Title Page Coupling specificity of NOP opioid receptors to pertussis toxin-sensitive Gα proteins in adult rat stellate ganglion neurons using small interference RNA

The opioid receptor-like 1 (NOP or ORL1) receptor is a G protein-coupled receptor whose endogenous ligand is the heptadecapeptide, nociceptin (Noc). NOP receptors are known to modulate pain processing at spinal, supraspinal and peripheral levels. Previous work has demonstrated that NOP receptors inhibit N-type Ca channel currents in rat sympathetic stellate ganglion (SG) neurons via pertussis toxin (PTX)-sensitive Gαi/o subunits. However, the identification of the specific Gα subunit that mediates the Ca current modulation is unknown. The purpose of the present study was to examine coupling specificity of Nocactivated NOP receptors to N-type Ca channels in SG neurons. Small interference RNA (siRNA) transfection was employed in order to block the expression of PTX-sensitive Gα subunits. RT-PCR results showed that siRNA specifically decreased the expression of the intended Gα subunit. Evaluation of cell surface protein expression and Ca channel modulation were assessed by immunofluorescence staining and electrophysiological recordings, respectively. Furthermore, the presence of mRNA of the intended siRNA target Gα protein was examined by RT-PCR experiments. Fluorescence imaging showed that Gαi1, Gαi3 and Gαo were expressed in SG neurons. The transfection of Gαi1-specific siRNA resulted in a significant decrease in Noc-mediated Ca current inhibition, while silencing of either Gαi3 or Gαo was without effect. Taken together, these results suggest that in SG neurons Gαi1 subunits selectively couple NOP receptors to N-type Ca channels. INTRODUCTION The opioid receptor-like 1 (NOP, or ORL1) receptor belongs to the opioid receptor subfamily of the G protein-coupled receptor (GPCR) superfamily. The heptadecapeptide nociceptin (Noc) is the endogenous NOP receptor ligand that mediates its effects by coupling NOP receptors to effectors via members of the Gαi/o family of heterotrimeric G proteins. The Gαi1-3 and GαoA/B protein subunits are pertussis toxin (PTX)-sensitive; GαoA and GαoB are splice variants while Gαi1-3 are not. Stimulation of NOP receptors by Noc results in inhibition of voltage-gated Ca channels, activation of G protein inwardly rectifying K (GIRK) channels and negative coupling to adenylyl cyclase (for review see Connor and Christie 1999; Mogil and Pasternak 2001; New and Wong 2002). NOP receptors have been shown to regulate pain processing as well as cardiovascular functions (Kapusta 2000; Mogil and Pasternak 2001). The Noc-mediated inhibition of N-type Ca channel currents occurs in a voltage-dependent and membrane-delimited manner (Larsson et al. 2000; Vaughan et al. 2001). More recently, it has been reported that NOP receptors, when expressed at a high density, are capable of forming a complex with N-type Ca channels that results in tonic inhibition of the channels in the absence of agonist (Beedle et al. 2004). Previously, we reported that adult rat sympathetic stellate ganglion (SG) neurons express NOP receptors that modulate N-type Ca channels following exposure to Noc (Ruiz-Velasco et al. 2005). Pretreatment with PTX abolished the nociceptin-mediated Ca current inhibition. Coupling specificity of GPCR and Gα subunits has been studied in a number of expression systems, including primary neurons and established cell lines. For instance, coupling of various GPCR with PTX-sensitive Gαi/o subunits has been examined in rat superior cervical ganglion (SCG) and hippocampal neurons by pretreating the cells with PTX and heterologously expressing mutationally modified PTX-resistant Gα mutants (Chen and Lambert 2000; Jeong and Ikeda 2000; Straiker et al. 2002; Kammermeier et al. 2003; Tian and Kammermeier 2006). The rescue of coupling between GPCR and effector (i.e. ion channels) indicates that the receptor is capable of coupling to the heterologously expressed PTX-resistant Gα subunit. Another technique used to study receptor-G protein coupling involves heterologous expression of PTX-sensitive Gα subunits fused to the C-termini of GPCR (Bertin et al. 1994). Under these conditions, a stoichiometric ratio of 1:1 between GPCR and Gα subunit is achieved (Moon et al. 2001). This approach has been applied successfully in various expression systems (for review see Seifert et al. 1999). Finally, the use of antisense oligonucleotides has also been shown to be an effective tool used to probe G protein coupling specificity between Ca channels with either Gα or Gβγ subunits (Hescheler et al. 1987; Kleuss et al. 1991; Kleuss et al. 1992; Gollasch et al. 1993). The purpose of the present study was to determine the specific PTX-sensitive Gα protein subunits that couple NOP receptors with N-type Ca channels in SG neurons by employing small interference RNA (siRNA). This is a new and powerful technique that provides an efficient means for blocking expression of a specific gene in a variety of cell types and organisms (Milhavet et al. 2003). In this report, we describe the transfection of acutely dissociated rat SG neurons with siRNA targeting each of the natively expressed PTXsensitive Gα subunits. Thereafter, the Noc-mediated modulation of N-type Ca channel currents was examined in freshly replated SG neurons. MATERIALS AND METHODS Stellate ganglion neuron isolation The experiments performed were approved by the Penn State College of Medicine Animal Care and Use Committee (IACUC). Adult rat SG neurons were prepared employing methods previously described (Ruiz-Velasco et al. 2005). Briefly, male Wistar rats (150-225 g) were sacrificed by CO2 anesthesia and decapitated using a laboratory guillotine. The SG was removed and cleared of connective tissue in ice-cold Hanks’ balanced salt solution. Thereafter, the SG neurons were enzymatically dissociated as described (Ruiz-Velasco et al. 2005). The isolated SG neurons were resuspended in Minimal Essential Medium (MEM), supplemented with 10% fetal calf serum, 1% Pen-Strep and 1% glutamine (all from Invitrogen, Carlsbad, CA). The neurons were plated onto 35 mm polystyrene tissue culture plates coated with poly-L-lysine and stored in a humidified incubator (95% air and 5% CO2) at 37C. The media in the cells was not replaced while the neurons were kept in culture. Gα Subunit Immunofluorescence and Cell Imaging For all imaging assays, acutely isolated SG neurons were cultured in 35 mm glass bottom dishes (MatTek Corp., Ashland, MA) coated with poly-L-lysine. SG neurons that were stained for expression of the PTX-sensitive Gα subunits (i.e. Gαi1-3 and Gαo) were kept in culture for 24 h (shown in Fig. 1), while the siRNA-transfected neurons were kept in culture for 110-120 h (shown in Figs. 3, 5 and 7). Prior to imaging, the cells were rinsed five times with 1X phosphate-buffered saline (PBS), then fixed with 2% formaldehyde and 2% sucrose for 20 min. The neurons were then permeabilized in a solution containing 0.05% TWEEN 20 (EMD Biosciences, San Diego, CA) and 5% goat serum (Vector Laboratories, Burlingame, CA) for 10 min at 37C. The cells were preincubated in 5% goat serum for 15 min at room temperature prior to addition of the primary antibody. Primary antibodies were diluted in PBS containing 5% goat serum. The primary antibodies employed were mouse anti-Gαi1 (1:200; Biomol, Plymouth Meeting, PA), mouse anti-Gαi2 (1:200; Biomol; Lab Vision, Fremont, CA; Santa Cruz Biotechnology, Inc., Santa Cruz, CA), rabbit anti Gαi2 (1:200; Abcam Inc., Cambridge, MA), chicken anti-Gαi3 (1:200; Chemicon International, Temecula, CA) and mouse anti-Gαo (1:200; Biomol). The mouse anti-Gαo monoclonal antibody was designed to recognize both Gαo splice variants. Neurons were incubated with the primary antibodies for 60 min at room temperature and placed in a rocker platform. Following the 60 min incubation period, the cells were preincubated in 5% goat serum for 60 min at room temperature prior to addition of the secondary antibodies. The secondary antibodies employed were Alexa Fluor 488 and Alexa Fluor 568 goat anti-mouse or goat anti-chicken (Invitrogen) at a final concentration of 3-5 μg/ml. Secondary antibodies were added to the cells for 45 min and placed on a rocker kept at approximately 4C. Finally, the cells were rinsed three times. In some experiments, micronuclear injection of acutely dissociated SG neurons with Gαi2 (0.2 μg/μl) and pEGFP (0.005 μg/μl) cDNA constructs was performed in order to obtain staining with the rabbit anti-Gαi2 antibody (Abcam). Images were obtained with a Nikon TE2000 microscope using a 20X objective, the XCite 120 (EXFO Life Sciences Group, Ontario, Canada) for illumination and acquired with an Orca-ER 1394 digital CCD camera (Hamamatsu Photonics, Japan) and iVision software (Biovision Tech., Exton, PA). Fluorescence images of Alexa Fluor 488-labeled neurons were obtained with a filter set containing an excitation filter at 480 ± 15 nm, a dichroic beam splitter of 505 nm (LP) and an emission filter at 535 ± 20 nm (B-2E/C, Nikon). The filter set (G-2E/C; Nikon) containing an excitation filter at 540 ± 15 nm, a dichroic beam splitter of 585 nm (long pass, LP) and an emission filter at 620 ± 30 nm was employed for Alexa Fluor 568-labeled neurons. Although the exposure time employed for cells stained with Alexa Fluor 488 and Alexa Fluor 568 was different, it remained constant for each group. The acquired images were pseudocolored with iVision software and were post-processed under the same conditions. Analysis of the red:green intensity ratios were performed with the ImageJ software package (Rasband, W.S., ImageJ, U. S. National Institutes of Health, Bethesda, Maryland, USA, http://rsb.info.nih.gov/ij/, 1997-2007). Optical sections of Silencer® Cy3-labeled-neurons were obtained with a 60X oil objective and the G-2E/C filter set, and collected in 1.0 μm steps covering the z-axis field employing the Pro Scan II (Prior Scientific, Cambridge, UK) motorized stage. The acquired images were processed with the Huygens Essential software package (Scientific Volume Imaging b.v., Hilbersum, The Netherlands) and pseudocolored with iVision software. Small interference RNA (siRNA) sequence design and transfection The siRNA sequences were chosen by employing a macro written by Stephen R. Ikeda, MD, Ph.D. (NIH/NIAAA) on Igor Pro software (Lake Oswego, OR) and based on criteria with a scoring system (10 being best and 1 worse) recently described (Reynolds et al. 2004). The target sequences for rat Gα subunits and scores were as follows: Gαi1, 5’-TAA CGG ACG TCA TCA TAA A-3’ (10) and 5’GAA TAG CAC AAC CAA ATT A-3’ (9) corresponding to nucleotide positions 1016-1034 and 482-500, respectively; Gαi3, 5’-TCA AGG AGC TCT ACT TCA A-3’ (9) corresponding to nucleotide positions 572-590; and GαoA, 5’TCA ACG ACT CTG CCA AAT A-3’ (9) and 5’-CCA TCT GCT TTC CTG AAT A-3’ (9) corresponding to nucleotide positions 446-464 and 854-872, respectively. This last siRNA sequence was also designed to target the GαoB splice variant as well. All siRNA oligonucleotides were purchased from Ambion/Applied Biosystems, Foster City, CA. In another separate set of experiments, labeled control siRNA, Silencer® Cy3-labeled Negative Control #1 (Ambion/Applied Biosystems), was used to assess the uptake efficiency and localization of siRNA oligonucleotides by SG neurons. Five h after plating the acutely dissociated SG neurons, the supplemented MEM media was replaced with premixed Opti-MEM® (Invitrogen), Lipofectamine 2000 (15-20 μl/ml, Invitrogen), 2 mM 2,3-butanedione monoxime and the appropriate amount of siRNA (10300 nM). The total volume of transfection media was 1 ml. After the fourth h of transfection, 1 ml of MEM (supplemented with 2% FBS and 1% glutamine) was added to each dish. Thus, the neurons were transfected with siRNA for a total of five h. Afterwards, the transfection media was removed and the dishes were rinsed three times with MEM (2% FBS, 