Intermediates in the constitutive and regulated secretory pathways released in vitro from semi-intact cells

The selective sorting of secretogranin II into heterogeneous vesicles larger than the constitutive vesicles suggests that the large vesicles are precursors of regulated secretory granules . The putative immature secretory granules differ from constitutive vesicles in several ways . The addition of GTPyS to the in vitro incubation only slightly reduced the amount of secretogranin II-containing vesicles formed in vitro, and did not have a dramatic effect on their density (data not shown) . The release of immature secretory granules from the semi-intact cells thus appeared to be less sensitive to GTPyS than was the release of constitutive secretory vesicles. A second difference is that after in vivo chases of 15 and 60 min, release of secretogranin II-rich vesicles was largely ATP independent . Their velocity and equilibrium density characteristics were determined during the same pulse-chase experiment used to characterize constitutive vesicles . We found that after 15-and 60-min in vivo chases, permeabilized cells released secretogranin II-labeled vesicles without need of an in vitro incubation (Fig . 7 , B and C, no reaction) . With or without an in vitro incubation, the released vesicles had a sedimentation velocity that increased with time of chase (Fig . 7 A-C) . The immature secretory granules released from the cells also increased in density with time of chase (Fig . 7 D-F and Fig . 8 A) . To determine the density of mature secretory vesicles, cells were labeled with [35 S]sulfate for 12 h and chased for an additional 12 h . The secretogranin II-containing vesicles were associated with the cell ghosts (PIM) and had an average density of 1 .20 g/ml (Fig . 8 A) . Thus, the secretory granule precursors that are released by semi-intact cells did not reach the density of mature granules by 60 min . Despite the differences between constitutive vesicles and immature secretory granules, the radioactivity recovered in both vesicle types peaked at 15 min of chase . Because ofthe major difference in their sedimentation velocities, it has been possible to resolve experimentally constitutive vesicles from immature secretory granules and compare their properties. Constitutive vesicles were rich in sulfated proteoglycans and not detectable unless permeabilized cells are incubated with ATP Their size, density, and ratio of secretogranin II to proteoglycans remained constant, independent of chase time (Fig. 8) . The sulfate-labeled Golgi membranes from which they may arise do not leak out of the permeabilized cells. The density ofthe Golgi peak fractions (1.178 t 0.002 g/ml) and the ratio ofsecretogranin II to proteoglycans (0.08 f 0.01) also remained constant during the chase. Immature secretory granules containing secretogranin II only showed ATP-dependent release immediately after pulse 60 min chase constitutive labeling. As the chase time increased, any ATP dependent release was presumably concealed by the release of preexisting immature secretory granules from cytoplasmic pools . The density of the immature granules was less than that of the Golgi membranes immediately after the pulse (Fig. 8 A) . The immature secretory granule fractions (4-12, Fig . 7 , A-C) contained 55-85 % of the released "S-secretogranin II, but also a minor fraction (30-40 %) ofthe released 'SS-proteoglycans . The ratio of secretogranin II to proteoglycans (0.05 t 0.02) immediately after the pulse was not significantly different from that of the Golgi fractions . Unlike t