Serine 363 Is Required for Nociceptin/Orphanin FQ Opioid Receptor (NOPR) Desensitization, Internalization, and Arrestin Signaling*

Background: Nociceptin/orphanin FQ opioid receptors (NOPR) are the least understood member of the opioid G protein-coupled receptor family. Results: NOPR serine 363 is required for receptor internalization, desensitization, and c-Jun N-terminal (JNK) phosphorylation. Conclusion: NOPR is regulated via GRK3 and Arrestin3 to control G-protein-dependent and -independent signaling. Significance: Understanding NOPR signaling and regulation could provide novel clues to the development of functionally selective opioid receptor ligands. We determined the role of carboxyl-terminal regulation of NOPR (nociceptin, orphanin FQ receptor) signaling and function. We mutated C-terminal serine and threonine residues and examined their role in NOPR trafficking, homologous desensitization, and arrestin-dependent MAPK signaling. The NOPR agonist, nociceptin, caused robust NOPR-YFP receptor internalization, peaking at 30 min. Mutation of serine 337, 346, and 351, had no effect on NOPR internalization. However, mutation of C-terminal threonine 362, serine 363, and threonine 365 blocked nociceptin-induced internalization of NOPR. Furthermore, point mutation of only Ser-363 was sufficient to block NOPR internalization. Homologous desensitization of NOPR-mediated calcium channel blockade and inhibition of cAMP were also shown to require Ser-363. Additionally, NOPR internalization was absent when GRK3, and Arrestin3 were knocked down using siRNA, but not when GRK2 and Arrestin2 were knocked down. We also found that nociceptin-induced NOPR-mediated JNK but not ERK signaling requires Ser-363, GRK3, and Arrestin3. Dominant-positive Arrestin3 but not Arrestin2 was sufficient to rescue NOPR-S363A internalization and JNK signaling. These findings suggest that NOPR function may be regulated by GRK3 phosphorylation of Ser-363 and Arrestin3 and further demonstrates the complex nature of G-protein-dependent and -independent signaling in opioid receptors.

ber of the opioid receptor family and least understood of the four receptor subtypes. NOPR receptors are widely expressed throughout the brain and spinal cord and are activated by the endogenous peptide nociceptin (1)(2)(3). Through coupling to the G-protein G␣ i/o , NOPR inhibits adenylate cyclase, and decreases N-type Ca V 2.2 calcium conductances (4 -6). The mechanisms of NOPR signal transduction are still under active investigation. NOPR receptor regulation has been only briefly studied and it is known that activation of NOPR with nociceptin produces rapid and robust receptor internalization over time (7). For the other three opioid receptor types it has been established that carboxyl-terminal serine residues are phosphorylated by G protein-coupled receptor kinases (GRK), followed by recruitment of endocytic machinery via the binding of arrestin. This process effectively results in receptor-G-protein inactivation and subsequent receptor internalization (2). However, little is known regarding the mechanisms of NOPR phosphorylation, internalization, and desensitization.
For many GPCR members, it is widely known that sustained agonist activation can lead to receptor phosphorylation and internalization (8). For the NOPR receptor the key amino acid residues, kinetics, and arrestin/GRK proteins involved remain unresolved. In addition, over the last decade, evidence suggests that arrestin-bound GPCR is not inactive but instead recruits a whole host of noncanonical signaling modules (9 -12). For example, for the -opioid receptor and -opioid receptors, ␤-arrestins (also called Arrestin2 and -3) are involved in p38 and ERK mitogen-activated protein kinase (MAPK) activation (10,13,14). Activation and recruitment of the third arm of the MAPK family, c-Jun N-terminal kinase (JNK) pathway has also been shown to require arrestin in some GPCR systems (15)(16)(17). JNK signaling has been demonstrated to play a role in the stress response, apoptotic cell signaling, neurodegeneration, and ischemia (18).
To assess how NOPR receptors are dynamically regulated, desensitized, and whether they activate arrestin-dependent MAPK pathways, we first determined key carboxyl-terminal amino acid residues required for NOPR internalization and desensitization of G-protein signaling. Here we show that the key residue for NOPR regulation is serine 363. We next expressed NOPR-YFP and NOPR-S363A-YFP into HEK293 cells and determined the kinetics and properties of receptor internalization, desensitization of Ca V 2.2 regulation, and cAMP inhibition. Furthermore, we assessed the role of serine 363 in GRK/Arrestin-mediated NOPR activation of ERK and JNK MAPKs. The results suggest that phosphorylation of NOPR at serine 363 by GRK3 acts to facilitate arrestin3 recruitment and receptor internalization, and that arrestin3 is required for NOPR activation of JNK signaling.

Chemicals
Nociceptin and Pertussis Toxin were purchased from R & D Systems (Bristol, UK). All drugs were dissolved in water unless otherwise indicated.

