Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

It is now recognized that there are at least three subtypes of the ₂-adrenoceptor, based on cloning of three distinct amino acid sequences from human (Regan et al., 1988; Kobilka et al., 1988; Lomasney et al., 1990; Weinshank et al., 1990), rat (Zeng et al., 1990; Harrison et al., 1991; Lanier et al., 1991), mouse (Chruschinski et al., 1992; Link et al., 1992), and guinea pig (Svensson et al., 1996) tissues. The three subtypes identified by these studies have a high degree of sequence homology between species; however, despite this homology, the rat clone designated as RG20 has distinct pharmacological differences from the human ₂C10. The most well recognized difference between these receptors is the low affinity of yohimbine and rauwolscine for the rat receptor. Pharmacological differences between the rat receptor and its human homolog, designated as the _2a-adrenoceptor, led to the designation of the rat receptor as having _2d-adrenoceptor pharmacology (Lanier et al., 1991). Subsequent studies have shown the single ₂-adrenoceptor cloned from the pig to have _2a-adrenoceptor pharmacology (Guyer et al., 1990), whereas the three ₂-adrenoceptors cloned from the mouse (Chruschinski et al., 1992; Link et al., 1992) and guinea pig (Svensson et al., 1996) include a receptor, homologous to the rat RG20 or human ₂C10, having _2d-adrenoceptor characteristics. Based primarily on antagonist affinity correlations in radioligand binding assays, the ₂-adrenoceptors found in native bovine tissues have been characterized as _2D-adrenoceptors (Simonneaux et al., 1991; Bylund et al., 1997). The high affinity of rauwolscine for the ₂-adrenoceptors in chicken pineal gland would suggest this system to have _2A-adrenoceptor pharmacology (Bylund et al., 1995).

Functional studies determining antagonist activity at prejunctional -adrenoceptors have been used to identify the ₂-adrenoceptor subtype in a variety of tissues from several species. Although there was initially some controversy regarding the subtypes involved in particular tissues, it now appears that prejunctional ₂-adrenoceptors in all rat tissues studied have _2D-adrenoceptor pharmacology (Trendelenburg et al., 1997). In contrast, the prejunctional ₂-adrenoceptors in human kidney correspond most closely to the pharmacology expected for the _2A-adrenoceptor (Trendelenburg et al., 1997). Consistent with the characteristics of the recombinant receptor, receptors in mouse brain and atria have _2D-adrenoceptor characteristics, as do those in guinea pig brain cortex, ileum, and atrium (Trendelenburg et al., 1997). Mouse knockout experiments, confirm the primary role of the _2D-adrenoceptor in the prejunctional control of sympathetic neurotransmission although the _2C-adrenoceptor also participates (Hein et al., 1999).

Since different tissues from a given species do not differ in their assignment between _2A- and _2D-adrenoceptor subtypes, it has been proposed that the _2a- and _2d-adrenoceptors are "species orthologs" (Bylund et al., 1995). Thus far, only three distinct ₂-adrenoceptors have been cloned from any species (human, rat, mouse, guinea pig), and the presence of both _2a- and _2d-adrenoceptors has not been clearly demonstrated in a single species. However, a consistent assignment of rabbit ₂-adrenoceptors between these two subtypes has not been possible. Radioligand binding studies in renal tubules of the rabbit showed a relatively high affinity for rauwolscine and low affinity for prazosin, consistent with the _2A-adrenoceptor subtype (Mohuczy-Dominiak and Garg, 1993). In contrast, similar studies in rabbit adipocytes showed a low affinity for yohimbine (Langin et al., 1990), comparable with that observed in rat adipocytes (Carpene et al., 1990) or platelets (Minuth and Jakobs, 1983), and comparison of affinity ratios for a series of antagonists were more consistent with the assignment of this tissue pharmacology as _2D-adrenoceptor mediated (Hieble and Ruffolo, 1996). Functional studies on presynaptic ₂-adrenoceptors in rabbit tissues consistently found _2A-adrenoceptor pharmacology (Trendelenburg et al., 1993, 1996; Limberger et al., 1995; Molderings and Göthert, 1995).

