Docking of Chromaffin Granules--A Necessary Step in Exocytosis? Theo Seh/ifer, Urs O. Karli, Felix E. Sehweizer and Max M. Burger Received April 9, 1987
KEY WORDS: chromaffingranules;docking;exocytosis. Putative docking of secretory vesicles comprising recognition of and attachment to future fusion sites in the plasma membrane has been investigated in chromaffin cells of the bovine adrenal medulla and in rat phaeochromocytoma (PC 12) cells. Upon permeabilization with digitonin, secretion can be stimulated in both cell types by indreasing the free Ca z +-concentration to #M levels. Secretory activity can be elicited up to 1 hr after starting permeabilization and despite the loss of soluble cytoplasmic components indicating a stable attachment of granules to the plasma membrane awaiting the trigger for fusion. Docked granules can be observed in the electron microscope in permeabilized PC 12 cells which contain a large proportion of their granules aligned underneath the plasma membrane. The population of putatively docked granules in chromaffin cells cannot be as readily discerned due to the dispersal of granules throughout the cytoplasm. Further experiments comparing PC 12 and chromaffin cells suggest that active docking but not transport of granules can still be performed by permeabilized cells in the presence of Ca 2 + : a short (2 min) pulse of Ca z + in PC 12 cells leads to the secretion of almost all releasable hormone over a 15 rain observation period whereas, in chromaffin cells, with only a small proportion of granules docked, withdrawal of Ca 2+ leads to an immediate halt in secretion. Transport of chromaffin granules from the Golgi to the plasma membrane docking sites seems to depend on a mechanism sensitive to permeabilization. This is shown by the difference in the amount of hormone released from the two permeabilized cell types, reflecting the contrast in the proportion of granules docked to the plasma membrane in PC 12 or chromaffin cells. Neither docking nor the docked state are influenced by cytochalasine B or colchicine. The permeabilized cell system is a valuable technique for the in vitro study of interaction between secretory vesicles and their target membrane. Dept. of Biochemistry,Biocenterof the University,Klingelbergstr.70, CH-4056, Basel.
The process of stimulated secretion--in contrast to the constitutive secretory pathway--crucially depends on the availability of secretory vesicles at their fusion sites whenever the cell can be stimulated. This availability might be guaranteed by constantly supplying the cell periphery with secretory vesicles and by docking them in a stable manner to the plasma membrane ready to fuse upon stimulation of the cell. The mechanism of docking and the molecules holding vesicles in such a docked position are not yet known. Docked secretory vesicles have best been described ultrastructurally in the unicellular ciliate Paramaecium by Plattner and his group (e.g. see Plattner, 1986). Secretory vesicles in this organism have a very distinct morphology. They are asymmetric, consisting of a body with a tip, which fits into preformed, easily recognizable docking sites. These sites, regularly distributed over the cell surface, are also the future fusion sites. They are characterized morphologically by intramembrane particles arranged as a rosette around the tip of the attached/docked secretory vesicle. These particles diffuse laterally during fusion and rearrange into the rosette upon docking of the next vesicle. The tip is connected to the plasma membrane by an electron-dense structure composed of proteinaceus material including an ATPase activity (Plattner et al., 1977). In addition, it fits into preformed pores in a meshwork composed of microfilaments aligned parallel to the plasma membrane. This extreme ultrastructural specialization of a docking site has never been described in higher cells. Stable docking of vesicles, although in an ultrastructurally less prominent way, has to be expected in other biological systems showing stimulated exocytosis: cortical granules attached to the plasma membrane and ready to fuse upon addition of Ca 2 § can easily be prepared from the eggs of sea urchins (Detering et al., 1977). At the other end of animal development, in the highly specialized neuromuscular junction, docking of neurotransmitter-containing vesicles can be observed at the presynaptic membrane in a regular array over the active zones (Heuser and Reese, 1981). Can it be assumed from these examples that docking of vesicles represents a crucial step in every stimulated secretory process? Cells secreting protein or peptide molecules have to transport secretory vesicles, after packaging of the contents, from the Golgi to the plasma membrane or to specialized domains of the plasma membrane. They are first confronted with the problem of sorting different vesicles destined to reach different subcellular locations. The mechanisms underlying this sorting process as well as the transport have not been elucidated. The second problem to be solved lies in the recognition of the correct target structure after transport. The recognition signal might be coupled to the structures responsible for positioning the transport/secretory vesicle at or near the plasma membrane. These structures might in addition also include components that can modify vesicles to make them ready for fusion with the plasma membrane, i.e. to render them fusion-competent.
