Introduction

Top Abstract Introduction Methods Results Discussion References

Galanin, a 29 amino acid neuropeptide, was first isolated in the pig intestine (Tatemoto et al., 1983) and has since been found to be widely distributed in the central and peripheral nervous systems (Melander et al., 1988) and endocrine system (Dunning et al., 1986). The amino acid sequence of human galanin has 90% homology with other species but shares little homology with other known neuropeptides, suggesting that galanin is member of a unique peptide family (Evans and Shine, 1991). Galanin has been implicated in the regulation of numerous biological functions including inhibition of glucose-induced insulin secretion, regulation of intestinal motility, the stimulation of growth hormone secretion, and the inhibition of leutinizing hormone-releasing hormone secretion (for reviews see Bartfai et al., 1992; Crawley, 1995). Centrally, galanin has been associated with several sensory and behavioral functions such as feeding, nociception, spinal reflexes and memory (for reviews see, Crawley, 1995; Leibowitz, 1995; Kask et al., 1997). Regarding its role in memory, galanin impairs working memory in rodents, and inhibits the release of acetylcholine in the hippocampus, as well as various other neurotransmitters throughout the brain. These influences, along with the observation of a hyperinnervation of galanin fibers in the human basal forebrain of Alzheimer's disease patients suggest a possible role for galanin in the cholinergic dysfunction characteristic of the disease (for review, see Crawley, 1996).

Galanin is thought to mediate these physiological effects via its high affinity interaction with G-protein-coupled receptor(s) and the subsequent activation of several intracellular signaling pathways including the inhibition of adenylyl cyclase (via G_i/o G proteins), the closure of dihydropyridine-sensitive L-type Ca⁺⁺ channels, and the opening of ATP-sensitive K⁺ channels (for reviews, see Bartfai et al., 1992; Kask et al., 1995b, 1997). Structure-activity studies using galanin fragments and chimeric peptides emphasize the importance of the N-terminal region for its binding to its receptor (Amiranoff et al., 1989). Although galanin binding sites have been studied in rat and human brain tissue and transformed cell lines (e.g., Melander et al., 1988; Lorinet et al., 1994; Walli et al., 1994; Heuillet et al., 1994; Deecher et al., 1995), the radioligands available are inadequate in discerning the known multiplicity of galanin receptor subtypes present in mammalian brain. Indeed, to date, there exists at least three high-affinity receptors for galanin. The first galanin subtype, GALR1, was cloned from a human Bowes melanoma and rat brain (Habert-Ortoli et al., 1994; Burgevin et al., 1995), whereas a second subtype, GALR2 was cloned from rat brain (Howard et al., 1997; Wang et al., 1997a; Smith et al., 1997b). As with the GALR1 receptor, the GALR2 receptor binds galanin with high-affinity and can negatively couple to adenylyl cyclase (Wang et al., 1997a); however, unlike the GALR1 receptor it can also activate phospholipase C (Smith et al., 1997b). Recently, a third galanin receptor subtype, GALR3, has been identified that is expressed in the periphery but reportedly not the rat brain (Wang et al., 1997b). In the absence of selective radioligands for galanin receptor subtypes, recombinantly expressed receptors serve as a useful tool to elucidate the pharmacological and biochemical properties of specific galanin receptor subtypes. In this study, we use an EBV oriP-driven episomal expression system to study the pharmacological and functional interactions of galanin, its fragments, and chimeric peptides with the human GALR1 receptor.

	Methods

Top Abstract Introduction Methods Results Discussion References

Materials. ¹²⁵I-human galanin (2200 Ci/mmol) and [³⁵S]GTPS (1000-1500 Ci/mmol) were purchased from New England Nuclear (Boston, MA). All peptides were purchased from Bachem Bioscience (King of Prussia, PA) or Sigma Chemical Co. (St. Louis, MO) including human and porcine galanin, galantide or M15, M35, M40, C7, Gal 1-16 and NPY. All other reagents not otherwise indicated were purchased from Sigma.

