J. Anim. Sci. 2005. 83:2339-2348
© 2005 American Society of Animal Science
ANIMAL GROWTH, PHYSIOLOGY, AND REPRODUCTION |
Characterization of porcine ß1- and ß2-adrenergic receptors in heart, skeletal muscle, and adipose tissue, and the identification of an atypical ß-adrenergic binding site1
M. N. Sillence*,2,
J. Hooper*,
G. H. Zhou
,
Q. Liu and
K. J. Munn*
* School of Agricultural and Veterinary Sciences, Charles Sturt University, Wagga Wagga, New South Wales 2678, Australia and
and
College of Animal Science, Nanjing Agricultural University, Nanjing 210095, P. R. China
 |
Abstract
|
|---|
The objective of this study was to characterize porcine ß1- and ß2-adrenergic receptors (ß1-AR and ß2-AR) in heart, skeletal muscle, and adipose tissue by measuring the binding of a radioligand to cell membrane fragments. In skeletal muscle (LM), [3H]CGP12177 labeled a homogeneous population of ß2-AR as evidenced by the rank order of affinity of catecholamines [(–)isoproterenol > (–)epinephrine > (–)norepinephrine], a high affinity of the binding site for the ß2-AR-agonist clenbuterol (equilibrium dissociation constant, Kd = 16 nM), and a low affinity of the binding site for the ß1-AR-antagonist CGP20712A (Kd = 21 µM). The affinity of ICI118551, a ligand selective for ß2-AR in other species, was uncharacteristically low in porcine LM (Kd = 441 nM), but was consistent with a value reported for the cloned porcine ß2-AR. In heart ventricle, ligand binding revealed a predominant population of ß1-AR, judged by the rank order of affinity of catecholamines [(–)isoproterenol > (–)epinephrine 
(–)norepinephrine] and high-affinity binding to CGP20712A (Kd = 40 nM). The Kd for ICI118551 (731 nM) was close to that observed at ß2-AR in LM, confirming that ICI118551 is not subtype-selective in the pig. Displacement studies using (–)propranolol, clenbuterol, and (–)isoproterenol revealed a second high-affinity binding site in the heart that was not a ß2-AR and could not be eliminated by guanosine 5'-triphosphate or guanylyli-midodiphosphate. In adipose tissue, an equal number of ß1- and ß2-AR was identified through the binding of clenbuterol and CGP20712A, whereas ICI118551 could not discriminate between these sites. In further experiments, we used 10 µM CGP20712A to eliminate ß1-AR binding and allow accurate Kd values to be determined at ß2-AR for nonselective ligands. Under these conditions, another binding site was observed that had a high affinity for (–)propranolol (Kd = 20 pM), which is inconsistent with ß3- or ß4-AR binding reported elsewhere. Our results indicate that porcine adipose tissue contains ß1-AR, ß2-AR, and an atypical binding site in the proportions 50, 34, and 16%, respectively, of the total binding sites labeled by [3H]CGP12177.
Key Words: Adipose Tissue • ß-Adrenergic Receptors • Heart • Muscle • Pigs
|
Introduction
|
|---|
Synthetic ß-adrenergic agonists are used in commercial pig production. Their repartitioning effects are mediated by a family of ß-adrenergic receptors (ß-AR), which contains up to four subtypes and has been subjected to many pharmacological studies (Caron and Lefkowitz, 1993
; Mersmann 1998, 2002). Subtypes of the ß-AR are well characterized in many mammalian species, but they have been difficult to delineate in porcine adipose tissue, either through studies on the adrenergic control of lipolysis (Liu et al., 1989) or direct ligand binding experiments (Coutinho et al., 1992; Mersmann and McNeel 1992; Mersmann et al., 1993). Part of the difficulty is that porcine ß-AR do not show the degree of selectivity for classic ß-AR ligands typical of other animal species (Coutinho et al., 1992; Liang and Mills, 2001; Mersmann, 2002). Another factor is that radioligands that are ideal for labeling ß-AR in some tissues and species produce variable or unreliable results in other circumstances (Mersmann and McNeel, 1992) in part because of the variable presence of non-adrenoceptor binding sites (Sillence and Matthews, 1994).
Through cloning of the porcine ß1- and ß2-AR (Liang et al., 1997; Cao et al., 1998), the tissue distribution of porcine ß-AR mRNA can be measured (McNeel and Mersmann 1995,1999), and the ligand binding profile of each subtype can be examined in isolation, using vectors such as Chinese hamster ovary (CHO) cells (Liang and Mills, 2001). Nonetheless, ligand binding affinities for these cloned receptors need to be reconciled with data obtained from native receptors in porcine tissues. Furthermore, the study of cloned receptors alone will not identify ß-AR subtypes that are yet to be discovered.
In the present study, we have taken a systematic approach to delineating porcine ß1- and porcine ß2-AR in adipose tissue by first studying the ligand binding profile of these receptors in heart and skeletal muscle, which were expected to contain a predominant or homogeneous population of each subtype. Thus, we sought to identify ligands that were highly selective for porcine ß1- and porcine ß2-AR, rather than to predict their selectivity from data obtained in other species.
|
Materials and Methods
|
|---|
Preparation of Heart, Adipose Tissue, and Skeletal Muscle Membranes
Six Yorkshire barrows (22 wk of age) were taken to a commercial abattoir, stunned using CO2, and killed. Tissue samples were collected from the neck (adipose tissue), the center of the skeletal muscle (LM), and the heart (both ventricles pooled) within 20 min of death. The tissues were frozen in liquid N2 then stored at –80°C until used for membrane preparation.