Cell Culture and Transfection of NOPR-YFP Expressing HEK293 Cells
HEK293 cells were grown in Dulbecco's modified Eagle's media/F-12 media supplemented with 10% fetal bovine serum containing 1ϫ penicillin/streptomycin (Invitrogen) and 400 g/ml of G418 to maintain selective pressure in NOPR-YFP expressing cells. HEK293 cells expressing human NOPR-YFP and NOPR-YFP C-terminal mutants were generated as previously described (10). Briefly, stable HEK cell lines expressing pcDNA3 containing NOPR-YFP, NOPR-S337/346/351A, NOPR-T362/S363/T365A, NOPR-T362/S363A, and NOPR-S363A were generated by transfecting HEK293 cells with identical amounts of cDNA (5 g) coding for each NOPR-type for 2-3 h using Superfect (Qiagen) reagent per the manufacturer's instructions and then placing the HEK293 cells under selective pressure with G418 (800 g/ml) for 3 weeks. Colonies of surviving cells were selected and grown into individual 100-mm cell culture plates under 400 g/ml of selective pressure for an additional 2-3 weeks. Cells were then FACS (Washington University FACS Sorting Facility) sorted for equal fluorescence between mutants and wild-type NOPR to further ensure equal receptor expression in each group. Sorted NOPR-YFP expressing HEK293 cells were then grown to confluence into larger T-75 flasks split and cryo-preserved for future use.

Calcium Channel Electrophysiology, Cell Culture, and Transfection
The coding region of human N-type Ca 2ϩ channel Ca V 2.2 subunit was cloned into plasmid pIRES2-EGFP (Clontech), under the CMV promoter. Plasmids encoding mCherry-tagged NOPR receptor (4 g) and Ca V 2.2 subunit (2 g) were co-transfected into stable HEK293 cells expressing the Ca 2ϩ channel auxiliary ␤ 1C and ␣ 2 ␦-1 subunits, using Lipofectamine 2000. One day post-transfection, cells were seeded on Matrigelcoated glass coverslips and recorded 24 -48 h later.
Electrophysiology-Transfected cells were identified by EGFP and mCherry fluorescence. Whole cell patch clamp recordings were performed at room temperature with a Multi-Clamp 700B amplifier (Molecular Devices). pClamp 10 (Molecular Devices) software was used to acquire and analyze data. Cell capacitance and series resistance were constantly monitored throughout the recording. The recording chamber was perfused with extracellular solution (1 ml/min) containing (in mM): 130 NaCl, 2 KCl, 2 CaCl 2 , 2 MgCl 2 , 25 HEPES, 30 glucose, pH 7.3, with NaOH, 310 milliosmole. The pipette solution contained (in mM): 110 CsCl, 10 EGTA, 4 ATP-Mg, 0.3 GTP-Na, 25 HEPES, 10 Tris phosphocreatine, 20 units/ml of creatine phosphokinase, pH 7.3, with CsOH, 290 milliosmole. Recording pipettes had Ͻ3.5 M⍀ resistance. Series resistance (Ͻ15 M⍀) was compensated by 80%. Current traces were corrected with on-line P/6 trace subtraction. Signals were filtered at 1 kHz and digitized at 10 kHz. Cells were held at Ϫ80 mV and depolarized from Ϫ80 to ϩ10 mV for 40-ms pulses every 10 s. After establishing baseline recording, cells were perfused with 1 M nociceptin while evoking Ca 2ϩ currents via depolarizing pulses. In some experiments, transfected cells were incubated in culture medium containing 1 M nociceptin at 37°C for 30 min. Cells were then perfused with extracellular solution for 5 min to wash off nociceptin before conducting whole cell patch clamp recordings. Data were calculated as percent inhibition of calcium currents and plotted.
cAMP Assay-HEK293 cells were transiently co-transfected with pGloSensor-22F cAMP plasmid (Promega E2301) and NOPR-YFP or NOPR-S363A containing plasmids using JetPrime transfection reagent (Polyplus-transfection SA, Illkirch, France) per the manufacturer's instructions. Cells were transfected and plated on 96-well tissue culture-treated plates (Costar) and allowed to recover overnight at 37°C, 5% CO 2 . The next day, media was replaced with 2% GloSensor reagent (Promega) suspended in CO 2 -independent growth medium (Invitrogen) and incubated at room temperature for 2 h. Baseline luminescence recordings were taken and cells were exposed to varying concentrations of nociceptin for 5-10 min before adding 10 M forskolin. Luminescent readings were taken ϳ20 min post drug addition using a SynergyMx microplate reader (BioTek, Winooski VT). Relative luminescent units were normalized to the maximal response evoked by forskolin while in the presence of nociceptin. Subsequent concentrationresponse curves were fit using standard nonlinear regression to obtain IC 50 values using GraphPad Prism (version 5.0d, GraphPad Software, San Diego CA). Triplicate data points were averaged per experiment. Data are expressed as mean Ϯ S.E.