In an attempt to clarify the pharmacology of the rabbit _2A/D-adrenoceptor, we have characterized the receptor of rabbit adipocyte and platelet using a radioligand binding assay, correlating the affinity for a diverse series of compounds with their affinity for recombinant human _2a-adrenoceptors and rat _2d-adrenoceptors.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Adipocyte Isolation. Mature New Zealand White rabbits were euthanized with pentobarbital anesthesia, and fat was removed from the omentum. The tissue was rinsed in ice-cold saline, placed into 50-ml plastic centrifuge tubes containing Krebs-Ringer-bicarbonate buffer, pH 7.5, containing 6 mM glucose, 35 mg/ml bovine serum albumin, and 1.5 mg/ml collagenase. Three to 4 g of tissue/6 ml of buffer was minced and then shaken at 60 cycles/min at 37°C for about 40 min. Following digestion, the mixture was stirred rapidly and then filtered through nylon mesh. The suspension was diluted further with an equal volume of buffer. Samples were spun-down quickly to separate fat cells from the remaining undigested tissue. The aqueous phase was aspirated off by inserting a large syringe into a plastic Pasteur pipette that had been cut to a blunt end. The cells were resuspended and centrifuged two additional times, and fresh buffer was added. Samples and reagents were kept at 37°C whenever possible.

Adipocyte Membrane Preparation. The isolated adipocytes were transferred into a plastic beaker with hypotonic lysis buffer containing: 2 mM Tris-HCl (pH 7.5), 2.5 mM MgCl₂, 1 mM KHCO₃, 0.3 mM EGTA, 5 µg/ml leupeptin, 0.1 mM benzamidine, and 0.1 mM phenylmethylsulfonyl fluoride at 20^oC to 23°C. The mixture was carefully stirred, and allowed to stand for several minutes before homogenizing with a Tekmar homogenizer (Tekmar-Dohrman, Mason, OH), for three 10-s bursts The suspension was centrifuged at 40,000g for 15 min at 15°C. This step was repeated twice, and the pellets were pooled and frozen in fresh lysis buffer.

Platelet Membrane Preparation. Mature rabbits, of either sex, were anesthetized with intravenous pentobarbital. Blood was collected via cardiac puncture, in 3.8% sodium citrate (1 ml/10 ml of blood). Samples were centrifuged at ambient temperature, 270g for 15 min. The platelet-rich plasma was carefully removed and centrifuged at 26,000g at 4°C for 10 min. The resultant platelet pellet was resuspended twice in buffer containing 0.05 M Tris, pH 7.4, 0.15 M NaCl, and 0.02 M EDTA and centrifuged at 16,000g for 10 min. The pellet was resuspended in an equal volume of ice-cold lysing buffer consisting of 5 mM Tris-HCl, 5 mM EDTA, pH 7.4. (i.e., 40 ml of plasma/40 ml of buffer). The suspension was kept on ice for several minutes before homogenizing with a Brinkmann polytron (Brinkmann Instruments Inc., Westbury, NY) for several 10-s bursts, at setting number 6. The preparation was centrifuged at 39,000g at 4°C for 15 min. The pellet was washed in binding buffer; 50 mM Tris-HCl, 10 mM MgCl₂ (pH 7.4), and centrifuged again at 39,000g for 15 min. Protein was measured using the Bradford method with bovine serum albumin as the standard. Samples were frozen in liquid nitrogen and stored at 70°C until assayed.

