Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

The endogenous catecholamines norepinephrine (NE) and epinephrine (EPI) exert their effects through a large family of adrenergic receptor (AR) subtypes. The three known -ARs, ₁-, ₂-, and ₃-, activate adenylate cyclase through the stimulatory G protein G_S, thereby increasing cyclic AMP (cAMP) levels in cells. NE and EPI have different relative potencies at the different -AR subtypes, with NE > EPI at ₁-ARs, NE < EPI at ₂-ARs, and NE = EPI at ₃-ARs (Bylund et al., 1994).

Some evidence suggests that -AR subtypes may activate G_S/adenylate cyclase with different efficiencies. In human heart, ₂-ARs couple more efficiently to adenylate cyclase than do ₁-ARs (Waelbroeck et al., 1983; Bristow et al., 1989) and are more efficient in causing contraction of single myocytes (del Monte et al., 1993). Some studies (Green et al., 1992) have also suggested that recombinant human ₂-ARs couple more efficiently to adenylate cyclase than do ₁-ARs in transfected fibroblasts, although others (Suzuki et al., 1992) found no differences with the same cells. In transfected mouse fibroblasts, human ₁-ARs coupled much less efficiently to adenylate cyclase than ₂-ARs, particularly when the subtypes were coexpressed (Levy et al., 1993). Green and Liggett (1994) identified a proline-rich region of the third cytoplasmic loop of the ₂-AR that may regulate coupling efficiency. However, subtype-specific differences in coupling efficiency appear to be dependent on cell phenotype (Rousseau et al., 1996), because there are reports of differences in some cell lines but not others (Zhou et al., 1995).

Previous studies in our laboratory have compared the coupling efficiencies of ₁- and ₂-ARs in rat C₆ glioma cells (Zhong and Minneman, 1993, 1995), where both native and recombinant ₁-ARs couple more efficiently than ₂-ARs to adenylate cyclase activation (Zhong et al., 1996). These results support the existence of functional differences between subtypes, although, in this case, the ₁-ARs are more efficiently coupled than ₂-ARs.

Many of these studies are complicated by the fact that -AR subtypes coexist in cells or tissues, making it difficult to isolate responses to a single subtype. Also, there has been no systematic comparison of the relative coupling efficiencies of ₁- and ₂-ARs in the same cellular background and how such coupling efficiencies might be altered by expression of multiple subtypes. Finally, the important question of which subtypes are involved in responses to specific agonists when closely related subtypes coexist has not been clearly addressed. In this study, we compare the relationship of ₁- and ₂-ARs, expressed alone or together, with agonist-induced cAMP formation in transfected GH₃ pituitary cells. This cell line normally expresses no detectable AR subtypes, allowing us to study responses to these receptors in the same cellular background.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. GH₃ pituitary cells were obtained from the American Type Culture Collection (Rockville, MD). Dulbecco's modified Eagle's medium (DMEM), horse serum, fetal bovine serum, and trypsin-EDTA were obtained from Atlanta Biologicals (Atlanta, GA); geneticin and LipofectAMINE were obtained from Gibco-BRL (Grand Island, NY); hygromycin B was obtained from Boehringer Mannheim (Indianapolis, IN); LacSwitch inducible expression system was obtained from Stratagene (La Jolla, CA); [2,8-³H]adenine (25Ci/mmol) was obtained from DuPont NEN (Boston, MA); ICI 118,551 was obtained from Cambridge Research Biochemicals (Macclesfield, UK); Ecolume was obtained from ICN Biomedicals (Costa Mesa, CA); ()-isoproterenol (ISO), 3-isobutyl-1-methylxanthine, cAMP, and other chemicals were obtained from Sigma Chemical Co. (St. Louis, MO). Dr. Curtis Machida (Oregon Regional Primate Center, Beaverton, OR; Machida et al., 1990) kindly provided the rat ₁-AR cDNA, and the rat ₂-AR genomic sequence was cloned in our laboratory (Zhong et al., 1996). Zinterol was kindly donated by Mead Johnson Pharmaceuticals (Evansville, IN) and CGP 20712A by Ciba Geigy (Summit, NJ).

Cell Culture. GH₃ cells were grown in DMEM (high glucose) plus 10% horse serum, 5% fetal bovine serum, 100 mg/l streptomycin, and 60 mg/l penicillin in a humidified atmosphere containing 7% CO₂ at 37°C. Confluent cells were subcloned at a ratio of 1:3 into Primaria flasks (Falcon). For cAMP assays, 200 µl of cells were subcultured per well into a 96-well plate. For radioligand binding assays, 10 ml of cells were added to 100-mm plates and grown until confluent.

