The adenosine A₁ receptor (A₁R) is a useful model of GPCR binding and activation. We have reported previously on the effects of A₁R expression level on the equilibrium binding properties of the receptor (Browning et al., 2000). The aim of this study is to examine whether the ligand binding properties of the A₁R-GFP(green fluorescent protein) and A₁-GFP-G _i fusion proteins (Bevan et al., 1999) differ from each other and the A₁R and whether these properties are sensitive to the receptor expression level.

CHO cell lines expressing high and low levels (ca 9 and < 1 pmole/mg protein) of the human A₁ receptor (HE and LE) have been described previously (Browning et al., 2000). Clonal CHO cell lines expressing either A₁R-GFP and A₁R-GFP-G _i fusion constructs (Bevan et al., 1999) at different levels were obtained after selection of individual cells by FACS sorting. A total of 24 and 23 cell lines expressing the A₁R-GFP and A₁R-GFP-G _i fusion proteins, respectively, were generated. Membranes were harvested as described previously (Browning et al, 2000). Radioligand binding studies were carried out for 120 min. at 23^o in a Na/Mg/Hepes buffer (Cohen et al., 1996). Because of the large number of cell lines, an efficient process was devised to determine the receptor binding properties. The total number of binding sites (Bmax) and the affinity (logK_A) of the antagonist [ ³H]DPCPX (1,3-dipropyl-8-cyclohexylxanthine) were determined using three concentrations of the radioligand. The binding affinities (log K_H and log K_L) of the high efficacy agonist CHA (N ⁶-cyclohexyladenosine) for the high and low affinity binding components (representing G-protein coupled and uncoupled receptor states) and the fractions of the high affinity component (fr_H) were measured in [ ³H]DPCPX(1nM)-CHA competition experiments using just three appropriately selected concentrations of CHA (3, 100 and 3000 nM).

The range of Bmax values for A₁R-GFP and A₁R-GFP-G _i(1-9 and 1-13 pmole/mg protein, n= 46 and 65, respectively) encapsulated those observed for HE and LE (8.7 ± 0.3 and 0.8 ± 0.1 pmole/mg, n=3). The mean log affinities of DPCPX (8.88 ± 0.01, n = 46, 8.72 ± 0.02, n = 65) and CHA (8.42 ± 0.06, n = 41, 8.47 ± 0.04, n = 61 for log K_H; 6.02 ± 0.14, n = 41, 6.05 ± 0.05, n = 63 for log K_L for A₁R-GFP and A 1R-GFP-G _i respectively) were independent of Bmax and very close to the values for A₁R [log K A 8.62 ± 0.09 (LE), 8.76 ± 0.03 (HE); log K_H 8.60 ± 0.08 (LE), 8.60 ± 0.03 (HE); log K_L 5.67 ± 0.24 (LE), 5.71 ± 0.05 (HE), n = 3]. However, independent of the receptor construct, fr_H was dependent on the expression level, being low (0.4-0.5) at Bmax values above 8 pmoles/mg protein and increasing progressively towards 1 at low Bmax.

It is concluded that the equilibrium binding properties of A₁R-GFP and A₁R-GFP-G _i are not sensitive to the presence of the tethered GFP or G _i . These constructs appear to interact as effectively with free G-proteins as A₁R. The properties of CHA-A₁R-G and the equivalent fusion protein complexes appear to switch from effectively a receptor: G protein stoichiometry of 1:1 to 2:1 as the concentration of receptor increases. This could be a manifestation of receptor dimerisation or oligomerisation.

Bevan N et al., (1999) FEBS Letters 462, 61-65.
Browning C et al., (2000) Br J Pharmacol 129, 42P.
Cohen FR et al., (1996) Br J Pharmacol 117, 1521-1529.