Constitutive and basal secretion from the endocrine cell line, AtT-20

([]) with cAMP. lated pathway, wild type AtT-20 ceils were labeled with a 1-hpulse of [35S]sulfate and chased for varying lengths of time. The [35S]GAGs were identified after gel electrophoresis as the ladder of radioactive material that migrated between the protein molecular weights of 2.3-14 kD (Fig. 2 B) . In wild type cells, 40% of the [3sS]GAGs were secreted into the medium with a half time of ,x,30 min, while the remainder of the GAGs was stored intracellularly (Fig. 3 A) . After a chase period, treatment with 8-Br-cAMP increased the amount of GAG chains released into the media twofold, suggesting that GAGs were indeed stored in the regulated secretory pathway (Fig. 3 C) . In contrast, variant cells released 98 % of their [35S]GAG chains, with a halftime also of • 30 min (Fig. 3 B) . No stimulation of secretion of the remaining 2 % of the GAG chains could be seen after the addition of 8-Br-cAMP (Fig. 3 D) . Longer incubation times of cells in xyloside and [35S]sulfate more extensively labeled the stored pools of [35S]GAG chains in wild type cells, since they have a half-time longer than 1 hour. No evidence of storage of GAG chains or their regulated release could be detected in the variant cells even under these conditions. Wild type AtT-20 cells were over twenty times more efficient at storing [35S]GAGs than the variant and the GAGs were stored as efficiently as trypsinogen, ACTH, and insulin. Either the variant did not have secretory granules into which GAGs may flow, or the GAGs could not enter the granules because of missing components of the sorting machinery. The apparent efficiency of sorting of the GAG chains into the regulated pathway is not consistent with the earlier assumption (Burgess and Kelly, 1984) that they can be used as a bulk phase marker.