Stimulated secretion of lysosomal enzymes by cells in culture.

F9 mouse teratocarcinoma and PyS-2 cells in culture incubated with monovalent cations in buffered sucrose solution (0.25 M) can secrete as much as 40% of their total lysosomal enzymes into the medium within 30 min. Longer incubation does not lead to further loss of enzyme, suggesting that only a certain fraction of lysosomes is capable of discharge. The simultaneous presence of sucrose and cation, each at the respective optimal concentrations of 0.25 and 0.15 M, is required for lysosomal discharge (i.e. twice isoosmolarity). The cells remain fully viable. Sodium ions are more effective than lithium and potassium ions, whereas amines and divalent cations are less effective. Other sugars including glucose can replace sucrose to varying extents. Secretion is accompanied by a rapid short-lived rise in the level of cAMP. Forskolin as well as agents that activate G protein such as cholera toxin, AlF4-, and vanadate ions also increase the rate of secretion. Sucrose-Na+ stimulation takes place independently of changes in influx or efflux of calcium ions or changes in the levels of extracellular or free intracellular calcium ions. Neomycin, an inhibitor of phospholipase C, has little effect on secretion. Our results suggest that the secretion observed is mediated by a cAMP-dependent mechanism involving G proteins. Calcium ions and phospholipase C appear to play little or no part in the activation process.


Stimulated Secretion of Lysosomal Enzymes by Cells in Culture"
(Received for publication, August 10, 1988)

Leonard Warren
From the Wistar Institute of Anatomy and Biology, Philadelphia, Pennsylvania 19104 F9 mouse teratocarcinoma and PyS-2 cells in culture incubated with monovalent cations in buffered sucrose solution (0.25 M) can secrete as much as 40% of their total lysosomal enzymes into the medium within 30 min. Longer incubation does not lead to further loss of enzyme, suggesting that only a certain fraction of lysosomes is capable of discharge. The simultaneous presence of sucrose and cation, each at the respective optimal concentrations of 0.25 and 0.15 M, is required for lysosomal discharge (Le. twice isoosmolarity). The cells remain fully viable. Sodium ions are more effective than lithium and potassium ions, whereas amines and divalent cations are less effective. Other sugars including glucose can replace sucrose to varying extents.
Secretion is accompanied by a rapid short-lived rise in the level of CAMP. Forskolin as well as agents that activate G protein such as cholera toxin, AlF;, and vanadate ions also increase the rate of secretion. Sucrose-Na+ stimulation takes place independently of changes in influx or efflux of calcium ions or changes in the levels of extracellular or free intracellular calcium ions. Neomycin, an inhibitor of phospholipase C, has little effect on secretion. Our results suggest that the secretion observed is mediated by a CAMP-dependent mechanism involving G proteins. Calcium ions and phospholipase C appear to play little or no part in the activation process.
Lysosomal enzymes, found both intracellularly and extracellularly, are capable of degrading lipids, carbohydrates, proteins, and nucleic acids. These enzymes play an important role in normal metabolic turnover, in the degradation of hormones and other molecules, and in defense against infection.
Secretion is generally quite slow in culture, especially as cells reach confluency, and only a small percentage of the total lysosomal enzyme content is lost from the cell each day (1). The rate of enzyme release from the cell can be increased moderately by a variety of agents, including weak bases such as aliphatic amines and NH,Cl, the extent of increase depending on the concentration of the agent and on the cell type (1)(2)(3)(4)(5). The agents appear to exert their effect after being concentrated in the lysosomes (6) and prelysosomal compartments where they elevate the pH, enhance binding of hydrolases to receptors, inhibit receptor utilization, and thereby disrupt the intracellular pathway for transport of acid hydrolases. Newly synthesized enzymes may be directly secreted * This work was supported by Grant CA-19130 from the United States Public Health Service and the American Cancer Society Grant RDP-ISH. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked ''advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. (7). Sucrose stimulates secretfon in organ culture of bone and cartilage (8), and sucrose together with NH, ions stimulate secretion in human fibroblasts (9). Release of lysosomal enzymes can also be effected by exposure of cells to immune complexes (10-lZ), zymosan, small fiber asbestos (13), and dextran sulfate (14).
