Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

Alpha-2 adrenergic receptors (A2AR) are one of the three major classes of receptors for norepinephrine and epinephrine (Bylund, 1988; Bylund et al., 1994). They are widely distributed in the body and play important roles in a variety of physiological and pathological processes, including regulation of blood pressure, nociception, locomotion, and processing of stressful stimuli (Ruffolo et al., 1993, 1995). A2AR have been used to ameliorate withdrawal symptoms from opiates and alcohol, as anesthetic adjuvants in surgery, and may be of some benefit in treating cognitive deficits in the elderly (Ruffolo et al., 1995). These actions and uses point to the importance of understanding the role of A2AR in central nervous system (CNS) function.

A2AR are members of the G protein coupled receptor superfamily and appear to interact primarily with G_i/o proteins (Chabre et al., 1994; Limbird et al., 1995). When A2AR are stimulated, GDP is released from the G_i/o protein complex, allowing GTP to bind in its place. This leads to the dissociation of the  from the subunits of the heterotrimeric G protein complex and the subsequent regulation of signal transduction systems within the cell. This activation of G proteins is one of the functional consequences of stimulation of A2AR and, as such, provides a measure of the functional activity of these receptors in tissue.

Relatively little is known about the development of central A2AR in general, and even less is known about their function during the developmental period. We examined A2AR functional activity during development using the [³⁵S]GTPS binding assay (Hilf et al., 1989; Sim et al., 1995). This approach has been used in the study of several receptors, including A2AR expressed in cultured cells (Tian et al., 1994; Gillison et al., 1997; Wise et al., 1997). We report here the functional linkage of A2AR to G proteins at birth and the subsequent alterations in the magnitude of this coupling during the first postnatal month. Increases in A2AR-agonist induced GTP-binding in forebrain generally parallel the increase in A2AR levels, but there are interesting discrepancies from this. In cerebellum and brainstem, the highest levels of both receptors and functional receptor-G protein coupling are present at birth and decline in parallel to relatively low levels postnatally.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. [³⁵S]GTPS (1000-1500 Ci/mmol) was obtained from New England Research Products (Boston, MA) and [³H]RX 821002 (53 Ci/mmol) from Amersham Life Science (Arlington Heights, IL). Atropine sulfate, carbamyl choline (carbachol), epinephrine bitartrate, dithiothreitol, GTPS, and baclofen were purchased from Sigma Chemical Company (St. Louis, MO). UK 14,304 was a gift from Allergan Pharmaceuticals (Irvine, CA). Rauwolscine HCl and RX 821002 HCl were purchased from Research Biochemicals Inc. (Natick, MA). Methadone was purchased from Mallinckrodt (St. Louis, MO) and GDP was purchased from United States Biochemical Corp. (Cleveland, OH). All other chemicals were research grade.

Animals. Adult female Sprague-Dawley rats, 185 to 250 g, (SASCO, Kingston, NY) were housed three to four per cage and fed ad libitum. Rat pups were bred in our colony. Litters were culled to nine pups and monitored for normal growth by body weight (Happe and Murrin, 1990). Brains were collected at P0 (day of birth), P7, P14, P21, and P28. All animal use procedures were in strict accordance with The National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the local Animal Care Committee.

Membrane Preparation. Adult female rats and rat pups at the designated ages were sacrificed by decapitation under halothane anesthesia, and brains were removed to ice and dissected into two regions (Fig. 1) designated forebrain and hindbrain (containing the cerebellum and brainstem). Tissue was homogenized in 20 volumes of ice-cold homogenization buffer (50 mM Tris-HCl, 3 mM MgCl₂, and 1 mM EGTA, pH 7.4) and centrifuged at 48,000g at 4°C for 10 min. The pellet was resuspended in 20 volumes of buffer and centrifuged at 48,000g for 10 min. The final pellet was resuspended in 20 volumes of incubation buffer (50 mM Tris-HCl, 3 mM MgCl₂, 1 mM EGTA, and 100 mM NaCl, pH 7.4). For A2AR agonist-stimulated [³⁵S]GTPS binding, we found that fresh membrane preparations were necessary to achieve detectable and consistent results and therefore all [³⁵S]GTPS-binding assays reported here used fresh tissue. Aliquots of resuspended membranes were stored at 80°C for receptor binding and protein assays. Protein levels were determined by the bicinchoninic acid method (Pierce Protein Detection System; Pierce, Rockford, IL; Smith et al., 1985).

