Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

The objective of the present was to investigate whether quantitative agonist-antagonist pharmacology could be carried out at the single-cell level on ₁-adrenoceptors (ARs). To this end, we used single cells harboring recombinant bovine _1a-ARs, of the same type that had been used to establish radioligand binding (RLB) and biochemical markers of cell signaling at the level of millions of cells (Schwinn et al., 1990, 1991, 1995).

₁-AR subtypes can couple to many different G proteins and effectors either directly or indirectly, depending on the tissue or cell type studied (for a review, see Guarino et al., 1996). The predominant effector in ₁-AR activation is considered to be activation of phosphoinositide-specific phospholipase C-, resulting in the formation of inositol-1,4,5-triphosphate (InsP₃) and diacylglycerol. In turn, either these messengers or other pathways result in an increase in the intracellular Ca²⁺ concentration ([Ca²⁺]_i) by release of either intracellular Ca²⁺ stores or its influx through plasmalemmal membrane channels, which are either voltage-operated Ca²⁺ channels (Ljung and Kjellstedt, 1987; Esbenshade et al., 1993) or receptor-activated Ca²⁺ channels (Han et al., 1992; Barritt, 1999).

Initially, the form of the Ca²⁺ signal transduction pathway was thought to be a discriminating criterion for classification of the ₁-AR subtypes (McGrath, 1982, 1985; Han et al., 1987). However, a large body of subsequent work has shown that this is not the case (McGrath et al., 1989; Wilson et al., 1991; for a review, see Guarino et al., 1996). It is now clear that the particular signaling pathway followed can vary according to the phenotype in which the native receptor is expressed and that fundamentally this depends on the presence and appropriate relationships of the components. For recombinant receptors, this may vary according to the cell system in which it is expressed (Kenakin, 1997).

Our aim in the present study was to use the Ca²⁺-sensitive fluorescent dye Fura-2 acetoxymethyl ester (AM) to define the nature of the [Ca²⁺]_i signal for a recombinant form of the _1a-AR. On achieving a sufficient understanding of this signal, we used [Ca²⁺]_i signals to evaluate functional responses to receptor activation. This allowed us to make a quantitative analysis of agonist and antagonist pharmacology at a recombinant ₁-AR in single cells. Our long-term reason for studying single cells is to establish a background for subsequent analysis of cells dissociated from heterogeneous native tissues. We hypothesized that the results from recombinant receptors would provide a more precise analysis than native systems in which more than one subtype might be present. We chose the _1a-AR because it has been shown, among the three subtypes, to be most effectively coupled to the production of InsP₃ and to be capable of instigating rises in [Ca²⁺]_i (Schwinn et al., 1991; Theroux et al., 1996). Rat-1 fibroblasts (R-1Fs) stably expressing the bovine _1a-AR (originally termed _1c-AR) were used because this system has been used in various comprehensive studies of ligand binding of the recombinant receptor (Schwinn et al. 1990, 1995) and in defining the intracellular second-messenger signaling pathways activated on stimulation with an agonist (Schwinn et al., 1991, 1995).

Microspectrofluorimetry analysis of the Ca²⁺ signaling mechanism activated during agonist stimulation of the bovine _1a-AR indicated that the dominant component of the [Ca²⁺]_i signal was due to rapid, short-lived release of stored Ca²⁺. The transient nature of this functional [Ca²⁺]_i signaling process allowed us to explore the issues and limitations involved in using a concentration-related but nonequilibrium agonist-mediated response to examine agonist-antagonist interactions with a recombinant form of the _1a-AR. The implications for analysis of native systems are discussed.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Cell Culture. R-1Fs stably expressing the bovine _1a-AR (Wise et al., 1995) were grown in monolayers in Dulbecco's modified Eagle's medium (DMEM) supplemented with 5% (v/v) newborn calf serum, 100 I.U./ml penicillin, 100 µg/ml streptomycin, and 1 mM L-glutamine in a 95% air and 5% CO₂ atmosphere at 37°C. Selection was maintained by adding geneticin (400 µg/ml) to the growth media.

RNA Extraction. Total RNA from R-1F culture samples was extracted using the RNAzol B method (Biogenesis Ltd., Bournemouth, UK) and treated with DNase I (0.1 U/µg RNA; Boehringer-Mannheim Biochemica, Mannheim, Germany) for 20 to 25 min at 37°C. The RNA samples were precipitated in ice-cold isopropanol for 15 to 20 min and resuspended in RNase-free water.

