 isoforms to interfere with the cell surface expression and subsequent signaling of the previously identified 7TM-rH3 isoforms, thus acting as dominant-negative isoforms.
An increasing number of GPCRs have been shown to exist as oligomeric complexes (for review, see, for example, Pfleger and Eidne, 2005), which are thought to be formed early during biosynthesis (Terrillon et al., 2003). In this study, we show also that the rH3AR is present as dimers or higher order oligomeric complexes in both transfected cells and rat brain. The oligomerization of GPCRs during biosynthesis and maturation seems crucial for proper exportation of receptors to the plasma membrane (for review, see Bulenger et al., 2005). The coexpression and formation of heterodimeric β2-ARs (Hague et al., 2004a) aids, for instance, the cell surface trafficking of olfactory GPCRs that are otherwise retained and degraded in the ER (Lu et al., 2003, 2004), as well as of the α1D-AR that normally is trafficked poorly to the cell surface (Uberti et al., 2005). The coexpression of differentially spliced GPCR variants may also aid the cell-surface expression, as shown for instance by the coexpression of the α1D-AR with α1B-ARs (Hague et al., 2004b). In contrast, certain alternatively spliced variants of, for instance, the α1A-AR (Cogé et al., 1999), the calcitonin receptor (Seck et al., 2003), and the dopamine D3 receptor (Karpa et al., 2000) seem to dimerize with their cognate full-length receptors and impede their cell surface expression because of mislocalization to an intracellular compartment. The coexpression of the 6TM-rH3 isoforms described herein not only reduced cell surface expression of the 7TM-rH3Rs but also consequently resulted in a reduced 7TM-rH3R-mediated signaling. Our data are therefore consistent with the reported findings on the coexpression of N-terminal truncated, dominant-negative mutants and wild-type V2 receptors, which results in the formation of heterodimers and reduced agonist binding, signal transduction, and cell-surface trafficking of the full-length V2 receptor (Zhu and Wess, 1998). Similar to our findings on the 6TM-rH3 isoforms, mutants of the α2-AR (Zhou et al., 2006), vasopressin V2 (Zhu and Wess, 1998), dopamine D2 (Lee et al., 2000), chemokine receptor CCR5 (Benkirane et al., 1997; Blanpain et al., 2000; Chelli and Alizon, 2001), gonadotropin-releasing hormone receptor (Brothers et al., 2004), and the platelet-activating factor (Le Gouill et al., 1999) receptors also impede the cell surface expression of their coexpressed wild-type counterparts, thus exhibiting trans-dominant-negative effects on wild-type receptor expression (Benkirane et al., 1997; Chelli and Alizon, 2001; Brothers et al., 2004), most likely through dimerization. It seems most likely that the 6TM-rH3 isoforms interfere with the functional expression of the 7TM-rH3Rs through heterodimerization. Although direct heterodimerization of the 6TM-rH3 isoforms with the 7TM-rH3Rs remains to be established, we have shown the rH3AR to be constitutively expressed as a dimeric receptor in the brain as well as upon heterologous expression.
GPCRs have been found to interact with accessory proteins, which can be critical for their biogenesis (for review, see Bermak and Zhou, 2001; Metherell et al., 2005). A conserved F(X)6LL motif, which is suggested to be important for the proper GPCR folding and subsequent export from the ER (Duvernay et al., 2004), is also present in the C-terminal domain of the 7TM-rH3Rs but is lacking in the 6TM-rH3 isoforms. Mutations in regions overlapping the F(X)6LL motif in the dopamine D1 receptor resulted in ER retention of the mutant receptor and loss of cell surface expression, and subsequent studies revealed that this region is important for a interactions with a specific ER-membrane-associated protein that regulates transport of GPCRs (Bermak et al., 2001). It is noteworthy that, instead of the F(X)6LL motif, the 6TM-rH3 isoforms possess an RXR ER retention signal. Taken together, these data suggest that the 6TM-rH3 isoforms lack protein-protein interactions with specific accessory proteins in the ER that are required for cell-surface expression. Although the localization of the 6TM-rH3 isoforms within the ER seems likely, this needs to be verified by future experiments. Nonetheless, the findings on GPR30 as an intracellular GPCR (Revankar et al., 2005) points out that the 6TM-rH3 isoforms may well have yet undiscovered intracellular functions in addition to their capability to retain the 7TM-rH3Rs intracellularly.
