Expression and characterization of the long and short splice variants of GS alpha in S49 cyc- cells.
The alpha subunit of the guanine nucleotide-binding regulatory protein GS mediates stimulation of adenylyl cyclase activity. This subunit, GS alpha, exists as two molecular weight forms, termed long and short, that differ by 14 or 15 amino acids. A physiological distinction between these two forms has yet to be defined. To compare the activities of these GS alpha isoforms, long and short forms of rat GS alpha were expressed in the cyc- variant of S49 murine lymphoma cells, which is deficient in endogenous GS alpha expression. By immunoblot analysis, the level of recombinant proteins in the clones expressing the long form of GS alpha was about twice that present in the clones expressing the short form of GS alpha or in the S49 wild-type cells. Both recombinant GS alpha proteins were sensitive to cholera toxin-catalyzed ADP-ribosylation, although the short form was labeled preferentially in both recombinant and S49 wild-type cell lines. In whole-cell assays, the clones expressing the long and short forms of GS alpha and the S49 wild-type cells gave comparable responses for stimulation of cAMP accumulation after challenge with (-)-isoproterenol, cholera toxin, or forskolin. In adenylyl cyclase assays with partially purified membranes, clones expressing the long form of GS alpha gave approximately twice the levels of cAMP in response to isoproterenol, guanosine-5'-O-(3-thio)triphosphate, NaF, or forskolin, compared with membranes from the clones expressing the short form of GS alpha or the S49 wild-type cells. However, when maximal adenylyl cyclase activity was normalized to the level of GS alpha protein in S49 wild-type cells, the cAMP productions were similar between all of the cell lines. In other membrane-based assays, the long and short forms of GS alpha were also equivalent in their dose response to isoproterenol and GTP, their kinetics of guanine nucleotide exchange and GTPase activity, and the induced high and low affinity states of the beta-adrenergic receptor in response to isoproterenol. In the latter radioligand binding analysis, membranes from the two clones expressing the long form of GS alpha consistently gave a greater proportion of the agonist high affinity state; however, this variation likely reflects the greater expression levels of GS alpha in these membranes. Thus, we conclude that the long and short forms of GS alpha expressed in S49 cyc- cells are very similar in their ability to stimulate adenylyl cyclase activity and to couple to beta-adrenergic receptors.