Stable expression of GALR1 receptors in HEK293 cells. A full-length human GALR1 cDNA was subcloned into the plasmid pH-a5-2 and transfected into 293EBNA cells (Invitrogen) using lipofectamine (Gibco BRL, Grand Island, NY). This plasmid also contained the CMV immediate early promoter to drive receptor expression and the EBV oriP for its maintenance as an extrachromosomal element, and the hph gene from Escherichia coli to yield hygromycin B resistance. Transfected cells were maintained in DMEM containing 10% fetal bovine serum at 37°C in a humid environment (5% CO₂) for 10 days. The cells were then adapted to spinner culture for bulk processing. On the day of harvest, cells were washed in PBS, counted, and stored at 80°C until used.

Membrane preparation. On the day of assay, pellets of whole cells (containing approximately 1 × 10⁸ cells) expressing the GALR1 receptor were thawed on ice and homogenized in 10 ml of tissue buffer (50 mM HEPES, 10 mM MgCl₂, 2 mM EGTA and 1 µg/ml each of aprotinin, leupeptin, and pepstatin, pH 7.0 at 23°C) using a Brinkman Polytron (PT-10, setting 6 for 10 sec). The homogenate was centrifuged at 48,000 × g for 12 min and the resulting pellet washed by repeating the homogenization and centrifugation steps. The final pellet was resuspended in tissue buffer to a working concentration of 0.1 mg/ml protein as determined by the method of Bradford (1976) using bovine serum albumin as the standard.

¹²⁵I-human galanin binding studies in membranes and whole cells. For the membrane binding assays, all ligands and reagents (e.g., guanine nucleotides) were prepared in assay buffer, which was identical to the tissue buffer except for the inclusion of 0.15 mM bacitracin and 0.1% w/v ovalbumin. Assays were conducted in disposable polypropylene 96-well plates (Costar Corp., Cambridge, MA), and were initiated by the addition of 100 µl membrane homogenate (containing 5-10 µg membrane) to 200 µl of assay buffer containing ¹²⁵I-human galanin and competing peptides. Specific binding was determined in the presence of excess porcine galanin (1 µM). The assay was incubated to equilibrium for 2 hr at 23°C as predetermined by kinetic analysis using filtration or the SPA. Reactions were terminated by rapid filtration using a cell harvester (Inotech Biosystems Inc., Lansing, MI) over GFF glass-fiber filters that had been presoaked in PEI (0.3% v/v). Filters were washed three times with 0.3 ml cold wash buffer (PBS, pH 7.0, containing 0.01% Triton X-100). Filters were dried at 23°C and then counted in a gamma counter at 80% efficiency. For competition experiments, wells received 100 µl assay buffer containing competing peptides (twelve concentrations in duplicate, 0.1-1 pM to 0.1-1 µM), 50 µl of 600 pM ¹²⁵I-human galanin and 150 µl membrane homogenate. For saturation studies, a mixed "hot/cold" human galanin saturation isotherm was employed to identify both high and low affinity sites of the GALR1 receptor. Twenty-four concentrations of human galanin ¹²⁵I-human galanin alone or with cold human galanin (noniodinated, up to 60 nM final) were tested in duplicate in a 300 µl volume. The "cold Scatchard" portion of the experiments was conducted as described by Bennett (1978) and included concentrations of ligand that overlapped with the "hot only" portion of the saturation isotherm.

Kinetic studies: scintillation proximity and filtration assays. A SPA was developed to assess the real-time association/dissociation kinetic profile of ¹²⁵I-human galanin at the human GALR1 receptor. First, however, specific ¹²⁵I-human galanin binding to wheat-germ agglutinin-polyvinyltoluene beads (Amersham Corp., Arlington Heights, IL) was optimized with respect to tissue concentration and amount of SPA bead. Second, a comparative analysis of SPA or traditional filtration methodology was undertaken to assess whether the different approaches yielded similar binding parameters in competition experiments.