On the one hand, the 6TM-rH3 splice variants may act as “antichaperones” inhibiting specific chaperones' activities or preventing their access to the 7TM-rH3Rs. On the other hand, the association of the 6TM-rH3 isoforms with the 7TM-rH3R isoforms in the ER may actively unfold or result in the misfolding of the protein complex. Because we find that the rH3D isoform does not interfere with the cell surface expression of unrelated GPCRs, it seems unlikely the 6TM-rH3 isoforms act through blocking either the ER or Golgi, or to promote ER-associated protein degradation. The precise mechanism underlying the action of the newly identified isoforms, however, remains unknown. Our data suggest that the regulation of the alternative rH3R mRNA splicing is a new and effective means for the regulation of H3R signaling. Functional (including constitutive) H3R activity may be regulated through the regulation of the splicing events underlying the occurrence of the various H3R isoforms. It is noteworthy that several cell signaling pathways, including the MAPK pathway, regulate mRNA splicing (reviewed in Shin and Manley, 2004). It is intriguing that the H3R activates the MAPK pathway (Drutel et al., 2001; Giovannini et al., 2003), arguing for the possibility of activation of splicing factors and hence, an autoregulation of the H3R activity.
Our studies also show that the expression pattern of the 7TM-rH3Rs and the 6TM-rH3 isoforms overlaps substantially in rat brain. Moreover, PTZ-induced seizures result in suppression of 6TM-rH3 (probe H3DEF) isoform mRNAs in particular brain regions, whereas mRNA levels of the isoforms with the full third intracellular loop (probe H3AD) are increased. A characteristic transient and short-living increase in the mRNA for the full-length third intracellular loop (probe H3AD), hippocampal CA3c area, followed by piriform cortex, amygdala, and hippocampal CA1 area is observed after systemic injection of kainic acid, a model of temporal lobe epilepsy (Lintunen et al., 2005), indicating a spatiotemporal correlation to progressing neuronal damage in this model of temporal epilepsy. Previous studies on PTZ have indicated damaged neurons in the rostral limbic cortex (both the orbital, agranular insular, and prelimbic), the lateral hypothalamus (in the vicinity of the rostral medial forebrain bundle), the bed nucleus of the stria terminalis, the claustrum, the hippocampal formation (CA3 and entorhinal cortex), and lateral thalamic nuclei 50 min after injection (Ben-Ari et al., 1981). The increases in H3R mRNA with full third intracellular loop as observed in this study after PTZ were observed significantly later than the reported damage begins (Ben-Ari et al., 1981), in agreement with the concept that the mechanisms of neuronal damage after kainic acid and PTZ are different. Although the piriform cortex has not been reported to suffer significant damage after PTZ (Ben-Ari et al., 1981), it seems not to be only a primary sensory area but because of its neuronal organization and associative fiber system may also be involved in the pathological mechanisms leading to seizures (Löscher and Ebert, 1996). Of the areas studied here, the piriform cortex showed a sustained decline in 6TM-rH3 mRNA expression. We observed a significant transient increase in H3R radioligand binding in piriform cortex at 6 h after PTZ concomitantly with the decline of H3DEF isoform mRNAs (data not shown). However, a similar strong correlation was not found in all areas where smaller changes were seen, suggesting that other factors in addition to mRNA ratios may affect receptor binding. The high susceptibility for induction of seizures by chemical or electrical stimulation and various studies addressing its role in seizure generation suggest that the piriform cortex can also function as an amplifier region to increase and propagate seizure activity induced in other limbic regions (Löscher and Ebert, 1996). Increased H3R activity (e.g., expression and translation) in this region could result in decreased glutamate release (Brown and Haas, 1999; Molina-Hernandez et al., 2001), because glutamatergic neurons exist in the piriform cortex (Riba-Bosch and Perez-Clausell, 2004), for control of overall neuronal activity in the region.
