Inhibition of metabotropic glutamate receptor signaling by the huntingtin-binding protein optineurin.

Huntington disease is caused by a polyglutamine expansion in the huntingtin protein (Htt) and is associated with excitotoxic death of striatal neurons. Group I metabotropic glutamate receptors (mGluRs) that are coupled to inositol 1,4,5-triphosphate formation and the release of intracellular Ca(2+) stores play an important role in regulating neuronal function. We show here that mGluRs interact with the Htt-binding protein optineurin that is also linked to normal pressure open angled glaucoma and, when expressed in HEK 293 cells, optineurin functions to antagonize agonist-stimulated mGluR1a signaling. We find that Htt is co-precipitated with mGluR1a and that mutant Htt functions to facilitate optineurin-mediated attenuation of mGluR1a signaling. In striatal cell lines derived from Htt(Q111/Q111) mutant knock-in mice mGluR5-stimulated inositol phosphate formation is also severely impaired when compared with striatal cells derived from Htt(Q7/Q7) knock-in mice. In addition, we show that a missense single nucleotide polymorphism optineurin H486R variant previously identified to be associated with glaucoma is selectively impaired in mutant Htt binding. Although optineurin H486R retains the capacity to bind to mGluR1a, optineurin H486R-dependent attenuation of mGluR1a signaling is not enhanced by the expression of mutant Htt. Because G protein-coupled receptor kinase 2 (GRK2) protein expression is relatively low in striatal tissue, we propose that optineurin may substitute for GRK2 in the striatum to mediate mGluR desensitization. Taken together, these studies identify a novel mechanism for mGluR desensitization and an additional biochemical link between altered glutamate receptor signaling and Huntington disease.

Huntington disease (HD) 4 is an autosomal-dominant neurodegenerative disorder manifested by symptoms of involuntary body movement, loss of cognitive function, and psychiatric disturbance, which inevitably leads to death (1)(2)(3)(4). The HD gene mutation consists of an unstable CAG repeat resulting in a polyglutamine expansion in the amino-terminal region of the huntingtin (Htt) protein, a ubiquitously expressed and evolutionary conserved protein (1). It is the polyglutamine expansion of the Htt amino terminus that is proposed to cause progressive widespread neuronal death in the neocortex and the striatum of HD patients. Although the precise function of Htt in cells is not completely understood, analysis of the proteins with which Htt interacts suggests that Htt plays a role in regulating clathrin-coated vesicle-mediated endocytosis, neuronal survival, vesicle transport, morphogenesis, calcium homeostasis, and transcriptional regulation (4).
Glutamate-mediated neurotoxicity has been postulated to play an important role in the pathogenesis and excitotoxic neuronal cell loss in HD (5)(6)(7)(8)(9). The receptors for glutamate are classified into two types: ionotropic and metabotropic (10). The ionotropic receptors comprise cation-specific ion channels that mediate fast excitatory glutamate responses and are subdivided into AMPA/kainate and NMDA receptors. There is a considerable body of evidence to support the idea that glutamate released from cortical afferents may regulate the excitotoxic damage observed in HD by activating both NMDA and AMPA/kainate receptors (5-7, 11, 12). In particular, there is increased sensitivity to NMDA receptor-mediated excitotoxic cell death in a YAC128 transgenic mouse model of HD (11) and NR2A and NR2B receptor gene variations modify the age of HD onset (12). More recently, Group I metabotropic glutamate receptors (mGluRs), which are G protein-coupled receptors (GPCRs) linked to the activation of phospholipase C, increase in intracellular inositol 1,4,5-triphosphate (IP 3 ) formation, and the release of intracellular Ca 2ϩ stores have been proposed to contribute to the underlying pathophysiology of HD. Specifically, the survival of R6/2 HD transgenic mice is significantly increased following treatment with mGluR5 antagonists (13). In addition, mutant Htt and Htt-associated protein 1 interactions with the IP 3 receptor result in altered mGluR5-stimulated Ca 2ϩ signaling (14). This increased Ca 2ϩ release is associated with increased apoptosis of medium spiny striatal neurons derived from a YAC128 transgenic mouse model of HD (15). Thus, glutamate signaling via both ionotropic and metabotropic receptors may be linked with Htt function and HD.