Transfection. GH₃ cells were cotransfected with the lac repressor vector (p3'SS) and the operator vector (pRSVNot) containing either the rat ₁- or ₂-AR coding sequence (Zhong et al., 1996) with LipofectAMINE. Cells were washed twice with DMEM, and 2 µg of DNA and 10 µl of LipofectAMINE were added in 1 ml of medium (DMEM plus 0.5 µM insulin, 100 µg/ml transferrin, 60 µM putrescine, and 30 nM selenium salt). Cells were incubated for 5 to 6 h at 37°C in 7% CO₂, and the medium was aspirated and replaced. Cells were selected with 200 µg/ml hygromycin B and 175 µg/ml geneticin, and resistant cells were screened for inducible receptor expression by radioligand binding. After subcloning by serial dilution, subclones with low basal and high inducible expression were propagated. For coexpression of both subtypes, a stable subclone containing the inducible ₂-AR cDNA (₂#9) was further transfected with the rat ₁-AR cDNA in the constitutively active vector pREP8 (Invitrogen), selected for resistance to histidinol (600 µg/ml), and screened by radioligand binding. Subclones were isolated with inducible expression of the ₂-AR and different levels of constitutive ₁-AR expression.

Radioligand Binding. ¹²⁵I-labeled cyanopindolol (¹²⁵ICYP) binding was performed as described previously (Zhong and Minneman, 1993). After washing with PBS (20 mM NaPO₄, 154 mM NaCl, pH 7.6), cells were resuspended and homogenized in PBS and centrifuged at 30,000g for 10 min, and the pellets were resuspended in PBS. Membranes were incubated with ¹²⁵ICYP (10-300 pM) in 0.25 ml PBS (1 h, 37°C) without or with competing drugs. Reactions were terminated by adding 10 ml Tris-HCl (10 mM, pH 7.4) and filtering through glass fiber filters (Schleicher & Schuell, no. 30). Filters were washed and counted in a gamma counter, and nonspecific binding was defined as binding in the presence of 50 µM ISO. ₁-AR binding was defined as the difference between binding in the absence and presence of 500 nM CGP 20712A, and ₂-AR binding was defined as the difference between binding in the presence of 500 nM CGP 20712A and 50 µM ISO (Dooley et al., 1986).

cAMP Accumulation. cAMP accumulation was measured by [³H]adenine prelabeling (Shimizu et al., 1969). Confluent cells in 96-well plates were prelabeled with [³H]adenine (2 µCi/well) for 2 h, and the medium was removed by submerging the plates three times in 500 ml of Krebs-Ringer-bicarbonate buffer (KRB; 120 mM NaCl, 5.5 mM KCl, 2.5 mM CaCl₂, 1.2 mM NaH₂PO₄, 20 mM NaHCO₃, 11 mM glucose, 0.029 mM CaNa₂EGTA, 37°C). Remaining KRB was removed by aspiration, and drugs were added in 200 µl of KRB containing 2 mM isobutylmethylxanthine. Cells were incubated at 37°C for 10 min, and reactions were terminated by adding 20 µl of 77% trichloroacetic acid and 10 µl of 5 mM cAMP. After sonicating for 10 s, 10-µl aliquots were counted for incorporation of [³H]adenine, and [³H]cAMP was isolated from the remaining supernatant by sequential Dowex and alumina chromatography (Salomon et al., 1974). Results are expressed as percent conversion of incorporated tritium into [³H]cAMP.

Data Analysis. Concentration-response curves were analyzed by nonlinear regression, and data are presented as means ± S.E. K_i values were calculated by the method of Cheng and Prusoff (1973). Student's t test was used to test significance of individual comparisons. Multiple dose-response curves were compared by one-way ANOVA, with a post hoc Neuman-Keul's comparison. P values < .05 were considered significant.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Induction of ₁- or ₂-AR Expression by Isopropyl -D-Thiogalactoside (IPTG). Subclones of GH₃ cells stably transfected with inducible vectors containing ₁- or ₂-AR coding sequences were isolated and screened for low constitutive and high IPTG-inducible receptor expression. The effect of increasing concentrations of IPTG on ₁- and ₂-AR-expressing subclones is shown in Fig. 1. Exposure to 1 mM IPTG for 48 h caused a 7- to 9-fold increase in receptor expression in the ₁#3 and ₂#9 subclones, respectively, with an EC₅₀ for IPTG of about 10 µM. Receptor density could also be titrated by varying the time of exposure to IPTG, with 50% maximal receptor expression occurring after about 10 h (data not shown).