Whole cell cAMP accumulation studies. cAMP accumulation studies were conducted in whole cells by radioimmunoassay. Before the peptide experiments were conducted, the assay was optimized with respect to temperature and cell density to yield a linear accumulation of forskolin-stimulated cAMP accumulation within a 20-min period. Cells were taken from spinner cultures and adapted to an adherent system by plating 40,000 cells/well onto 24-well culture plates using DMEM containing high glucose, 2 mM glutamine, 10% fetal calf serum and 250 µg/ml hygromycin B. After a 48-hr period, the high serum medium was replaced with DMEM containing 0.1% fetal calf serum and the cells were allowed to acclimate for 2 hr. Cells were preincubated with low serum DMEM containing 1 mM 3-isobutyl-1-methylxanthine for 30 min at 37°C, then exposed to various concentrations of galanin-related peptides (2 pM-100 nM) and 10 µM forskolin for an additional 20 min at room temperature. Reactions were terminated by aspirating the medium, adding cold radioimmunoassay buffer containing 1 mM 3-isobutyl-1-methylxanthine, and freezing the samples on dry ice. Intracellular cAMP levels were released by freezing and thawing and then measured using a single antibody radioimmunoassay kit (New England Nuclear; Boston, MA).

[³⁵S]GTPS binding assays. Agonist-stimulated [³⁵S]GTPS binding was examined using a modification of previously published methods (e.g., Williams et al., 1997). HEK293E membranes containing the human GALR1 receptor (15 µg protein/well) were preincubated for 1.5 hr (23°C) with galanin-related peptides in binding assay buffer containing 100 mM NaCl and 1 µM GDP. The incubates were then exposed to 0.5 nM [³⁵S]GTPS for 30 min. Optimal incubation periods and GDP concentrations were selected on the basis of preliminary experiments. Basal binding was defined in the presence of GDP and the absence of peptide, whereas nonspecific binding was determined in the presence of 10 µM GTPS. Reactions were terminated by rapid filtration as in the radioligand binding assays (minus the PEI presoaking of filters). Radioactivity bound to filter was determined by liquid scintillation spectroscopy.

Data analyses. The apparent equilibrium constants (K_D) and the maximal number of binding sites (B_max) from the saturation isotherm binding experiments and K_is from the competition experiments were calculated using the iterative nonlinear curve-fitting program (LIGAND) of Munson and Rodbard (Munson and Rodbard, 1980) or GraphPad Prism (San Diego, CA). IC₅₀ values were generated using the program Deltagraph (Monterey, CA) with the one-site "pseudo" Hill model (Graesar and Neubig, 1992): B = min + [(max  min)*IC₅₀^nH]/[I^nH + IC₅₀^nH]. For the membrane-based kinetic studies, the monophasic association phase was plotted as the rate of disappearance of reactant [free receptor; Ln(B_eq  B)] to determine k_obs from the line slope; where, B equals bound ligand at time t, B_eq is the bound ligand at equilibrium. For the whole cell assays, multiphasic kinetic profiles were analyzed by nonlinear regression using GraphPad Prism.

	Results

Top Abstract Introduction Methods Results Discussion References

Characterization of equilibrium binding of ¹²⁵I-human galanin to the human GALR1 receptor. Human HEK293E cells that stably express the EBNA1 support the episomal replication of plasmids containing the Epstein Barr virus origin of replication (EBV oriP). A 293 EBNA cell line expressing the human GALR1 receptor from an episomal plasmid was generated and scaled-up in spinner culture for detailed biochemical analysis in several weeks. This expression vector was stably maintained for at least 3 mo in culture.

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Influence of divalent and monovalent cations on specific 125I-human galanin binding to human GALR1 receptors in HEK293E membranes. Specific binding of 100 pM 125I-human galanin was measured in the presence and absence of increasing Na+, K+, Mg++ and Ca++ concentrations. Data are expressed as a percent of control specific binding (assay buffer without ions) for each ion. The data represent triplicate determinations per point from a single experiment that is representative of two independent experiments.