In the present study it is shown that inorganic monovalent cations play an essential role in the secretory process since cells in culture, not specialized to secrete, bathed in 0.25 M sucrose solution and exposed to inorganic cations, such as Na+, can secrete more than a quarter of their lysosomal enzymes within 30 min without loss of viability. This rapid rate of secretion has facilitated the study of factors involved in the process and is somewhat analogous to the explosive exocytosis of granules from mast cells (15) and cortical granules from sea urchin eggs (16). The complications of reuptake, synthesis, and degradation of lysosomal enzymes in studies lasting hours and days have been eliminated. By using this system, it has been shown that the secretion of lysosomal enzymes is dependent on cAMP and G protein function and is unaffected by shifts in concentration of intra-and extracellular calcium ions or on the activity of phospholipase C. While the underlying stimulatory process of the sucrose-Na+ procedure is not fully understood, it provides a convenient method to study a secretory mechanism, applicable to many kinds of normal and pathological cells.
Standard Assay-5 x lo5 cells in 3 ml of medium were grown in 35-mm plastic dishes (Costar) for 48 h at which time the cell number had increased to 2-3 X lo6. All experiments were carried out in duplicate and in some cases in triplicate. Each plate was washed three times with 3 ml of TM Sucrose (10 mM Tris chloride buffer, pH 7.1, 1 mM MgCl,, 0.25 M sucrose) and drained, and 0.5 ml of incubation mixture (TM Sucrose or TM Sucrose, 0.15 M NaCl with or without special factors) was added to each plate. Plates were incubated for 30 min at 37 "C on a rocker platform (Belco Glass Co., Vineland, NJ) 8835 moving at 8 cycles/ min. The bathing fluids were transferred to 1.5ml plastic tubes with pointed bottoms and centrifuged for 3 min in an Eppendorf centrifuge (13,500 X g), and the supernatants were removed. In order to determine the total enzyme activity of the system, cells were scraped off the dishes into 0.5 ml of TM Sucrose. Cells were transferred to plastic tubes containing pellets from centrifugation of the incubation medium. Plates were then washed with 0.25 ml of TM Sucrose and the second harvest of cells was transferred to the first. The cell suspension was homogenized in the plastic tube with 30 vigorous strokes using a pointed Teflon pestle.
In experiments with hybridomas or human granulocytes, cells were suspended in 0.3 ml of TM Sucrose with or without 0.15 M NaCI. After incubation, tubes were centrifuged for 5 min at 1000 X g, the supernatants were withdrawn, and 0.3 ml of TM Sucrose was added to the cell pellet, Cells were subjected to freezing and thawing three times and sonication for 30 seconds in a Branson Ultrasonic Cleaner.
Assay for Acid Hydrolases-The homogenate and supernatant were assayed for N-acetyl-P-D-glucosaminidase and other acid hydrolases using a para-nitrophenyl glycoside substrate. The assay vessel contained 20 mmol sodium citrate buffer, pH 4.8,3 mg of p-nitrophenyl glycoside (or phosphate), and enzyme (100 ~1 of supernatant or 10 pl of homogenate) in a final volume of 200 pl. After 1.5 h at 37 "C, 3 ml of a solution containing 0.3 M NaCl, 0.15 M KzC03 were added. Tubes were centrifuged at 1000 X g for 5 min, and absorbancies at 400 nm were determined. The percent of the total activity secreted by the cells of each 35-mm dish was calculated from the activities in the supernatant and homogenate. Sodium chloride (0.15 M) and sucrose (0.25 M) added to the assay vessels had no significant effect on enzyme activity ( n = 5). Protein was determined by the Lowry method (18).
Protease was measured by the extent of solubilization of radioactive material from [14C]hemoglobin. Substrate was labeled by exposure to ["Clformaldehyde followed by reduction with NaBH, (19). Lactic acid dehydrogenase in cell incubation medium and homogenate was measuredby the loss of absorbance at 340 nm of NADH upon addition of sodium pyruvate (20).
