Selective β₁-adrenoceptor blockade enhances the activity of the stimulatory G-protein in human atrial myocardium

Abstract

1. Chronic selective β₁-adrenoceptor (β₁AR) blocker treatment enhances the sensitivity of β₂-adrenoceptor (β₂AR) in human heart (Hall et al., 1990; 1991). To clarify the mechanism of the cross-sensitization between β₁AR and β₂AR, we determined whether the stimulatory G-protein (G_sα) function is increased in atria from β₁AR-blocker treated patients compared with non-β-blocked patients, and investigated whether this change is caused by an alteration of post-translational modification of G_sα protein.

2. G_sα function was determined by reconstitution of human atrial G_sα into S49 cyc⁻ cell membranes. In the reconstitution system, GTP_γS stimulated cyclic AMP generation in a dose-dependent manner. Upon 10⁻⁴ M GTP_γS stimulation, G_sα activity in the β₁AR-blocker, atenolol, treated group (78.2±10.3 pmol cyclic AMP mg⁻¹ min⁻¹ 10⁻³) was 65% higher than that in non-β-blocked patients (47.3±6.3 pmol cyclic AMP mg⁻¹ min⁻¹ 10⁻³, n=15, P=0.02).

3. Isoelectric point (pI) values of G_sα were measured by two dimensional gel electrophoresis (2D-E) and the amount of each isoform quantified by image analysis of a Western blot of the gel using specific antibody. Multiple isoforms of G_sα were detected by 2D-E with different pI values. There were no significant differences between the groups of patients in either pI values or the proportions of the acidic isoforms of G_sα to the main basic form (n=12, P>0.05).

4. The results suggest that chronic β₁AR-blockade enhances G_sα function in human atrium, and this may account in part for the hypersensitivity of β₂AR and other G_s-coupled receptors during β₁AR-blockade. The increased G_sα function is unlikely to be caused directly by blockade of protein kinase A phosphorylation of G_sα protein.

Introduction

β-Adrenergic receptor (βAR) antagonists are widely used for the treatment of ischaemic heart disease, hypertension and heart failure. To minimize the adverse effects of β₂AR-blockade, so-called cardio-selective (β₁AR selective) βAR-blockers were introduced. However, subsequent evidence has shown that human atrium and ventricle contain an appreciable proportion of functional β₂AR (Kaumann & Lemoine, 1987; Del Monte et al., 1993), and that, moreover, chronic treatment of patients with a β₁AR-selective blocker causes marked enhancement of β₂AR mediated atrial inotropic responses (Hall et al., 1990). The β₂AR hypersensitization by β₁AR blockade has been confirmed in vivo, where there is a 6 fold increase of the potency of salbutamol during intra-coronary infusion of this drug in patients receiving a β₁AR-blocker (Hall et al., 1991). This phenomenon may benefit the patients with chronic heart failure by improving the pumping function during β₁AR-blockade, but on the other hand, in patients with myocardial infarction the increased sensitivity to β₂AR stimulation is likely to increase heart rate and myocardial oxygen consumption, thereby increasing the risks of arrhythmia and the extent of infarction.

When the βAR is occupied by an agonist, the resulting conformational change of the receptor activates the stimulatory heterotrimeric G-protein (G_s) to bind guanosine triphosphate (GTP) to its α-subunit. The G_sα-GTP complex then stimulates adenylyl cyclase (AC) to generate the second messenger cyclic AMP. Cyclic AMP subsequently gives rises to the activation of protein kinase A (PKA) which can phosphorylate a series of proteins. Elements involved in G_s-mediated signal transduction cascade are obvious targets for clarifying the mechanism of cross regulation between receptors coupled to adenylyl cyclase. In support of this suggestion is the evidence that two other G_s-coupled receptors, the 5-HT₄-receptor and histamine H₂-receptor, mediate inotropic responses in human atria which are potentiated by chronic β₁AR-blockade (Sanders et al., 1995; 1996).