Docking of Chromaffin Granules
C H R O M A F F I N AND PC 12 CELLS The chromaffin cells of the adrenal medulla are amongst the best studied endocrine gland cells in terms of hormone synthesis, hormone storage and response to a variety of stimulating agents (for review, see Winkler and Westhead, 1980; Trifaro, 1982; Burgoyne, 1984; Baker and Knight, 1984; Carmichael, 1986). At the morphological level, one aspect hinting at the presence of docked vesicle, namely the presence of electron dense material between plasma membrane and secretory vesicle (the chromaffin granule) has been reported by Aunis et al. (1979) and by Burgoyne (1984). A specialized docking site represented by a characteristic arrangement of intramembrane particles similar to Paramaecium has not been found in these cells (Schmidt et al., 1983). A good indication for granules being docked comes from the ultrastructural localization of the granules in an established cell line originating from a rat phaeochromocytoma, PC 12, isolated by Greene and Tischler (1976). These cells contain fewer granules than the primary cells but most of them are aligned just underneath the plasma membrane (see Fig. la). Due to the larger size of the granules in the primary cells and therefore the smaller chance to section the contact point of granule and plasma membrane, it is difficult to determine on thin sections if there might exist a most peripheral row of docked granules, e.g. attached to the plasma membrane or held at a typical small distance from the plasma membrane (see Fig. lb). In addition, it has so far not been determined if granules can fuse with the plasma
Fig. 1. (a) Electronmicrographof rat phaeochromocytomacell (PC 12).Small dense-core granules are seen distributed preferentiallyjust underneath the plasma membranein a putativelydockedstate awaitingthe signal for fusion. (b) Electronmicrograph of primary bovinechromaffincell culturedfor less than a week. Dense-corechromaftingranulesare seendistributed throughout the cell. Peripheral row of potentially docked granules cannot be identified by such ultrastructural observation. Bars represent 0.5 #m.
Schiifer et al.
membrane only at predetermined microdomains or generally into the whole cell membrane. Secretion preceded by docking in a certain region may only occur in the cell in its physiological surrounding, but might get lost upon isolation and culturing of the cell. RESULTS F R O M DIGITONIN-PERMEABILIZED CELLS As the putative docking step is part of an intracellular process, a cell-flee system for reconstruction of the process in vitro and the dissection of components involved would be very helpful in understanding exocytosis. Chromaffin granules can indeed be obtained easily in high purity and large amounts, but the plasma membrane in the desired inside-out configuration has so far not been isolated. Thus, an in vitro model system for studying the interaction of chromaffin granules with the plasma membrane is unfortunately not available. A step in this direction, however, was the introduction of the plasma membrane permeabilization technique into the field of chromaffin cell biology by Baker and Knight (1978), Dunn and Holz (1983) and Brooks and Treml (1983). Using this approach the intracellular compartment becomes accessible by dialysis against the extracellular medium having the composition of interest. The advantage of the use of digitonin (Dunn and Holz, 1983) over the cell permeabilized cells (Sch~ifer et aI., 198'7). These large pores concomitantly are exposure to pore-forming toxins (Ahnert-Hilger et at., 1985; Baker et at., 1986) lies in the formation of larger pores allowing even the introduction of antibodies into such permeabilized cells (Sch~ifer et al., submitted). These large pores concomitantly are responsible for the leakage of the soluble cytoplasmic components of at least 134 kDa molecular weight (lactate dehydrognase) as shown by Holz and Senter (1985). Interestingly exocytosis in this system turned out to depend on MgATP and free Ca z § in #M concentrations only. Essentially the same result was obtained for exocytosis in permeabilized PC 12 cells (Peppers and Holz, 1986). This stringent requirement for MgATP and Ca z§ at physiologically relevant concentrations together with the insensitivity towards the loss of substantial parts of soluble cytoplasmic components make this approach attractive for the investigation of docking in these cells. In our studies we generally permeabilized cells with 20 pg/ml (chromaffin cells) or 10 #g/ml (PC 12 cells) of digitonin at room temperature for 10 min. In contrast to prior studies we washed off excess digitonin for increasing time intervals and only then stimulated the secretion of preloaded 3H-NA by raising the free Ca 2 § to 10 #M. As shown in Fig. 2 PC12 cells respond to Ca 2+ after wash periods of up to 30min, whereas stimulation during permeabilization always resulted in unsatisfactorally high background levels. Two observations were of interest for the question of docking of granules. First, up to 1 h after starting the permeabilization protocol granules were still seen located tightly at the plasma membrane of permeabilized by unstimulated cells, as shown for PC 12 in Fig. 3. Secondly, even after permeabilization and extended wash periods in chromaffin and PC 12 cells elevation of free Ca 2 + levels elicited exocytotic activity. Release of preloaded 3H-NA was indeed due to exocytosis as proven by the immunoelectron microscopical detection of ehromaffin granule membrane antigen incorporated into the plasma membrane of stimulated cells (Sch~ifer et al., 1987).