Cytosolic Ca2+ was measured in cells grown on coverslips (11 X 22 mm), incubated for 20 min at 37 "C in medium containing fura-2AM (5 mM). The coverslip was then washed three times with TM Sucrose and inserted into a holder adapted for a 1-cm cuvette containing stirred TM Sucrose (37 "C). The cuvette with the diagonally positioned coverslip was placed in a Spex Fluorolog 1681 0.22m Spectrometer furnished with a Spex 1699 Scanning Control Splitter. Excitation wavelengths were 340 and 380 nm, and emission wwelength was at 505 nm. Calcium ion concentrations were calculated by a DMIB Spectroscopy Laboratory Coordinator from the ratios of fluorescence excited by 340 nm to that excited by 380 nm. Calcium ion concentrations were determined using calibration measurements made in the presence of 2 mM ionomycin and 65 mM EGTA,' pH 8.3 (21,22).
Efflux Studies-Cells in each well of a Costar dish (24 wells) were incubated for 15 h in 1.5 ml of calcium-free medium containing 1 mCi of '5CaC12. Cells were washed twice with 2 ml of TM Sucrose, and in triplicate wells 1 ml each of TM Sucrose, TM Sucrose/CaClz (0.5 mM), TM Sucrose/NaC1(0.15 M), or TM Sucrose/NaC1(0.15 M), CaClz (0.5 mM) was added. At given times, radioactivity in O.1-ml aliquots of the medium was determined.
Cyclic AMP was measured in cells growing in 35-mm plastic dishes by radioimmunoassay. Cells were rapidly washed three times with 2 ml of TM Sucrose. One ml of TM Sucrose or TM Sucrose containing NaCl(O.15 M) was added. After incubation, the medium was removed and discarded and 1 ml of either 5% trichloroacetic acid or 0.1 N HC1 was added to extract cAMP (23). After 10 min at room temperature, extracts were transferred to a series of tubes. Trichloroacetic acid was removed by three extractions with water-saturated ether (5, 5, and 3 ml), and residual ether was removed by heating at 50 "C for 25 min (23). cAMP was assayed with a radioimmunoassay kit purchased from Amersham. Determinations at each time point were carried out in duplicate, and each extract was assayed in duplicate.

RESULTS
In the presence of 0.25 M sucrose solution, secretion by F9 cells increased with increasing concentrations of sodium ions, and maximum secretion was observed at approximately 150 The abbreviation used is: EGTA, [ethylenebis(oxyethyleneni-tri1o)ltetraacetic acid. mM NaCl (extracellular concentration) (Fig. 1). Approximately Vi of the total content of N-acetyl-b-D-glucosaminidase of the cell was secreted in 30 min (23.6 ? 6.3, n = 20). Other lysosomal enzymes such as acid protease, a-mannosidase, and @-galactosidase accompanied N-aCetyl-@-D-glUCOSaminidase. Surprisingly, acid phosphatase did not appear to respond to cation. The level of this enzyme in the medium was no higher than for the TM Sucrose control after 30 min ( Fig. 1). Similar results were obtained using PyS-2 cells.
Incubation of F9 and PyS-2 cells in TM Sucrose alone resulted in a relatively small amount of secretion of N-acetyl-@-Dglucosaminidase. Because the level of N-acetyl-&D-ghcosaminidase in F9 and PyS-2 cells was approximately 10 times that of the other lysosomal enzymes, N-acety~-~-D-glucosaminidase was almost always used as a marker. Chloride salts were routineIy used but similar results were obtained with sodium acetate, sodium isethionate, and sodium thiocyanate.
Time Course of Enzyme Accumulation-Enzyme accumulation in the medium was rapid, and after 30 min no further secretion was observed (Fig. 2). Because this observation was made eight times, a 30-min incubation was considered routine. Other lysosomal enzymes, a-mannosidase, @-galactosidase, and acid protease, accumulated at approximately the same rate and to the same extent. The lack of further secretion after 30 min suggests that only a certain defined fraction of lysosomal structures was capable of responding to cationic stimulation. In the absence of sodium ions, the level of Nacetyl-@-D-glucosaminidase in the medium increased only slightly ( Figs. 1 and 2).
Cell Viability-After a 30-min incubation in the presence or absence of cation, the cells remained adherent to the plastic surface, and the fraction of cells taking up the vital dye 25 ,  Under the phase contrast microscope ( X 250), PyS-2 and F9 cells in T M Sucrose or TM Sucrose/NaCl solution had slightly sharper and brighter edges, shrinking slightly and pulling away from each other. They remained attached and presented a cobblestone appearance. Granularity or vacuolation was not evident in cells bathed in either T M Sucrose or T M Sucrose/NaCl. Within 2 h of their return to medium, these cells were indistinguishable from control cells that had remained in culture medium. After 20 h, all cells were heavily confluent. WM-9, a cell which did not secrete, underwent the same series of changes.