Indirect evidence for the involvement of the coupling mechanism in the receptor ‘cross-talk' comes also from the findings that there is no increase in β₂AR density or affinity for ligands in β₁AR blocked right atria, nor any change in the cellular sensitivity to exogenously applied cyclic AMP analogues (Hall et al., 1990). However no quantitative change has been found following β₁AR-blocker treatment in the level of G-protein subunits (G_sα, G_iα, Gβ) (Ferro et al., 1993) or the mRNA of the four splice variants of G_sα and G_i2α (the principle cardiac G_i protein) (Jia et al., 1995; Monteith et al., 1995). A variety of post-translational modifications of G_sα have been reported, such as phosphorylation and ribosylation, some occurring in response to cyclic AMP stimulation (Yamane et al., 1993; Degtyarev et al., 1993; Linder et al., 1993; Pyne et al., 1993). These modifications therefore raise the possibility of a functional feedback regulation of G_s function without changes in its transcription.

The present study was designed to measure G_s protein function by reconstitution of G_sα protein from human atrium into S49 cyc⁻ lymphoma cell membranes. In this system, S49 cyc⁻ cells are genetically deficient in G_sα and therefore lack G_s-mediated adenylyl cyclase activity. The addition of solubilized G_sα to cyc⁻ cell membranes can restore G_s-mediated adenylyl cyclase activity. The rate of cyclic AMP generation theoretically reflects pure G_s function.

As a consequence of the activation of the G_s-mediated signal transduction pathway, the phosphorylation status and possibly function of numerous proteins may be increased by PKA. G_sα can be phosphorylated by PKA in vitro (Pyne et al., 1992). It is not clear whether this occurs in vivo. However we have reported that the serine⁺ variants of G_sα , which might be more prone to PKA phosphorylation, have a more acidic iso-electric point (pI) – consistent with in vivo modification by phosphorylation (Monteith et al., 1995). In the present study, two dimensional gel electrophoresis was employed to determine whether there is any change in the pI value of G_sα protein from human atria after chronic treatment with a β₁AR-blocker.

Methods

Patients

Right atrial appendage (RAA) was obtained immediately following its routine removal from patients undergoing coronary artery bypass grafting, at the time of institution of cardiopulmonary bypass. Premedication was with papaveretum and hyoscine; anaesthesia was induced with midazolam, fentanyl and propofol, with pancuronium as muscle relaxant. Propofol infusion was used for maintenance of anaesthesia. RAA were received from 54 patients; 30 samples were used for G_s functional assay, and the other 24 were used for two-dimensional gel electrophoresis. Half the samples for each assay were from patients receiving chronic β₁AR-blocker atenolol treatment (50 mg day⁻¹, over 1 month) at the time of surgery, and half from non-β-blocked patients. Although the choice was not randomized, it tended to reflect individual referring physicians' preference for β-blockade or Ca²⁺-blockade as secondary treatment in patients already receiving nitrates. Patients were matched as far as possible in terms of age, sex and diagnosis in both sets of studies (Table 1). Other drug therapy included aspirin, calcium channel antagonists, diuretics and nitrates. Tissue samples were immediately snap-frozen in liquid nitrogen as soon as they were removed during the heart operation and stored at −70°C until use. The use of tissue routinely removed and discarded during surgery was approved by the local research ethics committee without patients written consent being required.

Table 1.