Fig. 2. Release of preloaded 3H-norepinephrine (3H-NA) from digitoninpermeabilized PC 12 cells. Cells were permeabilized with 10 yg/ml of digitonin in a Ca2+-free high-potassium buffer (Baker and Knight, 1978) for 10min. When cells were stimulated with 10 #M free Ca 2 § during this step a high Ca 2 § independent background was consistently measured, probably due to damage of cells. When excess digitonin together with damaged cells was washed off for 10 or 30rain before stimulation for 15 rain, a Ca2+-dependent release was observed well above background after the prolonged wash period. Results are presented as mean_+ standard error mean. Measurements were performed in quadruplicate for this and all following figures.
Fig. 3. Electron micrograph of PC 12 cell permeabilized with 10#g/ml of digitonin for 10 rain and washed for 45 min in Ca: § buffer before fixation. Dense-core granules ready to fuse can still be observed docked at the permeabilized plasma membrane. Bar represents 0.5 #m.
Schiifer et al. a 50" ~.0" Z I -r
m 30"6 i I I
.d .d J J
b 30 84
/A ~" .4
Fig. 4. Influenceofcytochalasine B (a) and colchicine (b) on releaseof 3H-NA from permeabilizedPC 12cells. Indicated concentrations of drugs wereapplied during 10 rain permeabilization, 10 rain wash and 15 rain stimulation periods. Drugs have no inhibitory effect on secretion indicating that neither microfilamentsnor microtubules are responsiblefor the attachment of granules at docking sites.
Neither cytochalasine B nor colchicine had an influence on the release (Fig. 4a and b), implying that granules are not held in a docked position either by microfilaments or by microtubules. As permeabilization and wash steps were performed in the presence of 5 m M E G T A these results also suggest that this detergent-resistant docked state is neither dependent on Ca 2 + as such nor on the presence of proteins known to bind to membranes in the presence of Ca 2+ (Geisow e t al., 1986). C a / + might thus only be involved in the production of a "fusogen" triggering the fusion of firmly docked granules with the plasma membrane. The nature of the "fusogen" is not known but experiments performed by Frye and Holz (1985) showed that inhibition of phospholipase A 2 activity efficiently blocked secretion. Inhibition of secretion by
Docking of Chromaffin Granules
~ ]-Ca 2§ ~
Fln m _
Fig. 5. Influence of para-bromophenacylbromide on release of 3HNA f r o m permeabilized PC 12 cells. Indicated concentrations of drug were present during 10min permeabilization, 10 min wash, and 15 rain stimulation periods. Ca2+-dependent secretion can be totally inhibited indicating a crucial role of Ca2+-activated phospholipase A2 in the production of "fusogen".
z< 2O cr~
/ time (rain) Fig. 6. Time-course of Ca2+-dependent release of SH-NA from permeabilized PC 12 cells. Cells were permeabilized with digitonin for 10 min and washed in Ca2+-free buffer for another 10 rain. They were then stimulated with 10 #M Ca 2+ for indicated time periods.
p h o s p h o l i p a s e A 2 blockers m e p a c r i n e or p a r a - b r o m o p h e r a c y l b r o m i d e can also be observed in P C 12 cells (Fig. 5). Are there a d d i t i o n a l C a 2 + - d e p e n d e n t steps involved in the secretory process, possibly in the d o c k i n g step? W e have tried to a p p r o a c h this q u e s t i o n by the two following experiments.
Sch~ifer e t al. a
25" < 7
15 / / /
- C a 2~
/ / /
~10 ./ /
time (mini Fig. 7. Triggering of release of 3H-NA from permeabilized cells by a 2min Ca2+-pulse. PC 12 (a) and chromaffin cells (b) were permeabilized and washed as before. Buffer containing 10 pM Ca 2§ was added for 2 min and then exchanged for Ca 2§ or Ca 2+containing buffer as indicated in the figure. Broken lines denote the increase in background levels to be expected from the controls in Ca2+-free buffer. The difference between these background levels and the release measured after the 13 min period (arrows) represents the Ca2+-triggered release insensitive to withdrawal of Ca 2+ . Histograms on right side show 3H-NA released in a 15 min period without buffer change.