1% Pen-Strep, 1% glutamine). The neurons were incubated in MEM (2% FBS, 1% Pen-Strep, 1% glutamine) and returned to the incubator. The amount of serum was decreased from 10% to 2% to keep the growth of processes at a minimum. As the transfection protocol was developed, we found that it was necessary to transfect the neurons twice in order to obtain the loss of protein expression as measured with our immunofluorescence staining protocol. This approach has also been reported by others to be necessary for silencing G proteins (i.e. α, β and γ) in cultured cells (Krummins and Gilman 2006) and other signaling proteins in sympathetic neurons (Davidson et al. 2004). Thus, in the present study SG neurons were retransfected 48 h after the initial transfection. However, for the second transfection the incubation period was reduced. That is, the MEM was again replaced with premixed Opti-MEM®, Lipofectamine 2000, 2 mM 2,3-butanedione monoxime and the appropriate amount of siRNA (10-300 nM). Following the second h of transfection, 1 ml of MEM (supplemented with 2% FBS and 1% glutamine) was added to each dish. The total transfection period was 3 h. Thereafter, the cells were rinsed three times as described above and MEM (2% FBS, 1% Pen-Strep, 1% glutamine) was added to each dish and returned to the incubator. SG neurons were kept in culture for a total of 110-120 h at which time they were prepared for imaging (described above) or electrophysiological recording (described below). Electrophysiology and data analysis Whole-cell Ca channel currents of siRNA-transfected neurons kept in culture a total period of 110-120 hr were recorded at room temperature (21-24°C) employing the whole-cell patch-clamp technique. Recording pipettes were pulled from borosilicate glass capillaries (Corning 7052; Garner Glass, Claremont, CA) on a Flaming-Brown (P-97) micropipette puller (Sutter Instrument Co., Novato, CA), coated with Sylgard (Dow Corning, Midland, MI) and fire polished with a microforge. Ca channel currents were acquired with a patch-clamp amplifier (Axopatch 200B, Axon Instruments, Foster City, CA), analog filtered at 2-5 kHz (-3 dB; 4-pole lowpass Bessel filter) and digitized employing custom designed software (S5) on a PowerMacG4 computer (Apple Computer, Cupertino, CA) equipped with an 18-bit analog-to-digital converter board (ITC18, Instrutech Corp., Port Washington, NY). The cells’ series resistance and membrane capacitance were electronically compensated (8085%). Data and statistical analyses were performed with the Igor Pro and Prism 4 (GraphPad Software, Inc., San Diego, CA) software packages, respectively with P < 0.05 considered statistically significant. Graphs and current traces were produced with the Igor Pro, Canvas 8.0 (Deneba Software, Miami, FL) and GraphPad Prism (GraphPad) software packages. The internal pipette solution contained (mM): 120 N-methyl-D-glucamine, 20 tetraethylammonium hydroxide (TEA-OH), 11 EGTA, 10 HEPES, 1 CaCl2, 4 Mg-ATP, 0.3 Na2GTP, 20 mM HCl, and 14 tris creatine phosphate. The pH was adjusted to 7.2 with methanesulfonic acid and the osmolality was 293–302 mOsmol/kg. The external solution consisted of (mM): 145 TEA-OH, 140 methanesulfonic acid, 10 HEPES, 15 glucose, 10 CaCl2, and 0.0003 tetrodotoxin. The pH was adjusted to 7.4 with TEA-OH and the osmolality was 320–325 mOsmol/kg. Stock solutions of nociceptin (Noc; Tocris Cookson, Ellisville, MO), ω-conotoxin GVIA (Bachem, Inc, Torrance, CA) and noradrenaline (NE; Sigma Chem. Co.) were prepared in water and diluted in the external solution to the final concentration employed. SG neuron replating For electrophysiological recording studies, it was necessary to replate the SG neurons to eliminate the processes formed during extended culture times (i.e. 110-120 hr). As a result, space clamp conditions were improved. The neurons were incubated for 5 min in Earle’s Balanced Salt Solution (EBSS) containing 0.2 mg/ml collagenase Type D (Boehringer Mannheim, Indianapolis, IN). During the 5 min exposure to collagenase, the dishes were placed in an incubator shaker set at 38C and rotating at 50 rpm. Thereafter, the dishes were tapped several times to facilitate neuron detachment and 1 ml of MEM (10% FBS, 1% glutamine and 1% Pen-Strep) was added to each dish. The cells were then transferred to sterile 5 ml microtubes and an additional 3 ml of MEM (10% FBS, 1% glutamine and 1% Pen-Strep) were added to each microtube. After a six min centrifugation to pellet the neurons the supernatant was removed leaving approximately 150 μl of media remaining per microtube. The pellet of cells was gently broken up after the addition of 250 μl of MEM (10% FBS, 1% glutamine and 1% Pen-Strep). The cell suspension from each microtube was plated onto 35 mm dishes. Electrophysiological recordings of the neurons were begun at least 5 h after plating to allow for recovery from enzyme digestion. Single-cell reverse transcription-polymerase chain reaction (RT-PCR) analysis RT-PCR experiments were carried out employing the OneStep RT-PCR Kit (Qiagen, Valencia, CA) and the following primer sequences (Integrated DNA Technologies, Coralville, IA), employed previously (Kammermeier et al. 2003), with expected product sizes in parenthesis were used: GAPDH forward: CCA AAA GGG TCA TCA TCT CCG, GAPDH backward: AGA CAA CCT GGT CCT CAG TGT AGC (501 bp). Gαi1 forward: TGA CTA TGA CCT GGT TCT TGC TGA G, Gαi1 backward: ACA CTA CAT TCT CTG TTG CTG GGA G (482 bp). Gαi3 forward: TGA GTA AAG AGC CCA GGA TTG C, Gαi3 backward: CAA AGC AGT TCT GAC CAC CAA CC (453 bp). Gαo forward: CGT GGA GTA TGG TGA CAA GGA GAG, Gαo backward: AAG GTG AAG TGG GTT TCT ACG ATG (300 bp). In this set of experiments, the siRNA-transfected neurons required replating because of their strong adherence to the poly-L-lysine-coated dishes. Thus, the transfected cells were replated (described above) in new poly-L-lysine dishes and allowed to adhere to the bottom no more than 60 min prior to neuron collection. First, the cells