Gastrin release from dispersed gastrinoma cells: Effects of calcium and calcium lonophore (A23187)

JOURNAL

OF SURGICAL

RESEARCH

39, 331-337 (1985)

Gastrin Release from Dispersed Gastrinoma Ceils: Effects of Calcium and Calcium lonophore (A231 87)‘~~ EUGENE A. WOLTERING, M.D.,3 E. CHRISTOPHER ELLISON, M.D., THOMAS M. O’DORISIO, M.D., CHARLES HITCHCOCK, PH.D., KELLY ROBERTS, B.S., RALPH STEPHENS, PH.D., JOANN SPARKS, B.S., AND LARRY C. CAREY, M.D. Departments of Surgery, Pathology (Clinical Chemistry), and Medicine, The Ohio State University College of Medicine, Columbus, Ohio Submitted for publication November 10, 1983 Dispersed single-cell suspensionsof human gastrinoma tissue were incubated for 15,60, and 120 min in a calcium-containing medium, (0.1,2, 10 m&f) in calcium-free medium and in calcium-free medium containing the calcium ionophore A23 187 (0.01, 1, and 100&ml). Supematant and pellet (intracellular) gastrin levels were determined by radioimmunoassay. Supematant gas&in levels remained stable over 120 min in calcium chloride or calcium gluconate containing medium, while intracellular pellet gastrin approximately tripled during the same incubation period. Total gastrin (supematant plus pellet) approximately doubled during the 2-hr incubation in calcium. However, calcium (0.1,2, or 10 mM) failed to produce a dose-dependent rise in supematant, pellet, or total gastrin when compared to calcium-free incubates. Contrary to the expected gastrin response to calcium, supematant and pellet gastrin levels were higher in incubates in calcium-free medium than in calcium-containing incubates. A23 187 (0.01 or 1 m&ml) in a calcium-free medium decreasedsupematant gastrin while high dose ionophore (100 mcg/ml) increased supematant gastrin. All dosesof ionophore stimulated pellet and total gastrin levels. Thus, it appears that the clinical augmentation of gastrin levels, seenwith calcium challenge in vivo may not be solely due to changes in serum calcium. 0 1985 Academic press. hc.

role in the secretion of gastrin, several criteria should be fulfilled: (1) Gastrin release is deSerum gas&in levels in patients with the creased or abolished in the absence of extraZollinger-Ellison syndrome (ZE) rise in recellular calcium, (2) gastrin is released in a sponse to intravenous infusions of calcium dose-dependent fashion by increasesin extra(Passaro et al.) [ 141or in the presence of the cellular calcium, and (3) calcium ionophore high calcium levels that accompany hyper(A23187) must mimic the stimulatory effect parathyroidism [ 171.The mechanismsthrough seen with increasing extracellular calcium. which hypercalcemia increases serum gastrin We examined the role of calcium in the reare unknown, but it has been assumed to be lease of gastrin from an isolated single-cell the direct effect of calcium on the tumor cell. Calcium may stimulate gastrin synthesis, gas- suspension of gastrinoma tissue. We hypothesizedthat if the calcium-stimulated releaseof trin release, or both. gastrin was a direct result of calcium’s effect To demonstrate that calcium transport on the gastrinoma cell, then incubating gasacrossthe cellular membrane plays an essential trinoma cells with increasing extracellular cal’ Presented at the Annual Meeting of the Association cium should yield a dose-dependent increase in gastrin to the incubating media. Likewise, for Academic Surgery, San Antonio, Texas, October 3 lNovember 3, 1984 removal of calcium from the incubating media 2 Supported in part by the William Davis Fund, The should decreaseor abolish gastrin release.FiOhio State University, and the Department of Surgery, nally, we postulated that stimulation of gastrin Medical Research and Development Fund. release should be mimicked by the calcium 3 To whom reprint requestsshould be addressed:N-924 Doan Hall, Ohio State University Hospital, 410 W. 10th ionophore A23 187, an agent that increasesinAvenue, Columbus, Ohio 432 10-1228. tracellular calcium. INTRODUCTION

331

0022-4804/85 $1.50 Copyri@~t 0 1985 by Academic Press, Inc. All rights of reproduction in any form reserved.