Optineurin (OPTN) is one of a number of recently identified Httinteracting proteins (4). OPTN is a coiled-coiled protein that was first identified as a positive regulator of tumor necrosis factor-␣-mediated apoptosis (16). OPTN interacts with a variety of proteins that link it to the regulation of cellular morphogenesis and membrane trafficking (Rab8), vesicular trafficking (Htt), and transcription activation (TFIIIA) (17)(18)(19). Mutations in OPTN are responsible for ϳ17% of hereditary forms of normal tension glaucoma (20). The link between OPTN, apoptosis, and the retinal degeneration observed in glaucoma has lead to the suggestion that OPTN mutants may directly induce optic neuropathy leading to visual loss (20). This has lead to the suggestion that OPTN represents a common factor involved in protection against neuronal cell death. Here we show that OPTN is also a Group I mGluR-interacting protein that functions to inhibit mGluR G protein coupling to phospholipase C and IP 3 signaling. Moreover, we show that polyglutamine-expanded mutant Htt but not wild-type Htt potentiates the inhibition of mGluR signaling. The association of OPTN with both Htt and mGluRs may provide an additional biochemical link that may help shed light upon the relationship between glutamate-induced neurotoxicity and HD. Yeast Two-hybrid Screening-A human brain cDNA library in the pGAD10 vector, used in the two-hybrid screen, was purchased from Clontech. pGAD10 contains the ADH promoter expressing the GAL4 transactivation domain (amino acids 768 to 881). The yeast reporter strain Y190 harbors two reporter gene constructs. Two-hybrid interactions activate the transcription of the HIS3 gene, allowing screening for growth in the absence of histidine, and of the lacZ gene, allowing screening for ␤-galactosidase activity. Expression of the mGluR1-Ct-GAL4-binding domain fusion protein from pAS2-mGluR1-Ct in yeast strain Y190 was verified by immunoblotting with anti-mGluR1a antibodies (Upstate Biotechnology, Inc., Lake Placid, NY). Y190 was cotransformed with pAS2-mGluR1-Ct and the pGAD10-human brain cDNA library and plated at a density of 5 ϫ 10 4 colonies/plate on synthetic minimal medium lacking leucine and tryptophan (for plasmid selection) and histidine but containing 25 mM 3-aminotriazole. The plates were incubated at 30°C for 5-7 days. Approximately 2.5 ϫ 10 6 yeast transformants were screened. Colonies that grew were transferred to fresh selective plates and assayed for ␤-galactosidase activity by a filter assay. Yeast two-hybrid backcrosses were performed using the yeast reporter strain Y190 (MATa, leu2-3, 112 ura 3-5, trp1-901, his3-D200, ade 2-101) cotransformed with pGAD-OPTN (amino acid residues 202-246), and pAS2 mGluR1a-Ct using a modified lithium acetate method according to the manufacturer's instructions. Double transformants were plated on synthetic yeast drop-out medium lacking leucine, tryptophan, uracyl, and histidine in the presence and absence of 25 mM 3-aminotriazol and incubated 3-5 days at 30°C. Liquid ␤-galactosidase assays were conducted according to standard procedures.