Characteristics of patients

Cell culture and preparation of S49 cyc⁻ cell membranes

S49 cyc⁻ lymphoma cells were grown in suspension culture in HEPES buffered Dulbecco's modified Eagle's medium that had been supplemented with antibiotics (100 U ml⁻¹ penicillin and 100 μg ml⁻¹ streptomycin), antimycotic (0.25 μg ml⁻¹ amphotericin B) and 5% foetal calf serum. The population density was maintained at about 10⁶ cells ml⁻¹. 2.4×10⁹ cyc⁻ cells were harvested, the cells pelleted by centrifuging at 1000×g for 20 min at 4°C and washed twice with HEPES buffer (HEPES 50 mM, EDTA (pH 8.0) 5 mM). The cell pellet was suspended in HEPES buffer and homogenized in a ground glass homogenizer. The homogenate was centrifuged at 1000×g for 10 min at 4°C and the supernatant collected as supernatant 1. The pellet was re-homogenized in HEPES buffer and the supernatant collected as supernatant 2. Supernatants 1 and 2 were pooled, filtered through four layers of gauze and centrifuged at 48,000×g for 30 min at 4°C. The pellet was resuspended in HEPES buffer. Protein concentration was measured in 96-well plates using the Bio-Rad protein assay with bovine serum albumin as standard. The cyc⁻ cell membrane preparation with a final concentration of ∼2 mg ml⁻¹ was stored in separate aliquots at −70°C.

Cholate extraction of G_s protein from human right atrium

Right atrial appendage tissues (40–200 mg) were debrided of fat and connective tissue and homogenized in buffer A (170 μl 100 mg⁻¹ tissue, HEPES (pH 8.0) 10 mM, EDTA 1 mM, benzamidine 3 mM and 1 μg ml⁻¹ aprotinin). Sodium cholate and β-mercaptoethanol were added to the homogenate to give a final concentration of 1% and 20 mM respectively. After vortexing on ice for 60 min the homogenate was centrifuged at 100,000×g for 60 min at 4°C. Protein concentration of the supernatant was measured as above to be 10 μg μl⁻¹ on average, and the remainder aliquoted and stored at −70°C until use.

Reconstitution of G_sα activity in S49 cyc⁻ lymphoma membrane

Cholate extracts were diluted in buffer A. Frozen cyc⁻ membranes were resuspended in fresh HEPES buffer. Cyc⁻ membranes (40 μg) and RAA cholate extracts were incubated together in a 100 μl reconstitution mix containing cyclic AMP 1 mM, ATP 0.1 mM, creatine phosphate 20 mM, 40 U creatine phosphokinase, 40 U myosin kinase, GTP 0.05 mM HEPES 50 mM, EGTA 0.1 mM, MgCl₂ 10 mM, 0.1 mg ml⁻¹ BSA and 5 μCi [α-³²P]-ATP. After 40 min incubation in a water bath, the reaction was terminated by adding 100 μl of 40 mM ATP and heating at 90°C for 5 min. The cyclic AMP formed was isolated by sequential chromatography using a Dowex 50 cation exchange and neutral alumina column. Recovery of cyclic AMP was monitored by the addition of [³H]-cAMP to each tube. The [³²P]- and [³H]-cAMP eluted from the alumina were quantified by double-isotope liquid-scintillation spectroscopy. A concentration curve of Guanosine-5′-3-O-(thio)triphosphate (GTP_γS) stimulated adenylyl cyclase activities was examined in an atrial sample with triplicate determinations of each concentration of GTP_γS, and 10⁻⁴ M GTP_γS stimulated cyclic AMP generation with different amount of membrane protein (12–70 μg) was tested. Results were expressed as the amount of cyclic AMP formed per minute per milligram of protein.

Membrane preparation from human right atrium

Samples of right atrial appendage were debrided of fat and connective tissue. The remaining material was placed in ice cold buffer B (β-glycerophosphate 10 mM, HEPES 10 mM, benzamidine 2 mM, PMSF 1 mM, 5 μl mg⁻¹ tissue), minced with scissors and homogenized using a Polytron at maximum speed (setting 10) for 5×5 s. The homogenate was centrifuged at 2000×g for 10 min at 4°C. The resulting supernatant was centrifuged at 48,000×g for 15 min at 4°C to obtain a crude plasma membrane fraction that was resuspended in buffer B and stored in separate aliquots at −70°C. Protein concentration was measured as above.