A crude preparation of cell membrane fragments was obtained using the method described by Sillence et al. (1993). Portions of frozen tissue (5 to 10 g) were weighed, sliced, and suspended in 25 mL of ice-cold homogenate buffer (50 mM Tris 7.0, 250 mM sucrose, 1 mM EGTA; pH 7.4 at 4°C). The tissues were homogenized for 30 s and then centrifuged at 1,000 x g for 10 min. The supernatant fraction was filtered through gauze and recentrifuged at 10,000 x g for 15 min. The supernatant fraction was collected again and centrifuged a third time at 100,000 x g for 30 min. The supernatant fraction was discarded, and the pellet was resuspended in 1 mL of incubation buffer (50 mM Tris 7.7, 10 mM Mg Cl2, 150 mM NaCl; pH 7.4 at 37°C). Protein concentration was measured using the Bradford assay (BioRad, Hercules, CA) with BSA standards. Centrifugations were performed at 4°C, and the tissue preparation was kept in an ice-bath during other stages of processing. Samples of cell membrane suspension were stored at –80°C until assayed for ß-AR.
Radioligand Binding Assays
In preliminary experiments, we attempted to label ß-AR in LM using the well-characterized radioligand 125I-iodocyanopindolol ([125I]ICYP; Mersmann and McNeel, 1992; Sillence and Matthews, 1994; Liang and Mills, 2002). In agreement with earlier studies by Liang and Mills (2002), we found that [125I]ICYP seemed to label two binding sites, that accurate and repeatable Scatchard plots could not be produced (18 attempts), and that values for nonspecific binding (NSB) were unusually high (50 to 70% of total binding). Based on our previous experience in identifying lipophilic sites using [125I]ICYP that are not functional ß-AR (Sillence et al., 1993; Sillence and Matthews, 1994), we judged this radioligand to be unsuitable for the intended purpose. In all subsequent experiments, we used the more hydrophilic compound [3H]CGP12177 (44.5 Ci/mmol).
All reagents were prepared in incubation buffer, and polypropylene tubes (12 mm x 75 mm) were used throughout. Assays were performed in triplicate by incubating 100 µL of cell membrane suspension (incubation concentration = 0.5 to 0.6 mg of protein/mL) with 50µL of [3H]CGP12177 and 100 µL of incubation buffer. Maximum binding occurred when incubations were conducted for 1 h at 37°C in a shaking water bath. Separation of bound from free radioligand was achieved by filtering through presoaked glass-fiber filters (Whatman GF/B; Whatman Ltd., Maidstone, UK) using a 48-well cell harvester (Brandell, Gaithersburg, MD). Each tube was rinsed five times with 5 mL of ice-cold incubation buffer. Radioligand bound to the cell membrane fragments was retained on the filter papers, which were collected and soaked overnight in tubes containing 3 mL of scintillation fluid (Emulsifier-safe, Packard, Groningen, The Netherlands). The tubes were shaken thoroughly, and radioactivity was counted in a liquid scintillation counter that had an efficiency of 59% (LKB-Wallac, Turku, Finland).
Saturation Binding Studies
Six saturation experiments were performed using each tissue to obtain an accurate estimate of the equilibrium dissociation constant (Kd) of the radioligand. Eight concentrations of [3H]CGP12177 were used, ranging from 90 µM to 14 nM. Nonspecific binding was determined by replacing 50 µL of incubation buffer with 50µL of (–)propranolol at a final concentration of 2 µM.
Competition Studies
Equilibrium dissociation constants were determined for several ß-AR ligands. Each ligand was diluted over a range of concentrations then placed in competition with a fixed concentration of [3H]CGP12177 (11.4 nM). When the competing ligands were catecholamines, (+)ascorbic acid (2.8 mM) was included in the incubation buffer to decrease the rate of oxidation of these compounds. Although the binding of [3H]CGP12177 to porcine adipocyte membranes can be decreased by ascorbate and Na+ and Mg++ ions (Mersmann and McNeel, 1992), this was not observed in the present study, and all of the tubes contained the same concentration of these reagents.
Data Analysis
Data were analyzed using the nonlinear curve-fitting computer program LIGAND (Elsevier-BIOSOFT, Cambridge, UK). This program treats the affinity constant of the ligand(s), the maximum binding capacity of the sample, and the level of NSB as unknown variables, which are estimated from the binding data using an appropriately weighted least squares curve-fitting algorithm.
In the present study, an initial estimate of NSB was obtained by measuring the amount of radioligand bound in tubes containing 2 µM (–)propranolol, which would have saturated any ß1- or ß2-AR present. In the case of competitive displacement studies, this value would usually approximate to the valley observed toward the bottom of the displacement curve. The initial estimate for NSB, together with the quantity of radioligand bound in all other tubes, was entered into LIGAND. The program was then able to derive a more accurate estimate of NSB by performing a number of iterations to arrive at the line of best fit, as reflected by the lowest residual variance.