View larger version (89K): [in this window] [in a new window]   Fig. 1.   Autoradiograms of A2AR in sagittal sections. [3H]RX 821002 was used as radioligand to label A2AR in P5 (A) and adult rat brain (B). Note that early in postnatal development there is a higher density of A2AR in the brainstem (bs) and cerebellum (Cb) than in adult. In the forebrain, A2AR are found in white matter, e.g., corpus callosum (cc), and anterior commissure (ac) only during early postnatal development. In other brain regions such as the cortex (ctx), septum (sep), amygdala (amy), and olfactory bulb (OB), A2ARs increase in density with increasing age. The dashed lines indicate the approximate cuts used to dissect brains for forebrain and hindbrain membrane preparations.

Agonist-Stimulated [³⁵S]GTPS Binding Assay. The assay was performed according to Sim and colleagues (Sim et al., 1995) with minor modifications. Briefly, membrane preparations equivalent to 1 mg of wet weight (25-50 µg membrane protein) were incubated at 30°C for 1 h in incubation buffer (50 mM Tris-HCl, 3 mM MgCl₂, 1 mM EGTA, 100 mM NaCl, 2 µM GDP, and 1 mM dithiothreitol, pH 7.4) containing 0.05 nM [³⁵S]GTPS, in a volume of 1 ml. Nonspecific [³⁵S]GTPS binding was determined in the presence of 10 µM unlabeled GTPS. Agonists were used at a final concentration of 10 µM to determine receptor-stimulated [³⁵S]GTPS binding. UK 14,304 was used as agonist for A2AR, baclofen for GABA-B receptors, methadone for mu opiate receptors and carbachol for muscarinic acetylcholine receptors. Basal binding was determined in the absence of agonist. In fresh tissue, agonist-stimulated [³⁵S]GTPS binding increases linearly for at least 90 min (data not shown). The reaction was stopped by rapid filtration (Brandel Cell Harvester; Biomedical Research and Development, Gaithersburg, MD) through Whatman GF/B glass fiber filters (Whatman, Clifton, NJ), followed by three washes with 5 ml of ice-cold 50 mM Tris-HCl, pH 7.4. Filter disks were extracted overnight in Econo-Safe liquid scintillation cocktail (Research Products International Corp., Mount Prospect, IL), and bound radioactivity was determined by liquid scintillation spectrophotometry at 95% efficiency.

[³H]RX 821002 Binding Assay. A2AR levels were measured in membrane preparations using [³H]RX 821002 as ligand as described (O'Rourke et al., 1994) with minor modifications. Briefly, membrane preparations containing 100 µg of protein were incubated with 2.5 nM [³H]RX 821002 in 0.3 mM MgCl₂, 0.1 mM EGTA, 5 mM Tris-HCl (residual from membrane preparation), and 50 mM sodium phosphate, pH 7.4, for 1 h at room temperature. The suspensions were then filtered through GF/B glass fiber filters and washed 3 times with 5 ml of ice-cold 50 mM Tris-HCl, pH 7.4. Radioactivity was determined by liquid scintillation spectrometry in Econo-Safe at 45% efficiency. Nonspecific binding was determined by addition of 10 µM rauwolscine. Binding levels are expressed as femtomoles of [³H]RX 821002 bound per milligram of protein and were converted to B_max values using the formula B_max = (B × (K_d + [L]))/[L], where B is the measured binding, K_d is the affinity constant of [³H]RX 821002 (0.5 nM), and [L] is the [³H]RX 821002 concentration.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

A2AR Development

Comparison of neonatal and adult CNS A2AR, using receptor autoradiography with the antagonist [³H]RX 821002 (Fig. 1), demonstrates major changes in receptor distribution and density during postnatal development. In many brain regions, such as cortex, low levels of A2AR are present early in postnatal development and increase to much higher levels in the adult CNS. On the other hand, several brain regions, such as cerebellum and brainstem, have high levels of A2AR during the early postnatal period that decline to low levels in the adult. Forebrain receptor density undergoes a general increase, whereas hindbrain receptor density generally decreases during postnatal development. We therefore chose to study the forebrain and the hindbrain separately to examine the development of receptors and receptor-G protein coupling.