U p o n elevation of the free C a 2 +, secretion of 3 H - N A starts r a p i d l y a n d reaches a p l a t e a u after a b o u t 20 rain (Fig. 6). C o n s e q u e n t l y , we e x a m i n e d w h e t h e r a s h o r t initial presence (2 rain) of 10 # M free C a 2 + w o u l d be sufficient t o p r o d u c e the full r e s p o n s e seen before with C a 2§ being present d u r i n g the w h o l e 15 m i n s t i m u l a t i o n p e r i o d . Astonishingly, as p r e s e n t e d in Fig. 7a a n d b the result differed in P C 12 a n d in chromaffin cells. I n P C 12 cells the triggering of exocytosis for 2 m i n by C a 2 § followed b y a 1 3 m i n p e r i o d in Ca2+-free buffer resulted in an a l m o s t full response.
Docking of Chromaffin Granules
Fig. 8. Double stimulation experiments using permeabilized PC 12 cells. (a) The standard protocol, including a permeabilization (P) and wash (W) step in Ca 2+-free buffer ( - ) followed by a 15 min stimulation (St) with 10#M Ca 2+ (+), was extended in two different experiments: (b) First stimulation was performed during the 10min permeabilization step leading to a Ca a +-dependent release but also a high background, probably due to damage of cells. After a wash step in Ca z +-free buffer cells were stimulated a second time for 15min resulting in a second Ca 2+dependent release. (c) Cells were permeabilized for 10 min in Ca2+-freebuffer and only then stimulated the first time with 10 #M Ca ~+ for 10 rain. After a wash step in Ca 2+-free buffer cells were stimulated a second time for 15 min. In contrast to the results in (b) no second Ca 2+-dependent release could be elicited.
30 _-rm 20
~5 0 -~ i_
Contrastingly, in chromaffin cells the triggering for 2 min did not allow exocytosis to proceed further u p o n the removal of Ca z + during the second period. In view of the difference in the distribution of the granules--aligned along the plasma m e m b r a n e in PC 12 and mostly scattered t h r o u g h o u t the cytoplasm in chromaffin cells--one might conclude that Ca 2 + is necessary for docking of granules. This would only show up under these conditions in chromaffin cells, because all releasable granules seem to be docked already in P C 12 cells. F r o m the P C 12 experiment it is evident that a short trigger with Ca 2 + is sufficient to produce a "fusogen" active over a 15 min period. The limiting factor for secretion in the permeabilized cells would therefore be the availability of docked granules. C a n granules still be supplied to the putative plasma m e m b r a n e docking sites in permeabilized cells? We tried to answer this question by double stimulation experiments. Firstly, cells were permeabilized for 10 min in the presence of 10 # M Ca 2 + resulting in a Ca z +-dependent release of a H - N A (Fig. 8b, 9b). Cells were then washed in Ca 2 + -free buffer for 15 min and stimulated a second time with 10 # M Ca 2 + for 10min, causing another CaZ+-dependent release. D u r i n g permeabilization/ stimulation and the following wash a second population of granules could apparently be transported and docked to the plasma membrane. If, however, the experimental protocol was changed to a 10 rain permeabilization step prior to the first stimulation for 10 rain, a wash period again of 15 rain followed by a second stimulation for 10 min, the cells could not respond to the second stimulation a n y m o r e (Fig. 8c, 9c). Thus the mechanism responsible for the delivery of a second population of granules to the plasma m e m b r a n e seems to get lost with time after permeabilization, whereas the fusion of the first population of granules docked or ready to be docked seems to be
E~] -Co 2+ ] +Ca2 +
is st is
20 10 O
Fig. 9. Double stimulation experiments using permeabilizedchromaffincells.Experimentswere performed as describedin legendto Fig. 8.
insensitive to permeabilization and the loss of cytoplasmic components. What were the components lost which were involved in transport and docking of granules? Interestingly again the drugs cytochalasine B and colchicine had no influence on the second stimulation of the cells. In summary this experimental system offers the possibility to investigate various steps involved in the process of secretion. Hopefully, it will allow the dissection of factors responsible for steps as different as transport, recognition, docking and fusion of granules. The determination of the respective Ca 2 § and/or MgATP-requirements of the different steps, as shown in one example, might eventually also allow the elucidation of the regulation of a complex process like secretion. In particular the characterization of recognition and docking in terms of molecular structures involved might also give important insights into other recognition-interaction-fusion events operating in many different cellular processes, e.g. endocytosis, transfer of vesicles from ER to Golgi, transfer of vesicles from one Golgi stack to the next, and others involving the specific interaction of intracellular membrane components.