Materials
Plasmid Construction-The FLAG-mGluR1a and amino-terminal myc-tagged and wild-type GRK2 constructs were described previously (21,22). myc-tagged amino-terminal Htt mutant (Q138) and wild-type Htt (Q15) (amino acids 1-704) were constructed by digesting pEGFP-N1-Htt-Q138 1-704 or peGFP-N1-Htt-Q15 1-704, respectively, with BspEI followed by cloning the Htt amino-terminal fragments into pGEX4T1 (Amersham Biosciences) digested with XmaI. The pGEX4T-Htt constructs were then digested with BamHI and NotI. The Htt-containing fragments were cloned into pcDNA3myc. Full-length human OPTN was cloned by PCR amplification using a Quick Clone kit (Clontech). Amino-terminal myc-or HA-tagged wild-type OPTN and HA-tagged OPTN-H486R were constructed by PCR amplification and mutagenesis of human OPTN cDNA and inserted into the BamHI/ EcoRI sites of pcDNA3. The sequence integrity of each of the constructs was confirmed by automated DNA sequencing. The bait for the yeast two-hybrid screen was constructed as follows. The rat mGluR1a carboxyl-terminal tail (amino acids 841-1199) was PCR-amplified and cloned into pAS2-1 (Clontech) using BamHI and EcoRI restriction sites to generate pAS2-mGluR1-Ct. For expression of amino-terminal glutathione S-transferase (GST) fusion proteins in mammalian cells, the cDNA encoding the mGluR1a and angiotensin II type 1A receptor carboxyl-terminal tails (C-tails) were cloned into pEBG using BamHI and NotI restriction sites (23).
Cell Culture and Transfection-HEK 293 cells and COS7 cells were grown at 37°C in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and 100 g/l gentamicin. The mouse striatal cell lines expressing wild-type and mutant Htt were grown at 33°C in Dulbecco's modified Eagle's medium. Cells were transfected with plasmid cDNAs using a modified calcium phosphate method (24). After transfection (18 h), the cells were incubated with fresh medium, allowed to recover for 6 -8 h, reseeded into 6-, 12-, or 24-well dishes, and then grown for an additional 18 h prior to experimentation.
Primary Neuronal Cell Culture, Immunostaining, and Transfection-Neuronal cultures were prepared from cortex obtained from embryonic day-20 rat embryos. All animal procedures were approved by the University of Western Ontario Animal Care Committee. Cells were plated on poly-D-lysine-coated 50-mm glass coverslips in isolation medium for 4 h at 37°C and 5% CO 2 in a humidified incubator to permit cell attachment. Isolation medium was subsequently replaced with serum-free culture medium, and the cells were cultured with medium replenished every 3 days. Isolation medium consisted of minimum Eagle's medium with Earle's salts supplemented with 2 mM Glutamax/ glutamine, 10% heat-inactivated horse serum, 6 mg/ml glucose, 0.5 units/ml penicillin, 0.5 g/ml streptomycin, 10 M MK-801, and 25 mM KCl. Culture medium consisted of neurobasal medium supplemented with B-27, 2 mM Glutamax/glutamine, 0.5 units/ml penicillin, 0.5 g/ml streptomycin, 10 m MK-801, 25 mM KCl. Before experimentation, neurons were washed two times with HEPES-buffered saline solution (HBSS), and all of the experiments were performed using HBSS, which does not contain glutamate. Primary neuronal cultures 4 days in vitro were transfected with Effectene (Qiagen, Hilden, Germany) following the manufacturer's instructions (25). Confocal microscopy was performed using a Zeiss LSM-510 META laser scanning microscope with a Zeiss 63X or 100X NA 1.4 oil immersion lens and filters with emission wavelengths of 488 and 514 nm as described previously (26).