332

JOURNAL OF SURGICAL RESEARCH: VOL. 39, NO. 4, OCTOBER 1985 MATERIALS

AND

TABLE 1

METHODS

A 44-year-old man underwent resection of an isolated pancreatic gastrinoma. Preoperatively the patient’s basal gastrin ( 1700 pg/ml) rose to 2700 pg/ml in response to a calcium bolus (1 g over 1 min) and to 3200 pg/ml following a calcium infusion of 15 mg/kg elemental calcium over 4-hr [ 151.Postoperatively his serum gastrin returned to normal levels and has remained so for 1 year. Histologic analysis confirmed the diagnosis of a non-P islet cell tumor. There was no evidence of local invasion or distant metastasis at the time of surgery and the patient continues to have no evidence of persistent or recurrent disease. Immunohistochemical stains of the tumor for gas&in (ABC technique) showed that approximately 20% of the tumor cells contained gastrin [9]. The tumor contained 10.5 ng gastrin per gram wet tumor weight. Immediately after resection the tumor was placed in cold (4°C) tissue culture medium (MEM B, GIBCO Co., Grand Island, N. Y.) containing 2 mM calcium, 1 mM magnesium, dispersedby vigorous cross-blademincing, and further dispersed with 0.2% Type IV collagenase (Worthington Diagnostics, Freehold, N. Y.) and 50 pg/ml of DNAase (Sigma, St. Louis, MO.) for 45 min in a 37°C shaking water bath according to the method of Brown et al. [l-4]. Cells were washed and resuspended to a concentration of 1.7 X lo6 cells/ml in a calcium-free, 1 mA4 magnesium-containing medium (medium 199, GIBCO Co., Grand Island, N. Y.) containing 0.02 M Hepes and 0.2% bovine serum albumin, pH 7.4. Cell suspensions were analyzed by trypan blue exclusion, and initial cellular viability was 90-95%. Cellular viability remained 90-95% throughout the 2-hr incubation based on control tube samples. Aliquots (0.9 ml) of cell suspensions were added to 1.5 ml polypropylene vials preloaded with 0.1 ml of test solution to attain a final cell concentration of 1.5 X lo6 cells/ml. Table 3 shows the test solution and final drug concentrations used. Triplicate vials for each test compound and incubation period were pre-

CLINICAL GASTRIN LEVELS (pgjml) AFTER CALCIUM BOLUS AND CALCIUM INFUSION Calcium bolus” Time (min) 0

5 10 15 30 60

Serum gastrin

Serum calcium

1700 2700 2250 2100 1900 2000

4.7 5.2 5.1 5.1 5.0 5.1

Calcium infusionb Serum gastrin

Serum calcium

1700

5.2

3000 3200

5.9 6.2

’ Ninety-three milligrams elemental calcium (I g calcium gluconate) over 2 min. b Fifteen milligrams per kilogram elemental calcium (calcium gluconate) as infusion.

pared and incubated at 37°C in a shaking water bath for 15, 60, or 120 min, then centrifuged for 10 min at 4°C ( 1OOOg).The supernatant and cell pellet were separated and frozen at -70°C until assayed.Control tubes were used to ascertain cell count and viability over the 2-hr incubation period. Cell pellets were extracted by boiling in 1 ml of distilled water for 30 min and residual cell protein was assayed by the method of Lowry [ 121.Incubates containing 10 mM calcium developed a precipitate (calcium phosphate) which was discovered to interfere with the Lowry assay. Seventy-two tubes that had been incubated in calcium-free medium with (A23 187) or 0.1 mMcalcium yielded proteins ranging from 30 to 40 pugprotein/ml with a mean of 33.6 + 3.1 pg/ml of protein. This value was used to correct all supernatant, pellet, and total gastrins from picograms per milliliter to picograms per micrograms protein. All gastrin assayswere performed on 0.2 ml of supernatant or cell extract using the BectonDickinson gastrin RIA kit with a 2-hr incubation period. This assay has a sensitivity of 15 pg/ml and standard curves generated with calcium-free medium 199 or standard kit buffer or medium 199 with 0.1, 2, or 10 mM Ca were equivalent (Table 2). Furthermore, media containing test substanceswithout cells

WOLTERING

ET AL.: GASTRIN RELEASE FROM GASTRINOMA

CELLS

333

Incubation in Calcium-Free Media

TABLE 2 RECOVERY OF GASTRIN (pg/ml)”

Ca (mM)

Gas&in standard:

100

500

1000

0.0 0.1 2.0 10.0 Standard kit diluent

Gastrin recovered:

105 109 110 106

57 1 526 513 499

1058 1064 1059 995

107

577

1085

Note. Gastrin RIA intrassay variation +lO%. a Medium 199 with 0.2% bovine serum albumin.

showed no binding or cross reactivity in this assay.All gastrin results are reported as picograms (pg) per microgram (pg) of cellular protein. Changes in supematant, pellet, and total gastrin concentrations were compared between each of the test substances, their concentrations, and incubation intervals.

