0013-7227/78/1034-1245$02.00/0 Endocrinology Copyright © 1978 by The Endocrine Society
Vol. 103, No. 4 Printed in U.S.A.
Acute Stimulated Hormone Release from Cultured GH3 Pituitary Cells* RICHARD E. OSTLUND, JR., JOYCE T. LEUNG, SHIRLEY VAEREWYCK HAJEK, THOMAS WINOKUR, AND MARK MELMAN Metabolism Division, Washington University School of Medicine, St. Louis, Missouri 63110 ABSTRACT. Treatment of cultured rat pituitary GH3 cells with 50 mM KC1 in growth medium released 33% of cell PRL and 18% of cell GH with a half-time of 5 min. Hormone in the culture medium was increased 2- to 4fold over unstimulated levels. The response required calcium; barium and strontium, but not magnesium, could substitute for calcium. Low temperature completely inhibited hormone release, which was also reduced significantly by inhibitors of energy metabolism and by nitrogen. This acute response was similar in ionic requirements, hormones released, and time course to the acute effect of TRH. Like potassium stimulation, TRH
resulted in acute release of both PRL and GH. This contrasts with the finding that chronic TRH treatment reduced GH synthesis in GH3 cells. After a 10-min preincubation with potassium, subsequent short incubations with potassium released little hormone unless the cells were allowed to recover by incubation in normal medium for at least 2 h. This acutely releasable hormone pool seems to be located in a membrane-bound subcellular fraction, since GH3 cells did not discharge the cytoplasmic marker enzyme, lactate dehydrogenase, during potassium-stimulated hormone release. (Endocrinology 103: 1245, 1978)
M
and derivatives of cAMP (7-10). Since GH and P R L are stable in the culture medium, hormones secreted over several days are a reflection of synthesis (11, 12). Acute release of preformed hormone has also been reported from GH3 cells. T R H caused detectable P R L output into the medium after a 10-min incubation, and this rose to a 50% increase over basal levels by 60 min (9, 13). Incubation of GH 3 cells for 3 h with medium containing 50 mM KC1 caused a 70% depletion of intracellular GH and P R L (3). Although a statistically significant corresponding increase in the relatively large medium hormone content was not seen (unless the cells were pretreated with T R H ) , it was suggested that release of preformed hormone occurred. Because of our interest in the mechanism of hormone extrusion from secretory cells, the acute stimulated release of hormones from Received November 14,1977. Address requests for reprints to: Dr. Richard E. Os- GH 3 cells was studied. Both T R H and a high tlund, Jr., Washington University School of Medicine, concentration of potassium in the presence of Department of Internal Medicine, Metabolism Division, Barnes and Wohl Hospitals, 660 South Euclid Avenue, St. calcium were found to release substantial amounts of intracellular GH and P R L over Louis, Missouri 63110. * This work was supported by grants from the Ameri- minutes. T h e effect of T R H on acute GH can Diabetes Association, the Diabetic Children's Welfare Fund of the St. Louis Diabetes Association, and Grants release was qualitatively different from the chronic effect on hormone synthesis. AM-20421 and AM-20579 from the NIH.
OST secretory tissues require extracellular calcium for hormone release (1, 2). Net entry of calcium into the cell can be achieved by the addition of depolarizing amounts of potassium to the incubation medium. Hence, solutions containing potassium and calcium are able to stimulate endocrine secretion (1-3). Such a model may aid understanding of normal hormone secretion, since calcium also enters endocrine cells during physiological secretory stimulation (4, 5). GH3 cells are an established cloned tissue culture line derived from a rat pituitary tumor which have become an important source for studying the molecular biology of secretion (6, 7). GH and PRL are secreted in the basal state and respond appropriately to the administration over several days of stimulators and suppressors, such as TRH, cortisol, bromcryptine,
1245 The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 03:59 For personal use only. No other uses without permission. . All rights reserved.
OSTLUND, JR., ET AL.
1246
Materials and Methods
Endo • 1978 Vol 103 • No 4
Assays
were used 1 day after the last medium change. Cells were also grown in a 250-ml spinner vessel (Bellco) with Eagle's spinner medium (Microbiological Associates), which contains the same serum and antibiotics as monolayer cells (16). Cell density varied from 200,000-800,000/ml on the day of harvest. Half of the suspension was removed and replaced with fresh medium three times weekly.
Rat PRL and GH were determined by specific double antibody RIA with reagents provided by Dr. A. F. Parlow through the NIH Rat Pituitary Hormone Distribution Program. Twenty microliters of medium were added to an assay volume of 0.59 ml. Growth medium was found to reduce the binding of GH to antibody and, consequently, the medium content of all tubes in the GH assay was adjusted to 20 JLJ. Hormone measurement in cells was accomplished by treating the cell sonicate made in distilled water with sodium hydroxide to achieve a final concentration of 0.01 N. After 15 min at room temperature, the basic solution was diluted into the immunoassay tubes. Failure to treat with sodium hydroxide and immediately dilute often resulted in lower values for cell hormone recovery. Such treatment did not affect hormone standards. Lactate dehydrogenase (LDH) was determined spectrophotometrically and LDH units were calculated as described (17). The assay was performed in a 0.5-ml cuvette using 0.1 ml sample in order to increase sensitivity. A cell extract of LDH activity in 0.15 M NaCl was recovered with 92 ± 1.2% efficiency when added to growth medium, where it was stable with incubation for over 1 h at 37 C or overnight at 4 C.
Acute hormone release
Chemicals
Monolayer cells on 100-mm plastic dishes were washed six times with warm Puck's saline G (14) and then aspirated to dryness with a Pasteur pipet. Two milliliters of Ham's F-10 medium, containing serum, antibiotics, and 2 mM CaCl2 in addition to that in the Ham's formulation and serum (calciumaugmented growth medium), were added, and the cells were incubated at 37 C for 10 min in an atmosphere of 95% air and 5% CO2. Serum-free medium could also be used. After incubation, the medium was aspirated and frozen, and the cells were washed six times with 0.15 M NaCl at 4 C and frozen. Cell protein was determined by the method of Lowry et al. (15) after scraping the thawed cells into 2.5 ml distilled water and sonicating them for 1.5 min. Cells grown in suspension culture were washed by sedimentation for 2 min at top speed in a benchtop IEC clinical centrifuge and were resuspended immediately, since storage of the washed cells in a pellet resulted in reduced secretory response. Results are presented as nanograms of hormone released per mg cell protein ± SEM. Statistical comparisons were done with Student's two-tailed unpaired t test.
Chemicals were purchased from the following suppliers: TRH, Abbott; cortisol sodium succinate, Upjohn; ethylenebis-(oxyethylenenitrilo) tetraacetic acid, Eastman; strontium chloride, Merck; barium chloride, Mallinckrodt; calcium chloride, potassium chloride, and sodium chloride, Fisher Co.
Cells GH3 cells (CCL 82.1) were purchased from the American Type Culture Collection, Rockville, MD. The cells were cultivated in 35- or 100-mm plastic dishes with growth medium, consisting of Ham's F10 medium (Microbiological Associates) containing 15% horse serum and 2.5% fetal bovine serum (Gibco), 50 U/ml penicillin, 50 jug/ml streptomycin, and 1.25 jug/ml amphotericin B (Fungizone, E. R. Squibb and Sons, Inc.). The presence or absence of amphotericin did not affect hormone release. Ham's medium contains 127 mM sodium, 3.8 mM potassium, and 0.3 mM calcium as well as other components. Cultures were plated at 106 cells/100-mm dish and incubated at 37 C in a humidified atmosphere of 5% CO2 and 95% air for 7-21 days. The medium was changed twice weekly, and the cells
Results Acute hormone release in response to potassium and calcium Although GH3 cells have a high basal rate of PRL and GH secretion, a 2- to 4-fold increase in the release of both was seen when cells were incubated for 10 min with 50 mM KC1 in calcium-augmented growth medium compared to incubation in calcium-augmented medium alone (Table 1). This response was seen with GH3 cells grown in spinner suspension culture as well as cells attached in monolayers to plastic dishes. The release phenomenon as a function of time is plotted in Fig. 1. There was an initial
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 03:59 For personal use only. No other uses without permission. . All rights reserved.