NOPR-YFP Immunocytochemistry (ICC) and Confocal Microscopy
NOPR-YFP and NOPR-YFP C-terminal mutants were grown on poly-D-lysine coverslips in 24-well plates and placed in a 37°C 5% CO 2 incubator. Following drug treatment, cells were washed three times with PBS and then fixed in 4% paraformaldehyde for 20 min, washed in PBS 3 times, blocked for 2 h in 5% normal goat serum, 0.3% Triton X-100 in PBS (blocking buffer) at room temperature, and then incubated overnight with primary antibody in blocking buffer. The following primary antibody concentrations were used: goat anti-rabbit phospho-pJNK MAPK antibody (1:500, Cell Signaling, Beverly, MA) and anti-FLAG M2 antibody was used (1:1000, Sigma). Following primary antibody exposure and a 3 times wash in PBS, coverslips were exposed to secondary antibody, goat antirabbit Alexa Fluor 555 (1:1000, Molecular Probes, Eugene, OR) and goat anti-rabbit IgG 488 Alexa Fluor conjugate (1:400, Molecular Probes) diluted in blocking buffer and incubated with coverslips for 1 h at room temperature. Coverslips were mounted using VECTASHIELD (Vector Laboratories, Burlington, CA) and sealed with clear nail polish. Fluorescent YFP signals were excited at 488 nm, Alexa Fluor 555 fluorescent signals were excited at 543 nm detected and merged as appropriate. All imaging was performed within the Washington University Pain Center Confocal Imaging Center or the Washington University Bakewell Imaging Center. Images, cells, and treatment groups were chosen and analyzed in a blinded fashion. Fields of cells and 3-4 individual cells were chosen at random per coverslip, per treatment, in a triplicate fashion. Experimental "n" reflects an entirely new passage of cells and drug treatment set. Semiquantitative analysis of internalization of NOPR-YFP and mutants was calculated as previously described using Metamorph analysis algorithm for pixel intensity measurements (19) of internalized fluorescence (F) measures. To determine internalized percentages, equal cell shapes and sizes were always chosen; concentric circles around the fluorescence, back-ground internal fluorescence (untreated controls) or internalized (treated) portions of the entire cell were drawn in Metamorph, integrated pixel intensities were recorded for each using the Metamorph algorithm for integrating intensity and internalized receptors were calculated using Inside F/Total F ϫ 100.
Confocal Imaging-Stable HEK293 cells expressing the five different shRNA constructs listed above were plated on 24-well plates with 12-mm poly-D-lysine-coated coverslips (BD Bioscience) containing 2 g/ml of puromycin media at 50% confluence. Although cells are still in suspension, pcDNA3-NOPR-YFP was transiently transfected into cells using JetPrime per the manufacturer's instructions. 48 h after transfection cells were treated with 1 M nociceptin at specified time points. Cells were immediately washed in cold PBS and quickly fixed in cold 4% paraformaldehyde for 15 min while gently rocking. Following the previously outlined ICC protocol, cells were imaged using a FV300 confocal microscope (Olympus) under a ϫ40 oil objective. Representative images were taken for RFP, GRK2, ARR2, GRK3, and ARR3 at different time points and analyzed using Metamorph. For immunoblot experiments stable HEK293 cells expressing the five different shRNA constructs listed above were plated on 24-well plates at 50% confluence in 2 g/ml of puromycin media. Although cells are still in suspension pcDNA3-NOPR-YFP was transiently transfected into cells using JetPrime per the manufacturer's instructions. 48 h after transfection cells were treated with 1 M nociceptin at specified time points.
qRT-PCR-Total RNA was isolated using the RNAgem Tissue Plus kit (Zygem) and One Step qRT-PCR was performed on a Applied Biosystems 7500 real time cycler using QuantiTect SYBR Green PCR kit and QuantiTect primers targeting human GAPDH, GRK2, ARR2, GRK3, and ARR3 (Qiagen). C t values were collected in triplicate and relative ⌬⌬C t were analyzed using REST 2009 software (Qiagen).
Immunoblotting-Western blots for phospho-MAPKs were performed as described previously (9). Briefly, NOPR-YFP and NOPR-S363A-YFP (NOPR-S363A) expressing HEK293 cells were cultured as described above. Cells were serum-starved a minimum of 4 -6 h prior to drug treatment to avoid serum growth factor-induced MAPK activation. Cells were treated with drugs at various time points in cell culture medium at 37°C and then lysed in 350 l of lysis buffer containing 50 mM Tris-HCl, 300 mM NaCl, 1 mM EDTA, 1 mM Na 3 VO 4 , 1 mM NaF, 10% glycerol, 1% Nonidet P-40, 1:100 of phosphatase inhibitor mixture set 1 (Calbiochem), and 1:100 of protease inhibitor mixture set 1 (Calbiochem). Lysates were sonicated for 20 s and then centrifuged for 15 min (14,000 ϫ g, 4°C), the pellet was discarded, and sample supernatants were stored at Ϫ20°C. Protein concentration was determined by a Pierce BCA (Thermo Scientific) assay with bovine serum albumin as the standard. 20 g of total protein was loaded onto nondenaturing 10% bisacrylamide precast gels (Invitrogen) and run at 150 V for 1.5 h. For determination of molecular weights prestained molecular weight ladders (Invitrogen) were loaded along with protein samples. Blots were transferred to nitrocellulose (Whatman, Middlesex, UK) for 1.5 h at 30 mV, blocked in TBS, 5% bovine serum albumin for 1 h, incubated overnight at 4°C with a 1:1000 dilution of goat anti-rabbit phospho-pJNK MAPK antibody or goat anti-rabbit phospho-ERK 1/2 (Thr-202/Tyr-204) antibody (Cell Signaling) or mouse ␤-actin (1:5000, Abcam). Following overnight incubation, membranes were washed 4ϫ for 15 min in TBST (Tris-buffered saline, 1% Tween 20) and then incubated with the IRDye TM 800-and 700-conjugated affinity purified anti-rabbit or anti-mouse IgG at a dilution of 1:20,000 in a 1:1 mixture of 5% milk/TBS and Li-Cor blocking buffer (Li-Cor Biosciences, Lincoln, NE) for 1 h at room temperature. Membranes were then washed 4ϫ for 15 min in TBST, 1ϫ for 10 min in TBS to remove Tween 20 (which can cause high background fluorescence on the Odyssey imaging system), and analyzed as described below.