Saturation studies of ¹²⁵I-human galanin binding to the GALR1 receptors in membranes and whole cells. The saturable nature of specific binding of ¹²⁵I-human galanin to the GALR1 receptor in membranes is shown in figure 2 and described in table 1. Saturation isotherms were performed with the addition of ¹²⁵I-human galanin up to 5 nM, whereas from 5.0 to 60 nM, unlabeled cold human galanin was added along with a fixed concentration of 1.5 to 3.0 nM ¹²⁵I-human galanin. This strategy facilitated the determination of both high- and low-affinity binding states of the GALR1 receptor. ¹²⁵I-human galanin binding was highly specific and saturable, and revealed a heterogeneous population of binding sites (significant 2-site/state model, P < .05). Mean (±S.E.M.) equilibrium dissociation rate constants (K_D) equaled 48.9 ± 7.3 pM and 14.7 ± 4.3 nM for the high and low-affinity states of the receptor, respectively. The mean B_max values for the high- and low-affinity states of the receptor were 554.3 ± 24.9 fmol/mg protein and 2.07 ± 0.13 pmol/mg protein, respectively. Thus, the number of high-affinity sites represented approximately 21% of the total receptor pool (R_t). To provide additional binding evidence for coupling to a G-protein, a saturation experiment was conducted in the presence of a nonhydrolyzable GTP analogue, 200 µM Gpp(NH)p. Addition of Gpp(NH)p led to a reduction in specific counts and a shift to low affinity state that was best fit by a linear model (one-site/state) with an apparent K_D of 1.8 nM (data not shown).

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Saturation isotherms of 125I-human galanin to human GALR1 receptors in HEK293E membranes and whole cells. Saturation experiments were performed with the addition of 125I-human galanin up to 5 nM, whereas from 5.0 to 60 nM, unlabeled cold human galanin was added along with a fixed concentration of 1.5 to 3.0 nM 125I-human galanin as indicated in "Methods." A, Depicted is a saturation isotherm (and Scatchard transformation inset) using GALR1 membranes that is representative of four experiments. Nonlinear regression analysis (LIGAND) significantly fit (P < .05) a two site/state model better than a one site/state model. B, A representative saturation isotherm (and Scatchard transformation inset) using GALR1 whole cells. Depicted are data that are representative of three independent experiments (KDH = 419 pM, KDL = 5.8 nM; P = .07, LIGAND). As reported in "Results" and shown in table 1, a composite analysis (LIGAND) using data from three experiments (64 individual data points) was performed to adequately defined the low-affinity site. This analysis gave sufficient statistical power to yield a significant two-site/two state model (P < .001) with binding parameters (see table 1) that closely matched this individual experiment.

Kinetic analysis of ¹²⁵I-human galanin binding to the GALR1 receptor in membranes and whole cells: scintillation proximity and filtration assays. Scintillation proximity technology was used to assess the real-time association/dissociation kinetics of human galanin at the GALR1 receptor in membranes. Preliminary experiments revealed that a more than 85% specific signal was achievable at optimized tissue and SPA bead concentrations of 10 µg/well and 0.5 mg/well, respectively. Further, competition studies with porcine galanin and M40 yielded very similar IC₅₀ values when compared to values obtained using filtration methodology (see fig. 3).

View larger version (27K): [in this window] [in a new window]   Fig. 3.   Comparison of 125I-human galanin competition experiments in membranes using filtration vs. the scintillation proximity assay. Competition experiments were performed using various concentrations of porcine galanin and M40 as described in "Methods." Data are the mean values of duplicate determinations from one of two experiments that yielded similar results.