Reverse Transcription-Polymerase Chain Reaction (RT-PCR). Total RNA (~500 ng/sample) was reverse transcribed (25-µl reaction volumes) from the gene-specific antisense primer 5'-CCAGGTCCTTGTGCTGT-3' using Moloney murine leukemia virus reverse transcriptase (Stratagene, Cambridge, UK) for 1 h at 40°C. Resulting cDNA/RNA hybrid molecules were PCR amplified using an Omnigene Thermal Cycler (Hybaid Ltd., Middlesex, UK) in 25-µl reaction volumes. The 2-µl aliquots of completed RT products were combined with 2.85 µmol of each primer: sense, 5'-TTCTCCGTGAGACTGCT-3', which anneals to bases 769 to 785, and antisense; 5'-CCAGGTCCTTGTGCTGT-3', which anneals to bases 1107 to 1123 of the cloned bovine _1a-AR cDNA full-length sequence (Schwinn et al., 1990), and 2 U of Taq DNA polymerase (Promega, Southampton, UK) in a standard reaction mixture (total, 25 µl) containing 1× Taq polymerase buffer, 125 µM dideoxynucleoside triphosphates (dATP, dTTP, dGTP, dCTP), 1.5 mM MgCl₂, and 0.01 mg/ml acetylated BSA.

Radioligand Binding Competition Experiments. Estimates of binding affinity (pK_i) for the nonselective ₁-AR antagonist [³H]prazosin and the competitive ₁-AR antagonists prazosin, WB4101, and BMY7378 were made from displacement curves (by using 12 different concentrations of competing ligands in a total volume of 0.5 ml of Tris-HCl assay buffer) in membrane homogenates (0.05 mg/ml) stably expressing the bovine _1a-AR. [³H]Prazosin was used to label the ₁-AR, and nonspecific binding was determined in the presence of phentolamine (10 µM). All equilibrations were carried out for 30 min at 25°C, and the reactions were terminated by the addition of ice-cold Tris buffer. Using a Brandel cell harvester, bound ligand was separated from free ligand by rapid cold vacuum filtration over Whatman GF/C filters (Whatman International Ltd, Maidstone, UK). Concentrations of displacing agent producing 50% displacement of [³H]prazosin (IC₅₀) were interpolated with the use of nonlinear iterative curve-fitting methodologies performed by Prism (GraphPad Software, San Diego, CA) and converted into pK_i with the equation of Cheng and Prusoff (1973).

Measurement of [Ca²⁺]_i. Cells removed from culture flasks using trypsin/EDTA were washed by centrifugation-resuspension in fresh DMEM, and aliquots of this suspension were plated onto glass coverslips. Cells were grown overnight and then loaded (15 min at 37°C), with Fura-2 AM (1 µM). A rise or fall in [Ca²⁺]_i causes a corresponding effect in the Fura-2 fluorescence ratio recorded from cells loaded with this dye, and this allows receptor/voltage-mediated changes in [Ca²⁺]_i to be microspectrofluorimetrically monitored (Grynkiewicz et al., 1985). In the present study, Fura-2 fluorescence ratios (excitation wavelengths, 340 and 380 nm) were recorded at 4-Hz intervals from single cells at room temperature. Data were digitized and recorded directly to computer disk using an interface and associated software (Version 5.2) obtained from Cairn Research Ltd. (Faversham, Kent, UK).

Analysis of Data. Agonist-evoked [Ca²⁺]_i signals were quantified as the difference between the baseline resting ratio level and that attained at the peak response. The effect of antagonists on the CRC to Phe was analyzed by expressing the data as a fraction of the maximal peak response elicited by Phe in control solution alone. Agonist potency was expressed as pD₂ (log EC₅₀) value. Corresponding pD₂ and maximal stimulation (E_max) values are shown in Table 2.

Solutions. The physiological control saline solution contained 130 mM NaCl, 5 mM KCl, 1 mM CaCl₂, 1 mM MgCl₂, 20 mM HEPES, and 10 mM D-glucose, pH adjusted to 7.4 using NaOH. Composition of the high K⁺ depolarizing solution consisted of control saline solution with the following modifications: 5 mM NaCl, 140 mM KCl, and pH was adjusted to 7.4 with KOH. Ca²⁺-free solutions were prepared simply by omitting CaCl₂ from the control solution and adding the Ca²⁺-chelating agent EGTA (0.1 mM): this results in a contamination level of Ca²⁺ of ~0.02 µM (Miller and Smith, 1984).