It should be noted that, although the LIGAND program is an excellent tool to derive accurate estimates of binding variables, as well as to discriminate between displacement curves that reflect simple one-site binding kinetics and those which do not, no such program can distinguish between physiological receptors and binding artifacts, such as cooperativity, high-affinity NSB sites, and agonist-induced high-affinity receptor states. Thus, all radioligand binding data should be treated with caution, particularly data that indicate the presence of "atypical" binding sites. Details of the mathematical models used by the LIGAND program, together with a comprehensive discussion on the strengths and weaknesses of this method of analysis, have been published elsewhere (Munson and Rodbard, 1980).
Drugs and Chemicals
The compounds (–)norepinephrine (+)bitartrate, guanylylimidodiphosphate (GppNHp), and guanosine 5'-triphosphate (GTP) were purchased from Sigma Chemical (St. Louis, MO). The catecholamines (+)isoproterenol and (–)epinephrine were gifts from Sterling-Winthrop Research Institute (Rensselaer, NY). Ciba-Geigy (Basel, Switzerland) supplied CGP20712A [2-hydroxy-5[2-[[2-hydroxy-3-[4-[(1 methyl-4-trifluromethyl)1H-imidazole-2-yl]phenoxy]propyl]amino]ethoxy]-benzamine]. The compounds (–)propranolol, clenbuterol, and (±)propranolol were purchased from ICN Pharmaceuticals (Costa Mesa, CA). The radioligand [3H]CGP12177 [(–)-4-(3-t-butyl-amino-2-hydroxypropoxy)benz-imidazole-2-one] was purchased from NEN Research Products (Boston, MA). The antagonist ICI118551 [erythro-DL-1(7-methylindan-4-yloxy)-3-isopropylaminobutan-2-ol] was synthesized by G. Pegg and T. Badran at Central Queensland University (Rockhampton, Queensland, Australia).
|
Results
|
|---|
Skeletal Muscle
The specific binding of [3H]CGP12177 to LM membranes was saturable and of high affinity (Kd = 230 ± 33 pM; Figure 1A). Scatchard plots of the data were linear (Figure 1A), and the mean Hill coefficient was close to unity (0.95 ± 0.14), consistent with the presence of a single specific binding site. The mean density of binding sites was 58.5 ± 3.6 fmol/mg of protein. Nonspecific binding was not saturable and increased linearly with increasing concentrations of radioligand. Under the conditions used for competitive displacement studies, the mean NSB was 16%, with a range of 10 to 20%.
View larger version (16K):
[in this window]
[in a new window]
|
Figure 1. Specific binding of [3H]CGP12177 to cell membrane fragments from porcine skeletal muscle (Panel A), heart ventricle (Panel B), and subcutaneous adipose tissue (Panel C). Membranes were incubated with increasing concentrations of radioligand, and nonspecific binding was determined using 2 µM (–)propranolol. Each graph is representative of six experiments. Insets show Scatchard plot of these data, indicating a single binding site in each tissue.
|
|
The results of competitive displacement studies are shown in Table 1. The binding site showed >100-fold stereo-selectivity for the enantiomers of isoproterenol, consistent with the labeling of ß-AR rather than NSB sites. The rank order of affinity for the catecholamines was (–)isoproterenol > (–)epinephrine > (–)norepinephrine, which is consistent with a predominant population of ß2-AR (Figure 2a). The affinity of clenbuterol was high (Kd = 16 nM), which also is consistent with ß2-AR binding, whereas the affinity of ICI118551 (a ligand selective for ß2-AR in most species) was uncharacteristically low (Kd = 441 nM). Hill coefficients were close to unity for all compounds except CGP20712A, consistent with the presence of a predominant population of ß2-AR. The large Hill coefficient observed for CGP20712A (a compound selective for ß1-AR) may indicate positive co-operativity; however, these data should be treated with caution because it is difficult to obtain reliable binding estimates using any ligand that has such a low affinity for the receptor. The absence of any high-affinity binding with CGP20712A is consistent with the absence of a significant number of ß1-AR in porcine skeletal muscle.
View this table:
[in this window]
[in a new window]
|
Table 1. Equilibrium dissociation constants and Hill coefficients for agonist and antagonist binding to porcine ß-adrenergic receptors (AR) in skeletal muscle (LM) compared with values reported for recombinant porcine ß2-AR expressed in Chinese hamster ovary cellsa
|
|
Heart
The NSB of [3H]CGP12177 to membranes from heart ventricle was not saturable and ranged from 10 to 20%, with a mean of 16%. Specific binding was saturable and was of high affinity (Kd = 566 ± 44 pM), with a Hill coefficient close to unity (0.99 ± 0.40) and a receptor density of 112 ± 5 fmol/mg of protein (Figure 1B
). There was no indication of multiple binding sites using [3H]CGP12177 alone. The results of competition studies were consistent with porcine ventricle containing a predominant population of ß1-AR (Table 2). The rank order of affinity of the catecholamines was (–)isoproterenol > (–)epinephrine (–)norepinephrine (Figure 2B), and high-affinity binding was observed with the ß1-AR-selective ligand CGP20712A. The Kd for ICI118551 at porcine ß1-AR in heart ventricle (731 ± 26 nM) was close to the value observed at ß2-AR in skeletal muscle (441 nM), showing that ICI118551 is not subtype-selective in the pig.