1

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Kinetics of 125I-human galanin binding to the human GALR1 receptor using the scintillation proximity assay. Membranes pre-coupled to WGA-PVT SPA beads were incubated with 125I-human galanin (80-100 pM, at 23°C) and were counted using a Packard TopCount at increasing time periods. Dissociation was measured after the addition of 1 µM human galanin () or buffer alone (). Inset, Linear transformation of the association data depicted as the loss of reactant (free receptor) with time, where Beq equals binding at equilibrium (reversible + irreversible components) and Bt equals binding at time t yielded a mean (±S.E.M.) kobs value of 1.65 ± 0.073 × 108 min1 (t1/2 = 42 min). Data shown are mean triplicate determinations from one of three experiments having similar results.

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Kinetics of 125I-human galanin binding to human GALR1 receptors in whole cells. Shown are the () association phase and the () dissociation phase. Cells were incubated with 80 to 90 pM 125I-human galanin for 90 min (denoted graphically as time zero), and dissociation was observed after the addition of 1 µM human galanin. Data depicted are from one of four experiments performed in the presence of 100 mM NaCl. Substitution of isotonic NaCl with sucrose yielded qualitatively similar results.

Pharmacological characterization of ¹²⁵I-human galanin binding to the GALR1 receptor in membranes and whole cells. The pharmacology of ¹²⁵I-human galanin binding to the human GALR1 receptor in HEK293E membranes was examined by competition experiments using galanin and galanin-related peptides (see table 2). Consistent with the estimate that at 100 pM ¹²⁵I-human galanin, >96% of the sites labeled represent those of the high affinity state, the agonist competition curves were best fit by a one-site model with Hill coefficients near or at unity. Among the various peptides examined, human and porcine galanin, M35, and galantide (M15) had similar affinities for the GALR1 receptor whereas C7, M40 and the galanin NH2-fragment, Gal 1-16, were comparably an order of magnitude less potent in this assay. Human NPY was inactive at the GALR1 receptor at concentrations up to 10 µM. Similarly rank-ordered affinities were observed by whole cell competition experiments, except that the affinities for all of the peptides were noticeably weaker (6- to 14-fold) than those determined using GALR1 membranes. Indeed, the affinity separation between membrane and whole cells, respectively, for human galanin as determine by competition experiments (31 vs. 232 pM) was very similar to the binding affinities defined by saturation experiments (49 vs. 288 pM).

View this table: [in this window] [in a new window]   TABLE 2 Rank-order of potencies of galanin-related peptides for the inhibition of forskolin-stimulated cAMP production, stimulation of GTPS binding and inhibition 125I-human galanin binding in cells expressing human GALR1 receptors

Potency and efficacy of galanin-related peptides in augmenting cAMP production and GTPS binding. The ability of galanin-related peptides to inhibit forskolin-stimulated cAMP accumulation was examined by RIA in whole HEK293E cells stably expressing the human GALR1 receptor. Forskolin produced a maximal cAMP stimulation of 50 to 100 pmol/ml whereas basal values consistently averaged less than 10 pmol/ml. All peptides potently inhibited forskolin-stimulated cAMP levels between 70 to 80% with similar efficacy (see fig. 6). Intrinsic activities (IA) for the galanin fragment (Gal 1-16), chimeric peptides (M15, M35, M40, C7), and porcine galanin ranged from a mean value 0.85 to 0.95 when referenced against human galanin curve maxima (IA = 1.0). Therefore, all of the peptides behaved as high efficacy partial or full agonists in our system. The rank-ordered potencies of the peptides compare favorably well with the relative binding affinities (see table 2). Galanin, M35 and M15 were the most potent, whereas M40, C7 and galanin 1-16 were less potent in inhibiting forskolin-stimulated cAMP production. The IC₅₀s for the inhibition of cAMP production for each of the peptides, however, were all more potent than their respective binding affinities in whole cells. This apparent discrepancy between fractional receptor occupancy and fractional response raised the possibility of receptor reserve or "spare" receptors.