To test the viability of cells in the assay, an experiment was carried out under sterile conditions in which vessels were assayed as usual and the number of cells counted. The experimental bathing medium of a second set of dishes was removed, culture medium was added, and the dishes were incubated a further 20 h when the cells were assayed and counted. A third set of plates was not manipulated the first day but was assayed, and the cells were counted on the second day. After four experiments it was found that cells fully retained their ability to divide and to secrete lysosomal enzyme 20 h after the first secretory discharge, indicating that washing and incubating the cells did not result in cell death. However, the level of N-acetyl-0-D-glucosaminidase in cells that had secreted 24 h previously was only half that of controls. Despite this, the cells responded fully to the sucrose-Na+ stimulus on the second day, suggesting that they were capable of converting undischarged lysosomes of the first day to a form capable of responding to a secretory stimulus on the Other Cations-Sodium ions were considerably more effective than K' or Li' in stimulating N-acetyl-P-D-glucosaminidase secretion (Fig. 3). Ca2+, Mg", and NH: ions, at lower concentrations, resulted in a maximum secretion of 5% of the total enzyme in 30 min. Krebs solution, a physiological salt solution (118 mM NaCl, 4.8 mM KC1, 2.5 mM CaCl,, 1.2 mM MgS04, 25 mM NaHCO,), was no more effective in stimulating N-acetyl-0-D-glucosaminidase secretion than Na' alone. Spermine at a concentration of 5 mM was more effective than Ca2+, Mg2', or NH:, but led to only 8% secretion in 30 min. Maximum secretion occurred when the medium was pH 7.0. Secretory activity was reduced by 20% at pH 6.8 and 7.4.
Requirement for Sucrose or Other Sugars-The simultaneous presence of sucrose and cation was required for maximum N-acetyl-P-D-glucosaminidase secretion (Fig. 4). In the presence of 0.15 M NaC1, the rate of secretion increased with increasing concentrations of sucrose to a maximum at 0.25 M (isoosmolar), and higher concentrations of sucrose (0.4 or 0.5 M) did not increase the rate further (data not shown). The optimal combined concentration of sucrose and NaCl is, therefore, two times isoosmolar. Incubation media were tested containing various proportions of sucrose and NaCl whose combined concentration was isoosmolar. Secretion took place but was always less than 50% of maximum (data not shown).
Cells (F9 and PyS-2) were washed with solutions of various   Table  I resulted in very active stimulation. Disaccharides such as sucrose, lactose, and cellobiose as well as trisaccharides (melezitose and raffinose) were more effective than glucose. Glycerol, D-xylose, D-mannose, and D-galactose could not substitute for sucrose (data not shown).
Sensitivity of Enzyme Secretion to Temperature and Metabolic Inhibitors-The loss of enzymes from F9 cells in the sucrose-Na+ system is temperature-dependent as in human fibroblasts growing in culture medium (24). At 22 "C N-acetyl-P-D-glucosaminidase was secreted at approximately half the rate than at 38 "C. Secretion was greatly but not completely inhibited by metabolic inhibitors such as cyanide and azide (5 mM) which inhibited 60%. Sulfhydryl blocking agents such as sodium iodoacetamide (1 mM), sodium iodoacetate (10 mM), and N-ethylmaleimide (5 mM) inhibited approximately 50%. Higher concentrations of inhibitors did not increase inhibition.
Effect of State of Growth of F9 Cells on Enzyme Secretion-Among experiments, there was considerable variation from week to week in the maximal percent of total enzyme secreted in 30 min by F9 and PyS-2 cells, although results within each experiment (i.e. within each set of cells) were always consistent. The basis for the variability is unknown. To determine whether the extent of secretion might be influenced by the state of growth of the cell, several dishes of F9 cells were plated, and a set of dishes was assayed in T M Sucrose in the presence or absence of 0.15 M NaCl on each of 4 days. Secretion was highest 48 h after plating, during early log phase. Thereafter, the response to sucrose-Na+ declined (later log and plateau phase) (Fig. 5). Based on these results, cells were routinely seeded at 5 x lo5 cells/35-mm plastic dish (5.2 x lo4 cells/cm2), and the experiment was carried out after 48 h when there were approximately 2-3 X lo6 cells per dish. Cells plated at 2 X lo6 per dish secreted 25% less enzyme at 48 h after plating than after 24 h.