GH, HORMONE RELEASE
1247
TABLE 1. Acute potassium-stimulated hormone release PRL
GH
+KC1 -KCl +HC1 -KCl 22.6 ± 1.6 86.9 ± 10.3 108 ± 22 Monolayer cells 285 ± 23 71.4 ± 3.6 Spinner cells 42.3 ± 4.8 111 ± 3.5 203 ± 7.4 GH3 cells grown in either monolayer or spinner culture were washed and incubated 10 min with 2 ml calciumaugmented growth medium containing either 50 mM KCl (+KC1) or no addition (—KCl). At the end of the experiment, the medium was aspirated and analyzed for hormone content. Means and SES for triplicate determinations are presented. GH and PRL secretion was significantly increased by potassium in both spinner and monolayer cells (P < 0.01). Values are expressed as nanograms of hormone per mg cell protein • 10 min. TABLE 2. Effect of cations on potassium-stimulated PRL release 80
PRL released (ng/ %of no mg protein • 10 addition < '5 min) 6 2 60 100 None 24.9 ± 1.67 189 50 mM KCl 47.1 ±2.5" 106 50 mM NaCl 26.3 ±3.9 2 o) 40 50 mM KCl + 1 mM Na 18.1 ±0.4" 73 EGTA 20 283 50 HIM KCl + 1 mM Na 70.5 ±2.9" EGTA + 2 mM CaCl2 294 50 HIM KCl + 1 mM Na 73.1 ±3.5" EGTA + 2 mM BaCl2 20 40 60 80 222 50 mM KCl + 1 mM Na 55.4 ±0.7" MINUTES EGTA + 2 mM SrCl2 FIG. 1. Time course of PRL secretion. Confluent GH3 50 mM KCl + 1 mM EGTA 26.1 ±2.9 105 cells raised in 100-mm dishes were washed six times with + 2 mM MgCl2 saline G, and 5 ml calcium-augmented growth medium Triplicate 100-mm dishes of GH cells were washed with and without 50 mM KCl were added. The cells were and incubated with growth medium3 containing 0.5 mM placed in a 37 C CO2 incubator, and 50-fil aliquots were CaCl2 and the additions indicated for 10 min, after which taken for hormone assay at the Indicated times. Each the medium was removed and assayed. point represents the mean of duplicate dishes (control, ° P < 0.02, compared to no addition. Addition
• — • ; 50 mM KCl, O—O).
burst of PRL release, which was demonstrable as early as 2.5 min after the addition of 50 mM KCl and which became half-maximal at 5 min. After 20 min, there was little difference in continuing secretory rate between dishes treated with KCl and controls. The PRL present at time zero may represent secreted hormone adsorbed to the cells which is capable of release by the protein-containing incubation medium but not by the washing buffer. The incubation medium in these experiments was Ham's F-10, containing antibiotics, 15% horse serum, 2.5% fetal bovine serum, and 2 mM CaCl2 in addition to the 0.3 mM CaCl2 of Ham's salt formulation and that in the serum. The total calcium content measured by atomic absorption spectroscopy was 2.5 mM.
Effect of cations The specificity of potassium-stimulated hormone release was examined. As shown in Table 2, 50 mM NaCl in growth medium had no effect upon hormone release, suggesting that the osmotic effect of extra KCl does not contribute to hormone release. Extracellular calcium was essential for potassium to stimulate hormone release. The calcium chelator, EGTA, at 1 mM reduced potassium-stimulated PRL release to 38% of the expected release, a level lower than that seen in untreated controls (Table 2). However, this was not merely a toxic effect of EGTA, since potassium-stimulated secretion was restored if calcium was added in excess of EGTA. Excess strontium and barium also restored stimulated hormone
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 03:59 For personal use only. No other uses without permission. . All rights reserved.
OSTLUND, JR., ET AL.
1248
TABLE 3. The effect of calcium on stimulated PRL release into the medium Final Ca concentration
PRL release
(HIM)
-KC1 +50 mM KC1 100 ± 8 320 ± 11 1.5 93 ± 6 307 ± 14 2.5 120 ± 11 379 ± 26 4.5 119 ± 10 339 ± 20 8.5 99 ± 5 361 ± 16 GH3 cells were washed and incubated for 10 min with growth medium containing CaCl2 added to achieve the amounts indicated. Growth medium itself contained 0.5 mM CaCl2. Numbers from three to five dishes are expressed as a percentage of control release without added calcium or potassium. 0.5 No added CaCl2
TABLE 4. Acute hormone release by TRH Hormone released (ng/mg cell protein • 30 min)
PRL
GH
No addition 120 ± 2.3 139 ± 14 +50 nM TRH 273 ± 15° 375 ± 28" Triplicate 100-mm dishes containing cells were washed and 2.0 ml calcium-augmented growth medium were added with or without 50 nM TRH. The cells were placed in a CO2 incubator at 37 C for 30 min, after which the medium was removed and assayed. ° P < 0.01 difference from no addition.
release in EGTA-treated cells, but magnesium was without effect. The addition of up to 8 mM calcium to growth medium had little effect on hormone release in the presence or absence of excess KC1, but the greatest stimulated response occurred at 2.5 mM calcium (Table 3). Effects of TRH
Endo • 1978 Vol 103 • No 4
indicating that the GH synthetic machinery was capable of stimulation. TRH increased chronic PRL synthesis, as expected, to 206% of control. Hence, TRH rapidly released stored GH while inhibiting chronic synthesis of the same hormone. Calcium dependence of TRH-mediated hormone release Acute TRH-stimulated PRL and GH release was found to require extracellular calcium in a fashion very similar to acute potassium-stimulated release. In Table 6, it is seen that within 30 min TRH produced an 81% increase in PRL release into the medium. TABLE 5. Effect of chronic treatments on hormone synthesis jug Hormone/mg cell protein • 3 days GH PRL 49.7 ± 4.5 10.8 ± 2.3 35.0 ± 2.2" 22.2 ± 3.2° 223 ±42° Dishes (100 mm) of cells were incubated with the indicated material in growth medium for 3 days. The medium was aspirated and replaced, and the cells were incubated another 3 days. Hormones were measured in the medium from the final 3-day period. Numbers are from three to five experiments. a P < 0.05, compared to no addition. No addition 50 nM TRH 5 JUM Cortisol
TABLE 6. The effect of calcium on TRH-stimulated hormone release ng/mg Cell protein • 30 min Addition
PRL GH TRH is known to release PRL from both None 495 ± 58 604 ±45 rat pituitary (18) and GH3 cells (9). The data 894 ± 64° 50 nM TRH 1130 ± 66° of Table 4 confirm this, showing a 2.3-fold 1 mM EGTA 203 ± 35" 743 ± 118 661 ± 78 increase of PRL release in response to 50 nM 50 nM TRH + 1 mM 318 ± 42 EGTA TRH over 30 min. Unexpectedly, however, it 50 nM TRH + 1 mM 1007 ± 67" 1157 ± 161° was found that TRH also caused a 2.7-fold EGTA + 2 mM CaCl2 954 ± 71" 50 nM TRH + 2 mM 1120 ± 176" increase in GH release. CaCl2 In contrast to acute hormone release, the 2 mM CaCl 2 433 ± 73 768 ± 69 chronic synthesis of hormone can be measured Triplicate 35-mm dishes of cells were prepared and from medium hormone accumulation over a washed, as described in Materials and Methods. At time 3-day period (11, 12). TRH reduced chronic zero, 0.5 ml growth medium (0.5 mM CaCl2) was replaced synthesis of GH to 70% of the control value with or without additives. The cells were placed in a CO2 at 37 C for 30 min, after which the medium was (Table 5). However, a 4.5-fold increase in GH incubator removed and assayed. a synthesis was seen when cortisol was added, P < 0.05, compared to no additions.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 03:59 For personal use only. No other uses without permission. . All rights reserved.