Inositol Phosphate Formation-Inositol lipids were radiolabeled by incubating the cells overnight with 1 Ci/ml Co-immunoprecipitation-Co-immunoprecipitation of endogenous proteins from Hdh Q7/Q7 striatal cells was performed using lysis buffer (25 mM HEPES, pH 7.5, 300 mM NaCl, 1.5 mM MgCl 2 , 0.2 mM EDTA, 0.1% Triton X-100, 20 mg/ml leupeptin, 20 mg/ml aprotonin, and 20 g/ml phenylmethylsulfonyl fluoride). mGluR5 was immunoprecipitated from striatal cell supernatant (500 g of total protein) using 1.5 mg of mGluR5 rabbit polyclonal antibody (Upstate Biotechnology) and 1:75 OPTN rabbit polyclonal antibody. Co-immunoprecipitation experiments were also performed using 500 g of total cell lysate protein solubilized from HEK 293 cells transiently transfected with the various cDNA constructs as described in the legends to Figs. 1-8. The cells were solubilized in lysis buffer. FLAG-mGluR1a and FLAG-mGluR5a were immunoprecipitated with FLAG-Sepharose beads from cell lysates by overnight rotation at 4°C. The beads were washed four times with lysis buffer, and proteins were eluted in 3ϫ SDS sample buffer and then separated by SDS-PAGE. The membranes were blocked with 10% milk in wash buffer (TBS-T) (150 mM NaCl, 10 mM Tris-HCl, pH 7.0, 0.05% Tween 20) and then incubated with rabbit GRK2-, OPTN-, or mGluR5specific antibodies, as well as Htt-, myc-, FLAG-, or HA-specific monoclonal antibodies diluted 1:1000 in wash buffer containing 3% skim milk. The membranes were rinsed three times with wash buffer and then incubated with secondary horseradish peroxidase-conjugated donkey anti-rabbit IgG or anti-mouse IgG (Amersham Biosciences) diluted 1:2500 in wash buffer containing 3% skimmed milk. Membranes were rinsed twice with TBS-T and twice with Tris-buffered saline and were incubated with ECL Western blotting detection reagents. Intensities of immunoblot signals were determined with the ␤4.0.2 version of the data acquisition and analysis software from Scion Corp. (Frederick, MD).
Flow Cytometry Analysis of FLAG-mGluR1a Expression-Cells were placed on ice and washed two times in HBSS. Cells were incubated in a 1:500 solution of mouse anti-FLAG (Sigma) for 45 min. Antibody was washed off and replaced with 1:500 fluorescein isothiocyanate (Sigma) for 45 min followed by two washes in HBSS. Cells were incubated in PBS containing 0.5 mM EDTA for 5 min. Cells were removed from the plates and fixed in 1.6% formaldehyde in PBS. Cell surface immunofluorescence was measured by flow cytometry on a BD Biosciences FACScalibur.
Determination of Cell Viability by MTT Reaction-40 h post-transfection cells were incubated in fresh medium containing 5 mg/ml thiazolyl blue (MTT) (Sigma) for 3 h at 37°C. Viable cells convert the MTT to an insoluble purple formazan. Medium was replaced with dimethyl sulfoxide to solubilize cells, lysates were transferred to cuvettes, and cell viability of transfected cells was analyzed by spectrophotometry at 570 nm. Percentage cell viability was calculated as compared with that of the nontransfected control.
Hoechst Staining-40 h post-transfection cells were washed two times in PBS and incubated for 20 min in 3.6% paraformaldehyde. Cells were washed two times in PBS and incubated in 1:200 Hoechst stain (Sigma) for 45 min and washed twice with PBS. Cells were viewed using an Olympus IX70 inverted fluorescence microscope using a 20X lens with 0.40 NA. Apoptotic cells were defined as having pyknotic nuclei and exhibiting condensed chromatin. We used Northern Eclipse software to process the data presented as the number of apoptotic cells compared with the total number of cells.
Data Analysis-The means Ϯ S.D. or S.E. are shown for the number of independent experiments indicated in the legends to Figs. 1-8. GraphPad Prism software was used to analyze data for statistical significance, as well as to analyze and fit curves for quisqualate dose-response curves. Statistical significance was determined by either the Student's t test or the one-way analysis of variance with the post-hoc Tukey's multiple comparison test.

