- Email: [email protected]
The intra-molecular activation mechanisms of the dimeric metabotropic glutamate receptor 1 differ depending on the type of G proteins
Neuropharmacology 61 (2011) 832e841
Contents lists available at ScienceDirect
Neuropharmacology journal homepage: www.elsevier.com/locate/neuropharm
The intra-molecular activation mechanisms of the dimeric metabotropic glutamate receptor 1 differ depending on the type of G proteins Michihiro Tateyama*, Yoshihiro Kubo Division of Biophysics and Neurobiology, Department of Molecular Physiology, National Institute for Physiological Sciences, Myodaiji, Okazaki 444-8585, Japan
a r t i c l e i n f o
a b s t r a c t
Article history: Received 7 January 2011 Received in revised form 19 May 2011 Accepted 25 May 2011
Metabotropic glutamate receptor 1 (mGlu1) functions as a homodimer and activates not only the Gq but also the Gi/o and Gs pathways. Because of the dimeric configuration, different pathways could be activated either through the glutamate-bound subunit (cis-activation) and/or the other one (trans-activation). We here examined whether the intra-molecular activation mechanisms in the mGlu1 differ depending on the type of coupling G proteins, using various combinations of mGlu1 constructs that lack glutamate binding and/or G-protein coupling. The cis- or trans-activation alone was confirmed to trigger the Gq-coupled intracellular Ca2þ transient. In contrast, the Gi/o-coupled G protein-dependent inward rectifying potassium (GIRK) channels were not activated either through the cis- or trans-activation alone. When one subunit of dimeric mGlu1 lacked the G-protein coupling, a significant decrease in the glutamate-induced GIRK current density was also observed. As the G protein-coupling-deficient subunit did not decrease the cell surface expression of mGlu1 and the Gq-coupled Ca2þ transient, it was suggested that the coupling deficiency in one subunit of mGlu1 attenuates the Gi/o but not Gq coupling. In conclusion, multiple G-protein signaling was differentially activated by different intra-molecular activation mechanisms of the dimeric mGlu1. Ó 2011 Elsevier Ltd. All rights reserved.
Keywords: Metabotropic glutamate receptor Dual signaling TRPC3 GIRK
1. Introduction Metabotropic glutamate receptors (mGlu) are the key molecules for neuronal transmission and signaling (Nicoletti et al., 2010). mGlu are G protein-coupled receptors (GPCRs) that belong to family C, to which g-aminobutyric acid type B (GABAB) receptor also belongs. The family C GPCRs are dimers and have a large extracellular domain (ECD) where the ligand recognition sites locate (Nicoletti et al., 2010; Pin et al., 2003). The dimeric and/or oligomeric configuration has also been also reported in family A
Abbreviations: mGlu, metabotropic glutamate receptor; GPCR, G proteincoupled receptor; GABAB, g-aminobutyric acid type B; ECD, extracellular domain; 7TMD, 7 trans-membrane domain; ICD, intracellular domain; GIRK, G proteindependent inward rectifying potassium; TRPC3, transient receptor potential canonical 3; C1, C-terminal tail of GB1; C2KKTN, C-terminal tail of GB2 with KKTN; YFP, yellow fluorescent protein; HEK293T, human embryonic kidney 293T; CHO, Chinese hamster ovary; [Ca2þ]i, intracellular Ca2þ concentration; D[Ca2þ]i, maximal increase in the intracellular Ca2þ concentration; FRET, fluorescent resonance energy transfer; [cAMP]i, intracellular cAMP levels; PTX, pertussis toxin; [glutamate], glutamate concentration; A.U., arbitrary units; WT, wild type; FL-WT-C2, FLAGtagged WT-C2KKTN. * Corresponding author. Tel.: þ81 564 55 7823; fax: þ81 564 55 7825. E-mail addresses: [email protected] (M. Tateyama), [email protected] (Y. Kubo). 0028-3908/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.neuropharm.2011.05.031
GPCRs, but the formation of oligomers is not necessary for their signaling (Park et al., 2004). In contrast, the dimeric configuration is requisite for family C GPCRs to function as the signaling molecules (Pin et al., 2003). The ligand binding that occurs at the ECD should be transmitted to their 7 trans-membrane domain (7TMD) for activation of the G proteins. X-ray crystallographic analyses of mGlu have revealed that a glutamate stabilizes the closed conformation of one ECD, which has been suggested to cause dimeric rearrangement of the ECDs (Kunishima et al., 2000; Muto et al., 2007). The dimeric rearrangement was demonstrated in the intracellular domains (ICDs) (Tateyama et al., 2004), with some conformational changes of the 7TMD (Brock et al., 2007; Yamashita et al., 2008). The ligand-induced dimeric rearrangement has also been demonstrated in GABAB receptors (Matsushita et al., 2010), suggesting that the dimeric rearrangement is an important mechanism for the family C GPCRs to activate G proteins. Because of the dimeric configuration of the family C GPCR, G proteins might be activated via the ligand-bound subunit (cis-activation) and/or the other one (trans-activation). Alternatively, ligand binding on two subunits might be requisite for the G-protein activation. In the case of homodimeric mGlu5, the Gq pathway can be triggered either through the cis- or trans-activation alone (Brock et al., 2007). In addition, the heterodimeric GABAB receptor is accepted to activate the Gi/o proteins through trans-activation, since the receptor
M. Tateyama, Y. Kubo / Neuropharmacology 61 (2011) 832e841
consists of a ligand-sensing subunit of GB1 and a Gi/o coupling subunit of GB2 (Galvez et al., 2001; Margeta-Mitrovic et al., 2001). A subtype of mGlu, mGlu1a, is known to activate not only the Gq pathway but also the Gs and Gi/o pathways, which may be responsible for the diverse responses in neuronal cells (Kammermeier, 2009; Kitano et al., 2003; Sugiyama et al., 2008). The multiple signaling via mGlu1a is dependent on the mGlu1a expression system (Selkirk et al., 2001) and differentially affected by mutations or phosphorylation of the ICD of mGlu1a (Francesconi and Duvoisin, 1998, 2000). In addition, a difference in the activated conformation of mGlu1a has been suggested to alter the coupling profile of mGlu1 (Sheffler and Conn, 2008; Tateyama and Kubo, 2006). From these studies, it could be assumed that the difference in the intramolecular activation mechanisms, such as cis- or trans-activation, might differentially affect the Gs or Gi/o coupling of mGlu1a. In the present study, we confirmed that the Gq pathway via mGlu1 was activated either through the cis- or the trans-activation alone (Brock et al., 2007) and newly found that the Gi/o pathway was not triggered either through the cis- or the trans-activation alone. Similarly, the Gs pathway was not triggered through the trans-activation alone. We further investigated the effects of glutamate-binding or G protein-coupling deficiency in one subunit of mGlu1 on the multiple signaling. 2. Materials and methods 2.1. Constructs and expression system Point mutants of rat mGlu1a, mGlu2 and G protein-dependent inward rectifying potassium type 2 (GIRK2) channel were constructed by PCR with mutated primers and KOD plus ver. 2 Taq polymerase (Toyobo, Osaka, Japan). Transient receptor potential canonical type 3 (TRPC3) channel was isolated from a rat brain cDNA library by PCR. Fragments of the C-terminal tail of GB1 (C1) or GB2 with a retention signal KKTN (C2KKTN) were amplified by PCR with designed primer sets and then inserted after Arg857 using a SphI site of the mGlu1a constructs (Brock et al., 2007). The FLAG epitope was inserted into the N-terminal region of mGlu1-C2KKTN after the signal peptide sequence, as previously reported (Abe et al., 2003). The coding sequence of yellow fluorescent protein (YFP) was inserted between Ile685 and Leu686 at the 2nd intracellular loop of mGlu1-C1, i2-YFP-C1, by blunt-end ligation (Tateyama et al., 2004). DNA sequences of the constructs were confirmed and then subcloned into pCXN2 expression vector except for S196A-GIRK2, a mutant GIRK2 which is resistant to the inhibition by protein kinase C-dependent phosphorylation (Mao et al., 2004). The S196A-GIRK2 was subcloned into pIRES2-GFP (Clontech, Mountain View, CA, USA). Human embryonic kidney 293T (HEK293T) and Chinese hamster ovary (CHO) cells were transfected with the plasmid DNA using LipofectAMINE2000 (Invitrogen, Carlsbad, CA, USA) and seeded onto cover glasses. Experiments using HEK293T cells were carried out 24e48 h after transfection and those using CHO cells were done after 48e72 h. Before electrophysiological experiments or imaging, cells were incubated for more than 30 min in Hank’s balanced salt solution (Invitrogen) supplemented with 1 mM Ca2þ and 0.3 mM Mg2þ at room temperature.
833
2.3. Electrophysiology The Gi/o coupling of mGlu constructs was evaluated as the amplitude of the glutamate-induced inward currents through the GIRK2 channels. The Gq coupling of mGlu1 constructs was evaluated as the amplitude of the glutamate-induced inward current through the La3þ sensitive TRPC3 channels (Hartmann et al., 2008). Macroscopic currents were recorded by the whole cell patch clamp technique at room temperature (Tateyama and Kubo, 2008). The internal solution for recording of GIRK currents contained (in mM) 120 KCl, 5 K2-ATP, 10 NaCl, 3 EGTA, 10 HEPES, 0.1 CaCl2, 4 MgCl2, and 0.3 mM GTP (pH adjusted to 7.4 with KOH). The internal solution for recording of TRPC3 currents contained (in mM) 140 KCl, 4 Na2-ATP, 0.3 EGTA, 10 HEPES, and 5 MgCl2 (pH adjusted to 7.4 with KOH). The composition of the bath solution was the same as that for imaging. 140 mM Naþ ions were replaced with 140 mM Kþ ions for recording of GIRK currents and the Ca2þ ions were replaced with Ba2þ ions for recording of TRPC3 currents. After the glutamate-induced current reached a maximum, the mean amplitudes for 0.5 s were measured at a holding potential of 80 mV. The current amplitude induced by 1 mM glutamate was normalized by the cell capacitance to calculate the current density. To evaluate the glutamate sensitivity of the responses, the current amplitude evoked by various concentrations of glutamate ([glutamate]) was normalized to that evoked by 1 mM glutamate in each cell. The relationship between [glutamate] and GIRK current was fitted to a logistic curve (Origin8; OriginLab, Northampton, MA, USA). The voltagedependence of the glutamate-induced current was evaluated by applying a ramp protocol, 120 mV to 40 mV at a rate of 0.4 mV/ms, at 0.2 Hz. The current traces shown in figures were off-line smoothed by averaging 10 adjacent points (Origin 8) to clearly demonstrate the glutamate-induced changes. 2.4. Immunocytochemistry HEK293T cells were transfected with the same amount of cDNAs (1 mg) of mGlu1 constructs and S196A-GIRK2. Twenty-four hours after transfection, the immunocytochemical experiments were carried out, as previously reported (Tateyama and Kubo, 2008). The cells were fixed with 2% paraformaldehyde for 2 min at 4 C without Triton-X, which permeabilizes membrane, and then incubated with antibodies (1st antibody, FLAG-antibody clone M2, 4 mg/ml, SigmaeAldrich, Tokyo, Japan; 2nd antibody, Alexa-564-conjugated anti-mouse IgG, Invitrogen). Fluorescence images were captured with a BX-50 microscope (Olympus, Tokyo, Japan) equipped with an AxioCam CCD camera unit (Carl Zeiss, Jena, Germany) and analyzed with Aquacosmos software (Hamamatsu Photonics, Hamamatsu, Japan). To exclude fluorescence signals of Alexa-564 from the intracellular compartment, we first selected cells expressing GFP (excitation at 470e495 nm and emission at 510e550 nm, more than 6 arbitrary units (A.U.)) to determine the region for measurement. The fluorescence intensity of Alexa-564 (excitation at 520e550 nm and emission more than 580 nm) was measured at each pixel within the selected region. The intensity was normalized by the area to compare the surface expression level of mGlu1 constructs in each group. 2.5. Analysis and statistics All data are expressed as the means S.E., with n indicating the number of data. A statistical significance was estimated by unpaired Student’s t-test or Dunnet’s t-test; values of p < 0.05 were considered statistically significant.