GH3 HORMONE RELEASE
TRH-stimulated PRL release was reduced to 36% of the expected levels by EGTA, less than that of untreated controls, but was restored when calcium was added to excess of EGTA to the TRH-containing medium. Similar results were found for GH release. It should be noted that the basal release of PRL was reTABLE 7. Energy metabolism and acute hormone release Preincuba.. . , tion period
Inhibitor
(mm)
Potassiumstimulated PRL
release
(% control)"
4C° Spinner cells Monolayer cells
30 10
-2.7 ± 11 1.5 ± 4.5
Dinitrophenol (5 mM)
15
22.4 ± 11
KCN (5 mM)
15
36.6 ± 5.7
0
Nitrogen atmosphere 47.9 ± 9.6 30 Triplicate dishes were washed and incubated in calcium-augmented growth medium, as described in Materials and Methods, with and without the inhibitors. After the stated time, the medium was removed and replaced with fresh medium containing inhibitor with or without 50 mM KC1 and the cells were further incubated for 10 min. The medium was then removed and analyzed for hormone content. Experiments involving nitrogen were modified to eliminate even transient exposure to air. Nitrogen continuously flowed into a plastic centrifuge tube with two holes punched in the cap. At the end of preincubation, centrifugation to remove the medium was omitted and KC1 was added to the appropriate tubes through the second cap hole. At the completion of the experiment, secretion was stopped by placing the tubes at 4C. ° (Potassium-stimulated release — nonstimulated release) inhibitor present/(Potassium-stimulated release — nonstimulated release) inhibitor absent X 102. * Cells at 4 C and controls at 37 C were incubated in air with medium adjusted to pH 7.3 with dilute HC1. c Incubated in air or nitrogen at pH 7.3.
1249
duced to 41% of the expected levels by EGTA, while the basal release of GH was not significantly affected by EGTA. Energy dependence of stimulated hormone release Inhibitors of energy metabolism resulted in reduced potassium-stimulated hormone release, as shown in Table 7. Cells were preincubated with or without inhibitors for the specified time before being incubated in the presence or absence of 50 mM KC1 with or without inhibitors. The ability of potassium to stimulate PRL release in the presence of an inhibitor was expressed as a percentage of the effect of potassium to stimulate release in cells never exposed to inhibitor. Incubation at 4 C completely eliminated potassium-stimulated PRL release. In addition, the unstimulated baseline secretory rate was reduced by 4 C incubation to 25.6% and 6.4% of that seen at 37 C for spinner and monolayer cells, respectively. Dinitrophenol, potassium cyanide, and a nitrogen atmosphere all significantly inhibited potassium-stimulated release. Recovery of hormone-releasing capacity after potassium stimulation Refractoriness to further potassium-stimulated PRL release after the initial hormone burst is seen in Fig. 1. This was not due to total cell hormone depletion. Monolayer cells were stimulated 10 min with 50 mM KC1 and then hormones were measured in the medium and in the washed cells (Table 8). Net stimulated PRL release (defined as medium hor-
TABLE 8. Hormone balance with potassium stimulation PRL -KC1
GH +KC1
-KC1
+KC1
Hormone content (ng/mg) protein Medium 52.3 ± 3 213 ± 8 102 ± 18 320 ± 33 357 ± 23 253 ± 12 1078 ± 62 888 ±59 Cells Medium + cells 409 ±22 466 ± 19 1180 ± 75 1208 ± 48 8.7 Hormone release (medium x 102)/ 12.8 45.7 26.5 (medium + cells) (%) Net potassium-stimulated hormone 32.9 17.8 release (+KC1) - (-KC1) (%) Triplicate monolayer GH3 cells in 100-mm dishes were washed and stimulated for 10 min with 50 mM KC1, as described in Materials and Methods. After stimulation, the hormone content of the medium and the washed cells was determined.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 03:59 For personal use only. No other uses without permission. . All rights reserved.
OSTLUND, JR., ET AL.
1250
mone in the presence of 50 mM KCl less medium hormone in the absence of added KCl) was 32.9% of total PRL (medium plus cells). Similar calculations for GH showed a 17.8% release. To determine how fast the acutely releasable cell hormone pool is regenerated, the following experiment was performed. Dishes of GH3 cells were stimulated with 50 mM potassium chloride, allowed to recover by incubation in the absence of potassium chloride, and then restimulated with potassium chloride. The results are plotted in Fig. 2. During the initial stimulation, PRL release was 64.3 ± 10 ng/mg protein • 10 min higher in dishes receiving potassium chloride than in control dishes. When a group of dishes was then immediately restimulated with potassium chloride, the difference in PRL release between stimulated and nonstimulated dishes was reduced to 17 ± 6.7 ng/mg protein-10 min (P < 0.02). After 1 h of recovery in nonstimulating medium, potassium still had little effect, but partial restoration of the potassium response to approximately 40% of the expected response took place after 2 and 3 h of recovery incubation (26.6 ± 4.2 and 23.0 ± 5.8 ng/mg protein 10 min, respectively).
£J
2
< o ^ ~ o Z 2
0
1
2
3
RECOVERY TIME (hr.)
FIG. 2. Recovery of capacity to release PRL in response to KCl. Four sets of triplicate GH3 cells in 100-mm dishes were washed and stimulated with 50 mM KCl (see Materials and Methods) from time -10 min to time zero. At time zero, the medium was removed, the cells were washed twice with saline G, and the remaining fluid was aspirated. One set of cells was immediately restimulated with KCl for 10 min beginning at time zero. Other sets of cells were incubated for 1-3 h in medium not containing excess KCl, after which the medium was removed and the cells were washed twice with saline G and restimulated with KCl. Control cells received the same washings and medium changes, but no KCl was present at any time (control, • — • ; 50 mM KCl, O—O).
Endo • 1978 Vol 103 • No 4
TABLE 9. Release of LDH during acute potassium stimulation U/mg 627 ± 16
% Total LDH 100 ± 2.6
Total cell LDH content LDH released into medium during potassium stimulation 13 ± 3 2.1 ± 0.5 50 mM KCl No added KCl 23 ± 8 3.7 ± 1.3 LDH was determined on the cells and media from Table 8.
Release of LDH Release of the cytoplasmic marker enzyme, LDH, during potassium-stimulated hormone release was determined, and the results are presented in Table 9. The total cell LDH content was measured in the supernatant from cells homogenized in 2.5 ml 0.15 M NaCl and centrifuged at 40,000 X g for 20 min. During the 10-min potassium stimulation period, less than 2.1% of cell LDH was released into the incubation medium. Control dishes not containing added KCl released 3.7% into the medium. Both of these values are much less than the percentage of release of PRL and GH. Discussion We have found that PRL and GH are actively released from GH3 cells by potassium in the presence of calcium. The process had a half-time of 5 min. This phenomenon is consistent with the previously reported finding that GH3 cells contain less hormone after incubation with 50 mM potassium chloride for 3 h (3). By incubating for a much shorter time, hormone release relative to unstimulated controls was enhanced, and the medium hormone content was increased several fold. PRL release in the presence of both 50 mM potassium (Table 2) and 50 nM TRH (Table 6) was calcium dependent, being reduced below the release found in untreated dishes by an excess of the calcium chelator, EGTA. Release was restored if calcium was added in excess of EGTA. However, basal hormone release, especially of GH, had a significant component which was not calcium sensitive (Table 6). This is consistent with a previous report that calcium had little effect on basal hormone release in GH3 cells (24). Because of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 03:59 For personal use only. No other uses without permission. . All rights reserved.