View this table:
[in this window]
[in a new window]
|
Table 2. Equilibrium dissociation constants and Hill coefficients for agonist and antagonist binding to porcine ß-adrenergic receptors (AR) in heart ventricle compared with values reported for recombinant human ß1-AR expressed in Escherichia colia
|
|
In addition to the population of ß1-AR, a second binding site was revealed through the shallow or biphasic displacement of radioligand by (–)isoproterenol, propranolol, clenbuterol, (–)norepinephrine, and (–)epinephrine. A two-site model could not be fitted to all of the displacement curves; for example, the curve for (–)epinephrine reached a plateau at 80% binding, and an accurate Kd value for the high-affinity site was not obtained (Figure 2B). Nonetheless, two-site models gave a lower residual variance than a one-site model when describing the curves generated using (–)isoproterenol, clenbuterol, and propranolol (P < 0.05; Table 3). For example, the mean residual variance for propranolol curves was 63.3 for one-site models vs. 25.4 for two-site models. The additional high-affinity site accounted for 23% of the total population of binding sites when (–)propranolol was used as the displacing ligand.
View this table:
[in this window]
[in a new window]
|
Table 3. Equilibrium dissociation constants and relative binding site densities (Bmax) for the binding of agonists and an antagonist to high-affinity (pKdh) and low-affinity sites (pKdl) in cell membranes from porcine heart ventriclea
|
|
The high-affinity site bound to the ß2-AR agonist clenbuterol with a high Kd value of 22 nM, but CGP20712A, (–)isoproterenol, and (–)propranolol all yielded high Kd values, inconsistent with ß2-AR binding.
To eliminate the possibility that the second binding site represented an agonist-induced high-affinity state of the ß1-AR, competition studies using (–)isoproterenol were repeated in the presence of GTP and the non-hydrolyzable GTP analog GppNHp (Table 3). The results obtained with GTP were still consistent with the presence of two binding sites. With GppNHp, a two-site model gave a statistically better fit in only one experiment (P < 0.05), but Hofstee plots of all four studies were curvilinear, consistent with the presence of two binding sites (Figure 3). The observation that the high-affinity binding site was observed using the ß-AR antagonist (–)propranolol also is inconsistent with an agonist-induced high-affinity state.
View larger version (15K):
[in this window]
[in a new window]
|
Figure 3. Hofstee plot showing the pattern of displacement of [3H]CGP12177 (11.4 nM) from heart ventricle membranes by increasing concentrations of (–)isoproterenol. The curvilinear plot indicates the presence of multiple binding sites, despite the inclusion of guanylylimidodiphosphate to prevent agonist-induced high-affinity binding to adrenergic receptors. The plot is representative of four displacement experiments.
|
|
Adipose Tissue
Nonspecific binding of [3H]CGP12177 to adipose tissue membranes ranged from 20 to 30%, with a mean level of 26%. Specific binding was saturable and of high affinity (Kd = 233 ± 48 pM). The mean Hill coefficient was less than unity (0.77 ± 0.07), but a two-site binding model could not be fitted to the data successfully (Figure 1C). The mean receptor density was 211 ± 10 fmol/mg of protein.
The results of competition studies using adipose tissue are summarized in Table 4. Experiments using clenbuterol were conducted in the presence of Gpp(NH)p to eliminate agonist-induced high-affinity binding. Nevertheless, the pattern of displacement resulted in a low Hill coefficient (0.59 ± 0.01) and displacement curves that were described better by a two-site model than a one-site model (P < 0.05; Figure 4). The low Kd for clenbuterol (309 nM) corresponded to the value obtained for clenbuterol binding to ß1-adrenoceptors in heart ventricle. The high Kd value (6 nM) was closest to the Kd for clenbuterol binding to ß2-AR in LM (16 nM), but was also consistent with the atypical high-affinity site in heart (22 nM). The proportions of each binding site in adipose tissue were 49 and 51%, respectively.
View this table:
[in this window]
[in a new window]
|
Table 4. Equilibrium dissociation constants and Hill coefficients for the binding of ligands selective for ß1- and ß2-adrenergic receptors (AR) to cell membranes from porcine adipose tissuea
|
|
View larger version (12K):
[in this window]
[in a new window]
|
Figure 4. Competitive displacement of [3H]CGP12177 (11.4 nM) from cell membrane fragments of porcine subcutaneous adipose tissue by increasing concentrations of clenbuterol. Membranes were incubated in the presence of guanylylimidodiphosphate to prevent agonist-induced high-affinity binding to adrenergic receptors. The shallow displacement curve was fit best by a model that described two binding sites (putative ß1- and ß2-adrenergic receptors) in approximately equal proportions.
|
|
Competition studies using ICI118551 failed to reveal two binding sites, consistent with the lack of selectivity shown by this ligand for binding to ß1- and ß2-AR in heart and LM. In contrast, experiments using CGP20712A confirmed that this drug is highly selective between ß1- and ß2-AR. Fifty percent of CGP20712A binding was to a high-affinity site (Kd = 4 ± 0.11 nM), presumably the ß1-AR. The affinity of CGP20712A for the second site was too low to be modeled accurately by LIGAND; however, slightly <50% of the radioligand bound to this site was displaced at 100 µM CGP20712A, inferring a Kd of <2 µM, which is consistent with ß2-AR binding.