The abundance of mRNA coding for the 6TM-rH3 isoforms suggest that its production may have important biological implications. For example, in addition to its ability to interfere with cell surface expression of functional rH3Rs, which may arise from modification of the stability of the mRNA encoding functional 7TM-rH3Rs, the 6TM-rH3 isoform mRNAs might encode proteins with yet unidentified functions. Further analysis of the alternatively spliced products of the H3R gene is required to elucidate their biological significance.
In conclusion, we have identified three additional splice variants of the rH3R. The mRNAs of these isoforms are abundantly expressed in the brain and the expression pattern largely overlaps with that of the known rH3A-CR isoforms. Analysis of the sequence of these rH3D, rH3E, and rH3F isoforms reveals these isoforms to consist of 6TM domains. The 6TM-rH3 isoforms are retained intracellularly upon heterologous expression, and in subsequent pharmacological analysis studies we could detect no ligand binding or functional activity for these 6TM-rH3 isoforms. The 6TM-rH3 isoforms, however, selectively impede cell surface expression of the functional 7TM-rH3Rs. Moreover, the mRNA levels of the rH3 isoforms in rat brain are modulated by treatment with the convulsant PTZ. Although the functional significance and possible roles of these 6TM-rH3 isoforms in (patho)physiology remain to be established, these findings provide novel insight in the regulation of the histaminergic system in the brain.
Acknowledgments
This article is dedicated to Dr. Art A. Hancock, who passed away on November 11, 2005. Dr. Hancock was a great scientist with a warm and generous personality. Under his inspired leadership, Abbott Laboratories made many seminal contributions to the field of H3 receptors.
We thank Sarina M. Meusburger, Franca di Summa, Kim Retra, and José Antonio Arias-Montaño for expert assistance.
Footnotes
-
R.L. is the recipient of a PIONIER award of the Technologiestichting Stichting Technische Wetenschappen of the Nederlandse Organisatie voor Wetenschappelijk Onderzoek. Supported by the Academy of Finland (P.P.), and the Finnish Foundation for Alcohol Studies (P.P., M.L., and A.F.L.).
-
Article, publication date, and citation information can be found at http://molpharm.aspetjournals.org.
-
doi:10.1124/mol.105.019299.
-
ABBREVIATIONS: 7TM, seven transmembrane domain; GPCR, G-protein-coupled receptor; H3R, histamine H3 receptor; PTZ, pentylenetetrazole; HA, hemagglutinin; HEK, human embryonic kidney; GTPγS, guanosine 5′-O-(3-thio)triphosphate; APT, aminopotentidine; 125IPP, [125I]iodophenpropit; ELISA, enzyme-linked immunosorbent assay; FRET, fluorescence resonance energy transfer; BS3, bis(sulfosuccinimidyl)suberate; PCR, polymerase chain reaction; CREB, cAMP-responsive element-binding protein; tr-FRET, time-resolved fluorescence resonance energy transfer; ER, endoplasmic reticulum; 7TM-rH3R isoforms, the rH3A, rH3B, and rH3C receptors; 6TM-rH3R isoforms, the rH3D, rH3E, and rH3F isoforms.
-
↵1 Current affiliation: Department of Metabolic Diseases, Boehringer Ingelheim Pharma GmbH, Biberach, Germany.
-
↵2 Current affiliation: Institut François Magendie, Bordeaux, France.
-
↵3 Current affiliation: Galapagos Genomics BV, Leiden, The Netherlands.
- Received September 26, 2005.
- Accepted December 21, 2005.
- The American Society for Pharmacology and Experimental Therapeutics