Data Analysis
All Immunoblots were scanned using the Odyssey infrared imaging system (Li-Cor Biosciences). Band intensity was measured using Odyssey software following background subtraction and integrated band density in high-resolution pixels was calculated. For both pERK and pJNK blot quantitation, all subtypes of ERK (1 and 2) and pJNK (1, 2, 3) were quantified together, as there was no evidence using these epitope antibodies that JNK subtypes 1, 2, and 3 were differentially regulated by NOPR stimulation. All pERK/pJNK bands were normalized to ␤-actin, as an equal protein loading control. Data were then calculated to percentage of control or vehicle sample band intensity (basal, 100%) and plotted using GraphPad (GraphPad Prism 5.0) software. Concentration-response data were fit using nonlinear regression (Prism 5.0). Statistical significance was taken as p Ͻ 0.05, p Ͻ 0.01, or p Ͻ 0.001 as determined by the Student's t test or analysis of variance (ANOVA), followed by Dunnett's or Bonferroni post hoc tests where appropriate. Specific statistical tests and results are indicated in figure legends.

Serine 363 Is Required for Nociceptin-induced NOPR
Internalization-Previous studies with -, -, and ␦-opioid receptors have shown that C-terminal serines are crucial for mediating receptor desensitization and internalization (2). However, the key residue(s) involved in NOPR regulation have not been defined. Therefore, we generated mammalian expression constructs for YFP-tagged human NOPR and used site-directed mutagenesis to construct C-terminal NOPR-YFP serine/threonine to alanine mutants. Based on the highly conserved C-terminal cDNA sequence alignment between human, mouse, and rat ( Fig. 1A) we began by mutating groups of putative serine/threonine amino acid residues and screening for changes in nociceptin-induced (1 M) internalization (Fig.  1B). In HEK293 cells expressing NOPR-YFP, nociceptin caused a robust time-dependent receptor internalization peaking at 60 -90 min following agonist treatment (Fig. 1, B and C).
Mutation of C-terminal serine 337, 346, and 351 had no effect on nociceptin-induced (1 M, 60 min) NOPR internalization, with comparable internalization to the wild-type NOPR-YFP receptor (Fig. 1D). In contrast, mutation of the threonine 362, serine 363, threonine 365 sequence (TST) significantly blocked nociceptin-induced (1 M, 60 min) NOPR-YFP internalization (Fig. 1D) as did mutation of only the threonine 362serine 363 sequence. Finally, mutation of only serine 363 to alanine also caused a significant block of nociceptin-induced NOPR internalization (Fig. 1E) (**, p Ͻ 0.01 NOPR mutants versus NOPR-YFP, ANOVA, Dunnett's post hoc, n ϭ 4 independent experiments performed in triplicate sets). Together, these data suggest that mutation of serine 363 to alanine in NOPR is sufficient to block the ability of the endogenous agonist of nociceptin to cause receptor internalization, and strongly implicate this residue as a crucial site for GRK phosphoregulation of NOPR.
Serine 363 Is Required for NOPR Desensitization-Previous reports have demonstrated that NOPR receptors couple to inhibition of N-type Ca V 2.2 calcium currents and readily desensitize in a time-dependent manner in sensory neurons (5,6). To determine the pharmacological properties of NOPR and NOPR-S363A-mediated G-protein-mediated signaling and receptor desensitization, we measured the effects of nociceptin on NOPR-mediated cAMP inhibition and blockade of N-type Ca V 2.2 calcium currents in both NOPR-YFP and NOPR-S363A HEK293 cells. The Ca V 2.2 subunit of N-type calcium channels expressing the auxiliary ␤1C and ␣2␦-1 subunits were coexpressed with NOPR-YFP. We used patch clamp recording to measure whole cell Ca 2ϩ currents through N-type channels. The transfected cell was held at Ϫ80 mV and depolarized to ϩ10 mV for 40 ms to elicit inward current. Perfusion of 1 M nociceptin resulted in similar and significant inhibition of whole cell Ca 2ϩ currents in both NOPR-YFP and NOPR-S363A expressing HEK293 cells. Nociceptin induced comparable inhibition of Ca 2ϩ currents as previously reported (Fig. 2, A and B) (5). In contrast, pretreatment of cells with 1 M nociceptin (30 min, 37°C) prior to recording caused robust and significant desensitization (p Ͻ 0.001, n ϭ 11) of subsequent nociceptinmediated inhibition of Ca 2ϩ currents in NOPR-YFP expressing HEK cells (Fig. 2, A and B). However, in cells expressing NOPR-S363A, the same nociceptin pretreatment did not result in any subsequent desensitization of nociceptin-induced inhibition of Ca 2ϩ currents (Fig. 2, A and B).
We next determined whether mutation of serine 363 would also impact NOPR-induced inhibition of cAMP. Both NOPR-YFP and NOPR-S363A inhibited cAMP with equal potency and efficacy, suggesting that C-terminal serine 363 is not required for G-protein coupling to adenylate cyclase inhibition (IC 50 for