Membrane Receptor Binding. The developmental change in A2AR density was examined in forebrain and hindbrain membrane preparations throughout postnatal development. Consistent with the autoradiographic studies (Murrin et al., 1996), forebrain A2AR density increases and hindbrain A2AR density decreases with age (Fig. 2). The magnitude of change in forebrain receptor density from P0 to adult is dependent on expression of the data as either femtomole per milligram of protein (4-fold) or femtomole per milligram of wet weight (7-fold), as there is nearly a 2-fold increase from birth to adult in membrane protein content per milligram wet weight (Table 1). In forebrain there is a rapid increase in A2AR density from P0 to P14, when receptor density has essentially attained adult levels (Fig. 2A). In hindbrain there is a decrease in receptor density, falling to near adult levels by P14 (Fig. 2B).

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Development of A2AR in forebrain and hindbrain of rats. The density of A2AR was measured by [3H]RX 821002 binding in forebrain (A) and hindbrain (B), as described in Experimental Procedures. Nonspecific binding was determined with 10 µM rauwolscine. Binding levels were converted to Bmax as described in Experimental Procedures and are expressed as fmol/mg protein (solid lines) and fmol/mg wet weight (dashed lines), representing the mean ± S.E.M. of 7 to 11 animals at each age.

[³⁵S]GTPS-Binding Assay

To determine whether the appearance of A2AR is indicative of the onset of receptor function, we used the A2AR agonist-stimulated binding of [³⁵S]GTPS to assess the functional coupling between receptors and G proteins. Agonist-stimulated binding of [³⁵S]GTPS has been used previously to demonstrate functional coupling of receptors to G proteins in both membrane preparations and in slide-mounted tissue sections (Sim et al., 1995; Traynor and Nahorski, 1995; Gillison et al., 1997). We examined this approach as a potential functional assay for A2AR in both adult and neonatal rat brain.

In our initial studies, we found that storage of tissue or membrane preparations at 80°C led to low and variable levels of A2AR agonist-stimulated [³⁵S]GTPS binding. It has previously been reported that receptor coupling to G proteins, as determined by the exchange of bound GDP for [³⁵S]GTPS, can be altered by multiple freeze-thaw cycles of membrane preparations (Sim et al., 1995). It appears that the A2AR agonist-stimulated increase in [³⁵S]GTPS binding is quite sensitive to freezing, as binding in preparations from previously frozen tissue was considerably reduced relative to fresh tissue. On the other hand, stimulation of binding by agonists for other receptors (GABA-B, muscarinic acetylcholine, and opiate receptors) was only slightly reduced in frozen tissue compared with fresh tissue (data not shown). Based on these results, fresh tissue was used throughout the present study.

The [³⁵S]GTPS binding in membrane preparations from neonatal and adult rat forebrain is presented in Fig. 3. Basal [³⁵S]GTPS binding is inhibited more than 90% by the addition of excess unlabeled GTPS (1 µM). In all experiments agonists were used at 10 µM, which in preliminary studies gave maximum stimulation for each agonist. The A2AR agonist, UK 14,304, stimulates [³⁵S]GTPS binding 24% over basal in P5 forebrain and 18% over basal in adults (Fig. 3). The UK 14,304-stimulated binding is blocked by the A2AR antagonist, RX 821002, which alone has no effect. We also found that the agonist epinephrine (10 µM) was as effective as UK 14,304 in stimulating [³⁵S]GTPS binding (data not shown). Rauwolscine (10 µM), an A2AR antagonist without an imidazoline structure, was as effective as RX 821002 (10 µM) in blocking the stimulation produced by either agonist (data not shown). Binding characteristics in membrane preparations from P5 animals were qualitatively the same as found in adults, but the basal and agonist-stimulated [³⁵S]GTPS-binding levels were lower. Stimulation of forebrain muscarinic cholinergic receptors with carbachol increases [³⁵S]GTPS binding 26% in P5 animals and 22% in adults, an effect blocked by atropine (Fig. 3). As expected, both UK 14,304- and carbachol-stimulated [³⁵S]GTPS binding are dependent on Mg⁺⁺ (Sim et al., 1995).