ACKNOWLEDGEMENTS We wish to thank our colleague Dr Nick Sargent for helpful discussion and Ms Lilly Schwoerer for the preparation of the manuscript. This work was supported by the Swiss National Foundation (Grant 3.269-0.82 and 3.169-0.85).
Docking of Chromaffin Granules
REFERENCES Ahnert-Hilger, G., Bhakdi, S. and Gratzl, M. (1985). Minimal requirements for exocytosis. J. Biol. Chem. 260:12730-12734. Atlnis, D., Hesketh, J. E. and Devilliers, G. (1979). Freeze-fracture study of the chromaffin cell during exocytosis: Evidence for connections between the plasma membrane and secretory granules and for movements of plasma membrane-associated particles. Cell Tissue Res. 197:433-441. Bader, M. F., Thierse, D., Aunis, D., Ahnert -Hilger, G. and Gratzl, M. (1986). Characterization of hormone and protein reIease from a-toxin-permeabilized chromaffin cells in primary culture. J. Biol. Chem. 261:5777-5783. Baker, P. F. and Knight, D. E. (1978). Calcium-dependent exocytosis in bovine adrenal medullary cells with leaky plasma membranes. Nature 276:620-622. Baker, P. F. and Knight, D. E. (1984). Calcium control of exocytosis in bovine adrenal medullary cells. Trends Neurosci. 7:120-126. Brooks, J. C. and Treml, S. (1983). Catecholamine secretion by chemically skinned cultured chromaffin cells. J. Neurochem. 40:468-473. Burgoyne, R. D. (1984). Mechanisms of secretion from adrenal chromaffin cells. Bioehim. Biophys. Aaa 779:201-216. Carmichael, S. W. (1986). The adrenal medulla, Volume 4. Cambridge University Press, Cambridge. Detering, N. K., Decker, G. L., Schmell, E. D. and Lennarz, W. J. (1977). Isolation and characterization of plasma membrane-associated cortical granules from sea urchin eggs. J. Cell Biol. 75:899-914. Dunn, L. A. and Holz, R. W. (1983). Catecholamine secretion from digitonin-treated adrenal medullary chromaffin cells. 258:4989-4993. Frye, R. A. and Holz, R. W. (1985). Arachidonic acid retease and catechotamine secretion from digitonintreated chromaffin cells: Effects of micromolar calcium, phorbol ester, and protein alkylating agents. J. Neurochem. 44:265-273. Geisow, M. J., Fritsche, U., Hexham, J. M., Dash, B. and Johnson, T. (1986). A consensus amino-acid sequence repeat in Torpedo and mammalian Ca 2 +-dependent membrane-binding proteins. Nature 320:636-638. Greene, L. A. and Tischler, A. S. (1976). Establishment of a noradrenergic clonal line of rat adrenal phaeochromocytoma cells which respond to nerve growth factor. Proc. Natl. Acad. Sci. USA 73:24242428. Heuser, J. E. and Reese, T. S. (1981). Structural changes after transmitter release at the frog neuromuscular junction. J. Cell Biol. 88:564-580. Holz, R. W. and Senter, R. A. (1985). Plasma membrane and chromaffin granule characteristics in digitonintreated chromaffin cells. J. Neurochem. 45:1548--1557. Peppers, S. C. and Holz, R. W. (1986). Catecholamine secretion from digitonin-treated PC 12 cells. J. Biol. Chem. 261 : 14665-14669. Plattner, H. (1986). Synchronous exocytosis in Paramaeeium cells. In: Ceil Fusion (A. E. Sowers, Ed.), Plenum Press, New York. Plattner, H., Reichel, K. and Matt, H. (1977). Bivalent-cation-stimulated ATPase activity at preformed exocytosis sites in Paramaeeium coincides with membrane-intercalated particle aggregates. Nature 267: 702-704. Sch~ifer, T., Karli, U. O., Gratwohl, E. K.-M., Schweizer, F. E. and Burger, M. M. (1987). Digitoninpermeabilized cells are exocytosis-competent. J. Neurochem., in press. Schmidt, W., Patzak, A., Lingg, G., Winkler, H. and Plattner, H. (1983). Membrane events in adrenal chromaffin cells during exocytosis: a freeze-etching analysis after rapid cryofixation. Eur. J. Cell Biol. 32:31-37. Trffaro, J. M. (1982). The cultured chromaffin cell: a model for the study of biology and pharmacology of paraneurones. Trends Pharmacol. Sci. 2:389-392. Winkler, H. and Westhead, E. (1980). The molecular organization of adrenal chromaffin granules. Neuroscience 5:1803-1823.