were rinsed twice in a solution containing in mM: NaCl 130, KCl 5.4, 10 HEPES, MgCl2 0.8, CaCl2 10, glucose 15 and sucrose 15. Thereafter, neurons were collected into SigmaCote-coated (Sigma) glass capillaries (Corning 7052) with a final volume of 10 to 15 μl. The number of neurons collected per capillary ranged from 10 to 15. The cells were placed in PCR tubes containing reagents supplied by the manufacturer (Qiagen). The tubes were placed in a PCR thermocycler (Eppendorf) and subjected to the following protocol: 94C for 1 min, 58C for 1 min and 74C for 1.5 min, for a total of 44 cycles. The PCR products were analyzed by agarose gel electrophoresis, and visualized by ethidium bromide staining and UV illumination. Gel images were captured using a digital gel imaging system (Eastman Kodak Co., Rochester, NY) and processed with iVision software. RESULTS PTX-sensitive Gα subunits expressed in SG neurons couple N-type Ca channels to NOP receptors In the present study, we examined the coupling specificity between NOP receptors and PTX-sensitive Gαi/o subunits in rat SG neurons. N-type Ca channel currents are inhibited following Noc-mediated NOP receptor activation. Previous reports have shown that this channel subtype is the main carrier of Ca ions in SG neurons (Kukwa et al. 1998; Fuller et al. 2004). We have previously demonstrated that overnight pretreatment of SG neurons with PTX significantly decreases the Noc-mediated Ca current inhibition (Ruiz-Velasco, et al. 2005). The first set of experiments was performed to determine which PTX-sensitive Gα subunit was expressed in SG neurons. Figure 1 shows immunofluorescence images of acutely dissociated SG neurons exposed to Gαi1 (A), Gαi2 (B), Gαi3 (C) and Gαo (D) antibodies. Of the four subunits, Gαi2 immunofluorescence was not detected. Four different antibodies to Gαi2 were tested and no staining was observed with any of the antibodies. The inset in Figure 1B shows a phase image of the cells that did not stain for the Gαi2 antibody. Consequently, SG neurons were transfected with a Gαi2 cDNA construct and served as a positive control. The cells were co-transfected with the GFP cDNA construct in order to identify the cells that had been successfully transfected. Figure 1E shows an immunofluorescence image of an SG neuron stained with the Gαi2 antibody and the image in Figure 1F shows expression of GFP in the same neuron. The inset shown in Figure 1F is a phase image of the corresponding SG neuron expressing both constructs (indicated by arrow) and a separate cell that did not stain with the antibody or express GFP. These results indicate that SG neurons do not express Gαi2 and that NOP receptors employ one or more of the other PTX-sensitive Gα subunits to modulate Ca channel currents. siRNA of PTX-sensitive Gα subunits decreases mRNA expression We first optimized conditions to effectively transfect SG neurons with siRNA. To determine the uptake, cellular distribution and uptake efficiency of siRNA, SG neurons were transfected with a negative control Cy3-labeled siRNA for 5 hr. Figures 2Ai and ii show phase contrast and fluorescence images of SG neurons, respectively, following transfection with the Cy3-labeled oligonucleotide (300 nM). The fluorescence image demonstrates that all SG neurons internalized the labeled siRNA. The inset in Figure 2Aii is a deconvolved fluorescence image of a neuron transfected with Cy3-labeled siRNA showing the oligonucleotide is dispersed in the cytoplasm. With knockdown experiments, it was necessary to transfect the neurons a second time in order to significantly reduce Gα subunit expression, an approach reported to be successful (Krummins and Gilman 2006; Davidson et al. 2004). To test for the siRNA-mediated gene silencing of the Gα subunits, RT-PCR experiments were performed in control (Fig. 2B) and SG neurons following 110-120 hr transfection with 300 nM of either Gαi1 (Fig. 2C) or Gαi3 (Fig. 2D) or Gαo siRNA (Fig. 2E). The cells were replated in 35 mm dishes (as described in Methods section). Soon thereafter, 10 to 15 cells were collected in glass pipettes and RTPCR was performed. The primer pairs employed for each Gα subunit are described in the Methods section and have been successfully employed in other neuron types (Kammermeier et al. 2003; Tian and Kammermeier 2006). Fig. 2B shows a control gel for the three Gα subunits and the housekeeping gene, glyceraldehydes-3-phosphate dehydrogenase (GAPDH). The SG neurons transfected with Gαi1 siRNA showed that mRNA for this subunit was absent, while the message of the other three proteins was present (Fig. 2C). Similarly, Gαi3and Gαo-treated SG neurons showed downregulation of their respective mRNA and not of the other subunits tested (Fig. 2D and 2E). The gel images also show that in some cases the intensity of the Gα subunit mRNA was different (i.e. Gαi1 Panels A and D). Because the number of cells collected in each tube varied from 10 to 15, it is possible that the difference in band intensity is a result of greater mRNA loading in one lane than another. Nevertheless, these semi-quantitative results indicate that siRNA for each Gα subunit did not affect mRNA expression of the non-targeted PTX-sensitive Gα subunits and suggests that targeting was specific at the nucleotide level. These findings were observed in four separate rat preparations. siRNA directed to Gαi1, Gαi3 and Gαo reveals that NOP receptors couple to Gαi1 subunits to Ca channels In this set of experiments, we optimized conditions (i.e. siRNA concentration-response and time-to-silence) to transfect SG neurons with siRNA targeting Gαi1, Gαi3 and Gαo. Immunofluorescence imaging was initially employed to determine whether the targeted protein was in fact knocked down. Figure 3A shows immunofluorescence images of SG neurons stained for Gαi1 and Gαi3 in cells that were transfected with 10, 30, 100 and 300 nM siRNA directed to rat Gαi1. Following a 110-120 hr post-transfection period, Gαi1 expression diminished (as