Ser-363 in NOPR Regulation and Signaling
nociceptin-induced inhibition of cAMP at NOPR-YFP ϭ 0.704 Ϯ 0.12 nM, at NOPR-S363A ϭ 0.69 Ϯ 0.17 nM) (Fig. 3C). Treatment with 1 M nociceptin resulted in similar and significant inhibition of cAMP in both NOPR-YFP and NOPR-S363A expressing cells (Fig. 3, C and D). In contrast, pretreatment of cells with 1 M nociceptin (30 min, 37°C) prior to a subsequent nociceptin (1 M) challenge caused robust and significant desensitization (p Ͻ 0.01, n ϭ 4 -13) of nociceptinmediated inhibition of cAMP in NOPR-YFP expressing cells (Fig. 3D). However, in cells expressing NOPR-S363A, the same nociceptin pretreatment did not result in any subsequent desensitization of nociceptin-induced inhibition of cAMP ( Fig.  3D) (n ϭ 4 -13, p Ͻ 0.05). Taken together, these data suggest that mutation of serine 363 to alanine prevents NOPR-mediated desensitization to both G-protein ␣ (cAMP) and ␤␥ subunit (Ca 2ϩ channel) signaling, and further support the conclusion that serine 363 is a crucial C-terminal residue involved in the functional regulation of NOPR.
To assess the role of GRKs in NOPR internalization we knocked down expression of GRK2 and GRK3 and examined nociceptin-induced internalization of NOPR (1 M, 60 min). Nociceptin caused significant NOPR-YFP internalization (n ϭ 3, p Ͻ 0.05) in GRK2 siRNA expressing cells, but not in cells expressing GRK3 siRNA (Fig. 3, A and B). We used quantitative RT-PCR (qRT-PCR) of mRNA for GRK2 and GRK3 to confirm significant (Ͼ50%, p Ͻ 0.001) knockdown of mRNA for each kinase (Fig. 3C). We next determined the effect on NOPR internalization by using siRNA to knockdown arrestin2 and arres-tin3. Nociceptin (1 M, 60 min) caused significant internalization (n ϭ 3, p Ͻ 0.05) in Arrestin2 siRNA expressing cells, but not in cells expressing Arrestin3 siRNA (Fig. 3, D and E). Again, we used qRT-PCR of mRNA for arrestin2 and arrestin3 to confirm significant (Ͼ50%, p Ͻ 0.001) knockdown of expression for each arrestin subtype (Fig. 3C). Control siRNA (off target) transfected cells showed normal internalization of NOPR-YFP following agonist treatment, further confirming the selectivity of our GRK and arrestin siRNA knockdown approach (Fig. 3F).  DECEMBER 7, 2012 • VOLUME 287 • NUMBER 50

JOURNAL OF BIOLOGICAL CHEMISTRY 42023
Together, these data suggest that NOPR internalization requires GRK3 and arrestin3.
Nociceptin Induces Arrestin3 but Not Arrestin2 Membrane Translocation-We also examined whether nociceptin could preferentially induce membrane translocation of arrestin3 to further establish selective interaction of the nociceptin-engaged NOPR with the arrestin3 isoform. In NOPR-YFP expressing cells transfected with FLAG-arrestin3, arrestin3 dis-tribution is diffuse and cytosolic in untreated cells (Fig. 4A), but labeling becomes punctate and increasingly at the membrane after nociceptin treatment (1 M, 15, 30 min, n ϭ 3), indicative of specific recruitment to the membrane (Fig. 4A); whereas in NOPR-YFP cells transfected with FLAG-arrestin2, nociceptin failed to induce a shift in localization of arrestin2 to the membrane (1 M, 15, 30 min, n ϭ 3). Furthermore, in cells expressing NOPR-S363A and FLAG-Arrestin3 show a similar lack of significant membrane translocation of FLAG-Arrestin3 following nociceptin treatment at any time point (Fig. 4) (n ϭ 3).
We also utilized a novel arrestin3 mutant recently characterized as deficient in binding to both activated and inactive receptors to determine whether we could visualize its stable translocation at the cell membrane. This FLAG-arrestin3-KNC has 2 key phosphate-binding lysines and 10 residues that engage receptor elements all mutated to alanines (15). Interestingly, in cells transfected with both NOPR-YFP and FLAG-arrestin3-KNC, nociceptin treatment (1 M, 15, 30 min, n ϭ 3) did not promote translocation from the cytosol to the membrane of this FLAG-arrestin3-KNC, suggesting that binding of arrestin3 to NOPR may in part stabilize the arrestin3 at the membrane (Fig. 4 C). Together, these data suggest that arrestin3 translocates to the membrane in NOPR cells in response to agonist and further supports arrestin3 as a key isoform in nociceptin-induced NOPR regulation.