Binding Assays. All saturation assays were performed in duplicate, with at least 12 concentrations of [³H]RX821002. In the adipocyte assay, radioligand concentrations from 0.1 to 45 nM were used, with the protein concentration from 50 to 80 µg/tube in a 300-µl total volume. Lysed platelet membrane saturation assays used radioligand concentrations from 0.5 to 60 nM with approximately 50 µg of protein in a total volume of 200 µl. Membranes were incubated for 30 min at 23°C in 50 mM Tris-HCl, pH 7.4, 0.5 mM MgCl. Nonspecific binding was defined by 10 µM phentolamine. Specific binding was calculated as the difference between total and nonspecific binding. Incubations were terminated by vacuum filtration through Whatman GF/B glass fiber filters (Whatman International, Kent, UK) using a Brandell Cell Harvester (Brandel Inc., Gaithersburg, MD). K_D and B_max were calculated using the Accufit computer program (Acufil Software, Fullerton, CA) for nonlinear regression.

Determination of Affinity for Recombinant ₂-Adrenoceptors. As previously reported (Hieble et al., 1995), a stable cell line expressing recombinant human _2a-adrenoceptors was prepared using Chinese hamster ovary cells. A stable cell line of NIH-3T3 cells expressing the rat _2d-adrenoceptor was generously provided by Dr. Steven Lanier (University of South Carolina). Radioligand binding assays were performed using [³H]rauwolscine for the _2a-adrenoceptor and [³H]RX 821002 for the _2d-adrenoceptor.

Compounds. SK&F 104078 was synthesized by the Medicinal Chemistry Department of SmithKline Beecham Pharmaceuticals. BAM 1303 and ARC 239 were generously provided by Maruko Pharmaceutical Co. (Nagoya, Japan) and Karl Thomae (Biberach, Germany), respectively. Raubasine (Ajmalicine) was obtained from Fluka Chemical Co. (Buchs, Switzerland). Norepinephrine and all other adrenoceptor antagonists were obtained from Research Biochemicals, Inc. (Natick, MA). [³H]Rauwolscine and [³H]RX 821002 were obtained from Amersham Pharmacia Biotech (Piscataway, NJ).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

[³H]RX 821002 bound to ₂-adrenoceptors of rabbit adipocyte and platelet with a relatively high affinity. As shown in Figs. 1 and 2, the binding was saturable, with a K_D = 3.4 nM, and B_max = 750 fmol/mg of protein in the adipocyte preparation and a K_D = 2.6 nM and B_max = 360 fmol/mg in the platelet membranes. The K_D value for [³H]RX821002 in the adipocyte was similar to the value of 6.0 nM observed by Langin et al. (1990). Our B_max value was higher than observed by these authors (289 fmol/mg of protein) probably as a consequence of slight differences in the procedure used to prepare the adipocyte membranes.

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Typical experiment characterizing the binding of [3H] RX821002 to rabbit platelet membranes, showing saturable binding (curve fitted by nonlinear regression, bottom panel) and Scatchard analysis (line fitted by linear regression, top panel).

View larger version (16K): [in this window] [in a new window]   Fig. 2.   Typical experiment characterizing the binding of [3H]RX821002 to rabbit adipocyte membranes, showing saturable binding (curve fitted by nonlinear regression, bottom panel) and Scatchard analysis (line fitted by linear regression, top panel).

The ability of norepinephrine and 13-adrenoceptor antagonists to inhibit [³H]RX821002 binding in these tissues is shown in Table 1. The K_i values for all agents are very similar in these two tissues. Analysis of the data showed the best fit by assuming one site with the exception of norepinephrine in the adipocyte, which was fitted better with two-site parameters (K_i, high affinity = 16 ± 4 nM; K_i, low affinity = 3400 ± 1300 nM). To facilitate comparison with other compounds, and with data in the platelet membranes, a K_i value of 57 ± 12 nM, obtained by assuming one-site competition, was used for norepinephrine in the adipocyte preparation. Correlation of pK_i values in platelet and adipocyte yields a correlation coefficient (r²) of 0.97, with a slope of 0.93 for the least-squares regression line.

View this table: [in this window] [in a new window]   TABLE 1 Affinity of drugs for 2-adrenoceptors of rabbit adipocyte and platelet or for recombinant 2a- and 2d-adrenoceptors, as determined by inhibition of [3H]rauwolscine (2a) or [3H]RX 821002 (adipocyte, platelet, 2d) binding to membrane homogenates of cells or tissues Each value represents a mean of at least three determinations ± S.E.M. The Ki values are nanomolar concentrations.