GH3 HORMONE RELEASE
its calcium sensitivity, stimulated GH3 hormone release may more closely resemble that occurring in normal pituitary tissue. Acute potassium-stimulated hormone release was similar to the acute but not the chronic effects of TRH. TRH released both PRL and GH when incubated with cells for 30 min (Table 4), but it inhibited GH synthesis while stimulating PRL synthesis when incubated with cells for 6 days (Table 5). The acute TRH response was also dependent upon extracellular calcium (Table 6). The dual effect of TRH is consistent with other GH cell data. Tritiated TRH bound reversibly to cell surface receptors and could be easily displaced by unlabeled TRH after 90 min (23), but after 24 h, binding became irreversible and the increase in PRL synthesis persisted for at least a week even if TRH was withdrawn from the medium (7). Both potassium and TRH probably acutely increase cell membrane permeability to calcium, resulting in activation of the release mechanism. The chronic effects of TRH may then be exerted directly or indirectly on the cell nucleus or on translation of messenger RNA. The acute release of GH by TRH is similar to that reported in acromegalic subjects (19) and in vitro perfused normal rat hemipituitaries (26). Since GH3 cells are transformed, it should be expected that their mode of secretion might differ from normal. For example, GH3 cells have very few secretory granules and those are quite small (References 22 and 27, and unpublished observations). GH localized by immunological techniques in GH cells, is found in what appears to be endoplasmic reticulum and in the perinuclear area (22). Nevertheless, the hormone does not seem to be diffusely distributed in the cytoplasm. Acute stimulated secretion, therefore, may involve fusion of hormone-containing vesicles with the plasma membrane. This hypothesis is supported by our finding that release of the cytoplasmic marker enzyme, LDH, was not increased after potassium-stimulated secretion (Table 9). Hence, stimulated hormone release is not merely due to cell "leakage." Not all cell hormone is acutely releasable. In fact, less than half of the PRL and GH is extruded after potassium addition. What
1251
causes some immunoreactive hormone to be releasable is not known. It may be an alteration in the hormone-containing organelles, such as distance from or ability to fuse with the plasma membrane. Alternatively, the nonreleasable hormone may simply lack sufficient cytosol calcium as a secretory trigger because of changing plasma membrane cation fluxes over time. Our data demonstrate that storage of hormone and compartmentation of hormone within the cell are functions that can be performed by cells having very few classical secretory granules. These functions are probably achieved by means of tiny granules or membrane-bound hormone-containing vesicles, such as those of the endoplasmic reticulum or Golgi apparatus. Calcium seems to be the final common pathway of the acute stimulated release phenomenon. It is intriguing that barium and strontium, but not magnesium, will substitute for calcium in supporting potassium-stimulated release. This ion specificity is known to be associated with other secretory phenomena (2, 21) and with the ability to cause muscle contraction (20). Calcium may stimulate secretion by acting on a variety of contractile proteins, such as actomyosin or tubulin, known to be present in secretory tissues (25). The finding that potassium-stimulated hormone release was sensitive to cold, nitrogen, dinitrophenol, and KCN suggests that it was dependent upon cell energy stores. This is also consistent with the hypothesis that contractile proteins are involved in the acute release process. If so, the GH3 system with a theoretically limitless supply of cloned cells could prove very useful for the investigation of those proteins. References 1. Rubin, R. P., Calcium and the Secretory Process, Plenum Press, New York, 1974. 2. Douglas, W. W., Stimulus-secretion coupling: the concept and clues from chromaffin and other cells, Br J Pharmacol 34: 451, 1968. 3. Gautvik, K. M., and A. H. Tashjian, Effects of cations and colchicine on the release of prolactin and growth hormone by functional pituitary tumor cells in culture, Endocrinology 93: 793,1973. 4. Douglas, W. W., and A. M. Poisner, On the mode of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 03:59 For personal use only. No other uses without permission. . All rights reserved.
1252
5. 6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
OSTLUND, JR., ET AL.
acetylcholine in evoking adrenal medullary secretion: 16. increased uptake of calcium during the secretory response, JPhysiol 162: 385, 1962. Malaisse, W. J., Role of calcium in insulin secretion, 17. IsrJMedSci 8: 244, 1972. Tashjian, A. H., Jr., Y. Yasumura, L. Levine, G. H. Sato, and M. L. Parker, Establishment of clonal strains of rat pituitary tumor cells that secrete growth 18. hormone, Endocrinology 82: 342, 1968. Tashjian, A. H., Jr., and R. F. Hoyt, Jr., Transient controls of organ-specific functions in pituitary cells in culture, In Sussman, M. (ed.), Molecular Genetics and Developmental Biology, Prentice-Hall, Engle- 19. wood Cliffs, 1972, p. 353. Dannies, P. S., and A. H. Tashjian, Jr., Effects of thyrotropin-releasing hormone and hydrocortisone on synthesis and degradation of prolactin in a rat pitui- 20. tary cell strain, J Biol Chem 248: 6174, 1973. Dannies, P. S., K. M. Gautvik, and A. H. Tashjian, Jr., A possible role of cyclic AMP in mediating the 21.
Endo i 1978 Vol 103 , No 4
Bancroft, F. C, and A. H. Tashjian, Jr., Growth in suspension culture of rat pituitary cells which produce growth hormone and prolactin, Exp Cell Res 64: 125, 1971. Wacker, W. E. C, D. D. Ulmer, and B. L. Vallee, Metalloenzymes and myocardial infarction. II. Malic and lactic dehydrogenase activities and zinc concentrations in serum, N EnglJMed 255: 449, 1956. Rivier, C, and W. Vale, In vivo stimulation of prolactin secretion in the rat by thyrotropin-releasing factor, related peptides, and hypothalamic extracts, Endocrinology 95: 978, 1974. Coutant, G., M. Vandeweghe, and A. Vermeulen, Comparison of TRF, propranol-glucagon, insulin, and glucose stimulation tests in acromegaly, Horm Metab Res 9: 17, 1977. Caldwell, P. C, and G. Walster, Studies on the microinjection of various substances into crab muscle fibers, JPhysiol 169: 353,1963. Douglas, W. W., and R. P. Rubin, The effects of
effects of thyrotropin-releasing hormone on prolactin
alkaline earths and other divalent cations on adrenal
release and on prolactin and growth hormone synthesis in pituitary cells in culture, Endocrinology 98: 1147,1976. Gautvik, K. M., R. F. Hoyt, Jr., and A. H. Tashjian, Jr., Effects of colchicine and 2-Br-a-eryocryptinemethane-sulfonate (CB154) on the release of prolactin and growth hormone by functional pituitary tumor cells in culture, J Cell Physiol 82: 401,1973. Bancroft, F. C, L. Levine, and A. H. Tashjian, Jr., Control of growth hormone production by a clonal strain of rat pituitary cells, J Cell Biol 43: 432, 1969. Tashjian, A. H., Jr., F. C. Bancroft, and L. Levine, Production of both prolactin and growth hormone by clonal strains of rat pituitary tumor cells, J Cell Biol 47: 61, 1970. Dannies, P. S., and A. H. Tashjian, Jr., Release and synthesis of prolactin by rat pituitary cells are regulated independently by thyrotropin-releasing hormone, Nature 261: 707,1976. Puck, T. T., S. J. Cieciura, and A. Robinson, Genetics of somatic mammalian cells. III. Long-term cultivation of euploid cells from human and animal subjects, JExp Med 108: 945,1958. Lowry, O. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall, Protein measurement with the Folin phenol reagent, J Biol Chem 193: 265,1951.
medullary secretion, JPhysiol 175: 231, 1964. Masur, S. K., E. Holtzman, and F. C. Bancroft, Localization within cloned rat pituitary cells of material that binds antigrowth hormone antibody, J Histochem Cytochem 22: 385, 1974. Hinkle, P. M., and A. H. Tashjian, Jr., Receptors for thyrotropin-releasing hormone in prolactin-producing rat pituitary cells in culture, J Biol Chem 248: 6180, 1973. Gautvik, K. M., and A. H. Tashjian, Jr., Effects of Ca++ and Mg++ on secretion and synthesis of growth hormone and prolactin by clonal strains of pituitary cells in culture, Endocrinology 92: 573,1973. Ostlund, R. E., Jr., Contractile proteins and pancreatic beta-cell secretion, Diabetes 26: 245, 1977. Carlson, H. E., I. K. Mariz, and W. H. Daughaday, Thyrotropin-releasing hormone stimulation and somatostatin inhibition of growth hormone secretion from perfused rat adenohypophyses, Endocrinology 94: 1709,1974. Gourdji, D., B. Kerdelhue, and A. Tixier-Vidal, Ultrastructure d'un clone de cellules hypophysaires secretant de la prolactin (clone GH3). Modifications induites par l'hormone hypothalamique de liberation de l'hormone thyreotrope (TRF), C R Acad Sci (Paris) 274: 437,1972.
22.
23.
24.
25. 26.