NOPR Serine 363 to Alanine Mutation Prevents Biphasic Nociceptin-induced c-Jun N-terminal Kinase Phosphorylation (pJNK)-Recent studies have shown that GPCRs activate noncanonical signaling pathways including MAPK cascades in both
G-protein-dependent and -independent manners in biphasic patterns (10,13,21,22). Treatment of cells expressing NOPR-YFP with the selective NOPR agonist nociceptin caused a timeand concentration-dependent increase in pJNK (Fig. 5, A, C, and G) that was completely absent in untransfected control HEK293s (data not shown, n ϭ 3). However, HEK293 cells expressing mutant NOPR-S363A did not show a concentration-dependent increase in pJNK at 15 min and onward at time points following nociceptin treatment (Fig. 5, B, C, and G). The differences in response were not due to differences in expression (see "Experimental Procedures") because NOPR-YFP and NOPR-S363A HEK293 cells were generated to express equal levels of receptor using FACS sorting of equal surface fluorescence. Surprisingly, both NOPR-YFP and NOPR-S363A show equivalent nociceptin-induced ERK1/2 phosphorylation (pERK) in time-and concentration-dependent manners (Fig. 5,  D, E, F, and H) and equal potency and efficacy in cAMP inhibition. In addition, ERK and pJNK were activated with similar potency and kinetics in NOPR (non-YFP tagged) expressing HEK cells suggesting that the C-terminal YFP tag does not interfere with signaling of the receptor through these pathways (data not shown).
Dominant-Positive Arrestin3, but Not Arrestin2 Rescues NOPR-S363A-mediated pJNK and Receptor Internalization-To assess whether NOPR-S363A fails to activate JNK because agonistbound NOPR-S363A is unable to recruit arrestin and internalize as effectively as NOPR-YFP, we transfected the dominant-positive forms of arrestin3-FLAG (DP-Arr3) and arrestin2-FLAG (DP-Arr2) into NOPR-S363A HEK293 cells. This form of arrestin has FIGURE 2. Serine 363 in NOPR is required for agonist-induced receptor desensitization of coupling to N-type Ca 2؉ channels and inhibition of cAMP. A, representative N-type Ca 2ϩ current traces in response to square pulse depolarization (from Ϫ80 to ϩ10 mV for 40 ms) before and after 1 M nociceptin application in the exemplar cell from each experimental group. B, percentage of nociceptin-induced inhibition of Ca 2ϩ current (I Ca ) in cells expressing wild-type NOPR or the S363A mutant, with or without pre-treatment of 1 M nociceptin at 37°C for 30 min, respectively (n ϭ 8 -10 cells in each group, ***, p Ͻ 0.001, two-way ANOVA with post hoc Bonferroni test). C, concentration-response curves for nociceptin-induced inhibition of cAMP in NOPR-YFP (black) and NOPR-S363A expressing HEK293 cells (red) (n ϭ 5-10). IC 50 for nociceptin-induced inhibition of cAMP at NOPR ϭ 0.704 Ϯ 0.12 nM, at NOPR-S363A ϭ 0.69 Ϯ 0.17 nM. D, percentage of nociceptin-induced inhibition of cAMP in cells expressing wild-type NOPR or the S363A mutant, with or without pretreatment of 1 M nociceptin at 37°C for 30 min, respectively (n ϭ 4 -13, *, p Ͻ 0.05; **, p Ͻ 0.01; ***, p Ͻ 0.001, two-way ANOVA with post hoc Bonferroni test).
In NOPR-S363A expressing cells transfected with DP-Arr3 and then stimulated with 1 M nociceptin, the NOPR-S363A receptor was able to internalize and cause activation of pJNK (Fig. 6, A-D) in a similar fashion to NOPR-YFP receptors treated with nociceptin (1 M, 30 min, n ϭ 3). Successful transient transfection of DP-Arr3 was confirmed by increased FLAG immunolabeling in NOPR-S363A cells (ϳ20 -30% of total HEK population). In contrast, transfection of DP-Arr2 into NOPR-S363A expressing cells failed to rescue internalization (n ϭ 3). In addition, control transfection of wild-type arres-tin3 failed to induce a significant increase in NOPR-S363A activation of pJNK following nociceptin treatment (Fig. 6, B and D).
The effect of the DP-Arr3 transfection on NOPR-S363Amediated pJNK 1/2/3 activation was also quantified using immunoblotting. We found that transfection of DP-Arr3 activation following treatment with nociceptin significantly (*, p Ͻ 0.05, test, n ϭ 4 -6) restored late phase NOPR-S363A-mediated JNK signaling (Fig. 6B). Taken together, these results suggest that serine 363 phosphorylation and the subsequent recruitment of arrestin are required for maximal NOPR-YFP internalization and activation of pJNK.

NOPR-induced pJNK Requires Both G␣ i -protein-dependent and -Independent Arrestin3
Pathways-Because we found a biphasic phosphorylation profile for NOPR-induced JNK activation (Fig. 4) we determined if each phase was either G-protein or arrestin dependent. We found that treatment with pertussis toxin, an inhibitor of G␣ i signaling via ADP-ribosylation (300 ng/ml, 18 h) significantly blocked (p Ͻ 0.05) nociceptin-induced pJNK (1 M) at the 5-min time point in both NOPR-YFP and NOPR-S363A expressing cells (Fig. 7, A and B) (n ϭ 6 -11). These data suggest that the early phase of NOPR-mediated JNK signaling requires G␣ i . We next examined the second phase of NOPR pJNK signaling using siRNA knockdown of GRK2, GRK3, arrestin2, and arrestin3. Interestingly, we found that late phase (30 min) nociceptin-induced pJNK required GRK3 and arrestin3, but not GRK2 and arrestin2, because knockdown of GRK3 and arres-tin3 significantly attenuated (p Ͻ 0.05) nociceptin-induced pJNK activation in NOPR-YFP expressing cells (Fig. 7, C and  D). In addition, transfection of control siRNA had no effect on nociceptin-induced pJNK (Fig. 7, C and D). Together, these data suggest that GRK3 and arrestin3 are required for late phase nociceptin-induced pJNK activation, and further highlight the biphasic, bimodal nature of NOPR regulation and kinase signaling.