Effects of Other Agents on Enzyme Secretion-Ouabain produced only small variable effects, whereas valinomycin and amiloride at a wide range of concentrations had only little or no inhibitory effects on enzyme secretion. Inclusion of cytochalasin B (10 wg/ml), vinblastine sulfate ( 2 pg/ml), polymyxin B sulfate (20 pg/ml), cycloheximide (15 pg/ml), or leupeptin (20 wg/ml) in the incubation medium had little effect (not shown). Cells were preincubated with cytochalasin B and vinblastine sulfate for 4 h before being assayed. Mannose 6-phosphate and glucose 1-or 6-phosphate (10 mM) had no effect on the secretion of lysosomal enzymes in the standard assay in the presence or absence of 0.15 M NaC1. Secretion     by F9 cells was not altered when 10% heat-inactivated horse serum (37 "C, 4 h, pH 10) was included in the assay medium. Secretion by human fibroblasts in culture is also unaffected by serum (24). Enzyme Secretion as a Function of Cell Type-Like F9 cells, PyS-2, a mouse parietal yolk sac cell derived by differentiation of a teratocarcinoma cell (17), secreted 15 to 30% of its lysosomal enzyme in 30 min in a sucrose-Na+ system. Table  I1 gives the percentages of total N-acetyl-P-D-glucosaminidase secreted in the presence or absence of 0.15 M NaCl for several different lines of cells in culture. Cells were tested once or twice in duplicate and no attempt was made to establish optimal conditions. However, it is evident that a range of cells respond to the sucrose-Na+ stimulus to different extents without elevation of protein in the medium. While F9 and PyS-2 cells are, to date, the best secretors, some cells such as a human melanoma (WM-9) did not appear to respond to the sucrose-Na+ stimulus (not shown).

N-acefyl-P-~-glucosaminidase by various cell types Cells were incubated in TM Sucrose or TM Sucrose containing 0.15 M NaC1. The standard assay for secretion of N-acetyl-8-Dglucosaminidase (NAGA) is described under "Experimental
Non-involvement of Calcium Ions in Enzyme Secretion-An important early step in several exocytotic processes is a rise in intracellular calcium ions (16). To determine whether calcium played a role in our system, we analyzed the rate of enzyme secretion by cells incubated with the calcium iono-phores ionomycin or A23187 (1-25 p M ) in the presence or absence of CaC1, (0.2-3.0 mM). In the presence of calcium ions and ionomycin, the rate of secretion by PyS-2 cells was within 5.9% of control rates ( n = 3). The corresponding difference in the presence of A23187 was 9.7% (n = 5) for PyS-2 cells and 2.0% for F9 cells ( n = 2).
The rates of uptake of 45Ca into F9 and PyS-2 cells, bathed in TM Sucrose or TM Sucrose/NaC1(0.15 M), were found to be similar (<15% difference over 30 min). Sodium ions also had little effect on the rates of efflux of 45Ca from these cells.
Free intracellular calcium ion concentrations were measured in F9 and PyS-2 cells growing as monolayers on coverslips, by double-beam spectrofluorimetry using fura-2AM (21,22). Levels were measured in TM Sucrose before and after addition of NaCl (0.15 M) and CaCl, (0.5 mM). Upon addition of NaCl, a relatively small short-lived increase in the level of calcium ions occurred, lasting no more than 1 min followed by a decline, sometimes to a level slightly below the original (Fig. 6 A ) . This calcium must have been derived from intracellular stores. Addition of CaC1, (0.2-0.5 mM) in the presence or absence of NaCl caused a far greater increase in cytosolic Caz+ levels that persisted for at least 5 min (Fig. 6). Thus, no   correlation appears to exist between secretion and the level of cytosolic calcium ions in the sucrose-Na+ system, since addition of CaC1, to cells did not significantly affect the rate or extent of secretion. These results also show that cells in the presence of sucrose, with or without NaCl, can control the level of cytosolic Ca2+. On the other hand, phenothiazines such as trifluoperazine, chlorpromazine, and promethazine caused considerable but not complete inhibition (Table 111) at concentrations that did not cause lysis of cells (1 mM trifluoperazine). Although the phenothiazines bound to calmodulin and interfered with several calmodulin-dependent processes (25), including exocytosis (16,26), they had other effects (27). TMB-8, another calmodulin antagonist (28), did not inhibit secretion at a concentration of 0.01 mM but was inhibitory at 0.1 mM (Table 111).