Affinity of drugs for recombinant _2a- and _2d-adrenoceptors is shown in Table 1. As in our previous studies with human recombinant -adrenoceptors, affinity for the human _2a-adrenoceptor was measured by the ability of drugs to inhibit the binding of [³H]rauwolscine. The K_D for this ligand was 1.0 nM, and the B_max was 13 pmol/mg of protein. Since rauwolscine is known to have low affinity for the _2d-adrenoceptor (O'Rourke et al., 1994a), [³H]RX821002 was used as the radioligand for studies with this adrenoceptor subtype. Binding was of high affinity (K_D = 90 pM) and the B_max was 4 pmol/mg of protein. Several reports, using both native and recombinant ₂-adrenoceptors, have shown that ₂-adrenoceptor characteristics, such as receptor density and inhibition potencies for other antagonists, are highly similar when [³H]RX821002 is compared with [³H]rauwolscine or [³H]yohimbine (Langin et al., 1989; O'Rourke et al., 1994a; Halme et al., 1995). Figures 3 and 4 show the correlation between pK_i values in rabbit adipocyte with pK_i values in recombinant human _2a-adrenoceptors and recombinant rat _2d-adrenoceptors. A substantially better correlation (r² = 0.68, slope = 0.66) is obtained with the _2a-adrenoceptor subtype. Similar results (see Table 2) were observed for correlation of data from rabbit platelet with these two recombinant ₂-adrenoceptors.

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Correlation of binding affinity of drugs for 2-adrenoceptors in rabbit adipocyte membranes with drug affinity for Chinese hamster ovary cell membranes expressing recombinant human 2a-adrenoceptors. Ki values are from Table 1. Affinity was determined as the ability of drugs to inhibit binding of [3H]RX821002 (adipocyte) or [3H]rauwolscine (recombinant 2a-adrenoceptor) to membrane homogenates. Correlation coefficient (r2) and slope of regression line determined using either all data points or with norepinephrine (NE) omitted (data in parentheses).

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Correlation of binding affinity of drugs for 2-adrenoceptors in rabbit adipocyte with affinity for recombinant rat 2d-adrenoceptors. Affinity was determined as the ability of drugs to inhibit [3H]RX821002 binding to membrane homogenates. Ki values from Table 1. Omission of norepinephrine (NE) did not significantly influence correlation parameters (data in parentheses).

View this table: [in this window] [in a new window]   TABLE 2 Correlation of affinity of -adrenoceptor antagonists in rabbit platelet and adipocyte with their affinity at 2-adrenoceptors in other tissues or cells Data with norepinephrine, where available, are included in the correlation.

Many of the antagonists showed similar affinity for the _2a- and _2d-adrenoceptors, with potency ratios of 3-fold or less. The yohimbane alkaloids, yohimbine and rauwolscine, as well as the structurally related raubasine, showed >30-fold lower affinity for the _2d-adrenoceptor subtype. WB 4101, consistent with functional data (Trendelenburg et al., 1996), also showed selectivity for the _2a-adrenoceptor subtype.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

It is evident from the data in Table 1 that the pharmacologic characteristics of the ₂-adrenoceptors of rabbit adipocyte and rabbit platelet are identical The platelet was chosen as a comparison tissue for the adipocyte since the human platelet is a well characterized _2A-adrenoceptor system (Bylund et al., 1995) and if the rabbit adipocyte did indeed have _2D-adrenoceptor characteristics, a pharmacological difference between adipocyte and platelet might be expected.