DISCUSSION
The principle finding in this study was that NOP receptors are dynamically regulated through interactions at their carboxylterminal serine 363. Mutation of serine 363 to alanine prevented agonist-induced internalization, desensitization, and NOPR-mediated arrestin-dependent JNK MAPK signaling. Preincubation with nociceptin caused homologous NOPR desensitization of inhibitory cAMP signaling and Ca 2ϩ channel inhibition. This homologous desensitization was absent in NOPR-S363A mutant receptor expressing cells. We also found that siRNA knockdown of GRK3 and arrestin3 prevented nociceptin-induced NOPR-YFP internalization. Additionally, we show that NOPR-induced JNK phosphorylation requires serine 363, GRK3, and arrestin3. These findings suggest that NOPR activation of pJNK may require GRK3 phosphorylation of serine 363 in the carboxyl-terminal domain of NOPR and ensuing association with the scaffold protein arrestin3. This proposed model (Fig. 8) is consistent with previous data suggesting that arrestin binding to GPCRs may enable MAPK activation, and acts to regulate receptor function (10,20,22).
Agonist-induced receptor desensitization is a primary mechanism for the subsequent internalization and down-regulation of opioid receptor G-protein signaling (2,28,29). In many cases, one crucial distal end carboxyl-terminal serine residue is required for regulation of -, -, and ␦-opioid receptor function. Here we show in a similar fashion, that serine 363 is a crucial amino acid within NOPR necessary for both agonistinduced receptor internalization and homologous desensitization. We found that agonist-induced NOPR-YFP internalization peaked within 30 min, and that 30-min pretreatment of nociceptin prevents subsequent NOPR-mediated cAMP inhibition, as well as N-type Ca V 2.2 channel inhibition. This desensitization is consistent with the recently reported, nociceptininduced loss of NOPR-mediated N-type Ca V 2.2 channel inhibition in populations of sensory neurons (6). Our data provide novel information as to the specific GRK and arrestin likely to mediate these effects. However, further study in similar types of NOPR-expressing neuronal cultures and in vivo studies will be necessary to fully validate the NOPR regulatory properties we defined here.
We did not directly measure the ligand-induced phosphorylation state of NOPR nor did we directly assess the kinetics of the putative nociceptin-induced GRK3-mediated receptor phosphorylation; however, given our mutagenesis results, we now have a more complete picture of the key amino acid epitope likely to be important for NOPR regulation and phosphorylation. Additionally, knowing that serine 363 is necessary for NOPR regulation we are now equipped to generate phosphoselective C-terminal NOPR antibodies for discerning NOPR regulation in endogenous neuronal systems, as have been used for bothand -opioid receptors (29 -31). It is also important to note that the correlation between agonist-induced receptor phosphorylation, internalization, and desensitization is not always a direct relationship due to differences in ligand-receptor interactions and cell type. Thus, it will be necessary in future studies to fully characterize the kinetics of NOPR phosphorylation as they relate to the data in this study, and other reports, suggesting that rapid nociceptin-induced NOPR internalization is possible (Fig. 1B) (7).
Identifying the regulatory and desensitization mechanisms of NOPR will allow for a better understanding of how the NOP receptor communicates with N-type calcium channels in vivo, and how NOPR function is regulated following repeated nociceptin exposure. The specific mechanisms of NOPR regulation of N-type Ca 2ϩ channel function remains an active debate (5,6), although it is agreed upon by both groups that NOPR activation produces a 40% reduction in N-type Ca 2ϩ currents. (5,32). Our data in the present study corroborate these results and extend them by showing that NOPR serine 363 is an important player in desensitization of this G␤␥-mediated signaling path-