27.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 03:59 For personal use only. No other uses without permission. . All rights reserved.
Vol. 103, No. 4 Printed in U.S.A.
Acute Stimulated Hormone Release from Cultured GH3 Pituitary Cells* RICHARD E. OSTLUND, JR., JOYCE T. LEUNG, SHIRLEY VAEREWYCK HAJEK, THOMAS WINOKUR, AND MARK MELMAN Metabolism Division, Washington University School of Medicine, St. Louis, Missouri 63110 ABSTRACT. Treatment of cultured rat pituitary GH3 cells with 50 mM KC1 in growth medium released 33% of cell PRL and 18% of cell GH with a half-time of 5 min. Hormone in the culture medium was increased 2- to 4fold over unstimulated levels. The response required calcium; barium and strontium, but not magnesium, could substitute for calcium. Low temperature completely inhibited hormone release, which was also reduced significantly by inhibitors of energy metabolism and by nitrogen. This acute response was similar in ionic requirements, hormones released, and time course to the acute effect of TRH. Like potassium stimulation, TRH
resulted in acute release of both PRL and GH. This contrasts with the finding that chronic TRH treatment reduced GH synthesis in GH3 cells. After a 10-min preincubation with potassium, subsequent short incubations with potassium released little hormone unless the cells were allowed to recover by incubation in normal medium for at least 2 h. This acutely releasable hormone pool seems to be located in a membrane-bound subcellular fraction, since GH3 cells did not discharge the cytoplasmic marker enzyme, lactate dehydrogenase, during potassium-stimulated hormone release. (Endocrinology 103: 1245, 1978)
M
and derivatives of cAMP (7-10). Since GH and P R L are stable in the culture medium, hormones secreted over several days are a reflection of synthesis (11, 12). Acute release of preformed hormone has also been reported from GH3 cells. T R H caused detectable P R L output into the medium after a 10-min incubation, and this rose to a 50% increase over basal levels by 60 min (9, 13). Incubation of GH 3 cells for 3 h with medium containing 50 mM KC1 caused a 70% depletion of intracellular GH and P R L (3). Although a statistically significant corresponding increase in the relatively large medium hormone content was not seen (unless the cells were pretreated with T R H ) , it was suggested that release of preformed hormone occurred. Because of our interest in the mechanism of hormone extrusion from secretory cells, the acute stimulated release of hormones from Received November 14,1977. Address requests for reprints to: Dr. Richard E. Os- GH 3 cells was studied. Both T R H and a high tlund, Jr., Washington University School of Medicine, concentration of potassium in the presence of Department of Internal Medicine, Metabolism Division, Barnes and Wohl Hospitals, 660 South Euclid Avenue, St. calcium were found to release substantial amounts of intracellular GH and P R L over Louis, Missouri 63110. * This work was supported by grants from the Ameri- minutes. T h e effect of T R H on acute GH can Diabetes Association, the Diabetic Children's Welfare Fund of the St. Louis Diabetes Association, and Grants release was qualitatively different from the chronic effect on hormone synthesis. AM-20421 and AM-20579 from the NIH.
OST secretory tissues require extracellular calcium for hormone release (1, 2). Net entry of calcium into the cell can be achieved by the addition of depolarizing amounts of potassium to the incubation medium. Hence, solutions containing potassium and calcium are able to stimulate endocrine secretion (1-3). Such a model may aid understanding of normal hormone secretion, since calcium also enters endocrine cells during physiological secretory stimulation (4, 5). GH3 cells are an established cloned tissue culture line derived from a rat pituitary tumor which have become an important source for studying the molecular biology of secretion (6, 7). GH and PRL are secreted in the basal state and respond appropriately to the administration over several days of stimulators and suppressors, such as TRH, cortisol, bromcryptine,
1245 The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 03:59 For personal use only. No other uses without permission. . All rights reserved.
OSTLUND, JR., ET AL.
1246
Materials and Methods
Endo • 1978 Vol 103 • No 4
Assays
were used 1 day after the last medium change. Cells were also grown in a 250-ml spinner vessel (Bellco) with Eagle's spinner medium (Microbiological Associates), which contains the same serum and antibiotics as monolayer cells (16). Cell density varied from 200,000-800,000/ml on the day of harvest. Half of the suspension was removed and replaced with fresh medium three times weekly.
Rat PRL and GH were determined by specific double antibody RIA with reagents provided by Dr. A. F. Parlow through the NIH Rat Pituitary Hormone Distribution Program. Twenty microliters of medium were added to an assay volume of 0.59 ml. Growth medium was found to reduce the binding of GH to antibody and, consequently, the medium content of all tubes in the GH assay was adjusted to 20 JLJ. Hormone measurement in cells was accomplished by treating the cell sonicate made in distilled water with sodium hydroxide to achieve a final concentration of 0.01 N. After 15 min at room temperature, the basic solution was diluted into the immunoassay tubes. Failure to treat with sodium hydroxide and immediately dilute often resulted in lower values for cell hormone recovery. Such treatment did not affect hormone standards. Lactate dehydrogenase (LDH) was determined spectrophotometrically and LDH units were calculated as described (17). The assay was performed in a 0.5-ml cuvette using 0.1 ml sample in order to increase sensitivity. A cell extract of LDH activity in 0.15 M NaCl was recovered with 92 ± 1.2% efficiency when added to growth medium, where it was stable with incubation for over 1 h at 37 C or overnight at 4 C.
Acute hormone release
Chemicals
Monolayer cells on 100-mm plastic dishes were washed six times with warm Puck's saline G (14) and then aspirated to dryness with a Pasteur pipet. Two milliliters of Ham's F-10 medium, containing serum, antibiotics, and 2 mM CaCl2 in addition to that in the Ham's formulation and serum (calciumaugmented growth medium), were added, and the cells were incubated at 37 C for 10 min in an atmosphere of 95% air and 5% CO2. Serum-free medium could also be used. After incubation, the medium was aspirated and frozen, and the cells were washed six times with 0.15 M NaCl at 4 C and frozen. Cell protein was determined by the method of Lowry et al. (15) after scraping the thawed cells into 2.5 ml distilled water and sonicating them for 1.5 min. Cells grown in suspension culture were washed by sedimentation for 2 min at top speed in a benchtop IEC clinical centrifuge and were resuspended immediately, since storage of the washed cells in a pellet resulted in reduced secretory response. Results are presented as nanograms of hormone released per mg cell protein ± SEM. Statistical comparisons were done with Student's two-tailed unpaired t test.
Chemicals were purchased from the following suppliers: TRH, Abbott; cortisol sodium succinate, Upjohn; ethylenebis-(oxyethylenenitrilo) tetraacetic acid, Eastman; strontium chloride, Merck; barium chloride, Mallinckrodt; calcium chloride, potassium chloride, and sodium chloride, Fisher Co.
Cells GH3 cells (CCL 82.1) were purchased from the American Type Culture Collection, Rockville, MD. The cells were cultivated in 35- or 100-mm plastic dishes with growth medium, consisting of Ham's F10 medium (Microbiological Associates) containing 15% horse serum and 2.5% fetal bovine serum (Gibco), 50 U/ml penicillin, 50 jug/ml streptomycin, and 1.25 jug/ml amphotericin B (Fungizone, E. R. Squibb and Sons, Inc.). The presence or absence of amphotericin did not affect hormone release. Ham's medium contains 127 mM sodium, 3.8 mM potassium, and 0.3 mM calcium as well as other components. Cultures were plated at 106 cells/100-mm dish and incubated at 37 C in a humidified atmosphere of 5% CO2 and 95% air for 7-21 days. The medium was changed twice weekly, and the cells
Results Acute hormone release in response to potassium and calcium Although GH3 cells have a high basal rate of PRL and GH secretion, a 2- to 4-fold increase in the release of both was seen when cells were incubated for 10 min with 50 mM KC1 in calcium-augmented growth medium compared to incubation in calcium-augmented medium alone (Table 1). This response was seen with GH3 cells grown in spinner suspension culture as well as cells attached in monolayers to plastic dishes. The release phenomenon as a function of time is plotted in Fig. 1. There was an initial
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 03:59 For personal use only. No other uses without permission. . All rights reserved.