Materials and Chemicals. Cell culture plastics were supplied by Falcon. DMEM with sodium pyruvate, newborn calf serum, L-glutamine, penicillin, streptomycin, and trypsin/EDTA were purchased from Gibco Life Technologies (Paisley, Scotland). Fura 2-AM, phenylephrine HCl, HEPES, and EGTA were purchased from Sigma (Dorset, UK). WB4101, BMY7378 dihydrochloride, and prazosin HCl were supplied from Research Biochemicals Inc. (Natick, MA). [³H]Prazosin (0.2 nM; specific activity, 76 Ci/mmol) was ordered from Amersham Corp. (Arlington Heights, IL). The (R)-enantiomer of A-61603 was a gift from Dr. Michael Meyer (Abbott Laboratories, Abbott Park, IL). Moloney murine leukemia virus reverse transcriptase was purchased from Stratagene. Acetylated BSA, dideoxynucleoside triphosphates, and Taq DNA polymerase were obtained from Promega.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

RT-PCR

Amplification by RT-PCR of a 354-base pair region of the bovine _1a-AR coding sequence produced a PCR product that resolved as a single band on a 1.5% agarose gel (Fig. 1A, lanes 1 and 3). Nucleotide sequencing confirmed complete identity with nucleotide positions 769 to 1123 of the bovine cDNA clone (Schwinn et al., 1990). No PCR products were detected in the negative control samples, confirming that RT-PCR amplified products originated from reverse-transcribed mRNA rather than from contaminating genomic DNA (Fig. 1A, lanes 2 and 4).

View larger version (42K): [in this window] [in a new window]   Fig. 1.   A, Amplification by RT-PCR of a 354-base pair region of the bovine 1a-AR coding sequence. Lane L, 1-kb DNA ladder; lanes 1 and 3, PCR product from the RT-PCR amplified bovine 1a-AR cDNA; lanes 2 and 4, negative control samples. PCR product was electrophoresed on a 1.5% agarose gel stained with ethidium bromide. Arrowhead shows the band position of the RT-PCR product. B, displacement of [3H]prazosin (0.2 nM) binding to membranes prepared from R-1Fs expressing the bovine 1a-AR by increasing concentrations of WB4101 (, n = 4), BMY7378 (, n = 4), prazosin (, n = 4), Phe (, n = 4), and A-61603 (, n = 4). Assay points were determined in triplicate, and nonspecific binding was determined in the presence of phentolamine (10 µM). The pKi estimates for these compounds are listed in Table 1.

Radioligand Binding Experiments

Binding Affinity Estimates. Estimates of binding affinity (pK_i) for the agonists and competitive ₁-AR antagonists tested in the functional study are shown in Table 1. All of the agonist-antagonist displacement curves were monophasic. Test antagonists exhibited Hill coefficients that were close to unity. The agonists displayed values significantly lower than unity (Fig. 1B, Table 1) which could be due to negative cooperativity or complex two-stage binding processes, which are a common occurrence for complex ligands such as agonists. Calculated pK_i values for Phe and (R)-A-61603 were 5.69 and 7.47, respectively, so by this measure, A-61603 has 60-fold more affinity than Phe.

Ca²⁺ Signaling Mechanism. Depolarization. KCl (140 mM) failed to evoke any change in [Ca²⁺]_i (Fig. 2A). The involvement of voltage-operated Ca²⁺ channels in the regulation of [Ca²⁺]_i was therefore excluded.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Representative traces showing changes in [Ca2+]i in response to KCl and Phe in R-1Fs stably expressing the bovine 1a-AR. Experiments were replicated at least four times with similar results. A, change in [Ca2+]i elicited on exposure to a depolarizing concentration (140 mM) of KCl for 10 min in the presence of 1 mM Ca2+. B, effects of Phe (1 µM; 3-min pulse) on [Ca2+]i in the presence of external Ca2+ (1 mM). C, resolution of the initial Ca2+ release phase from the sustained Ca2+ entry phase. Cells were initially superfused with a Ca2+-free saline solution and exposed to Phe (1 µM) before external Ca2+ was restored to 1 mM for the indicated period. D, effects of the sequential application (5-min pulses) of Phe (1 µM) on [Ca2+]i release in the absence of extracellular Ca2+ to cells whose [Ca2+]i pool had been completely discharged before the second pulse of Phe was applied.