View larger version (22K): [in this window] [in a new window]   Fig. 3.   Agonist-stimulated [35S]GTPS binding in rat forebrain membranes. [35S]GTPS binding for P5 (A) and adult (B) are presented as percentage of increase over basal binding ± S.E.M. determined in triplicate for three animals at each age. UK 14,304 (UK), RX 821002 (RX), carbachol (Car), and atropine (Atr) were added at 10 µM before initiation of the assay by addition of 50 µg of membrane protein. Basal [35S]GTPS binding was 1627 ± 175 cpm for P5 animals and 4989 ± 103 cpm for adults. Nonspecific [35S]GTPS binding (1 µM unlabeled GTPS) was 366 ± 111 cpm for P5 animals and 356 ± 67 cpm for adults. For studies without Mg++ (Mg), 3 mM CaCl2 was substituted. In the absence of Mg++, basal binding was 755 ± 6 and 1166 ± 50 cpm for P5 and adult preparations, respectively.

The affinity of GTPS binding to G proteins was determined in forebrain from P0, P21 and adult animals, and in hindbrain from P0 animals by isotopic dilution of [³⁵S]GTPS with unlabeled GTPS (10¹⁰ to 10⁶ M; Fig 4). A one-site model was used throughout because a two-site model did not significantly improve the fit. No statistically significant differences in binding affinity are found among P0, P21, and adult forebrain and P0 hindbrain preparations for basal binding nor for A2AR agonist-stimulated GTPS binding (Table 2). The A2AR agonist-stimulated GTPS binding to G proteins has a higher affinity (K_d = 4.5 nM) than basal GTPS binding (K_d = 21 nM) at all ages and in both forebrain and hindbrain.

View larger version (28K): [in this window] [in a new window]   Fig. 4.   [35S]GTPS binding in forebrain and hindbrain membranes during development. Binding was measured at a single concentration of [35S]GTPS in the presence of increasing concentrations of unlabeled GTPS. Values are expressed as a percentage of control binding and were normalized to [35S]GTPS binding in the absence of unlabeled GTPS. Data are the mean of three to four separate experiments for P0 hindbrain (A), P0 forebrain (B), P21 forebrain (C), and adult forebrain (D). Basal binding (solid lines) and UK 14,304-stimulated binding (dashed lines) are shown in each panel. All binding data are best fit by a one-site analysis. No statistically significant difference in binding affinity was found between ages or regions for basal binding nor for A2AR-stimulated GTPS binding. GTPS affinities in basal binding and in UK 14,304-stimulated binding were significantly different, p <0.05 (see Table 2).

View this table: [in this window] [in a new window]   TABLE 2 GTPS binding in forebrain and hindbrain during development For specific UK 14,304-stimulated GTPS binding, basal binding (in the absence of agonist) was subtracted from binding in the presence of 10 µM UK 14,304. Results are mean ± S.E.M. for three to four experiments. Data were analyzed as described in Experimental Procedures. No significant differences were found among the Kd values within groups (basal or UK 14,304 stimulated). Kd values for agonist-stimulated [35S]GTPS binding were all significantly different from their corresponding basal controls. In subsequent studies, the composite Kd values, the mean of all determinations in each group, were used to determine Bmax values for each group (see Experimental Procedures).

Postnatal Development of [³⁵S]GTPS Binding

Basal [³⁵S]GTPS Binding. Basal [³⁵S]GTPS binding represents a large part of the total signal when examining agonist-stimulated binding at all ages. During postnatal development, basal [³⁵S]GTPS binding increases more than 3-fold in forebrain preparations and less than 1.5-fold in hindbrain preparations (Fig. 5). In the forebrain, basal [³⁵S]GTPS binding increases dramatically between P7 and P14, reaching adult levels at P14. On the other hand, hindbrain basal [³⁵S]GTPS binding remains low through P14 and then increases to adult levels by P21 (Fig. 5).