GH, HORMONE RELEASE
1247
TABLE 1. Acute potassium-stimulated hormone release PRL
GH
+KC1 -KCl +HC1 -KCl 22.6 ± 1.6 86.9 ± 10.3 108 ± 22 Monolayer cells 285 ± 23 71.4 ± 3.6 Spinner cells 42.3 ± 4.8 111 ± 3.5 203 ± 7.4 GH3 cells grown in either monolayer or spinner culture were washed and incubated 10 min with 2 ml calciumaugmented growth medium containing either 50 mM KCl (+KC1) or no addition (—KCl). At the end of the experiment, the medium was aspirated and analyzed for hormone content. Means and SES for triplicate determinations are presented. GH and PRL secretion was significantly increased by potassium in both spinner and monolayer cells (P < 0.01). Values are expressed as nanograms of hormone per mg cell protein • 10 min. TABLE 2. Effect of cations on potassium-stimulated PRL release 80
PRL released (ng/ %of no mg protein • 10 addition < '5 min) 6 2 60 100 None 24.9 ± 1.67 189 50 mM KCl 47.1 ±2.5" 106 50 mM NaCl 26.3 ±3.9 2 o) 40 50 mM KCl + 1 mM Na 18.1 ±0.4" 73 EGTA 20 283 50 HIM KCl + 1 mM Na 70.5 ±2.9" EGTA + 2 mM CaCl2 294 50 HIM KCl + 1 mM Na 73.1 ±3.5" EGTA + 2 mM BaCl2 20 40 60 80 222 50 mM KCl + 1 mM Na 55.4 ±0.7" MINUTES EGTA + 2 mM SrCl2 FIG. 1. Time course of PRL secretion. Confluent GH3 50 mM KCl + 1 mM EGTA 26.1 ±2.9 105 cells raised in 100-mm dishes were washed six times with + 2 mM MgCl2 saline G, and 5 ml calcium-augmented growth medium Triplicate 100-mm dishes of GH cells were washed with and without 50 mM KCl were added. The cells were and incubated with growth medium3 containing 0.5 mM placed in a 37 C CO2 incubator, and 50-fil aliquots were CaCl2 and the additions indicated for 10 min, after which taken for hormone assay at the Indicated times. Each the medium was removed and assayed. point represents the mean of duplicate dishes (control, ° P < 0.02, compared to no addition. Addition
• — • ; 50 mM KCl, O—O).
burst of PRL release, which was demonstrable as early as 2.5 min after the addition of 50 mM KCl and which became half-maximal at 5 min. After 20 min, there was little difference in continuing secretory rate between dishes treated with KCl and controls. The PRL present at time zero may represent secreted hormone adsorbed to the cells which is capable of release by the protein-containing incubation medium but not by the washing buffer. The incubation medium in these experiments was Ham's F-10, containing antibiotics, 15% horse serum, 2.5% fetal bovine serum, and 2 mM CaCl2 in addition to the 0.3 mM CaCl2 of Ham's salt formulation and that in the serum. The total calcium content measured by atomic absorption spectroscopy was 2.5 mM.
Effect of cations The specificity of potassium-stimulated hormone release was examined. As shown in Table 2, 50 mM NaCl in growth medium had no effect upon hormone release, suggesting that the osmotic effect of extra KCl does not contribute to hormone release. Extracellular calcium was essential for potassium to stimulate hormone release. The calcium chelator, EGTA, at 1 mM reduced potassium-stimulated PRL release to 38% of the expected release, a level lower than that seen in untreated controls (Table 2). However, this was not merely a toxic effect of EGTA, since potassium-stimulated secretion was restored if calcium was added in excess of EGTA. Excess strontium and barium also restored stimulated hormone
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 03:59 For personal use only. No other uses without permission. . All rights reserved.
OSTLUND, JR., ET AL.
1248
TABLE 3. The effect of calcium on stimulated PRL release into the medium Final Ca concentration
PRL release
(HIM)
-KC1 +50 mM KC1 100 ± 8 320 ± 11 1.5 93 ± 6 307 ± 14 2.5 120 ± 11 379 ± 26 4.5 119 ± 10 339 ± 20 8.5 99 ± 5 361 ± 16 GH3 cells were washed and incubated for 10 min with growth medium containing CaCl2 added to achieve the amounts indicated. Growth medium itself contained 0.5 mM CaCl2. Numbers from three to five dishes are expressed as a percentage of control release without added calcium or potassium. 0.5 No added CaCl2
TABLE 4. Acute hormone release by TRH Hormone released (ng/mg cell protein • 30 min)
PRL
GH
No addition 120 ± 2.3 139 ± 14 +50 nM TRH 273 ± 15° 375 ± 28" Triplicate 100-mm dishes containing cells were washed and 2.0 ml calcium-augmented growth medium were added with or without 50 nM TRH. The cells were placed in a CO2 incubator at 37 C for 30 min, after which the medium was removed and assayed. ° P < 0.01 difference from no addition.
release in EGTA-treated cells, but magnesium was without effect. The addition of up to 8 mM calcium to growth medium had little effect on hormone release in the presence or absence of excess KC1, but the greatest stimulated response occurred at 2.5 mM calcium (Table 3). Effects of TRH
Endo • 1978 Vol 103 • No 4
indicating that the GH synthetic machinery was capable of stimulation. TRH increased chronic PRL synthesis, as expected, to 206% of control. Hence, TRH rapidly released stored GH while inhibiting chronic synthesis of the same hormone. Calcium dependence of TRH-mediated hormone release Acute TRH-stimulated PRL and GH release was found to require extracellular calcium in a fashion very similar to acute potassium-stimulated release. In Table 6, it is seen that within 30 min TRH produced an 81% increase in PRL release into the medium. TABLE 5. Effect of chronic treatments on hormone synthesis jug Hormone/mg cell protein • 3 days GH PRL 49.7 ± 4.5 10.8 ± 2.3 35.0 ± 2.2" 22.2 ± 3.2° 223 ±42° Dishes (100 mm) of cells were incubated with the indicated material in growth medium for 3 days. The medium was aspirated and replaced, and the cells were incubated another 3 days. Hormones were measured in the medium from the final 3-day period. Numbers are from three to five experiments. a P < 0.05, compared to no addition. No addition 50 nM TRH 5 JUM Cortisol
TABLE 6. The effect of calcium on TRH-stimulated hormone release ng/mg Cell protein • 30 min Addition
PRL GH TRH is known to release PRL from both None 495 ± 58 604 ±45 rat pituitary (18) and GH3 cells (9). The data 894 ± 64° 50 nM TRH 1130 ± 66° of Table 4 confirm this, showing a 2.3-fold 1 mM EGTA 203 ± 35" 743 ± 118 661 ± 78 increase of PRL release in response to 50 nM 50 nM TRH + 1 mM 318 ± 42 EGTA TRH over 30 min. Unexpectedly, however, it 50 nM TRH + 1 mM 1007 ± 67" 1157 ± 161° was found that TRH also caused a 2.7-fold EGTA + 2 mM CaCl2 954 ± 71" 50 nM TRH + 2 mM 1120 ± 176" increase in GH release. CaCl2 In contrast to acute hormone release, the 2 mM CaCl 2 433 ± 73 768 ± 69 chronic synthesis of hormone can be measured Triplicate 35-mm dishes of cells were prepared and from medium hormone accumulation over a washed, as described in Materials and Methods. At time 3-day period (11, 12). TRH reduced chronic zero, 0.5 ml growth medium (0.5 mM CaCl2) was replaced synthesis of GH to 70% of the control value with or without additives. The cells were placed in a CO2 at 37 C for 30 min, after which the medium was (Table 5). However, a 4.5-fold increase in GH incubator removed and assayed. a synthesis was seen when cortisol was added, P < 0.05, compared to no additions.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 03:59 For personal use only. No other uses without permission. . All rights reserved.