Effect of Phe on [Ca²⁺]_i. Phe produced a concentration-related rise in [Ca²⁺]_i from a threshold of 10 nM to a maximum at 300 nM and a decline at concentrations beyond 1 µM (Fig. 4A). Responses were transient, declining after a peak despite the continued presence of Phe. This is illustrated at 1 µM Phe in Fig. 2B. In the presence of external Ca²⁺, Phe (1 µM; 3-min pulse) evoked a rapid rise of [Ca²⁺]_i that reached its maximum within 2.0 s and was followed by a rapid decline to a small plateau phase that remained sustained until Phe was withdrawn (Fig. 2B). Phe did not increase [Ca²⁺]_i in wild-type R-1Fs (n = 4, data not shown).

Role Played by External Ca²⁺ in [Ca²⁺]_i Response to Phe. Under Ca²⁺-free conditions, Phe evoked a transient rise in [Ca²⁺]_i without the sustained phase. Subsequent elevation of external Ca²⁺, in the continued presence of Phe, resulted in a maintained rise in [Ca²⁺]_i (Fig. 2C). In unstimulated cells, [Ca²⁺]_i remained unchanged by such a rise in external Ca²⁺ (n = 4; data not shown). If a response to Phe was elicited in the absence of external Ca²⁺ and Phe was removed, subsequent exposure to Phe elicited no response if external Ca²⁺ was withheld (Fig. 2D). Pretreatment of the cells with thapsigargin (1 µM), which depletes internal InsP₃-sensitive Ca²⁺ stores (Thastrup et al., 1990), completely abolished the Phe-evoked initial transient rise in [Ca²⁺]_i in Ca²⁺-free saline solution but had no effect on the sustained increase in [Ca²⁺]_i after the reintroduction of external Ca²⁺ (data not shown).

Reproducibility of Agonist-Mediated [Ca²⁺]_i Transient Responses. If extracellular Ca²⁺ was present, responses in a series of 6 pulses (30-s duration every 5 min) of 0.3 µM Phe were reproducible, indicating no interaction between responses with this protocol. Each Phe pulse evoked an essentially identical response in height (Fig. 3A) and time course (Fig. 3B). This provided a basis for subsequent construction of concentration-response curves and analysis of the time course of the response.

View larger version (23K): [in this window] [in a new window]   Fig. 3.   A, [Ca2+]i signals elicited by a series of 30-s pulses of 0.3 µM Phe that were delivered at intervals of exactly 5 min. B, superimposed [Ca2+]i transients (taken from A) showing the time course for the initial rate of [Ca2+]i release, corresponding peak rise, and fall in the Ca2+ signal evoked by each 30-s pulse of Phe (0.3 µM). C, superimposed [Ca2+]i signals comparing the time course for the initial rate of [Ca2+]i release and corresponding rise in [Ca2+]i evoked by increasing concentrations of Phe (traces 1-6, e.g., 0.01-3 µM). Each individual concentration of Phe was applied for 30 s and delivered at 5-min intervals. D, individual traces comparing the initial rate of [Ca2+]i release and peak rise in [Ca2+]i evoked by a supramaximal concentration of Phe (100 µM) under control conditions (trace 1), in the presence of a concentration of WB4101 that prevented the inverse phase of the concentration-response relationship and did not reduce Emax (e.g., 1 nM, trace 2) and a concentration that abolished the inverse phase and reduced Emax (e.g., 10 nM, trace 3).

Time Course of [Ca²⁺]_i Signal. Responses evoked by the lowest concentrations of Phe, 0.01 and 0.03 µM (e.g., traces 1 and 2 presented in Fig. 3C), did not reach a maximum or decline during the standard 30-s exposure. However, at higher concentrations, the start of the decline became gradually earlier as [Phe] was increased. Over the concentration range of 0.01 to 0.3 µM Phe (traces 1-4, Fig. 3C), the peak height and initial rate of rise of the peak [Ca²⁺]_i signal continued to increase despite the decline occurring sooner. Exposure to concentrations of Phe of more than 0.3 µM (dotted line, traces 5 and 6, Fig. 3C), however, caused a decrease in the peak height of the [Ca²⁺]_i signal. Essentially similar responses were elicited by (R)-A-61603 but at a lower concentration range (data not shown).