3. Results 3.1. Trans-activation alone triggers Gq but not Gi/o or Gs
2.2. Monitoring of the changes in [Ca2þ]i and [cAMP]i Changes in the intracellular Ca2þ concentration ([Ca2þ]i) in HEK293T cells were monitored every 3 s by using fura2-AM (Invitrogen), and the ratio of the fluorescence intensity (excitation 340 nm/380 nm, emission 535 nm) was calculated. Cells loaded with the fura2-AM were placed in a recording chamber filled with 250 mL of the bath solution containing (in mM) 140 NaCl, 1 CaCl2, 4 KCl, 0.3 MgCl2, and 10 HEPES (pH adjusted to 7.4 with NaOH). 250 mL of twice-concentrated glutamate stock solved in the bath solution was applied to the recording chamber by pipetting. Peak amplitudes of the ratio after glutamate application were obtained from cells expressing YFP, a transfection marker, and the maximal increase in the [Ca2þ]i (D[Ca2þ]i) was calculated by subtraction of the basal amplitude measured 0e30 s before glutamate application. To evaluate the Gs and Gq coupling of mGlu1 constructs in CHO cells, we simultaneously monitored the efficiency of fluorescence resonance energy transfer (FRET) of ICUE2, a probe for cAMP concentration (DiPilato et al., 2004; Dunn et al., 2006) and the emission ratio of indo-1 (excitation 340 nm, emission 405 nm/ 470 nm; Invitrogen) every 3 s. The intracellular cAMP concentration ([cAMP]i) was evaluated as the reversed ratio of FRET (excitation 436 nm, emission 470 nm/ 535 nm) and then the reversed ratios were normalized to the basal values (averaged value for 30 s before glutamate application). Cells that showed either D[Ca2þ]i > 0.05 or [cAMP]i > 1.05 were defined as glutamate-responsive cells for the analyses.
We first confirmed that the Gq pathway could be triggered through the trans-activation of mGlu1 alone, by using glutamatebinding-deficient R78L- and G protein-coupling-deficient F781SmGlu1a (Francesconi and Duvoisin, 1998; Jensen et al., 2000) (Fig. 1A). Homo-multimers of R78L and F781S could not increase the [Ca2þ]i upon 1 mM glutamate application, although they were well expressed on the plasma membrane (Fig. S1). In contrast, the hetero-multimer formed by co-expression of R78L- and F781SmGlu1a did increase the [Ca2þ]i (Fig. 1B and D). These results indicated that the Gq pathway can be triggered through the transactivation alone, because only the trans-activation pathway is available in the combination of R78L and F781S (Fig. 1A). Then the Gi/o coupling of the hetero-multimer was investigated by recording inward currents through GIRK2 channels in HEK293T cells, since mGlu1a activates the Pertussis toxin (PTX)-sensitive Gi/ o pathway in HEK293T cells (Kitano et al., 2003; Tateyama and Kubo, 2008). Activation of wild-type (WT) mGlu1a increased the
834
M. Tateyama, Y. Kubo / Neuropharmacology 61 (2011) 832e841
Fig. 1. The Gq pathway, but not Gi/o or Gs, was activated through trans-activation alone of mGlu1a. (A) Schematic drawings of the combinations of mGlu1a constructs. (B) Timelapse changes in [Ca2þ]i (upper traces) and GIRK current (lower traces) in HEK293T cells. [Ca2þ]i is represented as the ratio of fluorescence intensity (fura-2AM). The application timing of glutamate (1 mM) is indicated by arrows (upper panels) and black bars on the current traces (lower panels). (C) Averages of response amplitudes of mGlu1a constructs in HEK293T cells. Upper panel: Bars represent the maximal increases in [Ca2þ]i, (D[Ca2þ]i, n ¼ 66e117 cells). Lower panel: Bars represent the glutamate-induced GIRK current densities (n ¼ 4e10). (D) Time-lapse changes in [Ca2þ]i (upper panels) and [cAMP]i (lower panels) in CHO cells. [Ca2þ]i and [cAMP]i were simultaneously monitored in each cell. [Ca2þ]i is represented as the averaged ratio of the fluorescence intensity (indo-1) and [cAMP]i is represented as the normalized reciprocal FRET value (cAMP probe). The application timing of glutamate (1 mM) is indicated by arrows. (E) Averages of response amplitudes of mGlu1a constructs in CHO cells. Bars represent the glutamate-induced increase in D[Ca2þ]i (upper panel) and [cAMP]i (lower panel, averaged ratio between 230 and 250 s after glutamate application) in glutamate-positive cells (n ¼ 33e42 cells). *: p < 0.05 vs. WT; n.s.: not significant. The bars at right (C and E) represent the baseline fluctuation.
amplitude of GIRK current, whereas activation of the heteromultimer did not increase the current amplitude (Fig. 1B, lower traces). The glutamate-induced GIRK current density in cells transfected with R78L- and F781S-mGlu1a was significantly smaller than that in cells with WT-mGu1a (Fig. 1C). Gs coupling of the hetero-multimer was also analyzed by monitoring the
glutamate-induced increases in the [cAMP]i in CHO cells (Tateyama and Kubo, 2006). The activation of WT-mGlu1a increased the [cAMP]i, while activation of the hetero-multimer failed to increase it (Fig. 1D and E). These results clearly demonstrated that neither the Gi/o nor the Gs pathway could be triggered through the transactivation alone.
M. Tateyama, Y. Kubo / Neuropharmacology 61 (2011) 832e841
3.2. Cis-activation alone triggers Gq but not Gi/o Next we examined whether or not the multiple signaling pathways via mGlu1a are triggered through the cis-activation alone. For this purpose, the C-terminus of mGlu1a was replaced with those of GB1 (C1) and GB2 with an additional retention signal (C2KKTN). The endogenous and artificially added retention signals of C1 and C2KKTN have been reported to inhibit the surface expression of homo-multimeric mGlu-C1 and mGlu-C2KKTN, respectively (Brock et al., 2007). In the hetero-multimers, the retention signals would be masked by the C1eC2 interaction and thereby the heteromultimeric mGlu could be expressed on the plasma membrane. In fact, the glutamate-induced Ca2þ transients were not observed in cells transfected with either WT-C1 or WT-C2KKTN, but were detected in cells co-transfected with them both (Fig. 2AeC). The combination of WT-C1 and WT-C2KKTN also produced the Gi/o response in HEK293T cells (Fig. 2B and C) but did not exhibit the glutamate-induced Gs response in CHO cells (0 Gs-positive cells out of 8 glutamate-positive cells). The impaired Gs coupling might be
835
caused by the replacement of the C-tail with the C1 and C2KKTN, since the distal C-terminal tail of mGlu1 is requisite for the Gs coupling in CHO cells (Tateyama and Kubo, 2007). We thus could not investigate whether the Gs pathway could be triggered through the cis-activation alone, and hereafter investigated the dual signaling via mGlu1 (Gi/o and Gq) in HEK293T cells. Transfection of a double mutant construct, R78L/F781S-C1, did not produce any responses upon glutamate application. Co-transfection of the R78L/F781S-C1 with WT-C2KKTN restored the Gq but not the Gi/o responses (Fig. 2AeC, filled square), indicating that the Gi/o pathway was not triggered through the cis-activation alone. 3.3. The presence of one coupling-deficient subunit in the mGlu1 dimer attenuated the Gi/o coupling but not the Gq coupling The significant decreases in the GIRK current density might be caused by the glutamate-binding deficiency or G protein-coupling deficiency in one subunit of the mGlu1 dimer. We have addressed this issue by using chimeras, F781S-C1 and R78L-C2KKTN (Fig. 2D).
Fig. 2. The Gi/o coupling, but not the Gq coupling, was attenuated when one coupling-deficient subunit was included in the mGlu1 dimer. (A and D) Schematic drawings of the combinations of mGlu1-C1 and -C2KKTN constructs (upper panels) and the corresponding labels (lower panels). (B and E) Time-lapse changes in [Ca2þ]i (upper traces) and GIRK current (lower traces) in HEK293T cells. The application timing of glutamate (1 mM) is indicated by arrows (upper panels) and black bars on current traces (lower panels). (C and F) Averages of the response amplitudes of mGlu1 constructs. Upper panel: Bars represent the D[Ca2þ]i (n ¼ 56e89 cells (C), n ¼ 31e115 cells (F)). Lower panel: Bars represent the glutamate-induced GIRK current densities (n ¼ 6e17 (C), n ¼ 6e15 (F)). The Gq pathway, but not the Gi/o pathway, was triggered through cis-activation alone of mGlu1 (C). The Gi/o pathway, but not the Gq pathway, was attenuated when a coupling-deficient subunit was included within mGlu1 (F). *: p < 0.05 between the indicated groups; n.s.: not significant.
836
M. Tateyama, Y. Kubo / Neuropharmacology 61 (2011) 832e841
Transfection of either construct, F781S-C1 or R78L-C2KKTN, did not produce any responses upon glutamate application (Fig. 2F). But cotransfection with both constructs produced the Gq, but not the Gi/o, responses (Fig. 2E and F filled diamond), which is consistent with the results shown in Fig. 1C. Next, the roles of the binding- or coupling-deficient subunit were investigated. When a bindingdeficient subunit was co-expressed with the WT subunit of mGlu1 (WT-C1&R78L-C2KKTN), the phenotype was similar to that of WT. The binding-deficient combination activated both the Gq and Gi/o pathways (Fig. 2E and F, open diamond). On the other hand, when the coupling-deficient subunit was co-expressed with the WT subunit (F781S-C1&WT-C2KKTN), the Gq response was observed upon glutamate application but the increase in the GIRK current density was subtle (Fig. 2E and F, gray diamond). These results suggested that the presence of one coupling-deficient subunit in the mGlu1 dimer attenuated the Gi/o, but not the Gq, coupling of the receptor.
3.4. The presence of one coupling-deficient subunit in the mGlu1 dimer did not affect the Gq coupling of the receptor Gq coupling of the coupling-deficient combination (F781SC1&WT-C2KKTN) was also examined by recording whole cell currents through the Gq-coupled TRPC3 channels. Activation of the coupling-deficient combination elicited gradual increases in the inward currents (Fig. 3A). The glutamate-induced inward currents were sensitive to La3þ and their amplitudes were dependent on the concentration of glutamate. The [glutamate]eTRPC current amplitude relationship of the coupling-deficient combination was highly similar to that of the WT combination (Fig. 3B). These results showed that the presence of one coupling-deficient subunit in the mGlu1 dimer did not alter the glutamate sensitivity of the Gq response.
3.5. The presence of one coupling-deficient subunit in the mGlu1 dimer did not decrease the surface expression but attenuated the Gi/o coupling The glutamate-induced GIRK current density was significantly decreased when the coupling-deficient subunit was included in mGlu1. The decrease in the current density might have been caused by a decrease in the number of heterodimeric mGlu1s on the plasma membrane. The binding-deficient subunit has been reported not to significantly decrease the surface expression of the heterodimers (Kammermeier and Yun, 2005; Kniazeff et al., 2004), whereas it has been unclear whether the F781S mutation suppresses the surface expression of the heterodimeric mGlu1. We thus investigated the surface expression level of mGlu1 constructs immunocytochemically (Tateyama and Kubo, 2008). The FLAG tag was inserted into the ECD of WT-C2KKTN (FL-WT-C2) and labeled by the antibody under a membrane non-permeabilized condition (Fig. 4A, upper panels). In these experiments, GFP from the pIRESGFP vector containing cDNA of the S196A-GIRK2 channel was used as a successful transfection marker. When the FL-WT-C2 was transfected alone, the surface expression was not detected in GFPpositive cells (Fig. 4B, left images), but co-transfection of the WT-C1 or F781S-C1 with FL-WT-C2 induced the surface expression of the receptor (Fig. 4B, middle and right images). The intensities of Alexa564 in GFP-positive cells varied from cell to cell (Fig. S2), but the surface expression levels (Fig. 4D) as well as the GFP intensities (Fig. S2) were not significantly different between the two groups (Fig. 4D and S2). The averages of the FLAG-associated fluorescence intensity (A.U.) of WT-C1&FL-WT-C2 and F781S-C1&FL-WT-C2 were 19.2 2.3 (n ¼ 137 cells) and 17.9 1.7 (n ¼ 136 cells), respectively. If the intensity of fluorescence signals had reached the saturation level of our detection system, it would not have been possible to compare the values reliably. However, this possibility was ruled out, since the fluorescence intensity was clearly within
Fig. 3. The presence of one coupling-deficient subunit in the mGlu1 dimer did not affect the Gq coupling. (A) Schematic drawings of the combinations of mGlu1-C1 and -C2KKTN constructs (upper panels) and time-lapse changes in the Gq-coupled TRPC3 current (lower traces) in HEK293T cells. Inward currents through the La3þ blockable TRPC3 channels were induced by the indicated concentration of glutamate (mM, gray and black bars on the traces) in cells transfected with WT-C1&WT-C2KKTN (left) and F781S-C1&WT-C2KKTN (right). (B) Averages of response amplitudes of constructs. The presence of one coupling-deficient subunit in the mGlu1 dimer did not affect the Gq coupling of mGlu1. Increases in the current amplitude induced by the indicated concentration of glutamate were normalized to that by 1 mM glutamate (I/Imax). Bars represent the relative amplitudes of glutamate-induced currents (n ¼ 5e8 for each concentration). n.s.: not significant.