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Basal [35S]GTPS binding during postnatal development. Basal [35S]GTPS binding was determined in forebrain and hindbrain membrane preparations from 7 to 11 animals at each age. Data are mean Bmax ± S.E.M. in picomoles per milligram of membrane protein. Bmax was determined as described in Experimental Procedures.

A2AR-Stimulated [³⁵S]GTPS Binding. In forebrain membranes, A2AR-stimulated [³⁵S]GTPS binding increases nearly 4-fold from P0 to adult (Fig. 6A). The level of A2AR-stimulated [³⁵S]GTPS binding is low but detectable at P0 and P7. Between P7 and P14, agonist-stimulated binding increases dramatically, reaching peak levels at P14 to P21, and then declines slightly to adult levels (Fig. 6A). The developmental pattern for A2AR-stimulated [³⁵S]GTPS binding is similar to the increase in A2AR determined by ligand binding and also parallels the increase in basal [³⁵S]GTPS binding. Comparison of Figs. 2A and 6A shows that the rise in receptor density precedes somewhat the increase in A2AR-stimulated [³⁵S]GTPS binding. This is particularly evident at P7 and indicates that receptor-G protein coupling lags receptor appearance early in postnatal development. The overall magnitude of developmental increases in A2AR density and levels of A2AR-stimulated [³⁵S]GTPS binding are similar, about 4-fold. The A2AR-stimulated [³⁵S]GTPS binding, expressed as percentage of basal binding, remains relatively constant throughout the postnatal period (Fig. 6B).

View larger version (28K): [in this window] [in a new window]   Fig. 6.   Alpha-2 adrenergic agonist-stimulated [35S]GTPS binding during postnatal development. Specific A2AR-stimulated binding (10 µM UK 14,304) was analyzed by subtracting basal binding (no agonist). Agonist-stimulated [35S]GTPS binding was determined in forebrain (A and B) and hindbrain (C and D) for 7 to 11 animals at each age. Data are mean ± S.E.M. of Bmax in picomoles per milligram of membrane protein determined as described in Experimental Procedures (A and C). Data expressed as percentage of increase over basal binding are calculated from untransformed data and are mean ± S.E.M. (B and D).

[³⁵S]GTPS Binding Stimulated by Baclofen, Carbachol, and Methadone. For comparison with A2AR, we examined the development of functional receptor coupling to G proteins for the GABA-B, mu opiate, and muscarinic cholinergic receptor systems, using the agonists baclofen, methadone, and carbachol, respectively. In forebrain, agonist-stimulated [³⁵S]GTPS binding is similar at P0 for all three agonists (Fig. 7A). Baclofen-stimulated [³⁵S]GTPS binding increases over 6-fold by P21, whereas methadone- and carbachol-stimulated binding each increase about 4-fold. In forebrain, the developmental pattern of agonist-stimulated binding of [³⁵S]GTPS is remarkably similar for all three neurotransmitter systems (Fig 7A). There is little or no change in agonist-stimulated [³⁵S]GTPS binding between P0 and P7, but between P7 and P14 there is a dramatic increase for all three systems. After P14, binding levels increase slightly or stay the same through P21 and then decrease to adult levels. The developmental pattern of baclofen-, methadone-, and carbachol-stimulated [³⁵S]GTPS binding parallels both the increase in basal [³⁵S]GTPS binding (Fig. 5) and the increase in A2AR-stimulated [³⁵S]GTPS binding in the forebrain (Fig. 6A).