RESULTS
OPTN Interacts with Group I mGluRs-To identify novel signaling proteins that might interact with Group I mGluRs, we screened a human brain cDNA library using the C-tail domain of mGluR1a (amino acid residues 841-1199) as bait in the yeast two-hybrid system. We identified OPTN as a potential Group I mGluR interacting protein, and the specificity of the interaction was confirmed by growth on ϪLeu/ FIGURE 1. Identification of OPTN as a negative regulator of Group I mGluR signaling. A, yeast two-hybrid backcrosses between the mGluR1aCt fusion and OPTN (amino acid residues 202-246). SNF1/SNF4 interactions served as a positive control, and SNF1/mGluR1a-Ct and SNF4/OPTN crosses served as negative controls. Positive interactions display blue ␤-galactosidase staining. B, representative immunoblot from three experiments showing the co-precipitation of HA-OPTN with a GST fusion encoding the mGluR1a-Ct (GST-mGlu-Ct) but not with either GST or GST fusion encoding the angiotensin II type 1A receptor tail (GST-AT1AR-Ct) from COS7 cells. ϪTrp/ϪHis plates and ␤-galactosidase gene activity in the Y190 cells (Fig. 1A). Fig. 1A shows that the cotransformation of OPTN-(202-246) with the mGluR1a C-tail GAL4-binding domain but not OPTN-(202-246) with the SNF4 GAL4-binding domain fusion protein yielded yeast expressing ␤-galactosidase activity. To verify the specificity of the yeast two-hybrid interaction, we show that HA-tagged OPTN can be coprecipitated with an mGluR1a but not an angiotensin II type 1A receptor, C-tail GST fusion protein (Fig. 1B) and can also be co-immunoprecipitated from HEK 293 cells with either FLAG-tagged mGluR1a or mGluR5a (Fig. 1C). When transfected into primary cortical neurons, green fluorescence protein-tagged OPTN is localized to intracellular vesicular structures in both the soma and the dendritic arbor and is colocalized with internalized FLAG-mGluR1a in these vesicles (Fig.  1D). Taken together, these observations suggest that mGluR1a/OPTN interactions identified by yeast two-hybrid screening may be physiologically relevant.
OPTN Expression Inhibits mGluR1a Signaling-Having established that OPTN binds to Group I mGluRs, we assessed whether OPTN interactions with mGluR1a might influence the ability of the receptor to couple to phospholipase C␤ and stimulate increases in intracellular inositol phosphate concentrations. We find that the overexpression of OPTN in HEK 293 cells reduces the maximum velocity (V max ) for quisqualate-stimulated inositol phosphate formation to 61 Ϯ 8% of control without affecting the half-maximal effective concentration (EC 50 ) for the agonist (Fig. 2A). The prevailing mechanism considered to underlie the attenuation of GPCR signaling involves phosphorylation by G protein-coupled receptor kinases (GRKs) (27,28). However, GRK2 attenuates both agonist-stimulated and basal mGluR1a signaling in the absence of phosphorylation (21,22). Unlike what is observed for GRK2, OPTN expression in HEK 293 cells leads to increased basal mGluR1a activity (Fig. 2B). Thus, OPTN represents an alternative and novel regulator of mGluR signaling that, similar to GRK2, mediates the phosphorylation-independent inhibition of agonist-stimulated mGluR activity.
GRK2 and OPTN Compete for mGluR1a Binding-To test whether GRK2 and OPTN attenuate mGluR1a signaling by similar mechanisms, we examined whether the proteins might compete for mGluR1a binding. We found that, when expressed with OPTN, GRK2 displaces OPTN binding to mGluR1a and is preferentially co-precipitated with the receptor (Fig. 3A). In addition, the V max for quisqualate-stimulated mGluR1a inositol phosphate formation is attenuated by GRK2 to the same extent, independently of whether GRK2 is expressed either with or without OPTN (Fig. 3B). This observation suggests that at maximum expression levels GRK2 and OPTN have nonadditive effects on the inhibition of mGluR1a function. The observation that OPTN expression increases basal mGluR1a activity is likely related to the competition between endogenous GRK2 and OPTN for mGluR1a binding. Interestingly, although endogenous OPTN is expressed at similar levels in cerebellar, cortical, striatal, and hippocampal tissues, GRK2 is expressed at relatively low levels in the striatum (Fig. 3C), the brain region that is susceptible to neuronal degeneration in HD. GRK2 expression levels in the striatum are equivalent to GRK2 expression levels in HEK 293 cells (Fig. 3C). Thus, it is possible that OPTN replaces GRK2 to regulate mGluR signaling in striatal neurons.  