Involvement of Cyclic AMP-While sodium ions and sucrose alone were sufficient to elicit a secretory response, the process was enhanced by dibutyryl cAMP (Tables IV and V). Addition of Forskolin, a stimulator of adenyl cyclase (29,301, to F9 and PyS-2 cells almost doubled the rate of secretion, whereas an inactive analogue, dideoxyforskolin, at the same concentration had little effect (Table IV).
Cyclic AMP was measured in F9 and PyS-2 cells in the presence or absence of NaC1. Sodium ions appeared to cause a brief but definite elevation in the cAMP content of the cells (Fig. 7). Four-to fivefold increases were observed in 1 min, followed by a fall to control levels within 7 min. In three experiments, the levels of cAMP in PyS-2 cells, incubated in T M Sucrose or T M Sucrose/NaCl (0.15 M) for 30 min, were within 10% of one another.
Role of G Proteins in the Secretion of Lysosomal Enzymes-Treatment of cells with either cholera or pertussis toxin, which are known to ADP-ribosylate the a-subunit of G proteins (31), enhanced secretion (Table V). Addition of cAMP to cells already stimulated by pertussis toxin did not have a statistically significant effect. Presumably, cholera toxin stimulated G., whereas pertussis toxin abolished inhibition mediated by Gi (32,33). Secretion by glioma cells was stimulated by pertussis toxin, which also enhanced adenyl cyclase activity in these cells (33). When AlCL (30 FM) and NaF (10 mM) were added, forming AlF;, G protein function was stimulated (34,35), and in our experiments, secretion was clearly increased (Table VI). Many activities have been attributed to the vanadate ion (36), including the ability to stimulate G proteins. Adenyl cyclase activity of rat adipose cells was stimulated by vanadate ions (37), and in the present experiment vanadate stimulated secretion whether or not sodium ions and sucrose were present together or singly. In the standard assay, in seven experiments using PyS-2 cells, 1 mM vanadate stimulated secretion by 70% in the presence of NaCl and 75% in the absence of cations. It is unlikely that vanadate acted through its capacity to inhibit phosphatases (36) since NaF (50 mM), which is a phosphatase inhibitor as well as an inhibitor of phosphorylases and kinases (38), caused a 50% reduction of secretion ( n = 5).
Lack of Involvement of Phospholipase C in Enzyme Secretion-Two lines of experiments suggested that phospholipase C played little or no part in the secretory response in the sucrose-Na+ system. Neomycin (0.5-1.0 mM), an inhibitor of phospholipase C (39, 40), did not inhibit secretion (PyS-2 cells, control 31.0 f 2.3% secretion; with neomycin, 32.5 f 4.1% secretion; n = 5). Furthermore, cells grown overnight in the presence of the tumor promoter, phorbol myristate acetate M), and assayed in its presence secreted N-acetyl-B-Dglucosaminidase at the same rate and to the same extent as controls. If phospholipase C took part in the process, phorbol ester, an active substitute for diacylglycerol and one of the products of the reaction catalyzed by this enzyme (41), would be expected to exert a noticeable effect. DISCUSSION In the present study, rapid secretion of lysosomal enzymes by cells in culture was induced by the simultaneous presence of a monovalent cation and a sugar such as sucrose. Na+ was the most effective cation. Cells, which usually secrete lysosomal enzymes at a relatively slow rate, could be induced to discharge a third of their total enzyme content in 30 min. Despite the fact that cells were exposed to unphysiological conditions (2 x isoosmolar) for 30 min, the levels of the cytosolic marker, lactic acid dehydrogenase, and protein in the extracellular medium were unchanged. Cells remained viable and excluded a vital dye. When replaced in growth medium, they divided at an undiminished rate compared with untreated control cultures. After 24 h they were capable of repeating the process. Treated cells retained the ability to control the levels of cytosolic Ca2+ and cAMP and to respond in a graded manner to various stimulating and inhibitory substances. The secretory response was highly reproducible and was proportional to the concentration of cation and sucrose until there was no response. Sucrose-Na+ so enhanced the usually slow rate of exocytosis of lysosomal enzymes that the mechanisms involved could be conveniently examined in a manner similar to studies on the discharge of granules from mast cells (15) and fertilized eggs (16). Problems of uptake, synthesis, and degradation of enzymes, which might compli-cate the evaluation of a slow secretory process, were eliminated.