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Dose dependence for IPTG induction of 1-AR (top) or 2-AR (bottom) expression in 1- or 2-transfected GH3 pituitary cell subclones. Cells were exposed to the indicated concentrations of IPTG for 48 h, and total -AR density was determined by saturation analysis of specific 125ICYP binding. Affinity of 125ICYP was similar in all cases. Data represent means ± S.E. of three experiments performed in duplicate.

Pharmacological Characterization of Recombinant Receptors. Inhibition of ¹²⁵ICYP binding by the selective ₁-AR antagonist CGP 20712A, the selective ₂-AR antagonist ICI 118,551, and the selective ₂-AR agonist zinterol in ₁- and ₂-AR-expressing subclones is shown in Fig. 2. CGP 20712A was about 1000-fold more potent than ICI in GH_{3 1}-cells, whereas ICI and zinterol were each about 100-fold more potent than CGP 20712A in GH_{3 2}-cells. K_D values for these drugs at each subtype were similar to those reported for ₁ and ₂-ARs by other investigators (del Monte et al., 1993; Levy et al., 1993).

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Pharmacological characterization of 1- and 2-AR binding sites in membranes from 1#3 (top) or 2#9 (bottom) GH3 pituitary cells. Inhibition of 125ICYP (75 pM) binding by the 1-selective antagonist CGP 20712A (), the 2-selective antagonist ICI 118,551 (), and the 2-selective agonist zinterol () is shown. Each value is the mean ± S.E. of three experiments performed in duplicate.

Comparison of Multiple Subclones. Several stably transfected GH₃ cell subclones expressing each subtype were characterized (Table 1). The four ₁-AR-transfected subclones showed similar basal and IPTG-induced receptor expression, except for GH_{3 1}#19, where induced receptor expression was significantly higher. Basal and IPTG-induced expression were similar in the three GH_{3 2} subclones characterized (Table 1). Receptor expression was consistently higher for ₂- than for ₁-ARs in uninduced cells (₁ = 115 and ₂ = 547 fmol/mg) but was similar after IPTG induction (₁ = 1962 and ₂ = 2128 fmol/mg).

View this table: [in this window] [in a new window]   TABLE 1 Receptor density in 1- or 2-AR-transfected GH3 pituitary cell subclones before and after induction with 1 mM IPTG for 48 h Values are means ± S.E. of data from at least three experiments performed in duplicate.

Time Dependence of ₂-AR Induction and ISO-Stimulated cAMP Accumulation. The effect of increasing time of IPTG exposure on ₂-AR expression and ISO-stimulated cAMP accumulation in the ₂#9 subclone is shown in Fig. 3. Maximal receptor expression was observed after 48 h exposure to 1 mM IPTG. Induction of ₂-AR expression by IPTG exposure for 8, 24, and 48 h increased the potency of ISO in activating cAMP accumulation by 3-fold (p < .05), 7-fold (p < .01), and 11-fold (p < .001), respectively, without significantly changing maximal response (Fig. 3). In addition, small graded increases in basal cAMP levels (in the absence of agonist) were observed after receptor induction (2.00 ± 0.12-fold; n = 7). ₂-AR density (B_max) was determined in parallel plates and compared with the potency of ISO in increasing cAMP accumulation in the same cells or the maximum response to ISO (data not shown). There was a significant linear correlation (r² = 0.70) between log B_max and log EC₅₀ but not between B_max and maximum response (r² = 0.10).

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Time dependence of IPTG induction of 2-AR expression (top) and ISO-stimulated cAMP formation (bottom) in 2#9 GH3 cells. Cells were treated with 1 mM IPTG for the indicated times, and receptor density and ISO-stimulated cAMP formation were measured as described in the text. cAMP data were normalized to the maximal stimulation in cells not treated with IPTG (100%). Each value is the mean ± S.E. of data from three experiments performed in duplicate.