M. Tateyama, Y. Kubo / Neuropharmacology 61 (2011) 832e841
837
Fig. 4. A presence of one coupling-deficient subunit in the mGlu1 dimer did not decrease the surface expression but attenuated the Gi/o coupling. (A) Schematic drawings of the combinations of mGlu-C1 and FLAG-tagged WT-C2KKTN constructs (FL-WT-C2). (B) Images of GFP-positive cells and surface expression of the combinations indicated in A. The cell surface expression of constructs was detected immunocytochemically under a non-membrane permeabilized condition. FLAG tag was labeled by FLAG-antibody and then Alexa564-conjugated secondary antibody (lower images). Cells expressing the S196A-GIRK2 channels (GFP-positive cells in the upper images) were selected for further analyses. Scale bars represent 50 mm. (C) Voltage-dependence of the glutamate-induced inward current. Shown are typical currentevoltage relationships, obtained by a ramp protocol, in the absence (black) and the presence (red) of glutamate (1 mM). Currents were recorded from GFP-positive cells co-transfected with the indicated mGlu1 constructs. (D) Quantitative analysis of the surface expression (middle panel) and glutamate-induced GIRK currents density (lower panel). The upper panel indicates the combination of the transfected constructs. Bars represent the fluorescence intensities of the Alexa-564 (n ¼ 26e137 GFP-positive cells, A.U.) and increases in the glutamate-induced GIRK current density (at a holding potential of 80 mV, n ¼ 6e16). *: p < 0.05 between the indicated groups; n.s.: not significant.
the detection limit in cells transfected with FL-WT-C2 constructs (Fig. S3). We also compared the surface expression level of the heterodimers by labeling the surface membrane proteins with a membrane-impermeable biotin reagent. The fraction of the biotin-labeled mGlu1 constructs was not significantly different between cells transfected with F781S-C1&FL-WT-C2 and those transfected with WT-C1&FL-WT-C2 (Fig. S4). Whole cell currents were then recorded from the GFP-positive cells, under the same transfection condition as used for the immunocytochemical experiments. To eliminate possible influences by protein kinase C-dependent phosphorylation of the GIRK2 channel, Ser196 was mutated to alanine (Mao et al., 2004). Upon glutamate application, no increase in the inward currents was observed in cells transfected with FL-WT-C2, whereas the inward rectifying currents were increased in cells co-transfected with WTC1 and FL-WT-C2 (Fig. 4C, red lines). Activation of the heterodimeric receptor, F781S-C1&FL-WT-C2, induced the increases in the GIRK currents, but the current density was significantly smaller than that by the WT combination (Fig. 4D). These inhibitory effects of the coupling-deficient subunit were not observed on the Gq-coupled calcium transient: the maximal D[Ca2þ]i and the
[glutamate]eD[Ca2þ]i relationship were not changed by the coupling-deficient subunit (Fig. 5A). In contrast, the [glutamate]e GIRK current relationship in F781S-C1&FL-WT-C2 was attenuated when compared to that in WT-C1&FL-WT-C2 (Fig. 5B). Taken together, these results suggested that the presence of one couplingdeficient subunit in the mGlu1 dimer significantly attenuates the Gi/o coupling of the receptor but neither the surface expression nor the Gq coupling of the mGlu1. 3.6. The presence of one subunit tagged with YFP in the 2nd loop in the mGlu1 dimer impaired the Gi/o coupling but not the Gq coupling We have previously reported that the fusion of bulky fluorescent protein at the ICDs of mGlu1a disrupts the G-protein coupling without altering the surface expression or the glutamate-binding properties (Tateyama et al., 2004). We thus fused YFP at the 2nd loop of mGlu1-C1 (i2-YFP-C1, Fig. 6A), and investigated its effects on the Gq and Gi/o signaling of mGlu1. The Gq-coupled Ca2þ transients were observed upon glutamate application in cells transfected with i2-YFP-C1 and WT-C2KKTN (Fig. 6A), while the glutamate-induced increase in the GIRK current was not clearly detected (Fig. 6B).
838
M. Tateyama, Y. Kubo / Neuropharmacology 61 (2011) 832e841
Fig. 5. The presence of one coupling-deficient subunit in the mGlu1 dimer attenuated the Gi/o but not the Gq coupling. (A) Averages of D[Ca2þ]i. The amplitudes of glutamateinduced calcium transients were analyzed in cells transfected with WT-C1 and FL-WT-C2 (open square) and with F781S-C1 and FL-WT-C2 (filled square). Bars represent the glutamate-induced increase in D[Ca2þ]i. At all tested concentrations of glutamate, there was no difference in the D[Ca2þ]i (n ¼ 44e180 cells). (B) The normalized [glutamate]eGIRK current relationship. GIRK currents induced by various concentrations of glutamate were recorded to evaluate the doseeresponse relationship. For these analyses, cells of which the maximal GIRK current density was more than 1 pA/pF were selected. Circles represent the normalized amplitude of the glutamate-induced GIRK current (I/Imax). The doseeresponse curve was shifted rightward when a coupling-deficient subunit was included in the mGlu1 dimer. The EC50 (mM) values were 6.03 0.22 for WT-C1&FL-WT-C2 (open circles, n ¼ 7) and 11.26 0.61 for F781S-C1&FL-WT-C2 (filled circles, n ¼ 8). The slope factors were 2.16 0.13 for WT-C1&FL-WT-C2 and 1.84 0.18 for F781S-C1&FL-WT-C2.
These results were consistent with the previous results for transfection with F781S-C1&WT-C2KKTN (Figs. 2F and 5). 3.7. The presence of one coupling-deficient subunit in the mGlu2 attenuated the Gi/o coupling The Gi/o pathway is triggered through trans-activation alone in GABAB receptors (Galvez et al., 2001; Margeta-Mitrovic et al., 2001) and heteromeric mGlu2&4 receptors (Doumazane et al., 2010). In addition, fluorescent proteins fused at various ICDs of the GB1 subunit have been shown not to inhibit the Gi/o coupling of the receptor (Matsushita et al., 2010). Therefore, attenuation of the Gi/o coupling by the coupling-deficient subunit may not be generally observed in mGlus other than mGlu1. We thus investigated the effects of different activation mechanisms of mGlu2 on the GIRK current. Large inward currents were elicited by the application of 100 mM glutamate in cells transfected with WT-mGlu2 (Fig. 7A, top). These responses were not observed in cells transfected with a glutamate-binding attenuated mutant (R57A-mGlu2) (Malherbe et al., 2001) or a coupling-deficient mutant (F756S-mGlu2) (Doumazane et al., 2010) (Fig. 7A and S5). These effects were not due to decreases in the surface expression of the receptors, since R57A and F756S mutations did not inhibit the surface expression of the mGlu2 constructs (Table S1). Co-expression of R57A- and F756S-mGlu2 partially but not fully evoked the glutamate-induced GIRK current (Fig. 7A, bottom). The glutamateinduced current density in the combination was less than 50% compared to that in the WT-mGlu2 (at 100 mM glutamate, Fig. 7B). In addition, the activation time course of the GIRK current by the combination of R57A- and F756S-mGlu2 was clearly slower than that by WT-mGlu2, suggesting that the trans-activation alone is not sufficient for the intact Gi/o coupling of mGlu2. We then analyzed the coupling properties of specific combinations by constructing several mGlu2-C1 and -C2KKTN constructs. Although the mGlu2-C1 constructs were expected not to be expressed on the plasma membrane, glutamate-induced increases in the GIRK currents were observed in cells transfected with the mGlu2-C1 constructs (Fig. S6). Therefore, the mGlu2-C1 constructs could not be used to exclude the functional homodimer of mGlu2 from the plasma membrane. As a second choice, the cDNA content of F756S-mGlu2 was increased to a level 10-fold that of the WT, which would increase the population of the heterodimer without suppressing the expression level of
functional receptors on the plasma membrane (Table S2). We therefore used these constructs to investigate the qualitative aspects of the glutamate-induced GIRK currents (Fig. 8A, upper panels). Under these transfection conditions, the [glutamate]eGIRK current amplitude
Fig. 6. Insertion of YFP into one subunit of the mGlu1 dimer attenuated the Gi/o but not the Gq coupling. (A) Schematic drawings of the combinations of mGlu1 constructs. YFP was inserted into the 2nd intracellular loop of mGlu1-C1 (i2-YFP-C1). (B) Averages of the response amplitudes of mGlu1 constructs. The Gi/o pathway, but not the Gq pathway, was attenuated when YFP was inserted into the intracellular loops of one subunit of the mGlu1 dimer. Left panel: Bars represent the D[Ca2þ]i (n ¼ 58 cells for WT-C1&WT-C2KKTN, n ¼ 84 cells for i2-YFP-C1&WT-C2KKTN). Right panel: Bars represent the densities of glutamate-induced GIRK currents (n ¼ 7 for WT-C1&WTC2KKTN, n ¼ 8 for i2-YFP-C1&WT-C2KKTN). *: p < 0.05 between the indicated groups, n.s.: not significant.
M. Tateyama, Y. Kubo / Neuropharmacology 61 (2011) 832e841
839
Fig. 7. The Gi/o pathway was not fully activated through trans-activation of mGlu2 alone. (A) Schematic drawings of the combination of mGlu2 constructs (left panel) and typical GIRK current recordings (right panel). Shown are typical current traces, at a holding potential of 80 mV, upon application of 100 mM glutamate (white bars) followed by 1 mM glutamate (gray bars). Arrows indicate the current increases evoked by 100 mM glutamate, which were not sufficient to activate R57A-mGlu2 and F756S-mGlu2. (B) Averages of GIRK current density induced by 100 mM glutamate. The Gi/o pathway was partially activated through trans-activation of mGlu2. Bars represent the inward GIRK current densities (n ¼ 7e9). *: p < 0.05.
curve was significantly shifted rightwards by the combination of WTand F756S-mGlu2 (Fig. 8B, filled circle). These results suggested that the Gi/o coupling was also attenuated in mGlu2 when the couplingdeficient subunit was included in the mGlu2 dimer.
4. Discussion 4.1. Dimeric configuration of mGlu The dimeric configuration is a structural and functional characteristic of the family C GPCRs (Pin et al., 2003). In particular, the dimeric configuration of mGlu has been demonstrated both in heterogeneous expression systems and endogenous neuronal cells (Robbins et al., 1999). The time-resolved FRET analysis has also demonstrated that the mGlu1 dimer is not a promiscuous oligomeric form of GPCR such as might be observed in an overexpression system (Maurel et al., 2008). Hydrophobic interaction and the formation of a disulfide bond between the N-terminal ECDs are important to form an mGlu1 dimer (Robbins et al., 1999; Tsuji et al., 2000). Impairments of these interactions disrupted or attenuated the function of mGlu1. A mutation at the dimer interface of ECDs (I120A) did not alter the ligand binding or surface expression of the receptor but disrupted the function (Sato et al., 2003). Treatment of hippocampal slices with a reducing agent of DTT attenuated the function of type I mGlu receptors (Copani et al., 2000). Thus the dimeric configuration of mGlu1 is requisite for its signaling function and potentially confers variation in the intramolecular activation mechanisms.
4.2. Does the multiple activation mechanism of mGlu1 occur under physiological conditions? It has been reported that the mGlu1 subunits become couplingdeficient physiologically when the 2nd intracellular loops are masked by interaction with proteins, such as G protein-coupled receptor kinase (GRK) (Dhami et al., 2005). It could be assumed that one or two GRKs bind to the mGlu1a dimer. Based on the results in the present study, this could serve as a physiological mechanism to change downstream signaling. If two GRKs occupy two subunits of the mGlu1 dimer at the same time, neither the Gq nor the Gi/o could be triggered upon activation of mGlu1a. In contrast, when one GRK interacts with one subunit of the mGlu1a dimer, the Gi/o coupling but not the Gq coupling of the receptor would be attenuated, as the impairment of Gi/o coupling was observed with the insertion of bulky YFP into one subunit (Fig. 6). These differential effects on the downstream signaling might regulate the neuronal transmission, since PTX-sensitive Gi/o signaling via mGlu1 has been suggested to inhibit the voltage gated calcium channels in sympathetic and cerebellar Purkinje neurons (Kammermeier, 2009; Kitano et al., 2003). Further studies are necessary to fully elucidate the physiological significance of the multiple signaling and its regulation. 4.3. The presence of one coupling-deficient subunit in the mGlu dimer attenuated the Gi/o coupling The presence of one coupling-deficient subunit in the mGlu1 or mGlu2 dimer attenuated the Gi/o coupling (Figs. 2F, 4D, 5B, 6B, and 8). These results could be interpreted as meaning that one
840
M. Tateyama, Y. Kubo / Neuropharmacology 61 (2011) 832e841
In conclusion, multiple signaling pathways via the mGlu1 were differentially activated when only the trans-activation was available. Furthermore, the presence of one coupling-deficient subunit in the mGlu1 dimer attenuated the Gi/o coupling but neither the Gq coupling nor the cell surface expression. These results suggested that the intra-molecular activation mechanisms of the dimeric mGlu1 differ depending on the type of G protein. Acknowledgments We would like to thank Drs. J. Miyazaki for the pCXN2 expression vector, S. Nakanishi for the mGlu1a and mGlu2 cDNAs, J. Zhang and M.B. Feller for the ICUE2 cDNA, and L.Y. Jan for the GB1 and GB2 cDNAs. We also thank Drs. Y. Fukata and M. Fukata for their advice on the biotinylation experiments. Thanks are also due to Y. Asai for technical support. This work was supported partly by research grants from the Japan Society for the Promotion of Science (to MT and to YK), and from the Ministry of Education, Science, Sports, Culture, and Technology of Japan (to YK). Appendix. Supplementary material Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.neuropharm.2011.05. 031. Fig. 8. The presence of one coupling-deficient subunit in the mGlu2 dimer attenuated the Gi/o coupling. (A) Schematic drawings of the combination of mGlu2 constructs (upper panel) and typical GIRK current traces (lower panel). The amount of F756SmGlu2 cDNA used for transfection was 10-fold greater than the amount of WT cDNA, to reduce the population of the WT homodimer. Application of the indicated concentrations (mM) of glutamate elicited increases in the inward currents in cells transfected with WT (0.1, left panel) and WT with F756 (0.1:1, right panel). (B) The normalized [glutamate]eGIRK current amplitude relationship. The relative GIRK current amplitude (I/Imax) was plotted against the [glutamate] (n ¼ 4e8 for each point). The EC50 (mM) values were 3.92 0.10 for WT (open circle) and 6.31 0.07 for WT with a10-fold higher amounts of F756S-mGlu2 (filled circle) (p < 0.05). The slope factor was 1.87 0.08 for WT and 1.67 0.03 for WT with 10-fold higher amounts of F756S-mGlu2.