Functional Pharmacological Experiments

Agonist Potency. Typical traces recorded when the cells were stimulated noncumulatively with short exposures to increasing concentration of either Phe or (R)-A-61603 are depicted in Fig. 4, C and D. Phe and (R)-A-61603 both evoked concentration-dependent peak increases in [Ca²⁺]_i.

View larger version (32K): [in this window] [in a new window]   Fig. 4.   Concentration-dependent effects of Phe or (R)-A-61603 on [Ca2+]i release. Individual agonists were applied for 30 s and delivered at 5-min intervals. A, comparison of concentration-response data evoked by Phe (, n = 10) and A-61603 (, n = 6). Solid lines show sigmoid curves fitted to the pooled experimental data. B, initial rate of change in [Ca2+]i release evoked by each individual concentration of (R)-A-61603 (, n = 6) or Phe (, n = 10). C and D, representive traces illustrating the effects evoked by increasing concentrations of Phe or (R)-A-61603 on [Ca2+]i, respectively. E, representative recording showing the inhibitory effect plus the reversal of the inverse phase of the concentration-response relationship exerted by WB4101 (1 nM) on control Phe-evoked increases in [Ca2+]i. F, trace of [Ca2+]i against time illustrates the antagonism exhibited by BMY7378 (1 µM) (i.e., the response shown in Fig. 5C).

6

8

Antagonist Potency. Antagonism versus Phe by WB4101, prazosin, and BMY7378 was assessed across the concentration ranges indicated by their RLB affinities (Fig. 5). Corresponding quantitative parameters are depicted in Tables 1 and 2.

View larger version (35K): [in this window] [in a new window]   Fig. 5.   A, CRC to Phe in the absence () and presence of WB4101 (0.1 nM, n = 4, ; 1 nM, n = 4, ; and 10 nM, n = 5, ). Pooled data (mean ± S.E. for each test concentration) presented as fraction of maximal response evoked by Phe in control solution alone. G, Schild regression analysis of data presented in A; each data point represents an individual experiment. D, comparison of the initial rate of [Ca2+]i release evoked by each increasing concentration of Phe under control conditions (n = 10) and in the presence of WB4101 (0.1 nM, n = 4; 1 nM, n = 4; and 10 nM, n = 5). B, CRC to Phe constructed under control conditions () and in the presence of prazosin (1 nM, n = 4, ; 3 nM, n = 4, ; and 6 nM, n = 4, ). Data (mean ± S.E. for each test concentration, measured from pooled data) given as fraction of maximal response elicited by Phe under control conditions. H, Schild plot of data shown in B. E, initial change in the rate of [Ca2+]i release evoked by each concentration of Phe under control conditions (n = 10) compared with that elicited in the presence of each concentration of prazosin tested (1 nM, n = 4; 3 nM, n = 4; and 6 nM, n = 4). C, CRC to Phe generated in the absence () and presence of BMY7378 (0.1 µM, n = 4, ; 1 µM, n = 4, ; and 3 µM, n = 4, ). Pooled data (mean ± S.E. for each experimental condition) are expressed as fraction of maximal response evoked by Phe during superfusion with control solution alone. I, Schild regression analysis of data presented in C. F, comparison of the rate of [Ca2+]i release evoked by each concentration of Phe under control conditions (n = 10) and in the presence of BMY7378 (0.1 µM, n = 4; 1 µM, n = 4; and 3 µM, n = 4).

Time Course of Initial Rate of Change in [Ca²⁺]_i Release. Under control conditions, the initial rate of change in [Ca²⁺]_i evoked by (R)-A-61603 or Phe was essentially similar and concentration-dependent (Fig. 4B). With supramaximal concentrations of Phe or (R)-A-61603, this rate approached a maximum and did not exhibit an inverse concentration-response phase in contrast to the response measured as peak rise in [Ca²⁺]_i (compare Fig. 4B with Fig. 4A).