Effects of ₁- or ₂-AR Induction on ISO-Stimulated cAMP Accumulation. Concentration-response curves for ISO-stimulated cAMP accumulation were determined with and without induction of receptor expression by IPTG (1 mM, 48 h) in all ₁- or ₂-AR-expressing subclones, and the composite data are presented in Fig. 4. Increasing expression of either subtype by exposure to IPTG caused an increase in both basal cAMP accumulation and the potency of ISO in stimulating cAMP accumulation, without significantly changing the maximal response to ISO. Induction of either ₂- or ₁-ARs increased the potency of ISO by an average of 2- to 4-fold. Log EC₅₀ values were similar in ₁ (8.1 ± 0.27, control; 8.7 ± 0.24, induced) and ₂ (8.2 ± 0.16, control; 8.6 ± 0.22, induced) subclones. IPTG exposure increased basal cAMP levels in all subclones (data not shown) by an average of 61± 10% in the ₁ subclones and 57 ± 21% in the ₂ subclones. However, IPTG did not significantly increase maximal response in any subclone examined (data not shown).

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Composite data for effect of receptor induction with IPTG on ISO-stimulated cAMP accumulation in 1-transfected (top) or 2-transfected (bottom) GH3 cell subclones. Cells were incubated with (induced) or without (control) 1 mM IPTG for 48 h, and cAMP formation was measured as percent conversion of [3H]ATP to [3H]cAMP. Each value is the mean ± S.E. of data from all experiments with each of the subclones shown in Tables 1 and 2.

Comparison of Receptor Density and Responsiveness. Data on receptor density and ISO-stimulated cAMP accumulation from all ₁- or ₂-AR subclones examined were compiled and compared (Fig. 5). There was a significant relationship (r² = 0.6; p < .02) between receptor density (log B_max) and the potency of ISO in stimulating cAMP accumulation (log EC₅₀) for both subtypes (Fig. 5), with similar slopes of the regression lines (₁, slope = 0.86 ± 0.29; ₂, slope = 1.11 ± 0.33). However, there was no significant correlation between ₁- or ₂-AR density and the maximum response to ISO (data not shown).

View larger version (19K): [in this window] [in a new window]   Fig. 5.   Relationship of 1-AR (circles) or 2-AR (squares) density (log Bmax) to the potency (log EC50) of ISO in stimulating cAMP accumulation in transfected GH3 cell subclones. Receptor densities were increased in 1- or 2-transfected GH3 cells through induction with (closed symbols) or without (open symbols) IPTG. Correlation lines were determined by linear regression, with significant relationships observed for both subtypes (1, r2 = 0.60; 2, r2 = 0.59).

Characterization of GH₃ Cells Cotransfected with Both ₁- and ₂-ARs. Possible interactions between ₁- and ₂-ARs coexpressed in the same cells were studied by transfecting GH_{3 2}#9 cells (with IPTG-inducible ₂-AR expression) with a constitutively active vector coding for ₁-ARs. Antibiotic-resistant subclones were screened for ₁- and ₂-AR expression in the presence and absence of IPTG by radioligand binding. Inhibition of specific ¹²⁵ICYP binding by CGP 20712A and ICI was analyzed by nonlinear regression to determine the proportion of each subtype expressed before and after IPTG induction, as well as their affinities for the antagonists (Fig. 6). Two subclones were isolated with similar basal and inducible ₂-AR density but different densities of ₁-ARs (Table 2). In the absence of IPTG, ₂#9₁#5 cells expressed two to three times as many ₂- as ₁-ARs, whereas ₂#9₁#7 cells expressed approximately similar densities of the two subtypes (Table 2). Exposure to 1 mM IPTG for 48 h caused 4- to 7-fold changes in the ₂/₁ ratio that were caused exclusively by increases in ₂-AR expression. Neither subclone showed a significant change in ₁-AR expression after treatment with IPTG.

View larger version (20K): [in this window] [in a new window]   Fig. 6.   Inhibition of specific 125ICYP binding by the 1-selective antagonist CGP 20712A or the 2-selective antagonist ICI 118,551 in membranes from GH3 subclones 2#91#5 (top) or 2#91#7 (bottom) expressing both 1- and 2-ARs. Experiments were done in cells incubated in the absence (open symbols) or presence (closed symbols) of 1 mM IPTG for 48 h to induce 2-AR expression. Each value is the mean ± S.E. of data from three experiments performed in duplicate. , control + CGP 20712A; , IND (induced) + CGP; , control + ICI; , IND + ICI.