intact G protein-coupling subunit is not sufficient for the mGlu dimer to fully activate the Gi/o pathways. This interpretation is partly consistent with a recent study of dimeric dopamine type 2 receptors, in which the G-protein coupling of the dopamine receptor was shown to be negatively affected by the other receptor of the dimer, which lacked the G-protein coupling (Han et al., 2009). Two intact subunits may be necessary for the mGlu dimer to fully activate the Gi/o pathway. This assumption raised the possibility that two Gi/o proteins simultaneously interact with two subunits of the dimeric receptor for full activation of the downstream pathway. This appears unlikely, however, since two 7-TMDs of the dimeric receptor would not become active at the same time for the activation of G proteins (Brock et al., 2007; Han et al., 2009). One subunit of mGlu5, directly activated by a positive allosteric modulator, has been demonstrated to prevent the other subunit being activated (Brock et al., 2007). Both subunits of mGlu may not become simultaneously active and therefore two Gi/o proteins could not interact with two subunits of mGlu at the same time. As another possibility, both intact subunits in the mGlu dimer might be somehow involved in the coupling with single Gi/o proteins for its full activation. In contrast to the Gi/o coupling, one coupling-deficient subunit did not alter the Gq-coupled calcium transient induced by activation of mGlu1 (Figs. 2F, 3, 5A, and 6B). These differences suggested that the coupling configuration of mGlu1 for Gq is different from that of mGlu1 for Gi/o.
References Abe, H., Misaka, T., Tateyama, M., Kubo, Y., 2003. Effects of coexpression with Homer isoforms on the function of metabotropic glutamate receptor 1alpha. Mol. Cell. Neurosci. 23, 157e168. Brock, C., Oueslati, N., Soler, S., Boudier, L., Rondard, P., Pin, J.P., 2007. Activation of a dimeric metabotropic glutamate receptor by inter-subunit rearrangement. J. Biol. Chem. 282, 33000e33008. Copani, A., Romano, C., Di Giorgi Gerevini, V., Nicosia, A., Casabona, G., Storto, M., Mutel, V., Nicoletti, F., 2000. Reducing conditions differentially affect the functional and structural properties of group-I and -II metabotropic glutamate receptors. Brain Res. 867, 165e172. Dhami, G.K., Babwah, A.V., Sterne-Marr, R., Ferguson, S.S., 2005. Phosphorylationindependent regulation of metabotropic glutamate receptor 1 signaling requires g protein-coupled receptor kinase 2 binding to the second intracellular loop. J. Biol. Chem. 280, 24420e24427. DiPilato, L.M., Cheng, X., Zhang, J., 2004. Fluorescent indicators of cAMP and Epac activation reveal differential dynamics of cAMP signaling within discrete subcellular compartments. Proc. Natl. Acad. Sci. U.S.A. 101, 16513e16518. Doumazane, E., Scholler, P., Zwier, J.M., Trinquet, E., Rondard, P., Pin, J.P., 2010. A new approach to analyze cell surface protein complexes reveals specific heterodimeric metabotropic glutamate receptors. FASEB J. Dunn, T.A., Wang, C.T., Colicos, M.A., Zaccolo, M., DiPilato, L.M., Zhang, J., Tsien, R.Y., Feller, M.B., 2006. Imaging of cAMP levels and protein kinase A activity reveals that retinal waves drive oscillations in second-messenger cascades. J. Neurosci. 26, 12807e12815. Francesconi, A., Duvoisin, R.M., 1998. Role of the second and third intracellular loops of metabotropic glutamate receptors in mediating dual signal transduction activation. J. Biol. Chem. 273, 5615e5624. Francesconi, A., Duvoisin, R.M., 2000. Opposing effects of protein kinase C and protein kinase A on metabotropic glutamate receptor signaling: selective desensitization of the inositol trisphosphate/Ca2þ pathway by phosphorylation of the receptor-G protein-coupling domain. Proc. Natl. Acad. Sci. U.S.A. 97, 6185e6190. Galvez, T., Duthey, B., Kniazeff, J., Blahos, J., Rovelli, G., Bettler, B., Prezeau, L., Pin, J.P., 2001. Allosteric interactions between GB1 and GB2 subunits are required for optimal GABA(B) receptor function. Embo J. 20, 2152e2159. Han, Y., Moreira, I.S., Urizar, E., Weinstein, H., Javitch, J.A., 2009. Allosteric communication between protomers of dopamine class A GPCR dimers modulates activation. Nat. Chem. Biol. 5, 688e695. Hartmann, J., Dragicevic, E., Adelsberger, H., Henning, H.A., Sumser, M., Abramowitz, J., Blum, R., Dietrich, A., Freichel, M., Flockerzi, V., Birnbaumer, L., Konnerth, A., 2008. TRPC3 channels are required for synaptic transmission and motor coordination. Neuron 59, 392e398. Jensen, A.A., Sheppard, P.O., O’Hara, P.J., Krogsgaard-Larsen, P., Brauner-Osborne, H., 2000. The role of Arg(78) in the metabotropic glutamate receptor mGlu(1) for agonist binding and selectivity. Eur. J. Pharmacol. 397, 247e253. Kammermeier, P.J., 2009. C-terminal deletion of metabotropic glutamate receptor 1 selectively abolishes coupling to Galphaq. Eur. J. Pharmacol. 627, 63e68.
M. Tateyama, Y. Kubo / Neuropharmacology 61 (2011) 832e841 Kammermeier, P.J., Yun, J., 2005. Activation of metabotropic glutamate receptor 1 dimers requires glutamate binding in both subunits. J. Pharmacol. Exp. Ther. 312, 502e508. Kitano, J., Nishida, M., Itsukaichi, Y., Minami, I., Ogawa, M., Hirano, T., Mori, Y., Nakanishi, S., 2003. Direct interaction and functional coupling between metabotropic glutamate receptor subtype 1 and voltage-sensitive Cav2.1 Ca2þ channel. J. Biol. Chem. 278, 25101e25108. Kniazeff, J., Bessis, A.S., Maurel, D., Ansanay, H., Prezeau, L., Pin, J.P., 2004. Closed state of both binding domains of homodimeric mGlu receptors is required for full activity. Nat. Struct. Mol. Biol. 11, 706e713. Kunishima, N., Shimada, Y., Tsuji, Y., Sato, T., Yamamoto, M., Kumasaka, T., Nakanishi, S., Jingami, H., Morikawa, K., 2000. Structural basis of glutamate recognition by a dimeric metabotropic glutamate receptor. Nature 407, 971e977. Malherbe, P., Knoflach, F., Broger, C., Ohresser, S., Kratzeisen, C., Adam, G., Stadler, H., Kemp, J.A., Mutel, V., 2001. Identification of essential residues involved in the glutamate binding pocket of the group II metabotropic glutamate receptor. Mol. Pharmacol. 60, 944e954. Mao, J., Wang, X., Chen, F., Wang, R., Rojas, A., Shi, Y., Piao, H., Jiang, C., 2004. Molecular basis for the inhibition of G protein-coupled inward rectifier K(þ) channels by protein kinase C. Proc. Natl. Acad. Sci. U.S.A. 101, 1087e1092. Margeta-Mitrovic, M., Jan, Y.N., Jan, L.Y., 2001. Ligand-induced signal transduction within heterodimeric GABA(B) receptor. Proc. Natl. Acad. Sci. U.S.A. 98, 14643e14648. Matsushita, S., Nakata, H., Kubo, Y., Tateyama, M., 2010. Ligand-induced rearrangements of the GABA(B) receptor revealed by fluorescence resonance energy transfer. J. Biol. Chem. 285, 10291e10299. Maurel, D., Comps-Agrar, L., Brock, C., Rives, M.L., Bourrier, E., Ayoub, M.A., Bazin, H., Tinel, N., Durroux, T., Prezeau, L., Trinquet, E., Pin, J.P., 2008. Cell-surface protein-protein interaction analysis with time-resolved FRET and snap-tag technologies: application to GPCR oligomerization. Nat. Methods 5, 561e567. Muto, T., Tsuchiya, D., Morikawa, K., Jingami, H., 2007. Structures of the extracellular regions of the group II/III metabotropic glutamate receptors. Proc. Natl. Acad. Sci. U.S.A. 104, 3759e3764. Nicoletti, F., Bockaert, J., Collingridge, G.L., Conn, P.J., Ferraguti, F., Schoepp, D.D., Wroblewski, J.T., Pin, J.P., 2010. Metabotropic glutamate receptors: from the workbench to the bedside. Neuropharmacology. Park, P.S., Filipek, S., Wells, J.W., Palczewski, K., 2004. Oligomerization of G proteincoupled receptors: past, present, and future. Biochemistry 43, 15643e15656.
841
Pin, J.P., Galvez, T., Prezeau, L., 2003. Evolution, structure, and activation mechanism of family 3/C G-protein-coupled receptors. Pharmacol. Ther. 98, 325e354. Robbins, M.J., Ciruela, F., Rhodes, A., McIlhinney, R.A., 1999. Characterization of the dimerization of metabotropic glutamate receptors using an N-terminal truncation of mGluR1alpha. J. Neurochem. 72, 2539e2547. Sato, T., Shimada, Y., Nagasawa, N., Nakanishi, S., Jingami, H., 2003. Amino acid mutagenesis of the ligand binding site and the dimer interface of the metabotropic glutamate receptor 1. Identification of crucial residues for setting the activated state. J. Biol. Chem. 278, 4314e4321. Selkirk, J.V., Price, G.W., Nahorski, S.R., Challiss, R.A., 2001. Cell type-specific differences in the coupling of recombinant mGlu1alpha receptors to endogenous G protein sub-populations. Neuropharmacology 40, 645e656. Sheffler, D.J., Conn, P.J., 2008. Allosteric potentiators of metabotropic glutamate receptor subtype 1a differentially modulate independent signaling pathways in baby hamster kidney cells. Neuropharmacology 55, 419e427. Sugiyama, Y., Kawaguchi, S.Y., Hirano, T., 2008. mGluR1-mediated facilitation of long-term potentiation at inhibitory synapses on a cerebellar Purkinje neuron. Eur. J. Neurosci. 27, 884e896. Tateyama, M., Abe, H., Nakata, H., Saito, O., Kubo, Y., 2004. Ligand-induced rearrangement of the dimeric metabotropic glutamate receptor 1alpha. Nat. Struct. Mol. Biol. 11, 637e642. Tateyama, M., Kubo, Y., 2006. Dual signaling is differentially activated by different active states of the metabotropic glutamate receptor 1alpha. Proc. Natl. Acad. Sci. U.S.A. 103, 1124e1128. Tateyama, M., Kubo, Y., 2007. Coupling profile of the metabotropic glutamate receptor 1alpha is regulated by the C-terminal domain. Mol. Cell. Neurosci. 34, 445e452. Tateyama, M., Kubo, Y., 2008. Regulatory role of C-terminus in the G-protein coupling of the metabotropic glutamate receptor 1. J. Neurochem. 107, 1036e1046. Tsuji, Y., Shimada, Y., Takeshita, T., Kajimura, N., Nomura, S., Sekiyama, N., Otomo, J., Usukura, J., Nakanishi, S., Jingami, H., 2000. Cryptic dimer interface and domain organization of the extracellular region of metabotropic glutamate receptor subtype 1. J. Biol. Chem. 275, 28144e28151. Yamashita, T., Terakita, A., Kai, T., Shichida, Y., 2008. Conformational change of the transmembrane helices II and IV of metabotropic glutamate receptor involved in G protein activation. J. Neurochem. 106, 850e859.
