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

Since the regulated pathway-deficient variant could not store free GAG chains, this storage property (and release of free GAG chains after stimulation) in wild type cells must be because of the regulated pathway. In wild type AtT-20 cells, almost half of the [35S]GAG chains made in a 1-h pulse, remained stored in a releasable compartment at the end of a long chase (Fig. 3) . The storage of GAGs in the regulated secretory pathway was of surprisingly high efficiency, comparable to that found for ACTH, human growth hormone (Moore and Kelly, 1985) , trypsinogen (Burgess et al., 1985) , and insulin (Kelly et al., 1989) . Apparently, the GAG chains are not true bulk-phase markers as proposed by Burgess and Kelly (1984) . A tripeptide (N-octonoyI-ASN-TYR-THR-NH2) that behaves as a bulk-phase marker (Wieland et al., 1987) is excluded from the regulated secretory pathway in insulin-secreting cells (Gold et al., 1988) . GAGs may be sorted from the trans-Golgi network into the secretory granules either by a GAG-specific receptor, similar to the hormone-specific receptor (Chung et al., 1989) , or by association with a protein that is sorted. The latter mechanism has been invoked to explain sorting of an anti-secretogranin I antibody into secretory granules ). An alternative possibility is that in AtT-20 cells the machinery for synthesizing and sulfating GAG chains remain active in the immature granule after protein sorting events have occurred. Consistent with this alternative in our finding in AtT-20 cells that sulfated-labeled proteins and oligosaccharides appear to be sorted more efficiently than methionine or cysteine-labeled proteins (Matsuuchi, L., and R. B. Kelly, unpublished resuits) . Preferential secretion from parathyroid tissue of sulfate-labeled chromogranin A is also seen and was explained by proposing that sulfation occurs after the sorting of chromogranin into secretory granules (Gorr and Cohn, 1990) . In pulse-labeled PC12 cells, in contrast, no sulfation of immature or mature secretory granules was seen (Tooze and Huttner, 1990) . To resolve this point, similar experiments need to be done in wild type and variant AtT-20 cells.