The high degree of subtype selectivity of CGP20712A (>450-fold) was exploited in subsequent binding studies where CGP20712A (10 µM) was added to the incubation buffer. Based on the affinity constant for this ligand which is consistent between species, this concentration would be expected to occupy <33% of any ß2-AR present, while blocking 99.6% of any ß1-AR present. In this way, accurate Kd values could be obtained for the binding of nonselective ligands to ß2-AR in adipose tissue. These competition studies yielded a rank order of affinity of (–)isoproterenol > (–)epinephrine > (–)norepinephrine for the catecholamines, which is consistent with ß2-AR binding (Figure 2C). Nonetheless, Hill coefficients were low, suggesting multiple binding sites (Table 4). The degree of selectivity of the catecholamines for these sites was too low to allow accurate two-site models to be fitted; however, a two-site binding model was fitted successfully to the displacement curves generated by (–)propranolol (P < 0.01) in the presence of CGP20712A (Figure 5). The two binding sites formed 68 and 32% of the non-ß1-AR population. The larger site had a Kd of 10 nM, comparable with the value obtained at ß2-AR in skeletal muscle, whereas the smaller site had a Kd value of 20 pM, similar to that observed for propranolol binding to the atypical high-affinity site in heart ventricle. Thus, porcine adipose tissue contains at least three sites that bind to the ß-AR radioligand [3H]CGP12177.
View larger version (11K):
[in this window]
[in a new window]
|
Figure 5. Competitive displacement of [3H]CGP12177 (11.4 nM) from cell membrane fragments of porcine subcutaneous adipose tissue by increasing concentrations of (–)propranolol. Inclusion of CGP20712A (10 µM) eliminated binding to ß1-adrenergic receptors (approximately 50 to 60% of the total population). The biphasic displacement curve represents the remaining population of ß2-adrenergic receptors (equilibrium dissociation constant, Kd = 10 nM) and an additional binding site (Kd = 20 pM) that is yet to be characterized.
|
|
|
Discussion
|
|---|
Our results show that ß2-adrenoceptors are the predominant subtype in porcine LM, which is consistent with findings in LM from many other species as diverse as rats (Williams et al., 1984), cattle (Sillence et al., 1991), and cane toads (Javro, 1996). Even though mRNA for ß1-AR is present in porcine LM (McNeel and Mersmann, 1999), our results indicate that ß1-AR are not expressed in great abundance. In fact, ß1-AR were undetectable in LM by radioligand binding, as shown by the absence of any high-affinity binding sites for the ß1-AR-selective ligand CGP20712A. Our results do not rule out the presence of a small population of functional ß1-AR that is undetectable under these experimental conditions, as was found to be the case in bovine LM (Sillence and Matthews, 1994). Our finding is clearly at variance, however, with that of Liang and Mills (2002), who argued that porcine LM contains a predominant population of ß1-AR. This is a significant point, as it has fueled much debate over the mechanism of action of ß-AR agonists in pigs (Mersmann, 2002).
The results of Liang and Mills (2002) are based on ligand-binding studies using [125I]ICYP. This high-affinity radioligand is ideal under many circumstances, especially when the availability of membrane protein is limited. Researchers have used [125I]ICYP to characterize recombinant ß-AR in E. coli and CHO cells and have produced reliable estimates of binding affinity for a range of selective and nonselective ß-AR ligands in these systems (Marullo et al., 1989; Liang and Mills, 2001). When used in animal tissues, however, [125I]I-CYP has generated binding artifacts and is particularly prone to label lipophilic sites such as those identified in rat brown adipose tissue (Sillence et al., 1993) and soleus muscle (Roberts et al., 1993) and in bovine skeletal muscle and adipose tissue (Sillence and Matthews, 1994), which are not functional ß-AR.
We believe that the dominant binding site identified by Liang and Mills (2002) in porcine LM (and possibly in adipose tissue) was not the ß1-AR. Under the binding conditions used, Liang and Mills (2002) reported that the relative proportion of ß1- and ß2-AR in adipose tissue depended on the concentration of the radioligand. Thus, [125I]ICYP showed a marked selectivity between the putative ß1- and ß2-AR, which is inconsistent with its lack of selectivity between these cloned subtypes in CHO cells. Notwithstanding this anomaly, Liang and Mills (2002) attempted to derive Kd values for competing ligands on the basis of the [125I]ICYP Kd values observed in earlier studies using CHO cells, rather than those apparent in the native tissue. As those researchers acknowledge, determining Kd values in heterogeneous systems is extremely problematic when both the radioligand and competing ligand are selective. Another anomaly reported by Liang and Mills (2002) is that the Kd of BRL37344 for binding to putative ß2-AR was an order of magnitude lower than that observed in CHO cells. The binding conditions used by Liang and Mills (2002) were improved somewhat by blocking ß1-AR using CGP20712A, the same strategy we used in the present study; however, under these conditions Liang and Mills (2002) were still unable to generate acceptable Scatchard plots or decrease NSB levels, which were as high as 80% of total binding. Thus, whereas similar results were observed using [125I]ICYP in the present study and in the experiments reported by Liang and Mills (2002), our conclusions are different. We do not believe that reliable estimates of the proportion of ß1- and ß2-AR in porcine tissues can be obtained using this radioligand. Furthermore, as the present study was designed to characterize and quantify ß-AR, our use of the more hydrophilic radioligand [3H]CGP12177, which labeled a homogeneous population of ß2-AR in LM, seemed appropriate.
