To investigate the structure and function of the mineralocorticosteroid receptor (MR), one has to circumvent the major difficulty related to its very low abundance. For this purpose, the full-length human MR (hMR) has been produced using the efficient baculovirus system. The recombinant protein is overexpressed in Sf9 insect cells at a concentration of almost-equal-to 2 pmol/mg of protein, which is 50-100 times more than the concentration in aldosterone target tissues. It binds aldosterone with high affinity (K(d) almost-equal-to 1 nM) and clearly displays a mineralocorticosteroid specificity as evidenced by competition studies with steroid ligands and by the monoclonal anti-idiotypic antibody H10E interacting with the steroid-binding domain of MR. After [S-35]methionine labeling, a single polypeptide band at almost-equal-to 120 kDa is detected and further identified as hMR by immunoblotting with A4, an anti-peptide antibody. Sedimentation analyses show that the native form of MR is recognized by A4 and B174, an antibody to the 90-kDa heat shock protein, since they both induce a shift of the receptor from 9S to 11S sedimentation coefficient. These results clearly demonstrate that MR is a heterooligomer containing the insect equivalent of the 90-kDa heat shock protein. This 9S receptor complex is converted to a 4S form during high-salt gradient ultracentrifugation, suggesting that MR can undergo a complete in vitro transformation. Immunofluorescence studies indicate that hMR, which is almost exclusively a cytoplasmic protein in Sf9 cells, is translocated to the nucleus after aldosterone exposure. Therefore, the recombinant hMR seems to behave as the native receptor. Given the possibility of large-scale protein production, the baculovirus system should prove useful in studies of the molecular basis of MR function as a transcription factor.