We postulated that removal of calcium from the incubating medium should diminish or abolish gastrin response.However, gastrinoma cells incubated in a calcium-free, 1 rnM magnesium-containing media had higher levels of supernatant gastrin (Table 2), pellet gastrin (Table 4), and total gastrin (Table 5) than did the gastrinoma cells incubated in calciumcontaining media. Gastrin release was not augmented by incubation in calcium-free medium, rather gastrin releaseseemedto be unaffected by incubation in calcium-free media except in the earliest incubation period (Table 6). Incubation in Calcium Ionophore (A23187)

A23 187 (calcium ionophore) is known to increase calcium influx and to cause intracellular calcium redistribution [5, 61. We posRESULTS tulated that A23 187 in a calcium-free media might causea transient rise in free intracellular Incubation in Calcium Containing Media calcium and thus stimulate gastrin release.In Dispersed human gastrinoma cells incu- the dispersed cell suspension, low or interbated in calcium containing medium (0.1, 2, mediate concentrations of ionophore sharply 10 mM) showed stable levels of gastrin in the depressedsupernatant gastrin while high dose supernatant (approximately 20 pg/pg protein) ionophore increased supematant gastrin (Tathroughout the 120-min incubation period. ble 3). Pellet (intracellular) gastrin sharply inThe supematant gastrin values were similar creasedwhen gastrinoma cells were incubated whether the cells were incubated in calcium in A23 187 (Table 4). This effect was dose dechloride (CaC&) or calcium gluconate (CaGl) pendent in the early incubation period but this (Table 3). Pellet gastrin values increased by dose dependency was lost by 1 hr. Gastrin reapproximately 250% over the 2-hr period. lease was sharply suppressedby low or interCalcium concentration or calcium type (CaC12 mediate dosesof ionophore while the highest vs CaGl) did not produce significant differ- dose of ionophore produced releases values ences in intracellular gastrin (Table 4). Total approximating those seen in calcium-free or gastrin (supernatant plus pellet) doubled dur- calcium-containing incubates (Table 6). ing the incubation period, however, there was no calcium dosedependency in this rise (Table DISCUSSION 5). Releaseof gas&in (%) was calculated as supematant gas&in divided by total gastrin mulFailure of these gastrinoma cells to respond tiplied by 100 (S/T X 100). In the cells incu- to calcium with a sharp rise in supematant, bated in CaC12there was a dose-dependent in- cellular, or total gastrin might be related to creasein percentage gastrin releasein the 15 artifactual changesinduced by the mechanical min incubates but this effect was rapidly lost and enzymatic dispersion methods used to (Table 6). Incubation in calcium gluconate prepare single-cell suspensions. However, also produced increased releaseat 15 min but other investigators have shown that hormone this effect was not dose dependent (Table 6). or enzyme releasefrom other tissues prepared

Calcium free

83 + 10 142 xk 38 133 f 3.4

15 60 120

25 f 2.1 25 + 1.6 32 + 1.1

15 60 120

Time (min)

Calcium free

Time (min)

107 f 25 95+ 16 97 f 3.9

0.1

22 f 1.4 23 f 0.90 25 -c 0.90

2

36 f 7.0 96 iz 11 90+ 17

2

Calcium chloride (mM)

22 k 1.4 22 + 1.2 24 + 0.33

0.1

Calcium chloride (mM)

34 + 8.1 96k 12 86 f 4.6

10

TABLE 4

20 + 0.90 23 iz 2.9 24 + 3.3

0.1 23 -t 1.2 19 * 1.0 23 + 0.6

2

36 +- 6.1 92k 15 91 + 3.5

0.1

36 + 2.0 91 f 11 93* 17

2

Calcium gluconate (mM)

PELLETGASTRIN(~~/~L~PROTEIN)

24 f 1.1 20 + 0.33 23 + 0.33

10

10

22 f 1.7 22 + 2.5 22 + 0.6

10

34 + 4.0 93* 12 84 f 3.7

Calcium gluconate (mM)

TABLE 3

119 k 10 134 2 18.5 145 -+ 22.9

0.01

7 + 0.40 6 + 0.20 7 -t 0.60

0.01

125 + 13 127 + 9.2 109 j, 9.2

1

A23 187 (&ml)

7 f 0.20 5 f 0.30 6 + 0.33

1

A23 I87 (ps/ml)