Isoelectric focusing (IEF)

Two dimensional electrophoresis was performed according to the method of O'Farrel (1975) using the MINI-PROTEAN II 2-D CELL system. IEF gel monomer solution was made containing 9.2 M urea, 2% NP40, 0.018% ammonium persulphate, 0.025% TEMED, 2% pH 3–10 ampholyte, 1% pH 5–7 ampholyte, 1% pH 3.5–5 ampholyte and 1% pH 7–9 ampholyte. IEF tube gels were cast in glass capillary tubes of 10 cm long and 1 mm in diameter. Samples of crude plasma membrane were mixed with an equal volume of first dimensional sample buffer (9.5 M urea, 4% NP40, 5% β-mercaptoethanol, 0.4% pH 3.5–10 ampholyte, 1.6% pH 5–7 ampholyte) and incubated at room temperature for 10–15 min. The tube gel was pre-run at 200 V for 10 min, 300 V for 15 min and 400 V for 15 min with the upper chamber running buffer NaOH 20 mM (degassed) and lower chamber running buffer H₃PO₄ 10 mM. About 10 μg membrane proteins (25 μl) were loaded onto the tube gel. On the top of the sample, 30 μl sample overlay buffer (9 M urea, 0.8% pH 5–7 ampholyte, 0.2% pH 3–10 ampholyte) was loaded. The gel was run at 500 V for 10 min, 750 V for 3.5 h and 2 KV for 30 min. After IEF was completed, the tube gels were ejected into equilibration buffer (Tris-HCl (pH 6.8) 0.0625 M, 2.3% (w  v⁻¹) SDS, 2-mercaptoethanol 10 mM, 10% (w v⁻¹) glycerol) and frozen at −70°C until the second-dimension was run.

For measurement of the pH gradient along the tube gel, Pharmacia 2-D Electrophoresis Calibration Kit was used. Carbamylated creatine phosphokinase (CPK) protein was included with each atrial sample prior to iso-electric focusing and SDS–PAGE. The series of spots (approximately 34 spots) between pI values of 4.9–7.1 were visualized by Ponsau S staining of the nitro-cellulose membrane after Western blotting. The pI value of G_sα was calculated by comparison with these standard protein spots.

SDS–PAGE and immunoblotting

IEF gel strips were loaded directly on the top of the slab gel, excluding any air bubbles. The second-dimension separation was carried out at room temperature with a mini stacking gel (1×8×0.05 cm, 4%T) and separating gel (5×8×0.05 cm, 10%T) system. Voltage was set at 200 V and current at maximum.

Following electrophoresis, immunoblotting of G_sα protein was performed as described previously (Monteith et al., 1995). Briefly, proteins were electro-blotted onto a nitro-cellulose membrane for 1 h at 0.8 mA cm⁻² with LKB117-250 Novablot apparatus. After blocking the non-specific binding sites by immersing in 5% dried milk (Marvel) for 30 min, the membrane was incubated at room temperature with anti-G_sα antiserum (1 : 750) for 2 h. This antiserum was generated against a synthetic G_Sα C-terminal decapeptide (RMHLRQYELL) and its characterization has been described previously (Miligan & Unson, 1989; Ohisalo et al., 1989). After washing with 0.1% Tween-20 (v v⁻¹) made in Tris-buffer saline (TBS, NaCl 200 mM, Tris-HCl, (pH 7.4) 50 mM), the membrane was then incubated for 1 h with horseradish peroxidase-labelled goat anti-rabbit immunoglobulin (1 : 500 dilution). After the membrane was washed again, G-protein dots were visualized by incubating with enhanced chemiluminescence (ECL) detection reagents.

Quantification of G_sα protein

There was inadequate RAA protein to permit combined G_sα function and mass quantification except in four β-blocked and seven non β-blocked patients. Quantification was performed by 1-D Western blot, as previously described (Ferro et al., 1993), and results expressed in arbitrary optical density units. This permitted an assessment of G_sα specific activity in the 11 samples.