shown by decreasing fluorescence intensity) as siRNA concentration was increased. A final concentration of 300 nM was necessary to observe a significant decrease in Gαi1 expression. On the other hand, Figure 3A shows that in the same cells Gαi3 expression was unaffected by the Gαi1 siRNA treatment at all concentrations employed. The summary plot in Figure 3B shows the mean fluorescence intensity ratio (red:green) of control and Gαi1 siRNA-transfected neurons. The ratio was determined from all acquired images. Next, Ca currents were recorded in siRNA-transfected neurons. Figures 4A and 4B show time courses of Ca currents of a control and a Gαi1 siRNA-treated (300 nM) SG neuron, respectively, obtained before and during application of Noc (3 μM). The currents were evoked every 10 seconds with a depolarizing step pulse to +10 mV for 70 ms from a holding potential of -80 mV (fig. 4B top right). In the control neuron, the current amplitude was blocked by approximately 65% after Noc exposure. The superimposed Ca current traces before (a) and during (b) Noc exposure are shown to the right. Following recovery to the peak Ca current, the cell was exposed to 10 μM NE (current traces c and d). Application of NE resulted in inhibition of Ca currents as well. Figure 4B shows that Noc application to the Gαi1 siRNA-treated neuron did not result in significant coupling between Noc-stimulated NOP receptors and Ca channels. The Ca current amplitude was blocked by approximately 12% following Noc application (see current traces a and b on right). As a positive control, NE was next applied to the same neuron and Ca currents in a similar fashion as control neurons. Figure 4C is a summary graph of the Nocand NE-mediated Ca current inhibition in control and Gαi1 siRNA-transfected SG neurons. During Noc exposure, the inhibition of Ca currents was significantly (P < 0.01) lower than control SG neurons, while the effect of NE was not different in both group of neurons. As noted in the Methods section above, at least two siRNA sequences were employed separately for each Gα subunit. Similar results were obtained for both siRNA sequences and the results shown in Figure 4C were pooled. Specifically, the Noc-mediated Ca current inhibition was 25 ± 3.3, n = 26 and 28.7 ± 5.3%, n = 12 (P = 0.51) for the first and second siRNA sequences, respectively. It should also be noted that in 9 of the 38 siRNA-treated neurons with either nucleotide sequence, we did not observe a decrease in Ca current modulation following Noc application. That is, the Ca current inhibition was greater than 50%. We assume that the siRNA transfection was not as efficacious in silencing Gαi1 in this number of neurons (discussed below). In a separate set of experiments, control and Gαi1 siRNA-transfected neurons were exposed to ω-conotoxin GVIA (10 μM) to determine if N-type Ca channels were the main carriers of Ca ions in SG neurons, as has been previously reported (Kukwa et al. 1998; Fuller et al. 2004). In control neurons, exposure to the N-type Ca channel blocker resulted in Ca current inhibition of 62.4 ± 3.7% (n = 9). The Nocand NEmediated current inhibition measured in the presence of the toxin was significantly (P < 0.01) decreased to 14.8 ± 2.8 (n = 9) and 16.4 ± 4.7% (n = 6), respectively. Application of ω-conotoxin GVIA to transfected cells caused a 59.6 ± 4.5% (n = 11) inhibition of Ca currents. Subsequent exposure of these neurons to Noc and NE also resulted in a significant (P < 0.01) decrease of Ca current inhibition to 22.3 ± 4.8 (n = 11) and 20.9 ± 2.8% (n = 7), respectively. These results suggest that the Ca current inhibition mediated by activated NOP and α2-adrenergic receptors (α2-AR) occurs primarily through N-type Ca channels. This experimental paradigm was employed systematically to determine the coupling specificity of Gαi3 and Gαo and Ca channel modulation. Next, SG neurons were transfected with Gαi3 siRNA to determine whether silencing this subunit would also disrupt coupling of NOP receptors and Ca channels. Figure 5A shows the immunofluorescence images of control and neurons treated with increasing concentration of Gαi3 siRNA. As seen with SG neurons transfected with Gαi1 siRNA, 300 nM was the RNA concentration required to silence Gαi3 expression 110-120 hr post-transfection. The images shown in Figure 5A also demonstrate that Gαo expression was unaffected by siRNA transfection. The summary graph in Figure 5B shows that the fluorescence intensity ratio was significantly (P < 0.01) lower in siRNA-tranfected neurons than control cells. Figure 6A and 6B depict peak Ca currents as a function of time in a control and a Gαi3 siRNA-treated neuron, respectively. The Ca currents were evoked as described above for Figure 4. In the control neuron, application of Noc decreased Ca currents by approximately 64% (see current traces a and b on right). Exposure of the cell to NE following washout also resulted in inhibition of the Ca currents (traces c and d). The time course in Figure 6B shows that both Noc and NE exerted similar effects on Ca currents as the control group. Figure 6C summarizes the Nocand NE-mediated Ca current inhibition in control and siRNA-treated neurons. Unlike the Gαi1 siRNA-transfected neurons, Gαi3 protein knock-down produced no significant alteration on the Noc-mediated Ca current inhibition when compared to control neurons. Also, NE had similar effects in both group of cells. Again, the Ca current inhibition following Noc exposure was similar (P = 0.2) in both siRNA-treated neurons and the results were pooled (58.7 ± 4.5, n = 13 vs. 48.7 ± 5.4%, n = 7). The next set of experiments were designed to examine whether silencing of native Gαo subunits would also disrupt coupling of NOP receptors and N-type Ca channels. The Gαo siRNA concentration-response relationship in Figure 7A shows that 300 nM was also the most efficient concentration required to remove expression of this subunit. On the other hand, the fluorescence images demonstrate