View larger version (20K): [in this window] [in a new window]   Fig. 7.   GABA-B, mu opiate, and muscarinic receptor-stimulated [35S]GTPS binding during postnatal development. Specific agonist-stimulated binding was determined by subtracting basal binding from [35S]GTPS binding in the presence of 10 µM baclofen (), methadone (), or carbachol () for GABA-B, opiate, or muscarinic cholinergic receptors, respectively. Agonist-stimulated [35S]GTPS binding was determined in forebrain (A) and hindbrain (B) membrane preparations from 7 to 11 animals at each age. Data are expressed as picomoles per milligram of membrane protein and are mean ± S.E.M. of Bmax values determined as described in Experimental Procedures.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

Stimulation of A2AR reduces adenylyl cyclase activity and alters ion channel activity in the adult CNS. A2AR signal transduction is pertussis toxin sensitive, indicating coupling primarily through G_i/o proteins (Chabre et al., 1994; Limbird et al., 1995). In this report we use agonist-stimulated [³⁵S]GTPS binding to study the coupling of native A2AR to G proteins in the CNS. Agonist-stimulated [³⁵S]GTPS binding has been used to examine receptor-G protein coupling for several neurotransmitter receptor systems (Hilf et al., 1989; Gierschik et al., 1991; Lazareno et al., 1993; Lorenzen et al., 1993; Traynor and Nahorski, 1995; Newman-Tancredi et al., 1997; Waeber and Moskowitz, 1997), including the A2AR system in transfected cells (Tian et al., 1994; Wise et al., 1997), where receptor density is greater than in the CNS, and in RINm5f cells (Gillison et al., 1997). The characteristics of A2AR agonist-stimulated [³⁵S]GTPS binding in the current study are in general agreement with results in cell culture (Tian et al., 1994). This is the first study demonstrating the utility of this method for native A2AR in the CNS.

The present study provides evidence that A2ARs are functionally coupled to G proteins in the rat CNS throughout postnatal development based on the presence of A2AR agonist-stimulated [³⁵S]GTPS binding in membrane preparations. Our results show that norepinephrine and epinephrine can mediate cellular responses through A2AR as early as day of birth, and that developmental changes in A2AR agonist-stimulated [³⁵S]GTPS binding generally parallel changes in A2AR density and distribution.

We used tissue separated into forebrain and hindbrain for functional receptor-G protein coupling studies to allow comparison of brain regions of increasing and decreasing A2AR density during postnatal development. In forebrain, the predominant developmental pattern is a postnatal increase in A2AR density in regions expressing significant receptor levels in the adult. There are also a few regions in the forebrain that transiently express relatively high levels of A2AR at birth, but have few A2ARs in the adult. Functional coupling of A2AR to G proteins parallels the increase in receptor number determined by ligand-binding studies. In contrast to forebrain, transient expression of A2AR is a predominant developmental pattern in the hindbrain. The greatest receptor density is found at birth, followed by a rapid loss of receptors early in the postnatal period. Based on the parallel decline of A2AR agonist-stimulated [³⁵S]GTPS binding and A2AR binding in hindbrain, it appears that receptors expressed early in development, even in regions expressing few receptors in the adult, are nonetheless functionally coupled to G proteins. These results demonstrate that transiently expressed A2AR are capable of mediating norepinephrine- and epinephrine-stimulated effects during early postnatal development and suggest that these receptors may play a role during development distinct from that found in adults.

The postnatal increase in A2AR density in forebrain regions likely represents the maturation of the receptor system for noradrenergic function in the adult CNS. The coupling of A2AR to G proteins, as measured by [³⁵S]GTPS binding, generally increases in parallel with receptor number, although there is a delay in the increase of agonist-stimulated [³⁵S]GTPS binding compared with the increase in receptor number. This is most apparent at P7, when receptor density has doubled from P0 values (Fig. 2A) whereas agonist-stimulated [³⁵S]GTPS binding has increased only 25% (Fig. 6A). One interpretation of this pattern is that there is a delay in regulation of receptor-G protein coupling. The developmental changes in GTP-mediated shifts in the affinity state for A2AR agonist binding reported in rat cerebral cortex (Nomura et al., 1984) also suggest that regulation of A2AR-G protein-coupling changes during postnatal development.