View larger version (31K): [in this window] [in a new window]   Fig. 6.   Inhibition of forskolin-stimulated cAMP production by galanin-related peptides. HEK293E cells expressing human GALR1 receptors were exposed to forskolin and galanin-related peptides as described in "Methods." cAMP levels in the incubate were determined by radioimmunoassay. Data from multiple experiments were combined and composite curve-fits were generated for each peptide for graphic representation. However, potency values (IC50s) were generated for each peptide for each experiment, and the mean values of these multiple determinations are presented in table 2.

View larger version (32K): [in this window] [in a new window]   Fig. 7.   Potencies of galanin-related peptides in stimulating [35S]GTPS binding to cell membranes containing the human GALR1 receptor. Membranes were preincubated with galanin-related peptides and then incubated with [35S]GTPS. Because the levels of absolute [35S]GTPS stimulation showed some interassay variability, and because all of the peptides were full agonists, curves from multiple experiments were normalized to percent stimulation over basal levels and composite curve-fits were generated for each peptide for graphic clarity. However, EC50 values of the peptides for individual experiments were determined and mean values are depicted in table 2.

Agonist profile of galanin and M40 in the "zero" receptor reserve state: partial alkylation studies with phenoxybenzamine. Because intrinsic activity of an agonist ligand can be influenced by the level of receptor reserve ("spare" receptors) in a recombinant system, partial alkylation experiments were conducted to determine the intrinsic efficacy of M40 (a representative chimeric peptide) in a "zero" receptor reserve state. A preliminary binding experiment showed that 1 hr exposure to 10 to 100 µM PBZ reduced specific ¹²⁵I-human galanin binding in whole cells by 40-86% (see fig. 8). In the cAMP accumulation experiments, it was observed that 100 and 250 µM, but not 30 µM PBZ, sufficiently reduced GALR1 receptor density to the "zero" receptor reserve state since the potency of human galanin was fully shifted to the right and its maximal response (i.e., inhibition of forskolin-stimulated cAMP production) was attenuated (Kenakin, 1993). Under these conditions, M40 displayed an intrinsic efficacy equal to the endogenous full agonist galanin (see fig. 8).

View larger version (16K): [in this window] [in a new window]   Fig. 8.   Radioligand binding and forskolin-stimulated cAMP production by human galanin or M40 in GALR1 whole cells subject to irreversible receptor alkylation by phenoxybenzamine (PBZ). A, A dose-related reduction in whole-cell specific galanin binding was observed following pre-incubation (1 hr, 37°C) with PBZ. Data shown are triplicate mean specific counts from a single experiment. B and C, Adherent cells pretreated with 0, 30, 100 and 250 µM PBZ were stimulated with human galanin or M40, and cAMP levels were determined by RIA as described in "Methods." "Zero" reserve state is defined and denoted as the concentrations of PBZ (100 and 250 µM) at which the galanin response curve is shifted to the right and its maximal response (e.g., % control forskolin cAMP stimulation) is reduced. Data shown are mean triplicate determinations from one of three experiments having similar results.

	Discussion

Top Abstract Introduction Methods Results Discussion References

Our goal was to provide a comprehensive pharmacological and biochemical characterization of a recombinant human GALR1 receptor in HEK293E cells using an episomal expression system. Although the pharmacological, kinetic and functional properties of galanin receptor(s) have been evaluated to some extent in rat and human tissues and transformed cell lines (Heuillet et al., 1994; Lorinet et al., 1994; Walli et al., 1994; Deecher et al., 1995; Kask et al., 1995b), relatively little work has been performed specifically on the recombinant human GALR1 receptor (e.g., Sullivan et al., 1997) because its initial description (Habert-Ortoli et al., 1994). Moreover, the use of galanin as the radioligand, and the recent discovery of additional galanin receptor subtypes (rat GALR2 and GALR3; Howard et al., 1997; Wang et al., 1997a; Wang et al., 1997b) confounds some of the earlier pharmacology performed on postmortem brain tissue.