Effect of Antagonism on "Fade" and "Inverse Phase of Concentration-Response Relationship" . Preincubation of the cells with WB4101 (1 nM) prevented the response to 100 µM Phe from becoming smaller than the response to lower concentrations (compare trace 2 with trace 1, control; Fig. 3D). The time course showed a slowing of time to start of response, of initial rate of rise of [Ca²⁺]_i, and of time to onset of fade (Fig. 3D, compare trace 1 with 2). These effects were further exaggerated on exposure to a concentration of WB4101 (10 nM) that reduced E_max (Fig. 3D, compare trace 2 with 3).

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

The expression of cloned ARs in cells that lack endogenous ARs should simplify the analysis of receptor-response coupling and the pharmacological properties of a particular ₁-AR subtype. If quantitative agonist-antagonist pharmacology can be achieved in single cells harboring recombinant receptors, then this provides a well-defined baseline to identify functional receptors in cells isolated from native tissue or to identify the functional consequences of receptor mutations.

A system involving expression in cells that lack many of the signaling factors of native cells might be expected to show deviant properties. However, this study showed that reasonable quantitative pharmacology could be achieved. Moreover, some novel findings could be explained in terms of the properties of the nonequilibrium nature of the functional response and may shed light on previously unresolved issues in native tissue pharmacology.

The [Ca²⁺]_i signal in response to Phe consisted of two functionally distinct phases: an initial rapid transient phase due to mobilization of stored Ca²⁺ and a slowly developing smaller secondary phase due to receptor-regulated Ca²⁺ influx.

Schwinn et al. (1991) reported that the bovine _1a-AR evokes similar effects when stably expressed in HeLa cells, so [Ca²⁺]_i signaling pathway observed from the transfected R-1Fs was consistent with this report and with studies undertaken in rat portal vein myocytes (Lepretre et al., 1994) and a neural cell line (Esbenshade et al., 1993).

[Ca²⁺]_i is held at a constant low level by Ca²⁺ buffering within the cell and by the balance between Ca²⁺ release/influx into the cytosol and the opposing actions of the mechanisms that extrude Ca²⁺ from the cytosol. Receptor activation results in this balance being temporarily overcome. During short agonist exposures, the major effect is rapid mobilization of stored Ca²⁺ into the cytosol, which exceeds buffering and extrusion, causing a rise in [Ca²⁺]_i. Because the stores are finite, the release rate quickly declines to the point where it is exceeded by removal processes and [Ca²⁺]_i declines. This explains why the response shows "fade" during each 30-s exposure to agonist.

At different agonist concentrations, the time course of the fade varied in a simple manner consistent with the above interpretation. At low agonist concentrations, responses were slow to rise, and with the 30-s exposure, they did not reach a peak or fade, which is consistent with the Ca²⁺ stores not yet having emptied. As the agonist concentration increased, the responses rose more quickly, reached their peak, and faded earlier, as would be expected from the stores having been discharged more quickly. This pattern continued with each agonist up to an optimal concentration (0.3 µM for Phe), beyond which an additional action set in: the responses started to achieve a smaller peak.

Despite the reduced peak height, the initial rate of rise of the response continued to increase with concentration, so there was no loss of initial stimulus, and the peak continued to occur progressively earlier. This would not be expected from a simple run-down of the intracellular stores. It appears that the response has been limited by a reduced release from the stores before they have emptied. This explanation is supported by the observation that when the peak is reduced, the response takes longer to decline, as it might if Ca²⁺ release continued at a lower level for a longer time. A possible mechanism lies in a known negative feedback mechanism. High levels of [Ca²⁺]_i can inhibit InsP₃ receptors situated on the endoplasmic reticulum and the activity of its integral Ca²⁺ release channel (Goldbeter et al., 1990; Atri et al., 1993). Thus, a high agonist concentration will generate a high InsP₃ level, which in turn will produce a high local level of [Ca²⁺]_i (ahead of a general rise) capable of inhibiting Ca²⁺ release before the run-down of Ca²⁺ stores becomes the limiting factor. Regardless of the detailed mechanism, the key point is that a very "fast" response achieves a lesser peak than a slightly slower one. This was interesting to us because it can explain an aspect of the agonist-antagonist interaction that we found: namely, the abolition of the inverse phase of the concentration-response relationship in the presence of antagonists.