GH3 HORMONE RELEASE
TRH-stimulated PRL release was reduced to 36% of the expected levels by EGTA, less than that of untreated controls, but was restored when calcium was added to excess of EGTA to the TRH-containing medium. Similar results were found for GH release. It should be noted that the basal release of PRL was reTABLE 7. Energy metabolism and acute hormone release Preincuba.. . , tion period
Inhibitor
(mm)
Potassiumstimulated PRL
release
(% control)"
4C° Spinner cells Monolayer cells
30 10
-2.7 ± 11 1.5 ± 4.5
Dinitrophenol (5 mM)
15
22.4 ± 11
KCN (5 mM)
15
36.6 ± 5.7
0
Nitrogen atmosphere 47.9 ± 9.6 30 Triplicate dishes were washed and incubated in calcium-augmented growth medium, as described in Materials and Methods, with and without the inhibitors. After the stated time, the medium was removed and replaced with fresh medium containing inhibitor with or without 50 mM KC1 and the cells were further incubated for 10 min. The medium was then removed and analyzed for hormone content. Experiments involving nitrogen were modified to eliminate even transient exposure to air. Nitrogen continuously flowed into a plastic centrifuge tube with two holes punched in the cap. At the end of preincubation, centrifugation to remove the medium was omitted and KC1 was added to the appropriate tubes through the second cap hole. At the completion of the experiment, secretion was stopped by placing the tubes at 4C. ° (Potassium-stimulated release — nonstimulated release) inhibitor present/(Potassium-stimulated release — nonstimulated release) inhibitor absent X 102. * Cells at 4 C and controls at 37 C were incubated in air with medium adjusted to pH 7.3 with dilute HC1. c Incubated in air or nitrogen at pH 7.3.
1249
duced to 41% of the expected levels by EGTA, while the basal release of GH was not significantly affected by EGTA. Energy dependence of stimulated hormone release Inhibitors of energy metabolism resulted in reduced potassium-stimulated hormone release, as shown in Table 7. Cells were preincubated with or without inhibitors for the specified time before being incubated in the presence or absence of 50 mM KC1 with or without inhibitors. The ability of potassium to stimulate PRL release in the presence of an inhibitor was expressed as a percentage of the effect of potassium to stimulate release in cells never exposed to inhibitor. Incubation at 4 C completely eliminated potassium-stimulated PRL release. In addition, the unstimulated baseline secretory rate was reduced by 4 C incubation to 25.6% and 6.4% of that seen at 37 C for spinner and monolayer cells, respectively. Dinitrophenol, potassium cyanide, and a nitrogen atmosphere all significantly inhibited potassium-stimulated release. Recovery of hormone-releasing capacity after potassium stimulation Refractoriness to further potassium-stimulated PRL release after the initial hormone burst is seen in Fig. 1. This was not due to total cell hormone depletion. Monolayer cells were stimulated 10 min with 50 mM KC1 and then hormones were measured in the medium and in the washed cells (Table 8). Net stimulated PRL release (defined as medium hor-
TABLE 8. Hormone balance with potassium stimulation PRL -KC1
GH +KC1
-KC1
+KC1
Hormone content (ng/mg) protein Medium 52.3 ± 3 213 ± 8 102 ± 18 320 ± 33 357 ± 23 253 ± 12 1078 ± 62 888 ±59 Cells Medium + cells 409 ±22 466 ± 19 1180 ± 75 1208 ± 48 8.7 Hormone release (medium x 102)/ 12.8 45.7 26.5 (medium + cells) (%) Net potassium-stimulated hormone 32.9 17.8 release (+KC1) - (-KC1) (%) Triplicate monolayer GH3 cells in 100-mm dishes were washed and stimulated for 10 min with 50 mM KC1, as described in Materials and Methods. After stimulation, the hormone content of the medium and the washed cells was determined.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 03:59 For personal use only. No other uses without permission. . All rights reserved.
OSTLUND, JR., ET AL.
1250
mone in the presence of 50 mM KCl less medium hormone in the absence of added KCl) was 32.9% of total PRL (medium plus cells). Similar calculations for GH showed a 17.8% release. To determine how fast the acutely releasable cell hormone pool is regenerated, the following experiment was performed. Dishes of GH3 cells were stimulated with 50 mM potassium chloride, allowed to recover by incubation in the absence of potassium chloride, and then restimulated with potassium chloride. The results are plotted in Fig. 2. During the initial stimulation, PRL release was 64.3 ± 10 ng/mg protein • 10 min higher in dishes receiving potassium chloride than in control dishes. When a group of dishes was then immediately restimulated with potassium chloride, the difference in PRL release between stimulated and nonstimulated dishes was reduced to 17 ± 6.7 ng/mg protein-10 min (P < 0.02). After 1 h of recovery in nonstimulating medium, potassium still had little effect, but partial restoration of the potassium response to approximately 40% of the expected response took place after 2 and 3 h of recovery incubation (26.6 ± 4.2 and 23.0 ± 5.8 ng/mg protein 10 min, respectively).
£J
2
< o ^ ~ o Z 2
0
1
2
3
RECOVERY TIME (hr.)
FIG. 2. Recovery of capacity to release PRL in response to KCl. Four sets of triplicate GH3 cells in 100-mm dishes were washed and stimulated with 50 mM KCl (see Materials and Methods) from time -10 min to time zero. At time zero, the medium was removed, the cells were washed twice with saline G, and the remaining fluid was aspirated. One set of cells was immediately restimulated with KCl for 10 min beginning at time zero. Other sets of cells were incubated for 1-3 h in medium not containing excess KCl, after which the medium was removed and the cells were washed twice with saline G and restimulated with KCl. Control cells received the same washings and medium changes, but no KCl was present at any time (control, • — • ; 50 mM KCl, O—O).
Endo • 1978 Vol 103 • No 4
TABLE 9. Release of LDH during acute potassium stimulation U/mg 627 ± 16
% Total LDH 100 ± 2.6
Total cell LDH content LDH released into medium during potassium stimulation 13 ± 3 2.1 ± 0.5 50 mM KCl No added KCl 23 ± 8 3.7 ± 1.3 LDH was determined on the cells and media from Table 8.
Release of LDH Release of the cytoplasmic marker enzyme, LDH, during potassium-stimulated hormone release was determined, and the results are presented in Table 9. The total cell LDH content was measured in the supernatant from cells homogenized in 2.5 ml 0.15 M NaCl and centrifuged at 40,000 X g for 20 min. During the 10-min potassium stimulation period, less than 2.1% of cell LDH was released into the incubation medium. Control dishes not containing added KCl released 3.7% into the medium. Both of these values are much less than the percentage of release of PRL and GH. Discussion We have found that PRL and GH are actively released from GH3 cells by potassium in the presence of calcium. The process had a half-time of 5 min. This phenomenon is consistent with the previously reported finding that GH3 cells contain less hormone after incubation with 50 mM potassium chloride for 3 h (3). By incubating for a much shorter time, hormone release relative to unstimulated controls was enhanced, and the medium hormone content was increased several fold. PRL release in the presence of both 50 mM potassium (Table 2) and 50 nM TRH (Table 6) was calcium dependent, being reduced below the release found in untreated dishes by an excess of the calcium chelator, EGTA. Release was restored if calcium was added in excess of EGTA. However, basal hormone release, especially of GH, had a significant component which was not calcium sensitive (Table 6). This is consistent with a previous report that calcium had little effect on basal hormone release in GH3 cells (24). Because of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 03:59 For personal use only. No other uses without permission. . All rights reserved.