that expression of Gαi3 was unaffected. The fluorescence intensity ratio plotted in Figure 7B shows that transfection of SG neurons with Gαo siRNA resulted in decreased levels of Gαo subunits. The time courses of peak Ca currents shown in Figures 8A and 8B correspond to a control and an siRNA-treated SG neuron, respectively, before and after Noc and NE application. Exposure of Noc to the control and siRNA-treated neurons resulted in greater than 60% inhibition of the Ca currents (traces a and b, right side). The superimposed Ca current traces (a-d) for each neuron are shown to the right. When NE was applied externally to either cell, Ca current inhibition was also observed. The mean Nocand NE-mediated Ca current inhibition in control and siRNA-transfected cells is illustrated in Figure 8C. There were no significant differences between both groups of neurons. The NE-mediated Ca current inhibition values obtained in control and in all three groups of siRNA-transfected neurons were similar to those described earlier (Fuller et al. 2004). It should also be noted that there are two known Gαo splice variants (Hepler and Gilman 1992). One of the siRNA sequences employed targeted GαoA while the second sequence was designed to silence both GαoA and GαoB. As with the findings described above, the Noc-mediated Ca current inhibition magnitude was similar for both siRNA sequences employed (65.7 ± 4.8, n = 15, vs. 57.7 ± 2.8%, n = 11, P = 0.2). Finally, both Gαi1 and Gαo subunits were silenced to determine if α2-AR coupling to Ca channels would be altered. The NE-mediated Ca current inhibition was not significantly (P = 0.52) different between control and Gαi1/Gαo siRNA-transfected cells (53.8 ± 10.8, n = 4, vs. 46.0 ± 6.3, n = 7). These results suggest that α2-AR are capable of compensating for the silenced Gα subunits or SG neurons express splice variants that couple to all PTX-sensitive Gα subunits. In this study, coupling specificity of α2-AR was not further examined. Taken together, these findings demonstrate that Noc-activated NOP receptors couple to PTX-sensitive Gαi1 subunits to modulate N-type Ca currents, whereas α2-AR appear to employ more than one Gα subunit. DISCUSSION The goal of the present study was to determine whether signaling specificity occurred between Noc-activated NOP receptors and N-type Ca channels by employing RNA interference. Our efforts were focused at the NOP receptor–Gα subunit junction. The initial immunofluorescence results showed that of the five PTX Gα subunits identified (Gαi1-3 and the splice variants GαoA/B), the Gαi2 subunit was not expressed in the soma of SG neurons. This suggests that Gαi2 does not influence NOP signaling in rat SG neurons. However, the role of Gαi2 in coupling NOP receptors with Ca channels should not be discounted altogether as this G protein subunit is expressed in rat superior cervical ganglion (SCG) neurons (Kammermeier et al. 2003), and NOP receptor activation results in Ca current inhibition in SCG neurons (Larsson et al. 2000). The commercial Gαo antibody employed in these same experiments was generated to recognize the expression of both Gαo splice variants. The unavailability of antibodies that recognize each variant prevented us from identifying the specific Gαo subunit present in SG neurons. Therefore, Gα signaling specificity was examined for Gαi1, Gαi3 and GαoA/B subunits. Immunofluorescence assays demonstrated that 110-120 hr post siRNA transfection were sufficient to observe a significant decrease in expression of the targeted Gα without affecting the other PTX-sensitive Gα subunits. The results showed that siRNA directed toward the PTX-sensitive Gαi1 subunit led to a significant decrease in Ca channel modulation following Noc application when compared to control neurons. The magnitude of Noc-mediated Ca current inhibition was similar to that we previously reported to occur with PTX-treated SG neurons (Ruiz-Velasco et al. 2005). In addition, silencing of Gαi1 subunits did not affect the NE-mediated Ca current inhibition. On the other hand, when either Gαi3 or Gαo was silenced, coupling between NOP receptors and Ca channels was unaffected. Furthermore, the results from RT-PCR experiments showed that siRNA treatment affected expression of the intended siRNA target and not the other PTX-sensitive Gα subunits. Therefore, these findings suggest that coupling of NOP receptors and N-type Ca channels occurs specifically via the PTX-sensitive Gαi1 subunits. Similar observations have been reported to occur with the human δ opioid receptor (Moon et al., 2001). In that study, the opioid receptor cDNA was fused in-frame with PTX-resistant Gαi1 or Gαo subunits and GTPase activity was measured in HEK293 cells. It was found that cells expressing the Gαi1-receptor fusion construct exhibited a three-fold greater efficiency of agonist-stimulated GTPase activity than the Gαo-receptor construct. Conversely, in hippocampal neurons, GABAB, cannabinoid (CB1) and somatostatin receptors – all PTXcoupled GPCR – have been shown to be unable to couple with heterologously expressed PTX-resistant Gαi1 subunits (Straiker et al. 2002). The results presented in this study differ from those previously reported in another sympathetic neuron type by Brown and colleagues (Caulfield et al. 1994; Delmas et al. 1998; Delmas et al. 1999). The studies were performed with rat SCG neurons to examine coupling specificity between N-type Ca channels and α2-AR or M4 muscarinic acetylcholine receptors (mAChR) employing site-directed antibody injection or antisense RNA expression. It was shown that Gαo subunits are primarily responsible for mediating the inhibition of Ca channels. Similar findings also identified Gαo as the subunit coupling Ltype Ca channels and α-AR in rat pituitary GH3 cells (Hescheler et al. 1987; Kleuss et al. 1991). In the present study, silencing Gαo subunits did not appear to affect the coupling of NOP receptors and N-type Ca channels. The disparity between coupling