Reagents

Dulbecco's modified Eagle's medium was from Life Technologies (Paisley, Scotland, U.K.). Foetal calf serum was from Globepharm. [α-³²P]-ATP, [³H]-cAMP and enhanced chemiluminescence (ECL) detection reagents were from Amersham (Amersham International plc., U.K.). Anti-G_sα antiserum was kindly donated by Professor G. Milligan (Molecular Pharmacology Group, Department of Biochemistry, University of Glasgow, Scotland, U.K.). Horseradish Peroxidase-labelled goat anti-rabbit immunoglobulin was from DAKO (Denmark). Protein assay kit was from Bio-Rad (Bio-Rad Laboratories Ltd, Herculaes, CA, U.S.A.). 2-D Electrophoresis Calibration Kit was from Pharmacia (Pharmacia Inc, U.S.A.). Other chemicals were from Sigma (St. Louis, MO, U.S.A.).

Statistics

All data were expressed as mean±s.e.mean. Statistical comparison of the reconstituted G_sα activity and the pI value between β₁AR-blocked and non-β-blocked samples were by means of unpaired Student's t-test. The changes in the proportions of the acidic isoforms of G_sα to the main basic form after β₁AR-blockade was assessed by a multiple regression analysis incorporating patients' ages and estimated left ventricular function. A value of P<0.05 was considered significant.

Results

G_sα activity in cyc⁻ reconstitution assay

GTP_γS stimulation of adenylyl cyclase activity in cyc⁻ cell membranes was restored after combination with atrial extracts containing human G_sα. Cyclic AMP generation was dose-dependent up to the maximal concentration used of 10⁻⁴ M GTP_γS (Figure 1). 10⁻⁴ M GTP_γS-stimulated cyclic AMP generation varied linearly with the amount of membrane protein used up to 28 μg. Fifteen micrograms membrane protein was available from all RAA, and was used for the comparison of G_sα activity between β₁AR-blocked and non-β-blocked human atria. Cholate extracts were diluted before addition to the reconstitution mix, and in the absence of S49 cell membranes, restored negligible G_sα mediated adenylyl cyclase activity upon GTP_γS stimulation (data not shown). Reconstituted G_sα activity in the presence of 10⁻⁴ M GTP_γS was 78.2±10.3 pmol cyclic AMP mg⁻¹ min⁻¹ 10⁻³ in 15 atenolol treated patients, which was 65% greater than in 15 non-β-blocked patients, 47.3±6.3 pmol cyclic AMP mg⁻¹ min⁻¹ 10⁻³ (P=0.02, Figure 2).

Figure 1.

Dose-dependent increase of G_sα activity by GTP_γS stimulation in the G_sα reconstitution assay. Data are mean±s.e.mean of triplicate determinations from one patient's right atrial appendage.

Figure 2.

Comparison of atrial G_sα activity between β₁AR-blocked and non-β-blocked patients. G_sα activity was measured as the increase in adenylyl cyclase activity in S49 cyc⁻ cell membranes in the presence of 10⁻⁴ M GTP_γS. Data are represented as means of triplicate determinations of each patients' sample (open triangles) and means of data from 15 patients' right atrial appendage (solid triangle, mean±s.e.mean). ^*P=0.02.

In the 11 patients where the G_sα protein could be quantified, there was a trend for higher specific activity in the β₁-blocked atria (13.0±2.1 arbitrary units) than that in the non-β-blocked atria (7.3±2.3), although this was not a significant difference in the small number of samples.

2D-gel electrophoresis and immunoblotting

Typical 2D-patterns of G_sα from human atrium are shown in Figures 3 and 4. A total of 11 forms of G_sα were detected. Those at MW 52 KDa correspond to G_sα long isoform (G_sαL) and those at 45 KDa to G_sα short isoform (G_sαS). The pI for the G_sαL residues (pI 5.53±0.01) were less acidic than G_sαS (pI 5.19±0.01, P<0.001), which confirms the previous finding. The densest and most basic spot was taken as the reference point and assumed to be the unmodified G_sα.

Figure 3.

Sketchgraph shows the 2D-pattern of the G_sα protein from human atria. Membrane proteins were separated horizontally by iso-electric point and vertically by mass. G_sα was detected by immunoblotting with specific antibody. The dots were numbered from 1–11.

Figure 4.