In the present study and using this method, we were able to show that exocytosis of lysosomal enzymes involved G proteins and CAMP and that secretion did not depend on changes in the level of cytosolic Ca2' or on the activity of phospholipase C. In this regard, the lysosomal system differs from the exocytotic systems of mast cells (15), fertilized eggs (16), and other secretory systems (42).
This procedure was effective with several types of cells and may be of general applicability. A closer examination of the secretory mechanism of cells from individuals with various forms of lysosomal membrane transport (43, 44) and lysosomal storage disease (45) is possible. Do any of the known genetic defects causing these diseases have significant effects on the secretion of lysosomal contents? For instance, it will be of interest to examine in detail the exocytosis of lysosomal enzymes and free sialic acid from cells derived from individuals with Salla disease in which egress of sialic acid from the lysosome is defective (43).
Only a certain fraction of the total complement of lysosomal enzymes was secreted in 30 min, and no further loss of enzymes from the cells took place with extended incubation. Usually, 20-30% of the total was secreted but was never more than approximately 50%, whatever the stimulus. Up to 50% of the lysosomes, perhaps those fused with secretory vesicles, appeared to be competent to respond. The apparent division of lysosomes into two populations may permit us to define the change(s) in the lysosomal membrane necessary for the lysosomes to move to the surface and fuse with the plasma membrane during the secretory process. Since, CAMP, as we have shown, is an important mediator in lysosomal enzyme secretion, it is possible that a phosphorylation of proteins by activation of protein kinase A is essential. Phosphorylation of membrane proteins during secretion is now being investigated.
Acid phosphatase differed from all the other lysosomal enzymes in that it was not secreted. This observation is consistent with the data of Waheed et al. (46) who showed that the enzyme, transported to the lysosome as a transmembrane protein, was later cleaved to form a soluble enzyme. In PyS-2 and F9 cells cleavage may be delayed or, possibly, the lysosomes responding to the sucrose-Na' stimulus were recently formed and had phosphatase still attached to the lysosomal membrane.
Although sucrose was used routinely in this study, other sugars appeared to be at least as effective, while still others were inactive. Maximum activity was observed at a concentration of 0.25 M sucrose (isoosmolar) in the presence of isoosmolar Na' (0.15 M), i.e. twice overall isoosmotic strength. Combined concentrations greater or less than two times isoosmolar were not as effective. Although hyperosmolarity is probably an important part of the stimulating impulse, the process requires the combined presence of a monovalent cation and a polyhydroxy1 compound such as sucrose. Neither alone, whatever their osmolarity, is effective in inducing secretion.
It is known that sucrose and other sugars with M, > 220 are taken up by cells (24, 47, 48) and accumulate in the lysosomes. Water drawn into the lysosomes caused swelling and vacuolation of the cell (48, 49). However, we found that cytochalasin B, which inhibited uptake of sucrose into human liver cells in culture (50), had no effect on secretion in our system. It is quite possible that accumulation of sugar and water in the lysosomal vesicle is not associated with secretion. Perhaps, since the incubation period in our assay was rela-tively brief, secretion may have occurred before significant uptake of sucrose and water took place. It should be noted that D-ghCOSe ( M , 180) could substitute for sucrose in our system, despite the fact that this sugar can escape the lysosome and does not lead to vacuolation (6, 41, 51). More long term effects such as inhibition of endocytosis (52) and an increase in levels of lysosomal enzymes (6, 49, 53), which occurred after several hours of exposure to sucrose, probably did not have a bearing on the sodium-sucrose system described here.