Contents lists available at ScienceDirect
Neuropharmacology journal homepage: www.elsevier.com/locate/neuropharm
The intra-molecular activation mechanisms of the dimeric metabotropic glutamate receptor 1 differ depending on the type of G proteins Michihiro Tateyama*, Yoshihiro Kubo Division of Biophysics and Neurobiology, Department of Molecular Physiology, National Institute for Physiological Sciences, Myodaiji, Okazaki 444-8585, Japan
a r t i c l e i n f o
a b s t r a c t
Article history: Received 7 January 2011 Received in revised form 19 May 2011 Accepted 25 May 2011
Metabotropic glutamate receptor 1 (mGlu1) functions as a homodimer and activates not only the Gq but also the Gi/o and Gs pathways. Because of the dimeric configuration, different pathways could be activated either through the glutamate-bound subunit (cis-activation) and/or the other one (trans-activation). We here examined whether the intra-molecular activation mechanisms in the mGlu1 differ depending on the type of coupling G proteins, using various combinations of mGlu1 constructs that lack glutamate binding and/or G-protein coupling. The cis- or trans-activation alone was confirmed to trigger the Gq-coupled intracellular Ca2þ transient. In contrast, the Gi/o-coupled G protein-dependent inward rectifying potassium (GIRK) channels were not activated either through the cis- or trans-activation alone. When one subunit of dimeric mGlu1 lacked the G-protein coupling, a significant decrease in the glutamate-induced GIRK current density was also observed. As the G protein-coupling-deficient subunit did not decrease the cell surface expression of mGlu1 and the Gq-coupled Ca2þ transient, it was suggested that the coupling deficiency in one subunit of mGlu1 attenuates the Gi/o but not Gq coupling. In conclusion, multiple G-protein signaling was differentially activated by different intra-molecular activation mechanisms of the dimeric mGlu1. Ó 2011 Elsevier Ltd. All rights reserved.
Keywords: Metabotropic glutamate receptor Dual signaling TRPC3 GIRK
1. Introduction Metabotropic glutamate receptors (mGlu) are the key molecules for neuronal transmission and signaling (Nicoletti et al., 2010). mGlu are G protein-coupled receptors (GPCRs) that belong to family C, to which g-aminobutyric acid type B (GABAB) receptor also belongs. The family C GPCRs are dimers and have a large extracellular domain (ECD) where the ligand recognition sites locate (Nicoletti et al., 2010; Pin et al., 2003). The dimeric and/or oligomeric configuration has also been also reported in family A
Abbreviations: mGlu, metabotropic glutamate receptor; GPCR, G proteincoupled receptor; GABAB, g-aminobutyric acid type B; ECD, extracellular domain; 7TMD, 7 trans-membrane domain; ICD, intracellular domain; GIRK, G proteindependent inward rectifying potassium; TRPC3, transient receptor potential canonical 3; C1, C-terminal tail of GB1; C2KKTN, C-terminal tail of GB2 with KKTN; YFP, yellow fluorescent protein; HEK293T, human embryonic kidney 293T; CHO, Chinese hamster ovary; [Ca2þ]i, intracellular Ca2þ concentration; D[Ca2þ]i, maximal increase in the intracellular Ca2þ concentration; FRET, fluorescent resonance energy transfer; [cAMP]i, intracellular cAMP levels; PTX, pertussis toxin; [glutamate], glutamate concentration; A.U., arbitrary units; WT, wild type; FL-WT-C2, FLAGtagged WT-C2KKTN. * Corresponding author. Tel.: þ81 564 55 7823; fax: þ81 564 55 7825. E-mail addresses: [email protected] (M. Tateyama), [email protected] (Y. Kubo). 0028-3908/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.neuropharm.2011.05.031
GPCRs, but the formation of oligomers is not necessary for their signaling (Park et al., 2004). In contrast, the dimeric configuration is requisite for family C GPCRs to function as the signaling molecules (Pin et al., 2003). The ligand binding that occurs at the ECD should be transmitted to their 7 trans-membrane domain (7TMD) for activation of the G proteins. X-ray crystallographic analyses of mGlu have revealed that a glutamate stabilizes the closed conformation of one ECD, which has been suggested to cause dimeric rearrangement of the ECDs (Kunishima et al., 2000; Muto et al., 2007). The dimeric rearrangement was demonstrated in the intracellular domains (ICDs) (Tateyama et al., 2004), with some conformational changes of the 7TMD (Brock et al., 2007; Yamashita et al., 2008). The ligand-induced dimeric rearrangement has also been demonstrated in GABAB receptors (Matsushita et al., 2010), suggesting that the dimeric rearrangement is an important mechanism for the family C GPCRs to activate G proteins. Because of the dimeric configuration of the family C GPCR, G proteins might be activated via the ligand-bound subunit (cis-activation) and/or the other one (trans-activation). Alternatively, ligand binding on two subunits might be requisite for the G-protein activation. In the case of homodimeric mGlu5, the Gq pathway can be triggered either through the cis- or trans-activation alone (Brock et al., 2007). In addition, the heterodimeric GABAB receptor is accepted to activate the Gi/o proteins through trans-activation, since the receptor
M. Tateyama, Y. Kubo / Neuropharmacology 61 (2011) 832e841
consists of a ligand-sensing subunit of GB1 and a Gi/o coupling subunit of GB2 (Galvez et al., 2001; Margeta-Mitrovic et al., 2001). A subtype of mGlu, mGlu1a, is known to activate not only the Gq pathway but also the Gs and Gi/o pathways, which may be responsible for the diverse responses in neuronal cells (Kammermeier, 2009; Kitano et al., 2003; Sugiyama et al., 2008). The multiple signaling via mGlu1a is dependent on the mGlu1a expression system (Selkirk et al., 2001) and differentially affected by mutations or phosphorylation of the ICD of mGlu1a (Francesconi and Duvoisin, 1998, 2000). In addition, a difference in the activated conformation of mGlu1a has been suggested to alter the coupling profile of mGlu1 (Sheffler and Conn, 2008; Tateyama and Kubo, 2006). From these studies, it could be assumed that the difference in the intramolecular activation mechanisms, such as cis- or trans-activation, might differentially affect the Gs or Gi/o coupling of mGlu1a. In the present study, we confirmed that the Gq pathway via mGlu1 was activated either through the cis- or the trans-activation alone (Brock et al., 2007) and newly found that the Gi/o pathway was not triggered either through the cis- or the trans-activation alone. Similarly, the Gs pathway was not triggered through the trans-activation alone. We further investigated the effects of glutamate-binding or G protein-coupling deficiency in one subunit of mGlu1 on the multiple signaling. 2. Materials and methods 2.1. Constructs and expression system Point mutants of rat mGlu1a, mGlu2 and G protein-dependent inward rectifying potassium type 2 (GIRK2) channel were constructed by PCR with mutated primers and KOD plus ver. 2 Taq polymerase (Toyobo, Osaka, Japan). Transient receptor potential canonical type 3 (TRPC3) channel was isolated from a rat brain cDNA library by PCR. Fragments of the C-terminal tail of GB1 (C1) or GB2 with a retention signal KKTN (C2KKTN) were amplified by PCR with designed primer sets and then inserted after Arg857 using a SphI site of the mGlu1a constructs (Brock et al., 2007). The FLAG epitope was inserted into the N-terminal region of mGlu1-C2KKTN after the signal peptide sequence, as previously reported (Abe et al., 2003). The coding sequence of yellow fluorescent protein (YFP) was inserted between Ile685 and Leu686 at the 2nd intracellular loop of mGlu1-C1, i2-YFP-C1, by blunt-end ligation (Tateyama et al., 2004). DNA sequences of the constructs were confirmed and then subcloned into pCXN2 expression vector except for S196A-GIRK2, a mutant GIRK2 which is resistant to the inhibition by protein kinase C-dependent phosphorylation (Mao et al., 2004). The S196A-GIRK2 was subcloned into pIRES2-GFP (Clontech, Mountain View, CA, USA). Human embryonic kidney 293T (HEK293T) and Chinese hamster ovary (CHO) cells were transfected with the plasmid DNA using LipofectAMINE2000 (Invitrogen, Carlsbad, CA, USA) and seeded onto cover glasses. Experiments using HEK293T cells were carried out 24e48 h after transfection and those using CHO cells were done after 48e72 h. Before electrophysiological experiments or imaging, cells were incubated for more than 30 min in Hank’s balanced salt solution (Invitrogen) supplemented with 1 mM Ca2þ and 0.3 mM Mg2þ at room temperature.
833
2.3. Electrophysiology The Gi/o coupling of mGlu constructs was evaluated as the amplitude of the glutamate-induced inward currents through the GIRK2 channels. The Gq coupling of mGlu1 constructs was evaluated as the amplitude of the glutamate-induced inward current through the La3þ sensitive TRPC3 channels (Hartmann et al., 2008). Macroscopic currents were recorded by the whole cell patch clamp technique at room temperature (Tateyama and Kubo, 2008). The internal solution for recording of GIRK currents contained (in mM) 120 KCl, 5 K2-ATP, 10 NaCl, 3 EGTA, 10 HEPES, 0.1 CaCl2, 4 MgCl2, and 0.3 mM GTP (pH adjusted to 7.4 with KOH). The internal solution for recording of TRPC3 currents contained (in mM) 140 KCl, 4 Na2-ATP, 0.3 EGTA, 10 HEPES, and 5 MgCl2 (pH adjusted to 7.4 with KOH). The composition of the bath solution was the same as that for imaging. 140 mM Naþ ions were replaced with 140 mM Kþ ions for recording of GIRK currents and the Ca2þ ions were replaced with Ba2þ ions for recording of TRPC3 currents. After the glutamate-induced current reached a maximum, the mean amplitudes for 0.5 s were measured at a holding potential of 80 mV. The current amplitude induced by 1 mM glutamate was normalized by the cell capacitance to calculate the current density. To evaluate the glutamate sensitivity of the responses, the current amplitude evoked by various concentrations of glutamate ([glutamate]) was normalized to that evoked by 1 mM glutamate in each cell. The relationship between [glutamate] and GIRK current was fitted to a logistic curve (Origin8; OriginLab, Northampton, MA, USA). The voltagedependence of the glutamate-induced current was evaluated by applying a ramp protocol, 120 mV to 40 mV at a rate of 0.4 mV/ms, at 0.2 Hz. The current traces shown in figures were off-line smoothed by averaging 10 adjacent points (Origin 8) to clearly demonstrate the glutamate-induced changes. 2.4. Immunocytochemistry HEK293T cells were transfected with the same amount of cDNAs (1 mg) of mGlu1 constructs and S196A-GIRK2. Twenty-four hours after transfection, the immunocytochemical experiments were carried out, as previously reported (Tateyama and Kubo, 2008). The cells were fixed with 2% paraformaldehyde for 2 min at 4 C without Triton-X, which permeabilizes membrane, and then incubated with antibodies (1st antibody, FLAG-antibody clone M2, 4 mg/ml, SigmaeAldrich, Tokyo, Japan; 2nd antibody, Alexa-564-conjugated anti-mouse IgG, Invitrogen). Fluorescence images were captured with a BX-50 microscope (Olympus, Tokyo, Japan) equipped with an AxioCam CCD camera unit (Carl Zeiss, Jena, Germany) and analyzed with Aquacosmos software (Hamamatsu Photonics, Hamamatsu, Japan). To exclude fluorescence signals of Alexa-564 from the intracellular compartment, we first selected cells expressing GFP (excitation at 470e495 nm and emission at 510e550 nm, more than 6 arbitrary units (A.U.)) to determine the region for measurement. The fluorescence intensity of Alexa-564 (excitation at 520e550 nm and emission more than 580 nm) was measured at each pixel within the selected region. The intensity was normalized by the area to compare the surface expression level of mGlu1 constructs in each group. 2.5. Analysis and statistics All data are expressed as the means S.E., with n indicating the number of data. A statistical significance was estimated by unpaired Student’s t-test or Dunnet’s t-test; values of p < 0.05 were considered statistically significant.