This correlation between pK_i values in rabbit adipocyte with pK_i values in recombinant human _2a-adrenoceptors improves (r² = 0.89) if the values for norepinephrine are excluded. It is known that the ability of full agonists, such as norepinephrine, to inhibit the binding of an antagonist ligand, such as rauwolscine, to the recombinant _2a-adrenoceptor does not reflect its true affinity for the receptor, as measured either directly by the K_D for [³H]norepinephrine or by displacement of an agonist radioligand such as [³H]clonidine (Hieble et al., 1996). Similar results are observed at all three recombinant rat ₂-adrenoceptor subtypes (Tunstall et al., 1996) as well as in human platelets, where [³H]epinephrine binds with high affinity) (K_D = 2.5 nM; Garcia-Sevilla and Fuster, 1986), will produce potent inhibition of an agonist ligand such as [¹²⁵I]-para-iodoclonidine (K_i = 4.6 nM; Gerhardt et al., 1990) but requires much higher concentrations (K_i = 3.8 µM) to inhibit the binding of an antagonist radioligand such as [³H]yohimbine (Hui et al., 1991). If our data with norepinephrine in the adipocyte is dissociated into high and low affinity components, the high affinity K_i value (16 nM) is very similar to its K_D value for binding to the _2a-adrenoceptor (13 nM) and the low affinity (3.4 µM) is in the range observed for inhibition of antagonist radioligand binding to recombinant and native ₂-adrenoceptors. This dependence of K_i on ligand efficacy is not observed for partial agonists such as clonidine, which produce potent inhibition of the binding of either agonist or antagonist radioligands to the recombinant _2a-adrenoceptor (K_i = 10-26 nM; Hieble et al., 1996).

It has been suggested that this phenomenon results from a contribution of high and low affinity states (Tunstall et al., 1996) with agonist ligands labeling the high affinity state (Gerhardt et al., 1990). Consistent with this hypothesis, the addition of a guanine nucleotide analog produced a 5-fold reduction in the ability of norepinephrine to inhibit binding of [³H]RX821002 to the human _2a-adrenoceptor (Halme et al., 1995). While we observe an apparently monophasic, low affinity inhibition curve for norepinephrine against [³H]rauwolscine in membranes expressing the recombinant _2a-adrenoceptor, other investigators have observed high affinity inhibition, or biphasic inhibition, in a similar assay (Jansson et al., 1994; O'Rourke et al., 1994a; Halme et al., 1995). Since the K_i for norepinephrine against an antagonist radioligand seems highly sensitive to experimental conditions, in contrast to those for -adrenoceptor antagonists, it seems reasonable to assume that the differences in K_i for norepinephrine in rabbit adipocyte membranes and membranes expressing the _2a-adrenoceptor reflect differences in receptor environment, which may influence the contribution of high and low affinity states, rather than actual differences in pharmacologic characteristics of the receptors in the two preparations.

Table 2 shows the correlation between drug affinities in rabbit adipocyte and platelet with affinity in a variety of _2A- and _2D-adrenoceptor receptor models, including recombinant receptors, radioligand binding assays using native receptors and functional assays. The best correlation is consistently observed with tissues or receptors having _2A-adrenoceptor pharmacology, whether obtained or cloned from rabbit, human, or pig. Our adipocyte K_i values correlate extremely well (r² = 0.98, slope = 0.83) with those determined by Langin et al. (1990) in this tissue. Correlation parameters between our platelet and adipocyte data with the recombinant human _2a-adrenoceptors are identical, regardless of the radioligand (rauwolscine or RX 821002) used to label the recombinant receptor.

Correlation of drug affinities with receptors or tissues from rat or bovine sources having _2D-adrenoceptor pharmacology is poorer. In contrast, correlation with functional antagonist potency at prejunctional ₂-adrenoceptors in guinea pig atrium is in the range expected for the _2A-adrenoceptor subtype. Comparison of antagonist potency in this model with recombinant ₂-adrenoceptors showed the correlation to be almost as good with the human _2a-adrenoceptor (r = 0.82) as with the rat _2d-adrenoceptor (r = 0.90) (Trendelenburg et al., 1995).