View this table: [in this window] [in a new window]   TABLE 2 Comparison of -AR density and subtype ratio in GH3 pituitary subclones before and after 2-AR induction Values are means ± S.E. of data from 9 to 13 experiments performed in duplicate.

ISO-Stimulated cAMP Accumulation before and after IPTG Exposure in Subclones Expressing Both ₁- and ₂-ARs. Concentration-response curves for ISO-stimulated cAMP accumulation were determined in subclones coexpressing ₁- and ₂-ARs, with and without IPTG exposure (1 mM, 48 h; Fig. 7). Induction of ₂-ARs by IPTG exposure increased basal cAMP accumulation, increased the potency of ISO in activating cAMP accumulation, but had no significant effect on maximal response to ISO. IPTG-induced expression of ₂-ARs caused a doubling of basal cAMP levels in both subclones, although this is not easy to see in ₂#9₁#7 because of the large ISO stimulation (Fig. 7). The potency of ISO was increased about 3-fold in each subclone after IPTG exposure, although this increase was statistically significant only in subclone ₂#9₁#7 (Table 3). The maximal response to ISO was not affected by IPTG induction in either subclone, although normalized maximal response was lower in subclone ₂#9₁#5 than ₂#9₁#7, probably because of higher basal activity (3.2 versus 1.3%). Basal activity showed an unpredictable variation between experiments for both subclones (data not shown).

View larger version (17K): [in this window] [in a new window]   Fig. 7.   ISO-stimulated cAMP accumulation in GH3 subclones coexpressing both 1- and 2-ARs. Cells were pretreated without (open symbols) or with (closed symbols) 1 mM IPTG for 48 h in subclone 5 (top) or subclone 7 (bottom), and concentration-response curves for ISO-stimulated cAMP were determined. Data are expressed as fold-stimulation above basal values in uninduced cells (#5, control basal = 3.2 ± 1.3%; induced basal = 7.2 ± 1.6%; #7, control basal = 1.30 ± 0.25; induced basal = 2.10 ± 0.64%). Each value is the mean ± S.E. of data from four experiments performed in duplicate.

Effects of IPTG-Induced ₂-AR Induction on Responses to Selective Agonists. The effect of IPTG-induced increases in ₂-AR density on cAMP responses to the ₁-selective agonist NE and the ₂-selective agonist zinterol in cell lines expressing both subtypes are shown in Fig. 8, and the data are summarized in Table 3. ₂-AR induction caused a 7- to 11-fold increase in the potency of the ₂-selective agonist zinterol, a smaller increase in the potency of the nonselective agonist ISO and did not significantly alter the potency of the ₁-selective agonist NE. Basal cAMP accumulation increased an average of 79 ± 9% after ₂-AR induction in all experiments, although maximal responses did not change significantly for any of the agonists tested (data not shown).

View larger version (29K): [in this window] [in a new window]   Fig. 8.   NE-stimulated (top) or zinterol-stimulated (bottom) cAMP accumulation in GH3 pituitary subclones treated without (open symbols) or with (closed symbols) 1 mM IPTG for 48 h. Data from subclone 2#91#5 (left) or subclone 2#91#7 (right) are shown. Data are normalized to fold control basal as described in the legend to Fig. 7. Each value is the mean ± S.E. of at least three experiments performed in duplicate.

Isolation of ₁- or ₂-Mediated Responses with Selective Antagonists. Selective antagonists were used to determine the contributions of different subtypes to responses to each agonist. Concentration-response curves for ISO, NE, or zinterol were generated in the presence of CGP 20712A (500 nM) to block ₁-ARs or in the presence of ICI (500 nM) to block ₂-ARs. Figure 9 shows that the presence of CGP 20712A or ICI caused small 2- to 4-fold rightward shifts in concentration-response curves for ISO-stimulated cAMP accumulation in the ₂#9₁#5 cell line without affecting maximal response. Similar patterns were observed in the ₂#9₁#7 cell line (Table 3). After ₂-AR induction by IPTG exposure, the potency of ISO was increased, but both CGP 20712A and ICI caused similar small rightward shifts (Fig. 9).

View larger version (25K): [in this window] [in a new window]   Fig. 9.   Concentration-response curves for ISO-stimulated (top), NE-stimulated (middle), or zinterol-stimulated (bottom) cAMP accumulation in GH3 subclone 2#91#5 in the absence () or presence of 500 nM CGP 20712A () or 500 nM ICI (). Experiments were performed in cells after pretreatment without (control; left) or with (induced; right) 1 mM IPTG for 48 h. Each value is the mean ± S.E. of at least three experiments performed in duplicate.