Although the AA sequence and ligand binding specificity of ß2-AR are highly conserved among most mammals, the porcine ß2-AR is slightly unusual. Liang et al. (1997) decoded the gene sequence for the porcine ß2-AR and compared its structure with that of the human ß2-AR. Amino acid substitutions were identified in the porcine receptor at sites that are known in other species to be critical for ligand binding. A prediction of the tertiary structure of the receptor suggests several "kinks" in the porcine protein not observed in human ß2-AR. Furthermore, Liang and Mills (2001) reported an unusually low affinity for ICI118551 at recombinant porcine ß2-AR expressed in CHO cells. Despite using a different radioligand, our results at native porcine ß2-AR compare favorably with the CHO cell values, including the low affinity for ICI118551. This compound is significant, as it is one of the few antagonists that display a high degree of selectivity in binding between classical ß1- and ß2-AR, and so it has been used in many studies to delineate the two subtypes. Its lack of subtype selectivity in the pig has been suggested in several earlier studies (Coutinho et al., 1992; Mersmann and McNeel, 1992; Mersmann et al., 1993), but this may still account for some of the confusion in the literature over the character of porcine ß-AR in both muscle and adipose tissue.
Adrenergic receptors of the ß1-subtype predominate in cardiac tissue in most animals, but when Sato et al. (1997) used (–)isoproterenol to displace the radioligand [125I] ICYP from ß-ARs in porcine sub-endocardium, they observed two binding sites with Kd values for (–)isoproterenol of 54 nM and 1,100 nM. These data are strikingly similar to those obtained for (–)isoproterenol vs. [3H]CGP12177 in the present study (32 and 1,622 nM), supporting the existence of the two sites. The possibility that one of these sites is an agonist-induced high-affinity form of the ß1-adrenoreceptor could not be ruled out from the results of Sato et al. (1997), but it can be dismissed in the present study because the site was evident in the presence of GTP, GppNHp, and the antagonist (–)propranolol. In some animals, the heart contains ß2-AR. Prengle et al. (1996) reported ß1- and ß2-AR in porcine right atrium, whereas McNeel and Mersmann (1999) detected ß1-and ß2-AR mRNA in porcine left ventricle. Nonetheless, the possibility that the second site observed in the present study is a ß2-AR also can be ruled out. The compound CGP20712A, which is highly selective between porcine ß1- and ß2-AR, did not discriminate between the two binding sites in the heart, whereas (–)isoproterenol and (–)propranolol, which are nonselective for porcine ß1- and porcine ß2-AR, revealed the two sites. The high-affinity constants for the last two ligands also are inconsistent with those observed for ß3-AR binding in other species. It has been proposed that mammalian heart contains a fourth ß-AR subtype (Kaumann, 1997), which can be radiolabelled with [3H]CGP12177 and shows the same rank order of affinity for the catecholamines observed in the present study. However, CGP20712A is highly selective between the ß1-AR and the putative fourth cardiac ß-AR, which contrasts with the results of the present study. In common with the ß3-AR, the putative fourth cardiac ß-AR also is propranolol-resistant (Kd < 2 µM), whereas the atypical binding site we observed in the porcine ventricle was propranolol-sensitive (Kd = 23 pM). Thus, our observation is inconsistent with the ß2-AR, the ß3-AR, and the putative cardiac ß4-AR, and, with the limited data available at present, it is best described as an atypical ß-adrenergic binding site.
The existence of multiple ß-AR subtypes in porcine adipose tissue and in adipocytes is well-supported in the literature (Mersmann, 2002). Accurate Kd values have been difficult to obtain, however, because investigators have attempted to characterize these multiple subtypes simultaneously, and by reference to Kd values reported for other species, rather than from other porcine tissues. Liang and Mills (2002) took a more rational approach by first identifying selective ligands using recombinant porcine ß-AR (Liang and Mills, 2001) but, as discussed earlier, their choice of radioligand was unfortunate. In the present study, we tested three candidate drugs that show ß-AR subtype selectivity in other species. The ß2-AR-selective antagonist ICI118551 failed to show any selectivity between porcine muscle (ß2) and heart (ß1) receptors, in contrast to the 100- to 150-fold selectivity shown in humans (Mauriege et al., 1988; Marullo et al., 1989). Clenbuterol showed some preference for ß2-AR (50-fold), whereas CGP20712A showed marked selectivity (400-fold) for ß1-AR. Thus, the last two compounds can be used to determine the relative proportions of ß1- and ß2-AR in porcine tissues.