3. Results 3.1. Trans-activation alone triggers Gq but not Gi/o or Gs
2.2. Monitoring of the changes in [Ca2þ]i and [cAMP]i Changes in the intracellular Ca2þ concentration ([Ca2þ]i) in HEK293T cells were monitored every 3 s by using fura2-AM (Invitrogen), and the ratio of the fluorescence intensity (excitation 340 nm/380 nm, emission 535 nm) was calculated. Cells loaded with the fura2-AM were placed in a recording chamber filled with 250 mL of the bath solution containing (in mM) 140 NaCl, 1 CaCl2, 4 KCl, 0.3 MgCl2, and 10 HEPES (pH adjusted to 7.4 with NaOH). 250 mL of twice-concentrated glutamate stock solved in the bath solution was applied to the recording chamber by pipetting. Peak amplitudes of the ratio after glutamate application were obtained from cells expressing YFP, a transfection marker, and the maximal increase in the [Ca2þ]i (D[Ca2þ]i) was calculated by subtraction of the basal amplitude measured 0e30 s before glutamate application. To evaluate the Gs and Gq coupling of mGlu1 constructs in CHO cells, we simultaneously monitored the efficiency of fluorescence resonance energy transfer (FRET) of ICUE2, a probe for cAMP concentration (DiPilato et al., 2004; Dunn et al., 2006) and the emission ratio of indo-1 (excitation 340 nm, emission 405 nm/ 470 nm; Invitrogen) every 3 s. The intracellular cAMP concentration ([cAMP]i) was evaluated as the reversed ratio of FRET (excitation 436 nm, emission 470 nm/ 535 nm) and then the reversed ratios were normalized to the basal values (averaged value for 30 s before glutamate application). Cells that showed either D[Ca2þ]i > 0.05 or [cAMP]i > 1.05 were defined as glutamate-responsive cells for the analyses.
We first confirmed that the Gq pathway could be triggered through the trans-activation of mGlu1 alone, by using glutamatebinding-deficient R78L- and G protein-coupling-deficient F781SmGlu1a (Francesconi and Duvoisin, 1998; Jensen et al., 2000) (Fig. 1A). Homo-multimers of R78L and F781S could not increase the [Ca2þ]i upon 1 mM glutamate application, although they were well expressed on the plasma membrane (Fig. S1). In contrast, the hetero-multimer formed by co-expression of R78L- and F781SmGlu1a did increase the [Ca2þ]i (Fig. 1B and D). These results indicated that the Gq pathway can be triggered through the transactivation alone, because only the trans-activation pathway is available in the combination of R78L and F781S (Fig. 1A). Then the Gi/o coupling of the hetero-multimer was investigated by recording inward currents through GIRK2 channels in HEK293T cells, since mGlu1a activates the Pertussis toxin (PTX)-sensitive Gi/ o pathway in HEK293T cells (Kitano et al., 2003; Tateyama and Kubo, 2008). Activation of wild-type (WT) mGlu1a increased the
834
M. Tateyama, Y. Kubo / Neuropharmacology 61 (2011) 832e841
Fig. 1. The Gq pathway, but not Gi/o or Gs, was activated through trans-activation alone of mGlu1a. (A) Schematic drawings of the combinations of mGlu1a constructs. (B) Timelapse changes in [Ca2þ]i (upper traces) and GIRK current (lower traces) in HEK293T cells. [Ca2þ]i is represented as the ratio of fluorescence intensity (fura-2AM). The application timing of glutamate (1 mM) is indicated by arrows (upper panels) and black bars on the current traces (lower panels). (C) Averages of response amplitudes of mGlu1a constructs in HEK293T cells. Upper panel: Bars represent the maximal increases in [Ca2þ]i, (D[Ca2þ]i, n ¼ 66e117 cells). Lower panel: Bars represent the glutamate-induced GIRK current densities (n ¼ 4e10). (D) Time-lapse changes in [Ca2þ]i (upper panels) and [cAMP]i (lower panels) in CHO cells. [Ca2þ]i and [cAMP]i were simultaneously monitored in each cell. [Ca2þ]i is represented as the averaged ratio of the fluorescence intensity (indo-1) and [cAMP]i is represented as the normalized reciprocal FRET value (cAMP probe). The application timing of glutamate (1 mM) is indicated by arrows. (E) Averages of response amplitudes of mGlu1a constructs in CHO cells. Bars represent the glutamate-induced increase in D[Ca2þ]i (upper panel) and [cAMP]i (lower panel, averaged ratio between 230 and 250 s after glutamate application) in glutamate-positive cells (n ¼ 33e42 cells). *: p < 0.05 vs. WT; n.s.: not significant. The bars at right (C and E) represent the baseline fluctuation.
amplitude of GIRK current, whereas activation of the heteromultimer did not increase the current amplitude (Fig. 1B, lower traces). The glutamate-induced GIRK current density in cells transfected with R78L- and F781S-mGlu1a was significantly smaller than that in cells with WT-mGu1a (Fig. 1C). Gs coupling of the hetero-multimer was also analyzed by monitoring the
glutamate-induced increases in the [cAMP]i in CHO cells (Tateyama and Kubo, 2006). The activation of WT-mGlu1a increased the [cAMP]i, while activation of the hetero-multimer failed to increase it (Fig. 1D and E). These results clearly demonstrated that neither the Gi/o nor the Gs pathway could be triggered through the transactivation alone.
M. Tateyama, Y. Kubo / Neuropharmacology 61 (2011) 832e841
3.2. Cis-activation alone triggers Gq but not Gi/o Next we examined whether or not the multiple signaling pathways via mGlu1a are triggered through the cis-activation alone. For this purpose, the C-terminus of mGlu1a was replaced with those of GB1 (C1) and GB2 with an additional retention signal (C2KKTN). The endogenous and artificially added retention signals of C1 and C2KKTN have been reported to inhibit the surface expression of homo-multimeric mGlu-C1 and mGlu-C2KKTN, respectively (Brock et al., 2007). In the hetero-multimers, the retention signals would be masked by the C1eC2 interaction and thereby the heteromultimeric mGlu could be expressed on the plasma membrane. In fact, the glutamate-induced Ca2þ transients were not observed in cells transfected with either WT-C1 or WT-C2KKTN, but were detected in cells co-transfected with them both (Fig. 2AeC). The combination of WT-C1 and WT-C2KKTN also produced the Gi/o response in HEK293T cells (Fig. 2B and C) but did not exhibit the glutamate-induced Gs response in CHO cells (0 Gs-positive cells out of 8 glutamate-positive cells). The impaired Gs coupling might be
835
caused by the replacement of the C-tail with the C1 and C2KKTN, since the distal C-terminal tail of mGlu1 is requisite for the Gs coupling in CHO cells (Tateyama and Kubo, 2007). We thus could not investigate whether the Gs pathway could be triggered through the cis-activation alone, and hereafter investigated the dual signaling via mGlu1 (Gi/o and Gq) in HEK293T cells. Transfection of a double mutant construct, R78L/F781S-C1, did not produce any responses upon glutamate application. Co-transfection of the R78L/F781S-C1 with WT-C2KKTN restored the Gq but not the Gi/o responses (Fig. 2AeC, filled square), indicating that the Gi/o pathway was not triggered through the cis-activation alone. 3.3. The presence of one coupling-deficient subunit in the mGlu1 dimer attenuated the Gi/o coupling but not the Gq coupling The significant decreases in the GIRK current density might be caused by the glutamate-binding deficiency or G protein-coupling deficiency in one subunit of the mGlu1 dimer. We have addressed this issue by using chimeras, F781S-C1 and R78L-C2KKTN (Fig. 2D).
Fig. 2. The Gi/o coupling, but not the Gq coupling, was attenuated when one coupling-deficient subunit was included in the mGlu1 dimer. (A and D) Schematic drawings of the combinations of mGlu1-C1 and -C2KKTN constructs (upper panels) and the corresponding labels (lower panels). (B and E) Time-lapse changes in [Ca2þ]i (upper traces) and GIRK current (lower traces) in HEK293T cells. The application timing of glutamate (1 mM) is indicated by arrows (upper panels) and black bars on current traces (lower panels). (C and F) Averages of the response amplitudes of mGlu1 constructs. Upper panel: Bars represent the D[Ca2þ]i (n ¼ 56e89 cells (C), n ¼ 31e115 cells (F)). Lower panel: Bars represent the glutamate-induced GIRK current densities (n ¼ 6e17 (C), n ¼ 6e15 (F)). The Gq pathway, but not the Gi/o pathway, was triggered through cis-activation alone of mGlu1 (C). The Gi/o pathway, but not the Gq pathway, was attenuated when a coupling-deficient subunit was included within mGlu1 (F). *: p < 0.05 between the indicated groups; n.s.: not significant.
836
M. Tateyama, Y. Kubo / Neuropharmacology 61 (2011) 832e841
Transfection of either construct, F781S-C1 or R78L-C2KKTN, did not produce any responses upon glutamate application (Fig. 2F). But cotransfection with both constructs produced the Gq, but not the Gi/o, responses (Fig. 2E and F filled diamond), which is consistent with the results shown in Fig. 1C. Next, the roles of the binding- or coupling-deficient subunit were investigated. When a bindingdeficient subunit was co-expressed with the WT subunit of mGlu1 (WT-C1&R78L-C2KKTN), the phenotype was similar to that of WT. The binding-deficient combination activated both the Gq and Gi/o pathways (Fig. 2E and F, open diamond). On the other hand, when the coupling-deficient subunit was co-expressed with the WT subunit (F781S-C1&WT-C2KKTN), the Gq response was observed upon glutamate application but the increase in the GIRK current density was subtle (Fig. 2E and F, gray diamond). These results suggested that the presence of one coupling-deficient subunit in the mGlu1 dimer attenuated the Gi/o, but not the Gq, coupling of the receptor.
3.4. The presence of one coupling-deficient subunit in the mGlu1 dimer did not affect the Gq coupling of the receptor Gq coupling of the coupling-deficient combination (F781SC1&WT-C2KKTN) was also examined by recording whole cell currents through the Gq-coupled TRPC3 channels. Activation of the coupling-deficient combination elicited gradual increases in the inward currents (Fig. 3A). The glutamate-induced inward currents were sensitive to La3þ and their amplitudes were dependent on the concentration of glutamate. The [glutamate]eTRPC current amplitude relationship of the coupling-deficient combination was highly similar to that of the WT combination (Fig. 3B). These results showed that the presence of one coupling-deficient subunit in the mGlu1 dimer did not alter the glutamate sensitivity of the Gq response.
3.5. The presence of one coupling-deficient subunit in the mGlu1 dimer did not decrease the surface expression but attenuated the Gi/o coupling The glutamate-induced GIRK current density was significantly decreased when the coupling-deficient subunit was included in mGlu1. The decrease in the current density might have been caused by a decrease in the number of heterodimeric mGlu1s on the plasma membrane. The binding-deficient subunit has been reported not to significantly decrease the surface expression of the heterodimers (Kammermeier and Yun, 2005; Kniazeff et al., 2004), whereas it has been unclear whether the F781S mutation suppresses the surface expression of the heterodimeric mGlu1. We thus investigated the surface expression level of mGlu1 constructs immunocytochemically (Tateyama and Kubo, 2008). The FLAG tag was inserted into the ECD of WT-C2KKTN (FL-WT-C2) and labeled by the antibody under a membrane non-permeabilized condition (Fig. 4A, upper panels). In these experiments, GFP from the pIRESGFP vector containing cDNA of the S196A-GIRK2 channel was used as a successful transfection marker. When the FL-WT-C2 was transfected alone, the surface expression was not detected in GFPpositive cells (Fig. 4B, left images), but co-transfection of the WT-C1 or F781S-C1 with FL-WT-C2 induced the surface expression of the receptor (Fig. 4B, middle and right images). The intensities of Alexa564 in GFP-positive cells varied from cell to cell (Fig. S2), but the surface expression levels (Fig. 4D) as well as the GFP intensities (Fig. S2) were not significantly different between the two groups (Fig. 4D and S2). The averages of the FLAG-associated fluorescence intensity (A.U.) of WT-C1&FL-WT-C2 and F781S-C1&FL-WT-C2 were 19.2 2.3 (n ¼ 137 cells) and 17.9 1.7 (n ¼ 136 cells), respectively. If the intensity of fluorescence signals had reached the saturation level of our detection system, it would not have been possible to compare the values reliably. However, this possibility was ruled out, since the fluorescence intensity was clearly within
Fig. 3. The presence of one coupling-deficient subunit in the mGlu1 dimer did not affect the Gq coupling. (A) Schematic drawings of the combinations of mGlu1-C1 and -C2KKTN constructs (upper panels) and time-lapse changes in the Gq-coupled TRPC3 current (lower traces) in HEK293T cells. Inward currents through the La3þ blockable TRPC3 channels were induced by the indicated concentration of glutamate (mM, gray and black bars on the traces) in cells transfected with WT-C1&WT-C2KKTN (left) and F781S-C1&WT-C2KKTN (right). (B) Averages of response amplitudes of constructs. The presence of one coupling-deficient subunit in the mGlu1 dimer did not affect the Gq coupling of mGlu1. Increases in the current amplitude induced by the indicated concentration of glutamate were normalized to that by 1 mM glutamate (I/Imax). Bars represent the relative amplitudes of glutamate-induced currents (n ¼ 5e8 for each concentration). n.s.: not significant.