specificity observed between α2-AR and NOP receptors is possibly relevant regarding the critical function α2-AR play in regulating the overall sympathetic tone centrally and neurotransmitter release peripherally. The lack of Gα subunit specificity may be to preserve this crucial function of α2-AR. On the other hand, SG neurons provide the main sympathetic innervation to cardiac muscle. The discriminating coupling of NOP receptors in SG neuron may be associated with controlling blood pressure and/or heart rate (Kapusta, 2000). As previously mentioned, other techniques have been employed to study selective coupling between Gα subunits and GPCR in several cell systems that require heterologous expression of PTX-resistant Gα subunits alone or fused in-frame with the receptor. Thereafter, the rescue of coupling is tested. Coupling specificity has been reported to occur with δ opioid receptors (Moon et al. 2001), metabotropic glutamate receptor 2 (mGluR2; Kammermeier et al. 2003), mGluR6 (Tian and Kammermeier 2006), and GABAB, CB1, somatostatin receptors (Straiker et al. 2002). Although these studies have demonstrated coupling specificity, one drawback is that it can only be known which subunit is capable of coupling and not which subunit actually couples under physiological conditions. In the present study, with RNA interference, the heterologous expression of signaling proteins was circumvented and the role of natively expressed Gα subunits in Ca channel modulation could be determined. Thus, expression levels and nonspecific effects of heterologously expressed proteins were not a concern. For instance, one report has shown that coupling can occur between heterologously expressed Gα14, Gα16 and NOP receptors, yet both natively expressed Gα subunits are unable to do so with NOP receptors (Yung et al. 1999). Therefore, under certain conditions NOP receptors are capable of coupling to other nonPTX-senstive Gα subunits. Whether this occurs under physiological or pathophysiological conditions remains to be determined. Here we have also presented evidence suggesting that α2-AR expressed in SG neurons couple to more than one PTX-sensitive Gα subunit to modulate Ca channels. That is, when any of the three Gα subunits were silenced, the NE-mediated Ca current inhibition was similar to non-transfected neurons. This is not surprising given that two published studies have demonstrated that the NE-mediated Ca current inhibition in rat SCG neurons is mediated via Gαo (Caulfield et al. 1994) or Gαo and Gαi2 (Jeong and Ikeda 2000). Our data showed, however, that silencing Gαi1 alone or Gαi1 and Gαo in SG neurons did not disrupt coupling of α2-AR and Ca channels. Moreover, it has been reported that both Gαo and Gαi contribute to the α2-AR-mediated inhibition of Ca currents with a greater contribution of the former than the latter (Delmas et al. 1999). Interestingly, another report found that α2AR couple to all three Gαi protein subunits in COS-7 cells (Wise et al. 1997). Overall, the results imply that the non-silenced PTX-sensitive Gα subunits (i.e. Gαi3) are capable of compensating for the knocked down Gα protein such that α2-AR signaling remains unaltered in SG neurons. Alternatively, SG neurons express α2-AR splice variants that can couple to all PTX-sensitive Gα subunits. As pointed out above, although silencing of Gαi1 resulted in a significant decrease in Noc-mediated Ca current inhibition (n = 38), coupling (i.e. greater than 60% inhibition) was observed in nine siRNA-treated neurons. There may be several reasons for this. First, it could be that siRNA transfection in these neurons was not as effective as with most cells. Second, it may be that silencing Gαi1 in this group of neurons resulted in compensatory expression of other Gα subunits. In a recent report, silencing of specific Gαβγ subunits was performed in HeLa cells (Krummins and Gilman 2006). It was found that when some of the Gαi1 subunits were silenced, modest increases in expression of Gαi2 and Gαi3 occurred. Thus it may be that an increased expression of either of these subunits or a separate one was sufficient to overcome the loss of Gαi1 and allow coupling of NOP receptors and N-type Ca channels. If this is the case, then it would be expected that this compensatory effect be observed in all neurons transfected with Gαi1 siRNA. An explanation for this may be that although SG neurons are sympathetic, those that innervate sweat glands are cholinergic (Habecker and Landis 1994) while those that innervate the heart and lungs are adrenergic (Pardini et al. 1990; Fuller et al. 2004) and signaling of NOP receptors may differ. In the present study, however, we did not attempt to delineate the origin (i.e. retrograde label) of the SG neurons. Interestingly, the report by Krummins and Gilman (2005) showed that silencing each of the Gαi1-3 subunits alone or in combination did not affect the α2adrenergic-mediated adenylyl cyclase inhibition or forskolin-stimulated cAMP synthesis. On the other hand, knockdown of Gαi1 alone did significantly decrease the prostaglandin E1 (PGE1) receptor-mediated stimulated cAMP synthesis, suggesting that in HeLa cells signaling specificity occurs with PGE1 receptors and not with α-AR. A limitation of our experimental approach is that there is presently no available method to measure protein expression levels in single isolated neurons. In conclusion, this is the first study to demonstrate coupling specificity of NOP receptors with N-type Ca channels in SG neurons by successfully employing siRNA to knock-down specific PTX-sensitive Gαi/o subunits. Selective silencing showed that the Noc-activated NOP receptors couple to Ca channels via Gαi1 subunits and not through Gαi3 or Gαo proteins. Gαi2 protein subunits do not appear to be expressed in the soma of acutely dissociated SG neurons. In addition, silencing of any PTX-sensitive Gα subunit did not result in loss of coupling between α2-AR and Ca channels. Thus, specificity of Gα subunit coupling between NOP receptors and Ca channels neurons seems to be less promiscuous when compared to α2-AR in SG neurons. GRANTSThis study was supported by National Institute of Health Grant HL-074311 to V.R.