In the present study, the proportion of total ß-AR in adipose tissue represented by ß1-AR was approximately 50%, similar to the proportion reported by Coutinho et al. (1992), but somewhat less than the 70% reported for ß1-AR transcripts in porcine adipose tissue (McNeel and Mersmann, 1999) or for putative ß1-AR identified in adipocyte membranes by Liang and Mills (2002). A possible reason for this discrepancy is that our crude preparation would have had some contamination with tissue other than adipocyte membranes; however, preliminary studies in our laboratory using isolated adipocytes have not confirmed this (M. N. Sillence, unpublished data). Another possibility is that the population of porcine ß1-AR reported by Liang and Mills (2002) is an overestimate because of the methodological limitations of using [125I]ICYP as the radioligand.
As with the porcine ventricle, an atypical binding site also was present in adipose tissue. The relative proportion of ß1-, ß2-, and atypical binding sites was 50, 34, and 16%, respectively. The atypical site had a similar affinity for clenbuterol as did the ß2-AR, a relatively high affinity for (–)epinephrine and (–)norepi-nephrine, a high affinity for ICI118551, and a very high affinity for (–)propranolol. The atypical site is unlikely to be a ß3-AR, even though mRNA for the ß3-AR is present in porcine adipose tissue (McNeel and Mersmann, 1999). The Kd values for this binding site are all inconsistent with ß3-AR binding in other species, and our experiments were designed specifically using low concentrations of [3H]CGP12177 (11.4 nM) that would label only a minimal proportion of human ß3-AR (Emorine et al., 1989). Furthermore, Mersmann (1996) was unable to label ß3-AR in porcine adipose tissue using this radioligand, even though porcine ß3-AR mRNA was present. Thus, [3H]CGP12177 was the radioligand of choice, enabling the characterization of ß1- and ß2-subtypes without ß3-AR interference.
The atypical binding sites in adipose tissue and heart are similar in character, and in particular, both sites have a high affinity for (–)propranolol (Kd = 20 and 23 pM, respectively); however, high-affinity (–)propranolol binding was apparent in the presence of 10 µM CGP20712A, which blocked the atypical binding site in the heart. Thus, the cardiac and adipose tissue sites may be different. Receptors with such a high affinity for (–)propranolol have not been observed in our previous binding studies using cane toads, turkeys, rats, or cattle (Sillence et al., 1991, 1993; Javro, 1996). We have no evidence that the atypical binding site is a functional adrenergic receptor, and this needs further investigation.
In summary, porcine LM contains ß2-AR that are similar to ß2-AR in other species, except for a low affinity for ICI118551. Porcine ventricle contains a predominant population of ß1-AR, similar in character to those found in other species, but with proportionally lower affinity for the catecholamines (–)isoproterenol, (–)epinephrine, and (–)norepinephrine. Porcine adipose tissue contains a mixed population of ß1- and ß2-AR. Furthermore, porcine adipose tissue and heart both contain a population of atypical adrenergic binding sites, not attributable to ß1-, ß2-, or ß3-AR.
|
Implications
|
|---|
ß-Adrenergic agonists are used in commercial pig production to improve the efficiency of lean tissue growth. To maximize the benefits and minimize the adverse effects of ß-adrenergic agonists, it is desirable to understand the character and distribution of adrenergic receptors in various tissues. Porcine heart ventricle contains predominantly ß1 adrenergic receptors and LM contains predominantly ß2 adrenergic receptors, whereas adipose tissue contains approximately equal proportions of both subtypes. Adipose tissue and heart also contain an atypical site that binds to adrenergic compounds, which needs further investigation.
|
Footnotes
|
|---|
1 This research was funded in part by the Australian Centre for International Agricultural Research. We thank A. Ferraro for help with preparing this manuscript. 
2 Correspondence: P.O. Box 588 (61-2-69332205; fax: 61-2-69332995; e-mail: msillence{at}csu.edu.au).
Received for publication November 17, 2004.
Accepted for publication June 19, 2005.
|
Literature Cited
|
|---|
Cao, H., C. A. Bidwell, S. K. Williams, W. Liang, and S. E. Mills. 1998. Rapid communication: nucleotide sequence of the coding region for the porcine ß1-adrenergic receptor gene. J. Anim. Sci. 76:1720–1721.[Free Full Text]
Caron, M. G., and R. J. Lefkowitz. 1993. Catecholamine receptors: structure, function and regulation. Recent Prog. Horm. Res. 48:277–290.
Coutinho, L. L., W. G. Bergen, R. A. Merkel, and C. K. Smith. 1992. Quantitative characterization of beta-adrenergic receptor subtypes in porcine adipocytes. Comp. Biochem. Physiol. 101C(3):481–485.
Emorine, L. J., S. Marullo, M.-M. Briend-Sutren, G. Patey, K. Tate, C. Delavier-Klutchko, and A. D. Strosberg. 1989. Molecular characterization of the human ß3-adrenergic receptor. Science 245:1118–1121.[Abstract/Free Full Text]
Javro, C. A. 1996. Evaluation of the mechanisms by which animal growth promotants operating through ß2-adrenoceptors induce muscle growth: Methodological development and a study of clenbuterol. Ph.D. Thesis. Central Queensland Univ., Rockhampton, Queensland, Australia.