60 + 25 126 + 6.2 77 f 21.8

100

44 + 8.6 33 + 0.60 30 + 3.8

100

WOLTERING

ET AL.: GASTRIN

RELEASE

FROM

GASTRINOMA

CELLS

335

in a similar manner correlated well with clinical observations. Brown et al. [2, 41 studied dispersedhuman parathyroid tissue and noted that parathyroid hormone [PTH] release by normal parathyroid cells was suppressed by increasesin ambient calcium. However, PTH releaseby hyperplastic parathyroid tissue was more sensitive to increases in extracellular calcium than was neoplastic tissue from parathyroid adenomas. Brown et al. [2] further noted that dispersed single cell suspensions provided a convenient, homogeneous preparation that permitted study of the effects of various pharmacologic agentson PTH release. Williams et al. [ 18, 191and Gardner [7] used dispersed pancreatic acinar cells to study amylase release. These pancreatic acinar cells maintained synthetic function and apparently normal release mechanisms. Saxe et al. [ 161 demonstrated increases in immunoreactive insulin releasefrom dispersed human insulinoma in responseto glucose challenges, similar to the clinical insulin response to a glucose tolerance test. Lichtenberger et al. [lo] noted that gastrin releasefrom a single cell suspension of normal antral G cells increased in the presenceof low levels (O-O.3 n&f) of calcium, but was inhibited when the extracellular calcium concentration rose above 2.4 mM. In our model, apparently benign neoplastic gastrin-producing cells, prepared as an isolated single-cell suspension released greater amounts of gastrin into the supernatant in calcium-free medium than in calcium-containing (0. l- 10 mM) medium. Assuming that there is a chemical and electrical gradient favoring the movement of calcium into the cell [ 191,increasing extracellular calcium should increase cytosolic calcium. It would appear in our dispersed cell model that increasing cytosolic calcium suppressed supernatant, pellet, and total gastrin levels. We also had postulated that removal of calcium from the incubating medium should decreaseor abolish the gastrin response.Our results failed to support our postulate and, indeed, incubation of gastrinoma cells in calcium-free medium resulted in higher su-

336

JOURNAL OF SURGICAL RESEARCH: VOL. 39, NO. 4, OCTOBER 1985 TABLE 6 PERCENTAGE GASTRINRELEASE’ Calcium chloride (mW

Calcium gluconate bw

A23 187 (&ml)

Time (min)

Calcium free

0.1

2

10

0.1

2

10

0.01

1

loo

15 60 120

23 15 20

17 19 20

37 19 22

41 18 21

36 20 15

38 17 20

39 19 21

6 4 5

5 4 5

21 21 28

’ Calculated as supematant gastrin divided by total gastrin and multiplied by 100.

pematant, pellet, and total gastrin values when compared to calcium-containing media. The removal of calcium from the extracellular incubating media may result in an early redistribution of intracellular gastrin or subsequent depletion of intracellular gastrin [ I3 1. If one assumesthe latter to be true, our results may indicate that supematant, pellet, and total gastrin are augmented by low levels of intracellular calcium and inhibited by higher levels of intracellular calcium content. Conversely, if one assumesthat depletion of extracellular calcium is accompanied by an intracellular releaseof bound calcium there may be a temporary increase in intracellular free calcium in cells incubated in calcium-free medium. Incubation

in Calcium Ionophore

We had also postulated that incubation of gastrinoma cells in A23 187 would result in a dose-dependent increase in gastrin release based on ionophores known ability to increase intracellular calcium by augmentation of calcium influx and by redistribution of intracellular calcium [5, 61. Our results, however, again support the concept that increasesin intracellular calcium may suppresssupernatant, pellet, and total gas&ins. Cytotoxicity has been reported in cells exposedto high dose(20 m&I) A23 187 and this cell lysis may explain the increase in supematant gastrin and gastrin release seen in the highest doses of ionophore [lo]. To investigate the toxicity of A23 187 on gastrinoma cells, we subsequently incubated long-term tissue cells derived from this tumor with the high dose of ionophore for 2 hr and