The slope of the linear regression between platelet/adipocyte ₂-adrenoceptors was highest for comparison with _2A-adrenoceptor models (Table 2). However, this slope was generally between 0.6 and 0.7, rather than unity. These results are quite consistent between adipocyte and platelet, and between these tissues and a variety of binding and functional assays. The low slope of the linear regression is primarily a consequence of our finding prazosin and ARC 239 to be inactive (K_i > 40,000) on the platelet and adipocyte ₂-adrenoceptors, rather than the weak activity (K_i = 300-3000 nM) found by us and others at recombinant _2a- and _2d-adrenoceptors. Interestingly, Langin et al. (1990) also could not detect inhibition by prazosin of [³H]RX821002 binding to rabbit adipocytes.

Based on the lack of interaction of prazosin and ARC 239 with the ₂-adrenoceptors of either rabbit platelet or adipocyte, it is clear that _2B- or _2C-adrenoceptors are not present in either of these tissues. This is confirmed by lack of correlation of pK_i values in the rabbit tissues with affinity for recombinant _2b- or _2c-adrenoceptors for the 10 compounds for which affinity data at these subtypes is available (Table 2).

Although we confirm the results of Langin et al. (1990) showing that rauwolscine and yohimbine have relatively low affinity for the ₂-adrenoceptor of rabbit adipocytes, correlation of affinities for a diverse series of antagonists shows a better correlation with the recombinant human _2a-adrenoceptor, than with the rat _2d-adrenoceptor. The same pattern is shown with rabbit ₂-adrenoceptors from another tissue source, the platelet. Hence, all data with rabbit ₂-adrenoceptors, both from functional and radioligand binding data, point toward the rabbit as having _2A- rather than _2D-adrenoceptors, and supports the premise that the _2a- and _2d-adrenoceptors are species orthologs (O'Rourke et al., 1994b; Bylund et al., 1995), with the rabbit falling into the _2a-adrenoceptor group, along with man and pig. Although complete sequences are not available for any of the rabbit ₂-adrenoceptors, a partial sequence of the rabbit _2a-adrenoceptor has recently been reported (Molderings et al., 2000). This sequence, which includes most of the fifth transmembrane domain, shows the rabbit to have a cysteine at position 201, as observed in the human and porcine receptors, known to have _2A-adrenoceptor pharmacology, and differing from the rat and mouse, which have a serine at this position.

Data in both adipocyte and platelet would indicate that the affinity of the rabbit _2A-adrenoceptor for yohimbine and rauwolscine is lower than in human tissues containing this receptor subtype (e.g., HT-29 cells or human platelets) or for recombinant human or porcine _2a-adrenoceptors. It is certainly possible that, although the rabbit may be in the _2A-adrenoceptor group, its pharmacology still shows measurable differences from the human _2A-adrenoceptor. The amino acid sequences of the human _2a-adrenoceptor and rat _2d-adrenoceptor differ at 60 of 450 residues. Many of these differences (44) are in the third intracellular loop, and there are only eight differences contained in the seven putative membrane spanning regions, where agonists and antagonists are postulated to bind to the receptor. It has been postulated that most of the pharmacological differences between the _2a- and _2d-adrenoceptors can be attributed to heterogeneity at position 201 in transmembrane helix 5. Mutation of the Ser²⁰¹ in the mouse _2d-adrenoceptor to cysteine produces a 5-fold increase in the affinity for yohimbine (Link et al., 1992). However, subsequent studies with this mutant receptor have shown that the affinities for several other antagonists, including rauwolscine and WB 4101 remain as low as observed in the wild-type mouse _2d-adrenoceptor (Blaxall et al., 1994). This shows that other differences between human and mouse receptors contribute to the different pharmacology. Minor species differences in amino acid sequence of the _2a/_2d-adrenoceptor are observed, even between species that show similar pharmacology. For example, the human and pig _2a-adrenoceptor differ at 26 residues (four in or near transmembrane helices) and rat and mouse _2d-adrenoceptors differ at 14 residues (one in a transmembrane helix). Data from the partial rabbit sequence available (Molderings et al., 2000) would suggest that differences between the rabbit and human receptor are at least as great as between human and rodent receptors. Hence, it is possible that any of the amino acid substitutions may contribute to subtle differences in the pharmacology of these receptors.