The hypothesis that prolonged treatment of a constitutively active receptor with inverse agonists may lead to increased receptor density was tested for the alpha(2)-adrenoceptor (AR) inverse agonist (+)-RX 811059 at both the wild-type (WT) and Thr(373)Lys alpha(2A) ARs in CHO-K1 cells by monitoring [(3)H]RX 821002 and [(35)S]GTPgammaS binding responses. One-hundred micromolar KCl instead of NaCl in the [(35)S]GTPgammaS membrane binding assay favoured the detection of a high-magnitude constitutive alpha(2A) AR activity. Under this condition, (+)-RX 811059 was an inverse agonist [ E(max) (% vs. basal): Thr(373)Lys alpha(2A) AR (-52+/-2) > WT alpha(2A) AR (-31+/-6)] while atipamezole was a silent neutral antagonist for both WT and Thr(373)Lys alpha(2A) ARs. The B(max) value of [(3)H]RX 821002 binding sites to membranes of transfected CHO-K1 cells was <90% for the Thr(373)Lys alpha(2A) AR compared with the WT alpha(2A) AR (9.1+/-1.4 pmol/mg protein); K(d) values were similar (1.16+/-0.19 nM and 1.51+/-0.15 nM, respectively). Forty-eight-hours' pre-treatment of cells with either 0.1 microM (+)-RX 811059, 1 microM atipamezole or 1 microM of the efficacious agonist d-medetomidine increased the amount of [(3)H]RX 821002 binding sites of both WT (52%-59%) and mutant (306%-447%) Thr(373)Lys alpha(2A) ARs. The same alpha(2) AR ligands also prevented the loss of [(3)H]RX 821002 binding sites as induced by incubation of transfected CHO-K1 cellular membranes at 37 degrees C for 4 h (WT alpha(2A) AR) and 2 h (Thr(373)Lys alpha(2A) AR); 0.1 microM (+)-RX 811059 and 1 microM atipamezole caused an increase compared with the control amount of [(3)H]RX 821002 binding sites to the Thr(373)Lys alpha(2A) AR by 73% and 50%, respectively. In conclusion, no relationship was found between inverse agonism and alpha(2A) AR up-regulation. It is suggested that this is due to structural stabilisation of the alpha(2A) AR, irrespective of the nature of the ligand.