Htt Is Associated with mGluR1a-Because OPTN binds to Htt, we tested whether Htt could be co-precipitated in a complex with mGluR1a C-tail in an OPTN-dependent manner. We found that the amino-terminal domain (amino acid residues 1-704) of wild-type Htt (N-Htt Q15 ) can be precipitated with the mGluR1a C-tail in the absence of OPTN overexpression, presumably as a complex with endogenous OPTN that is coprecipitated with the mGluR1a C-tail (Fig. 4A). The overexpression of OPTN increases N-Htt Q15 coprecipitation with the mGluR1a C-tail by 2.7 Ϯ 0.7-fold (Fig. 4A). Full-length wild-type Htt Q15 also co-precipitates with FLAG-mGluR1a from HEK 293 cells (Fig. 4B). However, unlike what is observed for mGluR1a C-tail interactions, OPTN expres-sion does not significantly enhance the co-precipitation of either N-Htt Q15 or polyglutamine-expanded mutant Htt (N-Htt Q138 ) with the full-length FLAG-mGluR1a (Fig. 4C). However, we did observe a small increment in N-Htt Q138 binding to FLAG-mGluR1a (1.7-Ϯ 0.3-fold) following OPTN overexpression. In addition, there is no observable difference in the ability of either N-Htt Q15 or N-Htt Q138 to be co-immunoprecipitated with full-length HA-OPTN (data not shown).
Mutant Htt Enhances OPTN-mediated Inhibition of mGluR1a Signaling-Expression of either N-Htt Q15 or N-Htt Q138 in the absence of OPTN does not lead to a significant change in either the V max or the EC 50 for quisqualate-stimulated mGluR1a inositol phosphate formation  ( Fig. 5A). However, N-Htt Q138 expression increases the OPTN-dependent attenuation of the V max for quisqualate-stimulated mGluR1a inositol phosphate formation and shifts the EC 50 for quisqualate-stimulated inositol phosphate formation from 255 to 2.5 M when compared with cells expressing OPTN alone (Fig. 5B). In contrast, the expression of either wild-type or mutant N-Htt prevents the OPTN-dependent increase in basal mGluR1a activity (Fig. 5C). It is possible that the observed alterations in mGluR1a signaling in the presence of OPTN and N-Htt Q138 are not the consequence of either reduced cell surface receptor expression or increased cell death under the conditions in which we did the inositol phosphate formation experiments. When tested, we found that the overexpression of mGluR1a alone or with either Htt Q15 or N-Htt Q138 in the presence and absence of OPTN had no effect on cell surface FLAG-mGluR1a expression, apoptosis, or cell viability (Fig. 6). Taken together, these results suggest that the increase in the association of N-Htt Q138 and OPTN with mGluR1a augments OPTN-mediated antagonism of mGluR1a signaling rather than altering either cell expression or cell viability.
Impaired mGluR Signaling in Striatal Cells Derived from Hdh Q111/Q111 Mice-Because the co-expression of N-Htt Q138 augmented the attenuation of mGluR1a activity by OPTN, we examined quisqualate-stimulated inositol phosphate responses in striatal neuronal cell lines generated from mice homozygous for the knock-in of a polyglutamine expansion Htt mutant (Hdh Q111/Q111 ) and wild-type Htt (Hdh Q7/Q7 ) littermate embryos (29). Initial experiments demonstrated that both cell lines express equivalent levels of both OPTN and mGluR5 and do not express detectable levels of mGluR1 by immunoblot (Fig. 7A). In contrast, endogenous OPTN expression is not detectable in either HEK 293 cells or COS7 cells (data not shown). However, OPTN expression in both striatal cell lines was higher than COS7 cells transfected to overexpress OPTN (Fig. 7A). We were able to co-immunoprecipitate endogenous mGluR5 with endogenous OPTN from Hdh Q7/Q7 striatal cells (Fig. 7B). Consistent with what we observed in HEK 293 cells, the quisqualate dose response for endogenous mGluR-stimulated inositol phosphate formation was severely impaired in the Hdh Q111/Q111 striatal cells when compared with control Hdh Q7/Q7 striatal cells (Fig. 7C). Specifically, the V max for quisqualate-stimulated inositol phosphate formation was reduced by 39 Ϯ 3% in Hdh Q111/Q111 versus Hdh Q7/Q7 striatal cells, and the EC 50 obtained for quisqualate-stimulated inositol phosphate formation in Hdh Q111/Q111 striatal cells was increased 33-fold to 1.4 versus 43 nM in Hdh Q7/Q7 striatal cells. The inositol phosphate responses following the activation of two other G q -coupled GPCRs (muscarinic acetylcholine or serotonin receptors) were not altered between the two cell lines, suggesting that this does not represent a generalized impairment of phospholipase C signaling (Fig. 7D). Taken together, these observations indicate that mutant Htt overexpression leads to attenuation of Group I mGluR coupling to phospholipase C and inositol phosphate formation. Because this observation is only recapitulated in HEK 293 cells following the overexpression of OPTN with mutant Htt and Hdh Q111/Q111 striatal cells express relatively high levels of endogenous OPTN we suggest the possibility that OPTN may synergize with mutant Htt to attenuate mGluR signaling.