The temporal development of A2AR-stimulated [³⁵S]GTPS binding in forebrain is remarkably similar to the pattern of basal [³⁵S]GTPS binding. Especially notable is the rapid increase after P7. In addition, the forebrain development of G protein coupling to three other neurotransmitter receptors, GABA-B, opiate, and muscarinic cholinergic, also increases after P7. As opposed to basal and agonist-stimulated [³⁵S]GTPS binding, mu opiate, and muscarinic cholinergic receptor density increases extensively between P0 and P7 (Clendeninn et al., 1976; Coyle and Pert, 1976; Kuhar et al., 1980; Miyoshi et al., 1987). To our knowledge, GABA-B receptor binding has not been examined during development.

In the hindbrain the decrease in A2AR-G protein coupling contrasts with the G protein coupling to the three other neurotransmitter systems examined, all of which increase during postnatal development. The decline in A2AR density in hindbrain membranes, more than 3-fold, is greater than the 2-fold loss of agonist-stimulated [³⁵S]GTPS binding; this is consistent with the presence of spare receptors in the neonatal hindbrain. The difference could also be due to developmental changes in the expression of different A2AR subtypes and/or G protein subtypes in hindbrain. It remains to be determined whether A2AR loss is due to a loss of cells or whether receptor expression is turned off in cells that remain into adulthood.

The relationship of basal [³⁵S]GTPS binding to G protein levels is not clear. Basal binding likely represents [³⁵S]GTPS binding to a large number of different GTP-binding proteins. Nonetheless, pertussis toxin-sensitive G_i/o proteins comprise as much as 1% of total membrane protein in the CNS and represent the predominant subtype of GTP-binding proteins, at least in the adult brain (Sternweis and Robishaw, 1984; Harmouch et al., 1997). Given the quantity of G_i/o and the known favorable binding of [³⁵S]GTPS to G_i/o over Gs, it is likely that G_i/o is the major site of basal [³⁵S]GTPS binding. Basal [³⁵S]GTPS binding increases later in hindbrain than forebrain, as does binding stimulated by GABA-B, opiate, and muscarinic cholinergic agonists. The correspondence between basal [³⁵S]GTPS binding and agonist-stimulated [³⁵S]GTPS binding in both forebrain and in hindbrain suggests that the expression of G proteins may be rate-limiting for development of functional receptor-G protein coupling.

Although little is known about A2AR development or their role in regulating development, a transient perinatal expression of A2AR in some brain regions and a rapid postnatal increase in A2AR in other brain regions has been demonstrated by receptor autoradiography (Murrin et al., 1996). Similar developmental patterns in mRNA are seen using in situ hybridization (Wang and Limbird, 1997; Winzer-Serhan and Leslie, 1997; Winzer-Serhan et al., 1997a,b). Our results are in general agreement with previous studies (Murrin et al., 1996; Winzer-Serhan and Leslie, 1997; Winzer-Serhan et al., 1997a,b) indicating both postnatal increases and perinatal transient expression of A2AR in different brain regions. Our studies, however, do not distinguish different A2AR subtypes or small brain regions because of technical limitations of the assay. The transient expression of receptors in cerebellum and brainstem may indicate a specific role in development. Recently it has been suggested that embryonic A2AR expression is associated with apoptotic events, based on the anatomic and temporal correlation between A2AR mRNA expression and markers of apoptosis (Wang and Limbird, 1997). Likewise, based on the temporal and regional pattern of A2AR mRNA expression, it has been suggested that the perinatal increase in receptor density may serve specific roles in development, including neuronal migration, maturation of neurons, and mediation of sensory functions (Winzer-Serhan and Leslie, 1997; Winzer-Serhan et al., 1997a,b). All of these hypotheses require further study.

Receptor-coupling efficiency reflects the ability of a receptor to stimulate [³⁵S]GTPS binding to multiple G proteins upon agonist stimulation (Tian et al., 1994), and is estimated from the B_max of agonist-stimulated [³⁵S]GTPS binding and the B_max of ligand-binding sites. Based on our data, A2AR coupling efficiency is 6 to 10 mol of [³⁵S]GTPS bound/mol A2AR/h and does not change significantly with developmental age or between forebrain and hindbrain. The coupling efficiency of A2AR in our study is similar to results in PC-12 cells (Tian et al., 1994, 1996) but lower than studies in COS-7 cells (Wise et al., 1997) and RINm5F cells (Gillison et al., 1997). The discrepancies may be due, in part, to the expression of different levels and ratios of A2AR and G proteins in cultured and/or transfected cells as opposed to native tissue or to differences in G proteins expressed. Our coupling efficiency is also similar to that reported for several neurotransmitters systems (Hilf et al., 1989; Lazareno et al., 1993; Lorenzen et al., 1993; Traynor and Nahorski, 1995; Newman-Tancredi et al., 1997; Sim et al., 1997a) and lower than reported for some more robust systems (Gierschik et al., 1991; Tian et al., 1996; Gillison et al., 1997; Sim et al., 1997a).

