Newly synthesized Na,K-ATPase alpha-subunit has no cytosolic intermediate in MDCK cells.

Recently published data indicate that the alpha-subunit of Na,K-ATPase, a transmembrane protein of animal cell plasma membranes, is synthesized as a soluble precursor. In the present experiments we demonstrate that an apparent "soluble" form can indeed be detected in crude cytosolic fractions prepared by centrifugation from MDCK cells disrupted by sonication. We find, however, that this form has no precursor-product relationship with membrane-associated alpha-subunit. The quantity of unsedimentable alpha-subunit can be greatly diminished by increasing the centrifugal field employed to remove membranous vesicles from the cytosol fraction. Sonication of membrane pellets generates alpha-subunit which, like the "soluble" form, resists pelleting. Finally, cytosol fractions prepared from cells disrupted by sonication contain membrane-bound vesicles which can be immunoadsorbed on Staphylococcus aureus cells coated with a monoclonal antibody directed against alpha-subunit. We find, therefore, that the previously observed soluble precursor of alpha-subunit is actually a component of small unpelleted membrane vesicles generated by harsh disruption conditions. When cells are disrupted by less violent techniques no newly synthesized alpha-subunit can be detected in the cytosol fraction. We calculate that to escape detection under our experimental conditions a bona fide soluble precursor of alpha-subunit must have a cytosolic t1/2 less than 20 s. We conclude that the alpha-subunit is most probably inserted into the bilayer cotranslationally.