Single Nucleotide Polymorphism OPTN Variant Defective in Htt Binding-The OPTN gene was recently reported to possess both causal and risk-associated alleles for open angled glaucoma (20,30,31). We found that one missense single nucleotide polymorphism OPTN mutant, OPTN-H486R, demonstrated reduced association with N-Htt Q138 (52 Ϯ 20% of wild-type OPTN binding) but was unimpaired in either FLAG-mGluR1a or N-Htt Q15 binding (Fig. 8A). Therefore, we tested whether the H486R mutation might influence OPTN-mediated antagonism of mGluR1a signaling. OPTN-H486R reduced FLAG-mGluR1a-stimulated inositol phosphate formation in response to treatment with 30 M quisqualate to the same extent as wild-type OPTN when expressed either alone or with N-Htt Q15 (Fig. 8B). The expression of N-Htt Q138 increased wild-type OPTN-dependent inhibition of FLAG-mGluR1a-stimulated inositol phosphate formation from 49 Ϯ 2% to 27 Ϯ 1% of control inositol phosphate responses obtained in the absence of OPTN expression (Fig. 8B). However, OPTN-H486R expression significantly reduced the ability of N-Htt Q138 to synergistically antagonize FLAG-mGluR1a signaling (Fig. 8B). On the basis of these observations, we suggest that intact Htt Q138 /OPTN interactions are required for the synergistic antagonism of mGluR1a signaling.

DISCUSSION
In the present study, we provide evidence that OPTN can substitute for GRK2 to mediate the attenuation of mGluR signaling. To our knowledge, OPTN represents the first mGluR-interacting protein other than GRK2 that mediates the phosphorylation-independent uncoupling of mGluRs from heterotrimeric G proteins. OPTN is an Htt-interacting protein suggesting that Htt may be involved in the regulation of mGluR signaling. However, expression of wild-type or mutant huntingtin alone has no effect on mGluR1a coupling to (G q -mediated) increases in inositol phosphate formation. In contrast, when wild-type and mutant Htt are coexpressed with OPTN, only mutant Htt expression leads to an augmentation in OPTN binding to mGluR1a and increased antagonism of mGluR1a signaling. Surprisingly, in cells derived from the striatum of knock-in mice homozygous for the expression of a polyglutamine-expanded Htt mutant, mGluR signaling is severely impaired. Mutant Htt binding to OPTN appears to be essential in mediating the synergistic increase in OPTN-dependent G protein-uncoupling of mGluR1a. Taken together, these observations suggest that not only does OPTN have the capacity to substitute for GRK2 to attenuate mGluR signaling in striatal tissues but mutant Htt protein may further antagonize mGluR signaling in HD.