-V. REFERENCESArias A, Feinn R, Kranzler HR. Association of an Asn40Asp (A118G) polymorphism inthe mu-opioid receptor gene with substance dependence: a meta analysis. Drug AlcoholDepend 83: 262-268, 2006. Beedle AM, McRory JE, Poirot O, Doering CJ, Altier C, Barrere C, Hamid J, NargeotJ, Bourinet E, Zamponi GW. Agonist-independent modulation of N-type calcium channelsby ORL1 receptors. Nat Neurosci 7:118-125, 2004. Bertin B, Freissmuth M, Jockers R, Strosberg AD, Marullo S. Cellular signaling by anagonist-activated receptor/Gs alpha fusion protein. Proc Natl Acad Sci USA 91: 8827-8831,1994. Caulfield MP, Jones S, Vallis Y, Buckley NJ, Kim GD, Milligan G, Brown DA.Muscarinic M-current inhibition via G alpha q/11 and alpha-adrenoceptor inhibition of Cacurrent via G alphao in rat sympathetic neurones. J Physiol 477: 415-422, 1994. Chen H, Lambert NA. Endogenous regulators of G protein signaling proteins regulatepresynaptic inhibition at rat hippocampal synapses. Proc Natl Acad Sci USA 9: 12810-12815, 2000. Connor M, Christie MD. Opioid receptor signalling mechanisms. Clin Exp PharmacolPhysiol 26: 493-499, 1999.Davidson TJ, Harel S, Arboleda VA, Prunnell GF,Shelanski ML, Greene LA, TroyCM. Highly efficient small interfering RNA delivery to primary mammalian neuronsinduces microRNA-like effects before mRNA degradation. J Neurosci 24: 10040-10046,2004. Delmas P, Abogadie FC, Dayrell M, Haley JE, Milligan G, Caulfield MP, Brown DA,Buckley NJ. G-proteins and G-protein subunits mediating cholinergic inhibition of N-typecalcium currents in sympathetic neurons. Eur J Neurosci 10: 1654-1666, 1998. Delmas P, Abogadie FC, Milligan G, Buckley NJ, Brown DA. βγ dimers derived from Goand Gi proteins contribute different components of adrenergic inhibition of Ca channels inrat sympathetic neurons. J Physiol 518: 23-36, 1999. Fuller BC, Sumner AD, Kutzler MA, Ruiz-Velasco V. A novel approach employingultrasound guidance for percutaneous cardiac muscle injection to retrograde label rat stellateganglion neurons. Neurosci Lett 363: 252-256, 2004. Habecker BA, Landis SC. Noradrenergic regulation of cholinergic differentiation. Science264: 1602-1604, 1994. Hepler JR, Gilman AG. G proteins. Trends Biochem Sci 17: 383-387, 1992. Ikeda SR. Double-pulse calcium channel current facilitation in adult rat sympatheticneurones. J Physiol 439: 181-214, 1991. Jeong SW, Ikeda SR. Effect of G protein heterotrimer composition on coupling ofneurotransmitter receptors to N-type Ca(2+) channel modulation in sympathetic neurons.Proc Natl Acad Sci USA 97: 907-912, 2000. Kammermeier PJ, Davis MI, Ikeda SR. Specificity of metabotropic glutamate receptor 2coupling to G proteins. Mol Pharm 63: 183-191, 2003, 2003. Kapusta DR. Neurohumoral effects of orphanin FQ/nociceptin: relevance to cardiovascularand renal function. Peptides 21:1081-1099, 2000. Krummins AM, Gilman AG. Targeted knockdown of G protein subunits selectivelyprevents receptor-mediated modulation of effectors and reveals complex changes in non-targeted signaling proteins. J Biol Chem 281: 10250-10262, 2006. Larsson KP, Olsen UB, Hansen AJ. Nociceptin is a potent inhibitor of N-type Ca(2+)channels in rat sympathetic ganglion neurons. Neurosci Lett 296: 121-124, 2000. Milhavet O, Gary DS, Mattson MP. RNA interference in biology and medicine.Pharmacol Rev 55: 629-648, 2003. Mogil JS, Pasternak GW. The molecular and behavioral pharmacology of the orphaninFQ/nociceptin peptide and receptor family. Pharmacol Rev 53: 381-415, 2001. Moon HE, Bahia DS, Cavalli A, Hoffmann M, Milligan. Control of the efficiency ofagonist-induced information transfer and stability of the ternary complex containing thedelta opioid receptor and the alpha subunit of G(i1) by mutation of a receptor/G proteincontact interface. Neuropharmacol 41: 321-330, 2001. New DC, Wong YH. The ORL1 receptor: molecular pharmacology and signallingmechanisms. Neurosignals 11: 197-212, 2002. Pardini BJ, Lund DD, Schmid PG. Innervation patterns of the middle cervical-stellateganglion complex in the rat. Neurosci Lett 117: 300-306, 1990. Reynolds A, Leake D, Boese Q, Scaringe S, Marshall WS, Khvorova A. Rational siRNAdesign for RNA interference. Nat Biotechnol 22: 326-330, 2004. Ruiz-Velasco V, Ikeda SR. Multiple G-protein betagamma combinations produce voltage-dependent inhibition of N-type calcium channels in rat superior cervical ganglion neurons. JNeurosci 20: 2183-2191, 2000. Ruiz-Velasco V, Puhl HL, Fuller BC, Sumner AD. Modulation of Ca channels byopioid receptor-like 1 receptors natively expressed in rat stellate ganglion neuronsinnervating cardiac muscle. J Pharmacol Exp Ther 314: 987-994, 2005. Seifert R, Wenzel-Seifert K, Kobilka BK. GPCR-Galpha fusion proteins: molecularanalysis of receptor-G-protein coupling.[see comment]. Trends Pharmacol Sci 20: 383-389,1999. Straiker AJ, Borden CR, Sullivan JM. G-protein alpha subunit isoforms coupledifferentially to receptors that mediate presynaptic inhibition at rat hippocampal synapses. JNeurosci 22: 2460-2468, 2002. Tian L, Kammermeier PJ. G protein coupling profile of mGluR6 and expression of Galpha proteins in retinal ON bipolar cells. Vis Neurosci 23: 909-916, 2006. Vaughan CW, Connor M, Jennings EA, Marinelli S, Allen RG, Christie MJ. Actions ofnociceptin/orphanin FQ and other prepronociceptin products on rat rostral ventromedialmedulla neurons in vitro. J Physiol 534: 849-859, 2001. Wise A, Watson-Koken MA, Rees S, Lee M, Milligan G. Interactions of the alpha2A-adrenoceptor with multiple Gi-family G-proteins: studies with pertussis toxin-resistant G-protein mutants. Biochem J 321: 721-728, 1997. Yung LY, Joshi SA, Chan RY, Chan JS, Pei G, Wong YH. GalphaL1 (Galpha14) couplesthe opioid receptor-like1 receptor to stimulation of phospholipase C. J Pharmacol Exp Ther288: 232-238, 1999.