We chose the EBV oriP-driven episomal system since it can generate stable lines that express significant amounts of high-affinity receptors in a short time period (Horlick et al., 1997). Initial saturation experiments with ¹²⁵I-human galanin and GALR1 membranes revealed the presence of two distinct populations of the receptor with mean K_Ds equaling 48.9 pM and 14.7 nM for the high and low-affinity states of the receptor, respectively. Although the K_D of the high affinity state of the GALR1 receptor compares well with literature values, the use of mixed hot/cold saturation methodology facilitated the determination of a predicted, but previously unidentified, low-affinity binding state of the GALR1 receptor. Parallel experiments in whole cells also yielded two-site saturation isotherms, however, the affinity for the high-affinity state (288 pM) was 6-fold lower than that observed in membranes. Comparable affinity differentials (6- to 14-fold) between whole cell and membrane binding were observed for all of the galanin-like peptides in the competition experiments. The reasons for these differences are unknown, although intracellular GTP in whole cells may confer a significantly lower affinity for galanin-like peptides by discharging the high affinity states and increasing the preponderance of low affinity states. It is not likely caused by the inclusion of Na⁺ in the whole cell binding assay because unlike the known disruptive effects of Na⁺ on agonist binding in some 7TM receptors, galanin binding was largely refractory to changes in Na⁺ concentration.

We developed a SPA to evaluate the kinetic profile of human galanin at the GALR1 receptor. SPA is ideally suited for kinetic studies since receptor-ligand interactions can be monitored in real-time. Furthermore, since SPA obviates the need for separating bound from free ligand, it eliminates the potential of radioligand dissociation from low affinity sites that can occur with filtration methods. The kinetic profile of galanin for its receptor(s) has been a topic of discussion (e.g., Deecher et al., 1995) because a significant portion of galanin binding to membranes from brain (e.g., Lorinet et al., 1994; Walli et al., 1994; Deecher et al., 1995) and cell lines (e.g., Lagny-Pourmir et al., 1989; Sharp et al., 1989) shows slowly reversible or quasiirreversible kinetics. This quasiirreversible portion of binding has been hypothesized as being a "tight" agonist, high affinity state of the receptor whereas the reversible portion of galanin binding conceivably represents that of the low-affinity state (Deecher et al., 1995), although the GTP-insensitivity of the binding is inconsistent with a "tight" agonist state (Severne et al., 1987). We confirm with the recombinant receptor that approximately 50% of the binding show quasi-irreversible kinetics since this level of binding remains undiminished for up to 20 hr irrespective of whether dissociation was initiated with excess galanin or GTPS. The reversible and GTPS-sensitive portion of galanin binding to the GALR1 receptor likely represents binding to the high-affinity rather than low-affinity state, since having determined the K_DL and B_maxL for the GALR1 receptor in our study, only 3 to 4% of the labeled sites represent the low-affinity state under our conditions. The quasiirreversible site may simply represent a binding state that is an artifact of the membrane-based system because >90% of ¹²⁵I-human galanin binding was dissociable in the whole cell binding assay within 1 hr. The presence of GDP in live cells may afford normal receptor-G protein cycling not seen in membranes. Irrespective of the exact nature of the binding in membranes, the calculated rate constants from some equilibrium binding studies should be viewed as estimates since they were derived using traditional methods that assume reversible mass action.

We next examined the pharmacological profiles of galanin-like peptides in the membrane and whole cell binding assays and the cAMP and [³⁵S]GTPS functional assays. Competition binding experiments produced the following rank-ordered affinities: M35 = porcine galanin = human galanin = galantide (M15) > C7 = galanin 1-16 = M40  human NPY. This pharmacology compares very well with the profile reported for the recombinant human and the Bowes melanoma GALR1 receptor (Heuillet et al., 1994; Sullivan et al., 1997). Moreover, the rank-order of the binding affinities generally predict the relative potencies of the peptides in the functional assays. All of the peptides behaved as full agonists in the cAMP assays. These observations were largely supported by the GTPS binding assay; however, the potencies were several-fold weaker than those determined in the cAMP studies. This observation may, in part, be attributable to signal amplification of the more downstream second messenger in the case of the cAMP assays (e.g., Selley et al., 1997; Williams et al., 1997).