Effects of ₂-AR Induction on Agonist Potencies. Induction of ₂-ARs by exposure to IPTG significantly increased the potencies of ISO and zinterol but not NE (Table 3). When ₁-ARs were blocked by 500 nM CGP 20712A, ₂-AR induction increased the potencies of all agonists examined (Table 3), suggesting that ₂-ARs participate in responses to all three agonists when ₁-ARs are blocked by CGP 20712A. When ₂-ARs are blocked by 500 nM ICI, ₂-AR induction increased the potency of zinterol by 4- to 7-fold but not ISO or NE (Table 3), suggesting that NE and ISO function primarily through the ₁ subtype in the presence of ICI, whereas the response to zinterol still involves both subtypes.

View this table: [in this window] [in a new window]   TABLE 4 Calculated KD and predicted competitive antagonism for selective antagonists at 1- and 2-ARs in GH3 cells Mean KD values were derived from Figs. 2 and 6. The expected fold shift for each antagonist at each receptor subtype was calculated with the methods of Arunlakshana and Schild (1959).

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

Interactions between receptor subtypes coexisting on the same cells have received only modest attention over the years. Many cells express multiple receptors that can activate converging, diverging, or redundant signals, but little is known about how these signals are integrated and processed. We studied ₁- and ₂-ARs, which both activate G_S/adenylate cyclase and coexist on many cells, to systematically compare the coupling of closely related subtypes and determine how such coupling might be altered by a coexisting subtype.

GH₃ cells are one of the few cell lines that do not endogenously express any detectable ARs but do express G_S and adenylate cyclase (data not shown). Because they are relatively simple to transfect and propagate, and their rat origin is appropriate for studies on rat ARs, we used GH₃ cells for expression of these subtypes. An IPTG-inducible expression system was used to control receptor expression in single subclones, removing potential phenotypic differences that may complicate comparison of different subclones. The densities and ratios of subtypes were determined by radioligand binding, and coupling was determined by parallel measurements of whole-cell cAMP accumulation.

When ₁- or ₂-ARs were expressed alone in GH₃ cells, activation of each receptor by ISO increased cAMP accumulation. Increasing receptor density caused graded increases in agonist potency at both ₁- or ₂-ARs but did not change the maximal response to agonist. This suggests that basal receptor expression was sufficient for a receptor reserve, as has been observed previously in C₆ glioma cells (Zhong and Minneman, 1995). This supports the conclusion that expression of the adenylyl cyclase catalytic subunit or G_S may be limiting in cAMP accumulation (MacEwan et al., 1996).

Increasing the density of either ₁- or ₂-ARs in GH₃ cells also caused small increases in basal cAMP accumulation. Such agonist-independent constitutive activity has sometimes (Lefkowitz et al., 1993; Samama et al., 1993) but not always (Zhong et al., 1996) been observed with various receptors. Empty receptors were traditionally thought to require agonist binding for activation, but many receptors have now been reported to show some constitutive activity. When it occurs, such agonist-independent constitutive activity is often observed only at high receptor densities (Whaley et al., 1993; Gether et al., 1997).

When ₁- and ₂-ARs were each expressed in the absence of the other subtype, the relationship between receptor density and the potency of ISO in activating cAMP accumulation was similar for each subtype. This suggests that ₁- and ₂-ARs show no significant differences in coupling efficiency in GH₃ cells. Note that the data supporting this conclusion in Fig. 5 contain substantial variation. It is not clear how large a difference in coupling efficiency would be needed to become apparent with this approach. However, the large range of receptor densities and agonist potencies compared (100-fold) support the overall conclusion. Other studies (Green et al., 1992; Levy et al., 1993) have suggested that ₂-ARs couple more efficiently than ₁-ARs in some cases, whereas the reverse has also been observed (Zhong and Minneman, 1995; Zhong et al., 1996). This may be because of differences in cellular phenotype (Rousseau et al., 1996), particularly the presence or absence of regulatory molecules controlling signaling, such as RGS proteins (Berman and Gilman, 1998). In any case, our data suggest that differences in coupling efficiency are not intrinsic to ₁- and ₂-AR structure.