GH3 HORMONE RELEASE
its calcium sensitivity, stimulated GH3 hormone release may more closely resemble that occurring in normal pituitary tissue. Acute potassium-stimulated hormone release was similar to the acute but not the chronic effects of TRH. TRH released both PRL and GH when incubated with cells for 30 min (Table 4), but it inhibited GH synthesis while stimulating PRL synthesis when incubated with cells for 6 days (Table 5). The acute TRH response was also dependent upon extracellular calcium (Table 6). The dual effect of TRH is consistent with other GH cell data. Tritiated TRH bound reversibly to cell surface receptors and could be easily displaced by unlabeled TRH after 90 min (23), but after 24 h, binding became irreversible and the increase in PRL synthesis persisted for at least a week even if TRH was withdrawn from the medium (7). Both potassium and TRH probably acutely increase cell membrane permeability to calcium, resulting in activation of the release mechanism. The chronic effects of TRH may then be exerted directly or indirectly on the cell nucleus or on translation of messenger RNA. The acute release of GH by TRH is similar to that reported in acromegalic subjects (19) and in vitro perfused normal rat hemipituitaries (26). Since GH3 cells are transformed, it should be expected that their mode of secretion might differ from normal. For example, GH3 cells have very few secretory granules and those are quite small (References 22 and 27, and unpublished observations). GH localized by immunological techniques in GH cells, is found in what appears to be endoplasmic reticulum and in the perinuclear area (22). Nevertheless, the hormone does not seem to be diffusely distributed in the cytoplasm. Acute stimulated secretion, therefore, may involve fusion of hormone-containing vesicles with the plasma membrane. This hypothesis is supported by our finding that release of the cytoplasmic marker enzyme, LDH, was not increased after potassium-stimulated secretion (Table 9). Hence, stimulated hormone release is not merely due to cell "leakage." Not all cell hormone is acutely releasable. In fact, less than half of the PRL and GH is extruded after potassium addition. What
1251
causes some immunoreactive hormone to be releasable is not known. It may be an alteration in the hormone-containing organelles, such as distance from or ability to fuse with the plasma membrane. Alternatively, the nonreleasable hormone may simply lack sufficient cytosol calcium as a secretory trigger because of changing plasma membrane cation fluxes over time. Our data demonstrate that storage of hormone and compartmentation of hormone within the cell are functions that can be performed by cells having very few classical secretory granules. These functions are probably achieved by means of tiny granules or membrane-bound hormone-containing vesicles, such as those of the endoplasmic reticulum or Golgi apparatus. Calcium seems to be the final common pathway of the acute stimulated release phenomenon. It is intriguing that barium and strontium, but not magnesium, will substitute for calcium in supporting potassium-stimulated release. This ion specificity is known to be associated with other secretory phenomena (2, 21) and with the ability to cause muscle contraction (20). Calcium may stimulate secretion by acting on a variety of contractile proteins, such as actomyosin or tubulin, known to be present in secretory tissues (25). The finding that potassium-stimulated hormone release was sensitive to cold, nitrogen, dinitrophenol, and KCN suggests that it was dependent upon cell energy stores. This is also consistent with the hypothesis that contractile proteins are involved in the acute release process. If so, the GH3 system with a theoretically limitless supply of cloned cells could prove very useful for the investigation of those proteins. References 1. Rubin, R. P., Calcium and the Secretory Process, Plenum Press, New York, 1974. 2. Douglas, W. W., Stimulus-secretion coupling: the concept and clues from chromaffin and other cells, Br J Pharmacol 34: 451, 1968. 3. Gautvik, K. M., and A. H. Tashjian, Effects of cations and colchicine on the release of prolactin and growth hormone by functional pituitary tumor cells in culture, Endocrinology 93: 793,1973. 4. Douglas, W. W., and A. M. Poisner, On the mode of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 03:59 For personal use only. No other uses without permission. . All rights reserved.
1252
5. 6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
OSTLUND, JR., ET AL.
acetylcholine in evoking adrenal medullary secretion: 16. increased uptake of calcium during the secretory response, JPhysiol 162: 385, 1962. Malaisse, W. J., Role of calcium in insulin secretion, 17. IsrJMedSci 8: 244, 1972. Tashjian, A. H., Jr., Y. Yasumura, L. Levine, G. H. Sato, and M. L. Parker, Establishment of clonal strains of rat pituitary tumor cells that secrete growth 18. hormone, Endocrinology 82: 342, 1968. Tashjian, A. H., Jr., and R. F. Hoyt, Jr., Transient controls of organ-specific functions in pituitary cells in culture, In Sussman, M. (ed.), Molecular Genetics and Developmental Biology, Prentice-Hall, Engle- 19. wood Cliffs, 1972, p. 353. Dannies, P. S., and A. H. Tashjian, Jr., Effects of thyrotropin-releasing hormone and hydrocortisone on synthesis and degradation of prolactin in a rat pitui- 20. tary cell strain, J Biol Chem 248: 6174, 1973. Dannies, P. S., K. M. Gautvik, and A. H. Tashjian, Jr., A possible role of cyclic AMP in mediating the 21.
Endo i 1978 Vol 103 , No 4
Bancroft, F. C, and A. H. Tashjian, Jr., Growth in suspension culture of rat pituitary cells which produce growth hormone and prolactin, Exp Cell Res 64: 125, 1971. Wacker, W. E. C, D. D. Ulmer, and B. L. Vallee, Metalloenzymes and myocardial infarction. II. Malic and lactic dehydrogenase activities and zinc concentrations in serum, N EnglJMed 255: 449, 1956. Rivier, C, and W. Vale, In vivo stimulation of prolactin secretion in the rat by thyrotropin-releasing factor, related peptides, and hypothalamic extracts, Endocrinology 95: 978, 1974. Coutant, G., M. Vandeweghe, and A. Vermeulen, Comparison of TRF, propranol-glucagon, insulin, and glucose stimulation tests in acromegaly, Horm Metab Res 9: 17, 1977. Caldwell, P. C, and G. Walster, Studies on the microinjection of various substances into crab muscle fibers, JPhysiol 169: 353,1963. Douglas, W. W., and R. P. Rubin, The effects of
effects of thyrotropin-releasing hormone on prolactin
alkaline earths and other divalent cations on adrenal
release and on prolactin and growth hormone synthesis in pituitary cells in culture, Endocrinology 98: 1147,1976. Gautvik, K. M., R. F. Hoyt, Jr., and A. H. Tashjian, Jr., Effects of colchicine and 2-Br-a-eryocryptinemethane-sulfonate (CB154) on the release of prolactin and growth hormone by functional pituitary tumor cells in culture, J Cell Physiol 82: 401,1973. Bancroft, F. C, L. Levine, and A. H. Tashjian, Jr., Control of growth hormone production by a clonal strain of rat pituitary cells, J Cell Biol 43: 432, 1969. Tashjian, A. H., Jr., F. C. Bancroft, and L. Levine, Production of both prolactin and growth hormone by clonal strains of rat pituitary tumor cells, J Cell Biol 47: 61, 1970. Dannies, P. S., and A. H. Tashjian, Jr., Release and synthesis of prolactin by rat pituitary cells are regulated independently by thyrotropin-releasing hormone, Nature 261: 707,1976. Puck, T. T., S. J. Cieciura, and A. Robinson, Genetics of somatic mammalian cells. III. Long-term cultivation of euploid cells from human and animal subjects, JExp Med 108: 945,1958. Lowry, O. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall, Protein measurement with the Folin phenol reagent, J Biol Chem 193: 265,1951.
medullary secretion, JPhysiol 175: 231, 1964. Masur, S. K., E. Holtzman, and F. C. Bancroft, Localization within cloned rat pituitary cells of material that binds antigrowth hormone antibody, J Histochem Cytochem 22: 385, 1974. Hinkle, P. M., and A. H. Tashjian, Jr., Receptors for thyrotropin-releasing hormone in prolactin-producing rat pituitary cells in culture, J Biol Chem 248: 6180, 1973. Gautvik, K. M., and A. H. Tashjian, Jr., Effects of Ca++ and Mg++ on secretion and synthesis of growth hormone and prolactin by clonal strains of pituitary cells in culture, Endocrinology 92: 573,1973. Ostlund, R. E., Jr., Contractile proteins and pancreatic beta-cell secretion, Diabetes 26: 245, 1977. Carlson, H. E., I. K. Mariz, and W. H. Daughaday, Thyrotropin-releasing hormone stimulation and somatostatin inhibition of growth hormone secretion from perfused rat adenohypophyses, Endocrinology 94: 1709,1974. Gourdji, D., B. Kerdelhue, and A. Tixier-Vidal, Ultrastructure d'un clone de cellules hypophysaires secretant de la prolactin (clone GH3). Modifications induites par l'hormone hypothalamique de liberation de l'hormone thyreotrope (TRF), C R Acad Sci (Paris) 274: 437,1972.
22.
23.
24.
25. 26.
27.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 03:59 For personal use only. No other uses without permission. . All rights reserved.