M. Tateyama, Y. Kubo / Neuropharmacology 61 (2011) 832e841
837
Fig. 4. A presence of one coupling-deficient subunit in the mGlu1 dimer did not decrease the surface expression but attenuated the Gi/o coupling. (A) Schematic drawings of the combinations of mGlu-C1 and FLAG-tagged WT-C2KKTN constructs (FL-WT-C2). (B) Images of GFP-positive cells and surface expression of the combinations indicated in A. The cell surface expression of constructs was detected immunocytochemically under a non-membrane permeabilized condition. FLAG tag was labeled by FLAG-antibody and then Alexa564-conjugated secondary antibody (lower images). Cells expressing the S196A-GIRK2 channels (GFP-positive cells in the upper images) were selected for further analyses. Scale bars represent 50 mm. (C) Voltage-dependence of the glutamate-induced inward current. Shown are typical currentevoltage relationships, obtained by a ramp protocol, in the absence (black) and the presence (red) of glutamate (1 mM). Currents were recorded from GFP-positive cells co-transfected with the indicated mGlu1 constructs. (D) Quantitative analysis of the surface expression (middle panel) and glutamate-induced GIRK currents density (lower panel). The upper panel indicates the combination of the transfected constructs. Bars represent the fluorescence intensities of the Alexa-564 (n ¼ 26e137 GFP-positive cells, A.U.) and increases in the glutamate-induced GIRK current density (at a holding potential of 80 mV, n ¼ 6e16). *: p < 0.05 between the indicated groups; n.s.: not significant.
the detection limit in cells transfected with FL-WT-C2 constructs (Fig. S3). We also compared the surface expression level of the heterodimers by labeling the surface membrane proteins with a membrane-impermeable biotin reagent. The fraction of the biotin-labeled mGlu1 constructs was not significantly different between cells transfected with F781S-C1&FL-WT-C2 and those transfected with WT-C1&FL-WT-C2 (Fig. S4). Whole cell currents were then recorded from the GFP-positive cells, under the same transfection condition as used for the immunocytochemical experiments. To eliminate possible influences by protein kinase C-dependent phosphorylation of the GIRK2 channel, Ser196 was mutated to alanine (Mao et al., 2004). Upon glutamate application, no increase in the inward currents was observed in cells transfected with FL-WT-C2, whereas the inward rectifying currents were increased in cells co-transfected with WTC1 and FL-WT-C2 (Fig. 4C, red lines). Activation of the heterodimeric receptor, F781S-C1&FL-WT-C2, induced the increases in the GIRK currents, but the current density was significantly smaller than that by the WT combination (Fig. 4D). These inhibitory effects of the coupling-deficient subunit were not observed on the Gq-coupled calcium transient: the maximal D[Ca2þ]i and the
[glutamate]eD[Ca2þ]i relationship were not changed by the coupling-deficient subunit (Fig. 5A). In contrast, the [glutamate]e GIRK current relationship in F781S-C1&FL-WT-C2 was attenuated when compared to that in WT-C1&FL-WT-C2 (Fig. 5B). Taken together, these results suggested that the presence of one couplingdeficient subunit in the mGlu1 dimer significantly attenuates the Gi/o coupling of the receptor but neither the surface expression nor the Gq coupling of the mGlu1. 3.6. The presence of one subunit tagged with YFP in the 2nd loop in the mGlu1 dimer impaired the Gi/o coupling but not the Gq coupling We have previously reported that the fusion of bulky fluorescent protein at the ICDs of mGlu1a disrupts the G-protein coupling without altering the surface expression or the glutamate-binding properties (Tateyama et al., 2004). We thus fused YFP at the 2nd loop of mGlu1-C1 (i2-YFP-C1, Fig. 6A), and investigated its effects on the Gq and Gi/o signaling of mGlu1. The Gq-coupled Ca2þ transients were observed upon glutamate application in cells transfected with i2-YFP-C1 and WT-C2KKTN (Fig. 6A), while the glutamate-induced increase in the GIRK current was not clearly detected (Fig. 6B).
838
M. Tateyama, Y. Kubo / Neuropharmacology 61 (2011) 832e841
Fig. 5. The presence of one coupling-deficient subunit in the mGlu1 dimer attenuated the Gi/o but not the Gq coupling. (A) Averages of D[Ca2þ]i. The amplitudes of glutamateinduced calcium transients were analyzed in cells transfected with WT-C1 and FL-WT-C2 (open square) and with F781S-C1 and FL-WT-C2 (filled square). Bars represent the glutamate-induced increase in D[Ca2þ]i. At all tested concentrations of glutamate, there was no difference in the D[Ca2þ]i (n ¼ 44e180 cells). (B) The normalized [glutamate]eGIRK current relationship. GIRK currents induced by various concentrations of glutamate were recorded to evaluate the doseeresponse relationship. For these analyses, cells of which the maximal GIRK current density was more than 1 pA/pF were selected. Circles represent the normalized amplitude of the glutamate-induced GIRK current (I/Imax). The doseeresponse curve was shifted rightward when a coupling-deficient subunit was included in the mGlu1 dimer. The EC50 (mM) values were 6.03 0.22 for WT-C1&FL-WT-C2 (open circles, n ¼ 7) and 11.26 0.61 for F781S-C1&FL-WT-C2 (filled circles, n ¼ 8). The slope factors were 2.16 0.13 for WT-C1&FL-WT-C2 and 1.84 0.18 for F781S-C1&FL-WT-C2.
These results were consistent with the previous results for transfection with F781S-C1&WT-C2KKTN (Figs. 2F and 5). 3.7. The presence of one coupling-deficient subunit in the mGlu2 attenuated the Gi/o coupling The Gi/o pathway is triggered through trans-activation alone in GABAB receptors (Galvez et al., 2001; Margeta-Mitrovic et al., 2001) and heteromeric mGlu2&4 receptors (Doumazane et al., 2010). In addition, fluorescent proteins fused at various ICDs of the GB1 subunit have been shown not to inhibit the Gi/o coupling of the receptor (Matsushita et al., 2010). Therefore, attenuation of the Gi/o coupling by the coupling-deficient subunit may not be generally observed in mGlus other than mGlu1. We thus investigated the effects of different activation mechanisms of mGlu2 on the GIRK current. Large inward currents were elicited by the application of 100 mM glutamate in cells transfected with WT-mGlu2 (Fig. 7A, top). These responses were not observed in cells transfected with a glutamate-binding attenuated mutant (R57A-mGlu2) (Malherbe et al., 2001) or a coupling-deficient mutant (F756S-mGlu2) (Doumazane et al., 2010) (Fig. 7A and S5). These effects were not due to decreases in the surface expression of the receptors, since R57A and F756S mutations did not inhibit the surface expression of the mGlu2 constructs (Table S1). Co-expression of R57A- and F756S-mGlu2 partially but not fully evoked the glutamate-induced GIRK current (Fig. 7A, bottom). The glutamateinduced current density in the combination was less than 50% compared to that in the WT-mGlu2 (at 100 mM glutamate, Fig. 7B). In addition, the activation time course of the GIRK current by the combination of R57A- and F756S-mGlu2 was clearly slower than that by WT-mGlu2, suggesting that the trans-activation alone is not sufficient for the intact Gi/o coupling of mGlu2. We then analyzed the coupling properties of specific combinations by constructing several mGlu2-C1 and -C2KKTN constructs. Although the mGlu2-C1 constructs were expected not to be expressed on the plasma membrane, glutamate-induced increases in the GIRK currents were observed in cells transfected with the mGlu2-C1 constructs (Fig. S6). Therefore, the mGlu2-C1 constructs could not be used to exclude the functional homodimer of mGlu2 from the plasma membrane. As a second choice, the cDNA content of F756S-mGlu2 was increased to a level 10-fold that of the WT, which would increase the population of the heterodimer without suppressing the expression level of
functional receptors on the plasma membrane (Table S2). We therefore used these constructs to investigate the qualitative aspects of the glutamate-induced GIRK currents (Fig. 8A, upper panels). Under these transfection conditions, the [glutamate]eGIRK current amplitude
Fig. 6. Insertion of YFP into one subunit of the mGlu1 dimer attenuated the Gi/o but not the Gq coupling. (A) Schematic drawings of the combinations of mGlu1 constructs. YFP was inserted into the 2nd intracellular loop of mGlu1-C1 (i2-YFP-C1). (B) Averages of the response amplitudes of mGlu1 constructs. The Gi/o pathway, but not the Gq pathway, was attenuated when YFP was inserted into the intracellular loops of one subunit of the mGlu1 dimer. Left panel: Bars represent the D[Ca2þ]i (n ¼ 58 cells for WT-C1&WT-C2KKTN, n ¼ 84 cells for i2-YFP-C1&WT-C2KKTN). Right panel: Bars represent the densities of glutamate-induced GIRK currents (n ¼ 7 for WT-C1&WTC2KKTN, n ¼ 8 for i2-YFP-C1&WT-C2KKTN). *: p < 0.05 between the indicated groups, n.s.: not significant.
M. Tateyama, Y. Kubo / Neuropharmacology 61 (2011) 832e841
839
Fig. 7. The Gi/o pathway was not fully activated through trans-activation of mGlu2 alone. (A) Schematic drawings of the combination of mGlu2 constructs (left panel) and typical GIRK current recordings (right panel). Shown are typical current traces, at a holding potential of 80 mV, upon application of 100 mM glutamate (white bars) followed by 1 mM glutamate (gray bars). Arrows indicate the current increases evoked by 100 mM glutamate, which were not sufficient to activate R57A-mGlu2 and F756S-mGlu2. (B) Averages of GIRK current density induced by 100 mM glutamate. The Gi/o pathway was partially activated through trans-activation of mGlu2. Bars represent the inward GIRK current densities (n ¼ 7e9). *: p < 0.05.
curve was significantly shifted rightwards by the combination of WTand F756S-mGlu2 (Fig. 8B, filled circle). These results suggested that the Gi/o coupling was also attenuated in mGlu2 when the couplingdeficient subunit was included in the mGlu2 dimer.
4. Discussion 4.1. Dimeric configuration of mGlu The dimeric configuration is a structural and functional characteristic of the family C GPCRs (Pin et al., 2003). In particular, the dimeric configuration of mGlu has been demonstrated both in heterogeneous expression systems and endogenous neuronal cells (Robbins et al., 1999). The time-resolved FRET analysis has also demonstrated that the mGlu1 dimer is not a promiscuous oligomeric form of GPCR such as might be observed in an overexpression system (Maurel et al., 2008). Hydrophobic interaction and the formation of a disulfide bond between the N-terminal ECDs are important to form an mGlu1 dimer (Robbins et al., 1999; Tsuji et al., 2000). Impairments of these interactions disrupted or attenuated the function of mGlu1. A mutation at the dimer interface of ECDs (I120A) did not alter the ligand binding or surface expression of the receptor but disrupted the function (Sato et al., 2003). Treatment of hippocampal slices with a reducing agent of DTT attenuated the function of type I mGlu receptors (Copani et al., 2000). Thus the dimeric configuration of mGlu1 is requisite for its signaling function and potentially confers variation in the intramolecular activation mechanisms.
4.2. Does the multiple activation mechanism of mGlu1 occur under physiological conditions? It has been reported that the mGlu1 subunits become couplingdeficient physiologically when the 2nd intracellular loops are masked by interaction with proteins, such as G protein-coupled receptor kinase (GRK) (Dhami et al., 2005). It could be assumed that one or two GRKs bind to the mGlu1a dimer. Based on the results in the present study, this could serve as a physiological mechanism to change downstream signaling. If two GRKs occupy two subunits of the mGlu1 dimer at the same time, neither the Gq nor the Gi/o could be triggered upon activation of mGlu1a. In contrast, when one GRK interacts with one subunit of the mGlu1a dimer, the Gi/o coupling but not the Gq coupling of the receptor would be attenuated, as the impairment of Gi/o coupling was observed with the insertion of bulky YFP into one subunit (Fig. 6). These differential effects on the downstream signaling might regulate the neuronal transmission, since PTX-sensitive Gi/o signaling via mGlu1 has been suggested to inhibit the voltage gated calcium channels in sympathetic and cerebellar Purkinje neurons (Kammermeier, 2009; Kitano et al., 2003). Further studies are necessary to fully elucidate the physiological significance of the multiple signaling and its regulation. 4.3. The presence of one coupling-deficient subunit in the mGlu dimer attenuated the Gi/o coupling The presence of one coupling-deficient subunit in the mGlu1 or mGlu2 dimer attenuated the Gi/o coupling (Figs. 2F, 4D, 5B, 6B, and 8). These results could be interpreted as meaning that one
840
M. Tateyama, Y. Kubo / Neuropharmacology 61 (2011) 832e841
In conclusion, multiple signaling pathways via the mGlu1 were differentially activated when only the trans-activation was available. Furthermore, the presence of one coupling-deficient subunit in the mGlu1 dimer attenuated the Gi/o coupling but neither the Gq coupling nor the cell surface expression. These results suggested that the intra-molecular activation mechanisms of the dimeric mGlu1 differ depending on the type of G protein. Acknowledgments We would like to thank Drs. J. Miyazaki for the pCXN2 expression vector, S. Nakanishi for the mGlu1a and mGlu2 cDNAs, J. Zhang and M.B. Feller for the ICUE2 cDNA, and L.Y. Jan for the GB1 and GB2 cDNAs. We also thank Drs. Y. Fukata and M. Fukata for their advice on the biotinylation experiments. Thanks are also due to Y. Asai for technical support. This work was supported partly by research grants from the Japan Society for the Promotion of Science (to MT and to YK), and from the Ministry of Education, Science, Sports, Culture, and Technology of Japan (to YK). Appendix. Supplementary material Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.neuropharm.2011.05. 031. Fig. 8. The presence of one coupling-deficient subunit in the mGlu2 dimer attenuated the Gi/o coupling. (A) Schematic drawings of the combination of mGlu2 constructs (upper panel) and typical GIRK current traces (lower panel). The amount of F756SmGlu2 cDNA used for transfection was 10-fold greater than the amount of WT cDNA, to reduce the population of the WT homodimer. Application of the indicated concentrations (mM) of glutamate elicited increases in the inward currents in cells transfected with WT (0.1, left panel) and WT with F756 (0.1:1, right panel). (B) The normalized [glutamate]eGIRK current amplitude relationship. The relative GIRK current amplitude (I/Imax) was plotted against the [glutamate] (n ¼ 4e8 for each point). The EC50 (mM) values were 3.92 0.10 for WT (open circle) and 6.31 0.07 for WT with a10-fold higher amounts of F756S-mGlu2 (filled circle) (p < 0.05). The slope factor was 1.87 0.08 for WT and 1.67 0.03 for WT with 10-fold higher amounts of F756S-mGlu2.