failed to demonstrate changes in cellular viability; however, supematant LDH or cyclic AMP levels (as markers of cell permeability and viability) were not measured. Our study has failed to demonstrate a dosedependent rise in gastrin in either the supernatant or in the pellet when gastrinoma cells were incubated in 0.1 to 10 mM calcium chloride or calcium gluconate. Furthermore, removal of calcium from the incubating medium did not diminish or abolish the gastrin response;rather it seemedto enhance both supernatant and pellet gastrin levels. Finally, the addition of nontoxic levels of A23 187 (0.01 or 1 pg/ml) to a calcium-free medium failed to stimulate supematant gastrin or gastrin releasebut did increase pellet (intracellular) gastrin. The contrast between the rise in gastrin in responseto calcium stimulation in the patient and the apparent suppression by calcium of supematant, pellet, and total gastrin experimentally suggeststhat intermediate steps or a preexisting hormone-peptide milieu may be required for gastrin release in vivo and these prerequisites may not be present in the dispersedcell preparation. The increasesin pellet and total gastrin indicate that these cells are functioning and this model appearsto be suitable for the study of hormone releasenot only in gastrin but in a variety of other endocrine tumors [I 1, 151. REFERENCES 1. Brown, E. M., Hurwitz, S., and Aurbach, G. D. Prep aration of viable isolated bovine parathyroid cells. Endocrinology 99: 1582, 1976.

WOLTERING

ET AL.: GASTRIN RELEASE FROM GASTRINOMA

2. Brown, E. M., Gardner, D. G., Brennan, M. F., et al.

Calcium-regulated parathyroid hormone release in primary hyperparathyroidism. Amer. J. Med. 66: 923, 1979. 3. Brown, E. M., Gardner, D. G., and Aurbach, G. D. Effectsof the calcium ionophore A23 187on dispersed bovine parathyroid cells. Endocrinology 106: 133, 1980.

4. Brown, E. M., Dawson-Hughes, B. F., Wilson, R. E., and Adragna, B. Calmodulin in dispersed human parathyroid cells. J. Clin. Endocrinol. 53: 1064, 1981. 5. Chandler, D. E., and Williams, J. A. Intracellular divalent cation releasein pancreatic acinar cells during stimulus-secretion coupling. J. Cell Biol. 76: 386, 1978. 6. Church, J., and Zsoter, T. T. Calcium antagonistic drugs. Mechanism of action. Canad. J. Physiol. Pharmacol. 58: 254, 1980. 7. Ellison, E. C., Wokering, E. A., Howe, B., Nelson, K. P., Stephens, R., Carey, L. C., and O’Dorisio, T. M. Effectof calcium on vasoactiveintestinal peptide release from an islet cell tumor associated with the Vemer-Morrison syndrome. Surg. Forum 34: 227, 1983. 8. Gardner, J. D., Costenbader, C. L., and Uhlemann, E. R. Effect of extracellular calcium on amylase release from dispersedpancreatic acini. Amer. J. Physiol. 236: E745, 1979. 9. Hsu, Raine, S. M., and Fanger, L. Use of avidin-biotin

peroxidase complex (ABC) in immunoperoxidase techniques: A comparison between ABC and unlabeled antibody (PAP) Procedure. J. Histochem. C’ytochem. 29: 577, 1981. 10. Lichtenberger, L. M., Shaw, L. S., and Bailey, R. B. Influence of calcium on the release of gastrin from

CELLS

337

isolated rodent G cells. Proc. Sot. Exp. Biol. Med. 166:587,1981.

11. Lebrun, P., Malaisse, W. J., and Herchuelz, A. Modalities of ghclazide-induced Ca*+ influx into the pancreatic B-cell. Diabetes 31: 1010, 1982. 12. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193: 265, 195I. 13. Malaisse, W. J., Herchuelz, A., Levy, J., and Sener, A. Calcium antagonists and islet function. III. The possible site of action of verapamil. Biochem. Pharmacol. 26: 735, 1977. 14. Passaro,E., Jr., Basso,N., and Walsh, J. H. Calcium challenge in the Zollinger-Ellison syndrome. Surgery 72: 60, 1972.

15. Romanus, M. E., Neal, J. A., Dilley, W. G., Leight, G. S., Linehan, W. M., Santon, R. J., Fardon, J. R., Jones, R. S., and Wells, S. A., Jr. Comparison of four provocative testsfor the diagnosisof gastrinoma.Ann. Surg. 197: 605, 1983. 16. Saxe, A. W., Yoon, J. W., Gordon, P., and Brennan, M. F. Cell culture and in vitro studies of fresh and cryo preserved human insulinoma. In Vitro 18: 884, 1982.

17. Turbey, W. J., and Passaro,E. Hyperparathyroidism in the Zollinger-Ellison Syndrome. Arch. Surg. 210: 162,1972.

18. Williams, J. A. Regulation of pancreatic acinar cell function by intracellular calcium. Amer. J. Physiol. 238:6269,

1980.

19. Williams, J. A. The effect of ionophore A23187 on amylase release, cellular integrity and ultrastructure of mouse pancreatic acini. Cell Tissue Rex 186: 287, 1978.