Kaumann, A. J. 1997. Four ß-adrenoceptor subtypes in the mammalian heart. Trends Pharmacol. Sci. 18:70–76.[Medline]
Liang, W., C. A. Bidwell, S. K. Williams, and S. E. Mills. 1997. Rapid communication: molecular cloning of the porcine ß2-adrenergic receptor gene. J. Anim. Sci. 75:2824.[Free Full Text]
Liang, W., and S. Mills. 2001. Profile of ligand binding to the porcine ß2-adrenergic receptor. J. Anim. Sci. 79:877–883.[Abstract/Free Full Text]
Liang, W., and S. E. Mills. 2002. Quantitative analysis of ß-adrenergic receptor subtypes in pig tissues. J. Anim. Sci. 80:963–970.[Abstract/Free Full Text]
Liu, C. Y., J. L. Boyer, and S. E. Mills. 1989. Acute effects of beta-adrenergic agonists on porcine adipocyte metabolism in vitro. J. Anim. Sci. 67:2930–2936.
Marullo, S., C. Delavier-Klutchko, J. G. Guillet, A. Charbit, A. D. Strosberg, and L. J. Emorine. 1989. Expression of human ß1- and ß2- adrenergic receptors in E. coli as a new tool for ligand screening. Biotechnology (N.Y.) 7:923–927.
Mauriege, P., G. De Pergola, M. Berlan, and M. Lafontan. 1988. Human fat cell beta-adrenergic receptors: Beta-agonist-dependent lipolytic responses and characterization of beta-adrenergic binding sites on human fat cell membranes with highly selective beta1-antagonists. J. Lipid Res. 29:587–601.[Abstract]
McNeel, R. L., and H. J. Mersmann. 1995. ß-Adrenergic receptor subtype transcripts in porcine adipose tissue. J. Anim. Sci. 73:1962–1971.[Abstract]
McNeel, R. L., and H. J. Mersmann. 1999. Distribution and quantification of beta1-, beta2- and beta3-adrenergic receptor subtype transcripts in porcine tissues. J. Anim. Sci. 77:611–621.[Abstract/Free Full Text]
Mersmann H. J. 1996. Evidence of classic ß3-adrenergic receptors in porcine adipocytes. J. Anim. Sci. 74:984–992.[Abstract]
Mersmann, H. J. 1998. Overview of the effects of ß-adrenergic receptor agonists on animal growth including mechanisms of action. J. Anim. Sci. 76:160–172.[Abstract/Free Full Text]
Mersmann, H. J. 2002. Beta adrenergic receptor modulation of adipocyte metabolism and growth. J. Anim. Sci. 80(Suppl. 1):E24–E29.[Abstract/Free Full Text]
Mersmann, H. J., K. Akanabi, A. Shparber, and R. L. McNeel. 1993. Binding of agonists and antagonists to the porcine adipose tissue ß-adrenergic receptor(s). Comp. Biochem. Physiol. 106C:725–732.
Mersmann, H. J., and R. L. McNeel. 1992. Ligand binding to the porcine adipose tissue ß-adrenergic receptor. J. Anim. Sci. 70:787–797.[Abstract]
Munson, P. J., and D. Rodbard. 1980. LIGAND: a versatile computerized approach for characterization of ligand-binding systems. Anal. Biochem. 107:220–239.[Medline]
Prengle, A. W., K. G. Lindner, T. Anhaupl, J. Vogt, and K. G. Lurie. 1996. Regulation of right atrial ß-adrenoceptors after cardiopulmonary resuscitation in pigs. Resuscitation 31:271–278.[Medline]
Roberts, S. J., P. Molenaar, and R. J. Summers. 1993. Characterisation of propranolol-resistant [125I]I-iodocyanopindolol binding sites in rat soleus muscle. Br. J. Pharmacol. 109:344–352.[Medline]
Sato, S., N. Sato, R. J. Kudej, M. Uiechi, K. K. Asai, Y.-T. Shen, Y. Ishikawa, S. F. Vatner, and D. E. Vatner. 1997. ß-Adrenergic receptor signalling in stunned myocardium of conscious pigs. J. Mol. Cell. Cardiol. 29:1387–1400.[Medline]
Sillence, M. N., and M. L. Mathews. 1994. Classical and atypical binding sites for ß-adrenoceptor ligands and activation of adenylyl cyclase in bovine skeletal muscle and adipose tissue membranes. Br. J. Pharmacol. 111:866–872.[Medline]
Sillence, M. N., N. G. Moore, G. G. Pegg, and D. B. Lindsay. 1993. Ligand binding properties of putative ß3-adrenoceptors compared in brown adipose tissue and in skeletal muscle membranes. Br. J. Pharmacol. 190:1157–1163.
Sillence, M. N., G. G. Pegg, and D. B. Lindsay. 1991. Affinity of clenbuterol analogues for ß2-adrenoceptors in bovine skeletal muscle, and the effect of these compounds on urinary nitrogen excretion in female rats. Arch. Pharmacol. 344:442–448.
Williams, R. S., M. G. Caron, and D. Keifer. 1984. Skeletal muscle ß-adrenergic receptors: variations due to fibre type and training. Am. J. Physiol. 246:E160–E167.
Top
Abstract
Introduction
); and (–)norepinephrine (
). Each curve is representative of four competition experiments.