Recently, agonist-stimulated [³⁵S]GTPS binding has been adapted for autoradiographic use on slide-mounted tissue sections (Sim et al., 1995, 1997a) and applied to several neurotransmitter receptor systems (Sim et al., 1996a,b, 1997b; Sim and Childers, 1997; Waeber and Moskowitz, 1997). Autoradiography provides more detailed regional analysis of agonist-stimulated [³⁵S]GTPS binding than is possible with membrane preparations; however, we were unable to detect significant A2AR agonist stimulation of [³⁵S]GTPS binding over basal levels in adult brain sections (data not shown). In contrast, with the agonists baclofen, methadone, and carbachol we obtained autoradiographic results (data not shown) similar to previously published studies for GABA-B, mu opioid, and muscarinic receptors (Sim et al., 1995; Waeber and Moskowitz, 1997). Receptor density and coupling efficiency can limit the detection of [³⁵S]GTPS binding by autoradiography (Sim et al., 1997a). In this regard, A2ARs have similar densities to mu opioid (Clendeninn et al., 1976; Coyle and Pert, 1976) and muscarinic (Kuhar et al., 1980; Miyoshi et al., 1987) receptors in the CNS, and A2AR and muscarinic receptors have similar catalytic efficiencies. Therefore, neither of these factors provides a clear explanation for the inability to detect A2AR-stimulated [³⁵S]GTPS binding autoradiographically.

The specific heterotrimeric G protein coupled to a receptor may affect autoradiographic determination of [³⁵S]GTPS binding. For example, receptor coupling to G_s has not been detected by this method, probably due to the very slow dissociation of GDP from G_s (Sim et al., 1997a; Waeber and Moskowitz, 1997). Among receptors coupled to G_i/o proteins, serotonin 5-hydroxytryptamine_1A and 5-hydroxytryptamine_1B receptor-stimulated [³⁵S]GTPS binding can be demonstrated, but 5-hydroxytryptamine_IF receptor-stimulated [³⁵S]GTPS binding can not (Waeber and Moskowitz, 1997). The difference may be due to receptor subtype coupling to particular G_i/o subtypes, with some pairings more amenable to autoradiographic analysis than others (Waeber and Moskowitz, 1997). This requires further study. In membrane preparations we found that A2AR-stimulated [³⁵S]GTPS binding was very sensitive to freezing, to a much greater extent than found with other neurotransmitter receptor systems (data not shown). The use of frozen tissue sections for autoradiography also may contribute to the lack of success in obtaining A2AR-stimulated autoradiographic signal in the [³⁵S]GTPS-binding assay.

In summary, our data suggest that there may be important differences in the role of A2AR between the early postnatal period and adult rat brain. We have used agonist-stimulated [³⁵S]GTPS binding to examine A2AR function in rat brain during postnatal development. Our data show that A2AR are functionally coupled to G proteins throughout postnatal development and, therefore, are able to mediate signal transduction upon stimulation by norepinephrine and epinephrine. Functional A2AR in some brain regions are transiently expressed, suggesting a specific role in brain development. In many brain regions A2ARs are at low levels at birth, acquire adult receptor density by P14, and are functionally coupled to G proteins as soon as expression can be detected. Coupling of GABA-B, mu opiate, and muscarinic receptors to G proteins has also been demonstrated and these functions develop with a time course similar to A2AR coupling to G proteins and to the increase in basal [³⁵S]GTPS binding. These data suggest that a similar mechanism is involved in the development of functional receptor-G protein interactions for several different neurotransmitters and that G protein expression may be a limiting factor in this process.