Complex responses were observed when ₁- and ₂-ARs were coexpressed in GH₃ cells. Two different subclones with similar basal and inducible ₂-AR density but different constitutive ₁-AR expression were compared. IPTG-induced increases in ₂-AR density caused small (3-fold) increases in the potency of ISO in activating cAMP accumulation without changing the maximal response. Similar to our observations in cells expressing only a single subtype, IPTG also caused small increases in basal cAMP accumulation. The increase in ISO potency observed after increasing ₂-AR expression suggests that the response to ISO can be mediated primarily through the ₂-AR subtype. In support of this, the ₁-selective antagonist CGP 20712A caused only a small (4-fold) decrease in ISO potency, and this effect was not altered after increasing ₂-AR density after IPTG induction. Similarly, the ₂-selective antagonist ICI also caused only small (2-fold) decreases in ISO potency, demonstrating that ₁-ARs can activate this response when ₂-ARs are blocked. The effect of ICI was only slightly larger (3- to 7-fold) after IPTG exposure, consistent with the small increase in ISO potency caused by increasing ₂-AR density. Together, these experiments demonstrate that ₁- and ₂-ARs couple to cAMP accumulation in a completely redundant fashion when coexpressed in GH₃ cells. In fact additivity would not be expected given the large receptor reserve in this system. They also show that ISO activates ₁- and ₂-ARs with a similar potency in these cells, because neither ICI nor CGP 20712A substantially decreased ISO potency, suggesting that similar receptor reserves exist for the two subtypes.

We tried to isolate responses to individual subtypes with the selective agonists NE (₁) and zinterol (₂) alone and in combination with selective antagonists. NE and zinterol appeared to act primarily through their preferred subtype, because the ₁-antagonist CGP 20712A caused a larger decrease in potency for NE than zinterol, whereas the ₂-antagonist ICI showed the opposite profile. In addition, increasing ₂-AR density markedly increased the potency of zinterol but not NE, also suggesting that zinterol acts primarily through ₂-ARs and NE acts predominantly through ₁-ARs. However, NE appeared to show greater selectivity than zinterol, because CGP 20712A caused 50- to 80-fold shifts to the right for NE, whereas ICI caused only 2- to 4-fold shifts to the right for zinterol. Both zinterol and NE could still maximally increase cAMP even when their preferred subtype was competitively blocked, showing that both agonists could activate responses through the subtype for which they had a lower affinity. Comparison of different agonists can be complicated by differences in efficacy, which may result in differences in coupling efficiency (January et al., 1997). In addition, differential desensitization may occur with different agonists. However, the agonists used were chosen because of their similar high efficacy values. Thus, these results support the idea of redundancy between subtypes, in that selective agonists can exert their effects through either subtype, and the primary subtype mediating the response is the one activated at the lowest agonist concentration. This also suggests that subtype density and ratio will affect responses to both nonselective agonists such as ISO and highly subtype-selective agonists.

When ₁-ARs were competitively blocked by 500 nM CGP 20712A, IPTG-induced increases in ₂-AR density increased the potencies of ISO, NE, and zinterol, providing further evidence that ₂-ARs are involved in responses to all three agonists under these conditions. However, the effect of induction was different for each agonist, suggesting that ₂-ARs play different roles in responses to each agonist. As the proportion of ₂-ARs increased, their functional role in agonist-mediated responses also increased, clearly demonstrating that the subtype(s) mediating the observed response depend on the density and ratio of coexisting subtypes.

In conclusion, when ₁- or ₂-ARs are expressed in isolation in GH₃ cells, they show similar efficiencies in activating cAMP accumulation. This suggests that coupling differences between these subtypes are not intrinsic to the receptors themselves but may be due to cell-specific parameters. Pharmacological analysis showed that ₁- and ₂-ARs couple in a completely redundant manner to cAMP formation, with the subtype activated at the lowest concentration of agonist able to activate a maximal response. No additivity was observed between coexisting subtypes, and the coupling efficiency of each subtype was not appreciably altered by coexpression of the other subtype. This suggests that there is no competition or synergism between the two subtypes when they coexist on cells. Finally, subtype density and ratio were found to alter the potencies of both selective and nonselective agonists, which will influence the potencies of drugs and neurotransmitters in causing functional responses. Because of the difficulty in pharmacologically isolating responses to a single subtype within a mixture, defining the role of closely related subtypes in tissues expressing multiple subtypes will require development of new approaches and/or more selective drugs.