Comparison of two-dimensional pattern of G_sα in right atrial appendage between patients receiving β₁AR-blocker and non-β-blocker treatment. Four typical 2D-patterns of G_sα from each group are shown here. (A, B, C and D) β₁AR-blockade; (E, F, G and H) non-β-blockade.

The 2D-pattern of G_sα from the atria of 12 atenolol treated and 12 matched non-β-blocked patients were compared. The patterns were very reproducible. The pI value of the major isoforms of G_sαL and G_sαS were 5.53±0.01 and 5.19±0.01 for the β₁AR-blocked group, compared to 5.53±0.01 and 5.17±0.01 for the non-β-blocked group, (P>0.05, Figure 4). In order to determine whether the relative amount of these isoforms changed after β₁AR-blockade, the isoforms on the 2D-pattern from each individual were numbered from 1–11 (Figure 3) and were quantified by laser-densitometry, and the ratio of each dot to the main basic dot (number 1) calculated. The results are shown in Table 2. No significant differences were observed in the proportions of the acidic residues of G_sα relative to the main basic residue between these two groups of patients (P>0.05).

Table 2.

Quantification of G_sα isoforms in atria from β₁AR-blocked and non-β-blocked patients

Discussion

Previous studies showed that chronic β₁-blocker treatment causes a several fold increase in the potency of β₂AR-agonists and antagonists at adenylyl cyclase coupled receptors (Hall et al., 1990; 1991; 1993; Sanders et al., 1995; 1996). These observations have been made both in vitro and in vivo, the most striking being a 6 fold increase in potency of sulbutamol when infused into the right coronary artery of atenolol treated patients. This sensitization occurs with no increase in receptor number or occupation, or of cyclic AMP responsiveness (Hall et al., 1990; Brodde, 1991), pointing to enhanced coupling of the receptors to adenylyl cyclase through G_s. However in several studies of G_sα and β subunit mRNA and protein levels, no overall differences in levels were found (Ferro et al., 1993; Jia et al., 1995; Monteith et al., 1995). The present study shows that there is indeed an increase in G_s function in the atria of β₁AR-blocked patients, but fails to elucidate the mechanism.

There was reason to expect that β₁AR-blockade might influence post-translational modification of G-proteins. In the heart, the βAR–G-protein–adenylyl cyclase signal transduction pathway is normally under a high level of adrenergic stimulation, which results in tonic PKA activity. There is evidence that G_sα can be phosphorylated through PKA in vitro (Pyne et al., 1992), and it is plausible therefore that the PKA activation could cause G_sα to be phosphorylated in vivo. Chronic blockade of β₁AR interrupts the activation cascade and would therefore reverse any PKA stimulated phosphorylation of G_sα. Other kinds of post-translational modification of G_sα, such as palmitoylation and ADP-ribosylation, could also be affected in the same way, especially the latter which has been shown to be activated by cyclic AMP (Yamane et al., 1993; Degtyarev et al., 1993; Linder et al., 1993; Pyne et al., 1992). We have ourselves found indirect evidence that sympathetic tone in vivo does influence the post-translational modifications of G_sα identified by 2D-pattern changes. The finding was that the proportion of some of the acidic isoforms of G_sαL is much lower in saphenous vein than in the heart (Wang & Brown, 1996), consistent with the much lower sympathetic tone in the former. The lack of difference in the present study of pI values in atria of G_sα between β₁AR-blocked and non-β-blocked patients implies that G_sα is unlikely to be modified directly by cyclic AMP-dependent protein kinase A in vivo. It is conceivable that such phosphorylation has still occurred through the β₂AR, but in a small number of atria from patients receiving non-selective β-blockade, a similar 2D-pattern is still observed.

The precise mechanisms that result in the increase of G_sα activity remain intriguing. The reproducibility of the 2D-gel electrophoresis means that we could have readily detected changes in modification as great as the increase in G_s function present. The possibility of changes of some post-translational modifications that do not alter the pI values of the G_sα protein could not however be excluded.