An apparent mismatch between fractional binding occupancy (K_i) of the peptides in the whole cell binding assays and their fractional response (IC₅₀) in the cAMP assay under similar buffer conditions suggested the presence of receptor reserve. Receptor reserve can increase the relative intrinsic activity and potency of partial agonists (e.g., Kenakin, 1993, Adham et al., 1993). This issue potentially complicates the obvious discordance between the full agonist character of the chimeric peptides observed in our study and previous in vitro work (Bartfai et al., 1993; Gu et al., 1993; Heuillet et al., 1994; Kask et al., 1995a) and studies that demonstrate these peptides antagonize a variety of the known physiological effects of galanin (Crawley, 1995; Kask et al., 1995b, 1997). It is conceivable that the potentially higher receptor reserve of in vitro systems may obscure the true partial agonist character of some of these chimeric peptides since low-efficacy (partial) agonists would predictably antagonize the actions of a full agonist in vivo where a lower receptor reserve may be present. To assess this possibility, partial alkylation studies with phenoxybenzamine were conducted to assess the effects of M40, a representative chimeric peptide with an antagonist profile in vivo, in the cAMP assay under conditions of "zero" receptor reserve. M40 was as efficacious as human galanin when 86% or more of the receptors were "removed" and reserve was eliminated. These observations suggest that the antagonistic effects of M40 (and perhaps other chimeric peptides) in vivo are not mediated via direct interactions with the GALR1 receptor, assuming of course that the rat GALR1 receptor behaves like its human homologue.

A second explanation for this inconsistency may be that the antagonistic effects of the chimeric peptides in vivo are via blockade of a galanin receptor subtype(s) other than the GALR1 receptor. Two additional receptor subtypes have been cloned to-date, although data published thus far do not resolve this issue. A novel rat GALR2 receptor has been cloned (Howard et al., 1997; Wang et al., 1997a) and although its pharmacological and expression pattern is somewhat distinguishable from that of the GALR1 receptor (i.e., Gal (2-29) and [D-Trp2]galanin-(1-29) preferentially bind GalR2), the chimeric peptides were reported to be full agonists in activating phospholipase C (Smith et al., 1997b). The newest member of the galanin receptor family, the rat GALR3 receptor, was cloned and shown to have a more distinct expression pattern and pharmacology (Wang et al., 1997). Unlike the GALR1 and GALR2 receptors, the GALR3 receptor is reported to be exclusively expressed in the periphery in rat and is therefore an unlikely mediator of the functional effects of galanin and chimeric peptides. In terms of pharmacology, the chimeric peptides (M15, M35 and M40) and the NH2-terminal peptide Gal 1-16 have significantly weaker binding affinities at this receptor compared to their affinities at the GALR1 and GALR2 receptor, although the functional properties of these peptides have yet to be determined. Finally, a galanin receptor found in gastrointestinal smooth muscle cells (jejunum) may represent a different subtype because galanin and NH2-terminal galanin fragments, although displaying full agonism, are 1 to 2 orders of magnitude less potent at this receptor than human GALR1 receptor and activate (not inhibit) adenylyl cyclase (Gu et al., 1995; Juréus and Bartfai, 1997). Whether this receptor is the GALR2 or GALR3 receptor or indeed a novel subtype remains to be determined.

In summary, we have characterized the pharmacological and biochemical properties of the human recombinant GALR1 receptor. We have also highlighted the discrepancy between the agonist properties of chimeric peptides in vitro and their purported antagonist effects in vivo, and provide evidence that it is not explained by receptor reserve at the human GALR1 receptor. Continuing efforts at characterizing GALR1 and the newly discovered receptors may resolve this discrepancy as well as delineate the multiple physiological roles of galanin in the brain and periphery.