Ser-363 in NOPR Regulation and Signaling
way. Additionally, we found that NOPR-mediated G␣ i signaling to cAMP is dynamically regulated via this same serine 363. Given that NOPR receptors have been implicated in stress signaling, pain, and affective behavior, our molecular analysis should provide some foundation for further studies exploring endogenous nociceptin-NOPR activity and down-regulation in vivo.
Activation of MAPK signal transduction by GPCRs has been demonstrated for numerous GPCR classes and ligand types, in a wide variety of cell lines and endogenous systems (22,(33)(34)(35)(36). NOPR-mediated MAPK activation has not been examined in detail, although evidence does suggest that NOPR can activate ERK and p38 signaling (37,38). There is some evidence for this NOPR-mediated MAPK signaling to involve PKA and PKC, although future work has not followed up on these data. Evidence for NOPR-induced JNK activation has also been reported, with interesting data suggesting that NOPR may activate JNK via multiple mechanisms including pertussis toxinsensitive and -insensitive G-proteins, reminiscent of -opioidmediated pJNK signaling (39,40). Our results corroborate these data, showing similar nociceptin-induced pJNK kinetics and pertussis toxin sensitivity for early phase NOPR-mediated pJNK activation. However, we extend these findings and report that late phase NOPR-mediated pJNK requires GRK3 and arrestin3, suggesting that serine 363 is necessary for recruit-ment of this alternative NOPR-pJNK complex. Future studies will need to determine the additional proteins within this network, as well as the various microdomain differences between G-protein-dependent and -independent NOPR-induced JNK activation.
The function of arrestin in regulating opioid receptor desensitization, endocytosis, and phosphorylation is well established for the 3 original opioid receptor types (2). More recently, reports have shown that arrestins are necessary for ligand-directed signaling and functional selectivity at opioid receptors, and that they mediate vital opioid receptor-dependent behavioral effects (12,20,41). Our findings here suggest that NOPR shares some of these properties because we found siRNA knockdown of arrestin3 prevented NOPR-mediated pJNK. Dominant-positive arrestins have previously been reported to bind to the agonist-occupied receptor and mediate desensitization in the absence of receptor phosphorylation providing a useful tool for assessing the requirement of arrestin in signal transduction (10,23). In these reports, ␦and -opioid receptor phosphorylation-insensitive C-terminal mutants were internalized and desensitized in a similar way as wild-type when these dominant-positive arrestins were co-expressed. Using NOPR-S363A-YFP co-transfected with a dominant-positive form of arrestin3 (Fig. 6A), we found that nociceptin was able to cause internalization of NOPR-S363A and also cause activation  DECEMBER 7, 2012 • VOLUME 287 • NUMBER 50 of JNK. These data support the concept that NOPR-mediated JNK activation requires arrestin. Interestingly, a recent report has suggested that certain JNK isoforms (JNK3) can become activated via arrestin scaffolding in the absence of receptor (16). Understanding the biochemical properties of NOPR-induced phosphorylation of specific JNK isoforms will also be a key next step.

Ser-363 in NOPR Regulation and Signaling
Arrestin-mediated receptor endocytosis has been described for several GPCR classes, the rate of which depends on ligand, receptor, and cell type. Here we showed that NOPR internalization requires arrestin3 when the receptor is stimulated by nociceptin. However, it is important to note that arrestin2 may also interact with NOPR under different circumstances, includ-ing via other NOPR-selective agonists, or cell systems. The effects of GPCR functional selectivity are now widely reported, and our findings here open tantalizing possibilities for ligand bias at NOPR. Similarly, we identified a critical role for GRK3 and not GRK2 in NOPR endocytosis using siRNA technology. Although we did achieve significant knockdown of GRK2 and arrestin2 mRNA we cannot completely achieve full mRNA knockdown of each using this technology, so it is also plausible   6 -11, p Ͻ 0.05, t test). C, representative Western blots of nociceptin-induced pJNK (1 M, 30 min) in NOPR-YFP or NOPR-S363A co-expressed with control siRNA or siRNA for arrestin2 or arres-tin3. Western blots show increased nociceptin-induced pJNK in control and arrestin2 knockdown, but no increase in pJNK in arrestin3 knockdown. Bottom, ␤-actin (actin) loading control confirming equal protein loading in each sample. D, nociceptin-induced pJNK Ϯ S.E. data of NOPR-YFP-mediated pJNK in the presence and absence of various siRNAs including control, GRK2, GRK3, arrestin2, and arrestin3 (n ϭ 5, ***, p Ͻ 0.001).

Ser-363 in NOPR Regulation and Signaling
that residual expression may still play a role in NOPR regulation. Furthermore, we cannot completely rule out a role for GRK2 in regulation of NOPR when exposed to different ligands or in other cell types expressing different ratios of regulatory proteins. This ligand-dependent GRK subtype recruitment is consistent with some very recent studies examining -opioid and ␤-adrenergic receptor function (42,43). Additionally, arrestin scaffolding of the endocytic machinery is also well established. It is known that arrestin-clathrin interactions and Rab-GTPases are critical for these effects (44). Understanding how the NOPR-arrestin binding complex interacts with endocytic machine, and the kinetics of receptor degradation-recycling will be interesting avenues of further investigation.
The JNK MAPK pathway has been demonstrated to play a major role in environmental stress, inflammation, and cytokine pathway activation (45). JNK pathway activation has been recently shown to be involved in neuropathic pain responses in mice (46), and is well established to play a critical role in neurodegenerative disease progression (47). Both pathologies have also been linked to NOPR receptors in various contexts (48,49) suggesting that NOPR-mediated pJNK signaling could have important in vivo implications; however, conclusions based from transfected cells are limited, and overexpression studies may force signal transduction that is not evident physiologically. Regardless, our results presented here are internally consistent and the behavioral in vivo implications of this process will require new investigations.
The NOPR crystal structure was just recently solved using an antagonist peptide mimetic (compound-24) bound to the NOPR receptor (50). This important study carefully models the NOPR ligand binding pocket and the critical amino acid residues required for high affinity NOPR binding. This recent discovery taken together with our results showing the G-protein and arrestin-dependent diversity of NOPR signaling provides a vital roadmap for future studies exploring the potential for functionally selective NOPR ligands as have been recently reported for numerous other GPCR systems including other opioid receptor subtypes (2,40,51).
In conclusion, the findings presented in this work suggest that NOPR receptors are regulated via similar but unique mechanisms as compared with the other opioid receptor subtypes. Insights gained from this study are likely to provide additional understanding of the functional properties of the nociceptin-NOPR system and signaling complex, and may be relevant to other GPCR classes and system.