GRK2 is a ubiquitously expressed protein that is an essential GPCR regulatory protein that protects all cells against receptor overactivation    (27,28). However, GRK2 is most abundantly expressed in hematopoietic and brain tissues, and expression is particularly enhanced at synapses (32). Relative GRK2 expression levels are notably lower in striatal tissue when compared with other brain regions, suggesting that this important GPCR regulatory mechanism may be of limited effectiveness in striatal neurons. We find that GRK2 expression in striatal tissue is equivalent to GRK2 expression in HEK 293 cells, and we have demonstrated previously that mGluR expression leads to HEK 293 cell apoptosis (33). Thus, the observation that OPTN can substitute for GRK2 to mediate the attenuation of mGluR signaling is particularly relevant in the striatum in which OPTN may play a primary role in preventing pathophysiological release of Ca 2ϩ from intracellular stores in response to constitutive and agonist-induced mGluR signaling. Moreover, because OPTN is unlikely to promote ␤-arrestin binding to mGluRs, it is possible that the resensitization of mGluR signaling in the striatum is delayed, especially in the presence of mutant Htt.
The observation that mutant Htt facilitates OPTN activity and mGluR desensitization was surprising because the prevailing expectation was that Group I mGluR signaling via IP 3 leading to intracellular Ca 2ϩ release might be expected to enhance mutant Htt neurotoxicity (8, 9, 34 -36). Htt and Htt-associated protein 1 form a ternary complex with the IP 3 receptor (14), and upon transfection of these proteins into medium spiny striatum neurons Ca 2ϩ release from intracellular stores is significantly increased in response to treatment with threshold levels of the mGluR agonist 3,5-dihydroxyphenylglycine. Consistent with this observation, repetitive application of glutamate to medium spiny neurons derived from YAC128 transgenic mouse model of HD but not control mice also increased intracellular Ca 2ϩ concentrations and induced apoptosis (15). The excitotoxic actions of glutamate in these experiments were mediated by both Group I mGluRs and NR2B glutamate receptors. Thus, it is possible that the uncoupling of mGluR signaling represents an adaptive cellular response that protects against increased IP 3 receptor sensitivity in HD.
Recent studies have demonstrated that single nucleotide polymorphism OPTN variants are associated with hereditary forms of normal tension glaucoma (20,30,31). Alterations in mGluR signaling are also implicated in contributing to the etiology of glaucoma (37). However, it is unknown whether Htt might be involved in regulating retinal degeneration associated with glaucoma. Nonetheless, retina degeneration has been reported in both mouse and Drosophila models of HD, as well as in a small cohort of HD patients (38 -40). Consistent with a potential link between the molecular pathology of HD and glaucoma, we show here that a single nucleotide polymorphism OPTN variant linked to heritable glaucoma, OPTN-H486R, specifically exhibits reduced ability to bind to mutant Htt and antagonize mGluR signaling. Thus alterations in both mGluR signaling and OPTN function may be linked to both HD and glaucoma. Whether a link exists between glaucoma and HD remains to be determined.
Group 1 mGluRs play a dual role in regulating neurotoxicity and neuroprotection (41,42). Thus, as suggested above, the discordance between our observations and previous studies examining the role of polyglutamine-expanded mutant Htt in facilitating IP 3 -stimulated Ca 2ϩ from intracellular stores might be explained by the specific adaptation of mGluR signaling in HD models. It is possible that the uncoupling of mGluR5 from phospholipase C in striatal cells derived from Hdh Q111/Q111 mice represents a compensatory mechanism to protect against altered Ca 2ϩ homeostasis in striatal neurons. Unlike what might be expected for the dysregulation of Ca 2ϩ release in cells transfected with mutant Htt, the reduction in IP 3 formation in Hdh Q111/Q111 striatal cells is specific to mGluR signaling and was not observed for other G q -coupled GPCRs. Alternatively, the increased G protein uncoupling in Hdh Q111/Q111 striatal cells may abrogate the neuroprotective actions that have been described for Group I mGluRs. Thus, a loss of neuroprotective activity may exacerbate excitotoxic signaling by ionotropic glutamate receptors in HD. In summary, understanding the relative role of OPTN and mutant Htt in regulating the neurotoxic and neuroprotective signaling of mGluRs will be essential for the development of new therapeutic targets for the treatment of HD. Our data provide further evidence that mGluR5 is a potential target for pharmacological treatment of HD.