There was substantial overlapping of the G_sα function data between the two groups of patients, even though the means were significantly different. This overlapping, together with the modest mean increase in G_sα function raises the question whether this difference is sufficient to explain the 5–10 fold increase of the inotropic response of the atria from β₁AR-blocked patients (Hall et al., 1990; 1991). Since the signal transduction cascade is an amplification process, we would not expect 1 : 1 equivalence between G-protein activity and the final physiological response, but we have not attempted to measure the gearing. The increased G_sα function following β₁AR-blockade shows that receptor coupling to adenylyl cyclase is one site of receptor cross talk, but it is likely this is not the unique change which contributes to the increased β₂AR sensitivity. At the receptor level, chronic administration of bisoprolol, a β₁AR-selective blocker, to pigs has been reported to cause a reduction in left ventricular β-adrenergic receptor kinase (βARK) activity (Ping et al., 1995). The reduction of βARK expression was also observed in mouse model with long-term administration of atenolol (laccarino et al., 1998). Beyond G_sα in the transduction pathway, there are up to nine types of adenylyl cyclase that have been identified so far (Krupinski et al., 1989; Gao & Gilman, 1991; Ishikawa et al., 1992; Premont et al., 1992; Katsushika et al., 1992; Watson et al., 1994). Of these at least four types are present in the heart (AC IV, V, VI and VII) (Ishikawa & Homcy, 1997; Iyengar, 1993). Are these cardiac adenylyl cyclases coupled equally to different receptors or are some of the adenylyl cyclases switched on or off upon β₁AR-blockade? Determination of the gene expression of each subtype of the adenylyl cyclases upon β₁AR-blockade is ongoing. The multiple isoforms of the adenylyl cyclase present in the heart, and evidence of compartmentalization of cyclic AMP activation (Zhou et al., 1997) limit the value of comparison of the overall adenylyl cyclase activity between the two groups of patients. Cross-talk between the β₂AR and other receptors or cell signalling systems has been described in the organ bath. We believe these observations are not relevant to the β₂AR cross-talk we observe, which develops over several hours or days.

G_sα reconstitution study is a classic method to test G_sα function with the advantage of the availability of S49 cyc⁻ cell line genetically deficient in G_sα. The early experiments with the method by Ross and Gilman led to the discovery of G_sα (Haga et al., 1977; Ross & Gilman, 1977). However, when used to test G_sα function, the disadvantage is that cholate extraction causes subunit dissociation of G-protein. Therefore the method would under-estimate changes related to subunit interaction, for instance if reduced G_s activation during β₁AR-blockade results in a higher proportion of intact trimeric G_s-protein in the atrial membrane.

In summary, chronic β₁AR-blockade can enhance G_sα function in human atrium, explaining at least partially the sensitization of β₂AR mediated responses by β₁AR-blockade. The lack of difference in pI values of G_sα between β₁AR-blocked and non-β-blocked atria suggests that the increased G_sα activity is not due to a change in the phosphorylation status of G_sα protein consequent on the reduced activity of cyclic AMP-dependent protein kinase A.

Abbreviations

β₁AR

β₁-adrenoceptor

β₂AR

β₂-adrenoceptor

βARK

β-adrenergic receptor kinase

cyclic AMP

adenosine 3′ : 5′-cyclic monophosphate

CPK

carbamylated creatine phosphokinase

2D-E

two-dimensional gel electrophoresis

ECL

enhanced chemiluminescence

Gβ

β-subunit of G-protein

Giα

α-subunit of inhibitory G-protein

G_s

stimulatory heterotrimeric G-protein

G_sα

α-subunit of stimulatory G-protein

G_sαL

the long isoform of the stimulatory G-protein α-subunit

G_sαS

the shot isoform of the stimulatory G-protein α-subunit

GTP

guanosine triphosphate

GTP_γS

guanosin-5′-3-O-(thio)triphosphate

IEF

isoelectric focusing

PKA

protein kinase A

pI

isoelectric point

RAA

right atrial appendage

SDS–PAGE

sodium dodecyl sulphate-polyacrylamide gel electrophoresis