The equilibrium relationship between the agonist, antagonist, receptor pool, and the response depends on the fractional receptor occupancy required by the agonist to achieve a response and the dissociation equilibrium constants for that receptor, the agonist, and the antagonist. In turn, the dissociation equilibrium constants for each drug/receptor depends on the ratio of its association rate constant and its dissociation rate constant. However, when the antagonist is present at equilibrium (at sufficient concentration to occupy a high proportion of the receptors) and the agonist is added, the initial agonist association rate will depend to a great extent on the antagonist dissociation rate because the agonist can associate only with free receptors. This has the effect that the rate of formation of agonist-receptor complex will be slowed. Consequently, high agonist concentrations, which normally evoke peak transients that are smaller in magnitude than lower agonist concentrations, were able to generate their effect more slowly and thus escape the time limitation on their Ca²⁺ signal.

At high antagonist concentrations, the dominant influence becomes the reduced number of available sites. Therefore, the reduction in signal outweighs the advantage of its effect being spread out over time. This leads to the classic situation of the CRC moving right and down, characteristic of a response limited by reduction in receptor number rather than surmountable antagonism (Furchgott, 1966, 1972).

By defining the nature of the functional response, in terms of the Ca²⁺ signal, it has been possible to analyze the general nature of the agonist-response relationship for this recombinant subtype and to hypothesize a predictive relationship for the interaction of agonists and antagonists. During agonist stimulation, the peak [Ca²⁺]_i signal, which is due to rapid mobilization of stored Ca²⁺, was short-lived. The transient nature of the receptor-response coupling process therefore does not permit sufficient equilibration time to produce a true thermodynamic equilibrium between the agonist/antagonist and receptor, precluding accurate extraction of equilibrium constants from Schild regressions, which are valuable probes in evaluation of the kinetics of drug-receptor interactions and detection of nonequilibrium steady states. Nevertheless, we do know the affinities of the relevant drugs from RLB to the relevant receptors expressed in the same cells.

This thus is a very tightly defined system in which some of the uncertainties inherent in heterogeneous native systems are absent. It should be possible to validate or otherwise shed some light on the functional pharmacology of the recombinant _1a-AR and its putative relationship to native _1A-AR.

In terms of agonist selectivity, among the native ₁-AR subtypes, Knepper et al. (1995) found the highest ratio of potency of A-61603 to Phe for _1A-AR. This is confirmed by the present study, which showed a parallel concentration-response relationship for the two agonists separated by a 141-fold difference in potency and reflecting a 60-fold difference in RLB affinity versus [³H]prazosin in the same clone.

With regard to antagonist potency, our analysis predicts that competitive antagonists would become insurmountable as their concentration increased, but the threshold for antagonism should still lie near to their affinity constants. Taking this into account, the relative potencies found (WB4101  prazosin BMY7378) are consistent with the pattern for _1a-AR RLB (present results) or reexpressed _1a-AR responses measured by InsP₃ production (Schwinn et al., 1991, 1995). A small deviation from this is that WB4101 was more potent than prazosin at antagonizing nonequilibrium responses, although they had equal affinity in RLB. This is a fairly common observation in functional agonist/antagonist studies in tissues expressing native _1A-AR (McGrath, 1984). In the present study, this might reflect a slower dissociation rate for WB4101. This would be borne out by the abolition of the inverse phase, which was clearest for WB4101.

An interesting light was cast on the properties of BMY7378, which is considered to be selective for _1D-ARs relative to the other two subtypes (Goetz et al., 1995). Although its potency in this study was consistent with its relatively low affinity at _1A-ARs, its depression of the rate of rise of response was unexpectedly high, suggesting that it might dissociate slowly from the receptor. In situations in which this is an important factor, it might therefore owe its potency to its dissociation rate constant rather than its dissociation equilibrium constant (measured by RLB). Because there are few other antagonists that reliably distinguish between _1D- and _1A-ARs, this might be a significantly misleading factor. For example, a relatively high potency of this compound might indicate a nonequilibrium component within the response rather than an _1D-AR.

In conclusion, we analyzed both affinity and efficacy in a well-defined recombinant clone using a single-cell system in which the Ca²⁺ signal was characterized. The data give a pharmacological analysis that is internally consistent in terms of known data for the _1a-AR and bodes well for using such techniques to analyze the pharmacology in isolated single cells from heterogeneous native tissues where whole-tissue analysis lacks validity. The study also unveiled a form of fade that could be readily explained.