intact G protein-coupling subunit is not sufficient for the mGlu dimer to fully activate the Gi/o pathways. This interpretation is partly consistent with a recent study of dimeric dopamine type 2 receptors, in which the G-protein coupling of the dopamine receptor was shown to be negatively affected by the other receptor of the dimer, which lacked the G-protein coupling (Han et al., 2009). Two intact subunits may be necessary for the mGlu dimer to fully activate the Gi/o pathway. This assumption raised the possibility that two Gi/o proteins simultaneously interact with two subunits of the dimeric receptor for full activation of the downstream pathway. This appears unlikely, however, since two 7-TMDs of the dimeric receptor would not become active at the same time for the activation of G proteins (Brock et al., 2007; Han et al., 2009). One subunit of mGlu5, directly activated by a positive allosteric modulator, has been demonstrated to prevent the other subunit being activated (Brock et al., 2007). Both subunits of mGlu may not become simultaneously active and therefore two Gi/o proteins could not interact with two subunits of mGlu at the same time. As another possibility, both intact subunits in the mGlu dimer might be somehow involved in the coupling with single Gi/o proteins for its full activation. In contrast to the Gi/o coupling, one coupling-deficient subunit did not alter the Gq-coupled calcium transient induced by activation of mGlu1 (Figs. 2F, 3, 5A, and 6B). These differences suggested that the coupling configuration of mGlu1 for Gq is different from that of mGlu1 for Gi/o.
References Abe, H., Misaka, T., Tateyama, M., Kubo, Y., 2003. Effects of coexpression with Homer isoforms on the function of metabotropic glutamate receptor 1alpha. Mol. Cell. Neurosci. 23, 157e168. Brock, C., Oueslati, N., Soler, S., Boudier, L., Rondard, P., Pin, J.P., 2007. Activation of a dimeric metabotropic glutamate receptor by inter-subunit rearrangement. J. Biol. Chem. 282, 33000e33008. Copani, A., Romano, C., Di Giorgi Gerevini, V., Nicosia, A., Casabona, G., Storto, M., Mutel, V., Nicoletti, F., 2000. Reducing conditions differentially affect the functional and structural properties of group-I and -II metabotropic glutamate receptors. Brain Res. 867, 165e172. Dhami, G.K., Babwah, A.V., Sterne-Marr, R., Ferguson, S.S., 2005. Phosphorylationindependent regulation of metabotropic glutamate receptor 1 signaling requires g protein-coupled receptor kinase 2 binding to the second intracellular loop. J. Biol. Chem. 280, 24420e24427. DiPilato, L.M., Cheng, X., Zhang, J., 2004. Fluorescent indicators of cAMP and Epac activation reveal differential dynamics of cAMP signaling within discrete subcellular compartments. Proc. Natl. Acad. Sci. U.S.A. 101, 16513e16518. Doumazane, E., Scholler, P., Zwier, J.M., Trinquet, E., Rondard, P., Pin, J.P., 2010. A new approach to analyze cell surface protein complexes reveals specific heterodimeric metabotropic glutamate receptors. FASEB J. Dunn, T.A., Wang, C.T., Colicos, M.A., Zaccolo, M., DiPilato, L.M., Zhang, J., Tsien, R.Y., Feller, M.B., 2006. Imaging of cAMP levels and protein kinase A activity reveals that retinal waves drive oscillations in second-messenger cascades. J. Neurosci. 26, 12807e12815. Francesconi, A., Duvoisin, R.M., 1998. Role of the second and third intracellular loops of metabotropic glutamate receptors in mediating dual signal transduction activation. J. Biol. Chem. 273, 5615e5624. Francesconi, A., Duvoisin, R.M., 2000. Opposing effects of protein kinase C and protein kinase A on metabotropic glutamate receptor signaling: selective desensitization of the inositol trisphosphate/Ca2þ pathway by phosphorylation of the receptor-G protein-coupling domain. Proc. Natl. Acad. Sci. U.S.A. 97, 6185e6190. Galvez, T., Duthey, B., Kniazeff, J., Blahos, J., Rovelli, G., Bettler, B., Prezeau, L., Pin, J.P., 2001. Allosteric interactions between GB1 and GB2 subunits are required for optimal GABA(B) receptor function. Embo J. 20, 2152e2159. Han, Y., Moreira, I.S., Urizar, E., Weinstein, H., Javitch, J.A., 2009. Allosteric communication between protomers of dopamine class A GPCR dimers modulates activation. Nat. Chem. Biol. 5, 688e695. Hartmann, J., Dragicevic, E., Adelsberger, H., Henning, H.A., Sumser, M., Abramowitz, J., Blum, R., Dietrich, A., Freichel, M., Flockerzi, V., Birnbaumer, L., Konnerth, A., 2008. TRPC3 channels are required for synaptic transmission and motor coordination. Neuron 59, 392e398. Jensen, A.A., Sheppard, P.O., O’Hara, P.J., Krogsgaard-Larsen, P., Brauner-Osborne, H., 2000. The role of Arg(78) in the metabotropic glutamate receptor mGlu(1) for agonist binding and selectivity. Eur. J. Pharmacol. 397, 247e253. Kammermeier, P.J., 2009. C-terminal deletion of metabotropic glutamate receptor 1 selectively abolishes coupling to Galphaq. Eur. J. Pharmacol. 627, 63e68.
M. Tateyama, Y. Kubo / Neuropharmacology 61 (2011) 832e841 Kammermeier, P.J., Yun, J., 2005. Activation of metabotropic glutamate receptor 1 dimers requires glutamate binding in both subunits. J. Pharmacol. Exp. Ther. 312, 502e508. Kitano, J., Nishida, M., Itsukaichi, Y., Minami, I., Ogawa, M., Hirano, T., Mori, Y., Nakanishi, S., 2003. Direct interaction and functional coupling between metabotropic glutamate receptor subtype 1 and voltage-sensitive Cav2.1 Ca2þ channel. J. Biol. Chem. 278, 25101e25108. Kniazeff, J., Bessis, A.S., Maurel, D., Ansanay, H., Prezeau, L., Pin, J.P., 2004. Closed state of both binding domains of homodimeric mGlu receptors is required for full activity. Nat. Struct. Mol. Biol. 11, 706e713. Kunishima, N., Shimada, Y., Tsuji, Y., Sato, T., Yamamoto, M., Kumasaka, T., Nakanishi, S., Jingami, H., Morikawa, K., 2000. Structural basis of glutamate recognition by a dimeric metabotropic glutamate receptor. Nature 407, 971e977. Malherbe, P., Knoflach, F., Broger, C., Ohresser, S., Kratzeisen, C., Adam, G., Stadler, H., Kemp, J.A., Mutel, V., 2001. Identification of essential residues involved in the glutamate binding pocket of the group II metabotropic glutamate receptor. Mol. Pharmacol. 60, 944e954. Mao, J., Wang, X., Chen, F., Wang, R., Rojas, A., Shi, Y., Piao, H., Jiang, C., 2004. Molecular basis for the inhibition of G protein-coupled inward rectifier K(þ) channels by protein kinase C. Proc. Natl. Acad. Sci. U.S.A. 101, 1087e1092. Margeta-Mitrovic, M., Jan, Y.N., Jan, L.Y., 2001. Ligand-induced signal transduction within heterodimeric GABA(B) receptor. Proc. Natl. Acad. Sci. U.S.A. 98, 14643e14648. Matsushita, S., Nakata, H., Kubo, Y., Tateyama, M., 2010. Ligand-induced rearrangements of the GABA(B) receptor revealed by fluorescence resonance energy transfer. J. Biol. Chem. 285, 10291e10299. Maurel, D., Comps-Agrar, L., Brock, C., Rives, M.L., Bourrier, E., Ayoub, M.A., Bazin, H., Tinel, N., Durroux, T., Prezeau, L., Trinquet, E., Pin, J.P., 2008. Cell-surface protein-protein interaction analysis with time-resolved FRET and snap-tag technologies: application to GPCR oligomerization. Nat. Methods 5, 561e567. Muto, T., Tsuchiya, D., Morikawa, K., Jingami, H., 2007. Structures of the extracellular regions of the group II/III metabotropic glutamate receptors. Proc. Natl. Acad. Sci. U.S.A. 104, 3759e3764. Nicoletti, F., Bockaert, J., Collingridge, G.L., Conn, P.J., Ferraguti, F., Schoepp, D.D., Wroblewski, J.T., Pin, J.P., 2010. Metabotropic glutamate receptors: from the workbench to the bedside. Neuropharmacology. Park, P.S., Filipek, S., Wells, J.W., Palczewski, K., 2004. Oligomerization of G proteincoupled receptors: past, present, and future. Biochemistry 43, 15643e15656.
841
Pin, J.P., Galvez, T., Prezeau, L., 2003. Evolution, structure, and activation mechanism of family 3/C G-protein-coupled receptors. Pharmacol. Ther. 98, 325e354. Robbins, M.J., Ciruela, F., Rhodes, A., McIlhinney, R.A., 1999. Characterization of the dimerization of metabotropic glutamate receptors using an N-terminal truncation of mGluR1alpha. J. Neurochem. 72, 2539e2547. Sato, T., Shimada, Y., Nagasawa, N., Nakanishi, S., Jingami, H., 2003. Amino acid mutagenesis of the ligand binding site and the dimer interface of the metabotropic glutamate receptor 1. Identification of crucial residues for setting the activated state. J. Biol. Chem. 278, 4314e4321. Selkirk, J.V., Price, G.W., Nahorski, S.R., Challiss, R.A., 2001. Cell type-specific differences in the coupling of recombinant mGlu1alpha receptors to endogenous G protein sub-populations. Neuropharmacology 40, 645e656. Sheffler, D.J., Conn, P.J., 2008. Allosteric potentiators of metabotropic glutamate receptor subtype 1a differentially modulate independent signaling pathways in baby hamster kidney cells. Neuropharmacology 55, 419e427. Sugiyama, Y., Kawaguchi, S.Y., Hirano, T., 2008. mGluR1-mediated facilitation of long-term potentiation at inhibitory synapses on a cerebellar Purkinje neuron. Eur. J. Neurosci. 27, 884e896. Tateyama, M., Abe, H., Nakata, H., Saito, O., Kubo, Y., 2004. Ligand-induced rearrangement of the dimeric metabotropic glutamate receptor 1alpha. Nat. Struct. Mol. Biol. 11, 637e642. Tateyama, M., Kubo, Y., 2006. Dual signaling is differentially activated by different active states of the metabotropic glutamate receptor 1alpha. Proc. Natl. Acad. Sci. U.S.A. 103, 1124e1128. Tateyama, M., Kubo, Y., 2007. Coupling profile of the metabotropic glutamate receptor 1alpha is regulated by the C-terminal domain. Mol. Cell. Neurosci. 34, 445e452. Tateyama, M., Kubo, Y., 2008. Regulatory role of C-terminus in the G-protein coupling of the metabotropic glutamate receptor 1. J. Neurochem. 107, 1036e1046. Tsuji, Y., Shimada, Y., Takeshita, T., Kajimura, N., Nomura, S., Sekiyama, N., Otomo, J., Usukura, J., Nakanishi, S., Jingami, H., 2000. Cryptic dimer interface and domain organization of the extracellular region of metabotropic glutamate receptor subtype 1. J. Biol. Chem. 275, 28144e28151. Yamashita, T., Terakita, A., Kai, T., Shichida, Y., 2008. Conformational change of the transmembrane helices II and IV of metabotropic glutamate receptor involved in G protein activation. J. Neurochem. 106, 850e859.